Cellular and Subcellular Phosphate Transport Machinery in Plants

Phosphorus (P) is an essential element required for incorporation into several biomolecules and for various biological functions; it is, therefore, vital for optimal growth and development of plants. The extensive research on identifying the processes underlying the uptake, transport, and homeostasis of phosphate (Pi) in various plant organs yielded valuable information. The transport of Pi occurs from the soil into root epidermal cells, followed by loading into the root xylem vessels for distribution into other plant organs. Under conditions of Pi deficiency, Pi is also translocated from the shoot to the root via the phloem. Vacuoles act as a storage pool for extra Pi, enabling its delivery to the cytosol, a process which plays an important role in the homeostatic control of cytoplasmic Pi levels. In mitochondria and chloroplasts, Pi homeostasis regulates ATP synthase activity to maintain optimal ATP levels. Additionally, the endoplasmic reticulum functions to direct Pi transporters and Pi toward various locations. The intracellular membrane potential and pH in the subcellular organelles could also play an important role in the kinetics of Pi transport. The presented review provides an overview of Pi transport mechanisms in subcellular organelles, and also discusses how they affect Pi balancing at cellular, tissue, and whole-plant levels.


Introduction
Phosphorus (P) is an essential macronutrient for sustaining optimal plant growth, and plays a significant role in a diverse array of cellular processes, including energy production, redox reactions, photosynthesis, and phosphorylation/dephosphorylation-based reactions [1,2]. Soil-P exists in various organic and inorganic forms due to the decomposition of soil organic matter and weathering of parent rock materials [3]. However, the various P forms differ in their behavior in soils depending on the soil's pH and the size of mineral particles [3,4]. Among several soil-P forms, inorganic phosphate (PO 4 3− , H 2 PO 4 − , and HPO 4 2− , referred to as Pi hereafter) is the most readily accessible for plants; however, bioavailable Pi concentrations rarely exceed 1-10 µm in soil solutions [5][6][7]. In acidic soils, Pi is mostly like H + to avoid membrane hyperpolarization [11,12]. Indeed, investigations during Pi uptake demonstrated decreased cytoplasmic pH in a Catharanthus roseus cell suspension, but increased pH in the extracellular medium [13], as well as the acidification of cytoplasmic pH in the root hairs of Limmobium stoloniferum [14]. After Pi uptake into the root symplasm, Pi can go via different routes: (i) Pi enters the cell cytoplasm (metabolic pool), where the primary access form of Pi into organic molecules occurs via anhydride bond formation as the γ-phosphate group of ATP; (ii) Pi (H 2 PO 4 − or HPO 4 2− ) is secreted into the xylem for long-distance translocation to aerial parts of the plant; and (iii) Pi is stored in vacuoles for the maintenance of Pi homeostasis [11,12]. Additionally, Pi transport from the phloem to the xylem primarily occurs in the form of H 2 PO 4 − or HPO 4 2− ; however, organic Pi compounds such as hexose-phosphates and ATP are also detectable in phloem sap [11,12]. Significant progress was made in understanding Pi transport and utilization mechanisms, which are more or less conserved across the plant kingdom [5,6,15]. A few plant species have a unique ability to tackle adverse effects of Pi deficiency. For instance, Hakea prostrate from the Proteaceae family evolved in severely Pi-deficient soils of southwestern Australia; thus, H. prostrate is highly efficient in managing Pi deficiency as it possesses some unique features [16]. These include cluster roots for efficient Pi uptake, delayed greening, altered Pi allocation to ribosomes, changes in membrane lipid composition, highly efficient photosynthetic Pi use, efficient remobilization of Pi from older senescing leaves, and high-Pi-containing seeds for the initiation of life in Pi-deficient environments [16]. This observation suggests that subcellular Pi transport and its reallocation into various plant parts are important factors for maintaining plant growth under both Pi-repleted and Pi-depleted conditions. Recently, with progress in functional genomics, the roles of novel genes associated with subcellular Pi transport and regulation were investigated. The pH of the cytoplasm and subcellular compartments, as well as the intracellular membrane potential, influences Pi transport at the subcellular level. While pH affects the chemical species of Pi, membrane potentials determine the feasibility of Pi import/export [15] ( Table 1).

Figure 1.
