Gene-Wide Analysis of Aquaporin Gene Family in Malus domestica and Heterologous Expression of the Gene MpPIP2;1 Confers Drought and Salinity Tolerance in Arabidposis thaliana

The aquaporins (AQPs) are a family of integral membrane proteins involved in the transcellular membrane transport of water and other small molecules. A scan of the apple (Malus domestica) genome revealed the presence of 42 genes encoding putative AQPs. Based on a phylogenetic analysis of the deduced peptide sequences of the AQPs generated by Arabidopsis thaliana, poplar (Populus trichocarpa), and rubber (Hevea brasiliensis), the apple AQPs were each assigned membership of the five established AQP subfamilies, namely the PIPs (eleven members), the TIPs (thirteen members), the NIPs (eleven members), the SIPs (five members), and the XIPs (two members). The apple AQPs included asparagine-proline-alanine (NPA) motifs, an aromatic/arginine (ar/R) selectivity filter, and the Froger’s positions. The heterologous expression of MpPIP2;1 in A. thaliana was shown to enhance the level of tolerance exhibited against both drought and salinity.


Introduction
The aquaporins (AQPs) represent a family of integral membrane proteins, and form channels which allow the transport of water and other small molecules across membranes [1]. These proteins are produced by species across the phylogenetic spectrum, from microbes to plants and animals [2]. A typical aquaporin feature six transmembrane (TM1-TM6) helices (H1-H6) and five connecting loops (LA-LE); both their carboxylic and amino terminals lie on the cytoplasmic side, while two half helices formed the seventh TM helix by the opposite LB and LE dipping into the membrane. Given their general conservation across many AQPs, the asparagines-proline-alanine (NPA) motifs, the aromatic/Arginine (ar/R) selectivity filter formed by four residues (F58-H182-C191-R197 in AQP1) [3], and Froger's positions (P1-P5 residues, T116-S196-A200-F212-W213 in AQP1) [4], are considered to be important for function. AQPs are tetrameric proteins and each monomer is functional independently as a channel. Furthermore, the fifth channel, which forms through the middle of tetramer array, has been suggested to conduct gases, like CO 2 [5].
The survival and growth of a plant depends on its ability to maintain a sufficient level of tissue hydration. Proteins referred to as AQPs are known to represent an important component of the regulatory machinery used by plants for this purpose. AQPs have been shown to exert control over germination, since pea seeds imbibed in the presence of the AQP inhibitor mercury do not

Phylogenetic Analysis
The phylogenetic relationships between the set of MdAQPs with homologous proteins encoded by A. thaliana, poplar, and rubber (Hevea brasiliensis) is displayed in Figure 1. The analysis allowed the set of apple AQPs to be each assigned membership of one of the five plant AQP subfamilies, namely the MdPIPs (eleven members), the MdTIPs (thirteen members), the MdNIPs (eleven members), the MdSIPs (five members), and the MdXIPs (two members). The MdPIP members were further classified into the two subgroups, MdPIP1 and MdPIP2, the MdSIPs into the two subgroups, MdSIP1 and MdSIP2, and the MdTIPs into the five subgroups, MdTIP1-MdTIP5. The two MdXIPs belonged into the two subgroups MdXIP1 and MdXIP2, respectively, and the MdNIPs were divided into six subgroups, MdNIP1, 2, 4, 5, 6 and 7. On the basis of sharing a level of >90% similarity at the peptide level, 14

Chromosomal Location and Gene Structure
It was possible to map 40 of the 42 MdAQPs on 16 of the 17 apple chromosomes, but neither MdNIP1;2 nor MdTIP3;2 could be placed ( Figure 2). The sequence of each of the eleven MdPIPs featured three introns; all but one of the thirteen MdTIPs featured two introns (the exception was MdTIP1;1 in which only one intron was present); eight of the eleven MdNIP sequences were interrupted by four introns, with three introns present in both MdNIP5;1 and MdNIP5;2, and five in MdNIP5;3; three of the five MdSIPs harbored two introns, while neither MdSIP1;2 nor MdSIP1;3 featured any introns; finally, MdXIP1;1 had one intron while MdXIP2;1 included two introns ( Figure 3). The protein sequences were aligned by ClustalX and the phylogenetic tree was constructed by the Neighbor-Joining method (1000 bootstrap replicates) in the MEGA6 software. The subgroups are marked by a colorful background (orange for TIPs, purple for PIPs, red for NIPs, blue for XIPs and gray for SIPs).

