Aquaporin-Mediated Reduction in Maize Root Hydraulic Conductivity Impacts Cell Turgor and Leaf Elongation Even without Changing Transpiration^,

Abstract

Root hydraulic conductivity in plants (Lp_r) exhibits large variations in response to abiotic stimuli. In this study, we investigated the impact of dynamic, aquaporin-mediated changes of Lp_r on leaf growth, water potential, and water flux throughout the plant. For this, we manipulated Lp_r by subjecting roots to four independent treatments, with aquaporin inhibitors applied either to transpiring maize (Zea mays) plants grown in hydroponics or to detopped root systems for estimation of Lp_r. The treatments were acid load at pH 6.0 and 5.0 and hydrogen peroxide and anoxia applied for 1 to 2 h and subsequently reversed. First, we established that acid load affected cell hydraulic conductivity in maize root cortex. Lp_r was reduced by all treatments by 31% to 63%, with half-times of about 15 min, and partly recovered when treatments were reversed. Cell turgor measured in the elongating zone of leaves decreased synchronously with Lp_r, and leaf elongation rate closely followed these changes across all treatments in a dose-dependent manner. Leaf and xylem water potentials also followed changes in Lp_r. Stomatal conductance and rates of transpiration and water uptake were not affected by Lp_r reduction under low evaporative demand. Increased evaporative demand, when combined with acid load at pH 6.0, induced stomatal closure and amplified all other responses without altering their synchrony. Root pressurization reversed the impact of acid load or anoxia on leaf elongation rate and water potential, further indicating that changes in turgor mediated the response of leaf growth to reductions in Lp_r.

Leaf growth is an essential process for crop production and is subject to large temporal fluctuations with environmental conditions. There is accumulating evidence that a large part of the changes observed in leaf growth depends on water transport within the plant (Sperry et al., 1998; Bouchabke et al., 2006). It has also been shown that changes in leaf water potential induced by root pressurization can trigger rapid variations of leaf elongation rate in wheat (Triticum aestivum) and barley (Hordeum vulgare; Passioura and Munns, 2000). This raises the question of whether and to what extent low hydraulic conductivity within the plant can limit leaf growth.

After the stomata, the root system represents the largest resistance to water flow in the soil-plant atmosphere continuum (Steudle and Peterson, 1998). Root hydraulic conductivity (Lp_r) is affected by environmental stimuli such as drought, salinity, anoxia, low temperature, and nutrient availability (Zhang and Tyerman, 1991; Azaizeh et al., 1992; Birner and Steudle, 1993; Boursiac et al., 2005; Vandeleur et al., 2009). This ability to respond rapidly to fluctuating conditions suggests that Lp_r may participate in plant adaptation to diverse environments (Steudle, 2000). Aquaporins, a large family of water channel proteins located in plasma and intracellular membranes, are the main determinants of water flow across plant cells and tissues (Javot et al., 2003; Maurel et al., 2008). The dynamic changes in Lp_r in response to chemical or environmental stimuli may result from modifications of aquaporin abundance or activity (Carvajal et al., 1996; Tournaire-Roux et al., 2003; Boursiac et al., 2005). In particular, aquaporin regulation by phosphorylation, protonation, and relocalization in intracellular compartments has been reported in response to extracellular stimuli (Guenther et al., 2003; Tournaire-Roux et al., 2003; Vera-Estrella et al., 2004; Boursiac et al., 2008).

The first insights into the involvement of aquaporins in physiological processes such as cell enlargement, tissue differentiation, and organ movement have been obtained at the cell or tissue level (Hukin et al., 2002; Moshelion et al., 2002; Wei et al., 2007). It is still unknown to what extent changes in root aquaporin activity impact integrated physiological processes such as shoot growth of intact plants. The importance of aquaporins in controlling physiological processes in adult, transpiring plants is assumed to be rather limited; this is because the proportion of water transport controlled by aquaporins is believed to be much lower than that in slowly transpiring plants (Steudle and Frensch, 1996; Steudle and Peterson, 1998).

The manipulation of aquaporin activity offers the possibility to address this question. A classical approach is to alter the expression of aquaporin genes. Down-regulation of genes encoding aquaporins of the Plasma membrane Intrinsic Protein1 (PIP1) and PIP2 subfamilies in Arabidopsis and NtAQP1 in tobacco (Nicotiana tabacum) reduced the ability of these plants to recover after a water deficit treatment (Martre et al., 2002; Siefritz et al., 2002). However, genetic studies have been hindered by phenotypic compensation due to the functional redundancy of aquaporin isoforms in plants (Hachez et al., 2006b). The use of aquaporin inhibitors is a useful tool to investigate the role of root aquaporins in controlling leaf growth. Mercuric chloride (HgCl₂), which blocks aquaporins by binding of Hg²⁺ ions to Cys residues, has been widely used to evaluate the contribution of aquaporins to root water transport (Maggio and Joly, 1995; Carvajal et al., 1996). For instance, Lu and Neumann (1999) have observed that root exposure to 0.5 mm HgCl₂ immediately inhibited leaf growth in water-stressed rice (Oryza sativa) seedlings, thereby suggesting a role for aquaporins in controlling leaf growth. However, the signaling mechanisms involved in leaf growth inhibition remained unclear. Side effects of HgCl₂ application, such as the reduction of membrane potential in root cortex cells or an impaired cell respiration, restrict its usefulness in physiological studies (Wan and Zwiazek, 1999; Zhang and Tyerman, 1999). Manipulating the root environment is an alternative strategy to efficiently alter Lp_r and to assess the significance of such changes on leaf and/or shoot growth. The effects of varying Lp_r by root chilling or anoxia are correlated to leaf growth responses (Malone, 1993; Else et al., 1995, 2001), but the interpretation of these results remains controversial in the absence of precise hydraulic measurements.

The goal of this study was to determine whether alterations of root aquaporin activity can influence leaf growth in intact, adult plants via effects on Lp_r and cell turgor in the leaf elongation zone. In this work, we compared three chemical treatments that target aquaporin inactivation in roots via different mechanisms. Each of them could exert side effects, but provided that all treatments resulted in common responses, the role of aquaporins on leaf growth could be established. We have followed, with a high temporal definition, the consequences of experimentally induced changes in Lp_r on water flux, leaf water potential, and leaf elongation rate under different scenarios (three evaporative demands and pressurized or nonpressurized root systems). In addition, we have measured cell turgor in growing leaves using a pressure probe to investigate whether cell turgor responds to changes in root hydraulic conductivity and whether such changes could account for the control of leaf growth.

