SEARCH

SEARCH BY CITATION

Keywords:

  • Phosphate homeostasis;
  • phosphate signaling;
  • phosphate starvation;
  • post-translational control;
  • ubiquitin

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Transcriptional Control of Pi Starvation Responses
  5. Post-transcriptional Control of Pi Starvation Responses
  6. Ubiquitin-mediated Modulation of Pi Starvation Responses
  7. Pi-responsive UPS Components
  8. Summary and Outlook
  9. Acknowledgements
  10. References
  11. Supporting Information

Throughout evolution, plants have evolved sophisticated adaptive responses that allow them to grow with a limited supply of phosphate, the preferential form in which the essential macronutrient phosphorus is absorbed by plants. Most of these responses are aimed to increase phosphate availability and acquisition through the roots, to optimize its usage in metabolic processes, and to protect plants from the deleterious effects of phosphate deficiency stress. Regulation of these adaptive responses requires fine perception of the external and internal phosphate levels, and a complex signal transduction pathway that integrates information on the phosphate status at the whole-plant scale. The molecular mechanisms that participate in phosphate homeostasis include transcriptional control of gene expression, RNA silencing mediated by microRNAs, regulatory non-coding RNAs of miRNA activity, phosphate transporter trafficking, and post-translational modification of proteins, such as phosphorylation, sumoylation and ubiquitination. Such a varied regulatory repertoire reflects the complexity intrinsic to phosphate surveying and signaling pathways. Here, we describe these regulatory mechanisms, emphasizing the increasing importance of ubiquitination in the control of phosphate starvation responses.

  • image

[ Vicente Rubio (Corresponding author)]


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Transcriptional Control of Pi Starvation Responses
  5. Post-transcriptional Control of Pi Starvation Responses
  6. Ubiquitin-mediated Modulation of Pi Starvation Responses
  7. Pi-responsive UPS Components
  8. Summary and Outlook
  9. Acknowledgements
  10. References
  11. Supporting Information

Despite phosphorus’ (P) abundance on Earth's surface, it is one of the most limiting essential macronutrients for plants. This is because orthophosphate (a.k.a. inorganic phosphate, Pi), the form in which P is preferentially assimilated by plants, is mainly immobile and is distributed unevenly in soils, thus limiting its acquisition by plants (Holford 1997). To cope with growth under a low Pi supply, plants count on a battery of adaptive responses aimed at increasing Pi mobilization and uptake and protecting the plant against the adverse effects caused by Pi deprivation stress (reviewed in Vance et al. 2003; Franco-Zorrilla et al. 2004; Ticconi and Abel 2004; Vance 2010). Such responses include changes in root architecture and root hair development that increase the area of root surface available for soil exploration and Pi acquisition (López-Bucio et al. 2003; Péret et al. 2011). Thus, plants grown under low-Pi conditions display an increased root-to-shoot ratio and a higher number of long root hairs (Bates and Lynch 1996; Williamson et al. 2001). Additionally, lateral roots proliferate, whereas primary root growth is limited, enabling better exploration of the so-called “topsoil” and scavenging of Pi derived from organic decomposition (Lynch and Brown 2001; Williamson et al. 2001; López-Bucio et al. 2002). Some plant families have evolved much more sophisticated mechanisms to improve Pi acquisition. Thus, members of the Proteaceae family evolved proteoid roots (formed by clusters of lateral roots) that secrete high amounts of organic acids to the soil in order to dissociate Pi from insoluble precipitates (mainly Pi salts of Al, Fe, Mg and Ca; Dinkelaker et al. 1995; Watt and Evans 1999). Enhanced Pi uptake can be also accomplished by establishing mutualistic associations with mycorrhizal fungi, a phenomenon that occurs in 80% of plant species (reviewed in Harrison 2005).

Other changes occurring at the molecular level include increases in the expression of high affinity Pi transporters, which accumulate at the plasma membrane to enhance Pi uptake during Pi starvation (Raghothama 1999). Protein levels for these transporters at the plasma membrane are tightly regulated, since they can be degraded at the lytic vacuole when sufficient Pi levels are attained (Bayle et al. 2011). Pi-scavenging enzymes, such as phosphatases or ribonucleases, are also produced and secreted, together with organic acids and protons, to mobilize Pi from soil (Bariola et al. 1994; del Pozo et al. 1999; Vance et al. 2003; Tian et al. 2012). Adaptive responses also aim to economize Pi resources under prolonged Pi deprivation. These include utilization of alternative enzymatic reactions within the glycolytic and respiratory pathways that do not rely in Pi or adenylate as substrates, and changes in lipid composition of membranes, where phospholipids are substituted by galacto- and sulpholipids (Duff et al. 1989; Essigmann et al. 1998; Kobayashi et al. 2006). A severe drop in intracellular Pi levels also provokes harmful effects, such as a reduction in photosynthetic capacity (e.g. photoinhibition), that can be alleviated by increased biosynthesis of anthocyanins and other photoprotective pigments (Takahashi et al. 1991; Trull et al. 1997).

