Ascorbic acid (AA) fulfils many essential functions in plants. It is a key antioxidant and an important reducing substrate for a number of enzymes. The effects of low AA on plant architecture and leaf ultrastructure were studied in Arabidopsis thaliana mutants, which have constitutively moderately low (vtc1) or very low (vtc2) leaf AA contents compared with the wild type. Shoot development was comparable in all accessions over the first 14 d of growth. The production of primary roots was slightly different in vtc1, vtc2, and wild-type plants. However, the most notable difference was that a high proportion of the primary roots of the vtc2 plants grown on soil had lost the wild-type responses to gravity. The vtc mutants showed the antagonistic interaction between nitrate and sugar in the regulation of lateral root (LR) development that was observed in the wild type. However, the vtc2 mutants produced greater numbers of longer LRs than wild-type or vtc1 plants at all levels of nitrate. At later stages of development, the vtc rosettes were smaller than those of the wild type and the leaves showed intracellular structural changes that are consistent with programmed cell death (PCD). PCD symptoms such as nuclear chromatin condensation, the presence of multivesicular bodies, and extensive degradation and disorganization of the grana stacks were observed in 8-week-old vtc2 leaves and in 10-week-old vtc1 leaves. The data presented here illustrate the importance of tissue AA contents in regulating whole plant morphology, cell structure, and development.
programmed cell death
reactive oxygen species
Ascorbic acid (AA; vitamin C) is synthesized and accumulated in large amounts in leaves (Fry, 1998; Potters et al., 2000, 2004; Foyer and Noctor, 2005). It is a multifunctional metabolite with crucial roles in antioxidant metabolism, enzyme activity, and associated regulation. AA accumulation is not only important in stress protection but it is also considered to influence plant growth and development through effects on the cell cycle and cell elongation (Foyer et al., 2006). Plant organs provide essential vitamin C for the human diet. In animals, vitamin C is essential for the production of collagen, and is also considered to be crucial for health and well being, particularly with regard to protection against premature ageing (Partridge and Gems, 2002; Finkel and Holbrook, 2003). Similarly, low tissue AA contents are linked to premature senescence and programmed cell death (PCD) in plants (Barth et al., 2004, 2006; Pavet et al., 2005). Healthy ageing is a key objective in animals (Kurzweil and Grossman, 2005) and likewise the prevention of stress-induced premature senescence is an important target in crop improvement programmes. Hence, it is important to have a full understanding of how AA interacts with the processes that control plant growth and development (Foyer et al., 2006).
Very few AA-deficient mutant plant genotypes have been described in the literature. Arabidopsis thaliana mutants that are deficient in AA were isolated in screens detecting either enhanced ozone sensitivity (Conklin et al., 1996) or nitroblue tetrazolium-based screens (Jander et al., 2002) The vtc1 mutant was isolated in the ozone sensitivity screen; it was shown to have a defect in GDP-mannose pyrophosphorylase and to have <30% of the leaf AA present in the wild type (Conklin et al., 2000). This mutant provided strong evidence for the role of GDP-mannose in AA biosynthesis (Wheeler et al., 1998). The vtc2-1 and vtc2-2 mutant have lower levels of leaf AA than vtc1 (Conklin et al., 1996; Pavet et al., 2005). The vtc2 mutants are affected in GDP-L-galactose (GDP-L-Gal) phosphorylase , an enzyme that is at a branch point between AA synthesis and incorporation of l-galactose into polysaccharides (N Smirnoff, personal communication).
It has previously been shown that ectopic expression of L-gulono-1,4-oxidase (GLOase), the last enzyme of the AA biosynthetic pathway in animals, restores wild-type AA levels and the wild-type phenotype to the vtc1 and vtc2 mutants (Radzio et al., 2003). The five vtc mutant lines expressing GLOase had leaf AA contents that were equal to or higher than those of wild-type plants (Radzio et al., 2003) suggesting that the vtc1 and vtc2 phenotypes can be largely ascribed to low AA accumulation.
