Complexity and coordination of root growth at low water potentials: recent advances from transcriptomic and proteomic analyses

Authors


R. E. Sharp. Fax: +1 573 882 1469; e-mail: SharpR@missouri.edu

ABSTRACT

Progress in understanding root growth regulation and adaptation under water-stressed conditions is reviewed, with emphasis on recent advances from transcriptomic and proteomic analyses of maize and soybean primary roots. In both systems, kinematic characterization of the spatial patterns of cell expansion within the root elongation zone showed that at low water potentials, elongation rates are preferentially maintained towards the root apex but are progressively inhibited at more basal locations resulting in a shortened growth zone. This characterization provided an essential foundation for extensive research into the physiological mechanisms of growth regulation in the maize primary root at low water potentials. Recently, these studies were expanded to include transcriptomic and cell wall proteomic analyses of the maize primary root, and a proteomic analysis of total soluble proteins in the soybean primary root. This review focuses on findings related to protection from oxidative damage, the potential roles of increased apoplastic reactive oxygen species in regulation of wall extension properties and other processes, region-specific phenylpropanoid metabolism as related to accumulation of (iso)flavonoids and wall phenolics and amino acid metabolism. The results provide novel insights into the complexity and coordination of the processes involved in root growth at low water potentials.

INTRODUCTION

Under water stress conditions, some types of roots have the ability to continue elongation at low water potentials that completely inhibit shoot growth (Sharp & Davies 1979; Westgate & Boyer 1985; Sharp, Silk & Hsiao 1988). This is considered an important feature of plant adaptation to water-limited conditions that helps to maintain an adequate plant water supply (Sharp & Davies 1989; Ober & Sharp 2007). The ability to maintain elongation at low water potentials is pronounced in the primary root of a range of species, which helps seedling establishment under dry conditions by ensuring a supply of water before shoot emergence (Sharp et al. 1988; Spollen et al. 1993; van der Weele et al. 2000); this response is illustrated for maize and soybean seedlings in Fig. 1. Detailed understanding of the physiological mechanisms involved has been mostly limited to the primary root of maize, which has been studied extensively by Sharp and co-workers using protocols that allow precise and reproducible imposition of water deficits. Key findings in this system were reviewed by Sharp et al. (2004) and Ober & Sharp (2007) and, therefore, are summarized only briefly in this report. Recently, studies were expanded to the primary root of soybean in order to compare and contrast responses to water stress in the roots of important monocotyledonous and dicotyledonous crop species (Yamaguchi et al. 2009). In both cases, the research has taken advantage of a kinematic approach, that is, the study of spatial and temporal patterns of cell expansion within the root elongation zone (Erickson & Silk 1980; Sharp et al. 1988; Walter, Silk & Schurr 2009), which greatly facilitated the discovery of mechanisms involved in the response of root growth to water stress conditions. This review highlights recent advances gained from studies in which the kinematic approach was combined with transcriptomic and proteomic analyses both to build on the existing physiological foundation and to reveal novel insights into the complexity and coordination of metabolic processes involved in the growth responses to water stress of the primary roots of maize and soybean.

Figure 1.

Elongation rates of the primary root (solid circles) and shoot (open triangles) of maize and soybean seedlings at various water potentials; the data are plotted as a percentage of the rate under well-watered conditions. After germination, seedlings were transplanted into vermiculite at different water potentials (obtained by thorough mixing with different amounts of water) and grown at 29 °C and near-saturation humidity in the dark to minimize evaporative water loss. Accordingly, the seedlings were exposed to constant conditions of water potential. For roots, data were evaluated when elongation rates were steady; for shoots, data represent maximum elongation rates obtained after transplanting. The transcriptomic and proteomic studies reviewed in this report were conducted at a water potential of −1.6 MPa, which imposed a severe water stress treatment during which primary root but not shoot elongation occurred. Reproduced from Spollen et al. (1993) with permission from Bios Scientific Publishers.

