The cuttings of P. simonii (ca. 15 cm in length, 2 cm in diameter) were rooted in pots (10 L) filled with fine sand. The plants were cultivated in a glasshouse (natural light, 75% relative humidity) and irrigated with 50 ml nutrient solution (Hoagland and Arnon, 1950) (10 μM EDTA·FeNa, 5 μM MnSO₄·H₂O, 1 μM ZnSO₄·7H₂O, 1 μM CuSO₄·5H₂O, 30 μM H₃BO₃, 0.5 μM H₂MoO₄, 1000 μM MgSO₄·7H₂O, 1000 μM CaCl₂, 1000 μM Na₂SO₄, and 1000 μM NH₄NO₃, pH 6.5) containing Pi at one of three different levels (0 μM, 1000 μM, and 3000 μM KH₂PO₄) separately every other day. Each Pi treatment was applied to 12 plants for a period of 8 weeks. In addition, a phosphate-free solution was prepared by replacing KH₂PO₄ with K₂SO₄. Six plants for each Pi treatment were subjected to drought stress for 2 weeks before harvesting, and the other six plants for each treatment were not subjected to drought stress and served as controls. After the treatments, whole plantlets were harvested and separated into roots, stems, new leaves, and old leaves. Then, samples from the same tissues were pooled for each treatment, frozen in liquid nitrogen for further analysis.

To identify phosphate transporters in poplar, the reported protein sequences from both Arabidopsis and Oryza sativa were used as query sequence with E-value cutoff set as 1e⁻¹⁰ to perform a local BLASTP against P. trichocarpa JGI gene catalog (phytozome v11.0, https://phytozome.jgi.doe.gov/pz/portal.html), respectively. The choice of candidate phosphate transporters was based on the E-value, the sequence homology value (>60%) and the value of score (>450). All corresponding DNA and protein sequences of poplar phosphate transporters were obtained from Phytozome 11.0 (http://phytozome.jgi.doe.gov/pz/portal.html) and also verified with the Pfam database (http://pfam.xfam.org/). Various splicing variants of one gene and incomplete genes with short length were discarded. In addition, redundant sequences were removed, and the remaining sequences were used in further analyses.

For each putative protein, the grand average of hydropathicity (GRAVY) was calculated using ProtParam (http://web.expasy.org/protparam/). Further, TMpred (http://www.ch.embnet.org/software/TMPRED_form.html) was used to predict the transmembrane domains (TMDs) of each PHT protein. In addition, the subcellular location of each PHT was predicted using WoLF PSORT (http://www.genscript.com/wolf-psort.html). Multiple amino acid alignments of the PHTs were generated using ClustalW (Larkin et al., 2007). Phylogenetic tree was constructed using the neighbor-joining method with the 1000 bootstrap in MEGA 6 (http://www.megasoftware.net/).

The chromosomal locations and gene structures of the PHT genes were obtained from Phytozome 11.0. The chromosomal locations of the PHT genes were mapped with MapDraw program (Liu and Meng, 2003). The exon/intron organization of the individual PHT gene was drawn using Gene Structure Display Server (GSDS) program (Hu et al., 2015).

Promoter sequences located 2 kb upstream of the transcription start sites (ATG codons) of the Populus PHT genes were retrieved from the Populus trichocarpa genome v3.0 in Phytozome 11.0 (http://phytozome.jgi.doe.gov/pz/portal.html). The cis-regulatory elements were scanned in PLACE (Higo et al., 1999).

Total RNA was isolated from roots, stems, new leaves, and old leaves using a plant RNA extraction kit (R6827, Omega Bio-Tek, GA, USA) according to the manufacturer's instructions. The first-strand cDNA was synthesized using a PrimeScript RT Reagent Kit (DRR037S, Takara, Dalian, China) following the removal of trace genomic DNA using DNase I (E1091, Omega Bio-Tek). The specific primers (Table S1) for quantitative real time PCR analysis were designed by Primer Premier 6.0 (Premier Biosoft, Palo Alto, CA, USA). We performed PCR in a 20 μl reaction mixture containing 10 μl of 2X SYBR Green Premix Ex Taq II (Bioteke, China), 2 μl of cDNA, and 1 μl of 20 mM of each primer (Table S1) using a Roche LightCycle 96 machine (Roche, Germany). PCR amplification was performed under the following conditions: one cycle of 2 min at 95°C, followed by 45 cycles at 95°C for 10 s, 55°C for 20 s, and 72°C for 20 s. Actin2/7 was used as a reference gene (Brunner et al., 2004b). Three biological replicates with three technical replicates were assayed for each sample. Reactions for the reference gene were included in each plate. The relative expression levels of all the PtPHT genes were calculated using the 2^−ΔΔCT method (Livak and Schmittgen, 2001).

