ABA plays a central role in mediating the regulatory effects of nitrate on root branching in Arabidopsis

Authors

  • Laurent Signora,

    1. 1 Division of Crop Performance and Improvement, IACR-Rothamsted, Harpenden, Hertfordshire, AL5 2 JQ, UK, and2Centre for Plant Sciences, Leeds Institute of Biotechnology and Agriculture (LIBA), School of Biology, University of Leeds, Leeds, LS2 9JT, UK
    Search for more papers by this author
  • 1,2 Ive De Smet,

    1. 1 Division of Crop Performance and Improvement, IACR-Rothamsted, Harpenden, Hertfordshire, AL5 2 JQ, UK, and2Centre for Plant Sciences, Leeds Institute of Biotechnology and Agriculture (LIBA), School of Biology, University of Leeds, Leeds, LS2 9JT, UK
    Search for more papers by this author
    • Present address: Vlaams Instituut voor Biotechnologie, Department of Genetics, University of Gent, K. L. Ledeganckstraat 35, B-9000 Gent, Belgium.

  • 2, Christine H Foyer,

    1. 1 Division of Crop Performance and Improvement, IACR-Rothamsted, Harpenden, Hertfordshire, AL5 2 JQ, UK, and2Centre for Plant Sciences, Leeds Institute of Biotechnology and Agriculture (LIBA), School of Biology, University of Leeds, Leeds, LS2 9JT, UK
    Search for more papers by this author
  • and 1 Hanma Zhang 2,

    Corresponding author
    1. 1 Division of Crop Performance and Improvement, IACR-Rothamsted, Harpenden, Hertfordshire, AL5 2 JQ, UK, and2Centre for Plant Sciences, Leeds Institute of Biotechnology and Agriculture (LIBA), School of Biology, University of Leeds, Leeds, LS2 9JT, UK
    Search for more papers by this author

For correspondence (fax +44 (0) 113 233 2835; e-mail h.m.zhang@leeds.ac.uk).

Summary

The formation of lateral roots (LR) is a major post-embryonic developmental event in plants. In Arabidopsis thaliana, LR development is inhibited by high concentrations of NO3. Here we present strong evidence that ABA plays an important role in mediating the effects of NO3 on LR formation. Firstly, the inhibitory effect of NO3 is significantly reduced in three ABA insensitive mutants, abi4-1, abi4-2 and abi5-1, but not in abi1-1, abi2-1 and abi3-1. Secondly, inhibition by NO3 is significantly reduced, but not completely abolished, in four ABA synthesis mutants, aba1-1, aba2-3, aba2-4 and aba3-2. These results indicate that there are two regulatory pathways mediating the inhibitory effects of NO3 in A. thaliana roots. One pathway is ABA-dependent and involves ABI4 and ABI5, whereas the second pathway is ABA-independent. In addition, ABA also plays a role in mediating the stimulation of LR elongation by local NO3 applications.

Introduction

Lateral root (LR) formation plays a critical role in determining the architecture and the spatial arrangement of the root system in plants (Malamy and Benfey, 1997a). The development of LRs starts from an asymmetric transverse division of the pericycle cells adjacent to the two xylem poles of the parent root. This is then followed by a series of periclinal and transverse cell divisions leading to the formation of a dome-shaped LR primordium (LRP). The LRP then grows (via both cell division and cell expansion) through the overlaying cell layers and emerges from the parent root (via cell expansion). After emergence, the LRP undergoes an activation step to form a fully functional meristem (Malamy and Benfey, 1997b).