To coordinate all of the above-mentioned responses, plants require a Pi-monitoring system that involves the perception and integration of information on local and whole-plant Pi status (local and long-distance Pi signaling, respectively; reviewed in Franco-Zorrilla et al. 2004; Ticconi and Abel 2004: Doerner 2008; Rouached et al. 2010; Vance 2010; Yang and Finnegan 2010; Chiou and Lin 2011). Although the nature of Pi sensor(s) is unknown, many components of the Pi-starvation signaling pathway have been identified during the last decade. These regulatory components act at the transcriptional or post-transcriptional level to maintain plant Pi homeostasis and adequate responses to low Pi stress conditions. In this review, we do not intend to summarize knowledge on all these regulatory mechanisms, but instead wish to focus on the ever-increasing information involving the ubiquitin proteasome system (UPS) in the control of Pi homeostasis and Pi starvation adaptive responses.

Transcriptional Control of Pi Starvation Responses

  1. Top of page
  2. Abstract
  3. Introduction
  4. Transcriptional Control of Pi Starvation Responses
  5. Post-transcriptional Control of Pi Starvation Responses
  6. Ubiquitin-mediated Modulation of Pi Starvation Responses
  7. Pi-responsive UPS Components
  8. Summary and Outlook
  9. Acknowledgements
  10. References
  11. Supporting Information

Transcriptional control of the so-called Phosphate Starvation Responsive (PSR) genes greatly underlies the adaptive response program of plants to Pi-deficiency. According to this idea, early studies showed that expression of many genes involved in the control of the Pi starvation rescue system increases under Pi-deficient conditions (reviewed in Franco-Zorrilla et al. 2004). Subsequent expression profiling studies have further supported the complexity of the transcriptional control network underlying Pi-deficiency responses, which may be composed of at least two or more transcriptional programs (i.e. early and late Pi-starvation responses; Hammond et al. 2003; Wu et al. 2003). These studies have also shown that the over-all change in gene expression is conserved between distant higher plants, such as rice and Arabidopsis (Wasaki et al. 2003). In this regard, genome-wide transcriptional analyses have been used for monitoring the short-, medium- and long-term transcriptional responses of plants (mainly Arabidopsis) to Pi starvation, and the Pi-specificity of the corresponding transcriptional changes, using different Pi re-supply experimental designs (Misson et al. 2005; Morcuende et al. 2007; Müller et al. 2007; Lan et al. 2012). Additionally, distinction among groups of genes that are controlled by Pi deprivation either locally or systemically (i.e. long-distance responsive genes) was observed by means of “split root” techniques and transcriptomic assays (Thibaud et al. 2010). Recently, Bustos et al. (2010) contributed to the overall panorama by performing an organ (roots and shoots)-specific transcriptomic analysis under long-term Pi-deficiency conditions in Arabidopsis wild-type (WT) and in phr1 and phr1 phl1 mutants. These analyses elucidated the central role of the PHR1 transcription factor and its (partially) functionally-redundant homolog PHL1 in the control of transcriptional activation and repression responses to Pi starvation in Arabidopsis.

PHR1 and PHL1, together with their homolog proteins in Chlamydomonas, rice, and Phaseolus (PSR1, OsPHR2, and PvPHR1, respectively) belong to the MYB-Coiled Coil (MYB-CC) family of transcription factors (TF; Wykoff et al. 1999; Rubio et al. 2001; Valdés-López et al. 2008; Zhou et al. 2008; Bustos et al. 2010). PHR1 was identified in a screen for mutants displaying altered Pi starvation responses. Thus, phr1 mutants grown in low Pi conditions showed reduced expression of PSR genes, decreased accumulation of anthocyanins, and a reduced root/shoot ratio relative to WT plants. In addition, phr1 mutants displayed lower levels of free intracellular Pi in shoots under all Pi regimens (Rubio et al. 2001). Bustos et al. (2010) showed that these defects are enhanced in phr1 phl1 double mutants, where up to 80% and 60% of the Pi starvation-inducible genes in shoots and roots, respectively, display lower levels of expression compared to WT plants under–Pi conditions. Such major control in gene expression activation is in part exerted directly, as indicated by the fact that promoters of Pi starvation-induced genes are enriched in PHR1-binding sequences (P1BS), and by induction of PSR genes upon activation of a PHR1:GR fusion protein upon treatment with dexamethasone, even when protein translation is inhibited with cycloheximide (Bustos et al. 2010). Interestingly, PHR1/PHL1 direct targets include the majority of the genes that are systemically controlled by low Pi, as was shown by Thibaud et al. (2010).

Additional proteins belonging to different TF families also act as regulators of PSR gene expression, although to a lesser extent than PHR1/PHL1 family members. This is the case of Arabidopsis bHLH32, WRKY75, ZAT6, and MYB62 TFs, and of rice OsARF16 (Eulgem et al. 2000; Yi et al. 2005; Chen et al. 2007; Devaiah et al. 2007a, 2007b). Control of chromatin structure is also known to regulate Pi starvation responses. Thus, it was found that activity of ARP6, a nuclear actin-related protein that is part of the SWR1 chromatin remodeling complex, is required for proper deposition of Histone H2A.Z variant at a number of PSR genes (Smith et al. 2010). Therefore, plants lacking ARP6 function displayed constitutive Pi starvation responses, including increased length and number of root hairs, starch accumulation, a high level of phosphatase activity, and upregulation of a subset of PSR genes, when grown in Pi-rich media. These results suggest that, contrary to what it has been observed for the PHO regulon in yeast, H2A.Z deposition at PSR genes downregulates their expression (Smith et al. 2010).

