The Roles of Phosphorus and Nitrogen Nutrient Transporters in the Arbuscular Mycorrhizal Symbiosis

More than 80% of land plant species can form symbioses with arbuscular mycorrhizal (AM) fungi, and nutrient transfer to plants is largely mediated through this partnership. Over the last few years, great progress has been made in deciphering the molecular mechanisms underlying the AM-mediated modulation of nutrient uptake progress, and a growing number of fungal and plant genes responsible for the uptake of nutrients from soil or transfer across the fungal–root interface have been identified. In this review, we outline the current concepts of nutrient exchanges within this symbiosis (mechanisms and regulation) and focus on P and N transfer from the fungal partner to the host plant, with a highlight on a possible interplay between P and N nutrient exchanges. Transporters belonging to the plant or AM fungi can synergistically process the transmembrane transport of soil nutrients to the symbiotic interface for further plant acquisition. Although much progress has been made to elucidate the complex mechanism for the integrated roles of nutrient transfers in AM symbiosis, questions still remain to be answered; for example, P and N transporters are less studied in different species of AM fungi; the involvement of AM fungi in plant N uptake is not as clearly defined as that of P; coordinated utilization of N and P is unknown; transporters of cultivated plants inoculated with AM fungi and transcriptomic and metabolomic networks at both the soil–fungi interface and fungi–plant interface have been insufficiently studied. These findings open new perspectives for fundamental research and application of AM fungi in agriculture.


Introduction
Phosphorus (P) and nitrogen (N) are key nutrients that play major roles in crop production. However, there is not enough P in the soil that can be directly used by plants. On the one hand, most soil P is bound to organic molecules or mineral surfaces, or precipitated as insoluble phosphate (Pi); on the other hand, the mobility of P in soil is low, so it is more difficult for plants to obtain P [1]. In many ecosystems, Pi levels limit plant growth, which has a significant impact on agriculture, especially in areas where low-input agriculture is practiced. In addition, nitrate (NO 3 − ) and ammonium (NH 4 + ) and organic N in the forms of amino acids and peptides are the dominant forms of N that are available to plants, and plant N use efficiency is low, rarely exceeding 50% in most agricultural systems due to leaching and run-off losses and gaseous N emissions [2,3]. Because of the status of Pi and N in soil, there is a great demand for chemical fertilizers to improve nutrient deficiency [4]. According to the International Fertilizer Association (IFA) (https://www.fertilizer.org/ accessed on 1 September 2021, by the end of 2020, the global requirement of chemical fertilizers of N, P, and K was expected to reach 260 million tons. Excessive application of chemical fertilizer will not only increase the production cost but also cause a series of environmental problems. Plants have evolved a number of physiological changes to overcome scarce levels of Pi and N, and one of these strategies is to form symbiotic associations with AM fungi [5]. Over 80% of terrestrial plant species can form symbiotic relationships with AM fungi, and these partnerships began more than 450 million years ago [6]. Arbuscules are formed in cortical cells after plant roots were colonized by AM fungi, which are surrounded by a plant-derived periarbuscular membrane (PAM). The PAM and arbuscular membrane form the main symbiotic interface for bidirectional transfers of nutrients between plants and fungi [7]. The fungi facilitate plant uptake of mineral nutrients and the plants supply carbohydrates and lipids to fungi as a source of carbon for their metabolic needs [8]. AM fungi play important roles in terrestrial ecosystems and contribute to improving agricultural productivity; therefore, it is important to understand the regulation of nutrient transport and metabolism associated with the symbiotic state of AM fungi [9]. Previous studies showed that mycorrhizal rice could receive more than 40% of its N, and that 70% of the total P acquired by mycorrhizal rice was through the mycorrhizal pathway [10,11].
In the past, there has been an emphasis on the study of nutrient transport by AM symbiosis from a physiological perspective; however, research indicates an increase in the number of genes that may be involved in the transport of nutrients through the soil fungus interface or the fungus root interface [12]. In this review, we outline the current concepts (mechanisms and regulation) of nutrient exchange in this symbiotic relationship, with a focus on P and N transfer from the fungal partner to the host plant.

Development of AM Symbiosis
The history of mycorrhizae started with their description, interpretation, and naming by A.B. Frank in 1885 [13]. The formation of them is the result of long-term coevolution between the root system of plants and AM fungi, AM association requires the creation of a novel symbiotic interface within the root cells [2]. Development of AM symbiosis starts with a signal exchange between AM fungi and host plants [14], which can be divided into distinct stages: (1) At the onset of symbiosis, plants secrete strigolactones (SLs) to stimulate spore germination and hyphal branching near the root [15]. SLs, a class of carotenoid-derived terpenoid lactones, were initially characterized as germination stimulants for root parasitic plants such as witchweeds. After 40 years, SLs were identified to be root-derived symbiotic signals for AM fungi [16], this is also a very important factor that triggered the change in the development of AM fungi. (2) Once fungi perceive SLs, AM fungi release signaling molecules termed "mycorrhizal factors", including lipochitooligosaccharides (LCOs) and short chitin oligomers (Cos), to activate a common symbiotic signaling pathway [17].
(3) Then, fungal hyphae penetrate the root epidermis, invade cortical cells, and develop highly branched tree-like structures called arbuscules, where the nutrient exchange between fungus and plant takes place. Mycorrhizal colonized roots have two pathways for nutrient absorption: direct absorption through root epidermis or root hairs (direct pathway), and indirect absorption via AM fungi hyphae (mycorrhizal pathway) [18,19] (Figure 1). A recent study by Shi et al. [20] demonstrated that P starvation response transcription factors connect the mycorrhizal P uptake pathway with the direct P uptake pathway to improve P absorption.

P Uptake System in Plants: Lessons from Arabidopsis and Tomato
In the direct pathway, Pi is acquired mainly through low or high H + /Pi transporters on the epidermis. These phosphate transporters (PHTs) are either constitutively expressed or induced to be expressed in response to Pi deficiency. In roots, Pi is taken up through PHTs and loaded into the xylem for translocation to the shoots [21][22][23].
In Arabidopsis, PHTs involved in the transfer of Pi into plant cells have been identified, and these proteins have been classified into four families: PHT1, PHT2, PHT3, and PHT4 ( Figure 2) [24]. Nine transporters of the PHT1 family have been identified in Arabidopsis: AtPHT1.1-AtPHT1. 9. The functions of AtPHT1.1, AtPHT1.4, AtPHT1.5, AtPHT1.8, and AtPHT1.9 have been defined, and all of them are high-affinity PHTs that are localized in the plasma membrane and induced by Pi starvation. AtPHT1.1 and AtPHT1.4 are the most highly expressed PHTs in Arabidopsis [25]. The polypeptides encoded by AtPHT1.1 and AtPHT1.4 showed 78% similarity to each other and showed high amino acid sequence similarity to the high-affinity PHTs of Saccharomyces cerevisiae, Neurospora crassa, and Glomus versiforme [26][27][28]. AtPHT1.8 and AtPHT1.9 are highly expressed in roots and are also involved in Pi absorption [29]. In Arabidopsis roots and rosette leaves, Pi deficiency resulted in increased abundance of AtPHT1.5 transcripts, which are required for proper translocation of AtPHT1.5 Pi from roots to stems under these conditions. However, the expression of AtPHT1.5 under Pi replete conditions is different from that under Pi-deficient conditions. AtPHT1.5 is mainly expressed in shoots and is responsible for Pi transport to sinks and from shoots to roots under Pi-replete conditions [30]. AtPHT2.1, the only PHT2 family member in Arabidopsis, is a single-copy gene that shows shoot-specific expression independent of the external Pi concentration, and the localization of the AtPHT2.1 green fluorescent protein fusion protein indicates that it is present inside the chloroplast intima [31]. In Arabidopsis, PHT3 is also called MPT3, which has