SPATIAL PATTERN OF THE RESPONSE OF ELONGATION RATE TO WATER STRESS IN MAIZE AND SOYBEAN PRIMARY ROOTS

Knowledge of the spatial pattern of growth rates within tissues can be a powerful tool to investigate the effects of environmental variation on plant development (Erickson & Silk 1980; Walter et al. 2009). The advantage of this approach is exemplified by studies of the growth responses to water stress in primary roots of maize (Sharp et al. 1988; Liang, Sharp & Baskin 1997; Fan & Neumann 2004) and soybean (Yamaguchi et al. 2009). As shown in Fig. 2, the spatial distribution of the response of elongation rate to water stress is very similar in the two species, with elongation being preferentially maintained towards the root apex. Remarkably, local elongation rates are fully maintained in the early ontogenetic phases of growth even under severe water stress (water potential of −1.6 MPa). However, deceleration and cessation of expansion occur closer to the apex than in well-watered roots, resulting in a shortened elongation zone. These patterns allow the identification of three contiguous regions with distinct elongation characteristics: region 1, which encompasses the apical region in which elongation rates are maintained in water-stressed roots; region 2, in which elongation rates reach a maximum in well-watered roots but are progressively inhibited under water stress; and region 3, in which elongation decelerates in well-watered roots and is completely inhibited in water-stressed roots.

Figure 2.

(a) Displacement of marks away from the apex of maize primary roots during a 3.5 h period following marking for roots growing under well-watered (WW) or water-stressed (WS, water potential of −1.6 MPa) conditions (grown as described in Figure 1). The white lines indicate the vertical displacement of the root apices and of marks originally located at 5 and 10 mm from the apices during the growth period. Separation of marks, and hence tissue expansion, is apparent throughout the apical cm in the WW root. In the root growing at low water potential, in contrast, separation between marks occurred only in the apical region, illustrating that water stress resulted in a shortened elongation zone. Reproduced from Sharp et al. (1988) with permission from the American Society of Plant Biologists. (b) Displacement velocity as a function of distance from the root cap junction of primary roots of WW and WS maize seedlings (cv. FR697). The inset shows the profile of displacement velocity for WW and WS soybean primary roots (cv. Magellan). Relative elongation rates (h−1) are obtained from the derivative of velocity with respect to position. The maize and soybean velocity curves are reproduced from Sharp et al. (2004) & Yamaguchi et al. (2009) with permission from Oxford University Press and Wiley-Blackwell, respectively; the original data were calculated from root elongation rates and cell length profiles as described by Silk, Lord & Eckard (1989). Regions 1 to 3, as described in the text, are indicated. The green arrows indicate the comparisons between treatments within regions 1 and 2 and the comparison of WS region 2 with WW region 3 which were made in the transcriptomic and proteomic analyses reviewed in this article.

This spatial characterization provides an essential foundation for investigation of the mechanisms of growth regulation in water-stressed roots. Clearly, different mechanisms can be expected to underlie the distinct responses to water stress in the different regions, and investigation of mechanisms involved in the maintenance of elongation need to focus on the apical region. However, the inhibition of elongation in the basal region is also important to understand, because this is probably part of a coordinated response to allow the preferential utilization of limited resources (water and growth substrates) in the apical region. In addition to within-region comparisons, the kinematic analysis also allows consideration of effects that may be linked with developmental changes associated with the shortening of the elongation zone rather than to specific responses to water stress. Accordingly, as illustrated in Fig. 2, changes that are observed in region 2 of water-stressed roots can also be compared with region 3 of well-watered roots, which exhibits comparable elongation characteristics. This approach has been effectively utilized in recent transcriptomic and proteomic analyses to help distinguish specific responses of gene expression and protein abundance to water stress from changes that may merely reflect the occurrence of growth deceleration and tissue maturation closer to the apex (Fig. 3). These results are discussed in more detail below.

Figure 3.

Numbers of water stress-responsive maize transcripts (Spollen et al. 2008), maize cell wall proteins (Zhu et al. 2007) and soybean proteins (Yamaguchi et al. 2009) in region 1 (R1) and region 2 (R2) of the primary root elongation zone. The numbers of responsive transcripts and proteins in R2 which exhibited similar and significant responses when R2 of water-stressed (WS) roots was compared with region 3 (R3) of well-watered (WW) roots are indicated in parentheses and marked with asterisks; these changes are considered to be independent of developmental changes associated with the stress-induced shortening of the elongation zone and, therefore, are likely to be specific responses to water stress.

