Genome-Wide Identiﬁcation and Expression Analyses of the Aquaporin Gene Family in Passion Fruit ( Passiﬂora edulis ), Revealing PeTIP3-2 to Be Involved in Drought Stress

: Aquaporins (AQPs) in plants can transport water and small molecules, and they play an important role in plant development and abiotic stress response. However, to date, a comprehensive study on AQP family members is lacking. In this study, 27 AQP genes were identiﬁed from the passion fruit genome and classiﬁed into four groups (NIP, PIP, TIP, SIP) on the basis of their phylogenetic relationships. The prediction of protein interactions indicated that the AQPs of passion fruit were mainly associated with AQP family members and boron protein family genes. Promoter cis-acting elements showed that most PeAQPs contain light response elements, hormone response elements, and abiotic stress response elements. According to collinear analysis, passion fruit is more closely related to Arabidopsis than rice. Furthermore, three different fruit ripening stages and different tissues were analyzed on the basis of the transcriptome sequencing results for passion fruit AQPs under drought, high-salt, cold and high-temperature stress, and the results were conﬁrmed by qRT-PCR. The results showed that the PeAQPs were able to respond to different abiotic stresses, and some members could be induced by and expressed in response to multiple abiotic stresses at the same time. Among the three different fruit ripening stages, 15 AQPs had the highest expression levels in the ﬁrst stage. AQPs are expressed in all tissues of the passion fruit. One of the passion fruit aquaporin genes, PeTIP3-2 , which was induced by drought stress, was selected and transformed into Arabidopsis. The survival rate of transgenic plants under drought stress treatment is higher than that of wild-type plants. The results indicated that PeTIP3-2 was able to improve the drought resistance of plants. Our discovery lays the foundation for the functional study of AQPs in passion fruit. without water for 8 and 11 days. The results showed that under drought stress, the survival rate of the transgenic plants was higher than that of the wild-type plants.


Introduction
Passiflora (Passiflora edulis) is a perennial evergreen vine of the passiflora genus (passifloraceae), a tropical and rare fruit tree. Passion fruit is named after its fruit pulp, which contains the aroma of over 100 different fruits. It is native to central and northern South America, and is widely distributed throughout Central and South America, Australia, and Africa [1]. According to reports, there are about 520 species of passiflora, most of which are cultivated for flower viewing and gardening, and only about 60 species can be eaten fresh [2]. At present, the main countries in which passion fruit is grown are Brazil, Colombia, Ecuador, Australia, Vietnam, China, etc. The planting area is gradually

Identification of the Passion Fruit AQP Family
On the basis of HMM, protein BLAST searches, and the use of the MEME software, a total of 27 PeAQP members were identified and annotated from the passion fruit genome database. In addition, we analyzed the characteristics of the PeAQPs (Table 1); the length of the PeAQP CDS sequence ranged from 234 bp for PeNIP6-1 to 1794 bp for PeSIP2-1. The identified PeAQPs encoded proteins ranging from 78 amino acids in the case of PeNIP6-1 to 598 amino acids in the case of PeSIP2-1. With respect to the phosphorylation site analyses, 89% of PeAQPs were found to contain all three phosphorylation sites (Ser, Thr and Tyr). Among them, PeNIP6-1, PeSIP2-1, and PeSIP2-2 only contained the Ser and Thr phosphorylation sites. With respect to subcellular location prediction analysis, except for PeNIP6-1, located in the nucleus, all other AQP members were located on the cell membrane.

Evolutionary Characterization of AQP Genes
To study the classification and evolutionary relationship of AQP proteins in passion fruit, a phylogenetic tree was constructed with full-length protein sequences of the aquaporin gene family members in Arabidopsis, rice, and passion fruit ( Figure 1).
The exon-intron structure is an important evolutionary feature of genes. We also further analyzed the exon-intron organization of PeAQP family members. Most of the members had structures of 5′UTR (untranslated region) and 3′UTR, except that of PeTIP1-3 and PeNIP6-1, which had no UTR structure. In addition, PePIP2-2 and PeTIP2-2 Figure 1. Phylogenetic relationships of AQPs among the passion fruit, Arabidopsis, and rice. The tree was constructed by ClustalX and MEGA 7.0 using the neighbor-joining method with 1000 bootstraps. The four sub-families (TIPs, NIPs, SIPs, PIPs) were classified and displayed in different groups by different colored outer rings. The circle represents the AQPs in passion fruit; the five-pointed star represents the AQPs in Arabidopsis; the square represents the AQPs in rice.
The PeAQPs in passion fruit can be divided into four groups (PIP, NIP, TIP, and SIP). There are nine PIPs, seven NIPs, eight TIPs, and three SIPs. Among the PIP subfamily, nine members were divided into two groups: PIP1 with two members and PIP2 with seven members. Four groups were found for the TIP subfamily (TIP1 to TIP4), with four members in the TIP1 group, two members in each of the TIP2 and TIP3 groups, and one member in the TIP4 group. Four groups belonged to the NIP subfamily, with two members in the NIP1 group, three members in the NIP4 group, and one member in each of the NIP5 and NIP6 groups.
Because rice and passion fruit are more distantly related, as shown in the figure, they have fewer close pairings. For PIP subfamily genes, only one member, PePIP1-2, was the best orthologous match of OsPIP1-2 and OsPIP1-3. For NIP subfamily genes, PeNIP4-3 was closely related to OsNIP2-1 and OsNIP2-2. PeNIP5-1 was the most homogeneous gene of OsNIP3-1. For TIP subfamily genes, PeTIP1-1 was phylogenetically closest to OsTIP1-1.
The exon-intron structure is an important evolutionary feature of genes. We also further analyzed the exon-intron organization of PeAQP family members. Most of the members had structures of 5 UTR (untranslated region) and 3 UTR, except that of PeTIP1-3 and PeNIP6-1, which had no UTR structure. In addition, PePIP2-2 and PeTIP2-2 had only the 5 UTR structure.
phylogenetic tree was constructed based on PeAQPs and divided into four different color groups. A total of 15 motifs were denoted by different colored boxes. Untranslated (UTR) 5′ and 3′ regions are indicated by blue boxes, exon regions are indicated by yellow frames, and intron regions are indicated by black lines.

