Do pH changes in the leaf apoplast contribute to rapid inhibition of leaf elongation rate by water stress? Comparison of stress responses induced by polyethylene glycol and down-regulation of root hydraulic conductivity



    1. Laboratoire d'Ecophysiologie des Plantes sous Stress Environnementaux, UMR759 INRA/Montpellier SupAgro, 2, place Viala, 34060 Montpellier, France
    Search for more papers by this author
    • Current address: European Food Safety Authority (EFSA), Largo N. Palli 5/a, 43121 Parma, Italy.


    1. Ecologie Fonctionnelle & Biogéochimie des Sols, UMR1222 INRA/IRD/Montpellier SupAgro, 2, place Viala, 34060 Montpellier, France
    Search for more papers by this author

    1. Ecophysiologie et Génomique Fonctionnelle de la Vigne, UMR1287 Institut des Sciences de la Vigne et du Vin, 210 Chemin de Leysotte, CS 50008, 33883 Villenave d'Ornon, France
    Search for more papers by this author

    1. Laboratoire d'Ecophysiologie des Plantes sous Stress Environnementaux, UMR759 INRA/Montpellier SupAgro, 2, place Viala, 34060 Montpellier, France
    Search for more papers by this author

    Corresponding author
    1. Laboratoire d'Ecophysiologie des Plantes sous Stress Environnementaux, UMR759 INRA/Montpellier SupAgro, 2, place Viala, 34060 Montpellier, France
      T. Simonneau. Fax: +33 4 67 52 21 16; e-mail:
    Search for more papers by this author

  • This work was supported by a doctoral fellowship from INRA to C.E., and grant ANR-05-GPLA-034-02 from the French Agence Nationale de la Recherche to F.T. and T.S.

T. Simonneau. Fax: +33 4 67 52 21 16; e-mail:


We have dissected the influences of apoplastic pH and cell turgor on short-term responses of leaf growth to plant water status, by using a combination of a double-barrelled pH-selective microelectrodes and a cell pressure probe. These techniques were used, together with continuous measurements of leaf elongation rate (LER), in the (hidden) elongating zone of the leaves of intact maize plants while exposing roots to various treatments. Polyethylene glycol (PEG) reduced water availability to roots, while acid load and anoxia decreased root hydraulic conductivity. During the first 30 min, acid load and anoxia induced moderate reductions in leaf growth and turgor, with no effect on leaf apoplastic pH. PEG stopped leaf growth, while turgor was only partially reduced. Rapid alkalinization of the apoplast, from pH 4.9 ± 0.3 to pH 5.8 ± 0.2 within 30 min, may have participated to this rapid growth reduction. After 60 min, leaf growth inhibition correlated well with turgor reduction across all treatments, supporting a growth limitation by hydraulics. We conclude that apoplastic alkalinization may transiently impair the control of leaf growth by cell turgor upon abrupt water stress, whereas direct hydraulic control of growth predominates under moderate conditions and after a 30–60 min delay following imposition of water stress.


The signalling mechanisms associated with leaf growth responses to abiotic stresses, such as drought, salinity and cold, have been the subject of extensive research. This has resulted in the general agreement that the control of leaf growth involves both hydraulic and biochemical messages (Van Volkenburgh 1999; Boyer & Silk 2004). The latter, non-hydraulic messages, imply the generation or redistribution of biochemical (e.g. hormonal) signals that may necessitate variable time delays from minutes to days, depending on the signalling pathway, before they can impact on growth (Munns et al. 2000; Fricke et al. 2004; Vyroubalova et al. 2009). On the short term (minutes), leaf growth responses are thought to be dominated by hydraulic signals that rapidly propagate within the plant along the water continuum (Nonami, Wu & Boyer 1997; Fricke et al. 2006). However, the temporal contribution of the different signals that regulate leaf growth remains controversial (Else et al. 1995, 2001; Munns et al. 2000).