The first treatment used to alter Lp_r was acid loading of the solution surrounding the roots, which causes cytosolic acidification in root cortex cells. This triggers the closure of aquaporins due to the protonation of a conserved His residue (Tournaire-Roux et al., 2003). The second treatment was hydrogen peroxide (H₂O₂) application to the roots, which results in the inhibition of Lp_r in maize (Zea mays) by oxidative gating of aquaporins and/or their internalization (Ye et al., 2004; Aroca et al., 2005; Ye and Steudle, 2006; Boursiac et al., 2008). The third treatment was anoxia, an environmental stress that induces an inhibition of Lp_r in a large array of species through proton-induced closure of aquaporins (Zhang and Tyerman, 1991; Birner and Steudle, 1993; Else et al., 1995; Tournaire-Roux et al., 2003).

RESULTS

Cell Hydraulic Conductivity in Maize Root Cortex Was Decreased by Acid Loading

Figure 1 shows typical examples of the time course of hydraulic relaxation in root cortical cells following a change in turgor imposed by a cell pressure probe. Cells from the first and second outer cortical layers of untreated roots exhibited a short half-time of hydrostatic relaxation (t_1/2) of 2.9 ± 0.9 s (n = 9 cells; ±sd; Fig. 1B). These values were similar to t_1/2 values previously reported for maize root cortex cells (Hose et al., 2000). Acid load treatment of the root segments with propionic acid at pH 6.0 for 1 h significantly increased the half-time to 7.4 ± 2.0 s (n = 8 cells; P < 0.05; Fig. 1B). This represents a decrease in cell hydraulic conductivity (Lp_cell) from 7.7 ± 2.5 × 10⁻⁷ m s⁻¹ MPa⁻¹ (control) to 2.7 ± 1.2 ×10⁻⁷ m s⁻¹ MPa⁻¹ (acid load). The cell elastic modulus was not affected by acid loading (control, 4.4 ± 0.8 MPa; acid load, 4.1 ± 0.4 MPa). Turgor varied among measured cells and, on average, exhibited slightly but not significantly higher values upon acid loading (0.38 ± 0.09 MPa compared with 0.27 ± 0.04 MPa before acid load).

Figure 1.

Effects of acid load at pH 6.0 on pressure relaxation of cortical cells in maize roots. A, Typical examples of pressure probe recordings in root cortex cells of control and acid-loaded roots. A stationary turgor was reached within a few minutes after cell impalement with a cell pressure probe. Exo-osmotic and endo-osmotic water movements across the cell membrane were induced by quickly changing cell pressure using the probe. The probe was adjusted manually to maintain a constant volume until turgor had returned to a constant value. Hydrostatic relaxation was monitored in conditions where water exchange can mostly be accounted for by passage through the plasma membrane of the measured cell. B, Mean t_1/2 for cortex cells in control (n = 9) and acid-loaded (n = 8) root segments. Error bars indicate sd.

The Hydraulic Conductivity of Seminal Root Systems Was Reduced in Response to Acid Load, H₂O₂, and Anoxia Treatments

The effects of root treatments on Lp_r were studied using the suction technique applied to excised seminal root systems in hydroponics (Fig. 2). In a first series of experiments, a constant suction force (−0.02 MPa) was maintained at the upper section of the excised root system, and the resulting water flow was continuously monitored before and after root exposure to acid load, H₂O₂, or oxygen deprivation. All treatments induced a rapid reduction in root water flow, with half-times of approximately 15 min (Fig. 2, A–D). Experiments were replicated with different suction forces applied to the root systems. The resulting water flow per unit root area was linearly related to the applied suction force (r² ranging from 0.59 to 0.82). Lp_r was calculated as the slope of the regression line. Statistically significant reductions in Lp_r by 51%, 61%, and 54% were observed during the 1-h period following acid loading at pH 6.0, acid loading at pH 5.0, and H₂O₂ treatment, respectively (Fig. 2E). In the same manner, bubbling the nutrient solution with N₂, which induced a decrease in oxygen partial pressure from 95% to 10% in less than 10 min, caused a reduction in Lp_r by 31% (Fig. 2E). These effects were partially reversed when treatments were reversed after 60 min (Fig. 2E). No change in the osmotic potential of the xylem sap was detected after any of the treatments (data not shown). We conclude that the treatments were appropriate to induce rapid and appreciable reductions in Lp_r.

Figure 2.

Changes in root water flow (Jv; A–D) and in Lp_r (E) of hydroponically grown maize seedlings in response to various root treatments: acid load at pH 6.0 (AL6; A), acid load at pH 5.0 (AL5; B), 2 mm H₂O₂ (C), and anoxia (D). Jv and Lp_r measurements were made on individual excised maize root systems kept in nutrient solution. Water flow was induced across root systems by applying tensions between −0.02 and −0.06 MPa below atmospheric pressure to the detopped section. A to D, Typical time courses of Jv changes at a constant tension of −0.02 MPa. The treatments were applied at 0 min (black arrows) and reversed at 60 min (white arrows) either by washout of acid load and H₂O₂ (A–C) or rebubbling with air (D). During the root treatment, changes in Jv were induced by varying the applied tension. E, Lp_r was deduced from the slope of the Jv versus tension relationship; means with sd (error bars; n = 5–6 plants) were calculated before treatment (control), 60 min after the onset of treatment, and 60 min after its reversion.