Promoter Analysis of PeAQP Genes
Predicting the cis-regulatory elements in PeAQP gene promoters can help to better understand the transcriptional regulatory functions of gene family members ( Figure 4). Among them, the TATA-box and the CAAT-box could be found in 27 PeAQPs. Moreover, the promoter region of this family gene contained a large number of action elements related to abiotic stress, including drought-responsive elements (CAACTG), cold-responsive elements (CCGAAA), and salicylic acid-responsive elements (TCAGAAGAGG and CCATCTTTTT). Additionally, there were some hormone-related

Promoter Analysis of PeAQP Genes
Predicting the cis-regulatory elements in PeAQP gene promoters can help to better understand the transcriptional regulatory functions of gene family members ( Figure 4). Among them, the TATA-box and the CAAT-box could be found in 27 PeAQPs. Moreover, the promoter region of this family gene contained a large number of action elements related to abiotic stress, including drought-responsive elements (CAACTG), cold-responsive elements (CCGAAA), and salicylic acid-responsive elements (TCAGAAGAGG and CCATCTTTTT). Additionally, there were some hormone-related elements. Abscisic acid-responsive elements (ABREs) such as GACACGTGGC and ACGTG, MeJA-responsive motifs such as CGTCA and TGACG, gibberellin-responsive motifs (TATCCCA, CCTTTTG and TCTGTTG), and auxin-responsive elements (AACGAC) were present in abundance. This indicated that the PeAQP gene was regulated by stress response and was involved in plant growth and development.

Interaction Network Analysis of PeAQPs
To better understand the biological functions and regulatory networks involved in PeAQPs, the orthology-based method was used to predict the protein interaction relationships between them. The results showed that 13 PeAQPs had an orthologous relationship with Arabidopsis thaliana, and 13 interacting proteins were found. Most of the proteins that interacted with PeAQPs were aquaporin family genes, whereas the others included boron protein family genes (BOR1, BOR4) ( Figure 5); this is consistent with the function of aquaporins in transporting boron ions. These protein-protein interaction networks provide clues for studying the function of PeAQPs and a basis for further research of candidate genes. elements. Abscisic acid-responsive elements (ABREs) such as GACACGTGGC and AC-GTG, MeJA-responsive motifs such as CGTCA and TGACG, gibberellin-responsive motifs (TATCCCA, CCTTTTG and TCTGTTG), and auxin-responsive elements (AACGAC) were present in abundance. This indicated that the PeAQP gene was regulated by stress response and was involved in plant growth and development.

Interaction Network Analysis of PeAQPs
To better understand the biological functions and regulatory networks involved in PeAQPs, the orthology-based method was used to predict the protein interaction relationships between them. The results showed that 13 PeAQPs had an orthologous relationship with Arabidopsis thaliana, and 13 interacting proteins were found. Most of the proteins that interacted with PeAQPs were aquaporin family genes, whereas the others included boron protein family genes (BOR1, BOR4) ( Figure 5); this is consistent with the function of aquaporins in transporting boron ions. These protein-protein interaction networks provide clues for studying the function of PeAQPs and a basis for further research of candidate genes.

Homology Modeling
The 3D structures of these PeAQP proteins were further predicted using homolo gy-based modeling ( Figure 6). The structures of most of the PeAQPs were similar, and contained α-helices (the blue part), an extension chain (the red part), and an irregula curl (the orange part), except that of PeNIP4-1, PeNIP4-3, PePIP1-5, PeTIP3-1, PeTIP3-2 PeTIP4-1, PeSIP2-1, PeSIP2-2, and PePIP1-2, which only contained α-helices (the blue part) and the irregular curl (the orange part); PeSIP1-1 only contained α-helices (the blue part) and the extension chain (the red part); PeTIP1-1 and PeTIP2-1 only contained Figure 5. Predicted protein-protein interaction networks of PeAQP proteins with other passion fruit proteins using the STRING tool. The circles on the outside represent PeAQP proteins, and the other circles shows proteins that interact with PeAQPs. The gray connections between the circles represent the interactions between proteins.

