Grapevine Aquaporins: Gating of a Tonoplast Intrinsic Protein (TIP2;1) by Cytosolic pH

Grapevine (Vitis vinifera L.) is one of the oldest and most important perennial crops being considered as a fruit ligneous tree model system in which the water status appears crucial for high fruit and wine quality, controlling productivity and alcohol level. V. vinifera genome contains 28 genes coding for aquaporins, which acting in a concerted and regulated manner appear relevant for plant withstanding extremely unfavorable drought conditions essential for the quality of berries and wine. Several Vv aquaporins have been reported to be expressed in roots, shoots, berries and leaves with clear cultivar differences in their expression level, making their in vivo biochemical characterization a difficult task. In this work V. vinifera cv. Touriga nacional VvTnPIP1;1, VvTnPIP2;2 and VvTnTIP2;1 were expressed in yeast and water transport activity was characterized in intact cells of the transformants. The three aquaporins were localized in the yeast plasma membrane but only VvTnTIP2;1 expression enhanced the water permeability with a concomitant decrease of the activation energy of water transport. Acidification of yeast cytosol resulted in loss of VvTnTIP2;1 activity. Sequence analysis revealed the presence of a His131 residue, unusual in TIPs. By site directed mutagenesis, replacement of this residue by aspartic acid or alanine resulted in loss of pHin dependence while replacement by lysine resulted in total loss of activity. In addition to characterization of VvTn aquaporins, these results shed light on the gating of a specific tonoplast aquaporin by cytosolic pH.


Introduction
Plant growth and development are dependent on the tight regulation of water uptake and transport across cellular membranes and tissues at whole plant level. Depending on the environmental conditions and water balance, plants can modify the relative contribution of apoplastic and cell-to-cell water-flow pathways across the tissues to adjust the overall hydraulic conductivity [1]. Aquaporins are essential in the cell-to-cell pathway as their presence allows not only higher water permeability but also control and regulation of water flow [1], providing a fine tuning of the hydraulic conductivity of this pathway in response to biotic and abiotic stresses. With 28 members in Vitis vinifera [2], plants appear to express a much higher and diverse number of aquaporin homologues than mammalian or yeast. These proteins belong to the family of Major Intrinsic Proteins (MIPs). Based on their sequence similarity and on their main subcellular location plant aquaporins were organized in seven subfamilies: the plasma membrane intrinsic proteins (PIPs), the tonoplast intrinsic proteins (TIPs), the nodulin26-like intrinsic proteins (NIPs), the small and basic intrinsic proteins (SIPs), the GlpF-like intrinsic proteins (GIPs), the hybrid intrinsic proteins (HIPs) and the uncategorized X intrinsic proteins (XIPs) [3]. However, individual aquaporins cannot easily be assigned to homogeneous subcellular compartments, being reported that each organelle may be equipped with combinations of different isoforms and that, although predominant in one cellular localization, the same subfamily can be expressed differently according to plant tissue. This diversity suggests a putative role of these channels in different cell types or tissues for survival and development upon a wide range of conditions, but their specific physiological relevance still remains poorly understood [4]. Furthermore, aquaporin activity can be regulated by gating factors or mechanisms [5] such as osmotic solutes [6], pressure pulses [7], membrane tension [8,9], cytosolic pH and pCa 2+ [10,11] and post translational mechanisms like phosphorylation [12][13][14][15], glycosylation [16], methylation and acetylation [17] and heterotetramer formation [18]. Several stress factors were reported to affect plant aquaporin activity such as oxidative stress [19], NaCl [20], dehydration and excessive watering and chilling [21].