Reduction in Lp_r Had No Effect on Transpiration Rate, Stomatal Conductance, and Water Uptake under Low Evaporative Demand But Affected Xylem and Leaf Water Potentials

Under low evaporative demand (vapor pressure deficit (VPD) maintained at 1.35 kPa and photosynthetic photon flux density at 150 μmol m⁻² s⁻¹), neither plant transpiration nor stomatal conductance was affected by the acid-loading treatment at pH 6.0 (Fig. 3). The mean transpiration rate was 11 ± 1.38 mg s⁻¹ m⁻² before and 12.7 ± 1.4 mg s⁻¹ m⁻² at 60 min after the onset of root treatment (n = 6 plants; ±sd). Similar results were obtained for the H₂O₂ treatment (data not shown). Root water uptake was not affected either (11.9 ± 2.2 mg s⁻¹ m⁻² before and 11.7 ± 1.9 mg s⁻¹ m⁻² after the acid-loading treatment; n = 6 plants). Therefore, the reduction in Lp_r induced by acid loading was not associated with any significant change in water uptake or water loss. Consistently, the ratio of uptake to transpiration rate, calculated for each plant immediately before and 60 min after the onset of acid loading, was also unaffected (1.08 ± 0.07 and 1.05 ± 0.04, respectively). The ratios of greater than unity indicated that there was a slight accumulation of water in the plant, which is possibly related to plant growth. All root treatments reduced xylem and leaf water potentials progressively and in parallel during the first 60 min of treatment: reduced by 0.44, 0.5, 0.6, and 0.24 MPa in response to acid loading at pH 6.0, acid loading at pH 5.0, H₂O₂, and anoxia treatments, respectively (Fig. 4). Therefore, all treatments markedly affected plant water relations not only at the root but also at the shoot level without alteration in plant transpiration. However, a constant difference between xylem and leaf water potentials of about 0.15 MPa was maintained in all cases, consistent with the maintenance of water flow rate shown in Figure 3.

Figure 3.

Typical examples of the time course of changes in transpiration rate of intact maize plants in response to acid loading of the roots under two evaporative demands. Hydroponically grown maize plants were transferred to plastic containers with standard nutrient solution and placed on balances. The evaporative demand in the culture chamber was maintained either low (dashed line; 1.3 kPa VPD and 150 μmol m⁻² s⁻¹ photosynthetic photon flux density) or high (solid line; 2.8 kPa VPD and 400 μmol m⁻² s⁻¹ photosynthetic photon flux density). The container was continuously weighed, and weight loss data were stored in a data logger. Plant transpiration rate was calculated at intervals of 15 min and related to the leaf area. Acid load at pH 6.0 was applied to the roots at 0 min, as indicated by the dashed vertical line.

Figure 4.

Effects of root treatments on xylem (crosses, white circles, and gray symbols) and leaf water potentials (black and white squares) of maize plants grown in hydroponics. Treatments were acid load at pH 6.0 (A), acid load at pH 5.0 (B), 2 mm H₂O₂ (C), and anoxia (D). Xylem and leaf water potentials were measured on leaves 2 and 4 of plants kept at low evaporative demand in B to D and on plants kept at low (black symbols) or high (white symbols and dashed line) evaporative demand in A (as described in Fig. 3). Leaf 2 was wrapped in aluminum foil down to the base of the plant (including first leaves) at the end of the day preceding experiments to prevent transpiration and to reach water balance with the xylem at the root-leaf junction. For the sake of clarity, sd is not presented in A for plants subjected to high evaporative demand; mean sd was approximately 0.1 MPa. The treatments were applied at 0 min (black arrows) for 120 min. Data at 0 min correspond to xylem and leaf water potentials of untreated plants. Data represent mean values of n = 4 plants. Error bars indicate sd.

Increased Evaporative Demand, When Combined with Acid Load at pH 6.0, Induced Stomatal Closure and Amplified Xylem and Leaf Water Potential Responses

High evaporative demand (VPD maintained at 2.8 kPa and photosynthetic photon flux density at 400 μmol m⁻² s⁻¹) induced much higher transpiration rates in control plants compared with plants maintained under low evaporative demand (Fig. 3). In this case, acid loading at pH 6.0 dramatically decreased the transpiration rate and the water potentials (Fig. 4A). Xylem and leaf water potential stabilized at approximately −1.3 and −0.8 MPa, respectively, after 60 to 80 min of acid load.

Synchronous Fluctuations in Leaf Elongation Rate and in Turgor of Growing Cells Occurred after Acid Loading, H₂O₂, or Anoxia Treatment

Following acid loading or H₂O₂ treatment, the elongation rate of leaf 6 decreased for 30 to 40 min, with half-times of approximately 15 to 20 min when measured under low or high evaporative demand (Fig. 5, A–C). Anoxia induced a more gradual response of leaf elongation rate, which decreased for more than 100 min. Leaf elongation then stabilized at a significantly lower value (reductions of 48%, 79%, 75%, 58%, and 40% in response to acid loading at pH 6.0 and low evaporative demand, acid loading at pH 6.0 and high evaporative demand, acid loading at pH 5.0, H₂O₂, and anoxia treatments, respectively). The inhibition of leaf elongation rate by acid loading at pH 6.0 or anoxia treatment was partially reversed within 60 and 120 min after return to a standard nutrient solution or rebubbling with air, respectively. By contrast, no reversal was observed following acid loading at pH 5.0 or H₂O₂ treatment. We tested whether the decrease in leaf elongation rate observed after acid-loading treatments could be due to a propagation of propionate in the leaf elongation zone. To address this possibility, we used ¹³C-labeled propionic acid and analyzed via mass spectroscopy the abundance of ¹³C in root and leaf tissues (Fig. 6). No accumulation of ¹³C was observed in the elongation zone of leaf 6, suggesting that reductions in leaf elongation rate following acid-loading treatments were not due to significant accumulation of propionic acid molecules in the leaf elongation zone.

Figure 5.

Effects of root treatments on leaf elongation rate (LER; A–D) and turgor of growing cells of the same leaf (E–H). Treatments were acid load at pH 6.0 (A and E), acid load at pH 5.0 (B and F), 2 mm H₂O₂ (C and G), and anoxia (D and H). For all treatments, leaf elongation rate and turgor were measured simultaneously on the same plants (mean ± sd; n = 5 plants from at least two independent cultures). All treatments were performed at a low evaporative demand (1.3 kPa VPD; crosses and gray symbols). Additionally, in A and E, plants were treated with acid load at pH 6.0 in the dark with low VPD (0.8 kPa; black circles) and in the light with high VPD (2.8 kPa; white circles). In A, sd is not presented for clarity reasons; mean sd was approximately 0.4 mm h⁻¹. All treatments were applied at 0 min (black arrows), and acid load or H₂O₂ was washed out at 60 min (white arrows; A–C and E–G). Anoxia was reversed by rebubbling the solution with air at 120 min (white arrow; D and H).

Figure 6.