Collinearity Analysis
The collinearity diagrams on the PeAQPs were further analyzed. All genes were displayed on the chromosome circle except PePIP1-2. Nine collinearity pairs were found. Most PeAQP genes were found to have one paralogous gene, except for PePIP2-12, which had two homologous genes in passion fruit ( Figure 7).

Collinearity Analysis
The collinearity diagrams on the PeAQPs were further analyzed. All genes were displayed on the chromosome circle except PePIP1-2. Nine collinearity pairs were found. Most PeAQP genes were found to have one paralogous gene, except for PePIP2-12, which had two homologous genes in passion fruit ( Figure 7).
In addition, we also performed collinearity analysis on AQP members in passion fruit, Arabidopsis, and rice ( Figure 8). There were 12,406 and 3118 genes in Arabidopsis and rice, respectively, that had a collinearity relationship with the genes in passion fruit. Among the 27 members of the PeAQPs, each of them had a collinear relationship with members in Arabidopsis. This is consistent with the evolutionary relationship between passion fruit and Arabidopsis, which are both dicotyledonous plants. Among them, seven genes (PePIP2-12, PePIP2-2, PeTIP1-2, PeTIP1-1, PeTIP4-1, PeNIP1-1, and PeNIP1-2) had two homologous genes in Arabidopsis, whereas five genes (PePIP2-11, PePIP2-8, PeTIP2-1, PeTIP2-2, and PeSIP1-1) had three orthologous genes in Arabidopsis. In addition, three genes (PePIP2-8, PeNIP1-1, and PeNIP4-3) had one orthologous gene in rice. These results help us to further understand the evolutionary relationship of AQPs in passion fruit.

Collinearity Analysis
The collinearity diagrams on the PeAQPs were further analyzed. All genes were displayed on the chromosome circle except PePIP1-2. Nine collinearity pairs were found. Most PeAQP genes were found to have one paralogous gene, except for PePIP2-12, which had two homologous genes in passion fruit (Figure 7).  In addition, we also performed collinearity analysis on AQP members in passion fruit, Arabidopsis, and rice ( Figure 8). There were 12,406 and 3,118 genes in Arabidopsis and rice, respectively, that had a collinearity relationship with the genes in passion fruit. Among the 27 members of the PeAQPs, each of them had a collinear relationship with members in Arabidopsis. This is consistent with the evolutionary relationship between passion fruit and Arabidopsis, which are both dicotyledonous plants. Among them, seven genes (PePIP2-12, PePIP2-2, PeTIP1-2, PeTIP1-1, PeTIP4-1, PeNIP1-1, and PeNIP1-2) had two homologous genes in Arabidopsis, whereas five genes (PePIP2-11, PePIP2-8, PeTIP2-1, PeTIP2-2, and PeSIP1-1) had three orthologous genes in Arabidopsis. In addition, three genes (PePIP2-8, PeNIP1-1, and PeNIP4-3) had one orthologous gene in rice. These results help us to further understand the evolutionary relationship of AQPs in passion fruit.
In the early stage, we performed transcript sequencing of three different fruit ripening stages in passion fruit [32], on the basis of which we analyzed the expression levels of the AQP family members (Figure 11). On the basis of the results of the heat map, most of the genes exhibited their highest expression levels during the first period, and their expression levels gradually decreased with decreasing fruit moisture during the next two periods. There was also a small number of family members whose expression level increased as the fruit matured, such as PePIP1-2, PePIP2-7, and PeTIP2-2. We also performed qRT-PCR verification, and the results showed that this gene expression trend was consistent with the transcriptome result ( Figure 12). As the passion fruit ripened, its fruit moisture also decreased accordingly. This result indicated that the expression of AQPs decreased with increasing fruit maturity, which may be related to the water content in passion fruit.  Table S3. Data are means ± SD of n = 3 biological replicates. Means denote same letter are not significantly different at p < 0.05 as determined by Duncan's multip test.
In the early stage, we performed transcript sequencing of three different fr ening stages in passion fruit [32], on the basis of which we analyzed the express els of the AQP family members (Figure 11). On the basis of the results of the he most of the genes exhibited their highest expression levels during the first perio their expression levels gradually decreased with decreasing fruit moisture dur next two periods. There was also a small number of family members whose exp level increased as the fruit matured, such as PePIP1-2, PePIP2-7, and PeTIP2-2. W  Table S3. Data are means ± SD of n = 3 biological replicates. Means denoted by the same letter are not significantly different at p < 0.05 as determined by Duncan's multiple range test.  Table S2. Blue indicates a low expression level and red indicates a high expressi The heat map was generated using TBtools.  Table S2. Blue indicates a low expression level and red indicates a high expression level. The heat map was generated using TBtools.