Overview of phosphate (Pi) transport in Arabidopsis plants. The transport route is shown in four parts: uptake from soil to roots, transport from roots to shoots, unloading in shoots and subcellular organelles, and transport to seeds in the form of phytic acid. The high-affinity Pi (PHT1) family (PHT1;1 and PHT1;4) of transporters plays a major role in Pi uptake from soil to roots. The PHO1 protein increases root-Pi xylem loading, whereas PHT1;5 plays a key role in the retranslocation of Pi from shoots to roots, and Pi mobilization to reproductive organs. In plant cell, vacuoles act as the primary intracellular compartments for Pi storage, and SPX-MFS1 and SPX-MFS3/PHT5;1 mediate vacuolar Pi influx and efflux, respectively. Furthermore, Pi is metabolized and transported from leaves to seeds in the form of phytic acid by the ABC-MRP-type phytic acid transporter. The levels of PHT1, PHO1 and PHO2 transporters are regulated by miR399 and cis-NATPHO1;2 in xylem, and by miR399 and IPS1/AT4 in roots. ABC-MRP, ATP binding cassette-multidrug resistance-associated protein; AT4, Arabidopsis thaliana 4; IPS1, induced by phosphate starvation 1; cis-NATPHO1;2, cis-natural antisense transcript phosphate transporter 1;2; PHT, high-affinity phosphate transporter; PHR1, phosphate starvation response 1; Pi, phosphate; PHO1, phosphate transporter 1; SPX-MFS3, SYG1/PHO81/XPR1 major facility superfamily 3; SPX-MFS1, SPX major facilitator superfamily 1; VPT1, vacuolar phosphate transporter 1. Overview of phosphate (Pi) transport in Arabidopsis plants. The transport route is shown in four parts: uptake from soil to roots, transport from roots to shoots, unloading in shoots and subcellular organelles, and transport to seeds in the form of phytic acid. The high-affinity Pi (PHT1) family (PHT1;1 and PHT1;4) of transporters plays a major role in Pi uptake from soil to roots. The PHO1 protein increases root-Pi xylem loading, whereas PHT1;5 plays a key role in the retranslocation of Pi from shoots to roots, and Pi mobilization to reproductive organs. In plant cell, vacuoles act as the primary intracellular compartments for Pi storage, and SPX-MFS1 and SPX-MFS3/PHT5;1 mediate vacuolar Pi influx and efflux, respectively. Furthermore, Pi is metabolized and transported from leaves to seeds in the form of phytic acid by the ABC-MRP-type phytic acid transporter. The levels of PHT1, PHO1 and PHO2 transporters are regulated by miR399 and cis-NAT PHO1;2 in xylem, and by miR399 and IPS1/AT4 in roots. ABC-MRP, ATP binding cassette-multidrug resistance-associated protein; AT4, Arabidopsis thaliana 4; IPS1, induced by phosphate starvation 1; cis-NAT PHO1;2 , cis-natural antisense transcript phosphate transporter 1;2; PHT, high-affinity phosphate transporter; PHR1, phosphate starvation response 1; Pi, phosphate; PHO1, phosphate transporter 1; SPX-MFS3, SYG1/PHO81/XPR1 major facility superfamily 3; SPX-MFS1, SPX major facilitator superfamily 1; VPT1, vacuolar phosphate transporter 1. In view of this background information, the presented paper reviews our current understanding, gained mainly in Arabidopsis thaliana and rice (Oryza sativa), to provide an integrated view of Pi uptake from the soil and its transport mechanisms in/among various organs and subcellular organelles. Future research directions are discussed to develop suitable strategies for the development of crops better suited for growing under Pi-deficient conditions.

Phosphate Transporters and Their Role in Pi Acquisition, Translocation, and Remobilization in Various Organs
The transport of Pi across cell membranes is a critical stage in the regulation of Pi use. During the past few years, several transporter genes that mediate Pi transport processes were identified ( Figure 1). However, the functions of many Pi transporters still remain elusive. Arabidopsis and rice plants contain five high-affinity Pi transporter (PHT1, PHT2, PHT3, PHT4, and PHT5) families that are distinguished based on their protein sequences, locations, and functions [23]. Table 2 summarizes known transporters for uptake at the root surface, root-to-shoot translocation, and the unloading of Pi in shoots.

