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

Abstract

A series of transgenic lines of Nicotiana plumbaginifolia with modified expression of zeaxanthin epoxidase gene (ZEP) provided contrasting ABA accumulation in roots and xylem sap. For mild water stress, concentration of ABA in the xylem sap ([ABA]_xylem) was clearly lower in plants underexpressing ZEP mRNA (complemented mutants and antisense transgenic lines) than in wild-type. In well-watered conditions, all lines presented similar [ABA]_xylem and similar ABA accumulation rates in detached roots. Plants could, therefore, be grown under normal light intensities and evaporative demand. Both ZEP mRNA abundance and ABA accumulation rate in roots increased with water deficit in all transgenic lines, except in complemented aba2-s1 mutants in which the ZEP gene was controlled by a constitutive promoter which does not respond to water deficit. These lines presented no change in root ABA content either with time or dehydration. The increase in ZEP mRNA abundance in roots with decreasing RWC was more pronounced in detached roots than in whole plants, suggesting a difference in mechanism. In all transgenic lines, a linear relationship was observed between predawn leaf water potential and [ABA]_xylem, which could be reproduced in several experiments in the greenhouse and in the growth chamber. It is therefore possible to represent the effect of the transformation by a single parameter, thereby allowing the use of a quantitative approach to assist understanding of the behaviour of transgenic lines.

Nicotiana plumbaginifolia, abscisic acid, zeaxanthin epoxidase, root, xylem sap, ZEP, zeaxanthin epoxidase, aba2-s1, ABA-deficient ZEP mutant of N. plumbaginifolia, EpoxM-S, aba2-s1 mutant of N. plumbaginifolia complemented with a full-length ZEP cDNA in the sense orientation, EpoxW-S, wild-type N. plumbaginifolia transformed with a full-length ZEP cDNA in the sense orientation, EpoxW-AS, wild-type N. plumbaginifolia transformed with an antisense ZEP fragment, [ABA]_xylem, ABA concentration in the xylem sap, PPFD, photosynthetic photon flux density, RWC, relative water content, ψ_predawn, predawn leaf water potential

Introduction

Manipulating genes related to ABA synthesis promises to open useful avenues to dissect plant responses to environmental conditions because of the key role of this hormone in the control of several adaptive processes (Giraudat et al., 1994; Ingram and Bartels, 1996; Thompson et al., 1997). The role of ABA in the control of gene expression has been demonstrated by using mutants (Leung and Giraudat, 1998). Deficient mutants of maize or tomato have also been widely used for demonstrating the role of ABA in the control of growth or stomatal conductance (Nagel et al., 1994; Herde et al., 1997; Saab et al., 1992). Among the genes involved in ABA synthesis, ZEP has been cloned in Nicotiana plumbaginifolia and encodes for zeaxanthin epoxidase, which catalyses the conversion of zeaxanthin into violaxanthin (Marin et al., 1996). The N. plumbaginifolia aba2-s1 mutant, which is deficient in ZEP, accumulates very little ABA and exhibits a severe wilty phenotype (Marin et al., 1996). Transgenic plants of N. plumbaginifolia, expressing a sense or antisense ZEP cDNA under the control of the constitutive CaMV 35S promoter, display modified seed ABA content (Frey et al., 1999).

In addition to its role for protecting plants against near-lethal stresses, ABA has a key role in maintaining near-homeostasis of leaf water status when plants are subjected to mild water deficits or to changes in evaporative demand. In particular, it contributes to fine tuning stomatal conductance in such a way that day-time leaf water potential is largely maintained under water deficit in several species (Tardieu and Simonneau, 1998). ABA also contributes to a reduction in leaf expansion rate under water deficit, thereby reducing transpiration rates and depletion rates of soil water (Zhang and Davies, 1990; Socias et al., 1997, Ben Haj Salah and Tardieu, 1997). Mutants severely deficient in ABA cannot be used for analysing these processes, because they can only be grown in a moist atmosphere and do not survive mild water deficits. In contrast, the use of a series of transformed plants presenting a range of mild reductions in ABA synthesis rate can help in the quantitative analysis of the controls of ABA synthesis under water deficit, of stomatal conductance and plant water status.