Unlike the primary root, which is initiated during embryogenesis and whose development is predetermined, LRs are formed throughout the lifetime of a plant and display remarkable morphogenic plasticity. Many internal (developmental and metabolic) and external, particularly nutritional, factors affect LR development. A striking example is the effect of NO3 availability on LRs. In Arabidopsis thaliana, NO3 regulates LR development in two different ways. Firstly, low NO3 (≤ 1.0 mm) supply stimulates LR elongation. This stimulatory effect on LR formation is localized (only affecting roots directly in contact with the NO3 supply) and acts mainly on the elongation phase and not the early development. Secondly, high NO3 inhibits LR development. This inhibitory effect is systemic (affecting the whole root system), dose-dependent (usually noticeable at high concentrations, i.e. ≥ 10 mm) and developmentally regulated (occurring immediately after the emergence of LRP from the parent root). A phenotypic characteristic of the NO3 dependent inhibition is the reduction of visible (≥ 5 mm) LRs. Neither the early development (pre-emergence stages) of LRPs nor the elongation of established (post-meristem activation stage) LRs is affected by high-NO3 (Zhang and Forde, 1998; Zhang et al., 1999).

There is an antagonistic interaction between NO3 and sugar in the regulation of LR development. The inhibitory effect of high NO3 (10 and 50 mm) was significantly reduced when the sucrose supply in the medium was increased (Zhang et al., 1999, 2000). This C/N antagonism undoubtedly requires the accurate sensing of N and C status in the plants (Zhang et al., 1999).

As part of our continuing effort to elucidate the mechanisms underlying C/N antagonism in LR regulation, the inhibitory effect of NO3 was analysed in a number of sugar-sensing mutants, including sun6-2 (Dijkwel et al., 1997; Huijser et al., 2000), rsr1-1 (Martin et al., 1997) and mig3 (Pego et al., 2000). This led to the discovery that the sun6-2 mutant has reduced sensitivity to high-NO3-induced inhibition. During the course of our experiments, it was reported that this mutant carried a transposable element (En-1) inserted in the 5′ untranslated region of the ABI4 gene (Huijser et al., 2000). With this information an exploration of the relationship between ABA and the NO3-induced inhibition of LR development became essential. The results presented here show, for the first time, that ABA plays a crucial role in mediating NO3-dependent regulation of LR development.

Results

The inhibitory effect of nitrate on LR development is significantly reduced in the abi4-1, abi4-2 and sun6-2 mutants

The observation that the sun6-2 mutant is less sensitive to NO3-dependent inhibition of LR formation, together with the knowledge that this mutant carries an insert in the ABI4 gene, prompted us to test the association between nitrate inhibition and the mutation in the ABI4 gene. Seedlings of the abi4-1 mutant (Finkelstein et al., 1998) and the Col-0 control were grown on two KNO3 concentrations (1.0 and 10 mm). The resultant root morphology is shown in Figure 1. Although there was very little difference in overall root development between the control and the mutant at 1.0 mm KNO3, the abi4-1 seedlings produced significantly more visible LRs at 10 mm KNO3 than the Col-0 control (Figure 1c, d). This observation supports the hypothesis that the reduced LR inhibition at high-NO3 is associated with a mutation in the ABI4 gene.

Figure 1.

Comparison of the root morphology of Col-0 and abi4-1 seedlings on two different nitrate concentrations. The seedlings were grown either on 1 mm (a, b) or 10 mm (c, d) KNO3 for 7 days before photographing. The abi4-1 mutant had very similar lateral root growth as the control at 1.0 mm KNO3, but produced much more lateral roots than Col-0 at 10 mm KNO3.

To further confirm the above hypothesis, we took a quantitative approach to analyse the effect of NO3 supply on the growth of primary and lateral roots in two abi4 mutants, abi4-1 and abi4-2, and the Col-0 control. The abi4-2 mutant carries a single base pair deletion at position 277 of the ABI4 transcript and produces a truncated ABI4 protein of 109 amino acids instead of the wild-type form of 328 amino acids (Quesada et al., 2000). Seedlings of the mutants and control were grown on a range of KNO3 concentrations (0.1, 1.0, 10, 50 mm) for 7 (Col-0 and abi4-1) or 9 (abi4-2) days. The abi4-2 seedlings were grown for two extra days because their primary roots grew slower than those of the other two lines and required the extra 2 days to reach similar length as that of the 7-day-old Col-0 and abi4-1 seedlings. Despite such difference, primary root growth in all the three lines was largely unaffected by the level of KNO3 supply except at 50 mm, where a slower primary root growth was observed (Figure 2a) in all the lines.

