Does engineering abscisic acid biosynthesis in Nicotiana plumbaginifolia modify stomatal response to drought?

Transcription

1 Plant, Cell and Environment (21) 24, Does engineering abscisic acid biosynthesis in Nicotiana plumbaginifolia modify stomatal response to drought? C. BOREL, 1 A. FREY, 2 A. MARION-POLL, 2 F. TARDIEU 1 & T. SIMONNEAU 1 1 Laboratoire d Ecophysiologie des Plantes sous Stress Environnementaux (LEPSE), UMR 759 INRA-ENSAM, 2 Place Viala, 346 Montpellier Cedex 1, France and 2 INRA, Laboratoire de Biologie Cellulaire, 7826 Versailles Cedex, France ABSTRACT The consequences of manipulating abscisic acid (ABA) biosynthesis rates on stomatal response to drought were analysed in wild-type, a full-deficient mutant and four under-producing transgenic lines of N. plumbaginifolia. The roles of ABA, xylem sap ph and leaf water potential were investigated under four experimental conditions: feeding detached leaves with varying ABA concentration; injecting exogenous ABA into well-watered plants; and withholding irrigation on pot-grown plants, either intact or grafted onto tobacco. Changes in ABA synthesis abilities among lines did not affect stomatal sensitivity to ABA concentration in the leaf xylem sap ([ABA] xyl ), as evidenced with exogenous ABA supplies and natural increases of [ABA] xyl in grafted plants subjected to drought. The ABAdeficient mutant, which is uncultivable under normal evaporative demand, was grafted onto tobacco stock and then presented the same stomatal response to [ABA] xyl as wild-type and other lines. This reinforces the dominant role of ABA in controlling stomatal response to drought in N. plumbaginifolia whereas roles of leaf water potential and xylem sap ph were excluded under all studied conditions. However, when plants were submitted to soil drying onto their own roots, stomatal response to [ABA] xyl slightly differed among lines. It is suggested, consistently with all the results, that an additional root signal of soil drying modulates stomatal response to [ABA] xyl. Key-words: Nicotiana plumbaginifolia; abscisic acid; drought; stomatal conductance; xylem sap ph. Abbreviations: ZEP, zeaxanthin epoxidase; WT, wild-type N. plumbaginifolia; zep, the ABA-deficient N.p. mutant, previously named aba2 (Marin et al. 1996); MS-3 and MS- 8, two transgenic lines obtained by independent complementations of N. p. zep with a full-length ZEP cdna in the sense orientation; AS-7 and AS-15, two transgenic lines obtained by independent transformations of WT N.p. with an antisense ZEP fragment; [ABA] xyl, concentration of ABA in the xylem sap extruded from pressurized leaf; Y L, Correspondence: Thierry Simonneau. Fax: ; simonneau@ensam.inra.fr leaf water potential; Y pre-dawn, leaf water potential measured at pre-dawn; g s, stomatal conductance. INTRODUCTION Stomatal control of transpiration is a major process determining short-term response of the whole plant to drought. It directly influences soil water depletion rates, plant water potentials and fluxes of solutes in the xylem stream (Tardieu & Simonneau 1998). Stomatal response to drought is mediated by root-born abscisic acid (ABA) (Zhang & Davies 1991) and it has been suggested that plant response to water stress could be improved by manipulating rates of ABA biosynthesis. However, the consequences of such manipulations remain difficult to predict for several reasons. The role of ABA in plant response to soil drying can be altered or overcome by other mediators. Hydraulic control of stomatal conductance by leaf water potential or xylem embolism predominates in some adult trees (Fuchs & Livingston 1996; Triboulot et al. 1996; Salleo et al. 2). In herbaceous isohydric species, such as maize, stomatal response to ABA concentration in the xylem sap ([ABA] xyl ) can also be altered by changes in leaf water potential (Tardieu & Davies 1992; Tardieu & Simonneau 1998). In addition to increases in [ABA] xyl, other droughtinduced changes in the composition of the xylem sap can control stomatal response to soil drying. This includes changes in the concentration of ABA-glucose-ester as a potential source of free ABA (Cornish & Zeevaart 1984; Lehmann & Glund 1986; Dietz et al. 2), decreased concentrations of hormones such as cytokinins (Blackman & Davies 1985; Incoll, Ray & Jewer 199; Fubeder et al. 1992; Badenoch-Jones et al. 1996) or auxins (Dunleavy & Ladley 1995), and possibly changes in ion concentrations (Schurr, Gollan & Schulze 1992; Ruiz, Atkinson & Mansfield 1993; Ridolfi et al. 1996). Apoplastic ph has also been shown to increase with the depletion of soil water (Wilkinson et al. 1998) and change in apoplastic ph was suggested to control stomatal response to ABA in the early development of a drought stress when there is no detectable change in [ABA] xyl. Such a role of increased apoplastic ph on stomatal closure was clearly demonstrated in vitro (Wilkinson & Davies 1997) and in planta by comparing wild-type tomato and the ABA-deficient mutant flacca (Wilkinson et al. 1998) but remains to be evaluated in planta for other species. 21 Blackwell Science Ltd 477

2 478 C. Borel et al. Manipulation of the ABA biosynthesis pathway may indirectly influence stomatal aperture and transpiration rate via pleiotropic effects on leaf growth or photosynthesis rate. The biosynthetic pathway of ABA has been amply clarified (Cutler & Krochko 1999) and involves important physiologically active intermediates such as carotenoids (Demmig-Adams & Adams 1996) that can induce noticeable pleiotropic effects, notably in plants engineered in the very early steps of ABA synthesis. As an example, transgenic plants with modified carotenoids composition can be affected in their tolerance to high light (Niyogi, Grossman & Bjorkman 1998) and in their stomatal response to blue light (Zeiger & Zhu 1998). Other enzymes, such as those altering the synthesis of a molybdenum cofactor (Leydecker et al. 1995), influence the terminal steps of ABA formation but are not specific to the ABA synthesis pathway. Abscisic acid itself acts on leaf growth (Bacon, Wilkinson & Davies 1998) which can in turn control transpiration rate, modify leaf water potential, and influence stomatal response to ABA (Tardieu & Simonneau 1998). The following contribution evaluates the consequences of manipulating ABA biosynthesis rates on stomatal response to drought. A series of transgenic lines presenting a range of mild reductions in ABA synthesis rate has been obtained in Nicotiana plumbaginifolia by manipulating expression of the zeaxanthin epoxydase (ZEP) gene (Audran et al. 1998). ZEP catalyses two of the final committed steps of ABA synthesis (Marin et al. 1996). Reduced ZEP expression in roots (Borel et al. 21) and leaves (Frey et al. 1999) resulted in three- to five-fold lower [ABA] xyl in transgenic N. plumbaginifolia than in wild-type (Borel et al. 21). Despite these reductions, major pleiotropic effects usually induced by manipulating