It was previously shown that the leaves of vtc1 and vtc2 mutants have the same number of cells but they have modified elongation growth patterns (Pavet et al., 2005). Moreover, cell development is altered such that individual cells undergo PCD early in leaf development (Pavet et al., 2005). The vtc plants show enhanced pathogen resistance early in rosette development compared with the wild type. This is associated with constitutive expression of pathogenesis-related (PR) genes and with the movement of non-expressor of PR protein 1 (NPR1) into the nucleus (Pavet et al., 2005). The present study was designed to explore further the effects of low AA on plant architecture and cell structure. Seedling development and establishment was thus compared in the vtc1 and vtc2 mutants and the wild-type plants. Possible roles for AA in the control of primary root and lateral root (LR) development were compared, together with the effects of varying nitrate and sugar concentrations. The results provide evidence that the low AA levels present in the tissues of the vtc1 and vtc2 mutants have very little effect on primary root development. However, the vtc2 mutant shows a much more extensive elaboration of LR formation and root architecture consistent with a general build-up in tissue reactive oxygen species (ROS) levels due to low antioxidant defence. It was shown previously that the vtc leaf cells stopped elongation growth early in development, at about week 6 (Pavet et al., 2005). In the present study, these observations have been extended with an exploration of leaf cell ultrastructure during development. It is shown that the vtc2 mutant leaves show PCD symptoms early in the developmental programme. The vtc1 mutant also shows leaf PCD symptoms, but these are delayed compared with those of the vtc2 mutant leaves. In contrast, the wild-type leaves only show PCD symptoms at the last stages of senescence.
Materials and methods
Seeds of A. thaliana wild type (Col-0) and vitamin C-1 (vtc1-1) and vitamin C-2 (vtc2-1) mutants were obtained from Cornell University (Ithaca, New York, USA; Conklin et al., 2000). These seeds were used to produce an F4 backcrossed population.
Analysis of seedling growth
Analysis of early primary and lateral root development on agar plates
The experiments with uniform nitrate applications were carried out as previously described (Zhang and Forde, 2000; Signora et al., 2001). Sterilized seeds were placed in vertical 9 cm Petri dishes with 20–25 ml of basic medium consisting of 100 μM KCl, 40 μM MgSO4, 20 μM CaCl2, 22 μM NaH2PO4, 0.9 μM MnSO4, 0.09 μM KI, 0.97 μM H3BO3, 0.14 μM ZnSO4, 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 that was supplemented with 0.5% sucrose and 10 μM KNO3, and incubated in the growth room for 3 d to allow germination and initial seedling growth. The seedlings were then transferred to fresh plates (four seedlings per plate) containing different concentrations of KCl (control) or KNO3 (10 μM, 0.1, 1, 10, and 50 mM) and either 0.5 or 2% sucrose for the specified periods. Germination and primary root and LR growth were followed for 20 seedlings of each accession per treatment per experiment over 14 d. Seedlings were grown in a controlled-environment chamber at 25 °C with a 16/8 h light/dark regime. After 14 d growth, the length of the primary root and the length and quantity of LRs were quantified as described by Signora et al. (2001). The length of the primary root and LRs of each individual seedling were measured directly using a ruler. The relative frequency of LRs at developmental stage D within each 1 cm segment of the primary root was calculated and compared between the control and the mutants, as described by Signora et al. (2001).
Analysis of early primary and lateral root development on sand
Wild-type (Col-0), vtc1-1, and vtc2-1 seeds were germinated on sand with water for 6 d. They were then grown for a further 6 d on sand supplied with nutrient solution consisting of 10 μM NH4NO3, 100 μM KCl, 40 μM MgSO4, 20 μM CaCl2, 22 μM NaH2PO4, 0.9 μM MnSO4, 0.09 μM KI, 0.97 μM H3BO3, 0.14 μM ZnSO4, 2 nM CuSO4, 20.6 nM Na2MoO4, 2.1 nM CoCl2, and 3.6 μM Fe-EDTA in controlled-environment chambers that maintained a day/night temperature of 20 °C, a relative humidity of 70%, and a 10 h photoperiod (200 μmol quanta m−2 s−1). Some seedlings were removed at 10 d and photographed (Fig. 1A). Others were removed at 12 d and the length of primary roots was determined (Fig. 1B).
