Overexpression of PIP2;5 aquaporin alleviates effects of low root temperature on cell hydraulic conductivity and growth in Arabidopsis.

The effects of low root temperature on growth and root cell water transport were compared between wild-type Arabidopsis (Arabidopsis thaliana) and plants overexpressing plasma membrane intrinsic protein 1;4 (PIP1;4) and PIP2;5. Descending root temperature from 25°C to 10°C quickly reduced cell hydraulic conductivity (L(p)) in wild-type plants but did not affect L(p) in plants overexpressing PIP1;4 and PIP2;5. Similarly, when the roots of wild-type plants were exposed to 10°C for 1 d, L(p) was lower compared with 25°C. However, there was no effect of low root temperature on L(p) in PIP1;4- and PIP2;5-overexpressing plants after 1 d of treatment. When the roots were exposed to 10°C for 5 d, L(p) was reduced in wild-type plants and in plants overexpressing PIP1;4, whereas there was still no effect in PIP2;5-overexpressing plants. These results suggest that the gating mechanism in PIP1;4 may be more sensitive to prolonged low temperature compared with PIP2;5. The reduction of L(p) at 10°C in roots of wild-type plants was partly restored to the preexposure level by 5 mm Ca(NO(3))(2) and protein phosphatase inhibitors (75 nm okadaic acid or 1 μm Na(3)VO(4)), suggesting that aquaporin phosphorylation/dephosphorylation processes were involved in this response. The temperature sensitivity of cell water transport in roots was reflected by a reduction in shoot and root growth rates in the wild-type and PIP1;4-overexpressing plants exposed to 10°C root temperature for 5 d. However, low root temperature had no effect on growth in plants overexpressing PIP2;5. These results provide strong evidence for a link between growth at low root temperature and aquaporin-mediated root water transport in Arabidopsis.


INTRODUCTION
Low soil temperature is often a major factor restricting growth and yield of plants even if the soil is not frozen (Bonan, 1992;Wan et al., 1999Wan et al., , 2001. In plants that are sensitive to cold soils, growth reduction is accompanied by the reduction of water uptake, which usually starts within a few minutes following the temperature decrease (Bigot and Boucaud, 1996;Lee et al., 2005a). Low soil temperature which is accompanied by relatively high air temperature, with high transpirational demand, poses especially serious challenges to plants. Therefore, it is not surprising that low temperature tolerance has been often correlated with drought resistance (Janowiak et al., 2002;Aroca et al., 2003;Costa e Silva et al., 2007) and can be increased by treatments that facilitate root water uptake and (or) reduce plant water loss (Pérez de Juan et al., 1997;Aroca et al., 2005;Lee et al., 2008). Reduced water flux at low temperatures is thought to occur due to higher water viscosity (Muhsin and Zwiazek, 2002) and the inhibition of transmembrane water transport (Wan et al., 2001;Lee et al., 2005aLee et al., , 2005b. Contrary to chilling-sensitive cucumber plants, low root temperature had little or no effect on cell hydraulic conductivity (L p ) in roots of low temperature-tolerant figleaf gourd (Lee et al., 2005a(Lee et al., , 2005b(Lee et al., , 2008 suggesting that the ability of plants to maintain the transmembrane water flow may be among the key factors linked to chilling tolerance. Transmembrane water flow is regulated by aquaporins, which function as channels for water and other small neutral molecules (Echevarria et al., 1994;Ishibashi et al., 1994;Murata et al., 2000). Aquaporins have been classified into four different subfamilies based on subcelluar localization and sequence similarity including plasma membrane intrinsic proteins (PIPs), tonoplast intrinsic proteins (TIPs), nodulin-like intrinsic proteins (NIPs), small basic intrinsic proteins (SIPs) and the recently identified uncharacterized intrinsic proteins (XIPs) (Chaumont et al., 2001;Johansson et al., 2001;Quigley et al., 2001;Danielson and Johanson, 2008;Gupta and Sankararamakrishnan, 2009). Although low temperature was found to increase the transcript levels of some Arabidopsis aquaporins (Jang et al., 2004), the effects of low temperature on aquaporin expression are not always clear with down-regulation (Lian et al., 2004;Yu et al., 2005) of some aquaporins also reported. Even less is known about the effects of low temperature on aquaporin gating which regulates water flux through protein conformational changes (Lee et al., 2005a). Gating factors known to be involved in aquaporin regulation include phosphorylation and dephosphorylation (Johansson et al., 1996(Johansson et al., , 1998, cytoplasmic pH (Tournaire-Roux et al., 2003;Sutka et al., 2005;Alleva et al., 2006) and divalent cations (Gerbeau et al., 2002). Protein phosphorylation and dephosphorylation are common regulatory mechanisms through which many enzymes and receptor molecules are altered by environmental stimuli (Johansson et al., 2000;Kerk et al., 2002). Phosphorylation of the plasma membrane aquaporins was demonstrated to be responsible for the regulation of temperature-dependent opening of tulip petals (Azad et al., 2004). The phosphorylation of the Ser residue is catalyzed by the plasma membrane-associated Ca-dependent protein kinase (Johansson et al., 2000;Karlsson et al., 2003), which has been also implicated in temperature responses (Azad et al., 2004).
In the present study, we used genetically transformed Arabidopsis thaliana plants overexpressing PIP1;4 and PIP2;5 (Jang et al., 2007) to examine the role of these aquaporins in the responses of plants to low root temperature. The PIP1;4 and PIP2;5 aquaporins were chosen because of the reported increase in their expression levels in roots of plants exposed to low air temperature (Jang et al., 2004). In this study, we subjected Arabidopsis roots to low temperature (10 o C) while the shoots of plants were exposed to high transpirational demand conditions (23/21 o C day/night temperatures) to study the effects of low root temperature on L p and plant growth rates. We also used several inhibitors of protein phosphorylation and dephosphorylation to determine whether these processes may be involved in the responses of L p to low temperature. We hypothesized that 1) the impact of low temperature on root water transport involves aquaporin gating through the phosphorylation/dephosphorylation processes, and 2) overexpression of low temperature-responsive aquaporins PIP1;4 and PIP2;5 would help the plants maintain high L p values and, in consequence, high growth rates when their roots are exposed to low temperature.