ABA synthesis were avoided. Notably, direct consequences of genetic manipulations on stomatal response to drought were isolated from pleiotropic effects on leaf growth or photosynthesis rate (Marin et al. 1996; Borel 1999). A subset of N. plumbaginifolia transgenic lines together with wild-type and the full-deficient mutant zep were selected to examine in planta (i) if the rise in [ABA] xyl induced by soil drying accounted for changes in stomatal conductance in all lines, and (ii) whether altering ABA accumulation rates modified stomatal sensitivity to ABA. The relative contribution of ABA compared with other signals of drought on the reduction of stomatal conductance was tested by uncoupling changes in [ABA] xyl from that of soil water deficit. This was achieved by feeding exogenous ABA to whole plants or detached leaves. The same coupling between changes in xylem sap composition and soil drying was restored for all lines by grafting pairs of plants of different lines on the same N. tabacum stock. Grafting zep onto tobacco stock also enabled the analysis of its response to drought conditions, whereas ABAdeficient mutants cannot be submitted to drought stress unless grafted onto wild-type rootstock (Cornish & Zeevaart 1988; Fambrini et al. 1995). The combination of multiple experimental conditions (exogenous versus drought-induced ABA; detached leaves versus rooted plants; grafts on a common rootstock) together with the use of transgenic plants with altered rates of ABA synthesis offers a unique opportunity to evaluate the contribution of ABA in controlling stomatal response to soil drought. MATERIALS AND METHODS Gene constructions and plant transformations Nicotiana plumbaginifolia var. viviani and the stable zep (previously named aba2) mutant (aba2-s1, Marin et al. 1996) were transformed with different sense or antisense ZEP cdna constructs fused to a neomycin phosphotransferase sequence which conferred kanamycin resistance (Frey et al. 1999). Four transgenic lines together with wild type (WT) and zep were retained in this study: MS-3 and MS-8 were obtained by transformation of zep with a fulllength sense ZEP cdna; AS-7 and AS-15 were obtained by transformation of WT with a ZEP cdna in the antisense orientation. All ZEP constructs were placed between the CaMV 35S promoter and the nos3 end (details in Borel et al. 21). All transgenic lines contained a single copy of the ZEP construct (Borel 1999). Overall, all transgenic lines presented a reduction in their capacity to accumulate ABA in response to drought. This capacity has been evaluated for each line (except zep) by calculating the slope of the relationship between [ABA] xyl and leaf water potential measured at pre-dawn (Y pre-dawn ) on the same plant (Borel et al. 21). In comparison with WT, the capacity to accumulate ABA with drought was reduced to 28 6 ± 4% in MS-3 and MS-8, 28 9 ± 3% in AS- 7, and 22 4 ± 3% in AS-15 (Fig. 1). [ABA] xyl measured in well-watered plants was about three-fold lower in zep than in WT (Borel 1999), and dehydrating detached organs of zep did not increase ABA content either in leaves (Audran et al. 1998) or in roots (Borel et al. 21). Drought experiment with intact plants Seeds were germinated over a period of 4 to 6 weeks (25 C day/17 C night, 16 h photoperiod) on solid nutrient solution (agar 7 g L -1 in Hoagland N/1) (Borel 1999) with kanamycin for selection of transgenic plants. Green plantlets (2 3 of each line) were transferred to a greenhouse and transplanted into 6 dm 3 cylindrical pots ( 5 m high) in June 1996 (WT, MS-8), August 1996 (MS-3, AS-7, AS-15) and June 1997 (WT).The pots were filled with a 1 : 1 (v : v) mixture of sieved peat and clay-sandy-loam soil from a field near Montpellier (France), and they were wrapped in aluminium foil to minimize soil heating. Plants were watered daily with nutrient solution (Hoagland N/1) until the stem was about 5 mm high. Typical growth conditions in the greenhouse were 26 C (day)/18 C (night) mean air temperature, 7 9% mean daytime relative humidity and a mean daily cumulated photosynthetic photon flux density of 16 (for plants transplanted in August 1996 and measured in October 1996) to 21 mol m -2 d -1 (for plants transplanted in June 1996 and measured in August 1996). 21 Blackwell Science Ltd, Plant, Cell and Environment, 24,

3 Stomatal control in plants with modified ABA synthesis rates 479 Irrigation was withheld at the onset of flowering (when the stem was about 4 m high), and the soil was covered with white perlite to avoid heating and direct evaporation. Each day during soil drying, stomatal conductance (g s ) of three to five plants per line was measured on the twelfth leaf (first above the rosette) from 11 to 14 h (solar time) using a ventilated closed-cuvette (volume 1 dm 3 ; contact area 5 mm 2 ) coupled to a gas-exchange analyser (LI-62; LI-COR, Lincoln, NE, USA). Relative humidity inside the cuvette was maintained constant during measurement. Care was taken to ensure direct light exposure of the whole leaf for at least 3 min before and during measurement. Photosynthetic photon flux density (PPFD) was measured using the PPFD sensor (LI-19SA; LI-COR) of the LI-62 chamber, and ranged from 8 to 12 mmol m -2 s -1 during measurements. The boundary layer conductance in the cuvette was determined every second hour by inserting a replica of the measured leaves made of wet blotting paper. Immediately after the measurement of stomatal conductance, the leaf was excised, enclosed in a pressure chamber (Soil Moisture Equipment Corp., Santa Barbara, CA, USA) for measurement of leaf water potential (Y L ), and approximately 7 mm 3 of xylem sap were extracted by pressurizing the leaf at about 4 MPa above the balancing pressure. The sap was stored at -8 C for subsequent analysis of ABA and ph determination. Plants were then placed in a dark room until the following morning when pre-dawn leaf water potential was measured on a leaf adjacent to that used for the measurement of stomatal conductance. Some plants were depotted to check that roots had developed in the whole pot. An additional experiment was performed with 15 WT and 2 AS-15 plants that were transplanted in April 1998 and grown as above except that they were acclimated to an increased level of xylem ABA. From the seventh leaf appearance, daily irrigation was complemented with 2 mmol m -3 of (±)-ABA (Fluka, Buchs, Switzerland) added to the nutrient solution. At the end of the thirteenth leaf expansion (about 4 weeks later) irrigation was withheld, and measurements of g s, Y L, and extraction of xylem sap were performed at different levels of soil drying on the thirteenth leaf as described above. Drought experiment on plants grafted onto wild-type tobacco stocks Three successive experiments were performed on grafted plants in the greenhouse. In each experiment, a graft of one transgenic line among MS-8, AS-15 and zep was associated with WT (15 replicates of each graft type) onto a common Nicotiana tabacum cv. Xanthi (tobacco) stock plant. Tobacco was sown and grown in 6 dm 3 cylindrical pots ( 5 m high) filled with the peat : soil mixture described above, whereas N. plumbaginifolia plantlets were grown in 1 5 dm 3 pots filled with peat. All the plants were watered daily until the grafting. When the tobacco plants reached the flowering stage, their single shoots were trimmed below ABA concentration in xylem sap (mmol m 3 ) r 2 =. 