P Uptake System in Plants: Lessons from Arabidopsis and Tomato
In the direct pathway, Pi is acquired mainly through low or high H + /Pi transporters on the epidermis. These phosphate transporters (PHTs) are either constitutively expressed or induced to be expressed in response to Pi deficiency. In roots, Pi is taken up through PHTs and loaded into the xylem for translocation to the shoots [21][22][23].
In Arabidopsis, PHTs involved in the transfer of Pi into plant cells have been identified, and these proteins have been classified into four families: PHT1, PHT2, PHT3, and PHT4 ( Figure 2) [24]. Nine transporters of the PHT1 family have been identified in Arabidopsis: AtPHT1.1-AtPHT1.9. The functions of AtPHT1.1, AtPHT1.4, AtPHT1.5, AtPHT1.8, and AtPHT1.9 have been defined, and all of them are high-affinity PHTs that are localized in the plasma membrane and induced by Pi starvation. AtPHT1.1 and AtPHT1.4 are the most highly expressed PHTs in Arabidopsis [25]. The polypeptides encoded by AtPHT1.1 and AtPHT1.4 showed 78% similarity to each other and showed high amino acid sequence similarity to the high-affinity PHTs of Saccharomyces cerevisiae, Neurospora crassa, and Glomus versiforme [26][27][28]. AtPHT1.8 and AtPHT1.9 are highly expressed in roots and are also involved in Pi absorption [29]. In Arabidopsis roots and rosette leaves, Pi deficiency resulted in increased abundance of AtPHT1.5 transcripts, which are required for proper translocation of AtPHT1.5 Pi from roots to stems under these conditions. However, the expression of AtPHT1.5 under Pi replete conditions is different from that under Pi-deficient conditions. AtPHT1.5 is mainly expressed in shoots and is responsible for Pi transport to sinks and from shoots to roots under Pi-replete conditions [30]. AtPHT2.1, the only PHT2 family member in Arabidopsis, is a single-copy gene that shows shoot-specific expression independent of the external Pi concentration, and the localization of the AtPHT2.1 green fluorescent protein fusion protein indicates that it is present inside the chloroplast intima [31]. In Arabidopsis, PHT3 is also called MPT3, which has three members: AtMPT1, AtMPT2, and AtMPT3. PHT4 family has six members: five of the proteins are targeted to the plastid envelope, and the sixth resides in the Golgi apparatus, these genes are not responsive to Pi deprivation. The PHT4 gene is expressed in both roots and leaves, with AtPHT4.1 and AtPHT4.4 having the highest expression levels in the leaves. With the exception of AtPHT4.2 and AtPHT4.6, the other four members of this gene family are more abundantly expressed in leaves than in roots. Thus, the plastid-localized protein members of this family may function in both heterotrophic plastids and photosynthesis [33].
The PHT1 family in tomato consists of seven members: SlPT1-SlPT7 ( Figure 2) [34]. Transcripts of SlPT1 and SlPT7 are detected in large numbers in roots and leaves and, to a lesser extent, in stems and flowers as well as in fruits. The relatively high transcript levels of SlPT1 and SlPT7 in these sink tissues suggest that they may have evolved to meet the requirements for transporting Pi from source to sink [35]. The expression of SlPT2 and SlPT6 show relatively distinct tissue-specific profiles, with their transcripts intensively in roots and extremely faintly in some other tissues [36]. The expression patterns of SlPT3 and SlPT5 are relatively similar, and both genes are weakly expressed in all tissues. However, the transcript of SlPT4 is not detectable in all tissues examined [37]. The expression of SlPT1, SlPT2, SlPT3, SlPT5, SlPT6, and SlPT7 are significantly repressed under high-Pisupply conditions. The expression of these genes is significantly reduced in response to high P availability, which may be partly due to the significant increase in P concentration in tomato plants under these conditions [38].  The PHT1 family in tomato consists of seven members: SlPT1-SlPT7 ( Figure 2) [34]. Transcripts of SlPT1 and SlPT7 are detected in large numbers in roots and leaves and, to a lesser extent, in stems and flowers as well as in fruits. The relatively high transcript levels of SlPT1 and SlPT7 in these sink tissues suggest that they may have evolved to meet the requirements for transporting Pi from source to sink [35]. The expression of SlPT2 and SlPT6 show relatively distinct tissue-specific profiles, with their transcripts intensively in roots and extremely faintly in some other tissues [36]. The expression patterns of SlPT3 and SlPT5 are relatively similar, and both genes are weakly expressed in all tissues. However, the transcript of SlPT4 is not detectable in all tissues examined [37]. The expression of SlPT1, SlPT2, SlPT3, SlPT5, SlPT6, and SlPT7 are significantly repressed under high-Pi-supply conditions. The expression of these genes is significantly reduced in response to high P availability, which may be partly due to the significant increase in P concentration in tomato plants under these conditions [38].

Transport of P in AM Symbiosis
The transfer of Pi from AM fungi to host plants was revealed by using isotopically labeled substrates in a two-compartment system [39]. In AM-colonized rice, 70% of Pi was transported via the mycorrhizal pathway [10]. In mycorrhizal plants consisting of the rhizosphere and the hyphosphere, AM fungi take up Pi from the soil through Pi transporters located on the plasma membrane of extraradical hyphae. The absorbed Pi is rapidly converted to polyphosphate (poly-P) and isolated in tubular vesicles, where the IRM long-chain polyphosphate is broken down into [10] shorter chains by a vacuolar endopoly-phosphatase. Afterwards, polyphosphatase hydrolyzes the terminal residues of short-chain polyphosphate and releases Pi, which can be transferred to the host plant through the mycorrhizal interface [40][41][42] (Figure 1).

Pi Transport in AM Fungi
Pi is taken up by the mycelium of AM fungi, transferred to intraradicular fungal structures, and released into the periarbuscular space (PAS) containing arbuscule cells, and the transport proteins may be involved in symbiotic transport by the fungi. In particular, AM fungi have high-affinity Pi/H + transporters, which are homologues of the yeast PHO84 Pi transporter and belong to a family of major transporters similar to plant Pi transporters [3,43]. Four Pi transporters of AM fungi have been identified: Rhizophagus irregularis (GiPT), Funneliformis mosseae (GmosPT), Rhizophagus clarus (GvPT), and Gigaspora margarita (GigmPT) ( Table 1). They all share structural and sequential similarities with other plant and fungal high-affinity PTs [28,[43][44][45]. GvPT and GiPT transcripts were detected mainly in extraradical mycelium, indicating their role in acquiring Pi from the soil. GiPT expression was correlated with the external Pi concentration and over all Pi content of mycorrhizal roots. GmosPT had high sequence similarity (73% identity and 86% similarity with GiPT), and they were expressed in both extra-and intraradical mycelium, but not in germinating and dormant sporocarps. The phylogenetic analysis showed that GmosPT, GvPT, and GiPT groups have close sequences [43,44,46]. In addition, GigmPT is considered to be the major Pi sensor in Gigaspora margarita and is involved in the acquisition of Pi from PAS via upregulation of the phosphate signaling (PHO) pathway, as well as sensing changes in extracellular Pi through activation of the protein kinase A (PKA) signaling cascade [47].