Effects of low root temperature on relative growth rates
There were no significant differences in root and shoot relative growth rates between the different plant groups when exposed to 23 o C root temperature ( Fig. 1A and B).
However, when root zone temperature was decreased from 23 o C to 10 o C for 5 d, there was a sharp and statistically significant reduction of the root and shoot relative growth rates in plants of wild-type and those overexpressing PIP1;4 ( Fig. 1A and B). However, plants overexpressing PIP2:5 showed no significant differences in relative shoot and root growth rates at both root zone temperatures ( Fig. 1A

Hydraulic properties of root cells
Cell dimensions and the water relations parameters including turgor pressure (P), half-times of water exchange (T 1/2 ) and cell elasticity (ε) of the root cortex cells were similar in the wild-type and PIP overexpressing plants (Table I). The cell hydraulic conductivity (L p ) values were in the range of 6.2 to 9.0 × 10 -7 m s -1 MPa -1 (Tables I and II). Half-times of water exchange (T 1/2 ) values in wild-type (Table I; Fig. 2A and B) and overexpression plants (Table   I; Fig. 2C and D) were similar, ranging between 1 and 2 s. The addition of 100 µM HgCl 2 significantly increased T 1/2 by four-fold (Fig. 2B) and decreased L p values (Table II) in the wild-type plants but did not affect the stability of P (Fig. 2), demonstrating that mercury did not affect cell integrity in our experimental system. Similar changes, but of lower magnitude (two-fold or less) were recorded in PIP1;4 and PIP2;5 overexpressing plants (Table II; Fig. 2C and D).

Effects of low temperature on T 1/2 and L p
As the root temperature gradually decreased from 25 o C to 10 o C, T 1/2 values increased in root cortex cells of wild-type ( Fig. 3A) but remained unchanged in PIP overexpressing plants ( Fig. 3B and C). Furthermore, T 1/2 values in the wild-type plant did not recover after the root temperature was raised back from 10 C to 25 o C (Fig. 3A). Similar results were also observed for L p values (Fig. 4A). Short term treatment (for 1 d) of root at 10 o C did not affect L p in PIP1;4 and PIP2;5 overexpressing plants (Fig. 4B). However, when plants were exposed to 10 o C for 5 d, L p was significantly reduced in PIP1;4 overexpressing plants, but not in plants overexpressing PIP2:5 (Fig. 4B).