Chromosomal Location and Gene Structure
It was possible to map 40 of the 42 MdAQPs on 16 of the 17 apple chromosomes, but neither MdNIP1;2 nor MdTIP3;2 could be placed ( Figure 2). The sequence of each of the eleven MdPIPs featured three introns; all but one of the thirteen MdTIPs featured two introns (the exception was MdTIP1;1 in which only one intron was present); eight of the eleven MdNIP sequences were interrupted by four introns, with three introns present in both MdNIP5;1 and MdNIP5;2, and five in MdNIP5;3; three of the five MdSIPs harbored two introns, while neither MdSIP1;2 nor MdSIP1;3 featured any introns; finally, MdXIP1;1 had one intron while MdXIP2;1 included two introns ( Figure 3).

Conserved Residues in the Apple AQPs
The NPA motifs, ar/R filter, and Froger's positions were identified via a multiple alignment between the apple AQPs and other plant AQPs (Table 2). These conserved positions were critical for the substrate selectivity of AQPs. Both NPA domains were conserved in all MdPIP and MdTIP members, but the third residue of the first NPA in MdNIP5;1, MdNIP5;2, and MdNIP5;3 was serine rather than alanine, while in MdNIP5;1, MdNIP5;2, and MdNIP6;1, the third residue of the second NPA was valine rather than alanine; in MdNIP2;3 it was glutamate and in MdNIP5;3 it was isoleucine. The MdSIPs all carried a non-conserved third residue in the first NPA, while in addition, the first residue of the second NPA in MdSIP2;2 was serine rather than asparagine. Both the first and third residues of the first NPA of MdXIP1;1 were non-conserved and the third residues of the first NPA of MdXIP2;1 was valine. The ar/R filter sequence was well conserved within each subfamily, but varied between the subfamilies. Each of the PIPs carried the conserved sequence phenylalanine-histidine-threonine-arginine. The greatest diversity for this motif was present among

Conserved Residues in the Apple AQPs
The NPA motifs, ar/R filter, and Froger's positions were identified via a multiple alignment between the apple AQPs and other plant AQPs (Table 2). These conserved positions were critical for the substrate selectivity of AQPs. Both NPA domains were conserved in all MdPIP and MdTIP members, but the third residue of the first NPA in MdNIP5;1, MdNIP5;2, and MdNIP5;3 was serine rather than alanine, while in MdNIP5;1, MdNIP5;2, and MdNIP6;1, the third residue of the second NPA was valine rather than alanine; in MdNIP2;3 it was glutamate and in MdNIP5;3 it was isoleucine. The MdSIPs all carried a non-conserved third residue in the first NPA, while in addition, the first residue of the second NPA in MdSIP2;2 was serine rather than asparagine. Both the first and third residues of the first NPA of MdXIP1;1 were non-conserved and the third residues of the first NPA of MdXIP2;1 was valine. The ar/R filter sequence was well conserved within each subfamily, but varied between the subfamilies. Each of the PIPs carried the conserved sequence phenylalanine-histidine-threonine-arginine. The greatest diversity for this motif was present among the NIPs, where each of the tetrapeptides tryptophan-valine-alanine-arginine, glycine-serine-glycine-arginine, alanine-isoleucine-glycine-arginine, threonine-isoleucine-alanine-arginine, and alanine-valine-glycine-arginine was represented. With respect to the Froger's positions, there was also high conservation within each subfamily, but variability between subfamilies. While the P1 position was the only variable residue within the PIP and XIP subfamily members, both the P1 and P2 positions varied among members of TIP and SIP subfamilies, and the P1, P2, and P5 positions were were all non-conserved for NIP subfamily members.  When the abundance of transcripts generated from seven of the PIP2 genes was evaluated by applying a quantitative real-time PCR (qRT-PCR) assay to RNA extracted from M. hupehensis root tissue, PIP2;1 appeared to be the gene most strongly transcribed ( Figure 5). Since both drought and salinity stress are sensed by roots, PIP2;1 was chosen for a detailed functional analysis. In particular, the copy present in Malus prunifolia was selected, as this species provides a source of drought-tolerant rootstocks [22]. transmembrane helix 2; H5, transmembrane helix 5.