Grapevine (Vitis vinifera L.) is one of the oldest and most important perennial crops being considered as a fruit ligneous tree model system [22] in which the water status appears crucial for high fruit and wine quality, controlling productivity and alcohol level [23]. Furthermore, flooding of soils results in oxygen deprivation (anoxia) of plant roots during raining season, or after irrigation. One early response of plants to anoxia and other environmental stresses is the inhibition of root water permeability leading to a down-regulation of water uptake [14]. It was also reported that flooding induced anoxia led to reduced water uptake in roots. It was proposed that intracellular acidification, arising as a consequence of anoxia, would mediate inhibition of water channels [24]. Direct evidence for pH regulation was observed in Arabidopsis PIPs heterologously expressed in Xenopus laevis oocytes [14,25]. Moreover, it was shown that replacement of a histidine residue by an alanine or aspartic acid in loop D at position 197 prevented pH sensitivity [14,24]. V. vinifera PIPs (Chardonnay and Grenache cultivars) have shown water transport activity when individually expressed in Xenopus [21]. Several VvPIPs and VvTIPs have been reported to be expressed in roots, shoots, berries and leaves [2,26,27] with clear cultivar differences in their expression level [21], making their in vivo biochemical characterization a difficult task.
Due to its intrinsic low water plasma membrane permeability, yeast has been reported as a suitable system for heterologous individual aquaporin expression [28]. Moreover, the activity of heterologous water channels anchored in the membrane of a aqynull strain can be powerfully characterized by stopped flow spectroscopy. Yeasts can be maintained in a diversity of environments and even the composition of the interior of the cell can be controlled, creating a variety of conditions to study the specificity and regulation of putative water channels. The use of a fluorescence self-quenching methodology for assessing water transport in intact cells (without removing the cell wall) [29] represents an increased value for the study of yeast expressed plant aquaporins. This methodology opens new perspectives to measure water permeability in minimally disturbed cells that are quite stable during a rather long experiment [29].
In the present work, we describe the molecular cloning process of two VvPIPs (VvTnPIP1;1 and VvTnPIP2;2) and one VvTIP (VvTnTIP2;1) from V. vinifera cv. Touriga nacional and their characterization using the Saccharomyces cerevisiae heterologous expression system. The cultivar Touriga nacional was chosen as representative of Portuguese traditional cultivars with an important recent commercial value and deserving much interest from enologists throughout the world. Moreover, the cellular localization of each aquaporin was identified in yeast and the osmotic stress tolerance of these strains was evaluated. Finally, functional characterization (water permeability and activation energy of water transport) of VvTnTIP2;1 was performed, focusing on a putative aquaporin intracellular pH regulation mechanism.

Results
Cloning and molecular characterization of V. vinifera PIP1;1, PIP2;2 and TIP2;1 Oligonucleotide primers designed from 59UTR and 39UTR of V. vinifera cv. Cabernet sauvignon aquaporin PIP1;1, PIP2;2 and TIP2;1 sequences, available on the Grape Genome database from Genoscope, were used for amplification of aquaporin cDNAs from mRNAs extracted from V. vinifera cv. Touriga nacional callus cells. This procedure led to the isolation of four full length cDNAs encoding the corresponding putative aquaporins. A multiple alignment of the deduced amino acid sequences is shown in Figure S1. All the deduced sequences present the characteristic topology of six highly hydrophobic transmembrane spanning helices connected by five loops with the N-and C-terminal sequences on the cytosolic side of the membrane. The highly conserved SGxHxNPA sequence in the first half of the protein and the second highly conserved region in the second half of the protein, forming part of the MIP family signature, are also present [30].
The genes corresponding to the cDNAs cloned in this study were clearly identified as encoding two proteins belonging to the PIP subfamily and two belonging to the TIP subfamily. Each of the PIP aquaporins showed an obvious and specific resemblance with PIP1 and PIP2 V. vinifera sequences subfamily ( Figure S1). Accordingly, the phylogenetic analysis of these sequences clearly showed the consistent distribution of V. vinifera cv. Touriga nacional aquaporins, clustered within the group of V. vinifera PIP1, PIP2 and TIP2 aquaporins ( Figure S2). The nomenclature proposed for the grapevine aquaporins was established according to the results of multiple alignments with MIP genes from V. vinifera cv. Cabernet sauvignon, Pinot noir and Syrah ( Figure S2).