Several proteomic studies performed in different plant species indicate that there is a high correlation between transcriptional changes that occur in response to Pi deprivation, and variations in the abundance of the corresponding proteins (Li et al. 2007; Tran and Plaxton, 2008; Chevalier and Rossignol 2011; Lan et al. 2012). Such correlation is more evident in the case of genes that are highly upregulated in response to low Pi. However, in a significant number of cases, protein levels did not match gene expression variation, especially in the case of proteins whose abundance is reduced under Pi deprivation (Lan et al. 2012). These discordances between mRNA and protein levels highlight the relevance of post-transcriptional regulation during Pi starvation stress adaptation.

Post-transcriptional Control of Pi Starvation Responses

  1. Top of page
  2. Abstract
  3. Introduction
  4. Transcriptional Control of Pi Starvation Responses
  5. Post-transcriptional Control of Pi Starvation Responses
  6. Ubiquitin-mediated Modulation of Pi Starvation Responses
  7. Pi-responsive UPS Components
  8. Summary and Outlook
  9. Acknowledgements
  10. References
  11. Supporting Information

In this section, we briefly summarize different post-transcriptional mechanisms that, together with ubiquitination (the main focus of this review; see the following section), affect Pi uptake and long-distance Pi signaling, and modulate Pi starvation responses. Further details on the molecular basis of these mechanisms are provided by several excellent reviews (Doerner 2008; Rouached et al. 2010; Yang and Finnegan 2010; Chiou and Lin 2011; Péret et al. 2011).

Sumoylation

Similarly to ubiquitin, small ubiquitin-like modifier (SUMO) peptide is attached to protein targets by a specific enzymatic cascade. Modification of proteins with SUMO rarely triggers their degradation through the 26S proteasome, but alters protein activity, localization or interaction abilities (Ulrich 2005). In plants, sumoylation has been involved in the control of developmental processes, such as flowering control, and responses to different stresses, including cold, drought, pathogen attack and Pi starvation. Most of these regulatory effects of sumoylation have been obtained from the characterization of mutants in E3 SUMO-ligase SIZ1 (review by Miura and Hasegawa 2010). Among other pleitropic phenotypes, Arabidopsis siz1 mutants displayed enhanced local and systemic Pi starvation responses, such as reduced growth of the primary root, increased lateral root proliferation, and higher accumulation of anthocyanins. siz1 mutants also displayed upregulation of a subset of PSR genes under Pi-sufficient conditions (Miura et al. 2005). Interestingly, SIZ1 was able to sumoylate PHR1 in vitro. Substantiating this result, PHR1 was isolated together with SUMO-protein conjugates purified from Arabidopsis (Miller et al. 2010). SIZ1-mediated sumoylation of PHR1 may account for many Pi-related phenotypes found in siz1 mutants. However, it has recently been shown that SIZ1 negatively regulates root architecture responses to Pi limitation by controlling auxin-regulated gene expression and auxin patterning (Miura et al. 2011). These results indicate that SIZ1 may target additional yet-unknown proteins within the auxin-signaling pathway to control specific Pi starvation responses.

Pi transporter phosphorylation and trafficking

Certain Pi starvation responses are shared between organisms that rely on Pi availability in the external media (e.g., bacteria, fungi and plants), including increased expression and accumulation of high-affinity Pi transporters, aimed at facilitating Pi uptake into cells (Raghothama 1999). In plants, accumulation of high-affinity Pi transporters (PHT1 family members) at the plasma membrane (PM) is subject to tight regulation. Thus, upon Pi deprivation, expression of PHT1 genes is upregulated, and newly-synthesized PHT1 proteins are sorted from the endoplasmic reticulum (ER) to the PM into COPII-coated secretory vesicles (Bayle et al. 2011). Proper ER-to-PM trafficking of PHT1 proteins requires the function of PHF1 (PHOSPHATE TRAFFIC FACILITATOR 1), an ER-localized protein structurally related to SEC12 proteins, and is modulated by phosphorylation of specific residues at the C-terminal end of PHT1 proteins. PHF1 is apparently not necessarily used for COPII vesicle recruitment, but instead it may act as a packaging chaperone of incorrectly-folded PHT1 proteins (González et al. 2005; Bayle et al. 2011). Once at the PM, adequate PHT1 protein levels are maintained according to the Pi requirement of cells. Thus, under Pi-sufficient conditions, excess PHT1 proteins that are no longer necessary are removed from the PM by internalization into endosomes, and subsequent sorting and degradation into the vacuole. Under Pi-limiting conditions, PHT1 proteins are also subjected to endocytosis, although they are mainly redirected to the PM by endosome recycling processes, therefore allowing sustained Pi uptake rates.

miRNA-mediated long-distance Pi signaling and target mimicry

Pi homeostasis requires long-distance signaling mechanisms that allow integration of Pi status at different plant organ levels (reviewed in Doerner 2008; Chiou and Lin 2011). The nature of the signaling molecule(s) has been the focus of intense debate in the field. Recent studies have proposed that specific miRNAs (i.e., miR399 family members) can have this function based on key features of these molecules. Thus, miR399 genes are highly responsive to Pi starvation, their products can move from the shoot to the root through the phloem, and, moreover, they recognize as targets the transcripts of PHO2, a gene encoding a putative E2/E3 ubiquitin ligase that acts as a negative regulator of shoot Pi accumulation (Fujii et al. 2005; Aung et al. 2006; Bari et al. 2006; Chiou et al. 2006; Lin et al. 2008; Pant et al. 2008). Therefore, the proposed model is that, upon Pi deprivation, intracellular Pi levels will drop in the aerial part, triggering miR399 expression and transport to the root where they recognize specific target sites on the 5'UTR of PHO2 mRNA to promote its degradation through the RNA-induced silencing complex. As a result of PHO2 degradation, there is an increase in both Pi transport into root cells and Pi loading into the xylem, that aims to normalize Pi levels in the shoot (Figure 1).

