Elevated CO2 enhances stomatal responses to osmotic stress and abscisic acid in Arabidopsis thaliana

Authors

  • J. LEYMARIE,

    1. CEA/Cadarache – DSV – DEVM – Laboratoire des Echanges Membranaires et Signalisation, F-13108, Saint Paul lez Durance Cedex, France
    Search for more papers by this author
    • Present address: Max Planck Institut für Züchtungforshung, MDL, Carl-von-Linne-Weg 10, 50829 Köln, Germany.

  • G. LASCÈVE,

    1. CEA/Cadarache – DSV – DEVM – Laboratoire des Echanges Membranaires et Signalisation, F-13108, Saint Paul lez Durance Cedex, France
    Search for more papers by this author
  • A. VAVASSEUR

    1. CEA/Cadarache – DSV – DEVM – Laboratoire des Echanges Membranaires et Signalisation, F-13108, Saint Paul lez Durance Cedex, France
    Search for more papers by this author

Alain Vavasseur Fax: (33) 4 42 25 46 56; e-mail: vavasseur@dsvcad.cea.fr

Abstract


A, carbon assimilation rate
ABA, abscisic acid
Ci, intercellular space CO2 concentration
g, leaf conductance
WUE, water use efficiency

Carbon dioxide and abscisic acid (ABA) are two major signals triggering stomatal closure. Their putative interaction in stomatal regulation was investigated in well-watered air-grown or double CO2-grown Arabidopsis thaliana plants, using gas exchange and epidermal strip experiments. With plants grown in normal air, a doubling of the CO2 concentration resulted in a rapid and transient drop in leaf conductance followed by recovery to the pre-treatment level after about two photoperiods. Despite the fact that plants placed in air or in double CO2 for 2 d exhibited similar levels of leaf conductance, their stomatal responses to an osmotic stress (0·16–0·24 MPa) were different. The decrease in leaf conductance in response to the osmotic stress was strongly enhanced at elevated CO2. Similarly, the drop in leaf conductance triggered by 1 μM ABA applied at the root level was stronger at double CO2. Identical experiments were performed with plants fully grown at double CO2. Levels of leaf conductance and carbon assimilation rate measured at double CO2 were similar for air-grown and elevated CO2-grown plants. An enhanced response to ABA was still observed at high CO2 in pre-conditioned plants. It is concluded that: (i) in the absence of stress, elevated CO2 slightly affects leaf conductance in A. thaliana; (ii) there is a strong interaction in stomatal responses to CO2 and ABA which is not modified by growth at elevated CO2.

INTRODUCTION

Stomata regulate the transpiration flux according to environmental conditions. Among the parameters affecting stomatal aperture, abscisic acid (ABA) and CO2 are of major interest. ABA plays a crucial role in plant adaptation to water stress (Giraudat 1995). Synthesized with some delay, during a water stress, ABA induces a decrease in guard cell turgor resulting from a concerted modulation of ion channel activities (Blatt & Armstrong 1993; Pei et al. 1997), in order to limit water losses. CO2 is a second signal resulting in a reduction of stomatal aperture, but the mechanisms underlying this response are still debated (Willmer & Fricker 1996). It has been proposed that the apoplastic malate pool reflects the ambient CO2, and that a high malate concentration in the apoplast could lead to stomatal closure through the activation of anion channels in the guard cell plasma membrane (Hedrich et al. 1994). There are few studies and controversial results concerning a putative interaction between ABA and CO2 sensing in regulating stomatal movements. Mansfield (1976) observed a total independence of the ABA and CO2 responses in Xanthium strumarium. In contrast, the absence of CO2 has been reported to inhibit the effect of externally applied ABA in the same species (Raschke 1975) and in Solanum melongena (Eamus & Narayan 1989). Raschke & Hedrich (1985) have shown that the sensitization of guard cells to CO2 by ABA may depend on species and on the degree of stomatal aperture. More recently, an ABA-enhanced response to CO2 has been observed in soybean (Bunce 1998).

