Rice aquaporin OsPIP2;2 is a water‐transporting facilitator in relevance to drought‐tolerant responses

Abstract In rice (Oryza sativa), the PLASMA MEMBRANE INTRINSIC PROTEIN (PIP) family of aquaporin has 11 members, OsPIP1;1 to OsPIP1;3, and OsPIP2;1 to OsPIP2;8, which are hypothesized to facilitate the transport of H2O and other small compounds across cell membranes. To date, however, only OsPIP1;2, OsPIP2;1, and OsPIP2;4 have been demonstrated for substrate selectivity in their source plant (rice). In this study, OsPIP2;2 was characterized as the most efficient facilitator of H2O transport across cell membranes in comparison with the other 10 OsPIPs. In concomitant tests of all OsPIPs, four genes (OsPIP1;3, OsPIP2;1, OsPIP2;2, and OsPIP2;4) were induced to express in leaves of rice plants following a physiological drought stress, while OsPIP2;2 was expressed to the highest level. After de novo expression in frog oocytes and yeast cells, the four OsPIP proteins were localized to the plasma membranes in trimer and tetramer and displayed the activity to increase the membrane permeability to H2O. In comparison, OsPIP2;2 was most supportive to H2O import to oocytes and yeast cells. After de novo expression in tobacco protoplasts, OsPIP2;2 exceeded OsPIP1;3, OsPIP2;1, and OsPIP2;4 to support H2O transport across the plasma membranes. OsPIP2;2‐mediated H2O transport was accompanied by drought‐tolerant responses, including increases in concentrations of proline and polyamines, both of which are physiological markers of drought tolerance. In rice protoplasts, H2O transport and drought‐tolerant responses, which included expression of marker genes of drought tolerance pathway, were considerably enhanced by OsPIP2;2 overexpression but strongly inhibited by the gene silencing. Furthermore, OsPIP2;2 played a role in maintenance of the cell membrane integrity and effectively protected rice cells from electrolyte leakage caused by the physiological drought stress. These results suggest that OsPIP2;2 is a predominant facilitator of H2O transport in relevance to drought tolerance in the plant.