Effects of Ca(NO 3 ) 2, LaCl 3 , and protein phosphatase inhibitors on L p
Application of 1 mM LaCl 3 (calcium channel blocker) in the wild-type plants at 25 o C resulted in an over two-fold decrease in L p (Fig. 4C). The addition of 5 mM Ca(NO 3 ) 2 at 25 o C showed no effect on L p (Fig. 4C). However, when 5 mM Ca(NO 3 ) 2 was added at 10 o C, the value of L p was increased almost to the same level as the one measured at 25 o C (Fig. 4C).
Similarly, 1 mM Na 3 VO 4 and 75 mM okadaic acid increased L p when added to roots at 10 o C ( Fig. 4C).

Activation energy (E a ) for root water transport
www.plantphysiol.org on September 5, 2017 -Published by Downloaded from Copyright © 2012 American Society of Plant Biologists. All rights reserved.

8
Activation energy (E a ) for L p was 63 kJ mol -1 in the wild-type plants (Table III). In both PIP overexpression plants, E a values for L p were below 10 kJ mol -1 (Table III).
Relative expression profiles of the 13 PIP genes in Arabidopsis exposed to low temperature stresses Based on the expression levels at 23 o C, the 13 PIP genes in three-week-old wild-type The expression of low temperature response marker RD29A increased 40-folds at 10 o C after 1 h, and no significant changes in transcript level of actin was observed (data not shown).