Previous studies have shown that stomatal response to growth in high CO2 was highly variable, depending on species and experimental conditions (Drake, Gonzalèz-Meler & Long 1997). In the context of constant increase in atmospheric CO2, these interactions between ABA and CO2 sensing at the level of stomata must be better understood to accurately predict the effects of increasing CO2 concentrations on plant water use. Thus, the aim of this work was (i) to investigate the dependence of stomatal response upon osmotic stress and ABA in different CO2 concentrations in A. thaliana, a species of fundamental genetic interest (ii) to evaluate whether this relation was altered by the CO2 concentration prevailing during growth. To address these questions, changes in mean leaf conductance triggered by osmotic stress or ABA were monitored in A. thaliana by gas exchange techniques at different CO2 levels for plants grown at normal or double concentration of CO2. Additionally, dose–response curves of stomatal response to ABA from plants grown in normal or elevated CO2 were established in epidermal strip experiments.

MATERIALS AND METHODS

Plant material

Arabidopsis thaliana (L.) Heynh ecotype Landsberg erecta plants were grown individually in pots (65mm × 65mm × 70mm) of sand watered with half-strength Hoagland's solution (Epstein 1972) in a growth chamber (8 h light period, 22 °C; 16 h dark period, 20 °C; 70% relative humidity) at a CO2 concentration of 370 μmol mol–1 (control plants) or 740 μmol mol–1 (elevated CO2 plants). Light (250 μmol m–2s–1) was supplied by halogen lamps (HQI-TS, 150 W/NDL; Osram, München, Germany).

Whole plant gas exchange measurements

Four-week-old plants (leaf area: 6–8 cm2) were placed at a low vapour pressure deficit (VPD < 1 kPa) into an experimental chamber composed of independent shoot and root compartments with control of air humidity and temperature as previously described (Leymarie, Vavasseur & Lascève 1998). Water vapour pressures at the inlet and the outlet of the shoot compartment were continuously measured with two dew point hygrometers (Hygro M4; General Eastern, Woburn, MA, USA). Normal air (atmospheric air) was determined as containing 370 ±20 μmol mol–1 CO2. Different concentrations of CO2 (185, 555, 740 μmol mol–1) were obtained by adding pure CO2 to normal air or CO2-free air. The CO2 concentrations at the inlet and outlet of the plant chamber were monitored with a differential infra-red gas analyser (225 MK 3; ADC, Hoddesdon, UK). Light (8 h, 220 μmol m–2 s–1) was supplied by halogen lamps (HQI-TS 150 W/NDL; Osram, München, Germany). The leaf surface area was measured daily from enlarged photographs.

ABA (cis-trans, in methanol) and polyethylene glycol (PEG 3500) 6·7 or 10% w/w, resulting in a 0·16 and 0·24 MPa reduction in osmotic pressure, respectively (Wescor 5500; Wescor, Logan, Utah, USA), were added to the aerated nutrient solution 1 h after the onset of light. The methanol concentration in the solution never exceeded 0·001 v/v, and such a concentration in control experiments was found to have no effect on leaf conductance. One-third of the nutrient solution (70 mL) was changed twice a day.

Experiments with paradermal sections

Leaves from 4–5-week old plants (two plants for each experiment) grown either in normal air or in CO2-enriched air (740 μmol mol–1) were harvested in darkness at the end of the night period. Sections of abaxial epidermis tangential to the leaf surface, referred to as paradermal sections, were prepared using a razor blade and placed cuticle-up in Petri dishes containing a buffered solution [10 mM Mes/KOH; 10 mM KCl; 7·5 mM potassium iminodiacetate (K2IDA); pH 6·15] that was thermoregulated at 20 °C. The impermeant anion IDA was used to diminish the chloride concentration in the bathing medium since it has been shown that a high KCl concentration strongly reduces ABA and CO2 sensitivity in different species including A. thaliana (Wardle & Short 1981; Leymarie et al. 1998). Normal air or CO2-enriched air was continuously bubbled through the bathing medium (10 mM Mes/KOH; 10 mM KCl; 7·5 mM potassium iminodiacetate; pH 6·15). After 30 min in darkness, the paradermal sections were incubated 2 h under light (450 μmol m–2 s–1) at 20 °C. Then, ABA was added to the solution and measurements of stomatal apertures were performed after 3 h. The ABA was dissolved in methanol and the final methanol concentration in the solution never exceeded 0·001 v/v. The same amount of methanol was added to the controls. Stomata without underlying mesophyll contamination were observed at the edges of paradermal sections (Fig. 1) and used for measurements. Only ‘mature stomata’, whose ostiole length was higher than one-third of the length of stoma were taken into account. For each treatment at least 60 stomatal apertures (10 each from six different paradermal sections) were measured (Lascève, Leymarie & Vavasseur 1997). All experiments were replicated four times with different plants.