Figure 2.

The response of root growth to NO3 in the wild-type and the abi4-1 & abi4-2 mutants. Wild type (open bars), abi4-1 (shaded bars) and abi4-2 (filled bars) mutants were grown for 7 (Col-0 and abi4-1) or 9 (abi4-2) days on media containing a range of KNO3 concentrations. The length of the primary roots (a), the total length of LRs (b), the number (c) and the average (d) length of visible LRs were presented. The number of LR only included those that were ≥ 5 mm in length and clearly visible without the use of a microscope. The values and error bars are the means of 12–16 seedlings and the standard deviation, respectively. The experiment was repeated three times.

The mutants and the Col-0 control displayed clear differences in their LR growth in responses to KNO3 (Figure 2b, c and d). Firstly, although an increase in KNO3 from 0.1 to 1.0 mm had very little effect on LR growth in the control, it increased the total LR length by 109 and 48% in the abi4-1 and abi4-2 mutants, respectively. Secondly, the increase of KNO3 from 1.0 to 10 mm caused a 71% reduction of the total LR length, a 62% reduction of the average LR length and a 46% reduction of the number of visible LRs in the Col-0 control. However, these figures were reduced to 42, 38 and 14% in abi4-1 or 38, 28 and 13% in abi4-2, respectively, with the most significant reduction in the number of visible LRs (reduced to less than one-third of the control in both mutants). Thirdly, at 50 mm KNO3, almost no visible LRs were produced in the Col-0 control, whereas both abi4-1 and abi4-2 produced significant number of visible LRs. Furthermore, the difference of LR growth between the mutants and the control changes at different KNO3 concentrations. At 0.1 mm KNO3, the abi4-1 and abi4-2 mutants produced less LR growth. At higher KNO3 concentrations (1.0, 10 and 50 mm), however, this was reversed, the mutants producing greater total LR length (151 and 157% at 1.0 mm; 292 and 324% at 10 mm and 1173 and 563% at 50 mm; the figures in the brackets represent data of the mutants expressed as the percentage of the control), greater average LR length (120 and 115% at 1.0 mm; 193 and 214% at 10 mm and 1127 and 931% at 50 mm) and more visible LRs (123 and 143% at 1.0 mm; 196 and 228% at 10 mm and 640 and 320% at 50 mm) than the control. The higher the KNO3 concentration, the bigger the difference between the mutants and the control in LR growth.

The observation that the differences between the mutants and the controls increased at higher NO3 concentrations suggests that abi4 mutations reduce the inhibitory effect of high NO3 on LR growth. One approach to test this hypothesis is to establish whether the abi4 mutations affect LRs at the same developmental stage as high-NO3. It has been established that high NO3 inhibits LR development at a specific stage, namely immediately after the emergence from the parent roots, and does not affect the elongation of established LRs (Zhang et al., 1999). To test whether the abi4 mutations affect the elongation of established LRs, a ‘transfer’ experiment was performed. abi4-1 and Col-0 seedlings were first grown on either 1.0 or 10 mm KNO3 for 9 days to allow visible LRs to develop and then transferred onto medium containing either 1.0 or 10 mm KNO3. The growth rate of established LRs (> 5 mm in length) was then measured during the following 3 days. Comparable LR elongation rates were observed in Col 0 and in the abi4-1 mutant in all four transfer combinations, namely 1.0 to 1.0 mm (Figure 3a), 1.0 to 10 mm (Figure 3b), 10 to 1.0 mm (Figure 3c) or 10 to 10 mm (Figure 3d). These results indicate that the increased LR growth in the abi4-1 mutant is not due to a stimulation of the elongation rate of the established LRs. We have also observed no clear differences between the abi4-1 and abi4-2 mutants and the Col-0 control in LR development at the pre-emergence stages (data not shown), indicating that the stimulation of LR development in the abi4 mutants likely occurred at the same developmental stage as the inhibitory effect of NO3.