To dissect leaf growth limitations, a widely used biophysical model has been proposed that relates cell expansion to cell turgor (P), cell wall extensibility (m) and the minimal turgor (Y) below which expansion ceases (Lockhart 1965). While P is directly under hydraulic control, changes in the cell wall rheology (m and Y) are targets of biochemical mediators triggered by drought. For example, water deficit, induced by polyethylene glycol (PEG) and salinity, rapidly (within 20 min) decreases cell wall extensibility in the elongation zone of maize leaves (Neumann 1993; Chazen & Neumann 1994; Chazen, Hartung & Neumann 1995). Several signals potentially regulating cell wall extensibility have been examined, including the stress-induced alkalinization of the apoplast (Neumann 1993; Wilkinson & Davies 1997; Bogoslavsky & Neumann 1998; Wilkinson et al. 1998; Felle 2006). High water tension in the xylem, transmitted to the apoplast, could trigger the opening of mechanosensitive Ca2+ channels in the plasma membrane, resulting in the inactivation of H+-ATPases and alkalinization of the cell walls (Netting 2000). According to the acid growth theory, cell walls are less extensible at neutral than acidic pH (Van Volkenburgh & Boyer 1985; Rayle & Cleland 1992). Consistent with this theory, treatments aimed to modify leaf apoplastic pH, such as alkaline pH buffers and inhibitors of proton pumping, altered maize leaf growth (Van Volkenburgh & Boyer 1985; Neumann 1993; Bogoslavsky & Neumann 1998).

Two major obstacles hamper the dissection of what causes leaf growth reduction upon water stress. Firstly, water stress induces tightly co-ordinated changes in apoplastic pH, water potentials and cell turgor in growing zones of the leaf, and it remains difficult to distinguish their respective roles. Van Volkenburgh & Boyer (1985) correlated reduced leaf growth of maize plants subjected to soil drying for 1–5 d, with a reduced ability of the leaf to acidify an external solution. However, leaf water potential was also dramatically decreased by the water stress treatment and could have directly influenced growth by altering turgor. A second obstacle lies in the difficulty for measuring apoplastic pH in planta because of hindered access to the apoplast of growing leaf tissues: In dicots, growing leaves are rapidly coated with cuticular waxes, while in monocots leaf growing zones are hidden within surrounding sheathes of older leaves. Despite these difficulties, attempts to monitor changes in apoplastic pH in parallel with growth of leaves in response to various stresses have been carried out (Pitann, Kranz & Mühling 2009). Direct measurements with microelectrodes are usually restricted to coleoptiles of young seedlings where growing zones are accessible (Pitann et al. 2009), but low transpiration may limit the impact of water stress treatments in comparison to adult plants. Alternatively, xylem sap pH was analysed in several studies on adult plants as a proxy for leaf cell apoplast, because xylem vessels form a continuum with cell walls. The increased pH of xylem sap appeared as a typical response to soil drying for a wide range of species in several long-term studies that may reflect the changes in the apoplast of growing tissues (Hartung & Radin 1989; Gollan, Schurr & Schulze 1992; Wilkinson et al. 1998). However, the universality of xylem sap alkalinization as a short-term response across species has recently been questioned (Sharp & Davies 2009). Moreover, it is debatable whether xylem sap pH is representative for the pH of the apoplast of growing cell walls.

This study addresses the short-term impact of water stress on apoplastic pH and turgor of the growing leaf 6 in intact maize plants. We used an original combination of double-barrelled pH-selective microelectrodes and a cell pressure probe to continuously monitor apoplastic pH and turgor changes in the leaf elongation zone of intact maize plants for 60 min while exposing the roots to various stress treatments. We previously reported rapid growth inhibition of maize leaves in response to root exposure to anoxia and acid loading (Ehlert et al. 2009). Both treatments induced pronounced decreases in the root hydraulic conductivity (Lpr), and resulted in the decrease of plant water potentials in the root and shoot (Ehlert et al. 2009), probably linked to a change in aquaporin activity (Tournaire-Roux et al. 2003).