We identified a total of 42 putative PtPHT genes, including 12 PHO, 14 PtPHT1, 2 PtPHT2, 6 PtPHT3, and 8 PtPHT4 genes from the genome of P. trichocarpa according to the full AtPHT family in Arabidopsisf (Table 1). The lengths of the encoded proteins ranged from 314 amino acids to 802 amino acids, and their sequences contained 5 to 14 TMDs. The molecular weight (MW) of 42 putative proteins ranged from 33.5 to 93.5 kDa. The GRAVY value of putative phosphate transporter protein (PHT family) was positive and ranged from 0.128 to 0.795, while that of PHO family was negative ranging from −0.255 to −0.064.

The amino acid sequences of all of the PHT genes of P. trichocarpa were aligned with those of A. thaliana. The resulting phylogenetic tree contained five clusters with PHO, PHT1, PHT2, PHT3, and PHT4, respectively (Figure 1). Based on the phylogenetic analysis results, the P. trichocarpa PHT genes were named after the A. thaliana PHT genes (Figure 1, Table 1).

The PtPHT genes were marked on a physical map of Populus linkage groups based on the location information for PHTs in Phytozome 11.0. The Populus PtPHT genes showed a heterogeneous distribution pattern among the chromosomes (Figure 2). A total of 41 of the 42 PtPHT genes were localized to 15 of the 19 linkage groups of Populus (Figure 2). Only PtPHO12 was localized to unattributed scaffold fragments.

The lengths of the 42 PtPHT genes ranged from 945 to 2409 bp, and they contained 1 to 16 exons (Table 1). These genes were distributed across 15 of the 19 poplar chromosomes excluding chromosomes 3, 7, 11, and 13 (Table 1, Figure 2). In addition, the lengths of the 12 PtPHO genes ranged from 1890 to 2409 bp, and they contained 11 to 14 exons (Table 1). Most of these genes were localized to chromosomes 8 and 10 (Table 1). Further, the lengths of the 14 PtPHT1 genes ranged from 1311 to 1620 bp, and they contained 1 to 3 exons. Most of these genes were located on chromosomes 5 and 10 (Table 1). Two members of the PHT2 family, PtPHT2.1 and PtPHT2.2, had lengths of 963 and 1773 bp and contained 2 and 3 exons, respectively. The PHT3 family included six PtPHT genes in poplar, with lengths ranging from 945 to 1125 bp and containing 6 exons. Further, the PHT4 family included nine PtPHT genes, with lengths ranging from 864 to 1662 bp and containing 1 to 11 exons.

We further analyzed the exon/intron structures of the 42 PtPHT genes. The genes belonging to the same PHT family, with the exception of the PHT2 family, were well conserved in terms of their exon/intron structures, with similar exon lengths (Figure 3). Unexpectedly, compared with other members of the PHT4 family, which contained 7–16 exons, PtPHT4.6 possessed only one long exon, and its coding sequence was shorter. These substantial differences in gene structure are due to differences in the numbers of exons and introns among the various genes.

The 2 kb upstream regions of the 42 PtPHT genes were analyzed using PLACE signal scan program. A total of 234 putative cis elements were detected, many of which are involved in development, stress, or hormone activity (Table S2). Twenty-six of the detected cis elements were found in all of the PtPHT genes involved in energy, light, hormone, defense, carbon metabolism, and tissue-specific expression. For example, GATABOX, GT1CONSENSUS, and IBOXCORE are light-responsive cis elements, and CACTFTPPCAL is carbon metabolism related cis element (Gowik et al., 2004).

Some Pi-response and other stress-related cis regulatory elements including ABRE elements, helix–loop–helix elements, PHO-like, PHR1, TATA-box-like, and WRKY1 were detected in the upstream regions of PtPHT genes. AACACOREOSGLUB1, a WRKY1 element, and ACGTABREMOTIFA2OSEM, an ABRE element, were present in 22 and 7 PtPHT genes, respectively. In addition, A PHO-like element (AACACOREOSGLUB1) and a TATA-box-like element (TATABOX2) was detected in 22 and 34PtPHT genes, respectively; A PHR1 element (P1SB) was observed in 20 PtPHT genes; and a helix–loop–helix element, (CATATGGMSAUR) was present in 11 PtPHT genes. Furthermore, 17 PtPHT genes had typical cis elements that were present especially in the respective genes, and the total number of these cis elements was 32.