REGULATION OF MAIZE PRIMARY ROOT GROWTH AT LOW WATER POTENTIALS – PHYSIOLOGICAL BACKGROUND

Physiological research on mechanisms involved in growth responses of the maize primary root to water stress has focused on three primary areas: (1) the role of abscisic acid (ABA) accumulation; (2) the relationship of osmotic adjustment to root growth maintenance; and (3) modification of cell wall extension properties.

Growth-sustaining role of ABA accumulation

Although hormones are likely to play important roles in root growth regulation under water-stressed conditions, the involvement of most of these compounds has not been elucidated. The exception is the accumulation of ABA, which was investigated in maize primary roots by using ABA-deficient mutants and inhibitors of ABA synthesis to decrease endogenous ABA levels in seedlings growing at low water potentials. The results have changed the view of the role of ABA from the traditional idea that the hormone is generally involved in growth inhibition (reviewed in Sharp 2002). Instead, it was discovered that accumulation of ABA, which increases towards the root apex (Saab, Sharp & Pritchard 1992), is required for the maintenance of elongation in the apical region of the elongation zone at low water potentials (Saab et al. 1990; Sharp et al. 1994), and this response was shown to involve prevention of excess ethylene production (Spollen et al. 2000). In associated work, the action of ABA in maintaining root elongation was shown to involve changes in electrophysiology (Ober & Sharp 2003). The results indicated that under water stress, set points for ion homeostasis shifted in association with the maintenance of elongation rates in the apical region, and that ABA plays a role in regulating the ion transport processes involved in this response. Additional studies suggested that ABA accumulation might also play regulatory roles in the responses of both osmotic adjustment and cell wall extensibility in water-stressed maize roots, as detailed below.

Relationship of osmotic adjustment to root growth maintenance

The elongation zone of the maize primary root exhibits a substantial capacity for osmotic adjustment at low water potentials (Sharp, Hsiao & Silk 1990). By applying kinematic principles to characterize the deposition rates of water and solutes within the elongation zone, it was demonstrated that the rate of proline deposition (nmoles proline mm−1 length h−1) increased dramatically in the apical region under water stress; the resulting increase in proline concentration was responsible for as much as 45% (in the apical mm) of the total osmotic adjustment (see Fig. 7, inset; Voetberg & Sharp 1991). In contrast, in the basal region, the decrease in osmotic potential in water-stressed roots compared with well-watered roots could be accounted for by decreased rates of dilution of other solutes, particularly hexoses, as a result of growth inhibition (Sharp et al. 1990). The proline results provided the first demonstration of a major role for increased rates of solute deposition in the osmotic adjustment of growing regions in higher plants exposed to low water potentials. This response is more dynamic than would be expected if the osmotic adjustment were merely an inevitable accumulation of unused solute when growth is inhibited (Munns 1988). Accordingly, it was concluded that osmotic adjustment in the maize primary root elongation zone is likely to be a highly regulated process and an important adaptive response that helps to maintain root growth under water stress conditions.

Figure 7.

Outline of glutamate and proline metabolic pathways and water stress-responsive genes involved in those pathways. Enzymes which were differentially expressed in region 1 (R1) and/or region 2 (R2) of the primary root elongation zone in the maize transcriptome analysis (Spollen et al. 2008) are indicated. Blue or red boxes indicate the fold-change response (up- or down-regulated, respectively) of changes in transcript expression; white boxes indicate no statistically significant regulation under water stress. Dashed arrows indicate multiple steps. The metabolic map is based on Miflin & Habash (2002) and the KEGG Database (http://www.genome.jp/kegg/pathway.html). The inset shows that proline accumulates dramatically towards the apex of water-stressed (WS) compared to well-watered (WW) maize primary roots; the data are reproduced from Voetberg & Sharp (1991) with permission from the American Society of Plant Biologists. AlaAT, alanine aminotransferase; AS, asparagine synthetase; GAD, glutamate decarboxylase; GOGAT, glutamate synthase; GS, glutamine synthetase; PO, proline oxidase; P5CS, pyrroline-5-carboxylate synthetase; SSD, succinic semialdehyde dehydrogenase; KG, α-ketoglutarate.