| INTRODUCTION
Aquaporins (AQPs) are integral membrane proteins initially defined as "water channels" in all living organisms (Agre, 2004;Agre et al., 1993;Brown, 2017;Preston et al., 1992) but subsequently found to have a broader spectrum of cargo (substrate) selectivity among about 20 compounds (Laloux et al., 2018;Rhee et al., 2017). By the substrate-transporting role, AQPs participate in many physiological and pathological processes in animals (Brown, 2017;He & Yang, 2019;Bollag et al., 2020;Wang, Schoebel, et al., 2020;Wang, Zhang, et al., 2020) and plants (Laloux et al., 2018;Li et al., 2019;Singh et al., 2020;Tian et al., 2016). For example, if an AQP serves as a H 2 O-transporting channel, it will be associated with water relations and drought tolerance under most circumstances, possibly in all organs or throughout full life cycle, either in animals (Brown, 2017;de Laurentis et al., 2020;Li & Wang, 2017) or in plants (Hoai et al., 2020;Li & Wang, 2017;Plett et al., 2020;Vishwakarma et al., 2019). In plants, AQPs fall into five major phylogenic families, including the plasma membrane (PM) intrinsic protein (PIP) family. The PIP family is further divided into the PIP1 subfamily, which contains a varied number of PIP1 proteins from PIP1;1 to PIP1;12, and the PIP2 subfamily, which comprises several PIP2 isoforms from PIP2;1 to PIP2;12, in different plant species (Laloux et al., 2018). These proteins are believed to facilitate the transport of different substrates across PMs in an overlapping or redundant manner (Brown, 2017;Maurel et al., 2015aMaurel et al., , 2015b. However, substrates transported by only a small number of PIPs have been determined so far, indicating the imminent necessity to characterize substrate selectivity of most PIPs in most plant species, especially food crops (Laloux et al., 2018;Singh et al., 2020).
In food crops, rice (Oryza sativa) occupies a prominent position in global food security due to its vast production that is used to feed a huge population in the world. Rice also is representative of cereal crops with respect to the function of PIPs/AQPs in H 2 O transport tightly associated with water utility, osmotic response, and drought tolerance (Afzal et al., 2016;Barzana et al., 2020;Lee et al., 2003;Shekoofa & Sinclair, 2018;Vishwakarma et al., 2019). Drought tolerance is important not only to drylands agriculture (Ayadi et al., 2019;Zhang, Hu, et al., 2019) but also for flooding crops, typically like rice planted in the conventional agriculture (Ding et al., 2015(Ding et al., , 2019Grondin et al., 2016;Henry et al., 2012;Oladosu et al., 2019;Sriskantharajah et al., 2020;Vinnakota et al., 2016). In many countries that possess a large population with relatively little arable land, coastal wetland and salt lick have been reclaimed for rice planting (Chen et al., 2015;Zong et al., 2007;Mañosa et al., 2001;Withers, 2002).
Compared with other cereal crops, rice has more excessive transpiration from leaves and lower hydraulic conductance of roots (refer to Nada & Abogadallah, 2020). Therefore, rice is more sensitive to water deficit, to which the water-transporting role of AQPs/PIPs provides an effective tolerance for survival (Ding et al., 2015(Ding et al., , 2019Grondin et al., 2016;Oladosu et al., 2019;Vinnakota et al., 2016).
Functional multiplicity is a common characteristic of eukaryotic AQPs (Brown, 2017), while PIPs may be more multifaceted in rice than in other plants, especially the biological model Arabidopsis thaliana.
Rice has almost 36-fold larger genome size than Arabidopsis but does not have a homolog of AtPIP1;4 and AtPIP1;5 (Sakurai et al., 2005).