The specificities of tissues or organs may allow for the detailed elucidation of their functions. We measured the transcript levels of all of the PtPHT genes in 12 various tissues of Populus tremula using a public database, PopGenIE (http://popgenie.org/) (Figure 4). These tissues included roots, dormant cambium-phloem, mature petiole, freshly expanded leaves, expanding young leaves, mature leaves, suckers, dormant buds, dormant flowers, expanded flowers, expanding flowers, and mature seeds. Twenty-five PtPHT genes were highly expressed in the roots of P. tremula, suggesting that most PtPHT genes might be involved in Pi uptake. Six PtPHT genes, including PtPHO3, PtPHO4, PtPHO5, PtPHO6, PtPHT1.4, and PtPHT1.14, were highly expressed in more than six of the twelve investigated tissues. For example, PtPHO6 was highly expressed in nine tissues, including cambium, petiole, expanded freshly leaves, expanding young leaves, mature leaves, dormant buds, dormant flowers, expanding flowers, and mature seeds. In contrast, the expression levels of several PtPHT genes, including PtPHT4.3 and PtPHT3.3b, were very low in all of the investigated tissues. In particular, PtPHT1.9 and PtPHT1.10 were not expressed in any of the 12 tissues examined; this finding is not shown in Figure 4. Further, some PtPHT genes were only highly expressed in a few specific tissues. For example, PtPHT1.6 was exclusively expressed in the roots of P. tremula, while PtPHT1.5 and PtPHO9 were only highly expressed in the roots and suckers of P. tremula.

Real-time quantitative RT-PCR was also performed to analyze the transcript levels of all 42 PtPHT genes in the roots, stems, young leaves, and mature leaves of P. simonii. Transcripts of 35 PtPHT genes were detected in the examined tissues (Figure 5), but some of them were present at relatively low levels (such as PtPHT1.5/1.6, PtPHO1, and PtPHO2) (Figures 5A,D). Most of the PtPHT genes were confirmed to be expressed in the roots of poplar. Transcripts of 7 PtPHT genes (such as PtPHT1.3, PtPHT2.2, and PtPHT4.1a) were detected in all four tested tissues (Figures 5A–D). Most PtPHT genes were expressed in the roots of P. simonii at high levels. PtPHT1.8 and PtPHO7 exhibited increased transcript accumulation in the leaves compared to the roots (Figures 5A,D). In addition, PtPHT3.1a/b expression was similar in all four investigated tissues (Figure 5B).

To better understand the functions of PtPHT genes in relation to phosphate uptake, the transcription patterns of PtPHT genes in the roots of P. simonii were examined at varying phosphate levels. There were 33 PtPHT genes in total expressed in roots (Figure 6), but the expression of a few PtPHTs, such as PtPHT3.2a/b and PtPHO1/2/3, was very low (Figures 6B,D). In the roots, the expression of PtPHT1.12 was up-regulated and that of PtPHT1.5 was only slightly altered under phosphate starvation conditions; in addition, the expression of PtPHT1.2, PtPHT1.3, PtPHT2.2, PtPHT3.1b, and PtPHT4.1a was markedly up-regulated under low-phosphate conditions, and that of PtPHT3.3a, PtPHT4.1b, PtPHT4.5b, PtPHO8, and PtPHO9 was up-regulated under high-phosphate conditions (Figures 6A–D).

Most of the PtPHT genes were expressed at high levels in the roots of P. simonii in the presence of low phosphate levels, indicating that they are high-affinity genes. The accumulation of transcripts of many PtPHT genes was reduced when the level of H₂PO3- was increased in the roots of P. simonii, with the exception of PtPHO8 and PtPHO9.

To understand the transcriptional changes of PtPHT genes induced by drought, P. simonii plants supplied with different phosphate levels were subjected to drought stress for 2 weeks. PtPHT1.2 and PtPHO9 expression was found to be increased under drought conditions, irrespective of the phosphate level. In particular, PtPHT1.2 expression was significantly induced by approximately 90-fold. However, changes in the transcription of some PtPHT genes under drought stress were largely dependent on the phosphate level. For example, after 14 days of drought stress, the expression of PtPHT3.3b, PtPHT4.3, and PtPHT4.5a was significantly down-regulated, while that of PtPHT1.9, PtPHO4, and PtPHO9 was only slightly increased when no phosphate was supplied (Figure 7). However, the expression of these genes was up-regulated under drought stress in the presence of a low level of phosphate. Further, in the presence of a high level of phosphate, the expression of PtPHT1.9, PtPHT1.11, PtPHT1.3, PtPHT1.14, PtPHT2.2, PtPHT4.4, PtPHT4.6, PtPHO1, and PtPHO8 was increased under drought conditions. These data suggested that the changes in the transcription of some PtPHT genes under drought conditions were strongly influenced by the phosphate level.

