Overexpression of the Eucommia ulmoides Aquaporin, EuPIP1;1, Promotes Leaf Growth, Flowering and Bolting, and Stress Tolerance in Arabidopsis

Plasma membrane intrinsic protein (PIP) is one of the largest subfamilies of Aquaporins (AQPs) and plays an important role in plant growth and development, and resistance to abiotic stress. In this study, the full length of the EuPIP1;1 cDNA was cloned from Eucommia ulmoides using the rapid amplification of cDNA ends (RACE) method. The EuPIP1;1 gene was induced by drought treatment and expressed in all tested tissues, with the highest expression level in fruit. The subcellular localization showed that EuPIP1;1 was located in the plasma membrane. Constitutive overexpression of EuPIP1;1 in Arabidopsis thaliana could promote leaf growth and development, and accelerate bolting and flowering. Six genes related to growth and flowering (AtPIF4, AtTCP14, AtCRY1, AtCRY2, AtFCA and AtFT) were significantly up-regulated in transgenic lines. Further, EuPIP1;1 gene improved resistance to drought and salt stress in transgenic Arabidopsis. Under drought and salt stress treatment, the transgenic lines had a higher germination rate and accumulation of osmotic substances, lower membrane damage, and could maintain ion homeostasis. Our results suggest that EuPIP1;1 plays an essential role in plant growth and development and in the response to drought and salt stress.


Introduction
The global shortage of fresh water is one of our most severe ecological and agronomical problems [1]. Water scarcity can lead to land drying and salinization, which reduce plant growth and crop yields [2]. AQPs are thought to be the primary transporters of water, as well as small and uncharged solutes, through plant cell membranes [3]. Thus, AQPs play an important role in maintaining the balance of cellular water in plants [4]. AQPs are usually formed by an assembly of four monomers, each with six transmembrane spanning domains (H1-H6) and five connecting loops (LA-LE) located on the intra-(LB, LD) or extra-cellular (LA, LC, LE) sides of the membrane [5]. The LB and LE loops each carry a highly conserved asparagine-proline-alanine (NPA) motif, folding as half-membranespanning α-helices [6]. Plants can dynamically regulate plant water content and transport of small neutral molecules by regulating the abundance and activity of AQPs, thereby improving tolerance to external stresses [7]. Several studies have shown that AQPs not only play a vital role in regulating plant water balance and mineral nutrient transport [8], but also participate in leaf and petal expansion [9], fruit ripening [10], carbon and nitrogen fixation [11], and signaling [3].
The PIP subfamily is one of the largest subfamilies of AQP proteins [12]. This subfamily can be subdivided into two groups, PIP1s and PIP2s [13], which are different in

Cloning and Characterization of EuPIP1;1
Using the RACE method, a full-length of the aquaporin gene cDNA was isolated from E. ulmoides and designated as EuPIP1;1. The 1130 bp length of the EuPIP1;1 cDNA sequence contained an open reading frame of 861 bp (encoding 286 amino acids), a 75 bp 5 -noncoding region, and a 166 bp 3 -noncoding region (Figure 1a). The EuPIP1;1 and EUC12662 [27] cDNA sequences had a synonymous SNP (C-A) at base 752 of the coding region ( Figure S1). EuPIP1;1 contained two conserved 'NPA' (Asn-Pro-Ala) motifs ( Figure 2, orange box), the intrinsic protein-specific sequence of MIP family (Figure 2, red box), and conserved PIP plant sequences ( Figure 2, blue box). Phylogenetic analysis indicated that EuPIP1;1 was closely related to AQP7 (Camellia japonica), AQU18 (Camellia sinensis) and CsPIP1;3 (Camellia sinensis), and distantly related to rice and maize PIP proteins (Figure 1b). Protein homology-modeling predicted that EuPIP1;1 could form a homologous tetramer, which had high homology with the SoPIP2 sequence model (72.80%) (Figure 1c).

Expression Pattern of EuPIP1;1
Quantitative real-time PCR (qRT-PCR) analysis showed that the EuPIP1;1 transcript was expressed in the root, stem, leaves, and fruit, with the relative expression level being highest in the fruit, followed by the root, stem, and leaves ( Figure 3a). The EuPIP1;1 transcript level was up-regulated after 2 h PEG treatment and the highest expression level appeared after 4 h (Figure 3b). The result suggested that EuPIP1;1 was a drought-inducible gene, indicating that EuPIP1;1 could participate in the response to drought stress.