Both rice and Arabidopsis have the same number of members (PIP2;1 to PIP2;8) in the PIP2 subfamily, but the number of members is different in the PIP1 subfamily. The PIP2 subfamily is consisting of AtPIP1;1 to AtPIP1;5 in Arabidopsis but only has OsPIP1;1 to OsPIP1;3 in rice (Laloux et al., 2018). Presumably, OsPIPs have a higher degree of T A B L E 1 Rice PIPs characterized for substrate selectivity and physiological or pathological relevance  Kromdijk et al., 2020) and functional regulation mechanisms (Kukulski et al., 2005;Kirscht et al., 2016;Singh et al., 2020). While several PIPs have been characterized as transport channels for specific substrates in biological model plants like Arabidopsis and cereal crops like barley (Fox et al., 2017;Sadura et al., 2020), only three OsPIPs were studied with respect to substrate selectivity in their source plant (rice) ( Table 1). OsPIP1;2 was recently shown to support mesophyll CO 2 transport and phloem sucrose transport (Xu et al., 2019). This protein was identified as a physiologically relevant facilitator of CO 2 transport across PMs of rice cells based on gas exchange measurements performed on leaves of the wild-type (WT) and OsPIP1;2overexpressing rice plants (Xu et al., 2019 (Ding et al., 2019). In contrast, substrate selectivity of OsPIP1;1, 1;3, 2;2, 2;3, 2;5, 2;6, and 2;8 was tested in yeast but not in their source plant (Table 1).
We have extensively investigated plant PIPs with a main attempt to elucidate their implications in plant growth-defense tradeoffs . Emerging evidence demonstrates that several PIPs have the dual functions in both physiological and pathological processes, not only regulating plant growth and development  but also partaking in plant infection by and immunity against pathogens Ji & Dong, 2015;Li et al., 2019;Tian et al., 2016;Wang, Schoebel, et al., 2020;Wang, Zhang, et al., 2020;Zhang, Hu, et al., 2019). We compared OsPIPs in terms of physiological performances in response to the bacterial blight pathogen Xanthomonas oryzae pv. oryzae (Xoo) when it was infecting plants of the susceptible rice variety Nipponbare (Ji & Dong, 2015a;Wang, Zhang, et al., 2018;Li et al., 2019). We found that OsPIP1;4 can be used by Xoo to translocate a virulent effector from the bacteria into rice cells, wherein the bacterial effector activates a disease-susceptible gene to induce virulence in Nipponbare plants (Wang, Zhang, et al., 2018;Bian et al., 2019;Li et al., 2019;Zhang, Hu, et al., 2019). This finding confirms the recently proposed conceptional revolution that functions of AQPs far exceed the originally defined category only for substrate transportation (Hara-Chikuma et al., 2015;Afzal et al., 2016;Ha, 2017;Li et al., 2019). However, it was unclear whether the OsPIPs that associate with plant infection and immunity have a primary role for substrate transport in rice plants growing under regular conditions without pathogen infection.
In the present study, we analyzed OsPIPs with respect to their presence or absence in H 2 O transport across living cell PMs according to the original definition of "water channels" (Agre, 2004;Agre et al., 1993;Brown, 2017;Preston et al., 1992). We further determined the role of OsPIP2;2 in the plant resistance to a physiological drought stress caused by Polyethylene glycol 6,000 (PEG 6000 ). PEG with molecular mass 4,000-8,000 effectively induces physiological drought in a variety of plant species, and PEG 6000 has been mostly used (Arisha et al., 2020;Dong et al., 2005;Hajihashemi & Geuns, 2016;Huang et al., 2019;Ren et al., 2019;Tiwari et al., 2020;Zhang et al., 2017). While PEG induces drought syndromes from leaf wilting to plant collapse, plants in this process may defend themselves by activating the drought tolerance pathway to preserve water relations (Dong et al., 2005;Zhang et al., 2017). By cytological, physiological, and molecular assays, we show that OsPIP2;2 is a predominant PEG 6000 was used as an inducer of physiological drought stress (Dong et al., 2005;Dubois & Inzé, 2020) and applied to 15-day-old rice seedlings. Usually, PEG 6000 is prepared as an aqueous solution at a certain percent concentration (w/v) and applied in parallel to the deionized water control to incubate plants by immersing the root system. In this study, 15-day-old rice seedlings growing in the PINDSTRUP soil were taken out carefully by pushing the soil away using a finger, cleaned gently by soft washing with tap water to remove the surrounding soil scraps, and then placed into plastic tubes containing deionized water. After a 12-h acclimation, seedlings were moved into new tubes containing deionized water (control) or an aqueous solution of PEG 6000 justified to 0%, 5%, 10%, 15%, 20%, 25%, and 30% w/v, respectively. Seedlings were monitored in the subsequent 10 h by automatic photography at 10-min intervals. Chronological development of drought syndromes was analyzed.