image

Figure 1. Model depicting known roles of ubiquitin proteasome system (UPS) components in the control of local and systemic phosphate (Pi) signaling and of Pi-deficiency adaptive responses in Arabidopsis.  Conjugation (E2/E3; e.g. PHO2) and de-conjugation (DUB; e.g. UBP14) of ubiquitin molecules to target proteins recognized by specific E3s (E3; e.g. TIR1, FBX2 and NLA) are processes regulating Pi-deficiency adaptive responses, such as the enhancement of secondary root growth and the maintenance of Pi homeostasis. Both local responses (e.g. UBP14) and systemic signaling (e.g. PHO2, NLA) are also at least partially regulated by UPS components. UPS components are indicated in circular-shaped text boxes. Blue and brown text indicates tissue-specific activity for shoots and roots, respectively. Colored rectangular boxes indicate major Pi-related processes. Black and red lines indicate positive and negative regulation, respectively.

Download figure to PowerPoint

Additional studies added a new level of complexity to this regulatory mechanism by identifying non-coding RNAs that mimic miR399's targets (Franco-Zorrilla et al. 2007). In this regard, it was found that the products of the At4/IPS1 gene family (conserved in different eudicots, such as rice, tomato, Medicago and Arabidopsis) show partial complementarity to miR399. Thus, At4/IPS1 members share a 23-nt region that is complementary to the miR399 sequence, with the exception of a 2- or 3-nt mismatch in the predicted miRNA cleavage site. In this way, At4/IPS1 RNA molecules are able to compete with PHO2 transcripts for binding miR399. Importantly, imperfect pairing between At4/IPS1 and miR399 molecules prevents cleavage of At4/IPS1, resulting in miR399 sequestration and leading to stabilization of PHO2 transcripts (Franco-Zorrilla et al. 2007). Interestingly, At4/IPS1 and miR399 gene expression is induced by PHR1 under low Pi conditions. It has been suggested that At4/IPS genes may act as a negative feedback regulatory loop that limits miR399 function under fluctuating Pi supply conditions.

The regulatory mechanism mediated by At4/IPS1 genes was termed “target mimicry”, and since its identification in the context of Pi signaling in plants, it has also been found in animals, where it provides a framework for a dialogue among RNAs sharing the same miRNA binding sites (Rubio-Somoza et al. 2011).

Ubiquitin-mediated Modulation of Pi Starvation Responses

  1. Top of page
  2. Abstract
  3. Introduction
  4. Transcriptional Control of Pi Starvation Responses
  5. Post-transcriptional Control of Pi Starvation Responses
  6. Ubiquitin-mediated Modulation of Pi Starvation Responses
  7. Pi-responsive UPS Components
  8. Summary and Outlook
  9. Acknowledgements
  10. References
  11. Supporting Information

Ubiquitin (Ub) is a small peptide that acts as a post-translational modifier to virtually regulate all aspects of cell biology in eukaryotes, including cell division, growth, communication, movement and death (Hershko and Ciechanover 1998). In plants, Ub has been assigned to the control of many physiological, developmental and stress responses, such as phytohormone signaling, flowering, circadian clock function, plant defense, heat shock and cold stresses, DNA damage repair, nutrient deprivation and drought tolerance, and more (Smalle and Vierstra 2004). Ub is covalently linked to Lysine (Lys) residues in other proteins by specific enzymatic cascades. These cascades begin by transference of a Ub moiety from an E1 Ub-activating enzyme (E1) to an E2 Ub-conjugating enzyme (E2). E3 Ub-ligases (E3) represent the last step in the cascade, bringing together the E2 and the protein target that is then ubiquitinated. Therefore, E3s play a key role in ubiquitination by providing substrate specificity. Consecutive cycles of ubiquitination may result in polyubiquitination and subsequent recognition by the 26S proteasome for degradation (Hershko and Ciechanover 1998; Figure 2A). Despite the well-established function of ubiquitination in regulated proteolysis via the proteasome, recent evidence has shown that this mechanism can modulate protein function through additional means. In this context, it has been found that different poly-Ub chain conformations determine different fates for the targeted protein. Thus, substrates for proteosomal degradation are labeled with chains in which Ubs are linked via their Lys 48 or 11, whereas Ub chains linked via Lys 63, 29 or 6 modulate other aspects of protein functionality, such as localization, assembly, structure or enzymatic activity (Deshaies and Joazeiro 2009). A similar effect is found in the case of protein monoubiquitination, which is involved, for example, in endocytosis of membrane receptors and histone modification (Sigismund et al. 2004). However, it has recently been shown that monoUb can trigger proteasomal degradation of small protein targets (≤150 aas; Shabek et al. 2012). Target ubiquitination can be reversed by Ub hydrolases (i.e., deubiquitinases or Ub deconjugases), adding an additional level of complexity to the regulatory mechanisms involved in this posttranslational modification (Kim et al. 2003).