Overexpression of PeTIP3-2 Enhances Transgenic Arabidopsis Tolerance to Drought Stress
At present, the function of AQPs in plants has been discussed with respect to abiotic stress. In this study, we focused on PeTIP3-2, which is induced under drought stress ( Figure 14). The transgenic plants were grown in 1/2MS and mannitol-added medium (simulating drought stress) for 14 days. The results of GUS staining showed that the transgenic plants were darker under drought stress, indicating that the expression of PeTIP3-2 was induced under drought stress. We obtained three independent PeTIP3-2 transgenic Arabidopsis lines, named L1, L2, and L3. The transgenic and wild-type plants (WT) grown for 4 weeks were treated without water for 8 and 11 days. The results showed that under drought stress, the survival rate of the transgenic plants was higher than that of the wild-type plants.  Table S4. Data are means ± SD of n = 3 biological replicates. Mea denoted by the same letter are not significantly different at p < 0.05 as determined by Dunca multiple range test.
Finally, we also analyzed the expression of AQPs in the leaves, roots, stems, a fruits of passion fruit ( Figure 13). Among 27 members, 11 of them were mainly e pressed in the leaves (PeNIP1-1, PeNIP1-2, PeNIP4-2, PeNIP4-3, PePIP2-7, PePIP2-9, Pe IP2-10, PePIP2-11, PeSIP2-1, PeTIP1-2, and PeNIP4-3). Others were mainly concentrat in the leaves, roots, and stems, but to a lesser degree in the fruits, such as PePIP1-2, Pe IP2-7, and PeTIP2-2. The results showed that each gene was expressed in different part  Table S4. Data are means ± SD of n = 3 biological replicates. Means denoted by the same letter are not significantly different at p < 0.05 as determined by Duncan's multiple range test.  Table S5. Data are means ± SD of n = 3 biological replicates. Means denoted by the same letter are not significantly different at p < 0.05 as determined by Duncan's multiple range test.

Overexpression of PeTIP3-2 Enhances Transgenic Arabidopsis Tolerance to Drought Stress
At present, the function of AQPs in plants has been discussed with respect to abiotic stress. In this study, we focused on PeTIP3-2, which is induced under drought stress ( Figure 14). The transgenic plants were grown in 1/2MS and mannitol-added medium (simulating drought stress) for 14 days. The results of GUS staining showed that the  Table S5. Data are means ± SD of n = 3 biological replicates. Means denoted by the same letter are not significantly different at p < 0.05 as determined by Duncan's multiple range test. After 5 and 15 days of rehydration, most of the transgenic plants grew stronger than the wild-type plants. Physiological index measurements were performed on plants after being subjected to drought stress for 11 days and rewatering for 15 days. Under normal growth conditions, the MDA, ion leakage rate, PRO, and soluble sugar content of WT and transgenic plants were not significantly different. After drought stress and rewatering treatment, the MDA content and ion leakage rate in transgenic plants were lower, whereas PRO and soluble sugars were higher than those of WT. The results indicated that PeTIP3-2 can improve the drought resistance of plants.

Osmotic Stress-Induced Activity Analysis in Transgenic Tobacco
We further transiently transformed tobacco with PeTIP3-2 ( Figure 15). The control and infected leaves were soaked in 1/2MS and 1/2MS + 18% PEG solutions, respectively. The samples were taken at 0, 2, 4, 8, 12, 24, 36, and 48 h, and then subjected to GUS staining. The results showed that the GUS staining of the other treatment groups was deepened after PEG treatment except for the 0 h group, indicating that PeTIP3-2 could be induced by drought stress. Among them, the GUS staining of leaves was the darkest After 5 and 15 days of rehydration, most of the transgenic plants grew stronger than the wild-type plants. Physiological index measurements were performed on plants after being subjected to drought stress for 11 days and rewatering for 15 days. Under normal growth conditions, the MDA, ion leakage rate, PRO, and soluble sugar content of WT and transgenic plants were not significantly different. After drought stress and rewatering treatment, the MDA content and ion leakage rate in transgenic plants were lower, whereas PRO and soluble sugars were higher than those of WT. The results indicated that PeTIP3-2 can improve the drought resistance of plants.

Osmotic Stress-Induced Activity Analysis in Transgenic Tobacco
We further transiently transformed tobacco with PeTIP3-2 ( Figure 15). The control and infected leaves were soaked in 1/2MS and 1/2MS + 18% PEG solutions, respectively. The samples were taken at 0, 2, 4, 8, 12, 24, 36, and 48 h, and then subjected to GUS staining. The results showed that the GUS staining of the other treatment groups was deepened after PEG treatment except for the 0 h group, indicating that PeTIP3-2 could be induced by drought stress. Among them, the GUS staining of leaves was the darkest under 8, 12, and 24 h of treatment. This indicated that the gene was highly induced during these periods. under 8, 12, and 24 h of treatment. This indicated that the gene was highly induced during these periods.