DISCUSSION
Direct and continuous measurements of water relation parameters with the cell pressure probe demonstrated that aquaporin-mediated water transport is a major factor which is responsible for growth inhibition of plants growing in cold soils. These results also emphasize once again the pivotal role of aquaporins in stress responses in plants since many environmental stresses directly or indirectly upset water balance in plants (Fennell and Markhart, 1998;Holmberg and Bülow, 1998;Sutka et al., 2011).
Our results demonstrated that the shoot and root growth in Arabidopsis plants overexpressing PIP2;5 aquaporin was less sensitive to low root temperature compared with the wild-type plants. Low soil temperature is a common factor which inhibits plant growth (Wan et al., 2001;Lee et al., 2005a) and upsets plant water balance by reducing root water flux to the transpiring leaves (Wan et al., 1999(Wan et al., , 2001. Synchronization of leaf and root water transport in a plant is required to maintain stomatal conductance that is essential for CO 2 uptake (Fischer et al., 1998;Siefritz et al., 2002;Kaldenhoff et al., 2008). In the present study, higher root and shoot growth rates at 10 o C root temperature in plants overexpressing PIP2;5 clearly indicate that aquaporins are involved in the processes that limit growth of Arabidopsis plants under low root temperature conditions.
Our results indicate that the functional significance of the aquaporin isoforms which are weakly expressed at favorable growth temperatures, such as PIP1;4 and 2;5, increases 9 when the roots of Arabidopsis are exposed to low temperature. Similarly to the earlier report (Jang et al., 2004), in the present study, expression levels of PIP1;4 and PIP2;5 in the wildtype plants increased after 1 h and 24 h exposures of roots to 10 o C. We also found a severalfold increase in another weakly expressed aquaporin, PIP2;6, but only after 24 h of low root temperature treatment.
It is interesting that the overexpression of PIP1;4 was effective in maintaining high water permeability of root cortical cells at 10 o C for 1 d, but not when low root temperature was maintained for 5 d. This may explain why, in contrast to Arabidopsis plants overexpressing PIP2;5, the PIP1;4 overexpressing plants showed growth reductions when Alexandersson et al., 2010). However, they may also reflect possible differences in regulation mechanisms between PIP1;4 and PIP2;5 or in their water-transporting effectiveness, assuming no major differences in PIP1;4 and PIP2;5 membrane density in plants overexpressing the respective aquaporins.
Interestingly, the reduction of L p in the wild-type plants at 10 o C was reversed to as much as 70% of the pre-exposure level by the addition of 5 mM Ca(NO 3 ) 2 and protein phosphatase inhibitors (75 nM okadaic acid and 1 μ M Na 3 VO 4 ), pointing to protein phosphorylation and (or) dephosphorylation as likely gating mechanisms involved in the low root temperature response of aquaporins. Okadaic acid and vanadate are commonly used as inhibitors of protein phosphorylation (Cohen and Cohen, 1989;Gordon, 1991). We have also demonstrated that the Ca channel blocker, 1 mM LaCl 3 , significantly inhibited L p at 25 o C, further supporting the notion that aquaporin gating is linked to the Ca signal (Johansson et al., 1996;Németh-Cahalan and Hall, 2000;Carvajal et al., 2000;Azad et al., 2004).
Although several PIPs and TIPs are insensitive to Hg (Daniels et al., 1994;Biela et al., 1999), many aquaporins are blocked by Hg reagents (Maggio and Joly, 1995;Wan and Zwiazek, 1999) which are thought to bind to sulfhydryl groups resulting in a reversible alteration of protein conformation and channel closure (Zhang and Tyerman, 1999 of Hg on L p , likely since the inhibition of water transport by Hg is only partial and there are more water-transporting channels present in the PIP overexpressing plants (Jang et al., 2007;Lee et al., 2009). PIP1;4 and PIP2;5 do not have Cyst 189 which is known to be present in many Hg-sensitive aquaporins. If these PIP1;4 and PIP2;5 are indeed Hg insensitive, than the reduced Hg sensitivity of L p in both overexpression lines may be also due to their increased overall contribution to water transport. However, there are other cystein residues which these PIPs have (83,105,141,and 146 in PIP1;4 and 74,96,132,and 137 in PIP2;5) and which could potentially affect their Hg sensitivity. In higher concentrations, Hg can also inhibit water transport by acting as a non-specific metabolic inhibitor (Wan and Zwiazek, 1999;Zhang and Tyerman, 1999). In the present study, turgor pressure remained stable during the measurements suggesting that metabolic inhibition was not a major factor in the reduction of In the present study, high E a (63 kJ mol -1 ) were measured in the wild-type plants, demonstrating a high dependency of cell water transport on temperature in Arabidopsis roots.
Therefore, the 7-9-fold lower E a values measured in PIP1;4 and PIP2;5 overexpression lines provide strong evidence for the importance of these proteins in cell water transport at low temperature. Activation energy (E a ) depends on the nature of the rate-limiting barrier for water movement and on the energetic of the water pore interactions (Verkman et al., 1996).
The low E a (< 25 kJ mol -1 ) for water transport is among the typical features of membranes with water-transporting pores (Finkelstein, 1987;Niemietz and Tyerman, 1997) since water moving across a channel does not have to overcome a large energy barrier (Tyerman et al., 1999). On the other hand, water movement through the lipid bilayer would need to surmount the high energy barrier of water partitioning into hydrophobic lipid phase (46-63 kJ mol -1 ) (Tyerman et al., 1999). Most of the studies in higher plants and algae showed E a values ranging from 18-48 kJ mol -1 at the cell and membrane level (Hertel and Steudle, 1997;Niemietz and Tyerman, 1997;Gerbeau et al., 2002;Lee et al., 2005a). However, E a values as high as 186 kJ mol -1 were reported for Tradescantia virginiana leaf epidermal tissues (Tomos et al., 1981) and E a close to 100 kJ mol -1 was measured in cucumber root cortical cells (Lee et al., 2005a). These large differences in reported values may reflect the differences in measurement methods and/or in lipid bilayer properties of different plants and tissues, the 1 1 effects of which have not been extensively studied. High E a values, exceeding 57 kJ mol -1 , in plasma membrane vesicles from tobacco were used as evidence for diffusional water transport across the lipid matrix (Maurel et al., 1997). However, in the present study, L p of the wild-type plants showed not only high temperature sensitivity, but it was also sharply inhibited by HgCl 2 pointing to the significant involvement of aquaporins in water transport.
We suggest that the concept of E a for cell water transport may need to be revisited to consider likely differences in temperature sensitivities of different aquaporins which are involved in water transport processes of different plant species and plant tissues.

CONCLUSION
Overexpression of PIP1;4 and PIP2;5 in Arabidopsis was effective in alleviating the short-term effects of low root temperature on L p in root cells. However, contrary to the overexpression of PIP2;5, this alleviation was no longer present after 5 d of root exposure to 10 o C in plants overexpressing PIP1;4. The temperature sensitivity of cell water transport in Arabidopsis roots was reflected by the reductions in shoot and root growth rates in the wildtype and PIP1;4 overexpressing plants exposed to 10 o C root temperature for 5 d. However, low root temperature had no effect on root and shoot growth in plants overexpressing PIP2;5.
The inhibition of L p by low temperature could be partially prevented by the application of Ca(NO 3 ) 2 and by protein phosphatase inhibitors suggesting that the effect of low temperature on root water transport may involve the Ca-dependent protein phosphorylation/dephosphorylation processes. The results provide strong evidence for a link between growth responses to low root temperature and aquaporin-mediated root water transport in Arabidopsis and suggest that low temperature sensitivity of root water transport may be connected to the aquaporin phosphorylation/dephosphorylation processes. Our data point to the functional and biological significance of a low expression single PIP isoform in growth processes of plants exposed to stress conditions.