The Site of Apple PIP2 Expression
A search of the set of apple ESTs deposited in GenBank resulted in 685 hits for seven MdPIP2s, a subgroup of the M. domestica AQP gene family, the products of which are likely important regulators of water transport across the plasma membrane. The distribution of these hits among each PIP2 gene and the organ are shown in Figure 4. The most well represented gene was MdPIP2;4 (218 hits) and the least well represented was MdPIP2;5 (seven hits). The site of transcription of these genes can be inferred from the frequency of their transcripts' occurrence in the 76 cDNA libraries assembled from various organs. In the bud libraries, there were 116 such ESTs out of a total of 54,099 sequences; there were 104 out of 54,120 in the leaf libraries; there were 73 out of 35,380 in the stem libraries; there was 94 out of 12,679 in the root libraries; there were 117 out of 44,772 in the flower libraries; there were 176 out of 104,341 in the fruit libraries; and finally out of libraries constructed from in vitro cultured cells, there were 5 out of 5652. When the abundance of transcripts generated from seven of the PIP2 genes was evaluated by applying a quantitative real-time PCR (qRT-PCR) assay to RNA extracted from M. hupehensis root tissue, PIP2;1 appeared to be the gene most strongly transcribed ( Figure 5). Since both drought and salinity stress are sensed by roots, PIP2;1 was chosen for a detailed functional analysis. In particular, the copy present in Malus prunifolia was selected, as this species provides a source of drought-tolerant rootstocks [22].

The Abiotic Stress Tolerance of A. thaliana Plants Heterologously Expressing MpPIP2;1
A set of ten independent A. thaliana transgenics heterologously expressing MpPIP2;1 was obtained, and three of these were randomly selected to advance to the T3 generation ( Figure 6). The performance of the three transgenic lines was then compared with that of wild type (WT) Col-0 plants. The contrasting effect of drought stress on the transgenic and WT plants is illustrated in Figure  7. Under well-watered conditions, the growth of the transgenic lines was indistinguishable from that of WT plants. However, when water was withheld for 30 days, none of the WT plants remained viable as they were unable to maintain a sufficient level of leaf hydration (their relative water content fell to 10.5%); in contrast, many of transgenic plants survived, maintaining a leaf relative water content of about 20%. The post-stress survival rate of these latter plants was between 10.3% and 23.5%. A comparison of the leaf malondialdehyde (MDA) content showed that less of this stress marker accumulated in the transgenic plants than in WT plants; similarly, it was established that the relative electrolyte leakage of leaves sampled from the transgenic plants was lower than that of WT leaves. The activity of each of the enzymes superoxide dismutase (SOD), catalase (CAT), and peroxidase (POD), as well as the content of glutathione (GSH), were all greater in the transgenic plants than in the WT ones. Consistent with this result, the detached leaves of transgenic lines lost water slower than that of WT leaves. After 5 h of dehydration, the water loss rate for transgenic lines OE1, OE2, and OE3 were 17.6%, 8.8%, and 4.13% lower than that of WT, respectively.