| Quantification of OsPIP expression levels
Total RNA was isolated from leaves and subjected to real-time quantitative reverse polymerase chain reaction (qRT-PCR). All the qRT-PCR experiments were performed with the QuantStudio ® 3 Real-Time PCR Instrument (ThermoFisher Scientific) and using the constitutively expressed EF1α gene as a reference (Liu et al., 2011). Relative expression level of a tested gene was quantified as the tested gene to EF1α transcript quantity ratio, which was determined by the 2 ÀΔΔC T method (Livak & Schmittgen, 2011). 2.3 | OsPIP de novo expression

| Expression in oocytes
Based on gene expression analyses (Figure 1), OsPIP1;3, OsPIP2;2, OsPIP2;2, and OsPIP2;4 were further investigated. Complementary F I G U R E 1 Chronological changes in rice response to physiological drought stress caused by PEG 6000 applied in a range of concentration. Fifteen-day-old rice seedlings were incubated by root immersion in deionized water containing PEG 6000 added at the indicated concentrations. Images were obtained by automatic photography in 10 h at 10-min intervals and are partially shown here on the hour except one image in which seedlings incubated with the highest concentration of PEG 6000 had displayed drought syndromes. Arrowheads point the concentrations that already induced drought syndromes. Each image represents 105 plants treated proportionally with each of the 7 PEG 6000 doses and tested in 3 independent experiments DNAs (cDNAs) of OsPIPs were obtained by RT-PCR amplification of total RNAs isolated from leaves of 15-day-old rice plants. For use in transformation of African clawed frog (Xenopus laevis) oocytes, every OsPIP cDNA was fused to the enhanced green-fluorescent protein (eGFP) gene or a six-his tandem sequence by conventional recombination methods  between 495 and 520 nm using 488-nm argon-ion laser excitation with a Zeiss LSM700 laser scanning confocal microscope. Fusion proteins were analyzed by immunoblotting of membrane proteins isolated from the transformed oocytes using a previously described protocol (Jørgensen et al., 2016). Similar procedure was used in construction, de novo expression, and analyses of the OsPIP-his fusion protein except for laser scanning confocal microscopy (LSCM).