In addition to acid load and anoxia, we imposed water stress by root treatments with PEG at a concentration equivalent to moderate soil water deficit (−0.5 MPa). By comparing acid load, anoxia and PEG treatments, we have questioned to which extent pH and turgor changes contributed to leaf growth responses under changes in water status or flux. The high temporal resolution of the measurements allowed us to ascertain the chronological order of the signals and possibly the primary cause of leaf growth inhibition.


Plant growth conditions

Maize seeds (Zea mays L. cv. Dea) were germinated on wet filter paper in the dark at 24 °C for 2–4 d. Seedlings with up to 120-mm-long seminal roots and about 30-mm-long first leaves were then placed in tubes without bottoms, with their roots bathing in a continuously aerated solution composed of 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 2-[N-morpholino]ethanesulphonic acid (MES) (pH 5.5–5.8). The nutrient solution was renewed every third to fourth day during plant growth, which took place in a growth chamber at a vapour pressure deficit (VPD) of 0.8 kPa, with cycles of 14 h of light (400 µmol m−2 s−1 photosynthetic photon flux density) at 24 °C and 10 h of dark at 20 °C.

Acid load, anoxia and PEG treatments

Acid load, anoxia and PEG treatments were applied to the roots of intact plants for 120 min when leaf 6 was 15–20 cm in length. Acid load treatments consisted of the application of a weak acid (propionic acid) to the roots in order to decrease the cytosolic pH of root cells. This was achieved by substituting 20 mm propionic acid/potassium propionate for 20 mm KCl in the nutrient solution at constant pH 6.0 as described by Tournaire-Roux et al. (2003). Anoxia was induced by bubbling the nutrient solution with N2, and depletion of oxygen was monitored using an oxygen electrode. Hydraulic conductivity of the whole seminal root system was reduced by 51 and 31% upon acid load treatment and anoxia, respectively (Ehlert et al. 2009). PEG treatment was applied by replacing the standard nutrient solution by a nutrient solution maintained at constant pH and containing 195.4 g PEG L−1 solvent solution (PEG 6000; Sigma-Aldrich, St Louis, MO, USA) corresponding to a water potential of −0.5 MPa (Michel 1983) as verified with a vapour pressure osmometer (Vapro 5520; Wescor Inc., Logan, UT, USA).

Measurements of cell turgor and leaf elongation rate (LER)

Leaf growth and turgor measurements were carried out on the same plant under controlled environmental conditions as described by Ehlert et al. (2009). The VPD was set to 1.3 kPa and leaf temperature was maintained at 24 ± 1 °C. For LER measurements, linear variable differential transformers (LVDTs) (L100; Chauvin Arnoux, Paris, France) were connected to the tips of the growing leaf 6 with linen threads. The LVDTs were connected to a datalogger (LTD-CR10; Campbell Scientific, Leicestershire, UK), and data were averaged and stored every minute. The LER was calculated from the displacement of the LVDT during intervals of 20 min.

To measure the turgor of growing cells in leaf 6 with a cell pressure probe, a window of 12 mm length and 5 mm width was cut through leaves 2 to 5 at a distance of 2.5–3.5 mm from the leaf insertion point in the growing zone (Bouchabke, Tardieu & Simonneau 2006). The exposed leaf tissue was covered with a thin layer of vaseline to prevent water loss. Cells were impaled with a microcapillary filled with silicone oil and connected to a pressure transducer. Impalement caused the cell sap to invade the capillary that formed a meniscus in contact with the silicone oil. Turgor was determined as the pressure value of the circuitry when cell sap was pushed back into the impaled cell using a microsyringe. Readings were repeated up to three times following several back-and-forth movements of the meniscus (Bouchabke et al. 2006). Only repeatable measurements were kept. Before applying the treatment, the initial average cell turgor was calculated for each plant from five to six measurements on different cells. Cell turgor measurements were then made for 120 min after starting the treatments. Mean turgor values at each time point corresponded to five to seven plants per treatment with three to seven measurements made in different cells of the same plant.