To address the metabolic basis of the increase in proline deposition in the root elongation zone under water stress, proline synthesis, catabolism and transport rates were measured using radiolabelling flux analysis techniques (Verslues & Sharp 1999). The results indicated that the increase in proline deposition was primarily attributable to increased proline transport to the root tip. Additional work showed that ABA accumulation is required for the increase in proline deposition (Ober & Sharp 1994), suggesting that ABA may play a role in regulating proline transport to the apical region.

Modification of cell wall extension properties

The distinct responses of elongation rate in the apical and basal regions of the elongation zone under water stress conditions were shown to be associated with differential responses of cell wall extension properties (Wu et al. 1996). The possibility that an enhancement of cell wall extensibility may contribute to the maintenance of elongation in the apical region of water-stressed roots was initially suggested from the study of osmotic adjustment. Although the extent of osmotic adjustment was substantial, it was insufficient to maintain turgor at well-watered levels in roots growing under severe water stress. This was assessed directly by measuring the spatial distribution of turgor using a cell pressure probe, which showed that turgor was decreased by over 50% throughout the elongation zone of roots growing at a water potential of −1.6 MPa compared with well-watered controls (Spollen & Sharp 1991). In the apical region, the maintenance of elongation, despite the decrease in turgor, suggested that longitudinal cell wall extensibility increased under water stress, which was confirmed by direct assessment of cell wall extension properties. In the 0–5 mm region, acid-induced cell wall extension was markedly increased in water-stressed roots compared with well-watered roots, whereas in the 5–10 mm region, acid-induced extension was greatly inhibited under water stress (Wu et al. 1996). Further evidence that cell wall extension properties are inhibited in the basal region of the elongation zone of water-stressed maize roots was provided by Fan & Neumann (2004) and Fan et al. (2006).

Additional studies showed that extractable activities of two cell wall proteins with known or proposed wall loosening properties, expansins and xyloglucan endotransglycosylase/hydrolase (XTH), were substantially greater in the apical region of the elongation zone in water-stressed roots compared with well-watered roots (Wu et al. 1994, 1996). In addition, the susceptibility of the walls to exogenous expansins increased in this region of water-stressed roots, indicating that modifications of cell wall structure or chemistry that facilitated expansin accessibility or action also play an important role in the increase in extensibility. On the other hand, expansin activity also increased in the basal region of the elongation zone in water-stressed roots compared with well-watered roots (Wu et al. 1996), which was unexpected given the greatly decreased acid-induced extension in this region. This discrepancy is, presumably, also explained by modifications of cell wall structural composition that cause wall stiffening. This could not be confirmed by tests of expansin susceptibility, however, because the cell walls in this region did not respond to exogenous expansins in either well-watered or water-stressed roots.

Additional work showed that ABA accumulation was necessary for the increase in XTH activity in the apical region of the elongation zone in water-stressed roots (Wu et al. 1994), although in contrast, alterations of expansin gene expression at low water potentials did not appear to be ABA dependent (Wu et al. 2001).

KINEMATIC APPROACH TO TRANSCRIPTOMIC AND PROTEOMIC ANALYSES OF ROOT GROWTH RESPONSES TO WATER STRESS

The spatial characterization of the response of elongation rate to water stress in maize and soybean primary roots (Fig. 2) and the physiological knowledge gained to date provide a firm underpinning for functional genomics studies (Sharp et al. 2004; Poroyko et al. 2005). Transcriptomic and proteomic analyses provide powerful tools to investigate the molecular basis of plant responses to stress. However, when whole organs (or the entire plant) are analysed, the results provide only an averaged depiction of responses that may be highly variable in different regions of the plant. This concern is particularly important when growth responses are of primary interest because of the developmental gradient that occurs within growing regions. The shortening of the elongation zone in water-stressed roots highlights this issue; in addition to the normal developmental gradient that exists within the elongation zone (encompassing zones of both cell division and elongation, elongation only and maturation), the response of elongation rate to low water potential varies with distance from the root apex. Accordingly, it is not surprising that in several recent studies, water stress-induced changes in transcript expression and protein abundance were reported to be highly divergent in different regions of the elongation zone (Bassani, Neumann & Gepstein 2004; Sharp et al. 2004; Poroyko et al. 2007; Zhu et al. 2007; Spollen et al. 2008; Yamaguchi et al. 2009).