The comparative analysis of VvTnPIP1;1, VvTnPIP2;2 and VvTnTIP2;1 amino acid sequences from Touriga nacional cultivar with other similar proteins from other V. vinifera cultivars was also performed. The aquaporin sequences were mostly identical to all the similar proteins, excepting that a V276M substitution was found in VvTnPIP1;1 and that a M235I substitution was found in VvTnTIP2;1, both localized at the C-terminal region ( Figure S3).
Search for signatures characteristic of atypical substrates transport, as proposed by Hove et al. [31], was also performed. Interestingly, VvTnTIP2;1 sequence revealed all the characteristic residues indicating that besides water it may also permeate ammonia. As for VvTnPIP1;1 and VvTnPIP2;2, both present all the key residues of MIPs for H 2 O 2 transport.
To characterize the three putative V. vinifera aquaporins, plasmids harboring the corresponding ORFs, driven by the MET25 promoter and tagged with the GFP sequence at their 39 end were constructed ( Table 1). The plasmids were retrieved, their sequences confirmed by restriction analysis and sequencing and were used to transform S. cerevisiae aqy-null strain. For negative control, the same strain was transformed with the empty plasmid pUG35. Transformants containing VvTnPIP1;1 (LL0.P11), VvTnPIP2;2 (LL0.P22), and VvTnTIP2;1 (LL0.T21) were obtained. An additional mutated VvTnTIP2;1 in which glutamic acid E146 was replaced by a glycine residue was also cloned (LL0.T21.E146G). Representative colonies from each clone were used for heterologous expression studies.
GFP-tagging confirmed the plasma membrane localization of VvTnPIP1;1, VvTnPIP2;2, VvTnTIP2;1 proteins ( Figure 1), and the presence of similar amounts of the tagged proteins. Membrane localization and expression level was also confirmed for the mutated version of VvTnTIP2;1 harbored by LL0.T21.E146G (data not shown).
The regulation of expression levels by MET25 promoter was tested. The strains were grown in YNB supplemented with methionine and the fluorescence resulting from the presence of the aquaporin-GFP tagged proteins was drastically reduced, confirming the repression of gene expression under the control of MET25 promoter ( Figure 1).

V. vinifera PIP and TIP expression affects yeast growth under osmotic stress
The expression of the aquaporin genes inserted into pUG35 plasmid was not toxic to yeast, since the growth of all strains was similar to the strain transformed with the empty plasmid (LL0) both in YNB solid ( Figure 2) and liquid medium (data not shown).
Although the relationship between water permeation and osmotic stress is still not clear, previous reports show that the absence of water channels in yeast improved resistance to osmotic stress [32]. These results led us to examine the effect of plant aquaporin expression on yeast osmotic stress tolerance. The osmotic stress tolerance of strains expressing VvTnPIP1;1, VvTnPIP2;2 and VvTnTIP2;1 was examined.  Figure 2B). The behavior of LL0.T21.E146G, harboring the mutated version of VvTnTIP2;1 was similar to that of LL0. In general, the parallel effect of sorbitol was less evident.