| Expression in yeast
Each of the OsPIP-his fusion genes was cloned into the yeast binary vector pYES2 (Wang, Zhang, et al., 2020) with the help of restriction enzymes HindIII and EcoRI (TAKARA). The recombinant vector was transformed into competent cells of the wine-brewing yeast (Saccharomyces cerevisiae) strain NMY51 in transformation solution (.1 M LiCl, 30% w/v PEG 4000 ). The positive transformants were cultured in SD-Ura (synthetic dextrose minimal media without uracil) media with 2% w/v galactose overnight at 30 C and harvested in phosphate buffer solution (.2 mM, pH 7.4). Membrane proteins were isolated from the transformed yeast cultures using a previously described protocol (Hansen et al., 2017) and analyzed by immunoblotting. Yeast cells were observed between 495 and 520 nm using 488-nm argon-ion laser excitation with a Zeiss LSM700 laser scanning confocal microscope.

| Expression in tobacco plants
De novo expression of the selected OsPIPs (OsPIP1;3, OsPIP2;1, OsPIP2;2, and OsPIP2;4) in leaves of tobacco was performed by the agroinoculation method (Liu et al., 2011). Each of the OsPIP genes was combined at the N-terminus with the constitutive 35S promoter from cauliflower mosaic virus and at the C-terminus with the YEL-LOW-FLUORESCENT PROTEIN (YFP) gene from a previously used yeast vector . The fusion gene was inserted into the plant binary vector pCAMBIA1301 (Liu et al., 2011

| Expression in plant protoplasts
This was performed by the chemical mediation method (Shen et al., 2014). Protoplasts were isolated from strips of fully expanded leaves of 30-day-old tobacco seedlings or from stems and leaf sheaths of 30-day-old rice plants, using reagents from Sigma-Aldrich and a previously described protocol (Shen et al., 2014;Yoo et al., 2007).

| Water permeability measurements
Water permeability of the directly transformed rice and tobacco protoplasts, protoplasts from the transformed oocytes and yeast, and protoplasts from the OsPIP2;2-silenced and -overexpressing rice plants was measured by microscopy. These protoplasts were separately incubated in the ND96 solution, which provides a low osmotic gradient from exteriors to interiors of the incubated protoplasts. Ten minutes later, the osmotic water permeability coefficient (Pf) was determined by measuring volume changes of the protoplasts as previously described (Ding et al., 2019;Li et al., 2015).

| Proline measurement
Proline concentrations in tobacco and rice protoplasts were determined with a plate reader. Before testing, a standard curve was established using a commercial proline standard of known concentrations and the corresponding absorbance readings. To isolate proline from protoplasts, 5 ml of 3% sulfosalicylic acid solution was added into a tube containing a protoplast suspension and then the sample was setting for 10 min in a boiling water bath. The extraction solution was filtered into a clean test tube after cooling and the supernatant was regarded as a proline extract. This proline extract solution was mixed with equal volumes of glacial acetic acid and acid ninhydrin. After the mixture was phased automatically, solution of the upper phase was collected and centrifuged at 3,500g for 5 min.
Supernatant was assayed for absorbance at 520 nm in the plate reader. Proline concentration in the supernatant was estimated with reference to gradients of the absorbance by methylbenzene. The proline content was given as micrograms per gram protoplasts.

| Polyamine measurement
A previously described protocol (Zhu et al., 2020) was used to extract polyamines (PAs) from rice and tobacco protoplasts, respectively. Soluble conjugated PAs were calculated by subtracting the free PAs from the acid-soluble PAs. Precipitated protoplasts were suspended with 4 ml of v/v 5% cold perchloric acid (PCA) and incubated at 4 C for 1 h, followed by sonication in the presence of lyase and antifoam. The resulting suspension was centrifuged at 12,000g and 4 C for 10 min.
The supernatant was supplied with the internal standard 1,6-hexanediamine, and the mixture was centrifuge at 12,000 g and 4 C for 30 min. The resulting supernatant was blended with 6-N HCl at a 1:5 volume rate and hydrolyzed at 110 C for 18 h in flame-sealed glass ampules. After acid hydrolysis, HCl was evaporated by heating at 70 C, and the residue was suspended in 2 ml of 5% PCA, followed by centrifugation at 12,000g and 4 C for 30 min. The supernatant contained the acid-soluble PA fraction and free Pas liberated from PA conjugates. To obtain the insoluble bound PA, the pellets were rinsed four times with 5% PCA to remove any traces of soluble PAs and were suspended in 5 ml of 6 N HCl. This solution was hydrolyzed by the same procedure as described above. PAs recovered from the nonhydrolyzed supernatant, hydrolyzed supernatant, and the pellet were benzoylated as follows. An aliquot of the supernatant was treated with 2 ml of 2 N NaOH and 15 ml of benzoyl chloride, vortexed vigorously, and incubated for 30 min at 37 C. The reaction was terminated by adding 4-ml saturated NaCl solution. Thereafter, the benzoyl PAs were extracted with 3-ml cold diethyl ether. Finally, 1.5 ml of the ether phase was evaporated to dryness and redissolved in 1-ml methanol. PAs in the final solution were assayed by high-performance liquid chromatography (HPLC) (Agilent 1220, America).