Figure 1.

. Open stomata from Arabidopsis thaliana; (m), ‘mature’ stomata; (*), ‘immature’ stomata.

Statistical analyses

The significance of each effect was tested by one-way analysis of variance (ANOVA), or, when populations were not normally distributed or exhibited unequal variances, by one-way ANOVA on ranks. To test significance of interactions, data were subjected to a two-way ANOVA. Pairwise comparisons were carried out using Tukey's test. Analyses were performed using Sigmastat 2·0 (Jandel Corporation, Chicago, IL, USA).

RESULTS

Changes in CO2 concentration and leaf conductance in A. thaliana

Figure 2 presents the response of leaf conductance to a transition from normal air to CO2 concentrations ranging from half to twice the normal level of CO2. Decreased CO2 resulted in a substantial increase in leaf conductance (88 ± 17 mmol m–2 s–1) lasting for at least the three following photoperiods. In contrast, a doubling of CO2 transiently reduced the leaf conductance (–47 ± 7 mmol m–2 s–1 after 2 h), and leaf conductance returned, during the following photoperiod, to the level measured in normal air (control values in Table 1). At elevated CO2, assimilation rate was increased at an almost constant level of transpiration, and a higher WUE was found: 3·4 ± 0·2 μmol CO2 mmol–1 H2O, compared with 2·4 ± 0·2 in normal air.

Figure 2.

. Effect of a change in CO2 concentration on leaf conductance of A. thaliana. Typical responses to a change from normal air to different CO2 concentrations ranging from 185 to 740 μmol mol–1. The numbers in front of dashed lines refer to the values of leaf conductance at the onset of light. Black and white bars correspond to night and day periods, respectively.

Table 1.  . Effect of an osmotic stress on leaf conductance, g (mmol m–2 s–1) and assimilation rate, A (μmol m–2 s–1) according to CO2 concentrations. Plants were placed for 2 d in normal or elevated CO2 (370 and 740 μmol mol–1, respectively) before being subjected to osmotic stress. Measurements were taken after 6 h exposure to light 1 d before (control) and 24 and 48 h after addition of PEG. Each value is the mean of at least three independent experiments ± SE Thumbnail image of

The response of leaf conductance to an osmotic stress is enhanced at elevated CO2

The leaf conductance of plants grown in normal air or placed for 2 d in elevated CO2 were not significantly different from each other (P = 0·70). However, their response to an osmotic stress of 0·16 or 0·24 MPa, resulting from the application of 6·7 or 10% PEG 3500 in the nutrient solution, was markedly different, with the plants placed in high CO2 showing the strongest decrease in leaf conductance (Table 1). Figure 3 illustrates the modification of leaf conductance of a plant submitted to a decrease of 0·24 MPa in the osmotic pressure in the nutrient solution. The osmotic stress resulted in a decrease in the basal level of leaf conductance in both CO2 concentrations but the daily change in leaf conductance was affected more significantly at elevated CO2 (Figs 3 & 4). A moderate osmotic stress (0·16 MPa), which has no detectable effect on the light-induced shift in leaf conductance in normal air, triggered a significant inhibition under elevated CO2 (Fig. 4). When all data were considered, a two-way analysis of variance (ANOVA) indicated a significant interaction between the 2-d PEG treatment and CO2 concentration (P = 0·031). Even for a mild stress (0·16 MPa), this interaction with CO2 was significant (P = 0·039). The carbon assimilation rate was not significantly affected by the 0·24 MPa osmotic stress in normal air (P = 0·45) but a significant reduction occurred at elevated CO2 (P = 0·019) resulting in almost similar levels of carbon fixation in both atmospheric CO2 conditions (Table 1). It is interesting to note that WUE during the water stress was higher at high CO2 (7·0 ± 0·8) than in normal air (3·2 ± 0·2).

Figure 3.