Functional assessment of water transport
To analyze aquaporin activity, yeast strains expressing V. vinifera water channels were loaded with CFDA and challenged with a hypertonic sorbitol solution in a stopped flow device. The change in the fluorescence signal due to water efflux was used to calculate the osmotic permeability coefficient (P f ) and the activation energy (E a ) of water transport. As shown in Figure 3, the expression of VvTnTIP2;1 (LL0.T21) led to a large increase of the shrinking rate as compared to the non-expressing strain (LL0). However, expression of VvTnPIP1;1 (LL0.P11) and VvTnPIP2;2 (LL0.P22, not shown) did not affect the time course of water efflux. Figure 4 compares the P f values obtained at 9uC and the E a for the different strains. The enhance in water permeability conferred by VvTnTIP2;1 expression is clearly depicted (8.060.76 10 24 cm s 21 ), while no activity for VvTnPIP1;1 (1.060.26 10 24 cm s 21 ) nor for VvTnPIP2;2 (0.960.2610 24 cm s 21 ) expressing strains could be detected when compared to the LL0 strain (0.960.2610 24 cm s 21 ). The increase in permeability is consistent with the decrease in E a for the LL0.T21 strain (14.060.9 kcalmol 21 (58.263.7 kJ mol 21 ) for LL0 and 8.360.2 kcal mol 21 (34.760.8 kJ mol 21 ) for LL0.T21), therefore assuring that the pathway for water flow is being affected due to an increase in aquaporin activity. When methionine was added to the growth medium repressing the gene expression, the activity of VvTnTIP2;1 was diminished in the LL0.T21 strain (P f of 1.560.4610 24 cm s 21 ) with a concomitant increase in the E a (12.760.4 kcal mol 21  Cytosolic pH regulates V. vinifera TIP activity Previous reports indicate that some mammalian aquaporins are regulated by external pH [33] and that plants respond to conditions that lead to a decrease in cytosolic pH with reduction of water membrane permeability [34]. Also, the direct evidence for pH regulation of A. thaliana and V. vinifera PIP aquaporins heterologously expressed in Xenopus oocytes [14,25] led us to study pH regulation of V. vinifera cv. Touriga nacional water channels when cloned in yeast. The intracellular pH of yeast transformants was determined as indicated in Methods. An external acidification from pH 6.8 to 5.0 induced an intracellular pH (pH in ) decrease from 6.8 to 6.1. When cells were incubated with 4 mM benzoic acid at pH 5.0, a further marked drop in pH in from 6.1 to 4.8 was observed.
Acidification of the yeast cytosol reduced water permeability in the LL0.T21 strain as shown in Figure 5. The 10-fold higher P f found for LL0.T21 (10.761.3610 24 cm s 21 ) compared to LL0 (1.360.04610 24 cm s 21 ), decreased when pH in dropped from 6.8 to 6.1 (7.960.6610 24 cm s 21 ). At pH in 4.8, a further decrease in activity was detected (5.260.7610 24 cm s 21 ). This observed ca. 50% reduction of permeability was consistent with the ca. 50% increase in the E a values for the same experimental acidic conditions as shown in Table 2. The improved aquaporin activity of LL0.T21 strain was repressed by methionine incubation, reaching values similar to the negative controls at all pH in tested. Strains LL0 and the mutated LL0.T21E146G containing a non-  From these data we conclude that pH in rather than pH out regulates aquaporin activity, since the addition of 4 mM benzoic acid at pH out 5.0, changing pH in without changing pH out , resulted  in a decreased permeability. At pH 6.8, benzoic acid (8 mM) had no effect on pH in and no effect on permeability (not shown) supporting the idea that pH in and not benzoic acid itself affected aquaporin activity.

Role of histidine 131 residue in TIP activity
VvTnTIP2 sequences possess a histidine residue (His 131 ) in loop D, common in PIPs (His 193 in spinach, His 196 in tobacco, His 199 in strawberry plant and His 194 in A. thaliana aquaporin PIP2;1) but not in TIPs from other plants. Such histidine residues were reported to have a key function in pH-dependent aquaporin gating [14,15,24]. The presence of the His 131 residue at loop D of VvTnTIP2;1 together with a leucine in the following vicinity, as described for plant PIPs [35], led us to look for a possible role of this histidine in the pH-dependent regulation of VvTnTIP2;1 aquaporin activity. We introduced point mutations to replace His 131 by Asp (H131D), Ala (H131A), and Lys (H131K), and after confirming the membrane localization of the mutated VvTnTIP2;1 proteins, water transport activity was evaluated under conditions where yeast pH in was adjusted as indicated above.