Expression Pattern of EuPIP1;1
Quantitative real-time PCR (qRT-PCR) analysis showed that the EuPIP1;1 tran was expressed in the root, stem, leaves, and fruit, with the relative expression level highest in the fruit, followed by the root, stem, and leaves ( Figure 3a). The EuPIP1;1 transcript level was up-regulated after 2 h PEG treatment and the highest expression level appeared after 4 h (Figure 3b). The result suggested that EuPIP1;1 was a drought-inducible gene, indicating that EuPIP1;1 could participate in the response to drought stress. ulmoides treated with 20% PEG6000 for different time. (c) Subcellular localization of EuPIP1;1 in epidermal cells of tobacco leaves. Significant differences were determined by one way ANOVA followed by Duncan's multiple range test at p < 0.05.

Subcellular Localization of EuPIP1;1 Protein
To determine the subcellular localization of EuPIP1;1, the 35S-EuPIP1;1-GFP construct was transiently expressed in leaves of Nicotiana benthamiana. The result showed that the EuPIP1;1 protein was located on the plasma membrane, indicating that the protein belongs to the PIP subfamily ( Figure 3c).

Overexpression of EuPIP1;1 Promotes the Growth and Development of Arabidopsis
To evaluate the function of EuPIP1;1, an overexpression vector harboring EuPIP1;1 was transformed into Arabidopsis for heterologous expression. Generation of T3 from two transgenic lines P1 and P2 was selected for further analysis. The leaf size of the transgenic lines was significantly larger than that of the WT (Figure 4a,b). The leaf length of WT was 1.30 cm, and the leaf lengths of P1 and P2 were 1.54 and 1.70 cm, respectively (Figure 4b).
The width of the leaf of the transgenic lines was significantly wider than that of the WT (Figure 4a,b). By its leaf length, it can be seen that the leaf area of P1 and P2 was 1.44-and 1.74-fold that of the WT (Figure 4c). The qRT-PCR assay showed that two important genes related to growth and development, AtPIF4 [29] and AtTCP14 [30], were significantly upregulated in transgenic plants (Figure 4d). The expression levels of AtGIF2 [31] (related to leaf growth) and the genes related to root development, AtKUP4 [32] and AtLBD18 [33], in transgenic plants were not different from those of WT (Figure 4d). In addition, EuPIP1;1 could alter stomatal size and cell area in transgenic Arabidopsis. The stomatal size in P1 and P2 were significantly increased (49% increase in P1 and 37.7% in P2) (( Figure 5a,b), and the cell areas were 1.26-and 1.25-fold that of the WT, respectively (Figure 5c,d). These in epidermal cells of tobacco leaves. Significant differences were determined by one way ANOVA followed by Duncan's multiple range test at p < 0.05.

Subcellular Localization of EuPIP1;1 Protein
To determine the subcellular localization of EuPIP1;1, the 35S-EuPIP1;1-GFP construct was transiently expressed in leaves of Nicotiana benthamiana. The result showed that the EuPIP1;1 protein was located on the plasma membrane, indicating that the protein belongs to the PIP subfamily ( Figure 3c).