Determination of the osmotic potential of growing tissue

The osmotic potential of the growth zone was measured by using a vapour pressure osmometer (Vapro 5520; Wescor Inc.). A 40-mm-long tissue sample of the growth zone of leaf 6 was excised at a distance of 10 mm from the leaf insertion point. Tissue samples were placed in 1.5 mL reaction tubes and immediately transferred to liquid nitrogen. The tissue sap was extracted by centrifugation (5 min at 10 000 g). 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. Calibration was performed using paper discs soaked with 100, 290 and 1000 mm NaCl standard solutions (Wescor Inc.) before the osmotic potential of the tissue extracts was determined.

Manufacture of pH-selective double-barrelled microelectrodes

The initial steps of pH-selective double-barrelled microelectrode preparation were similar to those described by Miller & Zhen (1991). Two borosilicate microcapillaries (with filament) of 1.5 mm external diameter and 0.86 mm inner diameter (GC150F-10; Phymep, Paris, France) were glued to each other using two component glue (Araldite, Huntsman Advanced Materials, East Lansing, MI, USA). Previously, one end of one of the capillaries was slightly bent by heating it over a flame. During the microelectrode pulling procedure with a vertical micropipette puller (Narishige model, Tokyo, Japan), the microelectrode was manually twisted by 180° in less than 1 min. This resulted in an overall tip diameter of approximately 2–5 µm. One barrel of the microelectrode was silanized using 2% dimethyldichlorosilane (DCDMS, 40140; Fluka Chemicals, Gillingham, UK) in chloroform. The pH-selective sensor cocktail was prepared with 28% (w/w) polyvinyl chloride (PVC, Fluka Chemicals), 61.5% (w/w) Hydrogen Ionophore II cocktail A (95927; Fluka Chemicals) and 10.5% (w/w) nitrocellulose (Whatman International Ltd, Maidstone, UK) (Miller & Smith 1992). The components were mixed and dissolved in 200 µL tetrahydrofuran (THF). The silanized barrel was filled with approximately 20 µL of the cocktail. The microelectrodes were dried for at least 72 h at room temperature with their tips pointing downwards until the cocktail had formed a plug at the tip. The pH-selective barrel was backfilled with calibration solution, pH 6.0 (Table 1) (Walker, Smith & Miller 1995), and the reference barrel was backfilled with 100 mm KCl.

Table 1.  Compositions of solutions used to calibrate pH-selective microelectrodes for measurements in the leaf apoplast
Calibration solution (pH)KCl (mm)NaH2PO4·2H2O (mm)20 mm buffer
  1. KHP, potassium hydrogen phthalate; MES, 2-[N-morpholino]ethanesulphonic acid; MOPS, 3-[N-morpholino]propanesulphonic acid.


Measurement of apoplastic pH using pH-selective microelectrodes

As for turgor measurements, a small window was cut through leaves 2 to 5 at a distance of 2.5–3.5 mm from the leaf insertion point to gain access to the growing tissue of leaf 6. Water loss of the surface of the exposed leaf tissue was prevented by placing a small piece of water-saturated paper towel at the extremity of the window. The experimental device used for apoplast pH measurements was described by Plassard et al. (1999). Both the pH-selective barrel and the reference barrel were connected to microelectrode holders and plugged into a headstage (Axon, HS-2; gain × 0.0001 M, Axon Instruments, Union City, CA, USA) of a high impedance dual differential electrometer (Axoprobe 1A; Axon). The headstage was fastened to a one-dimensional micromanipulator (BE-8; Narishige). The microelectrodes were calibrated using standard pH buffers (Table 1), and microelectrodes with sensitivity higher than 50 mV per logarithmic unit of proton concentration were retained for pH measurements (Shabala & Lew 2002). Before the measurements in the leaf apoplast, a macroelectrode (3 mm diameter, A92135; Fisher Scientific, Illkirch, France ) was placed in the nutrient solution of the plant providing connection between the ground of the electrometer and the nutrient solution. The microelectrode tip was placed at approximately 10 µm from the surface of the elongation zone of leaf 6 using a video camera (Leica, Norcross, GA, USA). Using the micromanipulator, the tip of the microelectrode was brought into tight contact with the leaf apoplast. Electrometer output was recorded using Maclab R system (ADInstruments, Oxfordshire, UK). Data were displayed, analysed and stored using the Maclab 8e interface and the Maclab CHART software (version 3.3, both ADInstruments). The microelectrode was recalibrated after each successful measurement to check that the sensitivity remained higher than 45 mV per logarithmic unit of proton concentration.