Fast water uptake was observed when the variant H131D was expressed; however, aquaporin activity was not pH dependent (P f of (7.861.3), (5.761.4) and (6.561.0)610 24 cm s 21 for pH in 6.8, 6.1 and 4.8, respectively) ( Figure 5). The variant H131A showed a poor pH dependent activity compared to the wild-type protein (P f of (8.161.1), (7.060.9) and (5.861.0)610 24 cm s 21 for pH in 6.8, 6.1 and 4.8, respectively). The variant H131K lost the water transport ability through the mutated aquaporin for every pH tested (P f of (1.360.1), (1.060.1) and (1.060.2)610 24 cm s 21 for pH in 6.8, 6.1 and 4.8, respectively). These results were corroborated by a consistent pattern of E a values ( Table 2).

Discussion
Vitis vinifera, the grapevine originated in the eastern Mediterranean area, is generally cultivated in soils where water is the main limiting factor in agriculture, under conditions of drought and salinity stress. Although the adaptation of this plant to dry climates is well known, the molecular bases of water transport in grapevine still need further elucidation. Many genes encoding aquaporins were isolated from different plants including ligno-cellulosic species [22]. The recent release of Vitis vinifera genome allowed the identification of 28 putative aquaporin genes among which 8 coding for PIPs and 10 coding for TIPs have been found [2]. However, only the expression of a single or/and combination of each aquaporin isoforms in heterologous systems unequivocally allows the identification of its individual properties [36].
In this study we have cloned and individually characterized two VvTnPIPs and one VvTnTIP by heterologous expression in an aqynull yeast strain. Their sequences show all the typical characteristics of the corresponding sub-family for water permeation and additionally some other key residues that suggest permeation of other putative substrates [31]. Therefore, their specificity for atypical substrates should be further investigated.
The osmotic water permeabilities were assessed in intact yeast cells expressing these aquaporins by stopped-flow fluorescence using a volume sensitive dye. This technique enables the assessment of water transport in yeast cells with intact physiological functions of the cell wall, facilitating the biochemical characterization of aquaporins from foreign systems, such as plants, by heterologous expression. Taking advantage of this walled cell permeability assessment and in order to individually characterize aquaporins, the E a was calculated enabling to distinguish between the increase in aquaporin activity and the increase in lipid bilayer permeability [37]. It is well known that changes in pH affect the lipid bilayer permeability [24,38] possibly affecting accuracy of aquaporin activity measurements. Therefore, when comparing permeabilities measured at distinct pHs, a second criterion such as the E a should be taken into account. In this study, the E a values determined were in good agreement with the P f measured, indicating that the aquaporin activity rather than the bilayer fluidity is being evaluated.
All the cloned aquaporins were successfully expressed in a S. cerevisiae aqy-null mutant and their localization at the plasma membrane was confirmed by GFP fluorescence. Interestingly, the growth phenotypes of the yeast strains transformed with each of the two PIPs and one TIP were different and in accordance with previous studies showing that the absence of active aquaporins improve yeast osmotic tolerance [32]. Under osmotic stress conditions, the strains harboring the PIPs exhibited an improved growth, whereas the strain expressing VvTnTIP2;1 exhibited growth inhibition. Although the right localization of VvTnPIP1;1 and VvTnPIP2;2 at the yeast plasma membrane was confirmed, no increase in water conductance was observed. An incorrect insertion or even a defect in the oligomerization state of these aquaporins into the yeast plasma membrane may explain the lack of activity observed. However, since an improved growth phenotype under osmotic stress was depicted for strains expressing either VvTnPIP, one may speculate that these proteins might be involved in the transport of atypical substrates as suggested by the performed sequence analysis. Also, the possibility that PIPs coexpression is required for function, as previously reported for maize [39] and for Beta vulgaris [36], cannot be discarded. The permeability induced by VvTIP2;1 expression was roughly ten-fold higher compared to the aquaporin null strain, consistent with a lower E a and dependent on cytosolic acidification. A similar pH effect was reported using molecular dynamic simulations and point mutations on a conserved His residue in the