image

Figure 2. Phosphate (Pi)-responsive ubiquitin proteasome system (UPS) components.(A) Ubiquitination of a target protein requires an enzymatic cascade that catalyzes the activation (Ub-activating enzyme; E1), conjugation (Ub-conjugating enzyme; E2) and transference (Ub ligase enzyme; E3) of the Ub molecule to the substrate. Depending on Ub chain conformation, poly-ubiquitination of targets may trigger their degradation via the 26S proteasome. (B) Percentages of UPS genes, distributed in gene super-families, upregulated (yellow-scale bars) and downregulated (blue-scale bars) in wild-type (WT) seedlings after 7 d of Pi-deficiency. Changes in transcriptional behavior due to the double mutant phr1 phl1 are represented in different colors (see Conventions box). For each super-family, percentages are relative to the number of genes represented in the ATH1 array. Two different cut-off values were used for WT and mutants (2× and FDR < 0.05 and 1.5× and FDR < 0.1 for WT and phr1 phl1, respectively), Statistical significance was determined by binomial distribution analysis (•) (P < 0.01) and by t-test (*) (P < 0.05). Data from Bustos et al. (2010). (C) Pi content of shoots of 7-d-old seedlings grown in low Pi (50 μM KH2PO4/K2HPO4) and +Pi (500 μM KH2PO4/K2HPO4) solid Johnson media (see methods in Supplementary Materials). Color codes represent the three different genotypes used: pub27/28/29, phr1 and Col-0. Values and error bars are means ± SD.

Download figure to PowerPoint

Several reports have demonstrated that specific enzymes mediating Ub conjugation or deconjugation are involved in the control of adaptive responses to Pi starvation in plants (Figure 1). Below, we describe the best-known examples.

E3 Ub ligases as negative regulators of Pi responses

FBX2, a protein that contains both WD40 and F-box domains that could be used to recognize specific ubiquitination targets, was identified as a negative regulator of molecular, developmental, physiological and metabolic responses to Pi starvation in Arabidopsis (Chen et al. 2008). Indeed, a T-DNA insertion mutant in the corresponding gene, fbx2, displayed constitutive low Pi responses, such as higher levels of PPCK1 (Phosphoenolpyruvate Carboxylase Kinase 1) transcripts, a higher number of root hairs, and increased contents of intracellular Pi and anthocyanins under Pi-sufficient conditions (Chen et al. 2008). According to microarray data, FBX2 gene expression is not responsive to differences in Pi supply, which indicates that FBX2 function is controlled post-transcriptionally by a yet-unknown mechanism (Bustos et al. 2010). The molecular mechanism by which FBX2 modulates PSR is not clear either, since the identity of its potential ubiquitination substrates is unknown. Although PHR1 might be an ideal target candidate due to its central role in the control of PSR, no evidence for direct interaction between FBX2 and PHR1 was found (Chen et al. 2008). Another TF, bHLH32, was found to strongly interact with FBX2 in vitro. However, it is unlikely that bHLH32 is targeted by FBX2 activity, since both proteins act as negative regulators of similar Pi starvation responses, as observed in the analysis of fbx2 and bhlh32 mutants (Schiefelbein 2003; Zhang et al. 2003; Chen et al. 2007, 2008). It has been proposed that FBX2 acts as part of an SCF E3 Ub ligase that recruits target proteins for ubiquitination through its interaction with bHLH32 (Figure 1). The identification of the targets recognized by FBX2 will likely unveil novel regulatory elements within the Pi starvation-signaling pathway.

Crosstalk between auxin and Pi starvation signaling

Another F-box protein, Transport Inhibitor Response 1 (TIR1), also controls (to some extent) plant growth and developmental responses to Pi starvation. TIR1 acts as an auxin receptor that triggers ubiquitination of AUX/IAA (AUXIN/INDOLE-3-ACETIC ACID) proteins, and further degradation by the 26S proteosome (Gray et al. 2001; Dharmasiri et al. 2005; Kepinski and Leyser 2005). Removal of AUX/IAA proteins releases Auxin Response Factor (ARF) TFs that regulate expression of auxin-responsive genes (Tiwari et al. 2001). Interestingly, Pérez-Torres et al. (2008) reported that, contrary to WT seedlings, tir1–1 mutants failed to enhance lateral root production in response to Pi starvation, indicating that SCFTIR1-mediated auxin signaling is required for root architecture modification in response to low Pi availability. Additional data showed that changes in Pi availability do not cause differences in free IAA concentration in either the roots or shoots of WT seedlings (Jain et al. 2007; Pérez-Torres et al. 2008). Indeed, it was found that the increase in the formation and emergence of lateral roots under Pi starvation conditions is due to an enhancement in auxin sensitivity as a consequence of increased expression of TIR1 in Pi-deprived plants (Pérez-Torres et al. 2008). Therefore, upon Pi deprivation, accumulation of TIR1 would enhance degradation of AUX/IAA proteins, liberating ARF transcription factors and modulating the expression of genes involved in lateral root formation (Figure 1).