Pi Redistribution and Remobilization
AtPHT1;1, AtPHT1;5, AtPHT1;9 AtPHT1;5 plays a role in Pi translocation from source to sink organs. Pi redistribution across the vegetative organs. [24,[36][37][38] OsPHT1;4, OsPHT1;6, OsPHT1;8 OsPHT1;4 is involved in the remobilization of Pi from flag leaves to the panicles. OsPHT1;6 and OsPHT1;8 help in Pi remobilization from senescing leaves to young leaves and rice grains. [3] The PHT1 proteins are plasma membrane (PM) proton-coupled Pi-symporters that mediate Pi acquisition from the soil [23,27]. The PHT1 gene family has more members than other PHT families. For example, the Arabidopsis (At) AtPHT1 gene family contains nine Pi transporters (AtPHT1;1 to AtPHT1;9), among which AtPHT1;1 to AtPHT1;4 are mainly involved in Pi uptake from the soil to roots [9]. Gene expression analysis in Arabidopsis showed that AtPHT1;1, AtPHT1;2, AtPHT1;3, and AtPHT1;4 are mainly expressed in various types of root cells, suggesting that these transporters have similar and partially overlapping functions [36,39]. For example, AtPHT1;1 plays a major role when Pi supply is high, whereas AtPHT1;4 becomes predominant under Pi-deficient conditions in Arabidopsis [39]. Ayadi et al. [39] demonstrated that the Arabidopsis atpht1;1/atpht1;2/atpht1;3/atpht1;4 quadruple and the phosphate transporter traffic facilitator1(phf1)/atpht1;4 double mutants are impaired in Pi uptake from soil to roots. In addition, Wang et al. [27] reported that WRKY45 is a positive regulator of AtPHT1;1, and the Arabidopsis transgenic WRKY45-over-expressing lines showed induced AtPHT1;1 expression and increased Pi uptake. By contrast, the atpht1;1 mutants showed decreased Pi uptake in comparison with wild-type (WT) plants. In addition to the four major transporters (AtPHT1;1 to AtPHT1;4), the roles of other AtPHT1 proteins in Pi acquisition and transport cannot be neglected. For example, Nagarajan et al. [24] demonstrated that the Arabidopsis AtPHT1;5 plays a significant role in Pi translocation from source to sink organs. The authors reported that under low-Pi conditions, Arabidopsis atpht1;5-1 mutants exhibited a significant decline in Pi translocation into the shoots, and induced expression of several Pi starvation-response genes [e.g., At4, digalactosyldiacylglycerol synthase 1 (DGD1), and UDP-sulfoquinovose synthase 1 (SQD1)] [24]. However, under Pi-sufficient conditions, the Arabidopsis atpht1;5-1 mutants had higher shoot-Pi content than WT, suggesting that AtPHT1;5 plays a significant role in Pi homeostasis between the source and sink organs, in accordance with the Pi states and developmental cues [24]. In Arabidopsis, AtPHT1;6 and AtPHT1;7 regulate Pi translocation from leaves to pollens under Pi-sufficient conditions, while AtPHT1;8 and AtPHT1;9 control Pi translocation from roots to shoots, but not from the soil to roots, especially under Pi-deficient conditions [9,24,36].
The Arabidopsis phosphate transporter 1 (PHO1) gene family is another important Pi-transporter family, playing an essential role in long-distance Pi transport from roots to shoots, and is responsible for the regulation of Pi export from root epidermal and cortical cells into xylem vessels [40,41] (Figure 1). In Arabidopsis, AtPHO1 is mainly expressed in the lower part of the hypocotyl and the stellar cells of the roots, which is in agreement with its role in Pi transport to the xylem. It was reported that the atpho1 mutants were deficient in loading Pi from root epidermal cells into the xylem vessels, resulting in Pi deficiency in the shoots [40,42]. Among several AtPHO1 homologs in Arabidopsis, only AtPHO1 and AtPHO1;H1 could recover the defects of the atpho1 mutation, demonstrating that only AtPHO1 and AtPHO1;H1 are implicated in long-distance Pi transport from roots to shoots [32]. Recently, AtPHO1;H3 was shown to be involved in the suppression of root-to-shoot Pi transport under zinc-deficient conditions in Arabidopsis [33], while the roles of remaining transporter genes, AtPHO1;H2, and from AtPHO1;H4 to AtPHO1;H10, still need to be identified. In rice (Os), all three PHO1 (OsPHO1;1, OsPHO1;2, and OsPHO1;3) genes possess a cis-natural antisense transcript positioned at the 5 end of the genes, and only OsPHO1;2 is highly induced in the roots under Pi-deficient conditions [43]. Characterization of the ospho1;1 and ospho1;2 single mutants indicated that only ospho1;2 mutants had a significant decrease in Pi transport from roots to shoots, which was accompanied by high root-Pi and low shoot-Pi contents [43].