intracellular loop D of PIP aquaporins from tobacco [24] and A. thaliana [14]. The protonation of this specific His residue leading to the pore closure was correlated with a consequent loss of water transport through these proteins. The analysis of amino acid sequence of VvTnTIP2;1 revealed an equivalent His residue localized in loop D and facing the cytosol (His 131 ), suggesting that TIP2;1 might sense and be regulated by cytosolic pH; this idea was confirmed by the absence of pH regulation in AtTIP1;1 [14] and in VvTIP1;1 [25] which lack the correspondent His. To examine the molecular basis of VvTnTIP2;1 sensitivity to pH, we investigated the effects of point mutations in the H 131 residue that might be involved in this regulation. This aromatic basic and polar residue was individually mutated to an acidic and polar residue (Asp, H131D), a neutral and non-polar residue (Ala, H131A) and an aliphatic basic and polar residue (Lys, H131K). A marked decrease in water permeability was observed for the H131K mutant; in spite of belonging to the same amino acid class, the Lys residue shows a longer hydrocarbon chain with a higher pKa (10.53 for Lys vs. 6.10 for His), implying that not only a steric effect but also its permanent protonation at cytosolic pH may be responsible for the loss of water permeability. The H131D mutant showed a similar water channel activity for the same cytosolic pH, but with loss of pH sensitivity, a result that can be expected from its acidic lateral chain with a pKa of 3.86, preventing its protonation. As for the H131A mutant, this non-polar small amino acid was not expected to contribute to channel blockage; indeed, only a minor effect on pH sensitivity that cannot be assigned to the residue charge or structure was observed.
Evidence for aquaporin gating by cytosolic pH has been reported for B. vulgaris roots using an enriched fraction of tonoplast membrane vesicles [40]. In this case when pH dropped from 8.3 to 5.6, permeability decreased 42%. A similar inhibition was observed for VvTnTIP2;1 in the present study when cytosolic pH dropped from 6.8 to 4.8, the activity being specifically dependent on cytosolic pH. Conversely, AtTIP5;1 activity was shown to be dependent on extracellular pH [41]. On the contrary, when expressed in oocytes, VvTIP1;1 from Cabernet Sauvignon was shown to be insensitive to pH [25]. Apparently, different patterns of pH regulation can be found among tonoplast aquaporins.
In conclusion, these results suggest that His 131 in the D loop segment of VvTnTIP2;1 is involved in aquaporin gating, being the first clear evidence of a pH regulation for a TIP2;1. These data may provide a clue for a coordinated regulation of TIPs in plant tissues under anoxic pressure, as previously suggested [14,42]. Further studies will be required to clarify the role of V. vinifera aquaporins aiming to untangle the mechanisms involved in lignocellulosic plant water balance.

Plant Material
Vitis vinifera cv. Touriga nacional (kindly provided by S. Amâncio and S. Tavares, ISA-TUL) was used. Cell suspensions were obtained by adapting V. vinifera callus to liquid culture as described in [43]. Briefly, 4 g callus tissue was dispersed in 50 ml of liquid MS [44] (Duchefa Biochemie, Haarlem, NL) supplemented with 2.5 mM 2,4-D, 1 mM kinetin, 5 g/l PVP-40T, 20 g l 21 sucrose, pH 5.7, in 250 ml flasks. The cultures growing in the dark, at 25uC, in a rotary shaker at 100 rpm were sub-cultured weekly by diluting 25 ml culture into 25 ml of fresh medium.

Yeast strains, maintenance and growth conditions
The yeast strains and plasmids used in this work are listed in Table 1. All Saccharomyces cerevisiae strains are derivatives of S. cerevisiae 10560-6B MATa leu2::hisG trp1::hisG his3::hisG ura3-52 aqy1::KanMX4 aqy2::HIS3; (further indicated as aqy-null). S. cerevisiae aqy-null was used as recipient strain in complementation experiments with the plasmids listed in Table 1. Escherichia coli DH5a [45] was used for routine propagation of the plasmids. E. coli DH5a strain was routinely maintained in Luria-Bertani medium (LB) at 37uC; ampicilin (100 mg ml 21 ) and 5-bromo-4chloro-3-indolyl-ß-D-galactopyranoside (X-Gal, 4 mg ml 21 ) were used as supplements [46] when required. The recipient yeast strain was maintained in YPD medium (5 g l 21 yeast extract, 10 g l 21 peptone, 20 g l 21 glucose and 20 g l 21 agar). Transformant strains were maintained and grown in YNB medium without amino acids (DIFCO) with 2% (w/v) glucose (and 2% (w/v) agar for solid medium) supplemented with the adequate requirements for prototrophic growth [47]. When gene expression was to be reduced to its minimum level, an additional supplementation with 6 mM methionine was used.