. Typical responses of leaf conductance to an osmotic stress (0·24 MPa), resulting from the application of PEG 3500 (10% w/w) in the nutrient solution, plants being previously placed for 2 d in normal air (dotted line) or in 740 μmol mol–1 CO2 (solid line). Black and white bars correspond to night and day periods, respectively.

Figure 4.

. Daily change in leaf conductance observed before and after a 2-d osmotic stress (PEG 3500: 6·7%, 0·16 MPa or 10%, 0·24 MPa), plants being already placed for 48 h in normal air or in elevated CO2. The daily change in leaf conductance is taken as the maximal conductance minus the minimal conductance recorded during a photoperiod (mean of at least three replicates ± SE).

The response of leaf conductance to ABA is enhanced at elevated CO2

Application of 1 μM ABA in the nutrient solution of plants previously placed for 2 d in different CO2 concentrations resulted in a drop in leaf conductance after a time lag of about 45 min. The amplitude of this drop was strongly correlated to the ambient CO2 concentration (Fig. 5, Table 2). While the response was only transient in normal air, with a rapid recovery to the initial level of leaf conductance, at high CO2 the application of ABA induced a permanent reduction in leaf conductance for at least the three following photoperiods. As observed during an osmotic stress, the elevated CO2 levels also resulted in a higher ABA-induced drop in leaf conductance (Fig. 6). Indeed, there was a significant interaction between a 2-d ABA treatment and CO2 concentration in leaf conductance regulation (P = 0·018). In contrast, the assimilation rates were not significantly affected by 1 μM ABA treatment (P = 0·123).

Table 2.  . Effect of ABA on leaf conductance (mmol m–2 s–1) according to CO2 concentration. Plants were placed for 2 d in the indicated CO2 concentrations before application of 1 μM ABA. Measurements were taken after 6 h exposure to light 1 d before (control) and 24 and 48 h after addition of ABA. Each value is the mean of at least three independent experiments ± SE. 740*, plants were grown from germination at 740 μmol mol–1 CO2Thumbnail image of
Figure 5.

. Typical response of leaf conductance to an application of 1 μM ABA in the nutrient solution, plants being previously placed for 2 d in different CO2 concentrations (185, 370, 555, 740 μmol mol–1 CO2) except 740* plants which were grown from germination at 740 μmol mol–1 CO2. Numbers in front of dashed lines refer to leaf conductance at the onset of light. Black and white bars correspond to night and day periods, respectively.

Figure 6.

. Daily change in leaf conductance before and 48 h after application of 1 μM ABA according to CO2 concentration. Plants were placed in the various CO2 conditions 2 d before ABA application except 740* plants which were grown from germination at 740 μmol mol–1 CO2 (mean of at least three replicates ± SE).

Growth at elevated CO2 does not modify stomatal response to ABA

Additional experiments with plants fully grown from germination at 740 μmol mol–1 CO2 were conducted at elevated CO2 levels (Figs 5 & 6). In the absence of stress, these plants exhibited a leaf conductance in the range of those found for control plants in normal air (P = 0·352, Table 2). At elevated CO2, the assimilation rates (A), intercellular space CO2 concentrations (Ci) and the A/Ci ratio for plants grown in normal and elevated CO2 concentrations were not significantly different (P > 0·05, Table 3), indicating no evidence of reduction of photosynthetic efficiency in response to CO2 growth conditions. Moreover, the response to an application of ABA was not significantly different for plants grown either in air or double CO2 (ABA–culture interaction P = 0·687). Thus, the CO2 concentration prevailing during the growth period apparently did not induce any adaptation of the response of leaf conductance to ABA. Stomatal responses to ABA and CO2 according to CO2 growth conditions were also investigated in epidermal strips bioassays. The CO2 concentration during growth did not significantly affect stomatal density (Table 4). Whatever the growth conditions, in the absence of ABA, the stomatal apertures measured under light in normal air or elevated CO2 were not significantly different (P > 0·6, Fig. 7). For both growth conditions, the response to low concentrations of ABA was higher in elevated CO2. The only noticeable difference was that plants grown in double CO2 were slightly less sensitive to ABA at high CO2 (Fig. 7, inset) but there was no significant interaction culture × measure (P > 0·05).