An attempt has been made here to quantify the effects of genetic manipulations of ZEP gene expression on the response to water deficit of ABA accumulation in roots and xylem sap. The pioneering approach of Stitt and Schulze has been followed, which combined genetic manipulation and changes in environmental conditions for testing whether Rubisco controls the rate of photosynthesis (Stitt and Schulze, 1994).

Materials and methods

Gene constructions and transformation

The N. plumbaginifolia aba2-s1 stable mutant (aba2-s1), already described (Marin et al., 1996), had its ZEP gene interrupted by the Ac transposon footprint. Wild-type N. plumbaginifolia var. viviani (WT) and aba2-s1 mutants plants were transformed with different constructions, all carrying a neomycin phosphotransferase sequence, which confers kanamycin resistance. Three types of transgenic lines were used: Complemented aba2-s1 mutants (EpoxM-S) and transformed wild types (EpoxW-S) were obtained by placing a full-length ZEP cDNA (2.1 kb) in the sense orientation between a 35S promoter and a rbc3′ region. Leaf disc transformation and regeneration of kanamycin-resistant plants were performed as previously described (Marin et al., 1996). Antisense plants (EpoxW-AS) were obtained by introducing a construct sharing a 1.36 kb HindIII internal fragment in an antisense orientation under the control of the CaMV 35S promoter.

Independent transgenic lines were studied, and results are presented for one EpoxW-S line, three EpoxM-S lines (EpoxM-S2, EpoxM-S3 and EpoxM-S8) and one EpoxW-AS line. All the transgenic lines studied had a single insertion locus as revealed by the proportion of kanamycin resistance found in the progeny of primary transgenic plants (about 25% compared to 18–28%). EpoxM-S and EpoxW-AS plants could be cultivated in the greenhouse under normal relative humidity. They presented a wild-type phenotype when subjected to soil water deficit, excepting EpoxW-AS, which reversibly wilted at midday when subjected to high evaporative demand.

Soil drying experiments

Plants were sown in a glasshouse at Montpellier, France in April 1996 (WT, EpoxW-S, EpoxM-S2), July 1996 (WT and EpoxW-AS) and July 1998 (WT, EpoxW-S, EpoxM-S2 and EpoxW-AS). After in vitro germination of seeds and kanamycin selection, young green plantlets of transgenic plants were transferred in 2.5 dm³ pots (50 cm high) filled with a mixture 1:1 (v/v) of sieved peat and clay-sandy-loam soil from a field near Montpellier. Pots were wrapped up with an aluminum foil to minimize soil heating and watered daily with nutrient solution (Hoagland N/10) until the stem was 5 cm long. Watering was then withheld, and the soil was covered with white perlite to avoid heating and direct evaporation. Every day, 2 or 3 plants of each line were used to measure predawn leaf water potential of one fully expanded leaf using a pressure chamber (Soil Moisture Equipment Corp., Santa Barbara, USA), and approximately 70 mm³ of sap were collected by pressurizing the leaf at 0.5–0.7 MPa above balancing pressure. Sap samples were then stored at –80 °C for subsequent ABA analysis.

An additional experiment was carried out on WT (two sowing dates) and EpoxW-AS plants. Plants were grown in 6 dm³ pots (3 plants per pot) in a growth chamber (22/18 °C day/night, 14 h photoperiod, 500 μmol m⁻² s⁻¹ PPFD, and 50% RH_air). Watering was withheld and soil was covered with white perlite when the stem was 10 cm long. Predawn leaf water potential of mature leaves was measured every morning before the light was switched on and xylem sap was extracted. Plants were then detopped and the three root systems were rapidly removed from the soil. Water content was measured on a subsample of each root system, the remaining roots were rapidly washed in a mesh with water, divided into 7–9 samples, frozen and stored at −80 °C for subsequent ABA and mRNA analyses.