9 r 2 =. 72 r 2 =. 88 r 2 = Pre-dawn leaf water potential (MPa) Figure 1. Relationship between pre-dawn leaf water potential and concentration of ABA in leaf xylem sap measured before dawn on pot-grown Nicotiana plumbaginifolia submitted to soil drying. Each point represents coupled measurements on one leaf. Wild-type; complemented zep mutants (MS-3, closed symbols; MS-8, open symbols); antisense plants (AS-7, closed symbols and full line; AS-15, open symbols and dashed line). Straight lines and coefficients of correlation were obtained by linear regression for each line [except in, where the regression holds for both MS-3 and MS-8]. Dotted lines are intervals of confidence at 95% [in, only the upper limit was shown for AS-7, and the lower one for AS-15]. the tenth leaf and grafting was performed by incising the stem of tobacco at two locations, 1 mm below the fourth and tenth leaves, and inserting one plantlet of N. plumbaginifolia into each incision (a wild type and a transgenic plant). At this time the N. plumbaginifolia plantlets were four to five visible leaves. Grafts were wrapped in wet 21 Blackwell Science Ltd, Plant, Cell and Environment, 24,

4 48 C. Borel et al. tissues and covered by plastic bags for 7 d, after which leaves of the stock plant were trimmed and grafts developed without protection. Six to 1 weeks after the grafting, grafts presented seven to 1 newly grown and fully expanded leaves. Plants were then subjected to drought by withholding irrigation as described above for intact plants. Measurements of stomatal conductance, of Y L, and extraction of xylem sap for ABA analysis were carried out as described above, at varying soil dehydration levels, on the youngest fully expanded leaf (fifteenth or sixteenth) of each graft. ABA feeding into the xylem sap of well-watered plants Plants of WT, MS-8 and AS-15 were prepared as described above for intact plants, except that the pots (12 dm 3 ) were filled with peat, they contained three plants (one of each line) and were continuously irrigated (eight times a day) until the morning of experiments, when ABA was fed to the plants in the greenhouse. During growth, plants were trimmed to allow for the development of a unique stem. On the morning prior to measurements, a plastic funnel (1 mm high, 1 mm in diameter) was sealed with epoxy resin around the basis of the stem, and filled with degassed buffer solution ( 4 mol m -3 Ca(NO 3 ) 2, 2 mol m -3 KH 2 PO 4, ph 6 ) without ABA or with varying concentrations ( 1 8 mol m -3 ) of synthetic (±)-ABA (Fluka). The stem was drilled radially with a 1-mm-diameter bit below the surface of the solution to avoid cavitation. The funnel was covered to protect the solution against light and pollutants. The plants were then submitted to high PPFD (above 8 mmol m -2 s -1 ) for 2 h in the greenhouse to allow absorption of the ABA solution into the xylem stream. Measurements of stomatal conductance, of Y L, and extraction of xylem sap for ABA analysis were carried out on the first leaf above the funnel as described for the drought experiment. Detached-leaf experiments Leaves were sampled on well-watered plants (WT, MS-3, AS-7 and AS-15) that were transplanted in August 1996 and grown in the greenhouse in 1 5 dm 3 pots filled with peat (one plant per pot). When the plants were at the onset of flowering stage ( 4 m high stems), about 15 pots for each line were placed in a dark room. On the next day, the twelfth leaf of 1 plants per line was cut under water in reduced light conditions, and immediately placed into plastic vials ( 1 dm 3 ) containing 4 g of the degassed buffer solution described above, with varying concentrations (, 1, 2, 5, 1 mol m -3 ) of synthetic (±)-ABA (Fluka). Concentrations of (+)-ABA were calculated as half of the applied concentrations of (±)-ABA. Vials were sealed with foam and covered with parafilm. Leaves in the vials were maintained inclined at 45 to the horizontal, and exposed to direct light in the glasshouse (6 mmol m -2 s -1 PPFD, 27 C daytime air temperature, and about 5% relative humidity) or placed 7 m below metal halide lamps in the laboratory (8 mmol m -2 s -1 PPFD, 28 C air temperature, 4% relative humidity). Every 2 min during a 3 h period, vials were weighed and leaf temperature was measured using an infrared thermometer (IR-747 THI-3, Tasco, Osaka, Japan). Air temperature and relative humidity were measured every 2 s (HMP35A, Vaisala, Finland), averaged and stored in a data-logger every 6 s. The boundary layer conductance (g a ) was determined by measuring water evaporation from three wet replicas of the leaves made from blotting paper and using the same gravimetric method as for the determination of leaf evaporation rate. Areas of leaves and leaf replicas were measured at the end of the experiment. Stomatal conductance was calculated using the equation: J A = M T[ e( T )- e ] g g ( g + g ) w w L a a s a s where J w is the leaf transpiration rate (kg s -1 ), A is the leaf area (m 2 ), e(t L ) is the saturation vapour pressure at leaf temperature (Pa), e a is the water vapour pressure in the bulk air (Pa), T is the air temperature (K) and is the gas constant. ABA analysis and ph of the xylem sap Xylem ph was measured in sap samples with a microelectrode (INGOLD: Mettler-Toledo, Nänikon, Switzerland) after thawing and equilibration to room temperature. ABA concentration was then analysed in crude samples of xylem sap by radio-immunoassay (Quarrie et al. 1988) as previously described (Barrieu & Simonneau, 2). Specificity for ABA of the monoclonal antibody (MAC 252, provided by Dr S.A. Quarrie, Cambridge Laboratory, John Innes Centre, UK) was verified in xylem sap of N. plumbaginifolia by comparing radio-immunoassay of crude sap samples with radio-immunoassay of sap fractions recovered from thin layer chromatography (Borel 1999). Statistical analysis Fitting of non-linear relationships (stomatal response to [ABA] xyl ) and parameter estimation were carried out using the least squares method (GRG2 algorithm, Excel; Microsoft, Redmond, WA, USA). Differences among lines were tested by comparing the residual sums of squares for the individual fittings to each line, to the residual sum of squares for the common fitting to the whole of the lines, using an F-test (Borel et al. 21). RESULTS Common response of stomatal conductance to exogenous ABA was observed for whole plants and detached leaves of all lines Stomatal response to exogenous ABA was studied in wellwatered conditions in both whole plants and detached leaves. Perfusing well-watered plants with deionized and 21 Blackwell Science Ltd, Plant, Cell and Environment, 24,