In the current study, we conducted comprehensive analysis of PHT genes in poplar. A total of 42 PtPHT genes were retrieved from Populus genome (Phytozome 11.0, P. trichocarpa v3.0) with improved annotation, and they were classified into five clusters: PHT1, PHT2, PHT3, PHT4, and PHO. In a subsequent study, Loth-Pereda et al. (2011) reported the identification of 12 PHT1 family members in an earlier version of the Populus genome (v1), and we expanded upon their findings with the identification of 14 PHT1 family members in the latest version of Populus genome (v3). Notably, the PHT1 family is larger in poplar (14 genes) than in Arabidopsis (9 genes), barley (8 genes), rice (13 genes), and tomato (8 genes) (Muchhal et al., 1996; Rae et al., 2003; Liu et al., 2011; Chen et al., 2014).

The PHT2, PHT3, PHT4, and PHO families contained more members of PtPHT genes in poplar than that in Arabidopsis. One PHT2 family gene has been identified in Arabidopsis (Daram et al., 1999), while two (PtPHT2.1 and PtPHT2.2) were detected in poplar. These two genes are located on different chromosomes (Chr8 and Chr10), and the lengths of their encoded proteins greatly differ which is constituent with the previous report that AtPHT2.1 cDNA encodes a 61-kD protein that is structurally similar but different from the PHT1 family proteins in Arabidopsis (Daram et al., 1999). However, phylogenetic analysis showed that PtPHT2.1 and PtPHT2.2 share 100% similarity (Table S3). At least three AtPHT3 genes have been identified in Arabidopsis to date (Rausch and Bucher, 2002), while we found six PHT3 family genes in poplar. Further, six PHT4 family genes have been characterized in Arabidopsis (Guo et al., 2008), while 8 PtPHT4 genes were detected in the present study. Ten genes with homology to AtPHO1 are present in the Arabidopsis genome (Hamburger et al., 2002), and in this study, 12 PHO family genes were detected with similarities to the PHO1 protein in Arabidopsis.

The presence of a large number of genes within a family reflects the successive expansion and rearrangement of the genome by the extensive duplication and diversification that frequently occur over the course of evolution (Wang et al., 2007). The P. trichocarpa genome is duplicated and contains more protein-coding genes than the Arabidopsis genome; on average, there are 1.4 to 1.6 putative Populus genes for every Arabidopsis gene (Tuskan et al., 2006). Whole-genome duplication increases gene gains and losses, which greatly affects the amplification of members of gene families in the genome (Guo, 2013). The expanded genes in the PHT family, including PHT1, PHT2, PHT3, PHT4, and PHO, in poplar compared with Arabidopsis might be the result of gene duplication. Gene duplication has been reported in P. trichocarpa in many studies, including studies of the glutamine synthetase family, nitrate transporter gene family, and CCCH zinc finger family (Castro-Rodríguez et al., 2011; Chai et al., 2012; Bai et al., 2013). In the present study gene expansion and loss might also occur in the PHT family of P. trichocarpa similar to that of Arabidopsis. For example, two closely related orthologs of Arabidopsis, AtPHT4.1, AtPHT4.5, and AtPHO1, were found in Populus suggesting that duplications might be a partial reason for the expansion of PHT gene family in Populus. Forty-three PtPHT genes were localized to 15 poplar chromosomes, and some PHT genes were found to be clustered together as closely related genes (Figure 2). The presence of these closely related clusters suggests that these genes might evolve from regional duplications (e.g., PtPHO6, PtPHO7, PtPHO8, and PtPHO9).