Accumulation in leaf and root tissues of ¹³C-labeled propionic acid during an acid load at pH 6.0 treatment. Accumulation was analyzed by mass spectroscopy and expressed in ¹³C abundance (δ¹³C/₀₀). A typical acid root loading treatment was performed at pH 6.0 using 1 mm ¹³C-labeled sodium propionate and 19 mm unlabeled potassium propionate. ¹³C contents were monitored over time in root tissues (black circles), elongating leaf 6 tissues (white triangles), and nonelongating leaf tissues (white squares). Acid load was applied at 0 min for 120 min. Data represent mean ¹³C abundance values for n = 6 plants (±sd).

A decrease in turgor of growing cells paralleled that of leaf elongation rate in all treatments under low evaporative demand, with larger reductions for acid loading at pH 5.0 (45%) than for H₂O₂ and acid loading at pH 6.0 treatments (30% and 34%; Fig. 5, E–G). Anoxia had the most gradual effects on both cell turgor (18% reduction) and leaf elongation rate (Fig. 5, D and H), while leaf elongation rate and cell turgor remained constant in nontreated plants (data not shown). The osmotic potential of growing tissues of leaf 6, as determined by psychrometry, was not affected during the 60 min of all treatments (Table I). After the end of each treatment, turgor and leaf elongation rate exhibited similar time courses of response (i.e. a clear recovery in response to acid loading at pH 6.0 and anoxia and no recovery in response to acid loading at pH 5.0 and H₂O₂ treatment).

Table I.

Osmotic potentials (MPa) measured on leaf tissue extracts using a vapor pressure osmometer (WESCOR Vapro 5520)

Pieces of leaf tissue were sampled on maize plants before treatment (control) or 60 min from the start of root treatment (described in the text). Leaf samples were frozen and then centrifuged to collect the liquid leaf extracts. Values are means ± sd of eight or more samples on different plants.

Root Treatment	Leaf 6 (Growing Zone)	Leaf 6 (Mature, Emerged Zone)	Leaf 4 (Mature Zone)
No (control)	−0.93 ± 0.05	−0.74 ± 0.11	−0.73 ± 0.09
Acid load, pH 6.0	−0.95 ± 0.08	−0.78 ± 0.08	−0.79 ± 0.09
Acid load, pH 5.0	−0.96 ± 0.04	−0.77 ± 0.07	Not determined
Anoxia	−0.87 ± 0.04	−0.73 ± 0.05	−0.82 ± 0.1
H2O2	−0.94 ± 0.04	−0.76 ± 0.05	Not determined

Leaf elongation rate and turgor responses to acid loading at pH 6.0 were further analyzed under three transpiration regimes (Fig. 5, A and E). Both leaf traits were unaffected by root treatment in dark conditions. By contrast, the responses of leaf elongation rate and turgor were amplified by an increase in evaporative demand.

Overall, the synchrony between leaf elongation rate and turgor was remarkably conserved in all transpiration regimes and root treatments tested. Therefore, these data suggest that reductions in leaf elongation rate in response to chemical treatments on roots were mediated by hydraulic signals from the root to the shoot and were a consequence of reductions in turgor of growing leaf cells.

To further investigate the synchrony between hydraulic processes and the inhibition of leaf elongation rate in response to the decrease in Lp_r, we examined the relationships between mean reductions in Lp_r and changes in water potential gradients, turgor, and leaf elongation rate (Fig. 7). A surprising result was that the gradient of water potential between the nutrient solution and the mature leaves (as measured with a pressure chamber) increased more with root treatments than the root hydraulic resistance (experimental points were located above the 1:1 line in Fig. 7A), while an Ohm's law analogy would suggest simultaneous variations. However, this result did not apply to the water potential of the expanding zone. This water potential can be estimated by the sum of measured cell turgor and osmotic potential presented in Table I. Changes in total water potential 60 min after the start of the treatments were much lower in the expanding zone (decreased by 0.17–0.25 MPa depending on treatments) than in mature tissues (Fig. 4). This suggests that our treatments have triggered an increase in the resistance between the xylem and mature tissues, which did not apply to the growing zone located at the base of the leaf. Turgor of growing cells and Lp_r followed a common relationship in all treatments, supporting a physiological link between them (Fig. 7B). The same was true for the relationship between turgor and leaf elongation rate, which were linked by a common relationship throughout all treatments (Fig. 7C). Thus, there was a marked similarity between the effects of acid root loading, H₂O₂ treatment, and anoxia on Lp_r, turgor, and leaf elongation rate, which only differed in amplitude. The greater the inhibiting effect of a treatment on Lp_r, the greater the reduction in turgor and leaf elongation rate.

Figure 7.

Relationships between hydraulic variables and leaf elongation rate across treatments. A, Reduction in Lp_r plotted against the gradient of water potential between the nutrient solution and the leaf (ΔΨ_{leaf-solution}). B, Lp_r plotted against turgor in growing leaf cells. C, Turgor in growing leaf cells plotted against leaf elongation rate (LER). Changes in Lp_r or in water potential gradients are normalized by the initial values at 0 min (e.g. a value of 2 means a doubling at 60 min from the start of treatment compared with the initial value). Black diamonds, control (before treatment); gray circles, acid load at pH 6.0 under low evaporative demand; white circles, acid load at pH 6.0 under high evaporative demand; gray triangles, acid load at pH 5.0; gray squares, H₂O₂ treatment; crosses, anoxia. A and B, n = 4 plants; error bars indicate sd. C, Each point corresponds to the reductions in turgor and in leaf elongation rate measured on the same plant.

Root Pressurization Reversed the Effects of Acid Loading or Anoxia

Pressurization of the roots of an intact plant by means of a pressure chamber increases the water potential of the total medium-plant system. We used this method to test whether the observed reductions in leaf elongation rate were primarily caused by hydraulically mediated processes (Fig. 8). When root pressurization was applied to an intact plant that had received an acid load at pH 6.0 treatment for 30 min, the leaf elongation rate was restored to values recorded prior to the treatment (Fig. 8A). We measured water potentials in the mature, transpiring leaf 4 at 1 h after simultaneous pressurization and acid load treatment and obtained an average value of −0.43 ± 0.13 MPa, similar to those of nontreated plants (n = 3; ±sd). Similarly, no reduction in leaf elongation rate was observed when root pressurization was applied at the same time as anoxia to an intact plant (Fig. 8B).

Figure 8.