The rice OsPHT1 family comprises 13 Pi transporters [28], and with the exception of OsPHT1;3, 5, 7, and 12, all other OsPHT1 genes were studied in detail. Sun et al. [44] demonstrated that OsPHT1;1 was expressed in various shoot and root cells under Pi-sufficient conditions, and transgenic rice plants overexpressing OsPHT1;1 exhibited higher shoot-Pi content than WT, suggesting that OsPHT1;1 has a crucial role in Pi uptake and translocation under Pi-sufficient conditions. On the other hand, the low-affinity OsPHT1;2 and high-affinity OsPHT1;6 are strongly upregulated under Pi-deficient conditions in rice roots and shoots [35]. The authors suggested that OsPHT1;6 plays a broad role in Pi uptake from soil to roots, and probably in Pi translocation throughout the plant, whereas OsPHT1;2 mediates only Pi transport from roots to shoots [34]. OsPHT1;4 is expressed in roots, leaves, ligules, stamens, and caryopses, and is induced under long-term Pi-deficient conditions [28]. Transgenic OsPHT1;4-overexpressing rice plants displayed a significant increase in Pi contents in roots and shoots, whereas ospht1;4 mutants exhibited a decrease in Pi contents in the respective organs [28,29]. In addition, a gradual increase in the relative expression of OsPHT1;4 was observed in the embryos from 10 to 20 days after pollination when compared with that in the panicle axis and endosperm, suggesting a potential role of OsPHT1;4 in Pi acquisition during embryogenesis [29]. In rice, OsPHT1;4 is also implicated in Pi remobilization from flag leaves to panicles, while OsPHT1;6 and OsPHT1;8 are involved in Pi remobilization from senescing leaves to young ones and grains [28,45,46]. Additionally, using 33 Pi as a radiotracer, plausible involvement of OsPHT1;4, OsPHT1;6 and OsPHT1;8 in root-to-shoot Pi transport was also demonstrated [29,45,47]. A recent study by Jia et al. [47] reported that the transgenic rice OsPHT1;8-overexpressing lines exhibited a significant increase in both root-and shoot-Pi contents; however, the transgenic plants showed toxicity symptoms with a decrease in root and shoot biomass under highly Pi-supplied conditions. However, under limited-Pi conditions, OsPHT1;8-overexpressing lines exhibited normal root and shoot biomass similar to that of WT plants [47]. Thus, appropriate regulation of cytoplasmic Pi and/or organelle Pi levels is important for maintaining plant fitness under Pi-deficient conditions. Table 3 summarizes transporters involved in subcellular Pi transport. The Pi transport in various subcellular organelles is discussed below under separate subheadings.