Detached root experiments

Plants were grown in a growth cabinet (25/20 °C day/night, 13 h photoperiod, RH_air>60%). After in vitro germination and kanamycin selection, plantlets were transferred to 1.5 dm³ pots filled with sieved peat and irrigated daily with a modified Long-Ashton nutrient solution (Hewitt, 1963). When the sixth leaf appeared, plants with carefully washed roots were transferred to an aerated modified Long-Ashton solution changed every 10 d (4–6 plants in 10 dm³). On the day preceding the experiment, plants were placed in the dark at about 18:00 h. The following morning, the first root sample was collected in the dark. Plants were then detopped and whole root systems minus the tap roots were collected and blotted dry in tissue paper until no trace of water was visible. Roots were cut into approximately 20 mm long segments, mixed and divided into two batches. One batch was immediately sampled, subdivided into 24 subsamples placed in 10 cm³ tubes and incubated for 0–6 h. Another subsample was freeze-dried and weighed for relative water content determinations. The second batch was placed on a balance and dehydrated in a stream of warm air. When the desired relative water content was obtained, the batch of roots was subdivided into subsamples which received the same processing as moist roots. Calculation of relative water content were carried out considering that roots which had undergone no air-dehydration were at maximum hydration. During incubation, roots were maintained at a constant dehydration level in a moist atmosphere, following the procedure presented previously (Simonneau et al., 1998). After incubation, samples were frozen in liquid N₂ and stored at −80 °C. Half of the tubes were ground to powder in a mortar filled with liquid N₂ and then stored at −80 °C for mRNA analysis. Roots were crushed, reweighed and ABA was extracted by first boiling in distilled water (0.2 g dry matter in 5 cm³ water) for 7 min and shaking overnight in darkness at 4 °C. Supernatants were collected after 8 min centrifugation at 8800 g, and stored at −80 °C until ABA analysis.

ABA analysis

ABA concentration was analysed in crude samples of xylem sap or in aqueous root extracts by radioimmunoassay (RIA) (Quarrie et al., 1988), with the monoclonal antibody MAC 252 (provided by Dr SA Quarrie, Cambridge Laboratory, John Innes Centre, UK). By using thin layer chromatography (as described by Simonneau et al., 1998), it was possible to check that crude sap of N. plumbaginifolia was free of non-specific immunoreactive compounds and that ABA represented 75% of the total immunoreactants in aqueous root extracts (Borel, 1999).

Northern analysis

Total RNA was obtained from 0.5 g root after grinding in liquid nitrogen and phenol extraction as previously described (Audran et al., 1998). To check for equal loading, blots were rehybridized with a 0.5 kb cDNA fragment of 25S rRNA (Unfried and Gruendler, 1990).

Results

Genetic manipulations altered the ZEP mRNA levels differently among transgenic lines and wild-type in well-watered conditions. In leaves of well-watered plants, ZEP mRNA was 2.5 times more abundant in EpoxW-S plants than in WT. It was about 5–10 times less abundant in EpoxM-S and EpoxW-AS plants than in WT, respectively (Frey et al., 1999).

Modification of ZEP expression in transgenic lines altered the response of ABA concentration in the xylem sap ([ABA]_xylem) to soil water deficit. In all the plants studied, [ABA]_xylem measured at the end of the night increased with soil water deficit. This is shown in Fig. 1 by the relationships between [ABA]_xylem and predawn leaf water potential (an integrated indicator of soil water potential). For each genotype, a common linear relationship between predawn leaf water potential and [ABA]_xylem applied to different experiments carried out at different times of the year in the growth chamber or in the greenhouse (four experiments for WT, at least two experiments for transgenic lines). The sensitivity of the response of [ABA]_xylem to soil water deficit in a given genotype can, therefore, be estimated by the slope of the linear relationship of [ABA]_xylem to predawn leaf water potential.

Plants with an antisense construct (EpoxW-AS) accumulated ABA in the xylem sap in response to soil drying, but with a sensitivity of only 20% compared to that in WT (significant difference P<0.01) (Fig. 1d). Mutants transformed with a sense ZEP sequence (EpoxM-S2) also had a lower sensitivity than WT (40% of WT, significant difference P<0.01) (Fig. 1c). Sensitivity was also lower in EpoxM-S3 and EpoxM-S8 plants (35% of WT, data not shown). Wild-type plants transformed with a sense ZEP sequence (EpoxW-S) had the same sensitivity as that of WT over the whole range of predawn leaf water potential (Fig.1b).