Overexpression of EuPIP1;1 Promotes the Growth and Development of Arabidopsis
To evaluate the function of EuPIP1;1, an overexpression vector harboring EuPIP1;1 was transformed into Arabidopsis for heterologous expression. Generation of T 3 from two transgenic lines P1 and P2 was selected for further analysis. The leaf size of the transgenic lines was significantly larger than that of the WT (Figure 4a,b). The leaf length of WT was 1.30 cm, and the leaf lengths of P1 and P2 were 1.54 and 1.70 cm, respectively ( Figure 4b).
The width of the leaf of the transgenic lines was significantly wider than that of the WT (Figure 4a,b). By its leaf length, it can be seen that the leaf area of P1 and P2 was 1.44and 1.74-fold that of the WT (Figure 4c). The qRT-PCR assay showed that two important genes related to growth and development, AtPIF4 [29] and AtTCP14 [30], were significantly up-regulated in transgenic plants (Figure 4d). The expression levels of AtGIF2 [31] (related to leaf growth) and the genes related to root development, AtKUP4 [32] and AtLBD18 [33], in transgenic plants were not different from those of WT (Figure 4d). In addition, EuPIP1;1 could alter stomatal size and cell area in transgenic Arabidopsis. The stomatal size in P1 and P2 were significantly increased (49% increase in P1 and 37.7% in P2) (( Figure 5a,b), and the cell areas were 1.26-and 1.25-fold that of the WT, respectively (Figure 5c,d). These results suggested that overexpression of EuPIP1;1 could promote leaf expansion and could be related to increase in cell size and up-regulation of the expression level of growth-related genes. results suggested that overexpression of EuPIP1;1 could promote leaf expansion and could be related to increase in cell size and up-regulation of the expression level of growth-related genes.  Compared to WT, the transgenic Arabidopsis showed an early-flowering phenotype ( Figure 6a). To test whether the early-flowering phenotype in transgenic lines was related to the up-regulated expression levels of flowering-related genes, the transcriptional levels results suggested that overexpression of EuPIP1;1 could promote leaf expansion and could be related to increase in cell size and up-regulation of the expression level of growth-related genes.  Compared to WT, the transgenic Arabidopsis showed an early-flowering phenotype ( Figure 6a). To test whether the early-flowering phenotype in transgenic lines was related to the up-regulated expression levels of flowering-related genes, the transcriptional levels Compared to WT, the transgenic Arabidopsis showed an early-flowering phenotype (Figure 6a). To test whether the early-flowering phenotype in transgenic lines was related to the up-regulated expression levels of flowering-related genes, the transcriptional levels of four relation genes, AtCRY1, AtCRY2, AtFCA and AtFT, were analyzed [34]. The expression levels of these four genes in transgenic plants were elevated to different degrees. The transcriptional levels of AtCRY1 were 4.31-and 1.37-fold, AtCRY2 were 2.65-and 1.70-fold, AtFCA were 3.98-and 3.26-fold, and AtFT were 5.54-and 3.66-fold that of the WT, in the P1 and P2 lines, respectively ( Figure 6c). In addition, we examined the expression levels of the EuPIP1;1 gene in P1 and P2 lines. The results showed that the expression level of EuPIP1;1 in P1 was 3.17-fold that of P2 ( Figure 6b). We demonstrated that overexpression of EuPIP1;1 could accelerate bolting and flowering might be associated with up-regulated expression levels of flowering-related genes.
of four relation genes, AtCRY1, AtCRY2, AtFCA and AtFT, were analyzed [34]. The expression levels of these four genes in transgenic plants were elevated to different degrees. The transcriptional levels of AtCRY1 were 4.31-and 1.37-fold, AtCRY2 were 2.65and 1.70-fold, AtFCA were 3.98-and 3.26-fold, and AtFT were 5.54-and 3.66-fold that of the WT, in the P1 and P2 lines, respectively (Figure 6c). In addition, we examined the expression levels of the EuPIP1;1 gene in P1 and P2 lines. The results showed that the expression level of EuPIP1;1 in P1 was 3.17-fold that of P2 (Figure 6b). We demonstrated that overexpression of EuPIP1;1 could accelerate bolting and flowering might be associated with up-regulated expression levels of flowering-related genes.

Overexpression of EuPIP1;1 Improved Drought and salt Tolerance in Transgenic Arabidopsis
To investigate the effect of EuPIP1;1 on the tolerance to drought of transgenic Arabidopsis, the WT and transgenic Arabidopsis (P1 and P2) seeds were germinated on filter paper moistened with 0 or 200 mM mannitol for 7 days. On normal conditions, the germination rates of the transgenic plants for P1 and P2 were not significantly different to those of WT. Under mannitol treatment, the WT germination rate decreased to 13%, while the germination rate of P1 and P2 was significantly higher than that of the WT (52% for P1 and 85% for P2) (Figure 7a,b).
The root length of 17-day-old P1 and P2 seedlings was longer than that of the WT under normal and mannitol-treated conditions (Figure 7c,d). In normal, the root lengths of P1 and P2 were 2.12-and 2.11-fold that of the WT, respectively, while under mannitol treatment, the root length of P1 and P2 plants increased to 1.69-and 1.71-fold that of the WT, respectively (Figure 7c,d).