Vacuole: The Pi Storehouse
Vacuoles are considered as the storehouse of Pi in plant cells. Under sufficiently Pi-supplied conditions, 70-95% of the intracellular Pi is stored inside the vacuoles in the form of orthophosphate in vegetative cells, whereas in plant seeds, Pi is stored in specialized protein storage vacuoles in the phytate form [7]. Since Pi may not be available at optimal concentrations during the entire course of the plant's life cycle, excess Pi is taken up and stored in vacuoles under the conditions of high Pi availability [1]. Whenever there is a decline in Pi concentration in the cytosol, Pi supplies are operated by the vacuole-Pi pool [1]. Hence, optimized Pi influx and efflux from vacuoles are essential for maintaining Pi homeostasis in other organelles, tissues, and also at the whole-plant level. Recently, Liu et al. [1] reported the functional characterization of the Arabidopsis vacuolar phosphate transporter 1 (AtVPT1), an SYG1/PHO81/XPR1 (SPX) major facilitator superfamily3 (MFS3) domain protein, that mediates Pi influx from the cytoplasm into the vacuoles (Figure 2). Transgenic Arabidopsis atvpt1 mutant plants were not able to retain more Pi content in the vacuoles than WT plants under high-Pi conditions, suggesting that disruption of AtVPT1 reduced the adaptability of plants to the changed Pi availability [1]. In addition, AtVPT1 expression was detected in younger tissues under normal conditions, but the expression level increased drastically in older leaves under highly Pi-supplied conditions, suggesting that AtVPT1 functions in detoxification resulting from high Pi levels in older tissues [1]. Liu et al. [48] identified a vacuolar transporter, named AtPHT5;1, in Arabidopsis via analysis of Pi translocation across the tonoplast using 31 P-magnetic resonance spectroscopy. The authors found that AtPHT5;1 mediated Pi influx from the cytoplasm into vacuoles, and the AtPHT5;1-overexpressing lines accumulated more Pi in the vacuoles, leading to a decreased cytoplasmic-Pi level under Pi-repleted conditions (Table 3). Liu et al. [48] also heterologously expressed an AtPHT5;1 homolog, the rice OsSPX-MFS1 gene in yeast, and reported that OsSPX-MFS1 localized in the tonoplast of yeast cells, and mediated Pi influx into yeast vacuoles. By contrast, the transgenic rice OsSPX-MFS3-overexpressing lines exhibited a decrease in vacuole-Pi content, suggesting that OsSPX-MFS3 acts as a vacuolar low-affinity Pi transporter involved in Pi efflux from the vacuoles [19] (Figure 2). The storage capacity for minerals in the vacuoles of various leaf cell types is compositionally different, and the mechanisms and reasons underpinning this compartmentalization are still elusive. For instance, in Arabidopsis, potassium (K + ) and Pi accumulate preferentially in the vacuoles of bundle sheath and epidermal cells, whereas magnesium (Mg 2+ ) and Ca 2+ are stored at high concentrations (>60 mM) only in the vacuoles of mesophyll cells [49,50]. In addition, Ca 2+ and Pi do not co-localize at high concentrations in the same vacuoles, as this results in insoluble calcium phosphate complexes [49,50]. However, an opposite trend was observed in cereal monocots, including Sorghum bicolor and wheat (Triticum aestivum), in which Ca 2+ accumulated in the epidermal cells, whereas K + and Pi accumulated in the mesophyll or bundle sheath cells [49,50]. Therefore, the cellular location of a particular mineral element is vigorous within an individual plant; however, the cell type that accumulates a given element can vary among plant species. The Ca 2+ /H + -antiporter (CAX1) was reported as a key regulator for cell-specific storage of Ca 2+ to optimize transpiration, cell-wall expansion, and plant productivity [49,50]; however, identification of the mechanisms involved in the cell-specific storage of Pi is a future task. The storage capacity for minerals in the vacuoles of various leaf cell types is compositionally different, and the mechanisms and reasons underpinning this compartmentalization are still elusive. For instance, in Arabidopsis, potassium (K + ) and Pi accumulate preferentially in the vacuoles of bundle sheath and epidermal cells, whereas magnesium (Mg 2+ ) and Ca 2+ are stored at high concentrations (>60 mM) only in the vacuoles of mesophyll cells [49,50]. In addition, Ca 2+ and Pi do not co-localize at high concentrations in the same vacuoles, as this results in insoluble calcium phosphate complexes [49,50]. However, an opposite trend was observed in cereal monocots, including Sorghum bicolor and wheat (Triticum aestivum), in which Ca 2+ accumulated in the epidermal cells, whereas K + and Pi accumulated in the mesophyll or bundle sheath cells [49,50]. Therefore, the cellular location of a particular mineral element is vigorous within an individual plant; however, the cell type that accumulates a given element can vary among plant species. The Ca 2+ /H + -antiporter (CAX1) was reported as a key regulator for cell-specific storage of Ca 2+ to optimize transpiration, cell-wall expansion, and plant productivity [49,50]; however, identification of the mechanisms involved in the cell-specific storage of Pi is a future task. Table 3. List of transporters involved in phosphate transport in Arabidopsis and rice (Oryza sativa) at the subcellular level. ANTR1, putative anion transporter 1 (thylakoid Na + -dependent phosphate transporter); PHO, phosphate transporter; PHT, high-affinity phosphate transporter; PT, phosphate transporter; SPX-MFS3, SPX major facilitator superfamily 3.