The Abiotic Stress Tolerance of A. thaliana Plants Heterologously Expressing MpPIP2;1
A set of ten independent A. thaliana transgenics heterologously expressing MpPIP2;1 was obtained, and three of these were randomly selected to advance to the T3 generation ( Figure 6). The performance of the three transgenic lines was then compared with that of wild type (WT) Col-0 plants.

The Abiotic Stress Tolerance of A. thaliana Plants Heterologously Expressing MpPIP2;1
A set of ten independent A. thaliana transgenics heterologously expressing MpPIP2;1 was obtained, and three of these were randomly selected to advance to the T3 generation ( Figure 6). The performance of the three transgenic lines was then compared with that of wild type (WT) Col-0 plants. The contrasting effect of drought stress on the transgenic and WT plants is illustrated in Figure  7. Under well-watered conditions, the growth of the transgenic lines was indistinguishable from that of WT plants. However, when water was withheld for 30 days, none of the WT plants remained viable as they were unable to maintain a sufficient level of leaf hydration (their relative water content fell to 10.5%); in contrast, many of transgenic plants survived, maintaining a leaf relative water content of about 20%. The post-stress survival rate of these latter plants was between 10.3% and 23.5%. A comparison of the leaf malondialdehyde (MDA) content showed that less of this stress marker accumulated in the transgenic plants than in WT plants; similarly, it was established that the relative electrolyte leakage of leaves sampled from the transgenic plants was lower than that of WT leaves. The activity of each of the enzymes superoxide dismutase (SOD), catalase (CAT), and peroxidase (POD), as well as the content of glutathione (GSH), were all greater in the transgenic plants than in the WT ones. Consistent with this result, the detached leaves of transgenic lines lost water slower than that of WT leaves. After 5 h of dehydration, the water loss rate for transgenic lines OE1, OE2, and OE3 were 17.6%, 8.8%, and 4.13% lower than that of WT, respectively. The contrasting effect of drought stress on the transgenic and WT plants is illustrated in Figure 7. Under well-watered conditions, the growth of the transgenic lines was indistinguishable from that of WT plants. However, when water was withheld for 30 days, none of the WT plants remained viable as they were unable to maintain a sufficient level of leaf hydration (their relative water content fell to 10.5%); in contrast, many of transgenic plants survived, maintaining a leaf relative water content of about 20%. The post-stress survival rate of these latter plants was between 10.3% and 23.5%. A comparison of the leaf malondialdehyde (MDA) content showed that less of this stress marker accumulated in the transgenic plants than in WT plants; similarly, it was established that the relative electrolyte leakage of leaves sampled from the transgenic plants was lower than that of WT leaves. The activity of each of the enzymes superoxide dismutase (SOD), catalase (CAT), and peroxidase (POD), as well as the content of glutathione (GSH), were all greater in the transgenic plants than in the WT ones. Consistent with this result, the detached leaves of transgenic lines lost water slower than that of WT leaves. After 5 h of dehydration, the water loss rate for transgenic lines OE1, OE2, and OE3 were 17.6%, 8.8%, and 4.13% lower than that of WT, respectively.  A series of experiments were conducted to establish whether the constitutive expression of MpPIP2;1 in A. thaliana had any effect on the level of tolerance to salinity stress, as imposed by exposure to 0.3 M NaCl for 14 days (Figure 8). The transgenic plants maintained a superior leaf hydration status compared to the WT plants: Their respective relative water contents were >80% and 49%. While the survival rate of WT plants was 46.4%, that of the transgenic plants was >90%. Compared to WT leaves, those of the transgenic plants accumulated less MDA, developed a lower relative electrolyte leakage, exhibited a higher activity of SOD, POD, and CAT, and their GSH content was greater. A series of experiments were conducted to establish whether the constitutive expression of MpPIP2;1 in A. thaliana had any effect on the level of tolerance to salinity stress, as imposed by exposure to 0.3 M NaCl for 14 days (Figure 8). The transgenic plants maintained a superior leaf hydration status compared to the WT plants: Their respective relative water contents were >80% and 49%. While the survival rate of WT plants was 46.4%, that of the transgenic plants was >90%. Compared to WT leaves, those of the transgenic plants accumulated less MDA, developed a lower relative electrolyte leakage, exhibited a higher activity of SOD, POD, and CAT, and their GSH content was greater.