AM Symbiosis Affects Plant Pi Transport
The Pi uptake pathway of mycorrhizae may dominate Pi uptake in AM symbiosis, which is heavily dependent on AM-induced PHT1 members [48,49]. The transcript levels of many PHT1 transporter proteins decrease with increasing Pi levels, and the expression of a small subgroup of PHT1 transporter proteins in AM symbiosis is actually induced in mycorrhizal roots [3]. In AM symbiosis, two classes are known to be involved in Pi transport: mycorrhiza-specific Pi transporters and mycorrhiza-inducible Pi transporters [50].
Mycorrhizal-specific PHT1 members have been identified in many plants, and are expressed in response to symbiotic relationships (Table 1). Immunolocalization and expression studies on MtPT4 of Medicago truncatula showed that this transporter has a subcellular targeting effect on plant PAM. The rice homolog of MtPT4, OsPT11, showed a similar localization [51][52][53]. In addition, tracer studies confirmed that Pi was not delivered through the fungus during os-pt11 symbiosis [10]. Due to gene duplication, there are two orthologues of OsPT11 in tomato: SlPT4 and SlPT5. SlPT4 is exclusively expressed during symbiosis, unlike OsPT11, and it is dispensable for symbiotic P uptake.
Phylogenetic tree analysis showed that SlPT4 and StPT4 clustered with mycorrhizaspecific Pi transporter proteins from Medicago truncatula, indicating that two nonhomologous mycorrhizal response genes encoding Pi transporter proteins are coexpressed in Solanaceae [38]. In addition, extensive sequencing of StPT4-like clones and subsequent expression analysis in potato and tomato revealed the existence of a close homologous sequence between StPT4 and SlPT4, named StPT5 and SlPT5, respectively, representing a third Pi transport system in Solanaceae species, which is upregulated upon root coloniza-tion by AM fungi [34,37,38]. The AM-inducible Pi transporter of Lotus japonicus, LjPT3, is expressed in arbuscule-containing cells of the inner cortex. It is especially noteworthy that the LjPT3 gene is not orthologous to MtPT4, therefore, it is possible that other AMenhanced transporters may exist in Lotus japonicus [54]. In Populus trichocarpa, only PtPT10 transcripts were highly enriched in mycorrhizal roots, and promoter sequence analysis revealed conserved motifs in PtPT10 that were similar to other AM-inducible homologues. Mycorrhizal Pi transport in Astralagus sinicus was mediated by the specific Pi transporter AsPT4 but not AsPT1, the loss of AsPT4 function resulted in a block in symbiotic Pi uptake. Knockdown of AsPT1 also affected the growth of arbuscules, but did not alter Pi transfer in AM symbiosis, indicating compensatory effects between the two transporters [50]. Reverse genetic analysis showed that AsPT5 not only mediated Pi translocation and remodeled root system architecture, but was also critical for Astragalus sinicus arbuscule formation under moderately high Pi concentrations [55]. In addition, ZmPT6 is a mycorrhiza-specific Pi transporter gene of maize, and a PT6 mutant showed reduced mycorrhiza formation in maize roots [56]. Most interestingly, the AM-inducible transporter SbPT10 was only detected in roots colonized by AM fungi, not in leaf or stem tissue [57]. GmPT10 and GmPT11 obtained from Glycine max are induced upon fungal colonization; however, GmPT7 is only induced in the later stages of symbiosis [58]. In petunia, the expression of PhPT4 is mycorrhiza-specific, and PhPT3 and PhPT5 are mycorrhiza-inducible, both of them were expressed at low levels in nonmycorrhized roots [59]. VvPT1 and VvPT2, identified in the grape genome, code for putative proteins with a high level of similarity with a series of specific PHTs [60].
Mycorrhiza-inducible PHTs were strongly induced by AM symbiosis, but a small amount was expressed in uninoculated roots (Table 1). RNA hybridization assays in tomato using the StPT3 probe indicated that SlPT3 may be a homolog of StPT3 [34,37], and RNA localization and reporter gene expression indicated that StPT3 is expressed in root sectors where mycorrhizal structures are formed [61]. In solanaceous species, the AM-regulated PHT genes SmPT3 in eggplant, CfPT3 in pepper, and NtPT3 in tobacco have also been identified [62]. The accumulation of mycorrhizal maize ZmPT1, ZmPT3, ZmPT4, and ZmPT5 transcripts under low-Pi conditions was positively correlated with shoot biomass, which may be due to P accumulation [63,64]. Homologous PtPT8 is phylogenetically related to the AM-inducible PHT1 subfamily [65]. In Brachypodium distachyon, BdPT7, the orthologue of MtPT4, was highly induced in mycorrhizal roots, and its transcripts accumulated not only in mycorrhizal roots but also in noncolonized roots and leaves of Pi-starved plants. BdPT3, BdPT12, and BdPT13 are also induced during AM symbiosis, similar to OsPT13 [62,66]. In barley seedlings, AM colonization specifically upregulated the expression of HvPHT1;11, HvPHT1;11.2, HvPHT1;12, and HvPHT1;13.1/13.2 [67]. RT-PCR and in situ hybridization showed that the HvPT8 and TaPht-myc transporters had increased expression in roots colonized by AM fungi [68]. The expression of AM-inducible PHT1 genes (SbPT9, SbPT11, LuPT5, and LuPT8) in both root and leaf tissues indicates that these transporters not only play a role in mycorrhizal Pi uptake but also in Pi mobilization in leaves. SbPT8 also induced nonmycorrhizal roots under low-Pi conditions, suggesting a possible change from the direct Pi uptake pathway to the mycorrhizal Pi uptake pathway during the establishment of AM symbiosis [57]. Lotus japonicus allows mycorrhizal plants to take up Pi from their fungal partners and regulates morphogenesis in mycorrhizal plants, and LjPT4 may play an additional role in the root tip when AM fungi are absent [69].
Entry of Pi into plant cells via PTs through the plasma membrane requires protonation and deprotonation of the transporter, accompanied by conformational changes [19]. In recent years, AM-responsive HA genes have been identified in several plants and are considered key genes for the activation and regulation of the symbiotic interface secondary transport system (Table 1). Notably, MtHA1 from Medicago and OsHA1 from rice are the only two known HA genes whose expression is exclusively confined to specific root cells containing AM fungal structures [70,71]. In tobacco, two HA genes (PMA2 and PMA4) were found to be induced in cortical cells containing arbuscules of mycorrhizal roots [72].
A homolog of OsHA1 and MtHA1, SlHA8, was identified in tomato, and it was strongly and specifically induced to be expressed in roots colonized by the AM fungi. The SlHA8 promoter is able to drive GUS reporter gene expression in soybean and rice mycorrhizal roots, suggesting that AM-induced H + -ATPase gene expression is highly conserved in different mycorrhizal plants [73,74]. AM symbiosis has received increasing attention for its potential exploitation of the nutrients of crop plants, especially in sustainable agriculture. The function of the arbuscular plasma membrane H + -ATPase in energizing nutrient transfer may well be employed in crop improvement [71]. These findings offer new insights into the regulatory mechanism of mineral nutrient uptake by host plants from AM fungi.
AM colonization is usually accompanied by reduced expression of other Pi transporters, especially those involved in the direct pathway (Table 1). In rice, the expression of each gene in relation to AM symbiosis was examined. Although the mRNA levels of OsPT1, OsPT2, OsPT3, OsPT6, OsPT9, and OsPT10 expressed in roots declined during symbiosis, in Ospt11 Pi transporter mutants, the expression of OsPT2 and OsPT6 genes was not reduced, suggesting an elaborate regulation of Pi acquisition via the direct pathway by the mycorrhizal pathway [10,19,75]. This downregulation of Pi transporters has been observed in potato, and the abundance of the transcripts of StPT1 and StPT2 was reduced, probably as a result of the plant fungus interaction and/or an improved Pi status of the mycorrhizal roots [61,76]. Therefore, the root epidermis and AM fungi can balance the uptake of Pi.