Analysis of seedling establishment on soil
Wild-type (Col-0), vtc1-1, and vtc2-1 seeds were germinated under a propagator on Levingtons F25 compost for 7 d in controlled-environment chambers that maintained a day/night temperature of 20 °C, a relative humidity of 70%, and a 10 h photoperiod (200 μmol quanta m−2 s−1). They were then photographed (Fig. 2A). The percentage of plants with primary roots that did not penetrate the soil was then determined (Fig. 2B). Seedling establishment was also explored in sand as above, and showed similar characteristics to those of plants grown in soil (data not shown). Each experiment involved six pots with a minimum of 400 seedlings.
Plant growth in soil
Wild-type (Col-0), vtc1-1, and vtc2-1 seeds were germinated under a propagator on Levingtons F25 compost for 7 d and then transferred to pots containing the same compost in controlled-environment chambers maintained at a day/night temperature of 20 °C, a relative humidity of 70%, and a 10 h photoperiod (200 μmol quanta m−2 s−1) until senescence. Plants were photographed at 5 weeks (Fig. 3A) and leaf ascorbate measurements were performed as described previously (Pavet et al., 2005)
Leaf sections (1 mm2) were fixed for 2.5 h at 4 °C in 2.5% glutaraldehyde and 3% paraformaldehyde prepared in 0.1 M sodium phosphate buffer (pH 7.2). Tissue was then fixed for a further 2 h in 1% osmium tetroxide prepared in the same buffer. The samples were dehydrated in graded alcohol and embedded in Spurr's resin as described in Jones et al. (2002).
Ultrathin transverse leaf sections (50–60 nm) were prepared using a Reichert ultramicrotome, for transmission electron microscopy. The sections were stained with uranyl acetate and lead citrate. Ultrastructural examination of tissues was performed using both the Philip Tecnai 12 and Zeiss M10 electron microscopes.
Transverse semi-thin leaf sections (0.5 μm) were prepared as described above. The sections were stained with toluidine blue and viewed using a Leica DMR light microscope. Parameters of leaf morphology were quantified using the imaging software (Leica Q550 Win) as described previously by Olmos and Hellin (1998).
Wild-type, vtc1, and vtc2 plants were germinated and grown on sand or soil. Little variation in the rate of germination (data not shown) or seedling development was observed over this period (Fig. 1). After 10 d growth, the wild-type, vtc1, and vtc2 seedlings had similar phenotypes (Fig. 1A). The growth of primary roots in 12-d-old plants was significantly lower than the wild type in vtc1 (Fig. 1B) and significantly higher than the wild type in vtc2 plants (Fig. 1B). These differences in root growth cannot readily be explained in terms of AA availability as vtc1 and vtc2 seedlings have low tissue AA contents compared with the wild type (Pavet et al., 2005). While the growth of the shoot was similar in 7-d-old seedlings of each accession (Fig. 2A), the primary roots of a significant number of vtc2 seedlings failed to penetrate the soil but rather grew over the surface (Fig. 2B).
Effects of nitrate and sucrose on vtc root architecture
The effects of low AA on root architecture were studied in more detail in wild-type and vtc seedlings grown on agar containing a range of nitrate concentrations (0.1, 1.0, 10, and 50 mM) and two sucrose (0.5 and 2%) levels (Fig. 3). Root morphology was analysed after 14 d growth (Fig. 3). Primary root growth was largely unaffected by the NO3 concentration of the media when plants were grown on 0.5% sucrose (Fig. 3A). However, the primary roots of vtc2 mutants were shorter than those of the vtc1 mutants or the wild-type plants (Fig. 3A) at low (<10 mM) nitrate and 2% sucrose. At higher nitrate concentrations and 2% sucrose, primary root length was similar in all lines (Fig. 3D).