Germination and Root Elongation of MpPIP2;1 Transgenics Exposed to Either Salinity or Osmotic Stress
An experiment was conducted to establish whether the constitutive expression of MpPIP2;1 in A. thaliana had any effect on germination and/or root elongation in the presence of either salinity or osmotic stress (Figure 9). When the seed was imbibed in the absence of a stress agent (mannitol or NaCl), the rate of germination of both the WT and transgenic seed was high, and there were no significant differences between the germination rates of WT and transgenic seeds. However, in the presence of either 0.25 M mannitol or 0.15 M NaCl, the rate of germination of the WT seeds fell to just 10%, while that of the transgenic seeds remained >60%. Similarly, the ability of roots to elongate

Germination and Root Elongation of MpPIP2;1 Transgenics Exposed to Either Salinity or Osmotic Stress
An experiment was conducted to establish whether the constitutive expression of MpPIP2;1 in A. thaliana had any effect on germination and/or root elongation in the presence of either salinity or osmotic stress (Figure 9). When the seed was imbibed in the absence of a stress agent (mannitol or NaCl), the rate of germination of both the WT and transgenic seed was high, and there were no significant differences between the germination rates of WT and transgenic seeds. However, in the presence of either 0.25 M mannitol or 0.15 M NaCl, the rate of germination of the WT seeds fell to just 10%, while that of the transgenic seeds remained >60%. Similarly, the ability of roots to elongate was the same for the WT and transgenic seedlings under non-stressful conditions, but the extent of its inhibition by the presence of either mannitol or NaCl differed between the transgenic and WT seedlings. was the same for the WT and transgenic seedlings under non-stressful conditions, but the extent of its inhibition by the presence of either mannitol or NaCl differed between the transgenic and WT seedlings. (c) germination rate of transgenic and wild type seeds on different mediums for 7 days. The value presented was the mean ± SD of three replicates, and the bar with different letter was significantly different between plants at p < 0.05.

Discussion
The systematic scanning of the content of AQP-encoding genes in the apple genome reported here resulted in the identification of 42 such genes. AQPs make an important contribution to the way in which plants control their uptake of water, and hence represent a key component of their response to drought and osmotic stress [23]. Thus, gaining a full understanding of how apple plants regulate their water balance and adapt to drought and osmotic stress will likely involve revealing the function of many of this set of genes. There is already some experimental evidence which supports the participation of AQPs in the stress response of apple. According to Hu et al. (2003), the transcription of both MdPIP1a and MdPIP1b (here renamed as, respectively, MdPIP1;2 and MdPIP1;1) are up-regulated by osmotic stress [24]. The M. zumi homolog of MdPIP1;1 has been shown to be inducible by salinity (as well as by low temperature) stress [25]. Meanwhile, the constitutive expression of MzPIP2;1 (homolog of MdPIP2;4) in A. thaliana has a positive effect on drought tolerance and a small positive one on salinity tolerance [26], and the expression of MzPIP1;3 (homolog of MdPIP1;3) in tomato has been shown to enhance the plants' drought tolerance [27]. (c) germination rate of transgenic and wild type seeds on different mediums for 7 days. The value presented was the mean ± SD of three replicates, and the bar with different letter was significantly different between plants at p < 0.05.