Discussion
Because of its short growth cycle and unique flavor, passion fruit is becoming increasingly popular worldwide. Biotic and abiotic stresses affect its growth and development processes. Therefore, mining the genes and analyzing its mechanism of resistance is of great significance for improving the resistance of the passion fruit. AQPs play important roles in plant growth and developmental processes [16]. In the current research, we conducted a genome-wide identification of PeAQPs, focusing on abiotic stress and fruit maturity stage, which will provide new clues for further research on the function of AQP genes and the genetic improvement in the passion fruit.

The Evolutionary Characteristics of the AQP Family in Plants
In this research, the classification of AQPs in passion fruit, like most higher plants, such as Arabidopsis, maize, rice [27,28,33] and barley [34], is performed through a division into four subfamilies. In addition, the members of the AQP family in cucumber, soybean, sesame, and tomato can be divided into five subfamilies: PIP, NIP, SIP, TIP, and XIP [16].
According to several reports, plant AQPs have different biochemical characteristics related to their functions [35]. The results showed that the physical and chemical properties of different subgroup members are different to some extent. The difference is mainly due to the presence of basic residues in the C-terminal domain of AQP proteins, as also observed in sweet orange AQPs [36]. Related reports have shown that the subcellular localization of NIPs and PIPs in most plants usually takes place on the plasma membrane, whereas TIPs and SIPs are usually located on the tonoplast and endoplasmic reticulum [37,38]. In this research, the subcellular localization results of PeAQP showed that only PeNIP6-1 was predicted to be localized in the nucleus or mitochondria, whereas other members were on the cell membrane. This is consistent with the function of aquaporin in plants, which is mainly to transport water and other small molecules such as glycerol, CO2, and boron through biological membranes [4]. Among the TIP family members of melon, CmTIP3;1 and CmTIP5;1 are located in the mitochondria and chloroplasts, respectively [39]. In some plants, the aquaporin gene is located in two different regions, such as the ZmPIP1;2 genes in maize, which are located in the plasma membrane and endoplasmic reticulum (ER) [40]. In tobacco, the NtAQP1 gene has been reported to be located in the plasma membrane and the inner chloroplast membrane [41]. AtTIP3;1 is located in both the plasma membrane and the vacuolar membrane [42]. An increasing number of studies have shown that the subcellular localization of plant genes is also closely related to abiotic stress factors [43].

Discussion
Because of its short growth cycle and unique flavor, passion fruit is becoming increasingly popular worldwide. Biotic and abiotic stresses affect its growth and development processes. Therefore, mining the genes and analyzing its mechanism of resistance is of great significance for improving the resistance of the passion fruit. AQPs play important roles in plant growth and developmental processes [16]. In the current research, we conducted a genome-wide identification of PeAQPs, focusing on abiotic stress and fruit maturity stage, which will provide new clues for further research on the function of AQP genes and the genetic improvement in the passion fruit.

The Evolutionary Characteristics of the AQP Family in Plants
In this research, the classification of AQPs in passion fruit, like most higher plants, such as Arabidopsis, maize, rice [27,28,33] and barley [34], is performed through a division into four subfamilies. In addition, the members of the AQP family in cucumber, soybean, sesame, and tomato can be divided into five subfamilies: PIP, NIP, SIP, TIP, and XIP [16].
According to several reports, plant AQPs have different biochemical characteristics related to their functions [35]. The results showed that the physical and chemical properties of different subgroup members are different to some extent. The difference is mainly due to the presence of basic residues in the C-terminal domain of AQP proteins, as also observed in sweet orange AQPs [36]. Related reports have shown that the subcellular localization of NIPs and PIPs in most plants usually takes place on the plasma membrane, whereas TIPs and SIPs are usually located on the tonoplast and endoplasmic reticulum [37,38]. In this research, the subcellular localization results of PeAQP showed that only PeNIP6-1 was predicted to be localized in the nucleus or mitochondria, whereas other members were on the cell membrane. This is consistent with the function of aquaporin in plants, which is mainly to transport water and other small molecules such as glycerol, CO 2 , and boron through biological membranes [4]. Among the TIP family members of melon, CmTIP3;1 and CmTIP5;1 are located in the mitochondria and chloroplasts, respectively [39]. In some plants, the aquaporin gene is located in two different regions, such as the ZmPIP1;2 genes in maize, which are located in the plasma membrane and endoplasmic reticulum (ER) [40]. In tobacco, the NtAQP1 gene has been reported to be located in the plasma membrane and the inner chloroplast membrane [41]. AtTIP3;1 is located in both the plasma membrane and the vacuolar membrane [42]. An increasing number of studies have shown that the subcellular localization of plant genes is also closely related to abiotic stress factors [43].
The exon-intron structure is an important feature of genes. In this research, most family members contained 2-4 exons. The same pattern also exists in chickpeas. The research of Amit A. Deokar showed that most of the PIP, TIP, and NIP subfamily in chickpeas have 3, 2, and 4 introns, respectively [44]. Similar exon-intron patterns of the AQP family among species have also been reported in tomatoes, oranges, and soybeans [11,36,45]. In general, the intron-exon pattern is related to the conservation of gene family classification.