The production of LRs and the response to external nitrate were similar in the vtc1 mutant and the wild-type plants (Fig. 3B, C, E, F). The number and length of LRs measured 14 d after germination were decreased in wild-type and vtc1 and vtc2 mutant plants grown with 50 mM nitrate at lower sucrose concentration (Fig. 3B, C). The high-nitrate-dependent inhibition of LR formation was overcome by increasing the sucrose content of the media to 2% in all accessions (Fig. 3E, F). The vtc2 mutants produced a greater number of LRs in the media than the vtc1 mutants or the wild-type plants at all external KNO3 levels at both 0.5% (Fig. 3C) and 2% sucrose (Fig. 3F). On average, the vtc2 mutants produced twice as many LRs as the wild-type plants and vtc1 mutants (Fig. 3C, F). While the average LR length was similar in all lines at low nitrate levels, the average LR length was greatly increased in the vtc2 mutants at high nitrate levels >1 mM compared with the wild-type plants and vtc1 mutants.
Leaf cell structure
For the detailed analysis of leaf cell structure, plants were grown on soil to senescence. As observed previously, the shoots of the vtc1 and vtc2 mutants had substantially less AA and were noticeably smaller than those of the wild-type plants under these conditions (Fig. 4). Ultrathin transverse leaf sections were prepared from the leaves of these plants throughout development for comparisons of cellular structure using transmission electron microscopy. The following features were modified in vtc leaves compared with the wild type.
During the early stages of development, vtc2 leaves contained more chloroplasts per cell than either vtc1 or wild-type leaves. The numbers of wild-type chloroplasts per cell increased with tissue age, whereas the vtc1 chloroplast population remained unchanged throughout development (Fig. 5A). The number of chloroplasts per unit area decreased with development in all accessions (Fig. 5C). Wild-type and vtc1 chloroplasts were of a similar length and longer than vtc2 chloroplasts at each developmental stage (Fig. 5B). These differences in chloroplast length and size at each growth stage are also illustrated in Fig. 6. The relative starch content of wild-type and vtc1 leaves was higher than the amount observed in vtc2, at each stage of development. Although the percentage starch content declined relative to chloroplast age in all three A. thaliana accessions investigated, the reduction was more pronounced in 8–10-week-old vtc1 plants (Fig. 5D).
Grana stacks of vtc2 chloroplasts contained larger quantities of thylakoids than those present in the grana of vtc1 or the wild type (Fig. 7A, C). Grana stacks of vtc2 were typically composed of 8–12 thylakoids, while wild-type and vtc1 grana contained 5–8 thylakoids. Moreover, the grana structure and development in vtc1 and the wild type was shown to be similar during the first 8 weeks of growth (Fig. 6B, E, H). Extensive degradation and disorganization of the grana was observed in 8-week-old vtc2 rosette leaves (Fig. 6I). In vtc1 leaves, this evidence of degradation and disorganization could be seen in 10-week-old rosettes (Figs 6K, 7C).
Small dense spots of osmiophilic material located adjacent to the grana stack (plastoglobuli; arrowed Fig. 6) were observed in vtc1, vtc2, and wild-type chloroplasts. Interestingly, throughout development, vtc2 chloroplasts contained greater numbers of large, more densely packed plastoglobuli (Fig. 6C, F, I, L) than either vtc1 or wild-type chloroplasts. This was more evident in 8–10-week-old vtc2 plants (Fig. 6I, L). A few 10-week-old wild-type chloroplasts exhibited symptoms of senescence, such as dilation of the thylakoid lumen, reduced stromal density, lower numbers of grana and starch, and an increased quantity of plastoglobuli (Fig. 8C).