Summary and Phylogenetic Analysis of the PHTs
P is acquired by a mycorrhizal P transporter expressed in ERM and then translocated to IRM, and then taken up by plant cells through PHT proteins. In recent years, there has been a significant increase in our understanding of the physiology and molecular mechanisms of PHT proteins in AM symbiosis. In this part of the review, we discuss the key regulatory role played by PHTs in non-AM (for example, Arabidopsis and tomato) and in AM fungi formation. Sequence comparison between characterized PHTs from the nonmycorrhizal plants Arabidopsis and Brassica napu, the mycorrhizal plants Solanum Lycopersicum, Oryza sativa, and Nicotiana tabacum, and other AM plant species illustrates some interesting features of the PHT family ( Figure 3). The results showed that most mycorrhiza-specific and mycorrhiza-upregulated PHTs clustered into their respective subgroups, and these subgroups contained two mycorrhizal nonresponsive PHTs (NtPT5 and BdPT3) and did not contain any Arabidopsis and Brassica napus PHTs. Observations suggest that mycorrhizal plants have PHTs that are adapted to AM symbiosis. On the other hand, the IV subfamily mainly contains mycorrhizae-induced PHTs and four downregulated PHTs. These results indicate that the expression of mycorrhiza-induced PHTs is often accompanied by the downregulation of other PHTs. Interestingly, four mycorrhizae-specific PHT genes (AsPT5, SbPT10, GmPT11, and LjPT3) do not cluster with the mycorrhizae-specific subgroup and instead cluster with the other Pht1 members. This may be an evolutionary strategy to ensure a balance between mycorrhizal and direct uptake of P.
We covered the transport proteins involved in Pi fluxes from plants towards fungi, uptake from the soil, and exchange of P together; interestingly, we found P transporters to be less studied in different species of AM fungi. On the plant side, other PHT genes have thus far received less attention than PHT1 family genes in response to mycorrhizal plants. We may pay attention to this research in the future. These advances in the comprehension of PHTs will help underpin the development of crops with optimal P uptake efficiencies. We covered the transport proteins involved in Pi fluxes from plants towards fungi, uptake from the soil, and exchange of P together; interestingly, we found P transporters to be less studied in different species of AM fungi. On the plant side, other PHT genes have thus far received less attention than PHT1 family genes in response to mycorrhizal plants. We may pay attention to this research in the future. These advances in the comprehension of PHTs will help underpin the development of crops with optimal P uptake efficiencies.   (Table 1). Mycorrhizal-induced (blue) and downregulated Pi transporters (green) are highlighted in different color.