Chloroplast
The chloroplast is an important organelle which generates ATP, the prime source of energy for multiple reactions inside the cell, through photophosphorylation [60]. For efficient ATP synthesis, the reactants (ADP and Pi) must be transported into the chloroplasts, while the product (ATP) must be simultaneously exported out of the chloroplasts to avoid any feedback inhibition [18]. In Arabidopsis, a total of 16 plastidic phosphate translocator (pPT) genes are present, including those encoding the triose phosphate (TP)/phosphate translocator (TPT), the phosphoenolpyruvate (PEP)/Pi translocator (PPT), the xylulose-5-phosphate/phosphate translocator (XPT), and the glucose-6-phosphate/phosphate translocator (GPT) proteins [61] (Figure 2). In addition, some truncated versions of PPTs, GPTs, and phosphate translocator homologs (PThs) were also detected [61]. In Arabidopsis, two plastidic ADP/ATP translocator (AATP1) and H +/ Pi symporter (AtPHT2;1) were also identified and studied in detail [51,62] (Table 3). In order to maintain optimal rates of photosynthesis, triose-Pi is exported from chloroplast stroma to the cytosol by TPT in exchange for cytosolic Pi [63]. However, this transport activity depends on Pi availability, which is released during sucrose synthesis from the triose-Pi [63]. During the daytime, when sucrose synthesis slows down, Pi levels are reduced, leading to lower TPT activity and the diversion of triose-Pi into starch synthesis [64]. In addition, Pi release also maintains photosynthetic carbon assimilation [64]. Furthermore, if the sucrose biosynthesis pathway is impaired, Pi availability is reduced, leading to reduced ATP synthesis as found in the rice tpt mutants [63]. In addition to the TPT proteins, PPTs that mediate PEP/Pi exchange are present in plastid inner envelope membranes, and their encoding genes are exclusively expressed in both photosynthetic and heterotrophic tissues in Arabidopsis [61]. Since Pi transport in chloroplasts is tightly linked to the import and export of other metabolic intermediates, chloroplast-Pi homeostasis might play a role not only in maintaining ATP synthesis, but also in overall plant metabolism [61].
In Arabidopsis, members of the AtPHT2 and AtPHT4 families are expressed in chloroplasts [60]. Of the AtPHT2 family, AtPHT2;1 was found to be expressed predominantly in green tissues, and localize on the chloroplast inner envelope membrane [51]. Furthermore, Arabidopsis atpht2;1 mutants were shown to lack the ability to transport Pi inside the chloroplasts [51] (Table 3). Guo et al. [52] analyzed five members of the AtPHT4 family in Arabidopsis, and found that AtPHT4;1 and AtPHT4;4 are localized in leaf chloroplasts, AtPHT4;3 and AtPHT4;5 in shoot plastids, and AtPHT4;2 in root plastids. All AtPHT4 proteins were found to mediate H + -dependent Pi transport; however, the direction of the Pi transport was not conclusively demonstrated [65]. On the basis of the transport activity of expressed proteins in yeast, Guo et al. [65] suggested that AtPHT4 proteins might be involved in Pi export from plastids toward the cytoplasm. Irigoyen et al. [53] studied AtPHT4;2, and suggested that it may function in Pi export from plastids. The atpht4;1 mutants have growth defects which are associated with reduced Pi levels in stroma and decreased ATP synthase activity in chloroplasts [66] ( Figure 2).