Overexpression of EuPIP1;1 Improved Drought and salt Tolerance in Transgenic Arabidopsis
To investigate the effect of EuPIP1;1 on the tolerance to drought of transgenic Arabidopsis, the WT and transgenic Arabidopsis (P1 and P2) seeds were germinated on filter paper moistened with 0 or 200 mM mannitol for 7 days. On normal conditions, the germination rates of the transgenic plants for P1 and P2 were not significantly different to those of WT. Under mannitol treatment, the WT germination rate decreased to 13%, while the germination rate of P1 and P2 was significantly higher than that of the WT (52% for P1 and 85% for P2) (Figure 7a,b).
The root length of 17-day-old P1 and P2 seedlings was longer than that of the WT under normal and mannitol-treated conditions (Figure 7c,d). In normal, the root lengths of P1 and P2 were 2.12-and 2.11-fold that of the WT, respectively, while under mannitol treatment, the root length of P1 and P2 plants increased to 1.69-and 1.71-fold that of the WT, respectively (Figure 7c,d).
The tolerance to drought of WT and transgenic lines were evaluated in soil culture. Regarding transgenic Arabidopsis overexpressing EuPIP1;1, water was withheld from 40-day-old WT and transgenic lines (P1 and P2) for 30 days and then the plants were watered as normal for 7 days. The survival of transgenic tobacco plants was higher than that of the WT (Figure 7e). To further evaluate the effect of drought stress, plants were treated with 200 mM mannitol for 11 days. The results showed that the leaf growth of both transgenic and WT plants was inhibited, and the leaves of WT plants began to yellow and wilt, while the leaves of transgenic lines remained relatively healthy (Figure 7f). Due to one of the drought resistance indicators of plants, the Arabidopsis water loss rate was determined, and the water loss rate of transgenic lines was found to be lower than that of the WT (Figure 8a). The tolerance to drought of WT and transgenic lines were evaluated in soil culture. Regarding transgenic Arabidopsis overexpressing EuPIP1;1, water was withheld from 40day-old WT and transgenic lines (P1 and P2) for 30 days and then the plants were watered as normal for 7 days. The survival of transgenic tobacco plants was higher than that of the WT (Figure 7e). To further evaluate the effect of drought stress, plants were treated with 200 mM mannitol for 11 days. The results showed that the leaf growth of both transgenic and WT plants was inhibited, and the leaves of WT plants began to yellow and wilt, while the leaves of transgenic lines remained relatively healthy (Figure 7f). Due to one of the drought resistance indicators of plants, the Arabidopsis water loss rate was determined, and the water loss rate of transgenic lines was found to be lower than that of the WT (Figure 8a). To further study the effects of EuPIP1;1 overexpression on the physiology and biochemistry of transgenic Arabidopsis, the contents of malonaldehyde (MDA) and proline (PRO) in leaves of transgenic lines and WT plants were measured under mannitol treatment. The MDA and PRO contents of P1 and P2 were not significantly different, compared with WT, under normal conditions. However, under drought stress, the MDA contents of P1 and P2 were significantly lower than that of WT (Figure 8b), and the PRO content of transgenic lines was significantly higher than that of the WT (Figure 8c). These results indicated that the increasing tolerance to drought of transgenic Arabidopsis could be associated with an increase in osmotic adjustment substances and a reduction in membrane damage in transgenic plants under drought stress compared to WT.
To assess the implication of EuPIP1;1 in the abiotic stress response pathway, we analyzed the expression levels of three genes P5CS1 (which plays an important role in PRO biosynthesis [35]), AtHKT1, and AtSOS1 (which are involved in plant ion transport [36]). The results showed that the transcriptional level of AtP5CS1 in transgenic plants was not different from WT. However, under drought stress, AtP5CS1 was significantly up-regulated in the transgenic plants (Figure 8d). The expression levels of AtHKT1 and AtSOS1 were also up-regulated in transgenic plants by mannitol treatment (Figure 8e,f). These results implied that EuPIP1;1 could promote PRO biosynthesis by up-regulated AtP5CS1, and reduction of Na + by up-regulated AtHKT1 and AtSOS1, ultimately mitigating damage from drought stress. To further study the effects of EuPIP1;1 overexpression on the physiology and biochemistry of transgenic Arabidopsis, the contents of malonaldehyde (MDA) and proline (PRO) in leaves of transgenic lines and WT plants were measured under mannitol treatment. The MDA and PRO contents of P1 and P2 were not significantly different, compared with WT, under normal conditions. However, under drought stress, the MDA contents of P1 and P2 were significantly lower than that of WT (Figure 8b), and the PRO content of transgenic lines was significantly higher than that of the WT (Figure 8c). These results indicated that the increasing tolerance to drought of transgenic Arabidopsis could be associated with an increase in osmotic adjustment substances and a reduction in membrane damage in transgenic plants under drought stress compared to WT.
To assess the implication of EuPIP1;1 in the abiotic stress response pathway, we analyzed the expression levels of three genes P5CS1 (which plays an important role in PRO biosynthesis [35]), AtHKT1, and AtSOS1 (which are involved in plant ion transport [36]). The results showed that the transcriptional level of AtP5CS1 in transgenic plants was not different from WT. However, under drought stress, AtP5CS1 was significantly up-regulated in the transgenic plants (Figure 8d). The expression levels of AtHKT1 and AtSOS1 were also up-regulated in transgenic plants by mannitol treatment (Figure 8e, f). These results implied that EuPIP1;1 could promote PRO biosynthesis by up-regulated AtP5CS1, and reduction of Na + by up-regulated AtHKT1 and AtSOS1, ultimately mitigating damage from drought stress.