N Uptake System in Plants: Lessons from Arabidopsis and Tomato
Nitrate and ammonium are the main forms of N in soils [79]. To obtain N from soil, plants have evolved several N-absorbing systems. Understanding the molecular mechanisms of how plants absorb and assimilate N is a critical step to improve plant N use efficiency. To date, five transporter families are known to be involved in N uptake, distribution, or storage: ammonium transporters (AMT), nitrate transporter 1, peptide transporter family (NPF(NRT1/PTR)), nitrate transporter 2 (NRT2), slow anion channelassociated homologues (SLAC/SLAH), and chloride channel family (CLC). In this section, the current studies of N transporters in Arabidopsis and tomato are reviewed (Figure 4). plants have evolved several N-absorbing systems. Understanding the molecular mechanisms of how plants absorb and assimilate N is a critical step to improve plant N use efficiency. To date, five transporter families are known to be involved in N uptake, distribution, or storage: ammonium transporters (AMT), nitrate transporter 1, peptide transporter family (NPF(NRT1/PTR)), nitrate transporter 2 (NRT2), slow anion channel-associated homologues (SLAC/SLAH), and chloride channel family (CLC). In this section, the current studies of N transporters in Arabidopsis and tomato are reviewed (Figure 4). AtNPF6.3 (CHL1/AtNRT1.1) is the first identified nitrate transporter, functioning in nitrate uptake in roots and nitrate translocation from roots to shoots [80,81]. AtNPF2.7 (NAXT1) mainly expresses the cortex of mature roots and is implicated in root nitrate uptake [82]. AtNPF4.6 (NRT1.2/AIT1) encodes a constitutive component of the low-affinity nitrate uptake transporter [83]. AtNPF3.1 (AtNirt) encodes a pathogen-induced NO3 − /NO2 − transporter [84]. AtNPF7.3 (NRT1.5) is a low-affinity, pH-dependent bidirectional nitrate transporter, located in the plasma membrane and expressed in the periplasmic cells of the root whole near the xylem, responsible for xylem nitrate loading [85]. AtNPF7.2 (NRT1.8) is expressed predominantly in xylem parenchyma cells and plays an important role in plant removal of nitrate from xylem vessels [86]. AtNPF2.9 (NRT1.9) may facilitate loading of nitrate into the root phloem and enhance downwards transport of nitrate in roots [87]. AtNPF2.10 and AtNPF2.11 are expressed in leaves and silique walls, and regulate the loading of glucosinolates from the apoplasm into the phloem [88]. AtNPF2.3 is a constitutively Figure 4. Integrative illustration on physiological functions of those nitrate and ammonium transporters are well investigated using Arabidopsis and tomato as a model. HATS, high-affinity transport system; LATS, low-affinity transport system.
AtNPF6.3 (CHL1/AtNRT1.1) is the first identified nitrate transporter, functioning in nitrate uptake in roots and nitrate translocation from roots to shoots [80,81]. AtNPF2.7 (NAXT1) mainly expresses the cortex of mature roots and is implicated in root nitrate uptake [82]. AtNPF4.6 (NRT1.2/AIT1) encodes a constitutive component of the low-affinity nitrate uptake transporter [83]. AtNPF3.1 (AtNirt) encodes a pathogen-induced NO 3 − /NO 2 − transporter [84]. AtNPF7.3 (NRT1.5) is a low-affinity, pH-dependent bidirectional nitrate transporter, located in the plasma membrane and expressed in the periplasmic cells of the root whole near the xylem, responsible for xylem nitrate loading [85]. AtNPF7.2 (NRT1.8) is expressed predominantly in xylem parenchyma cells and plays an important role in plant removal of nitrate from xylem vessels [86]. AtNPF2.9 (NRT1.9) may facilitate loading of nitrate into the root phloem and enhance downwards transport of nitrate in roots [87]. At-NPF2.10 and AtNPF2.11 are expressed in leaves and silique walls, and regulate the loading of glucosinolates from the apoplasm into the phloem [88]. AtNPF2.3 is a constitutively expressed transporter whose contribution to NO 3 − translocation to the shoot is quantitatively and physiologically significant under salinity [89]. AtNPF6.2 (AtNRT1.4) regulates nitrate homeostasis in leaves, whose deficiency can alter leaf development [90]. AtNPF2.13 (NRT1.7) is expressed in leaf veinlet siliques and is responsible for silique loading of nitrate in the source leaf to allow nitrate translocation from older to younger leaves [91]. The lowaffinity nitrate transporter AtNPF2.12 (NRT1.6) is only expressed in reproductive tissue and is involved in delivering nitrate from maternal tissue to the developing embryo [92]. Another transporter expressed in seeds, AtNPF5.5, also affects N accumulation in embryos [93]. AtNPF1.1 (NRT1.12) and AtNPF1.2 (NRT1.11) are low-affinity nitrate transporters that are also involved in transferring nitrate from xylem to phloem [94]. AtNPF4.1 and AtNPF4.3 can transport abscisic acid [95]. AtNPF5.2 protects the plant against biotic and abiotic stresses and transports dipeptides and tripeptides [96]. AtNPF6.4 is induced by NO 3 − in the leaves, but its expression is inhibited in the roots. AtNPF8.1, AtNPF8.2, and AtNPF8.3 are localized at the plasma membrane, and facilitate the transport of dipeptides with high affinity [95,97]. SlNRT1.1 and SlNRT1.2 are the first N transporters identified in tomato that mediate nitrate uptake in roots [98]; they were also found to be low-affinity nitrate transporters, and their expression was root-cell-specific and regulated by N availability. SlNRT1.1 can improve nitrate uptake in grafted tomato plants under high N demand [99].
In contrast to the transporters of the NRT1 family, the NRT2 family mainly regulates the high-affinity transport system (Figure 4). With the exception of AtNRT2.7, all remaining NRT2 transporters interacted strongly with AtNAR2.1 [100]. Disruption of AtNRT2.1 and AtNRT2.2 reduces the inducible high-affinity transport system by up to 80%, and the constitutive high-affinity transport system is reduced by 30% [101]. The expression of AtNRT2.4 is induced under N starvation, and it is mainly expressed in the epidermis of lateral roots and the phloem of shoots [102]. AtNRT2.5 is a plasma membrane-localized high-affinity nitrate transporter protein that plays an important role in adult plants under severe N starvation [103]. AtNRT2.6 is involved in rhizobacterium-stimulated lateral root growth. SlNRT2.1 and SlNRT2.2 are highly similar in coding regions, the expression of SlNRT2.1 and SlNRT2.2 are restricted to roots, and the highest expression level of SlNRT2.1 occurs in the anthesis stage [104]. SlNRT2.3 formation was positively controlled by nitrate and negatively by ammonium, but not by glutamine, and it is expressed in tomato roots colonized by AM fungus [105]. The SlNRT2.4 gene was expressed in several tissues and organs with the lowest expression level. For subcellular localization, all NRT2 proteins were predicted to be in the plasma membrane [106].
In addition to the NRT1 and NRT2 families, some members of the CLC family also have nitrate transport capacity. In Arabidopsis, CLCa and CLCb act as proton-nitrate exchanges and are more selective for nitrate than for chloride [107].
Previous studies on phylogenetic analyses of the AMT gene family revealed two distinct subfamilies (Figure 4): the AMT1 subfamily (AMT1 cluster) and the AMT2 subfamily (AMT2/3/4 cluster) [108]. The AMT1 cluster gene encodes a protein with high affinity for NH 4 + transport. Transcriptome and RNA gel blot analyses showed that four of the six AMT homologues in Arabidopsis are expressed in roots and upregulated under N-deficient conditions [109]. AtAMT1.1 is a root NH 4 + transporter that confers approximately one-third of the overall high-affinity transport capacity of N-deficient Arabidopsis [110]. Similar to AtAMT1.1, AtAMT1.3 also accounted for approximately 30% of the overall ammonium uptake capacity in N-deficient Arabidopsis roots. Root ammonium influx in N-deficient plants was 60-70% lower in the atamt1.1 and atamt1.3 double mutants, indicating that AtAMT1.1 and AtAMT1.3 are functionally additive under N-deficient conditions [111]. However, AtAMT1.4 is specifically expressed in pollen, which contributes to N nutrients in pollen via NH 4 + uptake or retrieval [108]. AtAMT1.5 accounts for the remaining ammonium uptake capacity [109]. AtAMT2 may play a role in the transport of NH 4 + from the apoplast to the symplast [112]. SlAMT1.1 was the first ammonium transporter identified in tomato to function not only in ammonium uptake but also in ammonium translocation from roots to shoots; it was also found to be strongly induced under low N, and downregulated by drought and salt stress [98,113,114]. SlAMT1.2 was strongly induced by NH 4 + and NO 3 − , increased in leaves at the onset of light, and decreased when CO was elevated. In contrast to other AMT transporters, SlAMT1.3 is exclusively detected in the leaves and is lower in the light period, higher in the dark, and decreased with elevated CO [113]. Ruzicka et al. [115] identified two novel tomato ammonium transporter genes, SlAMT4 and SlAMT5. Quantitative real-time PCR (qPCR) analysis revealed that they were exclusively expressed in mycorrhizal roots, although they were not significantly regulated by NH 4 + treatments.