Photosynthetic photon flux density, VPD, air and leaf temperatures were monitored and maintained similar to those of the culture chamber in which turgor and LER measurements were made. The mean elongation rate of leaf 6 was checked in plants for which apoplastic pH was monitored using a ruler at different time points, and was similar to those of plants for which turgor and LER measurements were made.


Rapid and transient cessation of leaf growth in response to PEG contrasted with the gradual effects of acid load and anoxia

PEG application to the root medium at a concentration equivalent to −0.5 MPa water potential severely reduced the elongation rate of leaf 6 (Fig. 1a). LER reduction was very rapid, with almost total inhibition within the first 10 min following PEG application. Complete growth inhibition persisted until 30 min after the onset of PEG treatment, then LER progressively resumed at a lower rate than under control conditions. After 50–60 min of PEG treatment, LER recovered to approximately 0.7 mm h−1 (i.e. 30% of the original elongation rate), and remained at this reduced value during the subsequent 60 min of treatment. By contrast, progressive LER inhibition was observed from 10 to 50 min in response to acid load, and extended from 10 to 90 min after the onset of anoxia treatment. LER stabilized afterwards without recovery, remaining at 52 and 60% of control values for acid load and anoxia treatments, respectively. This resulted in the convergence of LERs for all three root treatments after 120 min of application.

Figure 1.

Impact of polyethylene glycol (PEG, −0.5 MPa, black squares), acid root load at pH 6.0 (white squares) and anoxia (white triangles) applied in the solution bathing the roots of intact maize plants on leaf elongation rate (LER) (a) and on turgor of growing cells in the elongation zone of the same leaf 6 (b). Treatments were applied at t = 0 min as indicated by the dashed line. LER and turgor data represent the mean value for n > 5 plants (±SE) from at least two independent plant cultures.

Turgor of growing cells was drastically reduced within 30 min of PEG treatment with only weak osmotic adjustment

All treatments induced a significant decrease in turgor of growing cells, measured directly with the cell pressure probe in the region with maximum elongation rate at the base of leaf 6 (20–40 mm beyond the leaf insertion point). In agreement with the effects of the treatments on LER, the most depressive effect on turgor was induced by PEG application (Fig. 1b). Overall, differences across treatments in the responses of cell turgor were similar to the differences in LER responses when considering their amplitudes and durations of change. Sixty minutes after the applications of PEG, acid load or anoxia, cell turgor was reduced by 55, 34 or 18%, respectively. As observed for LER, the depressive effect of PEG treatment on cell turgor was more rapid and transiently more pronounced than the effect of acid load or anoxia, but without the steep decrease and recovery observed for LER. Fifty minutes after the application of PEG, LER had already recovered, while turgor decrease was only halfway down to its minimal value. From 80 to 120 min after PEG application, turgor slightly recovered (to 58% of its initial value), which was not the case in anoxia and acid load treatments.

After approximately 120 min, cell turgor adjusted to similar values for all treatments, although with slight significant differences (from 0.18 ± 0.03 MPa to 0.25 ± 0.05 MPa).

Osmotic potential was measured on bulk tissues sampled in the elongation zone of leaf 6 between 10 and 50 mm from the leaf base. Osmotic potential remained stable in all conditions except between 60 and 120 min after the onset of PEG treatment when a slight, but significant, reduction was detected (Fig. 2). The decrease in osmotic potential (−0.14 MPa) during this time interval coincided in time with turgor recovery (Fig. 1b), but was greater than turgor increase (+0.05 MPa).

Figure 2.

Impact of polyethylene glycol (PEG, −0.5 MPa, black squares), acid root load at pH 6.0 (white squares) and anoxia (white triangles) root treatments on the osmotic potential of the growth zone of leaf 6. Legend details as in Fig. 1. Data represent the mean osmotic potential for n > 5 plants (±SE) from at least two independent plant cultures.