Mitochondria
In addition to chloroplasts, mitochondria also generate ATP through oxidative phosphorylation. Pi is imported into mitochondria through a Pi carrier (PiC), while ATP is exported from mitochondria via an ADP/ATP carrier (AAC) [67]. PiC acts as a Pi/H + -symporter (influx of Pi and H + ) or Pi/OH − -antiporter (influx of Pi, with efflux of OH − ), and also as an exchange route for matrix/cytosolic Pi [68]. Uncoupling proteins (UCPs) are located in mitochondrial inner membranes, and dissipate the proton gradient across the inner mitochondrial membrane without ATP synthesis, thereby affecting oxidative phosphorylation in plants [69]. Another transporter, namely the dicarboxylate transport protein (DTP), catalyzes the exchange of Pi and dicarboxylate compounds like malate, succinate and malonate, and can also influence Pi concentrations in mitochondria in peas (Pisum sativum) [68]. There are also mitochondrial membrane-localized Pi transporters (MPT) that are localized in the mitochondrial inner membrane, and function in the Pi/H + -symport into the mitochondrial matrix. Among the three members of the Arabidopsis AtMPT family, the functions of AtMPT2 and AtMPT3 in Pi transport are known. AtMPT3 overexpression in Arabidopsis alters plant growth by increasing ATP and reactive oxygen species (ROS) levels, and by boosting respiration rate [70]. Furthermore, PHT3 family members can also execute Pi transport into the mitochondria [57]. In Arabidopsis, three members of the AtPHT3 family, namely AtPHT3;1, AtPHT3;2, and AtPHT3;3, are expressed in stems, leaves, and flowers, and are assumed to be involved in Pi import into mitochondria [54,57] (Figure 2). Liu et al. [57] performed homology-based analysis of Arabidopsis and rice phosphate transporter (PT) genes, and identified OsPT15, 16,17,18,19, and 20 genes in a group with Arabidopsis AtPHT3 genes ( Figure 2). However, these rice OsPT proteins are localized in other organelles rather than the mitochondria [57]. Specifically, OsPT15, 17, 18, and 19 are localized in peroxisomes, OsPT16 in the endoplasmic reticulum (ER), and OsPT17 in the PM [57], which, together with the findings above, may suggest the functional divergence of these PT proteins in rice and Arabidopsis [54,57]. As such, mitochondrial Pi transporters of the PT family in rice are yet to be identified.

Cellular Movement of Pi Transporters Involves the ER
Phosphate transporter traffic facilitator 1 (PHF1) is a major regulator controlling the exit and movement of Pi transporters from the ER to the PM [71]. Ruan et al. [72] were able to modify the promoter of the rice transporter traffic facilitator 1 (OsPHF1), a major factor regulating the ER-exit of Pi transporters to the PM, by using the cis-regulator high-affinity phosphate starvation response 1 binding sequence (HA-P1BS). The rice OsPHF1-HA-P1BS transgenic lines exhibited moderate upregulation of the OsPHF1 gene, and improved Pi uptake efficiency under Pi-starvation conditions without adverse effects on growth in comparison with the transgenic plants constitutively overexpressing OsPHF1 that displayed Pi overaccumulation in the shoots, and necrosis in the leaves [72]. In Arabidopsis, all PHT1 genes also contain the P1BS cis-regulatory element in their promoter, whereas in rice, except for OsPHT1;1 and OsPHT1;4, other PHT1 members have the P1BS cis-element in their promoters. Additionally, OsPHT1;2 was characterized as a direct target of OsPHR2 [73]. Transgenic rice plants overexpressing OsPHF1 show remarkably higher Pi concentrations in shoots under both Pi-deficient and Pi-sufficient conditions than WT plants [74].
The trafficking of Pi transporters from the ER to the PM is regulated by phosphorylation mechanisms. A rice kinase subunit, casein kinase 2β3 (CK2β3), was found to interact with OsPT2 and OsPT8, and the holoenzyme, CK2α3/β3, was able to phosphorylate OsPT8 under Pi-sufficient conditions [75]. The phosphorylated OsPT8 could not interact with PHF1 for its movement from the ER to the PM under sufficiently Pi-supplied conditions; however, under Pi-starvation conditions, CK2β3 is degraded, and OsPT8 and PHF1 are able to interact. Therefore, under Pi-sufficient conditions, CK2α3/β3 acts as a negative regulator of PTs through the phosphorylation of PTs, thereby inhibiting the PHF1-mediated PT trafficking to the PM [75].