Table 3.  . Assimilation rate (A, μmol m–2 s–1) and A versus Ci (Ci in mmol mol–1) of plants grown in normal air or double CO2 before and after application of 1 μM ABA. All measurements were performed at 740 μmol mol–1 CO2Thumbnail image of
Table 4.  . Stomatal density (stomata mm–2) of plants grown at either 370 or 740 μmol mol–1 CO2. NS, not significant effect of growth condition on stomatal density. Values were obtained from the observation of 2200 stomata on 32 leaves from eight plants in each condition Thumbnail image of
Figure 7.

. Stomatal response to ABA in epidermal strip bioassays performed either in normal air (triangle) or in elevated CO2 (740 μmol mol–1, circle) with plants grown in normal air (open symbols) or CO2-enriched air (filled symbols). Inset, effect of elevated CO2 on stomatal response to ABA. ‘Delta high CO2’ means difference between stomatal aperture measured in normal and CO2-enriched air (dotted line, air grown plants; solid line, elevated CO2 grown plants). Data are the means of 240 measurements of stomatal apertures in four independent replicates. Error bar represent standard error of the mean with a confidence interval of 95%.

DISCUSSION

Stomatal response to CO2 in Arabidopsis

CO2 and ABA sensing in stomata is certainly located in the guard cell since guard cell protoplasts retain their ability to shrink in response to both signals (Gotow, Kondo & Syono 1982; Fitzsimons & Weyers 1986). In the integrated plant system, the stomata respond to the intercellular space concentration of CO2 determined by atmospheric concentration and by the assimilation rate of CO2 in the mesophyll (Mott 1988). On the basis of numerous experiments with various species, it is generally accepted that a doubling of atmospheric CO2 will result in an approximately 20–40% reduction in leaf conductance (Morison 1987; Beerling et al. 1996; Morison 1998). However, large variations in response to high CO2 have been observed according to the duration of exposition to CO2, the plant species (Bunce 1992; Beerling et al. 1996), and the growth conditions (Talbott, Srivastava & Zeiger 1996). In this study, while air-grown plants of A. thaliana were maintained in hydroponic conditions, we observed that lowering the CO2 concentration to half the normal concentration induced a permanent enhancement of leaf conductance. Conversely, a doubling of CO2 only provoked a transient 15% reduction in conductance followed by a rapid restoration of the initial level of leaf conductance. Similarly, high CO2 was poorly effective in inducing stomatal closure in bioassays conducted with epidermal strips from well-watered plants while in similar conditions the removal of CO2 induced a significant stomatal opening (Leymarie et al. 1998). Thus, in well-watered conditions, the stomatal behaviour was only slightly affected by a doubling of CO2 concentration.

Elevated CO2 enhances stomatal responses to osmotic stress and ABA

In contrast, the impact of high CO2 on stomatal conductance was more obvious after application of an osmotic stress. With plants transferred to elevated CO2, the short-term response to an osmotic stress was considerably enhanced compared with the response observed in normal air, revealing a strong interaction between CO2 sensing and water stress in stomatal regulation of gas exchanges in A. thaliana. These results underline the importance of the plant water status on stomatal sensitivity to CO2 and could at least partly explain the large variability reported on the effect of elevated CO2 on stomatal conductance. During drought, the sensitization of stomata to ABA by CO2 could enable the leaf conductance to follow more accurately changes in assimilation while saving water (Raschke 1987). It is interesting to note that when a short-term water stress was applied at high CO2, the leaf conductance was diminished by up to 67% while assimilation was only reduced by 25%. Such a limitation in leaf conductance at elevated CO2 allows a doubling of WUE.

ABA is certainly the major intermediate coupling sensing of water stress to stomatal closure, as demonstrated by the wilty phenotype of ABA-deficient mutant plants (Koornneef et al. 1982). In the present study, in whole plant experiments or in epidermal strip bioassays, high CO2 was able to enhance stomatal sensitivity to ABA. These observations strongly argue for a direct interaction between ABA and CO2 sensing at the guard cell level and fit well with the earlier observation of a CO2-enhanced ABA sensitivity in X. strumarium (Raschke 1975). The molecular basis of this interaction between responses to CO2 and ABA in guard cells remains to be elucidated, but calcium signalling in guard cells could well be a point of integration of these two signals. Supporting this hypothesis, exogenous ABA has been shown to trigger Ca2+ influx in guard cells (McAinsh, Brownlee & Hetherington 1990; Schroeder & Hagiwara 1990; Gilroy et al. 1991; McAinsh, Brownlee & Hetherington 1992; Lemtiri-Chlieh & MacRobbie 1994) and Ca2+ is also involved as a second messenger in the CO2 signal transduction pathway (Webb et al. 1996).