The apoplastic pH in the leaf elongation zone increased in response to PEG treatment, but not in response to acid load or anoxia

Figure 3 shows representative examples of measurements with pH-selective microelectrodes of pH changes in the apoplast of the elongation zone of leaf 6 in response to acid loading, anoxia and PEG treatments. In this study, measurements of apoplastic pH have proved difficult in the leaf elongation tissue potentially because of cell displacement by growth. This might have caused leaks of cell contents when a cell was damaged by the tip of the microelectrode. Therefore, measurements showing rapid and persisting variations in apoplastic voltage and pH were discarded. A minimum period of 5 min with stable pH measurements was required, before any stress treatment was applied.

Figure 3.

Typical examples of recordings from pH-selective microelectrodes in the apoplast of the leaf 6 elongation zone in response to polyethylene glycol (PEG) (a, b), acid load (c, d) and anoxia (e, f) treatments. Legend details as in Fig. 1. In a, c and e, the upper trace shows the recording of the pH-selective barrel and the lower trace shows the apoplastic potential. (b, d and f) The time-dependent change of the apoplastic pH as calculated from the traces in a, c and e.

The apoplastic voltage, represented by the lower trace in Fig. 3a, c and e, was unaffected by the three treatments. The pre-stress apoplastic voltage in the elongation zone ranged from −3 to −33 mV, with a mean of −18 mV. The upper trace of each figure shows the response of the pH-selective barrel. Because an ion-selective electrode senses the apoplastic voltage in addition to voltage caused by the activity of the ion of interest, in this study, the apoplastic pH was calculated by subtracting the apoplastic voltage from the pH-selective one (Ammann 1986).

The pre-stress apoplastic pH ranged from 4.1 to 5.9, with a mean of 5.1. Within 10 min after exposure to PEG, the response of the pH-selective barrel in the absence of any significant change in apoplastic voltage resulted in an apoplastic pH change from 5.2 to 6.2 during the first 60 min (Fig. 3b). In most cases, no significant pH change was detected in the presence of acid load or anoxia, although slight acidification was occasionally observed (Fig. 3f).

Figure 4 summarizes the pH measurements made on seven to nine plants for each treatment, grown in independent experiments. Consistent with Fig. 3, the mean apoplastic pH increased during 60 min of treatment from 4.9 ± 0.3 to 5.8 ± 0.2 after the onset of PEG treatment. By contrast, the pH of the apoplast was not significantly affected by acid load or anoxia.

Figure 4.

Time courses of leaf apoplastic pH measurements with pH-selective microelectrodes after the onset of polyethylene glycol (PEG) (black squares), acid root load at pH 6.0 (white squares) and anoxia (white triangles) root treatments at time t = 0 min. The pH data represent the mean value measured in the apoplast of the elongation zone of leaf 6 of n > 5 plants (±SE) from independent plant cultures.

LER and turgor concomitantly changed in response to all root treatments with slight, transient discrepancy upon PEG application

Combined measurements of turgor in growing tissues and LER enabled the estimation of changes in tissue extensibility using Lockhart's framework. Tissue extensibility can be estimated with the slope of a linear relationship that could be drawn between elongation rate and turgor in Fig. 5. Overall, the elongation rate was well coupled with turgor, suggesting a constant wall extensibility insensitive to root treatments, except during the first 20 min following PEG addition when growth collapsed while turgor remained intermediate. This observation indicates a transient cell wall hardening during this period of PEG application.

Figure 5.

Leaf elongation rate (LER) as a function of cell turgor measured with a cell pressure probe in the elongation zone of leaf 6 under control conditions (black diamonds) and at different times after the onset of polyethylene glycol (PEG) root treatment (grey circles, time = 20 min; black squares, time = 40–120 min), acid root load at pH 6.0 (white squares, time = 20–120 min) and anoxia (white triangles, time = 20–120 min). Each point represents mean values for n > 5 measurements on one same plant.