The ultrastructure of vtc2 organelles
Electron microscopy revealed a significant number of structural anomalies unique to vtc2. For example, during the early stages of development (∼4-week-old plants) and beyond, epidermal vtc2 cells were shown to contain annular plastids (Fig. 9A). Mature vtc2 leaves also contained a considerable number of convoluted chloroplasts, often with inclusions surrounding the cytoplasm and mitochondria (8–10-week-old plants; Fig. 9B, C). A large number of these convoluted chloroplasts surrounded peripheral reticulum, composed of small tubules (∼20 nm in diameter), frequently located near the chloroplast envelope (Fig. 8B) close to the sites of invagination (Fig. 9C).
Occasionally, vacuolar autophagy of vtc2 mesophyll chloroplasts was observed, in leaves of 10-week-old plants, always absorbed by the central vacuole (Fig. 8D, E). Autophagy of the chloroplast was often characterized by the disintegration of external membranes and disorganization of the grana (Fig. 8D, E).
A large number of vtc2 mitochondria contained amorphous inclusions, more frequently observed in 8–10-week-old plants (Fig. 9E). These inclusions were also observed in vtc1 mitochondria (Fig. 7C). In addition, a small number of vtc2 mitochondria present in the mesophyll cells of 8–10-week-old plants contained poorly developed cristae and a less dense matrix (Fig. 9F).
Multivesicular bodies were prevalent in mature vtc2 leaves (8–10-week-old plants), often associated with the central vacuole of mesophyll cells (Fig. 8A). Chromatin condensation was occasionally observed in the nuclei of 10-week-old vtc2 plants. The condensed chromatin was normally attached to the nuclear envelope (Fig. 9D).
Plant development programmes that govern the structure and form of shoots and roots are extremely plastic and respond to internal and external triggers (Deak and Malamy, 2005). ROS are important second messengers in the signalling cascades that underpin hormone-triggered plant growth responses, particularly LR formation (Foreman et al., 2003; Torres and Dangl, 2005). NADPH oxidase-catalysed production of ROS in the apoplast has a broad role in signalling, and is involved in plant responses to hormones, infection, and abiotic stress (Torres and Dangl, 2005). It has been shown previously that ROS generated by the RHD2 NADPH oxidase (atrbohC) are required for LR growth in A. thaliana (Foreman et al., 2003). Similarly, auxin promotes the generation of hydroxyl radicals (OH) in the apoplast, which mediate root gravitropism (Joo et al., 2001) and elongation growth (Schopfer, 2001). While much attention has focused on the roles of ROS as signals influencing plant responses to developmental cues, relatively little attention has been paid to the role of AA in the control of these processes (De Gara, 2004). This is surprising given that the AA pool exerts a profound influence on cellular redox signalling and plant growth responses (Pignocchi and Foyer, 2003; Foyer and Noctor, 2005a, b; Pignocchi et al., 2006). This study was therefore undertaken to explore the effects of low AA on seedling establishment and plant architecture by comparing root and shoot development in A. thaliana accessions with either a moderate (vtc1) or a large (vtc2) reduction in tissue AA contents. While even moderate reductions in leaf ascorbate (as in vtc1 rosettes, Fig. 4) lead to a decrease in shoot growth and biomass after the seedling establishment stage (compare the data in Figs 1 and 2 with those in Fig. 4), many of the more extreme effects of very low AA on plant growth and morphology were observed only in vtc2 mutants and not in the vtc1 plants (Figs 3, 5–9). Not all of the observed changes in plant growth responses and structure can be ascribed to AA deficiency (see, for example, Fig. 1B). Other changes such as early PCD events in leaves and LR development can be attributed to decreased AA availability. It was recently shown that manipulation of the AA pool in the apoplast alone by genetic manipulation of AA oxidase activities alters plant growth in response to hormone triggers (Pignocchi et al., 2006). The vtc1 mutant has no detectable AA in the leaf apoplast (Veljovic-Jovanovic et al., 2001). While there are no data on the AA pool in the apoplast of the vtc2 mutants, it is probable that this mutant is also deficient in apoplastic AA. Hence, at least some of the observation reported here can be attributed to the effects of low AA.