The ability of yeast strains to grow under osmotic stress was assessed on solid YNB medium supplied with sorbitol or KCl (pH 5) to the desired final concentrations. Cells were grown in liquid YNB medium with orbital shaking, at 28uC up to OD 640 <1, corresponding to 1610 7 cells/ml. Multi-well plates were prepared with serial 10-fold dilutions of the original culture and plates were inoculated with 3 ml drops using a replica platter for 96-well plates device and incubated at 28uC. Growth was recorded after 1 and 2 weeks.
Heterologous expression of V. vinifera aquaporins in S. cerevisiae RNA extraction and cDNA preparation -Total RNA was extracted from V. vinifera callus cells using RNeasy plant Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer protocol. All RNA samples were treated with RNase free DNase I (Qiagen, Hilden, Germany) and quantified using absorption of U.V. light at 260 nm. cDNAs were synthesized from mRNAs using Superscript III first strand Synthesis system for RT-PCR, priming with oligod(T)12-18 following manufacturer's instructions (Invitrogen), and were used for PCR reaction using the primers described in Table  S1. Full-length CDS were obtained using specific primers designed within the 59 and the 39 non-coding region of each cDNAs. PCR products were sequenced by automatic sequencing ABI Prism DNA sequencer (Perkin-Elmer), analyzed using Blast tools and compared with other plant known aquaporin sequences focusing specially on V. vinifera available sequences.
Cloning aquaporin genes -Forward and reverse aquaporin primers modified to incorporate restriction sites for XbaI (underlined) and ClaI (underlined), respectively, were used to amplify DNA fragments containing full-length ORFs encoding the respective aquaporin (Table S1). PCR amplification was carried out in an Eppendorff thermocycler with DNA polymerase from Finnzymes (annealing temperature according to the primer characteristics). The amplified products were sequenced, digested with ClaI and XbaI, purified using the purification kit ''GFX PCR DNA and Gel Band Purification'' (GE Healthcare) and cloned into the corresponding restriction sites of pUG35 digested by the same restriction enzymes, behind MET25 promoter and in frame with GFP sequence and CYC1-T terminator. Cloning was performed according to standard protocols described in [46]. Constructs were named according to Table 1. In addition, a mutated VvTnTIP2;1 in which glutamic acid E146 was replaced by a glycine residue G146 (TIP21.E146G) was also obtained. The plasmids were cloned into a DH5a E. coli strain, propagated, subjected to extraction and restriction analysis and sequenced.
Transformation of S. cerevisiae aqy-null strain was performed by the lithium acetate method described in [48]. Transformants were selected on YNB medium without uracil. E. coli plasmid isolation was performed by alkaline extraction as described [49] and modified [46]. For plasmid isolation from yeasts, the procedure described by [50] was used. Agarose gel electrophoresis and restriction site mapping were performed according to standard methods [46].
Cellular localization of V. vinifera aquaporins -Exponentially grown cells were spotted onto microscope slides and observed with an Olympus AX70 fluorescent microscope. For GFP visualizing, a U-MWB fluorescent cube was used with excitation filter 450-480 nm and barrier filter 515 nm.