Figure 3 summarizes the numbers of water stress-responsive transcripts and proteins in three studies of root growth responses to water stress, the results of which are synthesized in Figs 4 to 7: (1) a transcriptomic analysis of the maize primary root (Spollen et al. 2008); (2) a proteomic analysis of cell wall proteins (water-soluble and lightly ionically bound fraction) in the maize primary root (Zhu et al. 2007); and (3) a proteomic analysis of total soluble proteins in the soybean primary root (Yamaguchi et al. 2009). The maize studies revealed that the changes in gene expression and cell wall protein abundance in regions 1 and 2 of the elongation zone in response to water stress were largely region specific. This finding emphasizes the importance and effectiveness of the kinematic approach to transcript and protein profiling studies. Furthermore, the comparison of water-stressed region 2 with well-watered region 3 suggested that a large proportion of the stress-induced changes in region 2 might have been attributable to growth deceleration/tissue maturation associated with the shortening of the elongation zone, rather than to specific responses to water stress. Thus, although region 2 showed a greater total number of differentially expressed genes, the analyses suggested that region 1 exhibits a greater number of genes and cell wall proteins that are directly involved in regulating the water stress response. This result is consistent with the induction of a network of adaptive processes in both regions, but, in particular, for the maintenance of elongation in region 1. Figures 4 and 5 provide schematic representations of the predicted functions and interactions of several of the water stress-responsive proteins/genes and associated metabolites in the cell wall and cytosol in regions 1 and 2 of the elongation zone of the maize primary root.

Figure 4.

Schematic representation of predicted functions and interactions of water stress-responsive proteins/genes and associated metabolites in the cell wall and cytosol in region 1 of the elongation zone of the maize primary root. Major differences in response between regions 1 and 2 are indicated by shaded text. Non-italicized text and solid connecting arrows indicate responses that have been demonstrated in published studies; italicized text and dashed connecting arrows indicate responses that are hypothesized to occur but have not yet been demonstrated. Up-regulation under water stress is indicated by the short upward arrows; for ABA, proline and oxalate oxidase (OxO), the double arrows indicate that accumulation or activity in region 1 was greater than in region 2. *Flavonoid accumulation has not been determined in water-stressed (WS) maize roots; however, isoflavonoids were shown to accumulate in region 1 of the soybean primary root under water stress (see Figure 6; Yamaguchi et al. 2009). The inset shows increased apoplastic reactive oxygen species (ROS) in the epidermis of region 1 in WS compared to well-watered (WW) roots, as indicated by confocal microscopy of roots stained with H2DCF (2′,7′-dichlorodihydrofluorescein, green fluorescence), a membrane-impermeable ROS indicator. The right-hand diagram illustrates a transverse view of the root surface and focal planes. The increase in apoplastic ROS in WS roots is consistent with the increases in abundance of superoxide dismutase (SOD) and oxalate oxidase, which contribute to H2O2 production, and peroxidases (POX), which can also contribute to ROS production including the generation of hydroxyl radicals (OH) from H2O2 in the presence of superoxide (O2•−) and/or reductant (e.g. NADH). The ROS images are modified from Zhu et al. (2007) and reproduced with permission from The American Society of Plant Biologists.

Figure 5.

Schematic representation of predicted functions and interactions of water stress-responsive proteins/genes and associated metabolites in the cell wall and cytosol in region 2 of the elongation zone of the maize primary root. Major differences in response between regions 1 and 2 are indicated by shaded text. Non-italicized text and solid connecting arrows indicate responses that have been demonstrated in published studies; italicized text and dashed connecting arrows indicate responses that are hypothesized to occur but have not yet been demonstrated. Up-regulation under water stress is indicated by the short upward arrows. Apoplastic reactive oxygen species (ROS) signalling (as shown for region 1 in Figure 4) may also occur in region 2, but for clarity, this is not shown. The inset shows increased autofluorescence from lignin and other phenolic compounds under ultraviolet excitation in transverse sections from the centre of region 2 of water-stressed (WS) compared to well-watered (WW) roots (M. Yamaguchi & R.E. Sharp, unpublished data). OxO, oxalate oxidase; POX, peroxidase.