5 Stomatal control in plants with modified ABA synthesis rates 481 degased water (with no ABA) hardly modified the maximal value of stomatal conductance observed in intact plants just before perfusion ( 5 mol m -2 s -1 compared with 53 mol m -2 s -1 ). Feeding detached leaves with artificial sap (without ABA) resulted in slightly lower values than in intact plants with differences among lines ( 45 in WT, 32 in MS-3, 38 in AS-7 and 55 mol m -2 s -1 in AS-15). For further comparisons, relative stomatal conductance was calculated as the ratio of g s to maximal g s, the latter being determined for each line within each experiment in wellwatered conditions and without applied ABA. Supplying exogenous ABA to whole plants greatly increased the concentration of ABA in the xylem sap. The [ABA] xyl measured at the same time as stomatal conductance in ABA-fed plants increased up to 9 mmol m -3 whereas it was about 1 4 mmol m -3 in well-watered plants without exogenous ABA. Stomatal conductance negatively correlated with [ABA] xyl and was minimal when [ABA] xyl exceeded 2 mmol m -3 (Fig. 2a c). The response of stomatal conductance to exogenous ABA was similar in all lines. A unique decreasing exponential relationship fitted all data without significant difference among lines (P < 5) (Fig. 2). Similarly, although with more data scattering, detached leaves of the four studied lines presented a common decrease in relative g s as the concentration of ABA increased in the feeding solution (Fig. 3a c). Although data scattering hindered statistical analyses, the response of relative g s to exogenous ABA appeared similar for detached leaves and perfused plants. However, for all lines, stomatal closure at the highest ABA concentration was less pronounced in detached than in intact leaves. Stomatal conductance thus appeared slightly less sensitive to high [ABA] xyl in detached leaves (notably that of antisense lines) than in perfused plants (Figs 2 & 3). Grafts of WT and transgenic lines onto tobacco stocks, presented a common response of stomatal conductance to changes in [ABA] xyl induced by soil drying Relationships between g s and [ABA] xyl were analysed in three successive experiments performed on grafted plants. In each experiment, one transgenic line among MS-8, AS- 15 and zep was associated with WT on a same tobacco rootstock. The maximal value of g s measured in well-watered conditions differed among experiments but was similar between WT and the transgenic line within each experiment (1 mol m -2 s -1 when MS-8 and WT were grafted together, and 6 mol m -2 s -1 in all other experiments). In all cases, a common exponential decrease in g s (relative to maximal g s ) with increasing [ABA] xyl was noticed for all transgenic lines and WT (Fig. 4a c). A unique relationship was fitted without significant difference among lines (P < 5). Interestingly, zep, which could not be subjected to drought on its own roots, presented the same response of relative g s to [ABA] xyl as the other lines (Fig. 4c) when it was grafted onto tobacco stock. Stomatal conductance (% of control with no ABA) [ABA] xylem (mmol m 3 ) Figure 2. Stomatal conductance as a function of (+)-ABA concentration in leaf xylem sap ([ABA] xylem ) of well-watered Nicotiana plumbaginifolia fed with exogenous ABA. A funnel was sealed around the stem below the last fully expanded leaf, and filled with artificial sap containing ABA. Artificial sap was absorbed via a 1 mm hole drilled in the stem several hours before measurements in the greenhouse. Stomatal conductance (g s ) was expressed as percentage of the mean maximum value ( 53 mol m -2 s -1, mean of all lines) measured on plants perfused with degassed water containing no ABA. Wild-type; complemented zep mutant MS-8; antisense line AS-15. The solid line [same in, and ] was obtained by non-linear fitting to the whole set of data (individual fittings for each line were not significantly different, P < 5) and corresponds to the following equation: g s (% of control) = exp( [ABA] xylem ). 21 Blackwell Science Ltd, Plant, Cell and Environment, 24,

6 482 C. Borel et al. 1 Stomatal conductance of intact plants submitted to soil water deficit correlated with [ABA] xyl for all lines, but with a slightly higher sensitivity in transgenic lines than in WT Stomatal conductance (% of control with no ABA) Moreover, the response of relative g s to [ABA] xyl was similar in grafts subjected to soil water deficit onto tobacco rootstocks (Fig. 4) and in well-watered plants supplied with exogenous ABA (Fig. 2) ABA concentration (mmol m 3 ) Figure 3. Stomatal conductance of detached Nicotiana plumbaginifolia leaves as a function of concentration of (+)-ABA in assay solutions. The solutions were synthetic ABA in artificial xylem sap. Stomatal conductance is expressed as percentage of maximal value determined for each line on plants fed with artificial xylem sap only (no ABA). Values are means ± standarderrors of four observations on independent leaves. Wild-type; complemented zep mutant MS-3; antisense plants (AS-7, closed symbols; AS-15, open symbols). The solid line [same in, and ] was obtained by non-linear fitting to the whole set of data and corresponds to: g s (% of control) = exp( [ABA]) (no significant difference was observed between individual fittings to each line, P < 5). In experiments with intact plants, maximal g s measured in well-watered conditions differed among WT and transgenic lines: 1 55 mol m -2 s -1 for WT, 55 mol m -2 s -1 for MS-3 and MS-8 (without significant difference between both lines), 1 mol m -2 s -1 for AS-7, and 1 3 mol m -2 s -1 for AS- 15.As the soil dried, Y pre-dawn decreased from - 2 to about -1 MPa and g s decreased from maximal values to less than 1 mol m -2 s -1. A decrease in Y pre-dawn down to - 6 MPa induced stomatal closure in all lines. Concomitantly, mid-day [ABA] xyl increased up to 8 mmol m -3 in WT although it remained lower than 3 mmol m -3 in all transgenic lines. Stomatal conductance correlated with [ABA] xyl in all lines (Fig. 5a c). However, up to two-fold differences appeared among transgenic lines and WT (P < 5). Under-producing lines exhibited a higher sensitivity of g s to [ABA] xyl than WT. This conclusion was not statistically supported in MS-8, where scattering of data was observed for low values of [ABA] xyl (Fig. 5b). A higher stomatal sensitivity to [ABA] xyl was apparent in AS-7 and AS-15 for which almost all data points were distributed below the mean response of WT (Fig. 5c). Overall, half stomatal closure was observed at about 4 mmol m -3 ABA in the xylem sap of antisense transgenic lines, compared with 8 mmol m -3 in WT. For WT and AS-15, the experiments were repeated under two contrasted evaporative demands that varied with air vapour pressure