Effects of root pressurization on leaf elongation rate (LER) during acid load at pH 6.0 and anoxia treatments of the roots. The cylinders containing nutrient solution and the root systems of intact maize plants were inserted into a pressure chamber. A silicon seal was used to tightly fix the plant to the chamber cap. A, Acid load at pH 6.0 was applied to the roots at 0 min (black arrow). At 30 min, a hydrostatic pressure of 0.08 to 0.09 MPa was applied to the root system (white arrow) using a pressure chamber while acid root load was maintained. B, Root pressurization (white arrows) and anoxia (black arrow) were applied to the roots of intact maize plants. A first increase in pressure of 0.08 MPa was applied to the roots immediately after the oxygen concentration had decreased to less than 10% (0 min; white arrow). As leaf elongation rate was not fully restored by this pressure step, the pressure was further increased to 0.11 MPa (30 min; white arrow).

DISCUSSION

Changes in Lp_r Induce Cessation of Leaf Elongation by Decreasing Turgor of Growing Cells

In this study, four treatments altering Lp_r applied at two levels of evaporative demand had similar effects on the sequence of events leading up to a decrease in leaf elongation rate. All treatments induced reductions in water potential and cell turgor in the leaf growth zone. The similar patterns of the time course of the response of Lp_r, cell turgor, and leaf elongation rate to these treatments, and the consistency of the dose-dependent response, suggest that processes mediating the response of leaf elongation rate were of a hydraulic nature. Regardless of the cause of changes in Lp_r, its inhibition modulated water potentials in roots and shoots, with strongly correlated changes in turgor of growing cells and leaf elongation rate. This also provides evidence for the highly debated role of cell turgor in mediating short-term leaf growth responses to water-challenging treatments (Munns et al., 2000; Bouchabke et al., 2006). According to a recent study in parenchyma cells of maize leaf midribs (Kim and Steudle, 2007), a 4-fold decrease in Lp_cell could follow from a drop in cell turgor by about 0.2 MPa. This may amplify the consequences of changes in Lp_r on the water potential gradient in the growing zone. However, at the tissue scale, we did not detect any change in water potential gradient between xylem and bulk leaf upon Lp_r reduction in our experiments (Fig. 4), whereas the gradient between the solution and xylem (at the root-shoot junction) dramatically increased. This indicated that most of the changes in hydraulic conductivity occurred across the roots but not along the leaves.

Our results are consistent with the general view that leaf elongation rate is more sensitive to hydraulic challenges than stomatal conductance, photosynthesis, or the growth of other organs (Boyer, 1970; Westgate and Boyer, 1985; Saab et al., 1990). The sensitivity of leaf elongation to changes in Lp_r suggests that, under natural conditions, down-regulation of root aquaporins might contribute to leaf growth responses to stimuli sensed in the soil. Nevertheless, we considered the possibility that inhibition of leaf elongation rate was induced in our experiments by events other than a reduction in Lp_r. Metabolic changes in root cells due to weak acid or H₂O₂ treatment or a transport of these compounds via the xylem into the leaf might have directly affected leaf elongation rate. However, there are several arguments against this possibility. First, we used acid load, H₂O₂, and anoxia treatments, rather than HgCl₂, a common aquaporin inhibitor reported to affect membrane electrical potential or cell respiration (Wan and Zwiazek, 1999; Zhang and Tyerman, 1999). Second, acid loading and anoxia as well as H₂O₂ treatments rely on different mechanisms to reduce Lp_r (Tournaire-Roux et al., 2003; Ye and Steudle, 2006; Boursiac et al., 2008); therefore, the very similar responses induced by the treatments were probably not the result of toxic side effects of either cytosolic acidification or oxidative stress but rather a direct consequence of the effects on Lp_r. Third, we performed experiments that ruled out the possibility of an appreciable transport of the weak acid or its metabolites into the leaf following acid load treatment. In particular, no accumulation of ¹³C-labeled propionic acid was found in the elongation zone of leaf 6, whereas diffusion of this compound into the roots was clearly observed. Measurements with pH microelectrodes also showed that acid load and anoxic treatments of roots did not induce any change in apoplastic pH in the leaf elongation zone (data not shown). Fourth, neither leaf elongation rate nor turgor was affected by acid load in the dark. This remains consistent with our hydraulic signal hypothesis and excludes the presence of other side effects of acid load. Finally, the reversion by root pressurization of the effects of acid load at pH 6.0 and of anoxia on leaf water potential, cell turgor, and leaf elongation provided evidence that these changes were primarily due to hydraulic signals. In our experiments, these signals were specifically triggered by treatment-induced reductions in Lp_r.

Aquaporin Inhibitors Induced Rapid Changes in Hydraulic Conductivity in Maize Roots

Cell and root hydraulic conductivities were reduced by acid load in maize, as they are in Arabidopsis (Gerbeau et al., 2002; Tournaire-Roux et al., 2003). These decreases are mediated by cell acidosis and subsequent proton-dependent gating of PIP aquaporins (Tournaire-Roux et al., 2003). PIPs of maize are abundantly expressed in the plasma membrane of roots (Hachez et al., 2006a) and show a conserved cytosolic His residue that is critical in pH sensing. We noted, however, that the degree of inhibition of cell hydraulic conductivity by acid loading was less pronounced in maize than in Arabidopsis root cortex cells (Tournaire-Roux et al., 2003), possibly due to a higher intracellular pH buffering capacity or to a lower abundance of aquaporins in maize cortex cells. In support of the first hypothesis, acid loading at pH 5.0 was more efficient than acid loading at pH 6.0 in inhibiting Lp_r and fully blocked leaf elongation rate.

We also confirmed the previously reported decreases of maize Lp_r in response to H₂O₂ treatment (Aroca et al., 2005; Ye and Steudle, 2006). Ye and Steudle (2006) suggested that millimolar concentrations of H₂O₂, as applied in this work, may directly interfere with the activity of aquaporins by oxidative gating. Conversely, Boursiac et al. (2008) recently showed that, in Arabidopsis, the H₂O₂-induced decreases in Lp_r were mediated by a signaling cascade that involved the relocalization of PIP aquaporins into intracellular compartments. The amplitude and speed of Lp_r inhibition in maize roots, together with the inhibition of water transport at the cell level in our work and the work of Ye and Steudle (2006), support the idea that the H₂O₂ treatments result in an inhibition of aquaporins.

A composite transport model of water in the root has been proposed (Steudle and Frensch, 1996; Steudle and Peterson, 1998). This model accounts for variable contributions of transmembrane (where aquaporins may exert a control) and apoplastic (independent of aquaporin activity) pathways to the overall root water uptake, depending on the nature and the intensity of the driving force. In agreement with previous studies (Aroca et al., 2005; Ye and Steudle, 2006), our results suggest that aquaporins mediate a large proportion of water flow across maize roots under all evaporative demands tested.