Overexpression of EuPIP1;1 Improved Salt Tolerance in Transgenic Arabidopsis
To study the effect of EuPIP1;1 on the salt tolerance of transgenic Arabidopsis, the seeds of WT and transgenic lines (P1 and P2) were germinated on filter paper moistened with 0 or 50 mM NaCl for 7 days. Under salt stress treatment, the germination rate of P1 and P2 (by 84 and 89%, respectively) was significantly higher than that of WT (10%) (Figure 9a,b). The root length of the P1 and P2 seedlings was significantly longer (by 1.28-and 1.36-fold, respectively) than that of the WT under salt stress treatment (Figure 9c,d). To further evaluate the effect of EuPIP1;1 on the salt tolerance of transgenic Arabidopsis, 40-day-old WT, P1, and P1 Arabidopsis seedlings were watered in pots with 100 mM NaCl. After 6 days of treatment, the leaves of the WT plants started to turn yellow, while the leaves of the transgenic plants remained green. After 8 days of treatment, the leaves of the transgenic plants began to turn yellow, while the leaves of the WT plants turned purple and withered (Figure 9e). Similar to drought stress experiments, there were no significant differences in leaf MDA and PRO contents between WT and transgenic plants under non-stress treatment. Under salt stress treatment, the MDA contents of P1 and P2 were significantly lower than that of WT (Figure 9f), and the PRO content of P1 and P2 plants was significantly higher compared with that of the WT (Figure 9g). High levels of AtHKT1 and AtSOS1 expression have been observed in transgenic plants treated with salt stress (Figure 9h,i). These results indicated that EuPIP1;1 might promote root elongation, reduce membrane damage, increase the accumulation of osmotic substance, and up-regulate AtHKT1 and AtSOS1 to maintain ion homeostasis, ultimately improving salt resistance of transgenic Arabidopsis.