Cis elements may control the efficiency of promoters and thus regulate the expression of the genes that they control by interacting with the corresponding trans-regulatory factors (Liu et al., 2011). Studies of cis elements might play a crucial role in dissecting the functions of genes. A previous study has described the cis elements of PHT1 family members in poplar (Loth-Pereda et al., 2011), but the cis elements of all PHT gene family members in poplar have not yet been identified. In this study, the upstream regions of all the PtPHTs contain cis elements affecting energy, light, hormone, defense, carbon metabolism, and tissue-specific expression. These findings suggest that the expression of these PtPHT genes might be regulated by many factors, such as light-, hormone-, and defense-related factors. For instance, GATABOX and GT1CONSENSUS are cis elements required for light regulation (Lam and Chua, 1989; Terzaghi and Cashmore, 1995), and they have also been identified in the upstream regions of all OsPHT genes in rice (Liu et al., 2011). Further, putative PHO-like (CGCGTGGG) and TATA-box-like (TATAAATA) cis regulatory elements have been identified in the promoter regions of Arabidopsis genes (Hammond et al., 2003), and they were observed in 22 and 34 of the 42 PtPHT promoters, respectively. Alternatively, because the core sequence of the Arabidopsis PHO-like element (CACGTG) is also similar to that of the ABRE element (ABREOSRAB21), which is present in 6 of the 42 PtPHT promoters, proteins in the bZIP class of transcription factors might bind to them (Schindler et al., 1992). The presence of these cis elements in poplar PtPHT genes and their likely roles in regulating gene expression suggest that PtPHT genes may be involved in various stress responses in poplar.

Some cis elements function under phosphate deficiency, for example, the P1BS element can be bound to PHR1 (a MYB transcription factor) to potentially regulate the phosphate starvation response in Arabidopsis (Rubio et al., 2001), PHR1 has been suggested as a central integrator to role in the transcriptional regulation of phosphate starvation responses (Bustos et al., 2010). In the current study, P1BS-like elements were found in 20 of the 42 PtPHT promoters examined suggesting that expression of these genes might be influenced by Pi starvation. Similarly, P1BS-like elements have also been detected in OsPHT genes in rice (Liu et al., 2011). These elements are found to be in the promoters of phosphate regulated genes, not only in Arabidopsis but also in other plant species such as barley and rice (Schünmann et al., 2004; Liu et al., 2011), which suggests that they may participate in a conserved signaling pathway for the phosphate starvation response in plants. The existence of P1BS-like elements in poplar, rice, and Arabidopsis suggests that they may serve a similar role in both woody species and herbs.

The number of Pi transporters present in roots may reflect the complexity and significance of the process of Pi transport. In the current study, most PHT family members were expressed in the roots of P. tremula and P. simonii. In particular, most PHT1 family members were highly expressed in the roots of poplar, consistent with a previous study conducted by Loth-Pereda et al. (2011) suggesting that they may function in Pi uptake. Notably, most PHT1 genes in other plant species are also expressed in the roots. For example, eight of the nine PHT1 family genes in Arabidopsis are expressed in the roots (Bucher et al., 2001; Mudge et al., 2002), in addition, 8 PHT1 genes in tomato (Chen et al., 2014) and 10 of the 13 PHT1 genes in rice are expressed in the roots (Paszkowski et al., 2002). These results indicate that PHT1 family genes may play similar roles in Pi uptake in different species.

Homologous PHT1, PHT2, PHT3, PHT4, and PHO genes showed substantial differences in their regulation in different tissues. Nine PHT1 family genes were highly expressed in the roots of P. tremula, but the expression of the other five genes was very low, and two of these genes (PtPHT1.9 and PtPHT1.10) were not expressed in any tissues examined. PtPHT2.1 was highly expressed in the roots, cambium-phloem, mature petiole, and dormant buds, while PtPHT2.2 was mainly expressed in the roots, cambium-phloem, and mature leaves. Six homologous PtPHT3 genes exhibited substantial differences in their expression in the various tissues examined. Three of them (PtPHT3.1b, PtPHT3.2b, and PtPHT3.3a) were highly expressed in the roots of P. tremula, while the others (PtPHT3.1a, PtPHT3.2a, and PtPHT3.3b) were expressed at very low levels. Although most PHT4 family members, excluding PtPHT4.3 and PtPHT4.6, were highly expressed in the roots and leaves, PtPHT4.1 was also highly expressed in the cambium-phloem, dormant buds, and dormant flowers. Further, PtPHO5 and PtPHO6 were highly expressed in nine tissues examined, while PtPHO7 and PtPHO8 were exclusively highly expressed in only a few of the tissues. Expression of the PtPHT genes varied among the different tissues examined, suggesting that these genes may have different functions in Pi uptake, transport, and storage. These results suggest that the PHT genes may have different functions since they expressed in a wide range of tissues, which may be partly crucial for the growth of perennial species.