Ub deconjugases

The previous examples demonstrate that enzymes mediating Ub conjugation to protein targets play important roles in the control of Pi starvation responses. Likewise, deconjugation of Ub from targets should affect plant adaptation to Pi deprivation. This is the case of Ubiquitin-specific Protease 14 (UBP14), a Ub deconjugase (or deubiquitinase) that is required for regulation of Pi homeostasis at the posttranslational level, and which affects the root hair developmental program in response to Pi availability (Li et al. 2010). UBP14 had been previously described as being essential for the optimal functioning of the Ub/26S proteosome pathway during the early stages of plant development, through its roles in ubiquitin recycling and in the maintenance of an adequate balance between free ubiquitin molecules and poly-ubiquitin chains (Doelling et al. 2001). In fact, in the absence of UBP14, the embryos reach the globular stage, but are unable to produce viable seeds as a consequence of a limited number of cell divisions before the embryo arrest, resulting in an embryo-lethal phenotype (Doelling et al. 2001). By means of a screen of an ethyl methanesulfonate (EMS)-mutagenized Arabidopsis population, Li et al. (2010) isolated per1 (Pi deficient root hair defective1), a weak mutant allele of UBP14, which allowed further functional characterization of this gene. per1 plants displayed Pi-specific defects in root hair elongation under Pi starvation conditions, and a battery of phenotypes that resembled Pi-deficient plants under Pi-rich conditions. These results suggest that an adequate balance between Ub monomers and Ub chains is necessary for sustaining proper root responses to Pi deprivation (Li et al. 2010; Figure 1).

Ubiquitination pathway components controlling Pi homeostasis

As mentioned before, PHO2 protein acts as a negative regulator of Pi starvation responses by controlling Pi transport from roots and Pi allocation in shoots (Fujii et al. 2005; Bari et al. 2006). Thus, pho2 mutants accumulate excessive Pi in shoots (from 3- to 6-times the level in WT plants), which may cause toxic effects and leaf senescence (Delhaize and Randall 1995; Dong et al. 1998). PHO2, also known as UBC24, belongs to an atypical E2 Ub conjugase family that includes four members in Arabidopsis (UBC23–26; Bachmair et al. 2001; Kraft et al. 2005). UBC family members display sequence similarity to mammalian UBE2O (E2–230K) and BIRC6 (Apollon), two E2 enzymes that do not interact with E3 Ub ligases, but rather contain a chimeric E2/E3 domain that may allow them to directly target specific proteins for ubiquitination (Berleth and Pickart 1996; Hauser et al. 1998; Bartke et al. 2004; Hao et al. 2004). Both BIRC6 and UBE2O may have similar cellular localization, since BIRC6 localizes in the Golgi and Trans-Golgi network and in small vesicles, and UBE2O interacts with proteins associated with the endomembrane system, such as copine (Hauser et al. 1998; Tomsig et al. 2003). Interestingly, it has recently been shown that PHO2 also colocalizes with endomembrane compartments, where it may trigger ubiquitination and degradation of specific substrates involved in Pi transport to the shoot (Liu et al. 2012). The identity of one of these targets was recently unveiled by means of a screen for pho2 suppressors. Thus, it was found that mutations in PHO1, a transmembrane protein involved in the loading of Pi into the xylem, are able to rescue pho2 Pi levels to normal conditions (Poirier et al. 1991; Arpat et al. 2012; Liu et al. 2012). Interestingly, both PHO2 and PHO1 are expressed in vascular cells, where they colocalize at the endomembrane system (Hamburger et al. 2002; Arpat et al. 2012; Liu et al. 2012). Additional analyses demonstrated that PHO2 and PHO1 physically interact, and, moreover, that PHO2 is able to trigger degradation of PHO1 in transient expression assays in Nicotiana benthamiana leaves, although in a proteasome-independent manner. Further experiments have indicated that PHO1 proteolysis may occur in the vacuole, and is mediated by trafficking into multivesicular bodies (Liu et al. 2012). Therefore, PHO2 acts in the root as a negative regulator of Pi transport to the aerial part by promoting degradation of PHO1 under high Pi conditions (when PHO2 transcripts accumulate; Figure 1). However, whether or not PHO1 is directly ubiquitinated by PHO2 still needs to be demonstrated through further research.

Another UPS component, NLA, is involved in Pi homeostasis maintenance (Kant et al. 2011). NLA contains a RING domain, and has been proposed to act as an E3 Ub ligase. Additionally, it contains an SPX domain, which is present in many proteins involved in Pi sensing, and in transport in yeast and plants (e.g., PHO1 family members; Wang et al. 2004; Stefanovic et al. 2007). NLA ubiquitination targets have not been identified to date. However, it was found that Pi transporters accumulate in nla mutants, although whether this effect is direct or indirect remains unknown. Lack of NLA function causes an increase in Pi accumulation in shoots, similar to PHO2 mutation (Kant et al. 2011). Interestingly, high accumulation of Pi in these mutants depends on nitrate concentration. Thus, nla plants show enhanced Pi levels only under low-nitrate conditions, and pho2 displays higher increased Pi accumulation under these conditions. On the other hand, it was found that Pi starvation induces high expression of miR827, which displays complementarity to NLA transcripts and promotes their destabilization, allowing accumulation of PHT1 transporters and increased Pi uptake and transport to the shoot (Kant et al. 2011). Further characterization of the mechanisms that involve PHO2 and NLA function will surely shed light on the sophisticated regulatory crosstalk between Pi and nitrate signaling pathways.