Growth at elevated CO2 and stomatal responses

In general, growth at elevated CO2 affects stomatal density (Poole et al. 1996). A survey has shown that among 100 species examined 74% exhibited a reduction in stomatal density in response to growth at elevated CO2 (Woodward & Kelly 1995). In the present study, there was no significant effect of a doubling of CO2 on abaxial stomatal density in A. thaliana. Even though this observation does not support a direct effect of high CO2 on stomatal patterning, a long-term effect based on a selective advantage cannot be excluded. Furthermore, there was no evidence of a modification of stomatal sensitivity to CO2 resulting from growth conditions since stomatal conductance and A/Ci at 740 μmol mol–1 CO2 were similar for both air-grown and adapted plants. Such absence of stomatal acclimation to high CO2 has been observed in Phaseolus vulgaris (Radoglou, Aphalo & Jarvis 1992). In contrast, changes in stomatal sensitivity to CO2 resulting from growth at high CO2 have been described (Morison 1998) leading to a decreased sensitivity in Eucalyptus tetrodonta and Chenopodium album (Berryman, Eamus & Duff 1994; Santrucek & Sage 1996) and conversely to an enhanced sensitivity in Triticum aestivum (Tuba, Szente & Koch 1994). Moreover, the interaction of ABA and CO2 sensing was not affected by growth conditions since stomatal responses of air-grown or elevated CO2-grown plants to an application of ABA were almost similar. These results suggest that CO2 enrichment could have a large impact in enhancing stomatal response to drought stress.

Elevated CO2 and photosynthesis

Many studies have shown that photosynthesis acclimates to long-term exposure to elevated CO2 (Drake et al. 1997), although this response is not universal (Curtis & Wang 1998). There are at present few data concerning CO2 acclimation in Arabidopsis and only recent studies have shown that prolonged exposure to elevated CO2 altered the levels of expression of sucrose phosphate synthase (Signora et al. 1998) and Rubisco (Cheng, Moore & Seemann 1998). However, even after a 10-week-culture period at elevated CO2, the maximal photosynthetic activity was only slightly affected (Signora et al. 1998). In the present work, assimilation rate, stomatal conductance, and intercellular CO2 concentration measured at double CO2 were very similar for air-grown or double CO2-grown plants. Additionally, the increase in assimilation rate resulting from a transition from normal air to double CO2 was maintained for at least a week (data not shown). Thus, there was no evidence for a reduction of photosynthesis under elevated CO2 in our study but a complete analysis of the A/Ci relationship would be required to fully address this point. Such an absence of acclimation of photosynthesis to elevated CO2 has been observed in cotton plants during the exponential growth phase (Wong 1993), in young Phaseolus vulgaris plants (Radoglou et al. 1992) or in spring wheat (Garcia et al. 1998).

The present rise in global atmospheric CO2 level is expected to have important consequences on plant behaviour through modifications of the climate, changes in photosynthetic activity and modification of stomatal conductance (Field, Jackson & Mooney 1995) and patterning (Woodward 1989). It is now evident that plant acclimation to elevated CO2 is highly varied, depending on species and environmental conditions. For A. thaliana, in well-watered or stress conditions, the major effect of a doubling of ambient CO2 that we noticed was a higher WUE. However, long-term studies should be conducted to integrate adaptive processes and adverse effects such as a diminished heat loss from transpiration (Eamus & Jarvis 1989; Eamus 1991; Jarvis 1995).

Acknowledgements

The authors wish to thank Dr Daniel Plante for critical reading of the manuscript.

Footnotes

  1. Present address: Max Planck Institut für Züchtungforshung, MDL, Carl-von-Linne-Weg 10, 50829 Köln, Germany.

Ancillary