Figure 3.

Comparison of the elongation rates of established LRs in Col-0 (open bars) and the abi4-1 mutant (filled bars). Seedlings were grown on medium containing either 1.0 (a and b) or 10 mm KNO3 (c and d) for 9 days to allow LRs to develop and then transferred to either 1 (a and c) or 10 mm KNO3 (b and d). The values and error bars are the means of 14–16 seedlings and the standard deviation, respectively. The experiment was repeated twice.

The ABA synthesis mutants are also less sensitive to nitrate-induced inhibition on LR development

To establish whether the inhibitory effect of high NO3 on LR development requires ABA synthesis, this property was examined in a number of ABA synthesis mutants, including aba1-1 (Koornneef et al., 1984; Rock and Zeevaart, 1991), aba2-3, aba2-4 (Laby et al., 2000) and aba3-2 (Leon Kloosterziel et al., 1996). Seedlings of the mutants and their wild-type controls (the aba2-3 and aba2-4 mutants are in the Col background, the aba1-1 and aba3-2 mutant in the Ler background) were grown on four KNO3 concentrations (0.1, 1.0, 10 and 50 mm) for 7 (Col-0, aba2-3 and aba2-4) or 8 days (Ler-1, aba1-1 and aba3-2). Although the aba2-3, aba2-4 and aba3-2 mutants had very similar primary root growth to their respective controls (Figure 4a, c), they produced much more LR growth at all the four KNO3 concentrations (Figure 4b, d). In addition, the differences in LR growth between the mutants and their controls increased at higher KNO3 concentrations. At 1.0 mm KNO3, the total LR length produced by the aba2-3 or aba2-4 mutants was 211 or 243% that of the Col-0 control. However, it was increased to 377 and 319% or 1025 and 1113% that of the control at 1.0 or 10 mm KNO3, respectively. Similarly, the total LR length produced by the aba3-2 mutant was 230, 789 and 600% that of the control (Ler-1) at 0.1, 1.0 and 10 mm KNO3, respectively. All the aba mutants produced significant LR growth at 50 mm KNO3, whereas no visible LRs were produced in the controls on this medium. The increased differences in LR growth between the controls and the mutants at higher KNO3 concentrations indicate reduced sensitivity to NO3 inhibition in the mutants. The results obtained from aba1-1 were similar to those observed in the other three aba mutants, although the effect of aba1-1 mutation was slightly less pronounced than that of the other three mutations (data not shown). These results strongly suggest that the inhibitory effect of high NO3 requires ABA synthesis. However, the results also show that the inhibitory effect was not completely absent in the aba mutants, suggesting the presence of an ABA-independent pathway.

Figure 4.

The effect of NO3 supply on root growth in ABA synthesis mutants. Seedlings of Col-0 (open bars), aba2-3 (grey bars), aba2-4 (dark bars), Ler-1 (dashed horizontal bars) and the aba3-2 (hatched bars) were germinated and grown for 3 days on 0.1 mm KNO3 and then transferred onto mediums containing four different (0.1, 1.0, 10 and 50 mm) KNO3 concentrations for a further 7 (Col-0, aba2-3 & aba2-4) or 8 days (Ler-1 & aba3-2). Both the primary (a and c) and lateral roots (b and d) were measured and each point represents the data from 14 to 16 seedlings.