Site directed mutagenesis -For site directed mutagenesis, plasmid extracted from LL0.T21 was used as template for mutagenic PCR using primers described in Table S1. Each primer contained the proper nucleotide mutation and a 45 bp tail in order to promote homologous recombination. An internal fragment of TIP2;1 aquaporin cDNA was removed by digestion with XbaI and BstI restriction enzymes. The PCR mutagenized product and digested plasmid were purified as described above. S. cerevisiae was transformed simultaneously with both DNA fragments, allowed to recombine and selected in medium without uracil. Yeast colony PCR products were sequenced.

Cell volume assay
Equilibrium cell volumes were obtained by loading cells with CFDA under a fluorescent microscope equipped with a digital camera as previously described [29]. Cells were assumed to have a spherical shape with a diameter calculated as the average of the maximum and minimum dimensions of each cell. Sorbitol osmotic shocks of increasing tonicity were imposed on a microscope slide and an average of 6 pictures with 4-6 cells each were taken before (V o ) and within 10 to 40 s after the osmotic challenge (V ' ). The tonicity of the osmotic shock is defined as the ratio of the final to initial osmolarity of the outside medium, L = (osm out ) ' /(osm out ) o .

Measurement of osmotic permeability coefficient
The stopped-flow technique was used to monitor cell volume changes induced by osmotic shocks, in cells loaded with a concentration-dependent self-quenching fluorophore [29]. Cells were pre-loaded for 10 min at 30uC with the nonfluorescent precursor 5-and-6-carboxyfluorescein diacetate (CFDA, 1 mM in isosmotic solution) that is cleaved by intracellular nonspecific esterases generating the fluorescent form, expected to remain mainly in the cytoplasm. Although some of the probe may be either accumulated in the vacuole or exported to the medium [57], this effect can be neglected in face of the very rapid water flow [29]. As the cells shrink in response to osmotic changes, the concentration of the entrapped fluorophore increases with a change in the fluorescence output [29]. To avoid pH interference in fluorescence during the osmotic shock, cell suspensions and osmotic solutions were buffered with K + -citrate/KH 2 PO 4 50 mM at the selected pH.
Experiments were performed on a HI-TECH Scientific PQ/SF-53 stopped-flow apparatus, which has a 2 ms dead time, controlled temperature, interfaced with an IBM PC/AT compatible 80386 microcomputer. Four runs were usually stored and analyzed in each experimental condition. In each run, 0.1 ml of cell suspension (initial osmolarity (osm out ) o = 1.4 M) was mixed with an equal amount of hyperosmotic sorbitol solution (final tonicity = 1.5) to reach an inwardly directed solute gradient and induce an outward water flow responsible for cell volume change. Fluorescence was excited using a 470 nm interference filter and detected using a 530 nm cut-off filter. The time course of volume change was followed by fluorescence quenching CFDA. The recorded fluorescence signals were fitted to a single exponential from which the rate constant (k) was calculated. The osmotic water permeability coefficient, P f , was estimated from the linear relationship between P f and k [58], where V w is the molar volume of water, V o /A is the initial volume to area ratio of the cell population, and (osm out ) ' is the final medium osmolarity after the osmotic shock.

Activation energy of water transport
Stopped-flow experiments were performed at temperatures ranging from 9 to 37uC. The activation energy (E a ) of water transport was evaluated from the slope of Arrhenius plots (lnP f as a function of 1/T).

pH effect on aquaporin activity
In order to evaluate pH effect on aquaporin activity, adjustments of intracellular pH (pH in ) of yeast cells expressing V. vinifera aquaporins were undertaken by varying extracellular pH (pH out ) and by adding benzoic acid, which promotes intracellular acidification due to the passive diffusion of the non-dissociated form of the acid followed by dissociation inside the cell [59]. Cells were washed and incubated under three distinct conditions: pH out 6.8, pH out 5, and pH out 5 plus 4 mM benzoic acid in ice cold 1.4 M sorbitol and re-suspended in the same buffer with the same final biomass concentration in each suspension. Cells were kept on ice under these conditions for 90 minutes before stopped flow assays. pH in was calculated from the relative distribution of labeled 14 [C]propionic acid [60] using the same conditions described above.