Why Did Water Potentials Change with Hydraulic Conductivity without Changing Water Flux throughout the Plant?

Despite large reductions in Lp_r in response to root treatments, no associated changes in water uptake or plant transpiration rate were observed in this study under low evaporative demand. This contrasts with the assumption that the rate of water flow follows fluctuations in Lp_r. Concomitant changes in water flow and Lp_r were observed when excised roots were subjected to a stabilized gradient of water potential using special devices allowing the control of root depressurization or pressurization (Lu and Neumann, 1999; Quintero et al., 1999; Martinez-Ballesta et al., 2003). In intact transpiring plants, the water flux across the plant primarily follows changes in evaporative demand and in stomatal conductance, because stomata represent by far the dominant resistance barrier. The hydraulic conductivity of inner tissues can affect water flux either by becoming markedly limiting or by indirectly interfering with stomatal behavior. This was clearly not the case in our experiments, in which the low evaporative demand was maintained steady and the stomatal conductance was not affected by the application of aquaporin inhibitors on roots. Consequently, transpiration was steady and the water potential gradient between the root medium and the xylem at the leaf base was increased and counteracted the Lp_r reduction, at least under low evaporative demand.

This does not rule out the possibility that a reduction in Lp_r causes stomatal closure and thereby could reduce plant transpiration rate under other circumstances. This was observed in this study when Lp_r-reducing treatments were done under higher evaporative demand, which induced a dramatic decrease in leaf water potential, down to values similar to those observed in maize in field conditions (Tardieu and Simonneau, 1998). Previous studies of aspen (Populus species; Wan and Zwiazek, 1999) and pepper (Capsicum annuum; Martinez-Ballesta et al., 2003) also reported that, on a slightly longer term (2–2.5 h), exposure of the roots to HgCl₂ (0.05 and 0.1 mm HgCl₂, respectively) induced a significant decrease in stomatal conductance.

It seems surprising that water potential gradients from solution to mature leaves were more sensitive to root treatments than Lp_r. For example, the 2-fold increase in root hydraulic resistance (the reciprocal of Lp_r) induced by H₂O₂ treatment should have induced a no more than 2-fold increase in water potential gradient between leaf 4 and the nutrient solution. In the absence of appreciable changes in osmotic potential, this suggests that a gradient of water potential developed between the elongation zone at the base of the leaf 6, where turgor was measured, and the mature leaves. Cavitation may have developed in the xylem of mature leaves, consistent with the long-lasting decreases in leaf and xylem water potentials following root treatment.

Short-Term Effects of Anoxia on Leaf Elongation Follow the Same Hydraulic Cascade as Induced by Chemical Inactivation of Root Aquaporins

While an early and conserved response of plants to anoxia is the down-regulation of Lp_r (Birner and Steudle, 1993; Else et al., 1995; Tournaire-Roux et al., 2003), rapid stomatal closure and reductions in leaf elongation rate can also be observed, but the mechanisms underlying these responses are still debated (Neuman and Smit, 1991; Else et al., 1995, 2001; Dat et al., 2004). In anoxia-susceptible plant species, including maize, anoxia induces modifications in root metabolism or photosynthesis (Geigenberger, 2003). Smit et al. (1990) proposed along these lines that in poplar, transport of a root chemical stimulus to the shoot may be necessary to cause a reduction in leaf elongation. Yet, neither abscisic acid nor cytokinin concentrations were changed in growing leaf tissues. In tomato (Solanum lycopersicum), soil flooding resulted in a rapid reduction in leaf elongation and water potential, which could not be reversed by root pressurization (Else et al., 1995). This was interpreted as evidence that chemical, rather than hydraulic, messages were involved in this response. In castor bean (Ricinus communis), by contrast, the same authors found that the reduction in leaf growth induced by oxygen deprivation was caused by a negative hydraulic signal generated by a Lp_r reduction (Else et al., 2001).

In our study, in spite of the multitude of physiological alterations that are possibly triggered by anoxia, the cascade of hydraulic responses observed in oxygen-deprived maize plants was remarkably similar to that induced by aquaporin inhibitors. Although the effects of anoxia were less pronounced and slower than those of other root treatments, the synchrony of Lp_r reduction, decrease in cell turgor, and inhibition of leaf elongation were conserved. Furthermore, root pressurization reversed the short-term effect of anoxic stress on leaf elongation rate. These observations provide converging evidence that in maize, the reduction of Lp_r induced by anoxia was propagated as a decrease in leaf water potential and cell turgor, which in turn resulted in an inhibition of leaf elongation. Finally, differences in the time scales of response may partly reconcile the discrepancy in leaf growth responses raised by studies published to date. In this work, we concentrated on short-term plant responses to Lp_r changes. On this time scale, changes in turgor appear to determine most of the leaf elongation responses. In longer term studies, the hydraulic component may be overridden by osmotic adjustments (Van Volkenburgh and Boyer, 1985), growth-depressive effects of hormones such as cytokinins (Smit et al., 1990) or ethylene (Shiu et al., 1998), or changes in cell wall rheology.

In conclusion, this study provides evidence that changes in Lp_r can be crucial determinants for rapid plant responses to environmental stresses. Changes in cytosolic pH and H₂O₂ have recently emerged as cellular signals triggered by various external stimuli and mediating pronounced and rapid changes in Lp_r (Birner and Steudle, 1993; Tournaire-Roux et al., 2003; Lee et al., 2004; Aroca et al., 2005). Therefore, the cascade of hydraulic responses triggered by acid load or external H₂O₂ as described here could be of general significance for the response of plants to various environmental stresses in the short term. The relevance of hydraulic signaling and leaf growth response resulting from Lp_r reduction in a wider range of environmental stresses and on a longer term deserves further investigation.