Acid load, anoxia and PEG treatments were used to decrease leaf water potential in different ways. Their effects on turgor and leaf growth contrasted in amplitudes and durations of change. Anoxia had the slowest and least pronounced effect, acid load had a slightly more rapid and pronounced effect, and PEG had the most rapid and most depressive effect. The comparison of these three treatments therefore offered us the means to test conditions in which apoplastic pH was affected, and whether the pH changes were correlated to changes in LER. Based on turgor measurements in growing leaf cells for all three treatments, a common pH- or hydraulically mediated control for all conditions can be discussed.

pH Signals may contribute to rapid leaf growth inhibition in response to PEG, but not to anoxia or acid loading

No changes in apoplastic pH were observed in our study in response to acid load. Excess protons in the roots, resulting from the permeation and dissociation of the weak acid, if any, did not alter the apoplastic pH of the leaf elongation zone.

Root anoxia had no influence on maize leaf apoplastic pH either. This contrasts with the rapid alkalinization observed in the apoplast of mature barley leaves when roots were subjected to anoxia (Felle 2006). Rapid cytosolic acidification in the root cells (Roberts et al. 1984; Felle 1996; Tournaire-Roux et al. 2003) has been attributed to the depletion of ATP, resulting from oxygen deficiency and, consequently, to the reduced activity of the plasma membrane H+-ATPase (Gout et al. 2001). Owing to a lower H+ excretion activity, an opposite pH change (i.e. alkalinization) could be expected in the root apoplast (Felle 1996, 2006). Similar modifications of apoplastic pH could be expected in the leaf cell walls in response to oxygen deficiency in the roots. The absence of pH response in our study in maize therefore contrasts with previous reports in barley. This might be caused by a difference in residual oxygen availability to leaves between experimental conditions or to the diversity in mechanisms involved in stress responses between different plant species (Lu & Neumann 1998). A further explanation could be that apoplastic pH and its responsiveness to root stress in mature leaf tissue (as reported in barley) could differ from that of the leaf elongation zone (as observed in this study in maize). The apoplast of the mature leaf is a continuum, interconnected by mature xylem vessels, whereas the leaf elongation zone is separated by non-mature protoxylem vessels supplying water and ions to growing cells (Tang & Boyer 2002). Therefore, the pH of the elongation zone might be disconnected from that of the apoplast in the rest of the plant (Evert, Russin & Bosabalidis 1996; Martre, Durand & Cochard 2000). Furthermore, the leaf elongation zone exhibits a different metabolic activity (Bajji, Lutts & Kinet 2001) and potentially a different pH buffering capacity than mature tissue.

Our study showed that, in contrast to anoxia and acid load, PEG-induced water stress triggered a substantial increase in apoplastic pH of the leaf elongation zone, which may have reduced the cell wall extensibility in this zone. This is consistent with previous studies where concomitant variations in xylem sap pH and growth rate have been reported in response to water deficit (Wilkinson et al. 1998; Jia & Davies 2007). At a timescale of days, the inhibition of leaf growth by water deficit was related to inhibition of cell wall acidification (Van Volkenburgh & Boyer 1985). In maize roots, the spatial profile of growth under non-stressed conditions and the adaptation of root growth to water deficit appeared to correlate with the distribution of the rate of proton pumping into the cell walls and with the resulting apoplastic pH (Peters & Felle 1999; Fan & Neumann 2004).

A common hydraulic signalling in all treatments?

It is striking that cell turgor and LER decreased with similar kinetics and similar differences among treatments apart from a transient variation (0–20 min) of leaf growth in response to PEG treatment, which was not accompanied by changes in cell turgor (Fig. 5). Therefore, apart from these first 20 min upon PEG application, turgor remained a better candidate than apoplast pH to explain the changes in LER. This suggests that hydraulic signalling dominates the control of rapid leaf growth responses for most conditions analysed in this study, although other, unmeasured factors could also have contributed. Using two additional conditions to reduce the water availability for the plant, namely the reduction of the root hydraulic conductivity and the application of PEG to mimic soil water deficit, the present study provides further evidence that the regulation of short-term growth responses to reduced water availability depends on turgor changes to a large extent.