The inhibitory effect of NO3 on LR development involves a specific ABA signal transduction pathway

In addition to ABI4, four other genes in the ABA signalling pathway (ABI1, ABI2, ABI3 and ABI5) have been cloned (Finkelstein and Lynch, 2000; Finkelstein et al., 1998; Giraudat et al., 1992; Leung et al., 1994, 1997). To test whether the inhibitory effect of NO3 is mediated by specific ABA signal transduction mechanisms, we examined the effect of NO3 in the abi1-1, abi2-1, abi3-1 and abi5-1 mutants. Primary root growth in all the mutants was comparable to their respective controls (abi1-1, abi2-1 abi3-1 and the Ler-1 control, Figure 5a; abi5-1 and the Ws control, Figure 5c) at all the four KNO3 concentrations. The only exception was abi3-1, which had slightly lower primary root growth at 0.1 mm KNO3. The mutants showed rather diverse LR responses to increasing KNO3 concentrations. The abi1-1 mutant showed a nearly ‘wild-type’ response (Figure 5a, b), producing very similar LR growth at all four KNO3 concentrations. The abi5-1 mutant, on the other hand, showed an ‘abi4-like’ response and had increased LR growth relative to the Ws control, especially at the three higher KNO3 concentrations. The total LR length produced by the abi5-1 mutant was 117, 159 179 and 534% that of the Ws control at 0.1, 1.0, 10 and 50 mm KNO3, respectively (Figure 5d). The increased difference of LR growth between the Ws control and the abi5-1 mutant at higher KNO3 concentrations indicates that this mutant is also less sensitive to the NO3 inhibition. The abi2-1 and abi3-1 mutants produced more complicated responses to KNO3. The abi2-1 mutant produced significantly less LR growth at both 0.1 and 1.0 mm KNO3, slightly less LR growth at 10 mm KNO3 and similar LR growth at 50 mm KNO3 compared with the Ler-1 control. The abi3-1 mutant had a reduced LR growth (than the control) at 0.1 mm KNO3, a ‘wild-type-like’ LR growth at 1.0 and 10 mm KNO3 and an increased LR growth (than the control) at 50 mm KNO3. The complexity of the abi2-1 and abi3-1 responses to NO3 suggests that the ABI2 and ABI3 proteins are unlikely to play a direct role in mediating the inhibitory effect of NO3.

Figure 5.

The effect of NO3 supply on root growth in the ABA insensitive (abi) mutants. The primary (a and c) and lateral (b and d) root length of Ler-1 wild type (open bars), abi1-1 (shaded bars), abi2-1 (horizontal bars), abi3-1 (filled bars), Ws wild type (dashed horizontal bars) and abi5 mutant (hatched bars). Seedlings (14–16) were grown for 8 days before the measurement. The experiment was repeated twice.

It was noteworthy that the three ecotypes used in these experiments (Col-0, Ler-1 and Ws) displayed slightly different responses to NO3 supply (Figures 2a, 5b and d). The Ler-1 ecotype was slightly less sensitive to NO3 inhibition than the Ws or Col-0 ecotypes.

The localized stimulatory effect of NO3 is reduced in the aba1-1 mutants and increased in most of the abi mutants

It has been previously established that localized NO3 (1.0 mm) application stimulates LR elongation in the NO3-rich zone (Zhang and Forde, 1998; Zhang et al., 1999, 2000). To investigate whether ABA plays any role in mediating this stimulatory effect, we examined the responses of LRs in the aba and abi mutants to the localized NO3 treatment. To facilitate comparisons, we used the stimulation index (SI, see Experimental procedures) to define the degree of stimulation. Most of the abi mutants (except abi5-1) showed a significantly increased response to the localized KNO3-treatment (SI: Ler-1: 2.1, abi1-1: 4.7, abi2-1: 3.2 and abi3-1: 6.0; Col-0: 1.4, abi4-1: 2.6 and abi4-2: 3.4) (Figure 6a, b). It is interesting to note that although the degree of stimulation was increased in the abi4-2 mutant, the overall LR growth in this mutant was, in fact, reduced. On the other hand, the aba1-1 mutant produced more LR growth in both the KNO3- and KCl-treated middle segment (Figure 6a) than the Ler-1 control (Figure 6a). However, the stimulation index is reduced in this mutant. These results suggest a negative correlation between overall LR growth and responsiveness to the stimulatory effect of localized NO3.