N Assimilation in Arabidopsis and Tomato
For many plants, roots take up nitrate and assimilate it, where it is first reduced to nitrate-by-nitrate reductase in the cytoplasm and then further to ammonium by nitrate reductase in the plastids and synthetase (GS) in the plastids and cytoplasm, and then transported to the shoot. The ammonium derived from nitrate or directly ammonium uptake by ammonium transporters (AMTs) is further assimilated into amino acids via the GS/glutamine-2-oxoglutarate aminotransferase (GOGAT) cycle [116]. Plant GS occurs in most species as a single isoform in plastids (GS2) and as three to five isoforms localized in the cytosol (GS1). Cytosolic GS1 is important for primary NH 4 + assimilation in roots and for reassimilation of NH 4 + generated during protein turnover in leaves, whereas the dominating role of GS2 is in reassimilation of photorespiratory NH 4 + in chloroplasts and assimilation of NH 4 + derived from NO 3 − reduction in plastids [117,118]. Fd-GOGAT and NADH-GOGAT are two kinds of GOGAT species from higher plants, while Fd-GOGAT is derived from photorespiration and mainly assimilates ammonium in leaves. NADH-GOGAT is highly expressed in roots [119]. In Arabidopsis, GLN1;1, GLN1;2, and GLN1;4 are induced during leaf senescence, and GLN1;1 is located in the root surface, root tips, and root hairs. GLN1;2 and GLN1;3 are localized in the vasculature, and GLN1;2 is the only one that is significantly upregulated by ammonium. GLN1;3 is expressed in the root mature zone. GLN1;4 is expressed within the basal region of lateral root emergence. In Arabidopsis, two Fd-GOGATs have been identified: GLU1 and GLU2. GLU1 is expressed abundantly in the leaves, while GLU2 is mainly expressed in the roots [79]. In tomato, Liu et al. [120] identified six GS genes (SlGSI, SlGS2, and SlGS1.1-1.4) and two GOGAT genes (SlNADH-GOGAT and SlFd-GOGAT). These genes underwent species-specific evolution, and may have specific biological functions in vivo. SlGS1.1, SlGS1.2, SlGS1.3, and SlNADH-GOGAT may cooperatively play significant roles in primary N assimilation in roots.
Three enzymes also probably participate in ammonium assimilation, except for the GS/GOGAT cycle. Cytosolic asparagine synthetase (AS) catalyzes the ATP-dependent transfer of the amido group of glutamines to a molecule of aspartate to generate glutamate and asparagine [121] (Figure 5).

N Assimilation in Arabidopsis and Tomato
For many plants, roots take up nitrate and assimilate it, where it is first reduced to nitrate-by-nitrate reductase in the cytoplasm and then further to ammonium by nitrate reductase in the plastids and synthetase (GS) in the plastids and cytoplasm, and then transported to the shoot. The ammonium derived from nitrate or directly ammonium uptake by ammonium transporters (AMTs) is further assimilated into amino acids via the GS/glutamine-2-oxoglutarate aminotransferase (GOGAT) cycle [116]. Plant GS occurs in most species as a single isoform in plastids (GS2) and as three to five isoforms localized in the cytosol (GS1). Cytosolic GS1 is important for primary NH4 + assimilation in roots and for reassimilation of NH4 + generated during protein turnover in leaves, whereas the dominating role of GS2 is in reassimilation of photorespiratory NH4 + in chloroplasts and assimilation of NH4 + derived from NO3 − reduction in plastids [117,118]. Fd-GOGAT and NADH-GOGAT are two kinds of GOGAT species from higher plants, while Fd-GOGAT is derived from photorespiration and mainly assimilates ammonium in leaves. NADH-GOGAT is highly expressed in roots [119]. In Arabidopsis, GLN1;1, GLN1;2, and GLN1;4 are induced during leaf senescence, and GLN1;1 is located in the root surface, root tips, and root hairs. GLN1;2 and GLN1;3 are localized in the vasculature, and GLN1;2 is the only one that is significantly upregulated by ammonium. GLN1;3 is expressed in the root mature zone. GLN1;4 is expressed within the basal region of lateral root emergence. In Arabidopsis, two Fd-GOGATs have been identified: GLU1 and GLU2. GLU1 is expressed abundantly in the leaves, while GLU2 is mainly expressed in the roots [79]. In tomato, Liu et al. [120] identified six GS genes (SlGSI, SlGS2, and SlGS1. 1-1.4) and two GOGAT genes (SlNADH-GOGAT and SlFd-GOGAT). These genes underwent species-specific evolution, and may have specific biological functions in vivo. SlGS1.1, SlGS1.2, SlGS1.3, and SlNADH-GOGAT may cooperatively play significant roles in primary N assimilation in roots.
Three enzymes also probably participate in ammonium assimilation, except for the GS/GOGAT cycle. Cytosolic asparagine synthetase (AS) catalyzes the ATP-dependent transfer of the amido group of glutamines to a molecule of aspartate to generate glutamate and asparagine [121] (Figure 5).

Transport of N in AM Symbiosis
For N, the supply rate of N depends on the mineralization of organic N into inorganic N (NH 4 + or NO 3 − ) by microorganisms [122]. Determination of the 15 NO 3 − and 15 NH 4 + uptake rates indicated that AM fungi contributed more than the N uptake of the plants alone [123]. AM fungi maintains an extraradical mycelium that can extend several centimeters from the root. The IRM within the root are connected to the ERM and form a single continuum [124]. Many studies have reported that ERM can take up 42% of N via the mycorrhizal uptake pathway to plants [125,126]. Seventy five percent of the of N found in Zea maize leaves was taken up via the ERM of Glomus aggregatum [127]. According to current knowledge, AM fungi take up of N in the ERM is assimilated through the GS/GOGAT cycle, preferentially in the form of ammonium, and metabolizes it to arginine, which is the main form in which N is transported from the ERM to the intraradical mycelium (IRM) and broken down into urea and ornithine in the IRM. Ammonium, the product of urea hydrolyzation, is subsequently released to the symbiotic interface and taken up by plants [128][129][130].

N transport System in AM Fungi
AM fungi prefer the direct uptake of NH 4 + owing to the extra energy needed for the reduction of NO 3 − , which is required for nitrogen to incorporate into organic compounds [12]. Here, we report three fungal AMT genes that are attained for NH 4 + uptake by AM fungi (Figure 1). The first AMT of AM fungi, GintAMT1, characterized from R. irregularis, expressed in the ERM, encodes a high-affinity NH 4 + transporter [131]. Functional complementation in an AMT-defective yeast mutant showed that GintAMT2 encodes a functional NH 4 + transporter, and plasma membrane localization was revealed by polyclonal antibodies against GintAMT2. GintAMT1 and GintAMT2 are both expressed in ERM and IRM, which participate in NH 4 + uptake in soil solutions and may be involved in the recovery of NH 4 + leaked during fungal metabolism at the symbiotic interface [129]. GintAMT3 localizes to the fungal membrane and encodes a functional low-affinity transporter [125].
Nitrate uptake by the ERM of R. irregularis is probably coupled to a H + -symport mechanism [132]. Nitrate transporters have been shown to play vital roles in NO 3 − transport to the ERM. GiNT, identified from R. irregularis, was shown to be expressed in all AM fungi tissues (spores, arbuscules, ERM, and IRM) and could play an important role in establishing competition for NO 3 − between the plant and AM fungi at the symbiotic interface by regulating bidirectional fluxes ( Figure 1) [128,133].
AM fungi can also enhance the decomposition of N capture from complex organic material in soil, and hyphal growth of the fungal partner was increased in the presence of the organic material, independent of the host plant [134]. Amino acid transport systems have been studied extensively in higher plants, yeast, and filamentous fungi. Based on bioinformatics tools, polymerase chain reaction and heterologous expression systems have been used to characterize the Funneliformis mosseae amino acid permease (GmosAAP1) sequence. GmosAAP1 was expressed in ERM but not in IRM structures of plants treated with millimolar nitrate concentrations [135]. Genome-wide transcriptomic data obtained from R. irregularis were exploited, and RiPTR2 showed amino acid sequence and transmembrane domain profiles similar to those of members of the PTR2 family of fungal oligopeptide transporters. The RiPTR2 sequence was able to complement the growth defects of yeast mutants defective in the two well-studied dipeptide transporters. At least in the heterologous system, RiPTR2 was able to transport Ala-Leu, Ala-Tyr, Tyr-Ala, and other dipeptides [136].