Pi-responsive UPS Components

  1. Top of page
  2. Abstract
  3. Introduction
  4. Transcriptional Control of Pi Starvation Responses
  5. Post-transcriptional Control of Pi Starvation Responses
  6. Ubiquitin-mediated Modulation of Pi Starvation Responses
  7. Pi-responsive UPS Components
  8. Summary and Outlook
  9. Acknowledgements
  10. References
  11. Supporting Information

To evaluate the potential involvement of the UPS in the control of Pi starvation responses, we analyzed the transcriptional behavior of different gene super-families that compose the UPS in response to different Pi supplies. In addition, we determined the effect of the double mutant phr1 phl1 in the Pi response of UPS genes. To do this, we took advantage of microarray data produced in our laboratory corresponding to WT and phr1 phl1 roots and shoots grown under Pi-sufficient or long-term Pi deprivation conditions (Bustos et al. 2010). It should be kept in mind that, in this analysis, inherent limitations in the expression arrays, such as incomplete representation of gene families and the loss of information derived from unspecific probe sets, and/or low gene expression levels, represent challenges to reliable measurements (Benjamini and Hochberg 1995). Even if the representation of the gene super-families, grouped by enzymatic function (E1 and E2 were arbitrarily studied together due to the reduced number of E1 members), is heterogeneous in the ATH1 microarray, they are broadly represented (for an updated and detailed review of Arabidopsis UPS components see Sadanandom et al. 2012). Thus, we found probe sets corresponding to more than 75% of the members of almost all gene super-families studied, except for HECT and F-box families, whose representation was more limited (71.4% and 55.7%, respectively). Super-families such as E1/E2, Cullins, CDD (COP10, DET1, DDB1) components and DUB (Deubiquitinases) had almost all their members represented (>90%). However, as mentioned previously, there is an information gap derived from the nearly 300 F-boxes without corresponding ATH1 probe sets, which needs to be examined by alternative expression profiling techniques.

As shown in Figure 2B, the transcriptional behavior of the overall UPS components in response to Pi starvation differed from that found in the global ATH1 probe sets. Indeed, Pi-responsive UPS components were, in general, slightly underrepresented. This is in agreement with the lack of over-representation of PHR1 binding sites in the promoter sequences of the UPS component genes (data not shown). However, by splitting the analysis in different gene families, over-representation of upregulated genes in response to Pi-deficiency in shoots became apparent in the case of the U-box family (binomial distribution analysis in Figure 2B). This over-representation seems to be functionally relevant in the context of Pi homeostasis, as shown by analysis of a triple pub27/28/29 knockdown line which accumulated less intracellular free Pi under different Pi conditions, similar to phr1 mutants (Figure 2C). Interestingly, increased PUB28 gene expression in response to low Pi is controlled by PHR1/PHL1 genes, supporting the implication of this single-unit E3 family in the regulation of adaptive Pi-deficiency responses in Arabidopsis (Table S1).

In assessing the transcriptional effects of the double mutant phr1 phl1 on the expression of UPS components, it is evident that the post-transcriptional regulatory processes that involve the UPS in Pi-deficiency responses are mainly under the control of PHR1(-like) TFs. Thus, most of the Pi-responsive UPS components exhibited reduced expression in the double mutant phr1 phl1, both in shoots and in roots (Figure 2B), and displayed concordant dynamic ranges (Tables S1, S2). This transcriptional effect was independent of the representation of deregulated genes in each family. However, by comparing the overall transcriptional behavior of each gene super-family in phr1 phl1 and WT lines, it became evident that PHR1/PHL1 largely influences the expression of genes encoding substrate receptors (both E3 Ub ligase complex components and monomeric E3 enzymes; Figure 2B). Thus, PHR1/PHL1 was observed to significantly control Pi starvation-responsiveness of the U-box, F-box and DCAF gene families in both shoots and roots, and that of the RING finger and PHD families in shoots and roots, respectively. PHR1/PHL1 TFs also regulated expression of DUBs in Arabidopsis shoots, indicating that Ub deconjugation from specific protein targets and Ub recycling may play regulatory roles in the control of Pi starvation responses at the whole-plant level, and not specifically in roots as shown for UBP14 (Li et al. 2010).

In order to further assess the relevance of the UPS system in the control of Pi signaling, we took advantage of weak cul1–6 and cul3hyp Arabidopsis mutant lines (note that null mutants are embryo-lethal), which partially affect the assembly and function of two different E3 Ub ligases: the SCF and CUL3-BTB complexes, respectively (Moon et al. 2007; Thomann et al. 2009). Partial lack of CUL1 and CUL3 function should yield a reduction in the function of F-box and BTB proteins, including those that respond to Pi supply variations (Figures 2, 3). Using these lines, we analyzed the expression of different PSR genes under different Pi supply conditions by quantitative reverse transcription-polymerase chain reaction (qRT-PCR) (Figure 3A). Both cul1–6 and cul3hyp hypomorphic mutants displayed higher transcript levels of PSR genes relative to WT seedlings under Pi-deficient conditions. These results suggest that specific SCF and CUL3-BTB complexes, which are only partly active in the aforementioned mutants, control the abundance of positive regulators of PSR gene transcription. The identification of these targeted-transcriptional regulators will be of utmost importance.

image

Figure 3. Effect of reduced CUL1 or CUL3 function on Phosphate Starvation Responsive (PSR) gene expression under phosphate (Pi)-deficiency and Pi re-supply conditions.(A) Quantitative reverse transcription-polymerase chain reaction (qRT-PCR) expression levels, represented as log2 scale relative to wild-type (WT) +Pi values, of genes involved in Pi mobilization, miRNA regulation, sulpho- and galactolipid biosynthesis, and Pi transporter trafficking is shown for Arabidopsis WT plants (WT; grey graphs) and for mutants displaying reduced CUL1 (cul1–6; blue) or CUL3 (cul3hyp; orange) activity. Arabidopsis seedlings were grown in +Pi (1 mM) media for 7 d before they were transferred to Pi-lacking (-Pi) or fresh Pi-rich (+Pi) media for 4 d. For re-supply experiments, plants were grown as in –Pi but transferred to Pi-rich media for 24 h before harvesting. qRT-PCR was performed using TaqMan Universal ProbeLibrary System (UPL) with ACT8 as a housekeeping reference gene (for primer sequences and methods see Table S3 and Supplementary Materials). (B)cul1–6 and cul3hyp hypomorphic mutants are expected to be defective in the assembly of the SCF and CUL3-BTB E3 ubiquitin ligase complexes. Components of these complexes are depicted.