Endosomal complex required for transport (ESCRT) protein complexes assist in the trafficking of cargo proteins coming from the Golgi apparatus or the PM through intraluminal vesicles (ILV) of multivesiclar bodies (MVBs), prior to their release into the vacuolar lumen [76]. In a recent study, the cytosolic apoptosis-linked gene 2-interacting protein X (ALIX) was found to be associated with MVBs through the ESCRT-III subunit, sucrose non fermenting protein 7 (SNF7), for the trafficking of AtPHT1;1 into the vacuoles in Arabidopsis. The Arabidopsis alix-1 mutants showed reduced trafficking of AtPHT1;1 into vacuoles for degradation, and enhanced Pi-starvation responses [76]. Correct ER-to-PM trafficking of AtPHT1;1 protein requires the function of AtPHF1 [71,77]. Under sufficiently Pi-supplied conditions, a CK holoenzyme phosphorylates AtPHT1;1, inhibiting its exit from the ER [76]. Alternatively, AtPHT1;1 protein is sent to vacuoles for degradation to decrease its levels at the PM, resulting in reduced Pi uptake under Pi-sufficient conditions [76]. Therefore, the involvement of the ER is crucial in maintaining the optimum concentration of PHTs on the PM in Arabidopsis.

Golgi Bodies
The Golgi body engages in several important reactions in plant cells, including the trafficking of proteins, and carbohydrate synthesis. The Golgi compartment hosts several ATP-and nucleotide sugar-dependent reactions that lead to the net release of Pi [78]. This released Pi needs to be exported back to the cytoplasm for the maintenance of Pi homeostasis. However, there is a huge knowledge gap on the Pi transport system in Golgi bodies. Subcellular localization studies in Arabidopsis indicated that only one member of the AtPHT4 family (i.e., AtPHT4;6) was targeted to the Golgi apparatus [52] ( Figure 2). Mutation of this transporter led to inhibited growth and changes in the composition of the cell wall [58]. The reduction in AtPHT4;6 activity resulted in changes in Pi intracellular movement, and Golgi-mediated processes such as protein glycosylation [58]. AtPHT4;6 is responsible for the transport of Pi from Golgi bodies toward the cytosol [79] (Table 3). Transgenic Arabidopsis atpht4;6 plants exhibited dark-induced senescence, suggesting that AtPHT4;6 regulates the critical function of controlling cellular Pi levels for the proper functioning of primary metabolism [59] (Figure 2).

Conclusions and Future Research Directions
The identification and characterization of various Pi transport proteins led to a significant accumulation of knowledge about Pi uptake and movement in plants. There are still major gaps in understanding Pi localization and transport activities among subcellular organelles within a plant cell under the conditions of both Pi-sufficient and Pi-deficient conditions. Research aiming to clarify how Pi transport activities are coordinated among various subcellular organelles, as well as cell types, under various developmental and environmental conditions is a future task, and requires a substantial effort from the research community. Despite the crucial roles of SPX-MFS proteins as vacuolar Pi transporters in buffering the cytoplasmic-Pi concentration against variable Pi availability, the molecular identity of the regulatory mechanisms via which Pi is translocated across the tonoplast remains elusive. Therefore, identification of other proteins involved in the efflux and influx of vacuolar Pi is needed. Additionally, the role of autophagy as an alternative cellular mechanism in recycling Pi through the delivery of RNA into vacuoles requires more attention. Although Pi is released from the RNA degradation in vacuoles, the contribution of such a Pi pool to the cytoplasmic Pi metabolism, and its biological relevance remain to be elucidated. To better understand the dynamics of cellular Pi homeostasis in a physiological context, we need new non-destructive quantitative methods with high detection sensitivity for measuring/visualizing Pi and relevant Pi-compounds, and their content in various subcellular compartments. With the recent development in next-generation sequencing technologies, several Pi deficiency-responsive micro RNA (miRNA) members have been identified in various plant species, but their roles in Pi uptake and transport are still far from being comprehensive. Furthermore, future research is also needed to discover the precise inter-relationships between Pi and other mineral (major and minor) elements. The roles of other important players, like redox and hormone homeostasis, in the regulation of Pi uptake and transport also need to be delineated. The efficiency of Pi use within a plant system needs to be examined in depth, which would ultimately enable us to enhance nutritional values and the adaptation potential of crops to meet the food demand of the ever-increasing population. This would also decrease the dependency on fertilizers, and would help in sustaining agricultural yields in the near future.

Conflicts of Interest:
The authors declare no conflict of interest.