Discussion
PIPs play an important role in plant water relations. In our previous work, we proposed that overexpression of EuPIP1;2 could improve tolerance to drought and salt in transgenic tobacco [37]. In this study, we cloned the PIPI subgroup gene EuPIP1;1, which is localized on the plasma membrane. EuPIP1;1 is constitutively expressed in roots, stems, leaves, and fruits in E. ulmoides. Among the four tissues, EuPIP1;1 and EuPIP1;2 [37] both had the highest expression in fruits. This may be associated with the rapid development of the E. ulmoides fruit in August, which requires a lot of water and nutrients. However, unlike the expression of EuPIP1;2 in roots, which was significantly higher than that in stems and leaves, there was no difference in the expression of EuPIP1;1 in roots, stems, and leaves. Therefore, we speculated that EuPIP1;2 was mainly involved in the transport of water in the root [37], while EuPIP1;1 plays an important role in maintaining water transport in various tissues [38].
Gene expression studies in various plant species have shown that PIPs respond to environmental stimuli, and a number of PIPs transcripts were up-regulated under stressful conditions [39]. ScPIP1 [40] and ZxPIP1;3 [14] were up-regulated in roots under drought stress and salt treatment, and ScPIP2-1 in both below-and above-ground tissues were up-regulated under PEG and salt stress [38]. The use of transgenic plants with overexpressing or underexpressing PIP1 also supports the importance of PIP1 for tolerance to environmental stress. Arabidopsis plants expressing ScPIP1 [40] and CrPIP2;3 [41] had longer roots, which may have contributed to improved drought resistance. In this study, the expression level of EuPIP1;1 was up-regulated with PEG treatment. Our results for Arabidopsis support the view that increased AQPs levels may be associated with adaptation to water stress.
Excessive accumulation of intracellular Na + can not only cause ion toxicity and osmotic stress [42], but also interfere with the uptake of other ions and affect plant growth under salt stress [43]. Plants have multiple Na + transport systems to circumvent Na + toxicity and maintain Na + /K + homeostasis and osmotic regulation in cells, and, thus, ensure plant growth under salt stress [42]. Research has demonstrated that the HKT1 transporter and SOS1 are essential for Na + and K + homeostasis in plants [36]. Up-regulation of HKT1 and SOS1 contributes to Na + efflux and K + absorption; thus, maintaining and reconstructing cellular ion homeostasis and alleviating plant damage under drought and salt stress [42]. Studies have shown that overexpression of PIPs can increase HKT1 and SOS1 to improve tolerance to salt stress in transgenic plants [44]. In our study, HKT1 and SOS1 were upregulated in transgenic Arabidopsis, indicating that EuPIP1;1 could promote the expression of the two genes to maintain Na + /K + homeostasis and ensure plant growth under drought and salt stress.
The production of reactive oxygen species (ROS) in plants is a defense response to various stresses, while excessive amounts can cause damage [45]. As an osmotic agent and free radical scavenger, PRO can accumulate to protect cells from damage under abiotic stresses [35]. MDA is commonly used as a marker for ROS-mediated damage [46]. ScPIP1 [40], ZxPIP1;3 [14], MaPIP1;1 [47], EuPIP1;2 [37] etc., can promote the accumulation of PRO and reduce the production of MDA in transgenic plants, thereby keeping transgenic plants in a relatively healthy physiological state and improving tolerance to drought and salt. In our study, under normal conditions, the contents of PRO and MDA were not different between WT and transgenic plants. However, under drought and salt treatment, the PRO contents of P1 and P2 were significantly higher than those of WT, and the MDA content was lower than that of WT. It was suggested that EuPIP1;1 can maintain the stability of the cell membrane by increasing the accumulation of osmotic substances and reducing the content of membrane damage, thus improving the resistance to stress of Arabidopsis.
The higher expression of AQPs in growth, compared to non-growing, tissues is often associated with higher cell hydraulic conductivity (Lp) in growing tissue [48]. PIPs play an important role in regulating rapid transmembrane water flow during plant growth and are believed to be directly involved in cell growth [49]. In this study, the transgenic lines assessed both showed significant increase in stomatal size and cell area. This might be associated with overexpression of EuPIP1;1, which enhanced the water uptake in transgenic Arabidopsis and thereby promoted cell growth. Studies have shown that overexpression of PtoPIP1;1 in Arabidopsis could accelerate cell growth in leaves and roots and promote bolting and flowering by up-regulation of growth-and flower-related genes [49]. Similarly, in our study, the results showed that EuPIP1;1 accelerated the growth of leaves and roots, and promoted bolting and flowering in Arabidopsis. The expression level of the plant growth and development-related genes, AtPIF4 and AtTCP14, and the flowering related genes, AtCRY1, AtCRY2, AtFCA and AtFT, were significantly up-regulated in transgenic plants. These results indicated that overexpression of EuPIP1;1 could promote leaf development by expending cell size in transgenic Arabidopsis and up-regulating the transcriptional level of plant growth-and flowering-related genes, ultimately promoting the growth of the plant and accelerating bolting and flowering.
In conclusion, we found that EuPIP1;1 was induced by drought treatment and played an important role in maintaining water transport in various tissues. Overexpression of the EuPIP1;1 accelerated flowering and bolting, and promoted leaf development by expending cell size. The expression levels of flowering-related genes (AtCRY1, AtCRY2 and AtFCA) and hub genes in the regulatory networks underlying floral timing (AtFT) were significantly up-regulated in transgenic Arabidopsis. Moreover, overexpression of EuPIP1;1 enhanced the drought and salt tolerance in transgenic Arabidopsis by increasing accumulation of osmotic substances, reducing membrane damage and maintaining ion homeostasis. These results showed that EuPIP1;1 has potential use in improving plant stress tolerance.

Plant Materials and Growth Conditions
Ten-year-old E. ulmoides female plants from the Guizhou local variety, cultured in the Transgenic Plant Demonstration Base of Guizhou University, Guiyang (106 • 40 E, 26 • 24 N), were selected for gene cloning and tissue expression analysis [50]. Three-year-old plants, which germinated from the seeds of gene cloning tree, were used for drought treatment. N. benthamiana was used as the recipient for visualization of the subcellular localization of the protein. Arabidopsis 'Columbia-0' was used as wild-type (WT) and all transgenic lines were generated in the background of 'Columbia-0' in this study. All transgenic lines and WT Arabidopsis plants were grown at 23 • C in 16 h light/8 h dark cycles, the light intensities are 10,000 lux.