MATERIALS AND METHODS

Plant Growth Conditions

Maize seeds (Zea mays ‘Dea’) were germinated on a wet tissue in the dark at 24°C for 2 d. Seedlings were then placed in tubes with their roots bathing in a continuously aerated solution with the following composition: 0.25 mm CaSO₄, 0.8 mm KNO₃, 0.6 mm KH₂PO₄, 0.2 mm MgSO₄(7H₂O), 0.4 mm NH₄NO₃, 2 × 10⁻³ mm MnSO₄, 0.4 × 10⁻³ mm ZnSO₄, 0.4 × 10⁻³ mm CuSO₄, 0.2 × 10⁻³ mm Na₂MoO₄(2H₂O), 1.6 × 10⁻² mm H₃Bo₃, 0.04 mm Fe-EDDHA, and 2.5 mm MES, pH 5.5 to 5.8. The hydroponic solution was renewed every 3rd to 4th d. Plants were grown in a growth chamber at a VPD of 0.8 kPa, with cycles of 14 h of light (400 μmol m⁻² s⁻¹) at 24°C and 10 h of dark at 20°C.

Acid Load, H₂O₂, and Anoxia Treatments and Manipulation of Evaporative Demand

All root treatments were applied, for 60 to 120 min, to the root system of intact plants, which, unless otherwise stated, had a leaf 6 of 15 to 20 cm in length. During the treatments, plants were kept under the same conditions as described above. The evaporative demand was manipulated by changes in VPD (0.8, 1.3, or 2.8 kPa) and/or photosynthetic photon flux density (0, 150, or 400 μmol m⁻² s⁻¹). Measurements of leaf elongation rate, turgor, and water flux were made under identical environmental conditions except that light intensity was at 150 μmol m⁻² s⁻¹. Acid load treatments consisted of applying a weak acid to the roots and were imposed by substitution at constant pH (pH 6.0 or 5.0); 20 mm propionic acid/potassium propionate replaced 20 mm KCl in the nutrient solution, as described by Tournaire-Roux et al. (2003). A more acidic pH (pH 5.0) was also used to enhance diffusion of propionic acid into root cells. H₂O₂ was applied to plant roots at a concentration of 2 mm H₂O₂ in the nutrient solution. Anoxia was induced by bubbling the nutrient solution with N₂, and the depletion of oxygen was monitored using an oxygen electrode.

Measurements of Cell Hydraulic Conductivity of Root Cortex Cells

Cell pressure probe measurements on root cortex cells were performed as described by Gerbeau et al. (2002) and Tournaire-Roux et al. (2003). Seminal and nodal root tip segments of a length of 30 to 33 mm were excised from hydroponically grown maize plants and preincubated in a solution containing either 20 mm KCl and 10 mm MES, pH 6.0 (control), or 20 mm propionic acid/potassium propionate and 10 mm MES, pH 6.0 (acid load). The segments were laid horizontally on a filter paper and partially submerged by perfusion with the same solution. Pressure probe measurements were made at a distance of 21 to 27 mm from the root apex. Cells were impaled with micropipettes with an external tip diameter of 3 to 4 μm. After impalement of a root cortex cell, the meniscus was stabilized at its initial position. For hydrostatic relaxation measurements, a pressure change was applied to the cell and the meniscus was rapidly moved forward and backward and kept at a constant position during the relaxation procedure. Turgor and t_1/2 could be continuously monitored in individual cortical cells over periods of up to 10 min. The cell volumetric elastic modulus (ɛ) was obtained from pressure changes (ΔP) recorded after experimentally imposed volumetric changes (ΔV) according to the following equation (Javot et al., 2003):

where V (m³) is the mean cell volume.

Mean cell diameter and cell length values of 27.6 and 170 μm, respectively, were determined from longitudinal sections on three to four plants (Muller et al., 1998) and used to calculate mean cell volume (V) and cell area (A). The Lp_cell was calculated from the following equation:

where V (m³), A (m²), t_1/2 (s), and ɛ (MPa) are as defined above and P (MPa) and Ψ_ext (MPa) represent the cell turgor and the external osmotic water potential, respectively.

Lp_r Measurements in Whole Seminal Root Systems

A xylem sap flow was triggered in excised root systems by applying vacuum-induced subatmospheric pressures between −0.02 and −0.06 MPa, as described by Freundl et al. (1998) and Hose et al. (2000). The seminal root system was excised by sectioning the mesocotyl, which was then tightly fixed to silicon tubing using low-viscosity dental paste (President Light; Coltene Whaledent). To determine the rate of water flow across the root system, a water trap, made of a 2-mL tube filled with dry cotton, was sealed onto the tubing. Water flow was determined by weighing the xylem sap absorbed by a piece of cotton every 10 min; the cotton piece was renewed after each measurement. Free exudation was determined at atmospheric pressure, starting 30 min after root excision. Then, the tubing was connected to a vacuum port equipped with a precision tension gauge. Tension data were averaged and stored every 60 s using a data logger (Campbell Scientific; LTD-CR10). A two-valve system, driven by the data logger, automatically controlled the tension applied to the seminal roots. The water flux was subsequently measured for at least 60 min at three different suction forces ranging from −0.02 to −0.06 MPa. Acid load, H₂O₂, or anoxia treatment was then applied to the root system, and the solution was replaced by fresh nutrient solution after 1 h. Anoxia treatment was released by simply rebubbling the solution with air. The xylem sap was harvested for at least 60 min at the three different suction forces during and after the treatment. At the end of one experiment, the root system was dried from excess water using paper towels and weighed. The root surface was deduced from an independently determined relationship between the mass and surface of seminal root systems.

Measurements of Plant Water Transport Rates and Stomatal Conductance

Transpiration rate and stomatal conductance were measured on intact plants in a growth chamber under moderate evaporative demand (VPD of 1.3 kPa, photosynthetic photon flux density of 150 μmol m⁻² s⁻¹) and under high evaporative demand (VPD of 2.8 kPa, photosynthetic photon flux density of 400 μmol m⁻² s⁻¹). On the day before measurements were taken, individual maize plants were placed into 400-mL, 20-cm-high cylindrical containers filled with a nutrient solution containing 20 mm KCl, pH 6.0, and transferred into a growth chamber set at identical environmental conditions as described above. Transpiration measurements were obtained by recording changes in weight (of the cylinder containing the nutrient solution and the maize plant) on a balance (Precisa). Direct evaporation from the solution was avoided by covering the solution with a layer of oil and bubbling with water-saturated air. Weight loss data were averaged and stored every 10 s using a data logger (Campbell Scientific; LTD-CR10). Weight loss was plotted against time, and plant transpiration rate was calculated at intervals of 30 min by the slope of this relationship. Changes in weight caused by substitution of nutrient solution due to acid loading were excluded from calculations. Transpiration rates were divided by leaf area (A) calculated from leaf width (W) and length (L) as follows: A = 0.75 W × L. Additionally, a gas-exchange system (CIRAS; PPSystems) was used to continuously measure stomatal conductance on leaf 4 before and during the acid load treatments.