The more marked effects of the treatments on LER than on turgor are compatible with the widely used Lockhart model (Lockhart 1965) relating growth to turgor. In this model, growth is proportional to the difference between cell turgor (P) and cell wall yield (Y) below which growth stops (that could have been represented by a straight line relating LER and P in Fig. 5 with a non-zero intercept equal to Y on the P-axis). Our results fit a unique Lockhart model in most conditions except the transient cessation of growth upon PEG treatment. Excluding the corresponding points would give an overall linear relationship between LER and P with a P-axis intercept close to 0.1 MPa. Assuming that Y was maintained close to this value in most conditions, the mean decrease in turgor after 120 min for the whole set of treatments (from 0.33 to 0.2 MPa) corresponded to a mean reduction in (P − Y) from 0.23 to 0.1 MPa (i.e. a 57% decrease). With the assumption that only P mediated the response of leaf growth, the same 57% reduction for LER was expected from the Lockhart model, which was in good agreement with observed reduction (62%). However, as outlined in Fig. 5, our results also highlighted some transient discrepancies between observations and the Lockhart model, namely a rapid (within 10 min) leaf growth cessation induced by the PEG treatment without a marked change in turgor. This suggests that, in addition to a hydraulic control of leaf growth, other signalling mechanisms transiently superimposed their effects on Y or cell wall extensibility. Previous studies of short-term impacts of water stress on maize growth have reported such rapid reduction in cell wall extensibility in leaves (Chazen & Neumann 1994) or increase of the yield threshold Y in roots (Frensch & Hsiao 1994, 1995).

A cascade of signals interconnecting hydraulic and pH responses with different contributions depending on the timescale

Rapid leaf growth inhibition and simultaneous cell wall hardening observed in elongating maize leaves after the addition of PEG were hypothesized to be primarily generated by hydraulic signals (Chazen & Neumann 1994). Boyer's group defended that, on a very short timescale, growth was first inhibited by a decrease in the growth-induced water potential gradient between the xylem and the leaf elongation zone that causes the blockage of water entry in growing cells before any change in their wall properties occurred (Nonami et al. 1997; Tang & Boyer 2002, 2003; Tang et al. 2002). The rapid decrease in turgor that was observed in our experiments is consistent with such an inhibited water entry into the growth zone and could be per se the cause of growth cessation without any modification in cell walls. However, this hydraulic interpretation does not hold for the period of time immediately following the addition of PEG. Turgor was maintained at about 0.2 MPa during the 10 min period upon PEG application, while leaf elongation ceased. Other root treatments showed that this level of turgor was above the yield threshold Y of mildly treated plants (0.1 MPa as indicated above) and should have been accompanied by a LER of about 1 mm h−1 (Fig. 5). This indicates that a non-hydraulic effect of PEG rapidly blocked growth.

As a refinement to Boyer's interpretation, we propose that an early pH-mediated cell wall hardening was superimposed on decrease in turgor resulting in rapid growth inhibition. Our results indicated that hydraulic and pH signals occurred simultaneously. A clear causal link between hydraulic modification in the tissues and change in wall extensibility is still missing, giving rise to several interpretations. On a very short term (within 10 min), when the PEG treatment had the most pronounced impact on turgor, compared to acid load and anoxia, it is possible that strong negative tensions generated by PEG treatment would have exceeded a threshold and triggered pH changes. Netting (2000) argued that extracellular pH changes induced by water deficit could result from changes in activity of mechanosensitive Ca2+ channels which respond to negative tensions of the apoplastic water potential. Abscisic acid (ABA) could act as a second messenger triggered by changes in apoplastic pH, because pH gradients influence ABA redistribution (Bacon, Wilkinson & Davies 1998). A change in bulk leaf ABA content was not detected in the present study during the 120 min period of root treatment (not shown), but this does not rule out possible redistribution of the hormone.

Overall, our results suggest that changes in hydraulic components predominated in the mediation of short-term growth responses and that changes in apoplastic pH would have a transient, weaker role, perhaps only triggered by abrupt change in external water potential.