Figure 6.

Effect of localized supply of KNO3 on lateral root growth in the ABA synthesis and ABA insensitive mutants. Seedlings were initially germinated on medium containing 0.01 mm NH4NO3 for 4 days and then transferred to the segmented plate with the middle segment supplied with either 1.0 mm KCl or 1.0 mm KNO3 for a further 7 days. The total length of lateral roots in the KCl- (open bars) or KNO3-treated (filled bars) middle segment was presented. (a): Ler-1 and mutants (abi1-1, abi2-1, abi3-1 and aba1-1) in Ler-1 background; (b): Col-0 and mutants (abi4-1 and abi4-2) in Col background and (c): Ws and a mutant (abi5-1) in Ws background. The experiment was repeated twice and 14 seedlings were measured for each data point in each experiment.

The results obtained with the aba1-1 mutant are also interesting in that more LR growth was observed in the KCl-treated segment of the mutant than in the control segment exposed to KNO3 (Figure 6a). This suggests that LR growth in the ABA deficient background is constitutively stimulated and that the stimulatory effect of the localized NO3 treatment requires ABA synthesis.

Of the three ecotypes (Col-0, Ler-1 and Ws) studied in these experiments, Ler-1 was the most responsive (SI: 2.1) to the localized KNO3-treatment. Ws (SI:1.4) was the least sensitive of the ecotypes with Col-0 (SI: 1.4) falling in between.

Discussion

Post-embryonic LR formation displays remarkable plasticity and enables a plant to adapt to variable soil conditions and to exploit available resources efficiently. The developmental flexibility of LRs undoubtedly relies on the ability to integrate various developmental, metabolic and environmental signals. In this report, we have demonstrated that ABA, is part of the regulatory machinery governing LR development and that this plant hormone could play a key role in the integration of various regulatory signals.

ABI4 plays an important role in mediating the inhibitory effect of nitrate on LR development

In this study, we have provided three lines of evidence demonstrating that the inhibitory effect of NO3 on LR development is reduced in the abi4-1, abi4-2 and sun6-2 mutants. Firstly, the mutants produced more LR growth at higher NO3 concentrations than the Col-0 control (Figure 1b). This is consistent with a decrease in sensitivity to high NO3 inhibition. Since NO3 inhibition is most noticeable at higher NO3 concentrations, a decrease in sensitivity would also be most apparent at high NO3. Secondly, the abi4 mutants produced significantly more visible LRs, a result one would predict if sensitivity to NO3 were decreased. Thirdly, mutations in the abi4 gene act at the same developmental stage as the inhibitory effect of high NO3. Two observations support this conclusion: (1) the elongation of established LRs (Figure 3) and (2) the early (at pre-emergence stages) development of LRs are both not modified in the abi4 mutants.

The ABI4 gene encodes a putative AP2 domain transcription factor (Finkelstein et al., 1998). There is now considerable evidence that the ABI4 protein is involved in sugar signalling. The sun6-2 mutant is insensitive to both mannose-induced inhibition of seed germination and to repression of photosynthetic genes by sucrose (Huijser et al., 2000). The gin6 mutant, which carries a T-DNA insertion 2.0 kb upstream of the ABI4 gene and does not express the ABI4 transcript, was less sensitive to glucose inhibition (Arenas-Huertero et al., 2000). It is therefore feasible that the ABI4 protein plays some role in the antagonism between N and C in the regulation of LR development, but this remains to be unequivocally demonstrated.