Figure 6.

Outline of phenylpropanoid metabolic pathways and water stress-responsive genes/proteins involved in those pathways. Enzymes which were differentially expressed in region 1 (R1) and/or region 2 (R2) of the primary root elongation zone in the soybean total protein analysis (Yamaguchi et al. 2009) and the maize transcriptome analysis (Spollen et al. 2008) are indicated in green and purple, respectively. Blue or red boxes indicate the fold-change response (up- or down-regulated, respectively) of changes in protein abundance or transcript expression; white boxes indicate no statistically significant regulation under water stress. It should be noted that several O-methyltransferases (OMT) were regulated under water stress in the maize transcriptome analysis; based on sequence, the fold-change values shown are likely to represent ZRP4, which is thought to be involved in both lignin and ferulate synthesis (Barrière et al. 2007; Riboulet et al. 2009). Dashed arrows indicate multiple steps. The metabolic map is based on Winkel-Shirley (2001) & Barrière et al. (2007). Insets A and B show, respectively, that isoflavones accumulate towards the apex, whereas lignin staining (Mäule stain) increases towards the base of the elongation zone in water-stressed (WS) compared to well-watered (WW) soybean primary roots. Lignins are stained red-purple (guaiacyl-syringyl lignin) or brown (guaiacyl lignin). Asterisks in inset A denote significant differences between WS and WW values (P < 0.05). These profiles are consistent with the up-regulation of enzymes involved in isoflavonoid biosynthesis in region 1 and the high level of up-regulation of CCoAOMT in region 2 of water-stressed roots. The figures are modified from Yamaguchi et al. (2009) and reproduced with permission from Wiley-Blackwell. CCoAOMT, caffeoyl-CoA O-methyltransferase; CHI, chalcone isomerase; CHR, chalcone reductase; CHS, chalcone synthase; C3H, C3-hydroxylase; F3H, flavone 3-hydroxylase; GT, glucosyl transferase; HCT, hydroxycinnamoyl-CoA shikimate hydroxycinnamoyltransferase; IFR, isoflavone reductase; IFS, isoflavone synthase.

In the total soluble protein analysis in soybean, in contrast, the largest proportion of water stress-induced changes in protein abundance were common to regions 1 and 2 (Fig. 3). Moreover, only three of the total of 31 changes in region 2 did not exhibit similar and significant responses when region 2 of water-stressed roots was compared with region 3 of well-watered roots. Accordingly, the majority of the stress-induced changes in protein abundance in region 2 were likely to be specific responses to water stress. The differences in the soybean proteomic analysis to the results described for the maize transcriptome and cell wall sub-proteome analyses are probably the result of the comparative lack of resolution of the total soluble protein analysis (using two-dimensional gel electrophoresis), which primarily detected only the most abundant cytosolic proteins that were regulated under water stress. These proteins may mostly be required for processes of stress adaptation throughout the elongation zone.

Protection from oxidative damage – an important response throughout the elongation zone

The largest functional category of stress-responsive proteins in the soybean study (most of which were increased in abundance) was the control of reactive oxygen species (ROS) metabolism (Yamaguchi et al. 2009). In the maize analysis, consistently, the expression of transcripts for proteins involved in ROS metabolism tended to be up-regulated in both regions 1 and 2, in comparison to the predominantly region-specific patterns of expression of transcripts in other functional categories (Spollen et al. 2008). Abiotic stresses including water stress often result in increased ROS production, which may cause oxidative damage (Iturbe-Ormaetxe et al. 1998; Apel & Hirt 2004; Moller, Jensen & Hansson 2007). Accordingly, the commonality of the results related to control of ROS in the maize and soybean systems indicates that protection from oxidative damage is probably of general importance in the adaptation of the root elongation zone to water stress conditions, rather than being specifically associated with growth maintenance in region 1 or growth inhibition in region 2 (Figs 4 & 5).