It has been reported previously that the shoots of the vtc1 and vtc2 mutants were considerably smaller than those of wild-type plants (as shown in Fig. 4) and they attained less biomass (Pavet et al., 2005). It is shown here that this low AA-induced phenotype is not present at the seedling establishment stage, whether the plants are grown on agar, sand, or soil (Figs 1–3). These observations are consistent with previous findings from this laboratory showing that, although the vtc1 and vtc2 leaves had the same numbers of cells, they stopped elongation growth early in development (Pavet et al., 2005).
It was previously shown that both the vtc1 and vtc2 leaves display local lesions (Pavet et al., 2005). The data presented here confirm these observations at the microscopic level. The cellular changes reported here are consistent with PCD events. It is shown here that PCD markers are present early in the development of the vtc2 rosettes, which have a low steady-state level of leaf AA. Such markers are found in leaves following exposure to stress or ethylene, or upon initiation of the hypersensitive response (Kratsch and Wise, 2000; Selga et al., 2003). These PCD markers are observed later in the development of the vtc1 rosettes which have more moderate reductions in AA contents. By contrast, PCD markers are only observed in wild-type rosettes at the last stages of leaf senescence. These markers include condensation of chromatin, autophagy, mitochondria alterions, chloroplast invagination, or thylakoid disorganization (Ishikawa, 1996; Yun et al., 1996; Wu et al., 1997; Selga et al., 2003). A structure called the chloroplast peripheral reticulum, which is evident in vtc2 leaves, frequently appears in mesophyll cells as a consequence of chilling or water stress (Musser et al., 1984; Utrillas and Alegre, 1997; Kratsch and Wise, 2000).
The chloroplast peripheral reticulum is implicated in rapid transport of metabolites into or out of the chloroplast because of the large increase in the surface area of the chloroplast inner membrane that it provides (Laetsch, 1968; Kratsch and Wise, 2000). Moreover, the association of the chloroplast peripheral reticulum with the inclusion of the mitochondria supports the hypothesis of enhanced inter-organelle transport and communication as cells progress through the PCD pathway. The chloroplast peripheral reticulum was observed in 8–10-week-old vtc2 leaves, coincident with large numbers of individual dead cells in the vtc2 leaves (Pavet et al., 2005). Cytochemical studies have demonstrated that the peripheral reticulum can also play a role in calcium sequestration (Mosejev and Romanovskaya, 1988). In this regard, Kratsch and Wise (2000) proposed that the peripheral reticulum was involved in calcium modulation of PCD events affecting membrane permeability and metabolite transport across the chloroplast envelope.
The abundance of AA in the mutant leaves appears to influence chloroplast numbers as well as chloroplast function and morphology. While few consistent trends emerge comparing numbers of chloroplasts per cell between wild-type, vtc1, and vtc2 leaves, the data indicate that low AA favours greater numbers of chloroplasts. The vtc2 leaves which have the lowest AA content contained more chloroplasts per cell and larger quantities of thylakoids per chloroplast, particularly during the early stages of development. However, this increase in photosynthetic machinery does not appear to benefit the vtc leaves in terms of improved rates of photosynthesis. The rates of CO2 assimilation are similar in all accessions under optimal growth conditions. The regulation of photosynthesis is similar in wild-type and vtc1 leaves in the absence of stress (Pastori et al., 2003). However, the pathway of zeaxanthin synthesis which requires AA as a reductant is somewhat impaired in vtc2 leaves, and photosynthesis is much more susceptible to stress-induced inhibition (Mulle-Moule et al., 2002).