Discussion
The systematic scanning of the content of AQP-encoding genes in the apple genome reported here resulted in the identification of 42 such genes. AQPs make an important contribution to the way in which plants control their uptake of water, and hence represent a key component of their response to drought and osmotic stress [23]. Thus, gaining a full understanding of how apple plants regulate their water balance and adapt to drought and osmotic stress will likely involve revealing the function of many of this set of genes. There is already some experimental evidence which supports the participation of AQPs in the stress response of apple. According to Hu et al. (2003), the transcription of both MdPIP1a and MdPIP1b (here renamed as, respectively, MdPIP1;2 and MdPIP1;1) are up-regulated by osmotic stress [24]. The M. zumi homolog of MdPIP1;1 has been shown to be inducible by salinity (as well as by low temperature) stress [25]. Meanwhile, the constitutive expression of MzPIP2;1 (homolog of MdPIP2;4) in A. thaliana has a positive effect on drought tolerance and a small positive one on salinity tolerance [26], and the expression of MzPIP1;3 (homolog of MdPIP1;3) in tomato has been shown to enhance the plants' drought tolerance [27].
Both moisture and nutrient stress are initially sensed by a plant's roots. Based on its relatively high transcript abundance (inferred both indirectly from the frequency of its representation in EST libraries and directly through a qRT-PCR analysis), PIP2;1 is the PIP2 gene most strongly transcribed in the roots of the apple plant. This same gene has been shown, using a suppression subtractive hybridization method, to be up-regulated in response to moisture deficit [28], while two MdPIP2;1 ESTs have been identified in cDNA libraries developed from the roots of plants from which water had been withheld for a week (LIBEST_024527). Thus the evidence points to the conclusion that the product of PIP2;1 is an important component of the apple plant's response to moisture stress. This evidence has been strengthened by the demonstration here that heterologously expressing the gene in A. thaliana had a positive effect on the plant's tolerance of both drought and salinity.
A commonly observed plant response to abiotic stress is to accumulate reactive oxygen species (ROS), which become cytotoxic when present in excess [29]. When A. thaliana plants heterologously expressing MpPIP2;1 were exposed to either drought or salinity stress, both the MDA content and the relative electrolyte leakage of their leaves were below the levels shown by WT leaves; both of these traits are correlated with ROS-mediated cellular damage [30]. The maintenance of a non-damaging level of cellular ROS content is achieved both by the activity of a number of enzymes and the synthesis of antioxidant compounds [31]. Both the activity of the enzymes SOD, CAT, and POD and the content of the antioxidant compound GSH were higher in the transgenic than in the WT A. thaliana plants subjected to stress. The conclusion is that the product of MpPIP2;1 likely contributes to protecting the transgenic plants experiencing drought stress by enhancing their ability to control the accumulation of ROS. Similar conclusions have been drawn with respect to AQPs in a number of plant systems [32], so it is reasonable to propose that the product of MpPIP2;1, a gene which is strongly transcribed in the root of M. prunifolia, is an important determinant of the drought stress response when expressed in its native context.

Identification of the Set of AQP Genes in the Apple Genome
Apple AQP sequences were recovered from the NCBI Protein database (www.ncbi.nlm.nih.gov/ protein) by entering as a keyword search "(aquaporin OR MIP) AND Malus". The resulting hits were confirmed as genuine AQPs by submitting them to the NCBI Conserved Domain database (www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi). The sequences of identified apple AQPs were used as queries to search Malus x domestica Whole Genome v1.0 and GDDH13 Whole Genome v1.1 sequence for additional members with an E value less than 0.01. A phylogenetic analysis based on the deduced peptide sequences of AQPs encoded by A. thaliana, poplar, and rubber [33] was used assign the apple sequences to the five established AQP subfamilies. Multiple sequence alignments were carried out using ClustalX software [34], and an unrooted phylogenetic tree was constructed using MEGA6 software [35], applying the Neighbor-Joining method and 1000 bootstrap replicates.

Chromosomal Location, Gene Structure, and Protein Properties of Apple AQPs
The GDDH13 assembly was used to reveal the chromosomal location for each of the MdAQPs and to determine the intron/exon structure of each gene. The latter was visualized using GSDS software (bio.tools/GSDS) [36]. The pI and molecular weight of the deduced AQPs were predicted using the ExPASY program (web.expasy.org/compute_pi/). Transmembrane regions were detected using TMHMM software (www.cbs.dtu.dk/services/TMHMM/) [37], and subcellular localizations were predicted using WoLF PSOR software (wolfpsort.hgc.jp/) [38]. Sequences representing conserved domains, NPA motifs, the ar/R filter, and the Froger positions were manually identified, based on multiple sequence alignments of apple AQPs with heterologous AQPs.