| Data treatment
All experiments were repeated at least three times with similar results.
Quantitative data were analyzed using the commercial IBM SPSS19.0 software package (Shi, 2012). Homogeneity-of-variance in data was determined by the Levene test. The formal distribution pattern of the data was confirmed by the Kolmogorov-Smirnov test and P-P Plots.
Fisher's data-pairing test or Duncan's new multiple range test was performed along with analysis of variance (ANOVA) on data from at least three independent experiments, each involving three repetitions, unless specified elsewhere, such as when a leaf was treated as a statistical unit.

| RESULTS
3.1 | OsPIP2;2 highly responds to a physiological drought stress To identify OsPIP candidates implicated in water relations of rice, we analyzed expression levels of 11 OsPIP genes in plants of the rice variety Nipponbare following PEG 6000 treatment (Figure 1). We applied a range of PEG 6000 concentrations, 0%-30% at 5% gradients, to 15-day-old rice seedlings in liquid culture, and monitored F I G U R E 2 Expression levels of OsPIPs in rice plants temporally incubated in deionized water with and without supply with PEG 6000 . Rice seedlings already growing for 15 days in pot soil in a plant growth chamber were shifted into tubes containing deionized water. After 12-h acclimation, these plants were transferred into new tubes containing deionized water only or an aqueous solution of PEG 6000 at the indicated concentration. Insets show morphological changes of seedlings growing under the indicated conditions. Gene expression was analyzed by QRT-PCR performed with total RNAs isolated from the aerial parts of the plants 3 h after incubation. The constitutively expressed OsEF1α gene was used as a reference to quantify relative expression level of an OsPIP. Data show are mean values AE standard deviation (SD) estimates. The number of experimental repeats (n) = 6 independent experiments each involving 15 plants tested in three biological repeats. Different letters on graphs indicate significant differences by analysis of variance (ANOVA) and Fisher's test between the pairs of data (P = .001-1.2 Â 10 À7 ) chronological development of drought syndromes in the subsequent 10 h (Figure 1; Table 2). In that period, the lowest PEG 6000 dosage (5%) caused no responses, but drought syndromes were induced when PEG 6000 concentration was increased to 10% and higher ( Figure 1). Leaf lengthening and narrowing (Figure 2, insets) were found to be the first sign of drought syndromes, occurred mostly on the second leaves (Figure 2, insets) in 50-110 min after PEG6000 application at 10%-30% (Table 2). The formation of slender leaves was thought to indicate plant transition from the normal physiological status to drought response. This transition was followed by wilt of partial to all leaves and plant collapse in the end (Figures 1 and 2). All these syndromes occurred in 6 h after plant incubation with 25% and 30% PEG 6000 supplies, respectively (Table 2), but concentrations of 15% and 10% only caused partial leaves to wilt.
To evaluate the effects of the physiological stress on OsPIP expression, an aqueous solution of 15% or 30% PEG 6000 was applied to 15-day-old rice seedlings by immersing roots. These plants (+ PEG 6000 ) and those incubated in deionized water (ÀPEG 6000 ) were used 3 days later in OsPIP expression assessments by qRT-PCR. The qRT-PCR analyses using total RNA isolated from the leaves indicated that all OsPIPs responded to the physiological drought stress, displaying significantly (P = 1.2 Â 10 7 ) enhanced expression in contrast to the steady-state expression levels in control (Figure 2, graph).
Evidently, OsPIP2;2 exceeds all the other OsPIP genes in response to the physiological drought stress and displays the highest level of induced expression under the stress condition, indicating that OsPIP2;2 is likely to take part in water relations of rice.
3.2 | De novo expression of OsPIP2;2 enhances water permeability of oocyte and yeast membranes
Thus, the transformed protoplasts were incubated with 10% PEG 6000 , and proline and PA concentrations were measured 10 min later. At this time point, protoplast concentrations of both proline (Figure 6c)  (Figure 6c,d).
In addition to the physiological responses, cytological variations also occur in plants under a drought stress (Dong et al., 2004;Oladosu et al., 2019). We confirmed that the PEG 6000 treatment 3.6 | OsPIP2;2 conduces to membrane integrity of rice protoplasts In plants, one of PEG-induced responses is cellular ion efflux known as electrolyte leakage, which occurs due to injured integrity of the PMs (Dong et al., 2005). We analyzed whether the role of OsPIP2;2 in reducing rice protoplast collapse (Figure 8b) relates to its effect on electrolyte leakage from the protoplasts in response to PEG 6000 .
When the WT, OsPIP2;2-Si, and OsPIP2;2-OE plants incubated in deionized water were supplied with 30% PEG 6000 and measured 3 h later, electrolyte leakage from leaves was detected at different extents. Compared to the WT, the OsPIP2;2-OE plants had much less electrolyte leakage, but electrolyte leakage from OsPIP2;2-Si plants was significantly increased (Figure 9a). Conveying the conductivity values to thousand times of their reciprocals reflects integrity degrees of the protoplast membranes (Figure 9b), suggesting that OsPIP2;2 is conducive to maintenance of the membrane integrity in response to the physiological drought stress.
3.7 | OsPIP2;2 imparts activation of the drought tolerance pathway In rice, activation of the drought tolerance pathway by PEG treatment essentially involves induced expression of the pathway-marker gene COR413-TM1 and the pathway-regulatory genes OsABF1, OsPP48, and OsPP108 (Zhang et al., 2017). We determined that these genes were induced by PEG 6000 treatment to markedly express in leaves of the WT, OsPIP2;2-Si, and OsPIP2;2-OE plants ( Figure 10). As F I G U R E 9 The effects of OsPIP2;2 silencing and overexpression on cell membrane integrity of 15-day-old rice seedlings growing in the absence (À) and presence (+) of PEG 6000 treatment. OsPIP2;2-Si was considerably inferior to the WT in supporting PEG 6000 -induced expression of all the genes ( Figure 10). Clearly, OsPIP2;2 imparts the drought tolerance pathway toward activation in response to the physiological drought stress.