Download figure to PowerPoint

Summary and Outlook

  1. Top of page
  2. Abstract
  3. Introduction
  4. Transcriptional Control of Pi Starvation Responses
  5. Post-transcriptional Control of Pi Starvation Responses
  6. Ubiquitin-mediated Modulation of Pi Starvation Responses
  7. Pi-responsive UPS Components
  8. Summary and Outlook
  9. Acknowledgements
  10. References
  11. Supporting Information

Multiple regulatory mechanisms control the continuous perception and integration of information on local and whole-plant Pi status, and, therefore, the coordination of the adaptive responses of plants to Pi deprivation and the maintenance of Pi homeostasis. These mechanisms involve both the regulation of gene expression, which is greatly orchestrated by PHR1/PHL1 TFs, and the post-transcriptional control of gene product stability and function. Recent studies have highlighted the relevance of the ubiquitination and Ub-deconjugation pathways in the control of diverse Pi starvation responses (Figure 1). These mechanisms can trigger substrate degradation by two different proteolytic machineries: the 26S proteasome (e.g., TIR1-mediated destabilization of ARF19; Pérez-Torres et al. 2008), and the endocytic/vacuolar protein sorting pathway (e.g., targeted degradation of PHO1 mediated by PHO2; Liu et al. 2012).

Additional UPS components are likely to be involved in the control of plant adaptation to low Pi stress. In agreement with this idea, analysis of transcriptomic data presented here indicates that multiple E3 Ub ligases respond to Pi starvation at the gene expression level in a PHR1/PHL1-dependent manner (Figure 2B; Tables S1, S2). Furthermore, analysis of Pi responses in mutants displaying reduced function of SCF, CUL3-BTB, or of certain PUB E3 ligases, indicates that members of these UPS families modulate Pi accumulation in shoots or PSR gene expression (Figures 2C, 3A). Future studies will unquestionably shed light on the contribution of each of these E3s in the regulation of Pi signaling and low Pi responses. A big step forward in this direction would be provided by the identification and characterization of the specific Pi-responsive substrate receptors within each E3 family, and the protein targets they recognize.

(Co-Editor: Qi Xie)

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Transcriptional Control of Pi Starvation Responses
  5. Post-transcriptional Control of Pi Starvation Responses
  6. Ubiquitin-mediated Modulation of Pi Starvation Responses
  7. Pi-responsive UPS Components
  8. Summary and Outlook
  9. Acknowledgements
  10. References
  11. Supporting Information

Seeds for cul1–6 and cul3hyp mutants were kindly provided by Professors Mark Estelle and Pascal Genschik, respectively. Research in the laboratories of S.P., J.P.-A. and V.R. was supported by the Spanish Ministry of Economy and Competitiveness (MINECO) through different funding programs: the CONSOLIDER Program, Grant 2007–28317; S.P. and J.P.-A., the INNPACTO Program, Grant IPT-310000–2010-9; J.P.-A., the PLANT-KBBE Program, Grant EUI2008–03742; the “Fossi” project; V.R., EUI2008-03748; the “Transnet” project; J.P.-A. and the National Research Program, Grants BIO2011–29085; J.P.-A., BIO2008–04160, BIO2011–30546; S.P., and BIO2010–18820; V.R.). M.R.-T. received a Jae-Predoc fellowship from CSIC.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Transcriptional Control of Pi Starvation Responses
  5. Post-transcriptional Control of Pi Starvation Responses
  6. Ubiquitin-mediated Modulation of Pi Starvation Responses
  7. Pi-responsive UPS Components
  8. Summary and Outlook
  9. Acknowledgements
  10. References
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Transcriptional Control of Pi Starvation Responses
  5. Post-transcriptional Control of Pi Starvation Responses
  6. Ubiquitin-mediated Modulation of Pi Starvation Responses
  7. Pi-responsive UPS Components
  8. Summary and Outlook
  9. Acknowledgements
  10. References
  11. Supporting Information

Figure S1. Generation of pub27/28/29 mutants.

(A) Diagrammatic representation of T-DNA insertions in the PUB27 (line FLAG_104F05) and PUB28 (line SALK_101434) loci.

(B) RT-PCR analysis showing reduced expression of the PUB29 gene in two pub27/28 mutant lines transformed with an RNAi PUB29 construct (see Supplementary Materials).

Table S1. UPS components deregulated in the shoot in response to long-term Pi-deficiency. Data from Bustos et al. (2010).

Table S2. UPS components deregulated in the root in response to long-term Pi-deficiency. Data from Bustos et al. (2010).

Table S3. Primer sequences used for qRT-PCR experiments.

FilenameFormatSizeDescription
JIPB_12017_sm_suppmat.doc60KSupporting info item
JIPB_12017_sm_FigS1.ppt250KSupporting info item
JIPB_12017_sm_Tables.xls97KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.