The Sites of Apple PIP2 Expression
Apple ESTs were retrieved through a BLASTN search of the GeneBank database, using as search terms each of the MdPIP2 transcripts in turn. RNA was extracted from the roots of hydroponically-raised M. hupehensis seedlings (a triploid species characterized by facultative apomixis) which had formed 7-8 true leaves [39], and was processed for a series of qRT-PCR assays targeting seven PIP2 genes. The relevant primers were designed using Beacon Designer 8 are shown in Table 3.

Heterologous Expression of MpPIP2;1 in A. thaliana
A full length copy of MpPIP2;1 cDNA (JF834203.1) was PCR-amplified from an in-house pMD19-T-MpPIP2;1 plasmid using a primer pair listed in Table 3 [41]. This was used to generate the construct pCAMBIA2300-35S-MpPIP2;1, which was introduced into A. tumefaciens strain GV3101, and from thence into A. thaliana (ecotype Col-0) using the floral dip method [42]. Selection for transgenic products was carried out by culturing on a medium containing 50 mg·L −1 kanamycin. The abundance of MpPIP2;1 transcript produced in transgene homozygous T3 lines was evaluated using a qRT-PCR assay based on the primer pairs listed in Table 3.

Stress Tolerance Analysis
For the purpose of assaying in vitro germination, surface-sterilized WT and transgenic seeds were laid on either solidified Murashige and Skoog (1962) medium (MS) [43], MS containing 0.25 M mannitol, or MS containing 0.15 M NaCl, and held for seven days under a 16 h photoperiod at 23 • C. Root elongation was assessed by culturing pre-germinated seedlings for two weeks on vertically oriented plants containing MS, MS + 0.25 M mannitol, or MS + 0.15 M NaCl. To test for both drought and salinity tolerances, seedlings were grown in pots containing equal amounts of soil after stratified and generated in soil for 3 weeks. The water loss rate was measured by weighing the detached leaves at different time points. The plants were then either subjected to drought by the withholding of water for 30 days, or salinity stress by irrigation with 0.3 M NaCl solution [44,45]. The assays for relative leaf water content, leaf electrolyte leakage, and MDA concentration have been described elsewhere. To assay for SOD, CAT, and POD activity and GSH content, leaves was homogenized in phosphate buffer (pH 7.5) and the resulting supernatants recovered after centrifugation were tested using commercially available kits purchased from Nanjing Jiancheng Bioengineering Institute (Nanjing, China). The SOD Assay Kit was based on hydroxylamine method, the CAT Assay Kit was based on the method of ammonium molybdate, the POD Assay Kit was based on the hydrogen peroxide oxidation reaction, and the GSH assay kit is based on reduced glutathione reacting with 5,5 -dithiobis-2-nitrobenoic acid.

Statistical Analysis
Data were statistically analyzed using routines implemented in SPSS v17.0. Means were compared using Tukey's test, applying a 0.05 probability threshold to declare significance.

Conclusions
The present experiment has identified that the apple genome harbors 42 genes encoding putative AQPs. It is necessary for the future to comprehensively understand the process of water and small solutes flues through the membrane and related molecular-regulating mechanisms in the apple tree. In A. thaliana, plants were engineered to heterologously express one of these genes (MpPIP2;1), a gene which is strongly transcribed in the root of M. prunifolia. The transgenic plants exhibited an improved level of tolerance to both drought and salinity stress, which implies that in the apple, the product of PIP2;1 is an important component of the plant's response to moisture stress. This latter finding suggests a strategy for enhancing the stress tolerance of the apple using a molecular approach.

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