| DISCUSSION
Rice has a total of 33 AQPs including 11 PIP homologs (Sakurai et al., 2005) that mostly have not been characterized regarding substrate selectivity and specificities, as well as associated bioprocesses ( Previously, OsPIP2;1 was identified as a physiologically relevant CO 2 -transporting implementor in rice (Xu et al., 2019). In rice, moreover, both OsPIP1;2 (Ding et al., 2019) and OsPIP2;4 (Nada & Abogadallah, 2020) were characterized to support root hydraulic conductivity. It is unclear whether OsPIP1;2 and OsPIP2;4 also contribute to leaf hydraulic conductivity, which can be detected by cell pressure probing measurements (Hachez et al., 2011) as reliable as for the root . No evidence has been obtained to elucidate whether F I G U R E 1 0 The effects of OsPIP2;2 silencing and overexpression on expression of response genes regarded as molecular makers of the drought tolerance pathway. Rice seedlings already growing for 15 days in pot soil in a plant growth chamber were shifted into tubes containing deionized water. After 12-h acclimation, these plants were transferred into new tubes containing deionized water only or an aqueous solution of 30% PEG 6000 . Three hours later, gene expression was analyzed by QRT-PCR performed with total RNAs isolated from the aerial parts of the plants. The constitutively expressed OsEF1α gene was used as a reference to quantify relative expression level of an OsPIP. Data show are mean values AE SD estimates. Different letters on graphs indicate significant differences by analysis of variance (ANOVA) and Duncan's multiple new multiple range test of the data (P = .0001-1.7 Â 10 À11 ; n = 6 independent experiments each involving 15 plants tested in three biological repeats) an OsPIP functions for substrate transport in both roots and leaves, but it is frequent that a particular AQP plays a same role in different organs of plants (Gomes et al., 2009;Li et al., 2015;Maurel et al., 2008 (Gomes et al., 2009;Maurel et al., 2008;Péret, et al., 2012). Our results also agree with previous studies by Sakurai and colleagues (Sakurai et al., 2005(Sakurai et al., , 2008. They did not test substrate-importing functions of any OsPIPs in rice. Instead, they analyzed differential expression of 33 AQP-encoding genes including 11 OsPIPs, which constitute the full repertoire of AQPs in the rice genome, in different F I G U R E 1 1 Model of OsPIP2;2 functions in H 2 O transport and drought tolerance. When plants are growing regularly without drought stress, OsPIP2;2 functions to facilitate H 2 O transport in and out of the plant cells in response to a hydraulic gradient generated by natural metabolism in the apoplastic or cytoplasmic space. This function is assumed to play a role in maintenance of water relations. When rice plants incur osmotic stress from environment, such as the physiological drought stress caused by externally applied PEG, OsPIP2;2 turns to function to support drought tolerance possibly by physical and physiological regulatory mechanisms. In the assumed physical mechanism, the drought stress is inevitable to injure the cell membranes, causing electrolyte leakage for example, while the presence of OsPIP2;2 serves as an encountering force to help preserve the membrane integrity. The physiological mechanism, including increases in proline and polyamine concentrations, is used by OsPIP2;2 to maintain the cellular water homeostasis. Water homeostasis may also come from the role of OsPIP2;2 in modulating H 2 O transportation shuttles in and out of the cells, instead of a single direction, depending on hydraulic gradient changes by electrolyte leakage. Both physical and physiological mechanisms could be integrated to increase drought tolerance intensity. Abscisic acid (ABA) signaling may partake in the regulation of OsPIP2;2-mediated drought tolerance, which involves the ABA-responsive transcription factor COR413-TM1. The OsPIP2;2-dependent tolerance pathway is also likely to have a crosstalk with signaling by H 2 O 2 if it is induced by a particular stimulus (plant infection by a plant pathogen, for example) to accumulate in plant apoplastic spaces and to be transported by OsPIP2;2 to enter the cytoplasm organs of rice plants after chilling treatment (Sakurai et al., 2005). including increases in proline and PA concentrations (Dong et al., 2004;Xu et al., 2019;Zhu et al., 2020), which turn to enhance drought resistance.
Based on these analyses, we propose a model that hypothesizes how OsPIP2;2 execute its functions toward H 2 O transport and drought tolerance (Figure 11). In the model, OsPIP2;2 supports drought tolerance by physical and physiological regulatory mechanisms, which modulate the membrane integrity by controlling the molecular (H 2 O and electrolyte) trafficking and by regulating the intracellular responses against drought stress, respectively ( Figure 11).
In the tested drought-tolerant constituents, OsABF1 is a basic leucine zipper transcription factor that functions in plant responses to biotic stresses through the abscisic acid (ABA) signaling pathway.
Presumably, the OsPIP2;2-dependent drought tolerance is also subject to ABA signaling (Figure 11), which has been demonstrated to regulate plant tolerance to PEG-induced physiological stress (Dong et al., 2005). Furthermore, characterizing the possible role of OsPIP2;2 in H 2 O 2 transport will better explain the molecular basis of the membrane integrity. We recently proposed that H 2 O-transporting AQPs are also potential channels for H 2 O 2 trafficking (Wang, Schoebel, et al., 2020). In Arabidopsis, AtPIP1;4 plays triple roles in CO 2 , H 2 O , and H 2 O 2 transport across the PMs. H 2 O 2 is generated in the apoplast upon induction by pathogen infection, transported by AtPIP1;4 into the cytoplasm, and therefore participates in immunity signal transduction (Tian et al., 2016). If OsPIP2;2 has a role in H 2 O 2 transport, it may contribute to the membrane integrity by facilitating the stress-induced apoplastic H 2 O 2 into the cytoplasm and therefore reducing the oxidative pressure toward the cell membranes ( Figure 11). Verification of these hypotheses will be the subject for further studies.
In conclusion, the physiological, molecular, and cytological perfor-