The Function of AQP Promoters in Plants
Gene expression is regulated by promoters, which contain multiple cis-acting elements [47]. Promoters play an important role in regulating gene expression [48]. In this study, we selected the 2000 bp region upstream of the PeAQP CDS sequence and used PlantCARE to perform predictive analysis of the elements [49]. Most of the elements were light-responsive elements. It was found that light is one of the most important environmental factors regulating plant growth and development. Light regulates gene expression networks at many levels [50]. At the same time, there are related reports showing that light has an effect on the expression and function of the AQP family [51]. In addition to lightresponsive elements, cis-acting elements related to abiotic stress were also identified, such as drought-responsive, cold-responsive, and salicylic acid-responsive elements. Secondly, there are some elements related to plant hormones such as the abscisic acid responsiveness element (ABRE), the MeJA-responsive element, gibberellin-responsive motifs, and the auxin-responsive element. This suggests that PeAQPs are regulated by stress responses and associated with plant growth and development.

The Functional Diversity of AQPs in Plants
In this study, we performed transcriptome sequencing analysis of four different stress treatments: drought, high salt, cold, and high temperature. Among the members of the passion fruit AQP family, most members can respond to different abiotic stresses. For example, PePIP1-2 can respond to drought, high salt, and cold stress. PeTIP2-2 and PeTIP3-2 can be induced by and expressed in response to drought, high salt, cold, and high temperature. We performed qPCR detection and analysis of AQP gene expression in passion fruit under drought stress. Some members could be induced to express by drought stress, and one of the most highly induced genes, PeTIP3-2, was selected for the target gene with further functional verification. The results showed that PeTIP3-2 could be induced in transgenic plants under drought stress. Most reports indicate that AQPs in plants can respond to abiotic stress; the banana AQP gene MaPIP1;1 in transgenic Arabidopsis and banana can respond to drought and cold stresses [13,22]. MaPIP2-7 and MaSIP2-1 can respectively improve the drought resistance, salt, and cold tolerance in the transgenic bananas [4]. MaTIP1;2 in banana can improve the drought and salt tolerance in transgenic Arabidopsis [52]. CaTIP1-4, CaNIP1-4, and CaNIP1-8 in chickpeas can respond to high salt stress. CaTIP1-1 and CaPIP1-1 can improve the drought resistance and salt and high temperature tolerance of plants [44]. TaAQP7 and TaAQP8 of wheat can improve the drought and salt tolerance of transgenic tobacco [12,53]. Banana MusaPIP2;6 can be highly induced by salt stress [54].
In addition, we also analyzed the expression of AQP genes during three fruit ripening periods [32], and the results showed that most of the AQP genes exhibited their highest expression levels in the first stage. As the fruit matures and the degree of water loss in the fruit increases, its expression gradually decreases, which may be consistent with the function of AQPs in transporting water. Some studies have shown that AQP is related to plant development. In grapes, some PIP genes have their maximum expression levels 60 to 80 days after flowering, among which, the expression levels of VvPIP1-3 and VvPIP2-1 decreased after 80 days of flowering, which was related to the increase in water resistance of grape xylem [55]. For SlPIP1;2, SlPIP2;4, SlTIP1;1, and SlTIP2;1 in tomato, the expression level in the fruit remains high from 3 to 14 days after pollination [11]. FaPIP1 and FaPIP2 in strawberry are more highly expressed in the harder tissues, indicating that they may play a role in fruit firmness [9]. By detecting the tissue-specific and development-specific expression of tomato aquaporin genes, the contribution of aquaporin to water transport in leaves and fruit development is revealed [11].
Furthermore, the results of PeAQP expression in different plant tissues indicated that the members are expressed in each of the different plant tissues, and categorized into stems, leaves, roots, and fruits. The location of AQPs in different plants may be closely related to their function. For example, MaPIP1;1 in banana is mainly expressed in the roots. It has been found that the roots of MaPIP1;1 in transgenic Arabidopsis are longer than those in wild-type plants. The aquaporin gene CmPIP2;10 in melon has the highest expression in the roots, and it has been found that it can transport solutes such as urea. Additionally, CmNIP5;1 and CmNIP5;2 are most highly expressed in roots and leaves [39]. The PIP and TIP family members in wheat are mostly expressed in the leaves [17]. It has been found in chickpeas that AQPs have the highest expression levels in meristems with high water flux, such as leaves, shoots, and roots [44]. This corresponds to the main function of plant AQPs, which is to transport water. Water-tolerant and -sensitive wheat were studied with and without drought stress treatment, respectively. The expression of different AQP genes in the tolerant wheat roots decreased under drought stress, while expression was induced in the sensitive plants. This shows that AQPs play an important role in transporting water in the root system of wheat [17]. The results also showed that the PIP subfamily in cotton has the highest expression in the roots, leaves, and fibers, indicating that it is involved in water transport and CO 2 diffusion in cotton [16].