deficit: 9 and 2 2 kpa on two separate experiments with WT, and 9 and 3 1 kpa on two other experiments with AS-15. In both lines, the same relative response of g s to [ABA] xyl was obtained (not detailed) whatever the evaporative demand although minimal values of g s (less than 1 mol m -2 s -1 ) were obtained at different times from the last irrigation (4 d with the lowest evaporative demand as opposed to 7 d with the highest evaporative demand). Acclimating intact plants to elevated [ABA] xyl did not influence stomatal response to [ABA] xyl An experiment was performed on WT and AS-15 where plants were irrigated with 1 mmol m -3 of (+)-ABA during the development of the studied leaf (thirteenth from the plant basis). At the end of this acclimation period, the increase in [ABA] xyl measured at mid-day was significant in WT (36 ± 2 mmol m -3 in treated plants compared with 18 ± 4 mmol m -3 in control plants) but not significant in AS- 15 (26 ± 4 mmol m -3 ).Acclimation was followed by a period of soil drying without irrigation during which stomatal response to drought-induced ABA was studied in acclimated leaves. Maximal g s in acclimated plants was lower than in plants cultivated without ABA. However acclimation to elevated ABA did not affect the relative response 21 Blackwell Science Ltd, Plant, Cell and Environment, 24,

7 Stomatal conductance (% of well-watered control ) Stomatal conductance (% of well-watered control ) [ABA] xylem (mmol m 3 ) [ABA] xylem (mmol m 3 ) Figure 4. Stomatal conductance (g s ) as a function of ABA concentration in the leaf xylem sap ([ABA] xylem ) of grafted Nicotiana plumbaginifolia submitted to soil drying onto potgrown tobacco stocks. Each point represents coupled values of g s and [ABA] xyl measured on one leaf around midday in the greenhouse. Stomatal conductance was expressed as a percentage of the mean maximal value measured on grafts of each line onto well-watered tobacco stocks. Wild-type; complemented zep mutants MS-8; antisense plants AS-15 (open symbols) and zep (stars). Non-linear fitting did not significantly differed among lines (P < 5). The solid line [same in, and ] represents the non-linear fitting obtained in Fig. 5a for intact WT plants (not significantly different, P < 5) with the corresponding equation: g s (% of control) = exp( [ABA] xylem ). Figure 5. Stomatal conductance (g s ) as a function of ABA concentration in the leaf xylem sap ([ABA] xylem ) of intact potgrown Nicotiana plumbaginifolia submitted to soil drying. Each point corresponds to coupled values of g s and [ABA] xylem measured on the same leaf around midday in the greenhouse. Wild-type; complemented zep mutants (MS-3, closed symbols; MS-8, open symbols); antisense plants (AS-7, closed symbols; AS-15, open symbols). Cross-symbols in and correspond to measurements on plants irrigated with ABAenriched solution in order to increase [ABA] xylem during the whole development of the measured leaf. Stomatal conductance (g s ) was expressed as percentage of the mean maximum value determined for each line on well-watered plants (1 55 mol m -2 s -1 for wild-type, 5 for MS-3 and MS-8, 1 for AS-7, and 1 3 for AS-15). Three experiments (three sowing dates) were merged for wild-type and AS-15 (one sowing date for each other line). Exponential fitting was significantly different among lines (except MS-8 where data were scarce and scattered). For sake of comparison, the same solid line (fitted to WT data set) was shown in, and corresponding to: g s (% of control) = exp( [ABA] xylem ). 21 Blackwell Science Ltd, Plant, Cell and Environment, 24,

8 484 C. Borel et al. of g s to drought-induced ABA, either in WT and AS-15 (Fig. 5a & c, crossed symbols). In particular, acclimating the low-aba line AS-15 (Fig. 1) to elevated ABA levels did not decrease its stomatal sensitivity to [ABA] xyl. 1 The relationship between g s and [ABA] xyl was independent of Y L and of xylem sap ph Leaf water potential (Y L ) measured at the same time as g s on intact plants submitted to soil drying varied from - 7 to -1 4 MPa (Fig. 6d) with a concomitant decrease in Y pre-dawn from - 2 to -1 MPa. No significant differences in Y L appeared among lines when dehydrated to a given Y pre-dawn and experimented in similar conditions (i.e. sown and measured on the same day) with the exception of MS- 3 and MS-8, which presented lower but also very scattered values of Y L (Fig. 6d). The differences in Y L were more pronounced for a given line among experimental conditions than among lines in a given experimental condition. Notably, feeding detached leaves with artificial sap induced stomatal closure and consequently raised Y L from about - 8 to - 2 MPa (Fig. 6a & b). Intermediate Y L (-1 1 to - 5 MPa) was observed in plants grafted onto tobacco and submitted to soil drying (Fig. 6c). Contrary to intact plants, grafts of MS-8 exhibited slightly but not significantly higher Y L than did other lines. Within each experiment, stomatal conductance correlated with leaf water potential (Fig. 6). However, the relationship between g s and Y L was loose and varied considerably among experiments. Notably, in one experiment on WT plants that had been subjected to drought, all the leaves but one (twelfth) of half the plants were wrapped in aluminium foil. This increased Y L by approximately 2 MPa in comparison with control plants without an aluminium envelope (Fig. 7b).The relationship between g s and Y L notably differed between wrapped and control plants (Fig. 7b). Furthermore, in plants that had been subjected to drought (either intact or grafted, Fig. 6c & d), g s was positively correlated with Y L whereas it correlated negatively with Y L in ABA-fed plants (Fig. 6a). This indicated that g s was not directly controlled by Y L and that the correlation between g s and Y L was the result of their concomitant variation, which occurred differently in plants subjected to drought and in ABA-fed plants. Soil drying decreased Y L at the same time as g s declined. In contrast, feeding ABA in well-watered plants induced decreases in g s and hence reduced water flux through the plant, leading to a rise in Y L up to the value observed at pre-dawn in well-watered conditions (Fig. 6a). The effects of Y L and xylem sap ph on stomatal response to [ABA] xyl were further tested in WT and AS-15 by comparing the relationships between g s and [ABA] xyl for different intervals of Y L and of xylem sap ph (Fig. 8). Both in WT and AS-15, the relationship between g s and [ABA] xyl was similar whatever the values of Y L (from -1 2 to - 5 MPa) and of xylem sap ph (from 5 5 to 6 8) (Fig. 8). In AS-15, relationships fitted for the different Y L intervals did not differ (P < 5, Fig. 8b). Stomatal conductance of Stomatal conductance (% of control with no ABA) Stomatal conductance (% of well-watered control) (d) Leaf water potential (MPa) Figure 6. Relationship between stomatal conductance (issued from Figs 2 5) and leaf water potential measured at the same time (midday) on the same leaf of Nicotiana plumbaginifolia. Whole plants supplied with exogenous ABA; detached leaves infused with exogenous ABA; plants grafted onto tobacco submitted to soil drying; (d) intacts plants submitted to soil drying onto their own roots. Each symbol represents a different line (as in Figs 1 5). wild-type, open circles; MS-3, closed triangles; MS-8, open triangles; AS-7, closed squares; AS-15, open squares and zep, closed stars. WT appeared to be slightly less sensitive to [ABA] xyl for intermediate Y L (from - 8 to - 7 MPa), but the response of g s to [ABA] xyl was unique for other intervals of Y L including intervals below - 8 and above - 7 MPa (Fig. 8a). Xylem sap ph and Y L varied in the same ranges for WT and AS Blackwell Science Ltd, Plant, Cell and Environment, 24,