Full cDNA Cloning and Bioinformatic Analysis of EuPIP1;1
For gene cloning, total RNA was extracted from leaves following the method of Gambino, et al. [51]. Full-length cDNA sequence of EuPIP1;1 was cloned by the RACE method using the SuperScript II RT Kit (Invitrogen, Carlsbad, CA, USA), according to the manufacturer's instructions. The primers used for RACE are listed in Table S2.
For phylogenetic analysis, a multiple alignment was performed by DNAMAN 6.0 software (Lynnon Biosoft, QC, Canada), the phylogenetic tree was constructed using MEGA 11 software (Mega Limited, Auckland, New Zealand), and the maximum likelihood method with 1000 bootstrap replicates was used to determine the phylogeny in the neighborhood joining tree (NJ). The three-dimensional structure of the protein was modeled using the SWISS-MODEL server (https://swissmodel.expasy.org/ (accessed on 22 March 2022)).

Subcellular Localization of Protein
To construct the protein localization vector pCambia-35S-EuPIP1;1::GFP, the EuPIP1;1 full-length coding sequence without stop codon fused with GFP was inserted into the binary vector pCambia1300. Then, the vector was introduced into the Agrobacterium tumefaciens strain EHA105. Transient expression was performed using the leaves of N. benthamiana according to the method of Li [52]. The GFP fluorescence signal was observed with a laser confocal microscope (Leica TCS SP8 STED, Wetzlar, Lahn-Dill-Kreis, Germany) under excitation at 488 nm.

Quantitative Real-Time PCR
For tissue-specific expression, total RNA was extracted from leaves, roots, stems and fruit of 10-year-old E. ulmoides female plants using TRIzol reagent according to the method of Chen, et al. [37]. The RNA quality was detected by the absorbance at A260/A280 and A260/A230 (Table S1). Reverse transcription and quantitative RT-PCR (qRT-PCR) analysis were performed, as described by Zeng and Zhao [53]. Each test was repeated three times.
Three-year-old E. ulmoides plants were treated with 20% PEG6000 [40], and then the total RNA was extracted at 0, 1, 2, 6, 24, and 48 h from leaves using E.Z.N.A.™ Plant RNA Kit (Omega, Norcross, GA, USA). cDNA was reversed with the StarScript II Firststand cDNA Synthesis Mix with gDNA Remover Kit (GenStar, Beijing, China). Using the 2 × RealStar Green Fast Mixture (GenStar, Beijing, China) for qRT-PCR according to the manufacturers' protocols (two-step approach). All the data were calculated and analyzed from three independent samples. The EuActin gene was used as the internal control gene [54]. The 2 −∆∆Ct method was used to determine the relative expression level [55]. All primers for qRT-PCR were listed in Table S2.

Overexpression Vector Construction and Plant Transformation
The full-length coding sequence of EuPIP1;1 was amplified by chemical synthesis and inserted into the pCambia1301 vector ( Figure 10). The vector was introduced into the Agrobacterium tumefaciens strain EHA105, and then transformed into Arabidopsis via the floral dipping method. The seeds were then selected with hygromycin (50 mg/L), and the generation of T 3 was used for subsequent experiments.
into the binary vector pCambia1300. Then, the vector was introduced into the Agrobacterium tumefaciens strain EHA105. Transient expression was performed using the leaves of N. benthamiana according to the method of Li [52]. The GFP fluorescence signal was observed with a laser confocal microscope (Leica TCS SP8 STED, Wetzlar, Lahn-Dill-Kreis, Germany) under excitation at 488 nm.

Quantitative Real-Time PCR
For tissue-specific expression, total RNA was extracted from leaves, roots, stems and fruit of 10-year-old E. ulmoides female plants using TRIzol reagent according to the method of Chen, et al. [37]. The RNA quality was detected by the absorbance at A260/A280 and A260/A230 (Table S1). Reverse transcription and quantitative RT-PCR (qRT-PCR) analysis were performed, as described by Zeng and Zhao [53]. Each test was repeated three times.
Three-year-old E. ulmoides plants were treated with 20% PEG6000 [40], and then the total RNA was extracted at 0, 1, 2, 6, 24, and 48 h from leaves using E.Z.N.A.™ Plant RNA Kit (Omega, Norcross, Georgia, USA). cDNA was reversed with the StarScript II Firststand cDNA Synthesis Mix with gDNA Remover Kit (GenStar, Beijing, China). Using the 2 × RealStar Green Fast Mixture (GenStar, Beijing, China) for qRT-PCR according to the manufacturers' protocols (two-step approach). All the data were calculated and analyzed from three independent samples. The EuActin gene was used as the internal control gene [54]. The 2 −∆∆Ct method was used to determine the relative expression level [55]. All primers for qRT-PCR were listed in Table S2.