Relative growth rate in low temperature treatment
At the commencement of the experiment, 40 plants from each experimental group (wild-type, overexpressing PIP1;4 and overexpressing PIP2;5) were harvested as a reference (time 0) and the remaining 80 plants from each experimental group were divided into two groups and exposed to 10 o C or 23 o C root temperature for 5 d using a 15-L circulating water bath (Thermomix BU coupled to Frigomix U, B. Braun, Melsungen AG, Germany). For dry weight determinations, shoots and roots were harvested and dried in an oven at 80 o C for 72 h.
The relative growth rates of roots and shoots were calculated from the initial and final dry weights (Hoffmann and Poorter, 2002) (n=40).

Cell-pressure probe measurements
Cell-pressure probe measurements were conducted on the cortical cells of roots at 25-30 mm distance from the root apex after excising about 50-mm-long distal root segments.
Nutrient solution of the same composition as for solution culture was circulated along the roots during experiments. A cell-pressure probe was used to measure half-times of water exchange (T 1/2 ), elastic modulus (ε) and turgor pressure (  (cell volume and cell surface area), cell elastic modulus (ε) and osmotic pressure of cell were required (Steudle 1993). Cell dimensions were examined microscopically in thin root sections embedded in Spurr's resin (Spurr, 1969).

Root temperature control for cell pressure probe measurements and application of chemicals
To control the temperature, nutrient solution was pumped through a tube using a circulating water bath (F3, Haake, Berlin, Germany). The first cell-pressure probe measurements of T 1/2 were taken at 25 o C and then the temperature was decreased to 20 o C for 10 min and the measurements were taken again for 10 min. This was repeated at 15 and 10 o C when the last sets of measurements were taken. Root temperature during the measurements was monitored with the thermocouple that was placed near the root surface. The measurements lasted for about 110 min with the same cells measured for all temperatures.
For some of the roots, the measurements were conducted with ascending temperatures from 10 to 25 o C.
When cells had attained a steady P and T 1/2 at either 10 o C or 25 o C for 30 min, roots were treated with 5 mM Ca(NO 3 ) 2 , 1 mM LaCl 3 , 1µM Na 3 VO 4 or 75 nM okadaic acid which were added to the circulating solution to examine whether protein phosphorylation and Ca signals play a role in cell water transport responses to low root temperature. Also, 100 µM HgCl 2 was added to the circulating root medium to determine the contribution of mercurysensitive water transport.. Water relations parameters were continuously measured and compared before and after 20-30 min exposure to chemicals on the same individual cells.

Measurements of E a
Activation energy (E a ) for L p was calculated for all plants from the slopes of Arrhenius plots where the natural log of cortical cell hydraulic conductivity was plotted against the inverse of absolute temperature (T=283-298 K), since:

RNA extraction, reverse-transcription PCR and real time PCR
To investigate the effect of low temperature on the transcript level of each PIP gene, we monitored the expression of all 13 PIP genes in roots of the wild-type plants were grown in solution culture at 23 o C root temperature and then exposed to 10 o C for 1 h and 24 h. We   (Table S1), and quantified the abundance of PIP transcripts using real-time PCR analysis.
Total RNA was extracted from the frozen samples using the Plant RNeasy extraction kit (Qiagen, Valencia, CA, USA). The concentration of RNA was accurately quantified by spectrophotometer measurements using a NanoDrop 1000 apparatus, (NanoDrop, Wilmington, DE, USA) and cDNAs were obtained using a reverse-transcriptase kit (Qiagen).

Statistical analyses
Statistical analyses were performed using SigmaPlot V 11.0 (Systat Software, San Diego, CA, USA) at α = 0.05 confidence level. The data were analyzed using unpaired t-test to compare the water relations parameters, cell dimension, activation energy and aquaporin gene expression, and paired t-test was used for of HgCl 2 effects. ANOVA with Tukey's multiple comparison was used for growth rate and temperature effects on cell hydraulic conductivity.

Supplemental Data
Supplemental Table S1. Gene-specific primer pairs used in PCR.