9 Stomatal control in plants with modified ABA synthesis rates 485 Stomatal conductance (mol m 2 s 1 ) Days from last irrigation Leaf water potential (MPa) [ABA] xylem (mmol m 3 ) Figure 7. Stomatal conductance of wild-type N. plumbaginifolia plotted as a function of the time since irrigation was withheld; midday leaf water potential; and xylem ABA concentration ([ABA] xyl ). Stomatal conductance was measured around midday on a transpiring leaf while the rest of the plant was either also submitted to evaporative demand (closed symbols), or not transpiring (wrapped in aluminium foil, open symbols).. 7 < Y L <. 5 MPa. 8 < Y L <. 7 MPa. 9 < Y L <. 8 MPa 1. 2 < Y L <. 9 MPa 6. 6 < ph < < ph < < ph < < ph < < ph < Stomatal conductance (mol m 2 s 1 ) (d) [ABA] xylem (mmol m 3 ) [ABA] xylem (mmol m 3 ) Figure 8. Relationship between stomatal conductance and ABA concentration in leaf xylem sap ([ABA] xylem ) of pot-grown Nicotiana plumbaginifolia submitted to soil drying (same data as in Fig. 5a & b, except when leaf water potential or xylem sap ph were not determined). Several classes of leaf water potential (Y L ) (a, b) and xylem sap ph (c, d) were distinguished. Each point represents values of stomatal conductance, [ABA] xylem, Y L and xylem sap ph, from one leaf of wild-type (a, c) or the antisense line AS-15 (b, d). 21 Blackwell Science Ltd, Plant, Cell and Environment, 24,

10 486 C. Borel et al. WT response of stomatal conductance to [ABA] xyl was the same with various sources of ABA and under contrasting evaporative demands The response of relative g s to [ABA] xyl was similar for WT plants either intact (Fig. 5a), grafted onto tobacco (Fig. 4a) or fed with exogenous ABA (Fig. 2a). Additionally, g s response of intact plants to drought-induced ABA was conserved under various evaporative demands, whether it was varied by measuring plants on days with contrasting vapour pressure deficits (from 8 to 2 3 kpa), or by wrapping all leaves but one with aluminium foil (Fig. 7). A 5% decrease in stomatal conductance was observed at about 8 mmol ABA m -3 in the xylem sap whatever the treatment (grafted or intact plants, various evaporative demands) and the origin of ABA (artificially fed or drought-induced). DISCUSSION Manipulating capacities to synthesize ABA did not affect stomatal sensitivity to ABA in detached leaves and whole plants fed with ABA and in grafts submitted to soil drying onto tobacco stocks For all lines and for these three experimental conditions, relative g s always correlated with [ABA] xyl following a unique relationship (Figs 2 4). This was independent of the origin of ABA, either exogenously fed to intact plants and detached leaves, or endogenously supplied to grafts by the tobacco rootstocks in response to soil drying. The present study reports the first case where manipulation of ABA biosynthesis is shown not to influence intrinsic sensitivity of stomatal conductance to xylem ABA. Similar conclusions were suggested with reciprocal grafts of wild-type sunflower and the low-aba mutant w-1 (Fambrini et al. 1995). This excludes any strong genetic or physiological linkage between the capacity to synthesize ABA and the stomatal sensitivity to ABA, suggesting that both processes may be manipulated independently. Consistently, genes involved in stomatal response to ABA on one hand, and in the control of ABA synthesis rate on the other hand, continue to be cloned in abi and low-aba genotypes (Cutler & Krochko 1999; Li et al. 2) and to date, the function of these genes do not support any interaction between both processes. In the same way, plant capacity to synthesize ABA presumably does not interact with stomatal sensitivity of Arabidopsis to air humidity since the same sensitivity to air humidity has been evidenced in wildtype and the ABA-deficient mutant aba1 (Assmann, Snyder & Lee, 2). By contrast, stomatal insensitivity to light has been observed in ABA-deficient mutants (Nagel, Konings & Lambers 1994) including non-grafted zep whose stomata remain opened in darkness (Borel 1999). This abnormal phenotype was suggested to result from undeveloped stomata lacking a functional closing mechanism (Popova & Riddle 1996). Delivery of sufficient amounts of ABA during leaf development in ABA-deficient mutants is possibly needed for the restoration of a normal stomatal sensitivity to light and other environmental conditions. In the present study, stomatal sensitivity of the ABA-deficient zep mutant was analysed on leaves that received slight amounts of ABA from the tobacco rootstock throughout their development and therefore may differ from stomatal sensitivity of non-grafted zep plants. Relative g s (calculated as the ratio of actual to maximal g s measured on well-watered plants of each line within each experiment) was used to compare lines in different experiments because of the variability in maximal g s among experiments. The exact origin of this variability remains to be elucidated although it probably resulted from differences in leaf age across experiments and lines. Rapid decrease in stomatal conductance with leaf age has been evidenced after leaf growth ceased (Solarova 198) and such a behaviour has been observed in N. plumbaginifolia (not shown). Differences in vapour pressure deficit among experiments also slightly changed g s of well-watered plants, but with a much lower effect than leaf position and/or leaf age (Borel 1999). By contrast, it is unlikely that differences in transpiration rates at the plant level influenced maximal g s : covering all the leaves but one did not significantly modify the g s of well-watered plants (1 34 ± 13 mol m -2 s -1 in seven WT plants with only one leaf transpiring, compared with 1 35 ± 14 mol m -2 s -1 in 11 control plants). Whatever the cause of variability in maximal g s across experiments, the use of relative g s could not have biased comparisons between lines in all conditions. Notably, in experiments with the grafted plants, maximal g s was similar for all lines in a given experiment associating one of the transgenic line and the wild-type, and then comparison in