Overexpression Vector Construction and Plant Transformation
The full-length coding sequence of EuPIP1;1 was amplified by chemical synthesis and inserted into the pCambia1301 vector ( Figure 10). The vector was introduced into the Agrobacterium tumefaciens strain EHA105, and then transformed into Arabidopsis via the floral dipping method. The seeds were then selected with hygromycin (50 mg/L), and the generation of T3 was used for subsequent experiments.

Drought and Salt Tolerance Analysis of Transgenic Plants
For the analysis of germination, Arabidopsis seeds of two transgenic lines P1, P2, and WT were seeded on filter paper moistened with 0, or 200, mM mannitol and incubated for 7 days (23 °C, 16 h light/8 h dark). For root length assays, seeds were treated with 75% ethanol for 30 s, then sterilized with 10% NaClO solution for 5 min, and rinsed with sterile water 5 times. The seeds were cultured on half-strength Murashige and Skoog (1/2 MS) medium supplemented with 0, 200 mM mannitol or 50 mM NaCl for 17 days (23 °C, 16 h light /8 h dark), and then the lengths of the roots were measured.
The 40-day-old seedlings were treated without watering for 30 days, and then the plants were re-watered as normal for 7 d. The 50-day-old seedlings from WT and transgenic lines were used for water loss analysis, following the method described by Aharoni, et al. [56]. The percentage of water loss was calculated with the formula: water loss rate (%) = (FW -DW)/FW × 100. Leaves from 60-day-old Arabidopsis were treated with clear nail polish, as described previously [57]. Then, stomatal size (μm 2 ) was calculated using ImageJ 1.49 (National Institute of Health (NIH), Bethesda, USA) as the guard cell length × guard cell pair width from images collected at ×40 magnification [58]. Leaves from 60-day-old Arabidopsis were fixed in formalin-acetic acid-alcohol (FAA) and then cleared in chloral hydrate solution [49], and images were collected at ×40 magnification.

Drought and Salt Tolerance Analysis of Transgenic Plants
For the analysis of germination, Arabidopsis seeds of two transgenic lines P1, P2, and WT were seeded on filter paper moistened with 0, or 200, mM mannitol and incubated for 7 days (23 • C, 16 h light/8 h dark). For root length assays, seeds were treated with 75% ethanol for 30 s, then sterilized with 10% NaClO solution for 5 min, and rinsed with sterile water 5 times. The seeds were cultured on half-strength Murashige and Skoog (1/2 MS) medium supplemented with 0, 200 mM mannitol or 50 mM NaCl for 17 days (23 • C, 16 h light/8 h dark), and then the lengths of the roots were measured.
The 40-day-old seedlings were treated without watering for 30 days, and then the plants were re-watered as normal for 7 d. The 50-day-old seedlings from WT and transgenic lines were used for water loss analysis, following the method described by Aharoni, et al. [56]. The percentage of water loss was calculated with the formula: water loss rate (%) = (FW − DW)/FW × 100. Leaves from 60-day-old Arabidopsis were treated with clear nail polish, as described previously [57]. Then, stomatal size (µm 2 ) was calculated using ImageJ 1.49 (National Institute of Health (NIH), Bethesda, USA) as the guard cell length × guard cell pair width from images collected at ×40 magnification [58]. Leaves from 60-day-old Arabidopsis were fixed in formalin-acetic acid-alcohol (FAA) and then cleared in chloral hydrate solution [49], and images were collected at ×40 magnification.

MDA and PRO Detection
For MDA and PRO detection, the 40-day-old Arabidopsis seedlings were watered with or without 200 mM mannitol or 100 mM NaCl for 8 d every 2 days. The leaves were then collected to determine the contents of MDA and PRO. The PRO and MDA content were determined using the Ninhydrin Reaction PRO Detection Kits (Keming, Suzhou, China) and