Measurements of root water uptake rate required a slight modification of this experimental setup. A maize plant was held at its leaf base by a metal holder, which was placed on a second balance of the same type. Thereby, one balance measured the weight loss of the solution container only while the other balance measured the changes in weight of the plant. Root water uptake was calculated from weight loss of the solution container, and plant transpiration was obtained from the total weight changes of both balances. Mean root water uptake and transpiration rates were obtained from measurements on six plants under moderate evaporative demand (VPD of 1.3 kPa).

Measurements of Leaf Elongation Rate

Leaf elongation rate and turgor were simultaneously monitored on the same plant. For this, linear displacement transducers (LVDTs; L100; Chauvin Arnoux) were connected to the tips of the growing leaf 6 with linen threads (Bouchabke et al., 2006). The LVDTs were connected to a data logger (Campbell Scientific; LTD-CR10), and data were averaged and stored every 1 min. Leaf elongation rate was calculated from the displacement of the LVDT during intervals of 20 min. To check that changes in leaf elongation rate were not induced by changes in meristem temperature, a fine copper-constantan thermocouple (0.2 mm diameter) was inserted between the sheaths of leaves 1 and 2 at the height of the meristem (Sadok et al., 2007).

Cell Turgor Measurements

A cell pressure probe was set up in the growth chamber to permit turgor measurements in the elongation zone of leaf 6 while simultaneously measuring leaf elongation rate. Two hours prior to measurements, a window of 12 mm length and 5 mm width was cut through the bases of leaves 2 to 5 to access the elongation zone of leaf 6. The exposed zone of this leaf was covered with a thin layer of silicon gel to avoid any water loss that could have altered cell water status. Cell turgor was measured as described previously (Bouchabke et al., 2006). Consistent with the report of Bouchabke et al. (2006), we observed that leaf elongation rate was decreased by approximately 50% after cutting a window through the outer leaf layers and that it remained stable afterward. For this study, we retained plants for leaf elongation rate and turgor measurements that exhibited a leaf elongation rate higher than 1.5 mm h⁻¹ after window opening.

Cells were impaled with a microcapillary filled with silicon oil and connected to a pressure transducer. Readings were repeated up to three times following several back-and-forth movements of the meniscus formed at the contact of the cell sap with the silicon oil. The pressure was read each time the cell sap was pushed back to its location within the cell prior to impalement. Only repeatable measurements with less than 0.05 MPa min⁻¹ deviation during the successive readings were kept. Before applying the root treatment, the initial average cell turgor was calculated for each plant from five to six measurements on different cells. Cell turgor was then measured for 1 to 2 h after applying the treatments. Mean turgor values at each time point correspond to three to seven measurements made in different cells of the same plant. The mean turgor values presented in the figures were calculated for five to seven plants per treatment. After treating the plants for 1 or 2 h, the treatment was reversed by applying control nutrient solution or rebubbling the solution with air, and turgor measurements were pursued as described above.

Water Potential Measurements

Water potentials of leaves 2 and 4 were measured with a pressure chamber (Soil Moisture Equipment). On the day prior to measurements, the basal part of the plant including leaves 1 and 2 was wrapped in aluminum foil in order to prevent transpiration and to allow water potential equilibrium between leaf 2 and its insertion point (Supplemental Fig. S1). Thus, the water potential of leaf 2 was considered as indicative of the water potential within the xylem (Simonneau and Habib, 1991), whereas measurement on leaf 4 characterized the water potential of fully developed transpiring leaves. Prior to the root treatments, initial water potentials of leaves 2 and 4 were measured for five to six control untreated plants. To exclude possible diurnal fluctuations, water potentials were also measured in five to six control plants at 120 min (i.e. at the end of the measurement series).

Osmotic Potential Measurements

The osmotic potential of elongating (leaf 6) leaf tissue was measured using a vapor pressure osmometer (Vapro 5520; Wescor). Leaf tissue samples of 40 mm length were excised, placed in 1.5-mL reaction tubes, and immediately transferred to liquid nitrogen. The tissue sap was extracted by centrifugation (5 min at 10,000g). Filter paper discs (10 mm diameter) were soaked with 10 μL of the extracts and introduced into the chamber of the osmometer to determine the osmotic potential.

Root Pressurization of Intact Plants

The root systems of intact maize plants were inserted into a customized pressure chamber similar to previously described devices (Passioura and Munns, 1984). Plants were tightly fixed to the chamber cap by silicon seals using a screw. Pressures between 0.08 and 0.1 MPa were applied to the root system either directly or 30 min after the onset of acid load at pH 6.0 or anoxia. The chamber was equipped with a precision pressure sensor connected to a data logger (Campbell Scientific; LTD-CR10). Pressure data were averaged and stored every 60 s. A two-valve system, driven by the data logger, allowed the automatic control of the pressure applied to the root system. Leaf elongation rates were monitored before and during root pressurization as described above. After 60 min of acid loading, the water potential of leaf 4 was measured.

Analytical Procedure for ¹³C Analysis in Root and Leaf Tissue by Mass Spectroscopy

¹³C-labeled propionic acid (¹³CH₃CH₂CO₂H; Sigma) was used in an acid-loading experiment at pH 6.0 at a final concentration of 1 mm sodium propionate. Segments of 40 mm length were excised from the root apex, at a distance of 10 mm from the leaf 6 insertion point (elongating tissue) and at the tip of leaf 6 (nonelongating tissue). Tissue sampling was made before the treatment and 120 min after the application of the acid-loading treatment. Tissue samples from six plants at each time point were used for calculating the mean ¹³C content. The sample preparation for mass spectrometry analysis was done as described by Lejay et al. (1997). Briefly, the tissues were dried, ground, and combusted. The ¹³C content (δ¹³C/₀₀) of the released CO₂ was then measured using an online mass spectrometer (GV Instruments). Isotopic signatures were expressed as the ratio δ¹³C/₀₀ relative to carbon standards as described previously by Maros et al. (2006).

Supplemental Data

The following materials are available in the online version of this article.

• Supplemental Figure S1. Comparison of water potentials measured on leaves 2 and 4 previously wrapped in aluminum foil to prevent transpiration.