The differential expression of PtPHT genes in different tissues might be related to their subcellular locations (Table 1). For example, two PHT2 family members were expressed in the leaves of P. simonii, in agreement with a previous study demonstrating that the AtPHT2.1 gene, which is located in the chloroplast, is predominantly expressed in green tissues in Arabidopsis (Liu et al., 2011). In addition, the expression patterns of PtPHT4 genes in both roots and leaves, and the phosphate transport activities as well as their subcellular locations, indicate that PHT4 proteins play roles in transferring Pi between the cytosol and chloroplasts, heterotrophic plastids, and the Golgi apparatus (Guo et al., 2008). For example, AtPHT4.6 transports Pi out of the Golgi lumenal space to recycle Pi released from glycosylation processes in Arabidopsis (Cubero et al., 2009). In addition, the PHO1 protein has been reported to be localized to stellar cells of the roots as well as the lower part of the hypocotyl in Arabidopsis (Ribot et al., 2008). In the present study, PHO family members were found to be highly expressed in many of the tissues examined, suggesting that they have diverse functions in loading Pi to different parts of plants.

The transcription of PtPHT genes differed at different Pi levels. Upon Pi starvation, PtPHT1.9, PtPHT1.11, PtPHT1.12, PtPHT3.3b, PtPHT4.3, and PtPHT4.5a expression was up-regulated in P. simonii (Figure 5). The up-regulation of some PtPHTs, such as PtPHT9 and PtPHT11, under Pi starvation is consistent with the findings of a previous study of P. trichocarpa conducted by Loth-Pereda et al. (2011). Most PHT1 family genes expressed in roots show up-regulation in phosphate deprived Arabidopsis plants; in particular, PtPHT9 and PtPHT11 in poplar are highly homologous to AtPHT8 and AtPHT9, which have been shown to be strongly up-regulated during Pi starvation (Mudge et al., 2002). In rice, high-affinity Km values of OsPHT9 and OsPHT10 for Pi transport have been reported, as well as specific induction of their expression by Pi starvation (Wang et al., 2014). Therefore, some of these PHT transporters likely play roles in Pi scavenging under conditions of low soil Pi availability. PtPHT1.2, PtPHT1.3, PtPHT2.2, PtPHT3.1b, and PtPHT4.1a expression was markedly up-regulated under low-Pi conditions, suggesting that their expression might be induced under these conditions. The transcript levels of a few PHT genes in Arabidopsis, such as AtPHO1, AtPHO;H1, and AtPHO;H10, were increased by Pi deficiency (Ribot et al., 2008); however, the expression of most PHO family members was not induced by Pi deficiency in poplar, suggesting that PHT genes might have different functions in woody species and herbs. PtPHT3.3a, PtPHT4.1b, PtPHT4.5b, PtPHO8, and PtPHO9 expression was up-regulated under high-phosphate conditions, indicating that these proteins might be low-affinity transporters. In addition, some PHT4 family members might be low-affinity transporters, consistent with the relatively high Pi levels that have been reported in the cytosol and other subcellular compartments (Mimura, 1999).

The expression of PHT genes varied at different Pi levels, suggesting that PHT family members are likely to encode both high- and low-affinity transporters. For example, OsPHT2 shows low-affinity characteristics on the mM scale in oocytes (Ai et al., 2009). By contrast, a Km value of 19 μM has been reported for HvPHT1 (barley high-affinity transporter, Preuss et al., 2010). AtPHT1.1 has been demonstrated to be a high-affinity transporter in a gene over-expression study (Km value of 3.1 μM, Mitsukawa et al., 1997), in addition to HvPHT1.1 (9.06 μM) and HvPHT1;6 (385 μM) in barley (Rae et al., 2003). The affinities of PHT genes in poplar require further study, including assessment of the heterologous transformation of Xenopus oocytes and gene over-expression studies using plant cell cultures.

The expression levels of some PtPHT genes, such as PtPHT1.2 and PtPHO9, were up-regulated under drought conditions, irrespective of the Pi level, suggesting that their expression was not affected by the Pi level under these conditions. However, the changes in expression of the other PtPHT genes, including PtPHT3.3b, PtPHT4.3, and PtPHT4.5a, under drought stress were largely dependent on the phosphate level (Figure 6). Especially, the upregulation of PtPHT genes induced by a low level of Pi may contribute to drought tolerance of poplar plants. These results suggest that under drought conditions, Pi uptake in plants likely changes in association with alterations in PHT gene expression. This altered uptake in the presence of different levels of phosphate might be related to the drought resistance of plants.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.