ABA plays direct role in mediating the inhibitory effect of nitrate on LR development

We have previously observed that NO3 has two opposing effects on LR development: a localized stimulatory effect and a systemic inhibitory effect. The stimulatory effect is mediated by a nitrate-inducible transcription factor, ANR1. Down-regulation of ANR1 expression resulted in a negative linear relationship between NO3 concentration and LR growth (Zhang and Forde, 1998; Zhang et al., 1999). This led to the hypothesis that the inhibitory effect of NO3 on LR growth is dose-dependent and occurs at all NO3 concentrations (Zhang et al., 1999, 2000). Of particular note is the observation that the aba mutations showed the ‘opposite’ effect (at the lower KNO3 concentration range, 0.1–1.0 mm) to ANR1 down-regulation (Zhang and Forde, 1998; Figure 4). The aba mutations increase (rather than decrease) LR growth with the increase of NO3 concentrations, indicating that ABA plays a major role in mediating the inhibitory effect of NO3.

The results presented here show conclusively that the inhibitory effect of NO3 requires ABA synthesis. One possible mechanism enabling this response is that NO3 enhances ABA synthesis. Measurements of the root ABA content of seedlings grown on a range of NO3 concentrations will help to establish whether this is the case. Consistent with this hypothesis is our recent observation that exogenous ABA can inhibit LR development in A. thaliana (Zhang, Signora and Foyer, unpublished data).

The inhibitory effect of NO3 also involves an ABA-independent pathway. This is clearly illustrated by the fact that the inhibition does not disappear completely in the aba mutants (Figure 4). Total LR length in the presence of 50 mm KNO3 was less than 20% that obtained at 1.0 mm KNO3 in all four aba mutants tested. Although this may be due to residual capacity for ABA synthesis in the mutants (Rock and Zeevaart, 1991), it is probable that there is also an ABA-independent pathway involved in the NO3 inhibition response.

The inhibitory effect of NO3 on LR development involves distinct ABA signal transduction pathways

The results presented in Figure 5 show that the inhibitory effect of NO3 on LR development involves specific ABA signal transduction components. Among the abi mutants, only abi4-1 and abi5-1 showed clear reductions in the degree of NO3 inhibition. Interestingly, the abi mutants showed a similar division in their sensitivity to glucose-dependent inhibition of early seedling development. Wild-type A. thaliana seedlings cannot develop green and extended cotyledons in the presence of 7% glucose. Of the abi mutants tested in the present study (abi1-1, abi2-1, abi3-1, abi4-1, and abi5-1), only abi4-1 and abi5-1 were insensitive to inhibition by glucose (Arenas-Huertero et al., 2000). It is tempting to suggest that this similarity reflects overlap in the signal transduction pathways linking the glucose-dependent inhibition of seedling development to NO3-dependent inhibition of LR development.

ABA and LR elongation

The stimulatory and inhibitory effects of NO3 on LRs occur at different developmental stages and are mediated by different mechanisms (Zhang et al., 1999, 2000). The results obtained in some of the experiments presented here are consistent with this view. For example, in the abi5-1 mutant, the inhibitory effect of NO3 was reduced, whereas the stimulatory effect of localized NO3 treatments was not affected. The abi1-1, abi2-1 and abi3-1 mutants, on the other hand, had an increased response to the stimulatory effect of localized NO3 treatment, but showed no significant changes in their response to the inhibitory effect. However, our results indicate that ABA may also play a role in mediating the stimulatory effect of NO3. Firstly, the stimulatory effect of NO3 clearly requires the capacity to synthesize ABA, indicated by a reduced stimulation in the aba1-1 mutant (Figure 6a). One possible explanation is that ABA suppresses LR growth and the effect of the localized NO3 is to remove this suppression. Secondly, the stimulatory effect of the localized NO3 treatment was enhanced in most of the abi mutants (except abi5-1). The mechanism whereby ABA mediates both the stimulatory and the inhibitory effects of NO3 is not known. Similarly, the nature of the overlap allowing the two different roles of ABA remains to be elucidated.