Identification of AQP Genes in Passion Fruit
The passion fruit genome sequences were retrieved from Phytozome V12.1 (https:// ngdc.cncb.ac.cn/search/?dbId=gwh&q=GWHAZTM00000000) (accessed on 22 November 2021). Meanwhile, the AQP protein domain (PF00230) was downloaded from the PFAM database (http://pfam.xfam.org/) (accessed on 3 December 2021) and was applied to identify the candidate passion fruit AQP members. Additionally, blast analyses with all AQPs from the top three with the highest comparison values were used to further identify possible AQPs in the passion fruit database. Furthermore, the protein sequences identified by both of the above methods were integrated and parsed by manual editing to remove redundancies. The Arabidopsis and rice AQP protein sequences were retrieved and downloaded from the ensemble plant database (https://plants.ensembl.org/index. html) (accessed on 18 January 2022). The full-length protein sequences of the AQPs in passion fruit, Arabidopsis, and rice were aligned using ClustalX2 software (https:// clustalx.software.informer.com/2.1/) (accessed on 22 January 2022). After the ALN file was generated, MEGA6 software (https://www.megasoftware.net/) (accessed on 19 January 2022) was used to build the neighbor-joining (NJ) phylogenetic tree [56]. The criteria were adopted with the pairwise deletion option and the Poisson correction model. A bootstrap test was performed with 1000 replicates. Finally, the PeAQP protein motifs were analyzed with the MEME tool (https://meme-suite.org/meme/tools/meme) (accessed on 19 January 2022) to compare whether the above PeAQPs had a common structural domain, and the exception members were removed. The members of the PeAQPs were finally considered by the screening of the above several methods.

Gene Identification, Gene Structure, and Chromosomal Locations
The protein iso-electric point (PI), molecular weight (MW), and the molecular formula for all candidate family members were computed using the ProtParam website (http: //www.expasy.org/tools/protparam.html) (https://meme-suite.org/meme/tools/meme) (accessed on 11 February 2022). The NetPhos 3.1 Server (http://www.cbs.dtu.dk/services/ NetPhos/) (accessed on 11 February 2022) was used to predict the protein phosphate site. Their subcellular localization was predicted using the WoLF PSORT software (https: //www.genscript.com/wolf-psort.html) (accessed on 15 February 2022). The gene structure map was obtained using TBtools (Guangzhou, Guangdong, China) [57]. The combined visualization of the phylogenic tree, the motifs, and the gene structures of the PeAQPs were also obtained using TBtools, as were the visualizations of the chromosome localization and syntenic relationships [57].

Analysis of cis-Acting Elements of PeAQP Genes
A total of 2000 bp of genomic DNA sequence upstream of the transcriptional start site in each PeAQP was obtained from the PlantCARE database (http://bioinformatics.psb. ugent.be/webtools/plantcare/html/) (accessed on 18 February 2022), which was used to identify the cis-acting elements in the promoter region of PeAQPs [49].

Analysis of Gene Protein Interaction Regulatory Network
First, the orthovenn2 tool was used (https://orthovenn2.bioinfotoolkits.net/home) (accessed on 22 February 2022) to identify the orthologous pairs between PeAQPs and AtAQPs. Second, the interaction networks in which PeAQPs were involved were identified based on the orthologous genes between the passion fruit and Arabidopsis using the AraNetV2 (http://www.inetbio.org/aranet/) (accessed on 3 March 2022). The STRING (http://string-db.org/cgi) (accessed on 5 March 2022) database and the predicted interaction network were displayed using Cytoscape software (https://cytoscape.org/) (accessed on 5 March 2022).

Gene Collinearity Analysis
The Multicollinearity Scanning Toolkit (MCScanX) (Fuzhou, Fujian, China) was used to analyze gene duplication events between species and within the same species [57].

Plant Materials and Growth Conditions
Healthy and virus-free passion fruit seedlings of purple fruit varieties were chosen. They were grown in soil under a growth chamber (30 • C; 200 µmol·m −2 ·s −1 light intensity; 12 h light/12 h dark cycle; 70% relative humidity) to a height of about 1 m and with 8-10 functional leaves, which were subjected to various abiotic stress treatments. For drought stress analysis, water was withheld from the treatment group to obtain soil moisture of 50% and 10%. For salt stress tolerance analysis, the seedlings were treated with 300 mM Nacl solution after 3 and 10 days. For the cold treatment, the plants were treated at 0 • C for 20 h and 48 h, respectively. For the high temperature treatment, the plants were treated at 45 • C for 2 h, 4 h, and 24 h.