N Transport Systems on the Plant Side
Plant transporters located in PAM are responsible for the capture of nutrients from the periarbuscular apoplast and their delivery into the cytoplasm of cortical cells [12] (Table 2). LjAMT2.2 is the first plant AMT gene characterized to be involved in N uptake during AM symbiosis, and has been shown to be the highest upregulated gene in a transcriptomic analysis of Lotus japonicus roots upon colonization with Gigaspora margarita [137]. In tomato, the mycorrhizal specific transporter SlAMT4 was 68% identical to LjAMT2.2 [137], and SlAMT5 shares 64.4% amino acid identity with SlAMT4 [115]; RT-PCR indicated that five soybean genes (GmAMT1.4, GmAMT3.1, GmAMT4.1, GmAMT4.3, and GmAMT4.4) were upregulated in root colonization with AM, and promoter reporter analysis indicated that the most abundantly transcribed gene, GmAMT4.1, showed specific expression in arbuscule cortical cells [138]. In poplar, PtAMT1.1 and PtAMT1.2 were mycorrhiza-inducible in AM roots [139]. In Sorghum, the relative gene expression of SbAMT3;1 and SbAMT4 were significantly (70 and 20 times, respectively) higher in roots colonized by AM fungi than in nonmycorrhized roots, and SbAMT3;1 and SbAMT4 genes are expressed in root cortical cells, which makes them ready to accommodate arbuscules [140]. In M. truncatula, MtAMT2;3, MtAMT2;4, and MtAMT2;5 transcripts increase significantly during AM symbiosis [141].
Aquaporin-mediated membrane transport of ammonia has already been well analyzed, and aquaporins could be a component of the low-affinity ammonia transport system. An AM-specific Nod 26-like intrinsic protein, MtNIP1 (Table 2), showed strong induction expression during mycorrhization and possibly facilitated the cellular uptake of ammonia [142]. Laser microdissection revealed that the second NIP gene was expressed in the surrounding hyphae-containing cortical cells as well, and the two NIPs might be involved in the uptake of N into host cells [143].
AM fungi can influence NO 3 − uptake of plants by regulating the transcript levels of NRTs (Table 2). AM-induced NO 3 − transporters have been identified in a variety of plant species, including Medicago truncatula [6], Lotus japonicus [137], and grapevine [144]. When the roots were colonized by AM fungi, the expression of SlNRT2.3 extended to the inner cortical cells and the transcript levels of SlNRT2.3 in AM-colonized roots were higher than in noncolonized controls, which may mediate the positive effects of AM fungi on NO 3 − uptake from soil and NO 3 − distribution to the host [105]. Drechsler et al. [145] investigated the transcriptional regulation of 82 rice NPF genes in response to colonization by the AM fungi R. irregularis in roots of plants grown under five different nutrition regimes, and the expression of the NPF6.4, NPF2.2, and NPF1.3 genes was strongly induced in mycorrhizal roots and depended on the composition of the fertilizer solution. Wang et al. [11] proposed that OsNPF4.5 is a low-affinity NO 3 − transporter, and mycorrhizal colonization strongly induced the expression of OsNPF4.5 in rice roots, which is exclusively expressed in the cells containing arbuscules. The orthologues of NPF4.5 in maize (ZmNPF4.5) and sorghum (SbNPF4.5) were also found to be strongly upregulated in AM fungal-colonized roots [2]. However, the roles of these AM-induced NRTs in symbiotic NO 3 − transfer are still far from well understood due to the lack of precise identification of their subcellular localization and transport activities. Members of the plant NPF can transport not only NO 3 − but also oligopeptides across the plasma membrane [95]. Guether et al. [146] demonstrated that peptide transporter (LjPTR) transcripts were found exclusively in arbuscule cells of Lotus, which suggested that the corresponding transporter may be involved in N transfer from fungus to plant before or after arbuscule disintegration. AM fungi can also take up substantial amounts of amino acids, such as glycine, glutamic acid, glutamine, and aspartic acid. The LjLHT1.2 gene, encoding an LHT1-type amino acid transporter, was also consistently expressed in cortical cells of AM roots, where transcripts were localized mainly in arbuscular cells but also in the noncolonized cells of the root cortex [147][148][149].

Summary and Phylogenetic Analysis of N Transporters
This coordinated and specific expression of ammonium and nitrate transporters in mycorrhizae-colonized cortical cells suggests the crucial importance of fungal N transfer in plants. Although less studied, N is also a nutritional determinant of the interaction. In this part of the review, we focus on the roles of N transporters in AM fungi and on the plant side of mycorrhizal N transporter response. We also summarize the recent advances in N uptake, assimilation and translocation in AM symbiosis, and sequence comparison between N members from the nonmycorrhizal plants Arabidopsis and Brassica napus, the mycorrhizal plants Solanum Lycopersicum, Oryza sativa, and Nicotiana tabacum, and other AM plant species (Figure 6). In subgroups III, all NH 4 + transporters were induced by AM except AtAMT2. Arabidopsis cannot be colonized by AM fungi, indicating that the function of AMT genes in one clade is conserved. OsNPF4.5, SbNPF4.5, and ZmNPF4.5 were also found to cluster together, and they were strongly upregulated in roots colonized by AM fungi, suggesting that the symbiotic NO 3 − uptake route might be conserved in different plant species ( Figure 6). Nevertheless, this subfamily also has many other plant nitrate transporters that might be an evolutionary strategy to ensure a balance between mycorrhizal N uptake and symbiotic N fixation. Such inferences deserve further attention. transporters that might be an evolutionary strategy to ensure a balance between mycorrhizal N uptake and symbiotic N fixation. Such inferences deserve further attention.
We cover the transport proteins involved in N fluxes from plants towards fungi, to the uptake from the soil and exchange of N. Interestingly, similar to P transporters, N transporters are less studied in different species of AM fungi. On the other hand, on the plant side, N transporter genes have thus far received less attention than P transporter genes in response to mycorrhizal plants. We may pay attention to N research in the future.  Table 2). Mycorrhizal-induced transporters are highlighted in pink color.

Regulation of Nutrient Exchange by P and N in AM Symbiosis
The availability of soil-derived nutrients has long been recognized as one of the important environmental factors controlling mycorrhizal phenotypes [150]. P and N are two major nutrient elements needed by plants. AM fungi help host plants absorb more Pi and N from the soil; in turn, Pi and N regulate AM symbiosis. Maintaining proper Pi homeostasis is important for plant growth and development, as either too-low or too-high Pi concentrations in plant cells can harm plants during symbiosis, and plants must integrate Pi status with fungal colonization and arbuscule development to maintain beneficial in-  Table 2). Mycorrhizal-induced transporters are highlighted in pink color.
We cover the transport proteins involved in N fluxes from plants towards fungi, to the uptake from the soil and exchange of N. Interestingly, similar to P transporters, N transporters are less studied in different species of AM fungi. On the other hand, on the plant side, N transporter genes have thus far received less attention than P transporter genes in response to mycorrhizal plants. We may pay attention to N research in the future.