ZEP is involved in ABA accumulation in roots of plants subjected to soil drying. When whole plants were subjected to various degrees of soil drying, root RWC decreased sharply with predawn leaf water potential in the range from −0.4 to −0.6 MPa and remained close to 0 in drier soils (Fig. 2). Root ABA content increased almost linearly with predawn leaf water potential in the whole range studied (Fig. 2). ZEP mRNA levels slightly increased in roots in parallel with root ABA content (Fig. 3b). They were abundant in leaves compared to roots of the same plant, but no increased accumulation of ZEP mRNA was noticed in leaves when predawn leaf water potential decreased (Fig. 3a).

When ZEP expression was modified, as in antisense plants EpoxW-AS, ABA still accumulated in roots when whole plants were subjected to soil drying (Table 1), but with lower concentrations than in WT plants at both mild and severe water deficits (predawn leaf water potential of −0.4 and −0.8 MPa, respectively). Root ABA content and [ABA]_xylem were affected in the same proportion in antisense plants compared to WT (Fig. 1a, d; Table 1).

ABA synthesis in detached roots subjected to different RWC differs between transgenic lines, depending on the promoter driving ZEP expression. Typical changes with time in the concentration of ABA in incubated roots are presented in Fig. 4, for WT and aba2-s1 plants. When just washed and immediately frozen, turgid roots contained substantial amounts of ABA, about 0.4 nmol g⁻¹ DM in aba2-s1 and WT. Initial concentrations were similar in the other transgenic lines (not shown). Root ABA concentration sharply increased with incubation time in wild-type roots subjected to dehydration. It increased much more slowly in turgid roots of WT and in both turgid and dehydrated roots of aba2-s1 mutants. The rate of ABA accumulation was calculated as the slope of the regression line of ABA content with time during the first 3 h of incubation. They are presented in Fig. 5 for all genotypes studied at the two root water statuses.

The rate of ABA accumulation was similar in turgid roots of all genotypes except the aba2- s1 mutant in which it was slower (Fig. 5a). In contrast, both accumulation rate of ABA and relative abundance of ZEP mRNA markedly differed between genotypes in dehydrated roots. As expected, soil dehydration increased neither the accumulation rate of ABA nor the relative abundance of ZEP mRNA in EpoxM-S2 plants in which the ZEP gene was under the control of a constitutive promoter. The same results were obtained in EpoxM-S3 (data not shown), and in the aba2-s1 mutant in which the ZEP gene was disrupted. Genotypes in which the gene was under the natural promoter (i.e. WT, transformed WT (EpoxW-S1) and antisense constructs (EpoxW-AS)), accumulated more ABA in dehydrated than in turgid roots (Fig. 5). Accumulation of ABA and abundance of ZEP mRNA were slightly lower in EpoxW-S1 than in WT. An unexpected result was that accumulation rate of ABA was high in roots of antisense plants.

Fig. 1.

Relationship between the concentration of ABA measured in xylem sap at predawn ([ABA]_xylem) and predawn leaf water potential measured on the same plant in wild-type N. plumbaginifolia (WT, A), primary (B, open symbols) and homozygous (B, closed symbols) transgenic EpoxW-S plants, complemented mutants (EpoxM-S2, C) and antisense EpoxW-AS plants (D). Solid lines and coefficients of correlation were obtained by linear regression for each line. Dashed lines are intervals of confidence at 95%.

Fig. 2.

Changes in ABA content (A) and RWC (B) with predawn leaf water potential in roots of wild-type plants submitted to soil drying. Roots were sampled at predawn on each plant, after measurement of leaf water potential. Root RWC was calculated supposing that water content was maximal in roots of plants with predawn leaf water potential higher than or equal to −0.3 MPa.

Fig. 3.

Expression of ZEP mRNA in leaves (A) and roots (B) of wild-type whole plants subjected to soil drying. Roots and one leaf were sampled at predawn on the same plant, after measurement of predawn leaf water potential (root ABA content and root RWC of these samples are represented in Fig. 2). Ethidium bromide staining of 25S RNA is shown as a control. Relative ZEP mRNA abundance was determined using 25S rRNA as a standard, and setting 1 for relative ZEP mRNA abundances in roots and leaf of the plant with the highest predawn leaf water potential (−0.3 MPa).

Fig. 4.