The data presented here show extensive degradation and disorganization of the grana in 8-week-old vtc2 rosette leaves (Fig. 4I), which have only ∼10% of the AA compared with the wild type at the same stage of development (Pavet et al., 2005). This early degradation of the chloroplasts is clearly linked to low AA, as it occurs later in rosette development (at 10 weeks) in vtc1 leaves that have 30% of the wild-type leaf AA (Figs 4K, 5C) and later still in wild-type leaves.
While chloroplasts from all lines contained plastoglobuli (Fig. 4), the vtc2 chloroplasts contained greater numbers of densely packed plastoglobuli than either vtc1 or wild-type chloroplasts. There is evidence that the number and size of plastoglobuli increase substantially in chloroplasts following exposure to certain stresses, such as ozone (Oksanen et al., 2001), virus infection (Hernandez et al., 2004), chilling (Nordby and Yelenosky, 1984), freezing (Nordby and Yelenosky, 1985), salinity (Hernandez et al., 1995), and drought (Munnë-Bosch et al., 2001). It is presumed that plastoglobuli have a function in the storage of thylakoid components such as lipids, plastohydroquinone, and tocopherol (Murphy, 2001). However, they may also function in the synthesis and recycling of lipophilic products arising from oxidative metabolism during stress. Recent proteomic analyses have revealed that Arabidopsis leaf plastoglubuli contain seven fibrillins and 25 proteins that are probably involved in the metabolism of molecules derived from isoprenoid and lipid pathways, as well as carotenoid cleavage (Ytterberg et al., 2006).
The results presented here provide evidence that low AA influences root morphology and architecture as well as exerting effects on the shoot. Two clear effects were observed. First, the primary roots of vtc2 plants with the lowest AA levels displayed a high frequency of poor gravitropic response (Fig. 2), and, secondly, they produced a greater number of LRs (Fig. 3). LR production and development in Arabidopsis are influenced by developmental and metabolic cues as well as by external stimuli (Zhang and Forde, 2000). The low levels of AA in vtc2 plants affected the growth of the primary root, an organ that is initiated during embryogenesis and whose development is pre-determined. The poor gravitropic responses exhibited by the vtc2 roots illustrate the importance of controlled oxidation in the ROS-mediated auxin-dependent tropic responses (Joo et al., 2001). It has recently been shown that oxidation of the apoplastic AA pool in transgenic tobacco plants with constitutive overexpression of AA oxidase was associated with loss of the auxin response (Pignocchi et al., 2006). As stated previously, the vtc mutants have no AA in the apoplast (Veljovic-Jovanovic et al., 2001). Hence, the data presented here emphasize the importance of the redox state of the apoplast in the modulation of plant growth responses to hormones such as auxin.
Very low AA levels also altered the formation of LRs, which are formed throughout the lifetime of a plant and display remarkable morphogenic plasticity (Zhang and Forde, 2000; Signora et al., 2001). The formation of LRs in A. thaliana is stimulated by low nitrate and inhibited by high concentrations of nitrate, an effect mitigated by increasing sugar in all three lines. Hence, low AA has little detectable effect on the carbon and nitrogen signalling that controls root architecture (Zhang and Forde, 2000; Signora et al., 2001). The enhanced LR growth observed in vtc2 is entirely consistent with the role of ROS in LR growth. ROS produced at specific sites by activation of the AtrbohC NADPH oxidase are required for LR growth. Since in its role as an antioxidant AA removes ROS, it follows that low AA availability in roots will favour root hair formation and growth. Hence low AA might be predicted to favour more elaborate root architecture with better anchorage in the soil and improved uptake of minerals. The increased root area might be beneficial when the plant is exposed to abiotic stress and may serve to compensate to some extent for the relatively poor resistance of the shoot to suboptimal environmental conditions, for example, by allowing better water and nutrient uptake.
This work was funded by a Biotechnology and Biological Sciences Research Council grant BB/C51508X/1. EO is grateful for a Mobility Grant of Researcher from the Spanish Government, Ministerio de Educacion y Ciencia (PR2004-0361). Rothamsted Research receives grant-aided support from the BBSRC (UK).