absolute and relative g s led to the same conclusion. Additionally, expressing stomatal conductance relative to the maximal value observed in well-watered plants resulted for the wild-type in a unique relationship between relative g s and [ABA] xyl in all experimental conditions, including intact plants. This strongly supports the use of relative g s to compare experiments with contrasting maximal g s. Similar responses of stomatal conductance to [ABA] xyl among lines and under contrasting conditions reinforce the key role of ABA in controlling g s response to drought in N. plumbaginifolia In plants submitted to soil drying, leaf water potential, [ABA] xyl, and xylem sap ph measured either at pre-dawn or at the same time ( mid-day ) as g s, all co-evolved with the decrease in stomatal conductance, but only mid-day [ABA] xyl correlated with relative g s with a unique relationship for contrasting experimental conditions (Fig. 7). A unique relationship between mid-day [ABA] xyl and relative g s also held for wild-type leaves whether they were attached to plants grafted onto tobacco stocks that had been subjected to drought, or attached to intact plants and submitted to changes in stress-induced or exogenous ABA (Figs 2a, 21 Blackwell Science Ltd, Plant, Cell and Environment, 24,

11 Stomatal control in plants with modified ABA synthesis rates 487 3a & 4a). The relationship was still conserved (i) among lines despite contrasting time-courses of [ABA] xyl after withholding irrigation, and (ii) for WT plants whether they transpired from all or only one of their leaves (Fig. 7). This makes it unlikely that any other component of the xylem sap significantly influences the control of g s by ABA in N. plumbaginifolia. Consistently, no effect of xylem ph was detected in stomatal response to [ABA] xyl of WT and AS-15 (Fig. 8c & d). Xylem ph weakly varied with drought in N. plumbaginifolia, compared for example with the two-unit change induced by drought in tomato plants (Wilkinson et al. 1998). This may explain why ph influence on stomatal conductance in the present experiments with N. plumbaginifolia was not as important as suggested in tomato (Wilkinson et al. 1998) or in Commelina communis (Thompson et al. 1997). However high concentrations of exogenous ABA failed to reduce stomatal conductance of detached leaves to the same extent as comparable concentrations of endogenous ABA did in plants grafted onto tobacco or in intact plants. In this latter case, the whole composition of the xylem sap, which originated from the tobacco stock or the intact roots, was influenced by the drought treatment, whereas only [ABA] xyl was varied in ABA-fed leaves. Thus, any chemical signal of soil drying, which may have accompanied the action of ABA on stomatal conductance can explain why ABA-fed leaves were less sensitive to ABA than plants subjected to drought. The role of Y L, which modulates stomatal sensitivity to ABA in maize (Tardieu & Davies 1992) and apparently controls stomatal conductance in some woody species (Fuchs & Livingston 1996), was clearly excluded in N. plumbaginifolia (Fig. 8a & b). Co-evolutions of Y L and g s during drought led to relationships between these two variables (Fig. 6c & d), but the relationship notably changed when evolutions of g s and Y L were uncoupled. This was achieved in WT by covering all but one leaf of plants that were subjected to drought (Fig. 7b), or by injecting ABA in all lines (Fig. 6a & b). In WT, g s correlated with [ABA] xyl following a unique relationship for all treatments except detached leaves (Figs 2a, 4a, 5a & 7c), whereas relationship between g s and Y L notably differed between intact and ABA-fed plants (Fig. 6) and between wrapped and control plants (Fig. 7b). Strictly, this conclusion should be restricted to the experienced range of Y L which may appear limited (- 5 to -1 2 MPa) compared with other species (see Tardieu & Simonneau 1998) but this corresponded to a wide range of evaporative demands (vapour pressure deficit from 9 to 3 1 kpa, all experiments included). Nicotiana plumbaginifolia exhibited the same response of g s to drought as other anisohydric species in which stomatal sensitivity to ABA was also shown to be independent of Y L (Tardieu, Lafarge & Simonneau 1996; in sunflower; Borel et al. 1997; in barley). Such stomatal control was proposed to be at the origin of the anisohydric behaviour that characterizes species which exhibit parallel reductions in midday Y L (measured under non-limiting PPFD for stomatal opening) and in Y pre-dawn as soil water is depleted (Tardieu & Simonneau 1998). Consistently with this inter- pretation, typical decreases in midday Y L from - 7 to -1 4 MPa were observed in N. plumbaginifolia when Y pre-dawn decreased from - 2 to -1 MPa, without marked differences among lines (Borel 1999). Drought-induced accumulation of ABA in the xylem sap of WT N. plumbaginifolia was intermediate between that of maize (isohydric) and of sunflower (anisohydric) (Tardieu & Simonneau 1998). The anisohydric behaviour was conserved in all lines of N. plumbaginifolia (Borel 1999) although transgenic lines accumulated much less ABA in response to drought than maize. Therefore, differences in drought-induced accumulation of ABA were unlikely to be responsible for the contrast between isohydric and anisohydric species. Slight differences appeared among lines in stomatal responses to drought-induced ABA, which did not result from differences in Y L, neither in ph, nor in acclimation to elevated ABA levels before drought Stomatal conductance of intact plants appeared to be less sensitive to drought-induced changes of [ABA] xyl in under-producing lines than in WT (Fig. 5). This contrasted with the unique stomatal response to [ABA] xyl that was observed in ABA-fed plants and, more intriguingly, in plants of different lines grafted onto the same tobacco rootstock (Figs 2 & 4). It can be questioned whether this contrast resulted from acclimation of under-producing lines to reduced levels of ABA when they were cultivated on their own roots. Acclimation may have occurred during leaf development in well-watered conditions, although ABA accumulation rates were not very contrasted between WT and transgenic lines in well-irrigated plants (Borel et al. 21). Different time-courses of ABA accumulation among lines more likely occurred during soil drying. Acclimation of wild-type and one of the transgenic lines (AS-15) to modified levels of ABA during the whole development of the measured leaf resulted in a conserved stomatal response to [ABA] xyl (Fig. 5, crossed-symbols). Therefore, acclimation of stomata to slight differences in ABA supply induced by the genetic manipulations of ZEP levels is unlikely to explain the differences among lines in stomatal responses of intact plants to [ABA] xyl. Such a result, and the fact that low-aba lines conserved the same intrinsic stomatal sensitivity to ABA, also suggest that changes in stomatal sensitivity to ABA following a dehydration rehydration cycle (see Peng & Weyers 1994) are probably not attributable to acclimation to transient rises in ABA content, provided that results in N. plumbaginifolia could be extrapolated to other species. Neither leaf water potential, nor xylem ph could explain the differences in stomatal responses among lines. First, stomatal sensitivity to ABA was independent of Y L (in all lines, Figs 6, 8a & b) and of xylem ph (in WT and AS-15, Fig. 8c & d; not tested in other lines). Second, neither Y L nor xylem ph noticeably differed in intact plants among lines. 21 Blackwell Science Ltd, Plant, Cell and Environment, 24,