Time-course of ABA content in detached roots of WT (A) and aba2-s1 mutant (B) when incubated in moist air at 24 °C in darkness. Roots were collected on adult plants cultivated in hydroponic conditions (and at saturating air humidity for aba2-s1), and subdivided into two batches. One batch was separated into eight subsamples that were immediately incubated for different period times (open symbols, RWC 100%). The second batch of roots received the same treatment, except that it was rapidly dehydrated at 40% RWC just before incubation (closed symbols).

Fig. 5.

Effect of rapid dehydration on ABA accumulation rates (A) and ZEP mRNA abundance (B) in detached roots of wild-type, mutant and different transgenic lines of N. plumbaginifolia. Rates of ABA accumulation (A) were calculated as the slope of the regression line between root ABA content and time over 3 h incubation time periods. Roots were either maintained at 100% RWC (open columns) or dehydrated and maintained at 40% RWC (hatched columns) during the incubation period. ZEP mRNA abundance was determined in roots samples from the same experiment (B) and represents average abundance (between 2 h to 3 h 30 incubation times) relative to ZEP mRNA level at the beginning of incubation. Bars indicate standard errors.

Table 1.

Root ABA content for wild-type (WT) and transgenic (EpoxW-AS) N. plumbaginifolia

Roots were sampled before dawn on plants submitted to mild (predawn leaf water potential (ψ_predawn)=−0.4±0.1 MPa) or severe soil water deficit (ψ_predawn=−0.8±0.1 MPa). Means of 3–5 measurements on different plants±standard errors.

ψpredawn	Root ABA content (nmol g−1DM)
	WT	EpoxW-AS
−0.4 MPa	0.128±0.076	0.023±0.026
−0.8 MPa	0.855±0.087	0.270±0.032

Discussion

Transformed lines have contrasting sensitivities to drought of ABA accumulation in roots and xylem sap, but unaltered baseline accumulations of ABA

The range of plants presented here had several useful characteristics in the analysis and modelling of stomatal control. In well-watered plants, the accumulation rate of ABA in roots and the concentration of ABA in the xylem sap did not differ between WT and transgenic plants (except in the deficient mutant). Plants could, therefore, be grown in normal evaporative demands and light intensities in the greenhouse and in the growth chamber. This behaviour contrasts with that of other mutants, such as the sitiens and flacca mutants of tomato, where concentrations of ABA in roots and shoots are greatly lowered in well-watered plants (30% and 10% of WT; Cornish and Zeevaart, 1988), thus rendering them incapable of growth in normal environmental conditions.

In spite of this common baseline [ABA]_xylem, the sensitivity of ABA accumulation to water deficit had a large variability across lines in the range of predawn leaf water potential observed in natural conditions. At a predawn leaf water potential of −1.5 MPa, [ABA]_xylem ranged from 200 (EpoxW-AS) to 850 μmol m⁻³ (EpoxW-S) corresponding to an 80% difference in sensitivity to water deficit between these lines. This difference is similar to that observed between several species (e.g. 300 versus 800 μmol m⁻³ at −1.5 MPa in maize and sunflower, Tardieu and Simonneau, 1998). All transgenic lines accumulated ABA in response to water deficit, again in contrast to flacca and sitiens (Cornish and Zevaart, 1988). Plants could therefore be subjected to severe water deficits without dying, so their behaviour could be analysed quantitatively. It is noteworthy that the most contrasting genotypes have been presented here, but other transgenic lines have intermediate sensitivities to predawn leaf water potential (50–60% of the sensitivity of WT).

Other plant characteristics, such as the relationship between photosynthesis and stomatal conductance, stomatal density or leaf expansion rate of well-watered plants did not appreciably differ between transformants (Borel, 1999). Interference effects are therefore less likely than in mutants whose architecture and water relations largely differ from those of WT. Likewise, the fact that the range of plants was obtained with independent events of transformation and three strategies helps avoid errors in interpretation.

Induction of the ZEP expression is needed for ABA accumulation in roots, but with a difference in behaviour between detached roots and whole plants

As expected, a decrease in root RWC caused neither increase in ZEP mRNA abundance nor increase in ABA accumulation rate in detached roots when the ZEP gene was controlled by the 35S promoter (EpoxM-S). This promoter presents a constitutive expression (Odell et al., 1985; Kay et al., 1987) and is not induced by dehydration. The same result was observed in the mutant in which the ZEP gene was disrupted. In contrast, ZEP expression was increased in transgenic lines with the WT promoter when ABA accumulation rate was increased by root dehydration. This correlation between ABA accumulation rate and ZEP mRNA induction suggests that induction of this gene is necessary for drought-induced ABA synthesis in detached roots conditions.