RNA Extraction, Transcriptome Sequencing, and qRT-PCR
We froze the samples, and the total RNA was extracted from samples with different abiotic stress treatments and fruit maturity stages, and from the roots, stems, and leaves of the passion fruit, using the plant RNA isolation kit (Fuji, China, Chengdu) with three biological replicates. The cDNA was used for transcriptome sequencing analysis and qRT-PCR. Transcriptome data under different abiotic stress conditions and at different fruit maturity stages were used in this study. Primer sequences were designed using the Primer 5.0 tool. The expression of PeAQPs was detected by quantitative real-time polymerase chain reaction (qRT-PCR) analysis using SYBR ® Premix Ex Taq™ (TaKaRa, Japan, Tokyo) chemistry on Lightcycle-480 (Roche). Relative expression levels were calculated using the 2 −∆∆Ct method and normalized to the PeAQPs.

Heat Map
The genome-wide microarray data of PeAQPs in response to drought, salt, cold, and high-temperature stress and at various fruit maturity stages are shown in Tables S1 and S2. The transcript data of PeAQPs were analyzed using TBtools software (https: //bio.tools/tbtools) (accessed on 22 January 2021). The normalized expression data were used to generate a heatmap using TBtools software [57].

Cloning and Vector Construction of PeTIP3-2
The full-length cDNA of PeTIP3-2 was amplified from the passion fruit purple fruit varieties by reverse transcription-polymerase chain reaction (RT-PCR) based on the sequence information in the passion fruit genome database (https://bigd.big.ac.cn/gwh/) (accessed on 1 January 2021). Healthy passion fruit seedlings were used to construct a single-stranded cDNA template. PCR products were cloned into the pMD19-T vector (Promega, Madison, WI, USA) and sequenced on an ABI PRISM310 Genetic Analyzer (PerkinElmer Applied Biosystems, Foster City, CA, USA). The full-length cDNA of PeTIP3-2 was assessed by DNA-MAN software (https://www.lynnon.com/dnamanxl.html) (accessed on 22 January 2022) and BLAST (http://ncbi.nlm.nih.gov/blast) (accessed on 22 January 2021). The PeTIP3-2 ORF, including engineered NcoI/SpeI restriction sites, was obtained using gene-specific primers. The PCR products were inserted into the pCAMBIA1304 expression vector to generate a PeTIP3-2-GFP fusion protein under the control of the CMV35S promoter.

Plant Transformation and Generation of Transgenic Plants
The pCAMBIA1304-PeTIP3-2 construct was transferred into the Agrobacterium strain EHA105. Transgenic Arabidopsis was generated using the floral DIP-mediated infiltration method [60]. Seeds from T0 transgenic plants were selected on 1/2MS medium containing 25 mg/L of hygromycin B. Homozygous T3 lines were used for further functional investigation.

Drought and PEG Stress Treatment in WT and Transgenic Plants
The plants were subjected to drought stress tolerance analysis experiments. The wildtype and transgenic plants were grown at 23 • C for 4 weeks; water was withheld for 8 days and 11 days, and then they were given 5 days and 15 days of recovery, respectively, before being observed and photographed.
The 60-day-old tobacco leaves were used for transient expression experiments and treated with the detached leaves. Leaf discs with a diameter of 0.5 cm were cut out and floated in 1/2 MS liquid medium supplemented with 18% (w/v) PEG 6000 (osmotic stress) at 25 • C for 0, 2, 4, 8, 12, 24, 36, and 48 h, respectively. The leaf discs floating in 1/2 MS liquid medium were considered the control.

GUS Activity Detection
The 15-day transgenic Arabidopsis and transiently transformed tobacco in normal and mannitol-containing medium were stained with GUS and then photographed. Fresh samples (small plants) were placed in X-Gluc solution (GBT, St. Louis, MO, USA) [61] for histochemical analysis. The seedlings were placed in GUS staining solution containing 50 mM sodium phosphate (pH 7.0), 0.5 mM potassium ferricyanide, 0.5 mM potassium ferrocyanide, 10 mM EDTA, 0.1% Triton X-100, and 1 mM X-Gluc, then incubated at 37 • C for 24 h. Fluorometric determination of GUS enzyme activity was conducted using 4-methylumbelliferyl glucuronide.

Conclusions
Aquaporins in plants can transport water and small molecules, and are widely reported to play a role in the abiotic stress resistance of plants. We identified 27 aquaporin members from the genome of the passion fruit. Analyses were performed with respect to evolutionary tree, structural domains, promoter cis-acting elements, and inter-species, and intra-species collinearity. PeAQPs are able to respond to abiotic stresses such as drought, high salinity, cold, and high temperature. This is expressed in different organizations of the passion fruit. As the fruit matures, the expression of PeAQPs gradually decreases, indicating that it is related to water transport. We focused on the gene PeTIP3-2, which is induced by drought stress. Under drought stress, PeTIP3-2 transgenic Arabidopsis exhibited significantly deeper GUS staining than the control. The survival rate and resistance of transgenic Arabidopsis were higher than those of the wild type. These results indicate that PeTIP3-2 may be used as a candidate gene for the development of drought-tolerant crops. Our research can provide a basis for further understanding the function of PeAQPs and analyzing their mechanism of action in improving plant stress resistance.