12 488 C. Borel et al. A root-sourced signal of soil-drying other than ABA is suggested to modulate stomatal sensitivity to ABA in intact plants The coupling between drought-induced changes in [ABA] xyl and other changes in the chemical composition of the xylem sap probably differed among lines since ABA synthesis abilities were differently altered. This may be at the origin of the slight divergences observed in intact plants among lines (Fig. 5). An unidentified root-sourced signal of soil drought, transported in the xylem stream, possibly interacted with ABA. Such a signal must be of a chemical nature, since Y L did not participate to the regulation of g s in N. plumbaginifolia. The existence of a signal of soil drying co-acting with ABA on stomatal conductance has been previously questioned by Blackman & Davies (1985) who suggested an effect of cytokinins, and by Wilkinson & Davies (1997) who emphasized the influence of slight drifts in xylem sap ph on the redistribution of ABA between apoplast and symplast. No effect of xylem sap ph on stomatal response to ABA was detectable in the present experiments with N. plumbaginifolia (Fig. 8c & d). Conversely, all the results agreed with the possible combined (antagonist) actions of cytokinins (CK) and ABA, as proposed by Cheikh & Jones (1994) and Stoll, Loveys & Dry (2). So far as root signals were supplied by the same root system (as in grafts that were subjected to drought), ABA and CK were presumably the same in all lines, resulting in the same time-course of ABA/CK ratio among lines as drought develops and therefore a common response of g s to [ABA] xyl could be expected. CK were also expected not to modify the g s response to [ABA] xyl among lines in ABAfeeding experiments: detached leaves did not receive CK at all, and whole plants of all lines probably conserved a unique CK level since well-watered conditions were maintained during ABA feeding. By contrast, in intact plants submitted to drought, differential effects of CK on g s response to [ABA] xyl were expected among lines, because of concomitant decreases in CK for all lines (as described in other species: Bano et al. 1993; Munns & Sharp 1993) but differential accumulations of ABA. Soil water deficit had to be more severe in order to reach a given [ABA] xyl in underproducing lines than in WT. Associated with this given [ABA] xyl, the CK level was therefore probably lower (and hence, so was stomatal conductance) in low-aba-lines than in WT, which could explain the lower stomatal conductance in transgenic than in WT N. plumbaginifolia (Fig. 5). Strong consistency of the results with an assumed role of CK merits further studies, considering that any other mediator of drought stress with a similar action on stomata would have caused the same responses. ACKNOWLEDGMENTS C.B. was supported by the Ministère Français de l Education Nationale et de la Recherche Scientifique et Technique (Grant no. 9518).We are grateful to Anne Frey for providing transgenic lines. A special mention is due to Benoît Suard, for the clever high hygrometry compartment in the greenhouse and for taking care of its electronic components. Finally, the authors are also grateful to Philippe Barrieu for friendly assistance with ABA assays. REFERENCES Audran C., Borel C., Frey A., Sotta B., Meyer C., Simonneau T. & Marion-Poll A. (1998) Expression studies of the zeaxanthin epoxidase gene in Nicotiana plumbaginifolia. Plant Physiology 118, Assmann S.M., Snyder J.A. & Lee Y.-R.J. (2) ABA-deficient (aba1) and ABA-insensitive (abi1 1, abi2 1) mutants of Arabidopsis have a wild-type stomatal response to humidity. Plant, Cell and Environment 23, Bacon M.A., Wilkinson S. & Davies W.J. (1998) ph-regulated leaf cell expansion in droughted plants is abscisic acid dependent. Plant Physiology 118, Badenoch-Jones J., Parker C.W., Letham D.S. & Singh S. (1996) Effect of cytokinins supplied via the xylem at multiples of endogenous concentrations on tranpiration and senescence in derooted seedlings of oat and wheat. Plant, Cell and Environment 19, Bano A., Dörffling K., Bettin D. & Hahn H. (1993) Abscisic acid and cytokinins as possible root-to-shoot signals in the xylem sap of rice plants in drying soil. Australian Journal of Plant Physiology 2, Barrieu P. & Simonneau T. (2) The monoclonal antibody MAC252 does not react with the (-) enantiomer of abscisic acid. Journal of Experimental Botany 51, Blackman P.G. & Davies W.J. (1985) Cytokinins, abscisic acid and the control of plant water balance. Acta Horticulturae 171, Borel C. (1999) Modélisation de la synthèse d ABA et du contrôle stomatique en cas de déficit hydrique chez des plantes transgéniques affectées dans la synthèse d ABA. Ph.D. Thesis. ENSA- Montpellier, France. Borel C., Audran C., Frey A., Marion-Poll A., Tardieu F. & Simonneau T. (21) N. plumbaginifolia zeaxanthin epoxidase transgenic lines have unaltered baseline ABA accumulations in roots and xylem sap, but contrasting sensitivities of ABA accumulation to water deficit. Journal of Experimental Botany 52, in press. Borel C., Simonneau T., This D. & Tardieu F. (1997) Stomatal conductance and ABA concentration in the xylem sap of barley lines of contrasting genetic origins. Australian Journal of Plant Physiology 24, Cheikh N. & Jones R.J. (1994) Disruption of maize kernel growth and development by heat stress. Role of cytokinin/abscisic acid balance. Plant Physiology 16, Cornish K. & Zeevaart J.A.D. (1984) Abscisic acid metabolism in relation to water stress and leaf age in Xanthium strumarium. Plant Physiology 76, Cornish K. & Zeevaart J.A.D. (1988) Phenotypic expression of wild-type tomato and three wilty mutants in relation to abscisic acid accumulation in roots and leaflets of reciprocal grafts. Plant Physiology 87, Cutler A.J. & Krochko J.E. (1999) Formation and Breakdown of ABA. Trends in Plant Science 4, Demmig-Adams B. & Adams W.W. (1996) The role of xanthophyll cycle carotenoids in the protection of photosynthesis. Trends in Plant Science 1, Dietz K.J., Wichert K., Sauter A., Messdaghi D. & Hartung W. (2) Extracellular b-glucosidase activity in barley involved in the hydrolisis of ABA glucose conjugate in leaves. Journal of Experimental Botany 51, Blackwell Science Ltd, Plant, Cell and Environment, 24,