However, the relationship between ZEP mRNA and ABA accumulation in roots was less straightforward when whole plants were subjected to soil drying. While decreasing RWC to 40% of control caused a 6-fold increase in ZEP mRNA abundance in detached roots (Fig. 5b), the increase was only 2-fold in roots of whole plants subjected to soil drying (Fig. 4b). Large accumulation of ABA in roots was observed in both cases, although units differ between experiments (ABA amounts in whole plant, ABA accumulation rate in detached roots). This suggests that plants may respond differently to rapid and progressive stress. It could be hypothesized that in response to rapid stress, root xanthophyll was rapidly metabolized, so it was necessary to produce them again. On the other hand, during a progressive stress a slight increase of the flux of xanthophyll metabolized for ABA synthesis could be sufficient, since it was standing during the whole stress period. Then, a slight increase of the ZEP mRNA level could be sufficient, possibly in combination with post-transcriptional regulation of this gene or of other genes implicated in ABA synthesis. For example, the induction of the gene encoding NCED (which catalyses the cleavage of 9-cis-epoxycarotenoids) has been reported in leaves in response to water stress (Tan et al., 1997) or drought (Burbidge et al., 1997; Qin and Zeevaart, 1999).

Unexpectedly, ABA accumulation in detached roots was similar in EpoxW-AS and WT (Fig. 5), while ABA content in roots of whole plants markedly differed between EpoxW-AS and WT (Table 1). One possible interpretation is based on differences in stress-induced expression of ZEP between detached roots and roots of whole plants, while expression level of antisense mRNA (under the control of the 35S promoter) remained unchanged in all conditions (as did the expression of ZEP mRNA, also controlled by 35S, in EpoxM-S2, Fig. 5b). High ZEP expression levels induced by water stress in detached roots (as for WT, Fig. 5b) possibly overcame any difference caused by expression of antisense mRNA between WT and EpoxW-AS, consistent with the results of Fig. 5. In contrast, ZEP expression level was low in roots of whole plants even when submitted to soil drying (Fig. 3b), so that expression of antisense mRNA in EpoxW-AS was likely to have altered ZEP activity more strongly in roots of whole plants than in detached roots. Neither ZEP levels nor protein activity were analysed to test this hypothesis.

Alterations of ZEP expression level can be expressed as a parameter in the predicting model of [ABA]_xylem during drought

The response was linear in the eight studied transgenic lines (presented in Fig. 1 or not shown). The series of transgenic lines presented here therefore behaved as other species for which linear responses of [ABA]_xylem to predawn leaf water potential were also observed (Wartinger et al., 1990; Gallardo et al., 1994; Socias et al., 1997; Tardieu and Simonneau, 1998). The effect of every transformation can therefore be expressed via the slope of this linear relationship, thereby allowing the prediction of [ABA]_xylem of any transgenic line at any soil water status. This opens the possibility of an integrated model of stomatal control in transgenic lines, thereby reconstructing a complex behaviour from a limited number of equations (Tardieu and Simonneau, 1998).

This linear relationship between predawn leaf water potential and [ABA]_xylem was not necessarily expected in lines where the ZEP gene was under the control of a constitutive promoter, because neither the expression of this gene nor the accumulation rate of ABA were increased by dehydration of detached roots. Part of the explanation could involve the difference between short-term and longer term root dehydration as discussed above. Part of the increase in [ABA]_xylem of droughted plants could also be due to a lower dilution of ABA in xylem sap on the previous day (Tardieu et al., 1992, 1993).

Acknowledgments

We thank Philippe Barrieu and Benoît Suard for technical assistance, and Isabelle Constant for enjoyable help. Charlotte Borel was supported by the Ministère de l'Education Nationale et de la Recherche Scientifique et Technique (Grant 95108).

Author notes

3

To whom correspondence should be addressed. Fax: [phone]. E-mail: [email protected]