Regulation of Nutrient Exchange by P and N in AM Symbiosis
The availability of soil-derived nutrients has long been recognized as one of the important environmental factors controlling mycorrhizal phenotypes [150]. P and N are two major nutrient elements needed by plants. AM fungi help host plants absorb more Pi and N from the soil; in turn, Pi and N regulate AM symbiosis. Maintaining proper Pi homeostasis is important for plant growth and development, as either too-low or too-high Pi concentrations in plant cells can harm plants during symbiosis, and plants must integrate Pi status with fungal colonization and arbuscule development to maintain beneficial interactions [151]. Balzergue et al. [152] and Adeyemi et al. [153] reported that high levels of P fertilizer suppress mycorrhizal root colonization, which could be explained by the disruption of the symbiotic interaction of AM fungi by high P availability in soil. P could also inhibit AM fungi root colonization by suppressing the expression of plant symbiotic genes, especially genes encoding carotenoid and lactone biosynthesis enzymes. High levels of Pi directly inhibit spore germination by reducing the biosynthesis of strigolactone and symbiotically associated Pi transporters [153]. Plants constantly sense and signal Pi status in response to the environment. Pi starvation response 1 (PHR1) in Arabidopsis and its orthologues in other species play key roles in these processes by regulating Pi signaling and Pi homeostasis, and activate the expression of a broad range of Pi starvation-induced (PSI) genes by binding to the P1BS element in Pi-deficient conditions to improve plant Pi acquisition [154]. SPX proteins (named after the Saccharomyces cerevisiae SYG1 and Pho81 proteins and the mammalian Xpr1) have emerged as key sensors and signaling regulators of cellular Pi status in plants, and SPX proteins bind the affinity of PHRs to P1BS elements through protein-protein interactions in Pi-sufficient conditions, securing plant P homeostasis [20,151]. Recent mining of the published genomic and transcriptomic data from AM fungi detected the presence of genes encoding proteins containing the SPX domain [155]. Wang et al. [151] demonstrated that Medicago proteins SPX1 and SPX3 regulate Pi homeostasis and root colonization. Both AM symbiosis and light exposure increase nutrient uptake and utilization in plants, especially P. On the other hand, P content and the transcripts of AM-specific PTs were increased by R light under AM conditions. Light acts as a signal moving from shoots to roots in a phyB-HY5-dependent manner to regulate SL synthesis in roots. In particular, R light promotes P uptake of AM plants via both CCD7-dependent and CCD7-independent pathways [7].
It is commonly accepted that P appears to be a major regulator of AM symbiosis establishment and efficiency. In contrast, the inhibitory effect of plant N status seems to be more controversial; some studies have shown that low N stimulates mycorrhiza formation, while other studies obtained different results [156,157]. An increasing number of results suggests that N appears to be another regulator for the maintenance of the mutualistic functioning of AM symbiosis. Nouri et al. [158] reported that starvation for nitrate reversed the inhibitory effect of Pi on AM, suggesting that nutrient starvation triggers a major AM-promoting signal to counteract the effect of high Pi. Wang et al. [11] observed that, compared with 0.25 mM NO 3 − conditions, rice and sorghum supplied with either 2.5 mM or 5.0 mM NO 3 − results in increased AM colonization. Pan et al. [159], via gradients of long-term N addition in a Mongolian steppe, showed that low to moderate N, and N:P ratio increases were able to increase AM fungi parameters. Furthermore, Nanjareddy et al. [160] showed that NO 3 − can tremendously modify mycorrhizal morphology and behavior along with plant morphology.
Recent evidence has directly demonstrated that a complex interplay occurs between N and P homeostasis [155]. P and N deficiencies had cumulative effects on AM formation; LPN (combined low P and low N) increased AM levels and significantly induced the mycorrhizal marker genes MtPT4 and MtBCP1 in Medicago truncatula. Split-root experiments further showed that AM formation in LPN plants is systematically controlled not only by P but also by N, which supports the interaction between P and N in AM symbiosis [156]. Meanwhile, downregulating PTs in Medicago truncatula inhibits arbuscule development, which appears to be countered by low-N conditions. This response to low N seems to involve AMT2;3, and the simultaneous mutation of this AMT and PT restores the inhibition of AM fungi colonization. This indicates that AMT-mediated N transfer to the symbiotic interface also serves as a signal to regulate mycorrhizal colonization, in some cases going beyond the function of PT [141,161,162].

Concluding Remarks and Future Perspectives
To cope with nutrient deficiency, plants have evolved several promising strategies, and AM symbiosis is one of them. To date, an increasing number of plant and fungal genes have been identified and functionally characterized that are responsible for the transport of nutrients from the soil or across the intraradical symbiotic interfaces. This review highlights some key examples to show the integrated roles of nutrient transfers in AM symbiosis. Transporters belonging to the plant or AM fungi can synergistically handle the transmembrane transport of soil nutrients to the symbiotic interface for further plant acquisition. These transporters can also promote or inhibit the colonization of AM symbiosis based on the nutrient status of plants and soil. Sequence comparison between characterized P and N transporters from the nonmycorrhizal plant Arabidopsis and Brassica napus, Solanum lycopersicum, Oryza sativa, Nicotiana tabacum, and other AM plant species illustrates that plant species capable of forming mycorrhizas possess specialized types of nutrient transporters adapted to AM symbiosis. On the other hand, in contrast to the repressed expression of the Pi transporter protein gene responsible for the direct uptake pathway, AM symbiosis tends to upregulate NO 3 − transporter genes, and this difference may be partly attributed to the fact that plants require more N than P.
Although much progress has been made to elucidate the complex mechanism for the integrated roles of nutrient transfers in AM symbiosis, many questions still remain to be answered.
(1). P and N transporters are less studied in different species of AM fungi. Some mycorrhizalspecific nutrient transporter genes do not cluster with mycorrhizal-specific subgroups. It is tempting to speculate that this might be an evolutionary strategy that guarantees the balance between mycorrhizal uptake and direct uptake, and such an inference deserves further attention. (2). The involvement of AM fungi in plant N uptake is not as clearly defined as that of P. A better understanding of the mechanism and regulation of N uptake assimilation, translocation, and transfer to the host is important for potential applications of AM fungi. (3). Coordinated utilization of P and N is crucial for plants to maintain nutrient balance and achieve optimal growth. It has been proposed that the relative availability of soil P and N determines whether mycorrhizal benefits outweigh their costs. Increasing evidence suggests that P and N uptake and transport control mycorrhizal functioning. More studies are necessary to understand the role of mycorrhizal uptake pathways in P and N uptake. (4). Many studies regarding transporters have been conducted on model plants, and the results have been useful in elucidating key aspects, as a next step into bringing science and agriculture. However, the symbiotic status of cultivated plants with AM fungi has not been insufficiently studied. (5). In recent years, benefitting from the rapid progress in "omics" studies for both symbiotic partners, great progress has been made in our understanding of the P and N uptake mechanisms in AM symbiosis. Future research should address the analysis of transcriptomic and metabolomic networks at both the soil-fungi interface and fungi-plant interface to open up new perspectives in depicting a bigger picture regarding symbiosis-mediated nutrient signaling regulatory networks.