Experimental procedures

Basic medium and growth conditions

The basic growth medium consisted 100 μm KCl; 40 μm MgSO4; 20 μm CaCl2; 22 μm NaH2PO4; 0.9 μm MnSO4; 0.09 μm KI; 0.97 μm H3BO4; 0.14 μm; 2 nm CuSO4; 20.6 nm Na2MoO4; 2.1 nm CoCl2; 3.6 μm Fe-EDTA, 0.5 g l−1 MES (pH 5.7) and 1% agar-agar. Seedlings were grown vertically on the surface of the solidified agar medium in a growth chamber at 20°C with a 16 h light : 8 h dark regime. Seeds were sterilized by a 20 min incubation in 20% (v/v) commercial bleach/ethanol followed by five washes with ethanol.

Uniform nitrate treatment

The experiments with uniform nitrate application were carried out as previously described (Zhang et al., 1999). Sterilized seeds were first placed on basic medium supplemented with 0.5% sucrose and 10 μm NH4NO3 and incubated in the growth room for 3 days to allow germination and initial seedling growth. The seedlings were then transferred to fresh plates (four seedlings per plate) containing different concentrations of KNO3 (0.1, 1, 10 or 50 mm) and 2% sucrose for a specified period. The length of the primary and lateral roots of each individual seedling was measured directly using a ruler. LRs, which were 5 mm or longer, were defined as ‘visible’.

Elongation of established LRs

For measuring the growth rates of established LRs, seedlings were first grown on medium containing either 1.0 or 10 mm KNO3 for 9 days to allow LRs to develop and then transferred to fresh medium (four seedlings per plate) with either 1.0 or 10 mm KNO3. The length of individual LR was measured daily for 3 days after the transfer. An ‘established’ LR is defined as the one longer than 5 mm at the time of transferring.

Monitoring of early LR developmental stages

The early LR developmental stages were examined as previously described (Zhang et al., 1999). LR primordia were classified according to the stage of development (four stages: Stage A < 3 cell layers; Stage B > 3 cell layers to pre-emergence; Stage C postemergence to < 5 mm and Stage D > 5 mm LRs). The relative frequency of each of the four developmental stages within each 1 cm segment of the primary root were calculated and compared between the control and the mutants.

Localized NO3 treatment

The experiments of localized NO3 treatment were carried out as previously described (Zhang and Forde, 1998). The seeds were incubated on the basic medium supplemented with 10 μm NH4NO3 and 0.5% sucrose for 4 days to allow the primary roots to reach 2 cm. The seedlings were then transferred to segmented plates (four seedlings per plate) with the middle segment supplemented with 1.0 mm of either KNO3 (treatment) or KCl (control). The length of individual LR in the middle segment was measured using a ruler. The stimulation index was calculated as the LR length in the KNO3-treated segment divided by that in the KCl-treated segment.

Acknowledgements

We thank Professor Koornneef of Wageningen University for providing the aba1-1 and aba3-2, Dr Gibson of Rice University, Houston, USA for the abi2-3 and abi2-4, Professor Smeekens of University of Utrecht for the sun6-2 and mig3, Dr Finkelstein of University of California at Santa Barbara for the abi1-1, abi2-1, abi3-1 and abi5-1, Professor Micol of Universidad Miguel Hernandez for the abi4-2, Dr Graham of University of York for the cai181 and Professor Frommer of Universität Tübingen, Germany for the rsr1-1 seeds. Thanks are also due to the Nottingham Arabidopsis Stock Centre for supplying seeds of Arabidopsis thaliana ecotype Columbia (Col-0), Landsberg (Ler-1), Wassilewskija (Ws) and the abi4-1 mutant. H.Z. is supported by a research fellowship from the University of Leeds and a Research Grant from the Royal Society. An EU Marie Curie Postgraduate Fellowship awarded to LIBA supports I.D.S.

Ancillary