Water stress-regulated transcripts and proteins related to control of ROS metabolism in the maize and/or soybean studies included several ROS-scavenging proteins, for example, catalase, glutathione transferase and peroxidases. In addition, several metal-chelating proteins including metallothioneins and ferritins were up-regulated in both regions 1 and 2, indicating that control of free metal ions is also important for adaptation. Metallothioneins chelate heavy metal ions such as iron and copper, whereas ferritin exclusively sequesters iron. Free iron and copper ions catalyse the Fenton reaction that generates hydroxyl radicals, the most harmful ROS, from hydrogen peroxide. In water-stressed soybean roots, the increased abundance of ferritin proteins was shown to effectively sequester more iron and thereby prevent excess free iron throughout the elongation zone (Yamaguchi et al. 2009).

Excess ROS can oxidize cell components including proteins, lipids and sugars, and can also trigger apoptosis-like cell death (Solomon et al. 1999; Gechev et al. 2006). Accordingly, the up-regulation of several proteinase inhibitors in regions 1 and/or 2 of both maize and soybean [including cysteine proteinase inhibitor (maize and soybean), subtilisin–chymotrypsin inhibitor (maize) and trypsin inhibitors (soybean)] is likely to help prevent the degradation of oxidized proteins, allowing time for recovery [possibly by glutathionation involving the up-regulation of glutathione transferase activity; Moons (2005)]. Cysteine proteinase inhibitor activity may be particularly important in the inhibition of apoptosis (Solomon et al. 1999; Belenghi et al. 2003).

The results also suggested that supplemental protection from excess ROS occurs preferentially in region 1 of water-stressed roots. Both isoflavonoids in soybean (Fig. 6) and proline in maize (Fig. 7) were shown to accumulate substantially and preferentially towards the apex of roots growing at low water potentials. Maize does not have the enzymes for isoflavonoid synthesis that were up-regulated in soybean, but up-regulation of enzymes for flavonoid synthesis (Fig. 6) suggests that flavonoids probably accumulate in the elongation zone of water-stressed maize roots; this has not yet been assessed. Both (iso)flavonoids and proline have antioxidant activity (Jovanovic et al. 1994; Nerya et al. 2004; Kruk et al. 2005; Kaul, Sharma & Mehta 2008) in addition to other potential functions in stress adaptation (discussed in the following sections). Accordingly, the preferential accumulation of these compounds in the apical region of water-stressed roots is likely to help in protection from excess ROS, thereby contributing to the maintenance of elongation in this region.

Taken together, the co-regulation of ROS-scavenging proteins, proteinase inhibitors and metabolites with antioxidant activities suggests that a complex and coordinated set of proteins is regulated under water stress to protect the root elongation zone from oxidative damage under water stress conditions (Figs 4 & 5). It is interesting to note that flavonoid synthesis (Ithal & Reddy 2004), proline accumulation (Ober & Sharp 1994) and the expression of several ROS-scavenging proteins (Seki et al. 2002) including catalase (Guan, Zhao & Scandalios 2000) and ferritins (Lobreaux, Hardy & Briat 1993) have been reported to be regulated by ABA, suggesting that ABA accumulation may play an important role in cellular protection from ROS-induced oxidative damage in water-stressed roots of both maize and soybean. Consistent with this hypothesis, the maize vp14 mutant, in which ABA accumulation under water stress is restricted (Sharp 2002), exhibits increased cytosolic ROS levels in the primary root elongation zone under water-stressed conditions (I.J. Cho, M. Sivaguru & R.E. Sharp, unpublished data). As detailed above, accumulation of ABA is essential for the maintenance of elongation in the apical region of the maize primary root at low water potentials; whether ABA is similarly required for the adaptation of soybean primary roots to water stress has not been investigated. The relationship of ABA's potential involvement in ROS metabolism in water-stressed roots to its above-mentioned role in preventing excess ethylene production (Spollen et al. 2000) is under investigation; notably, ROS production and ethylene synthesis have been shown to be interrelated in other systems (Overmyer et al. 2000; Moeder et al. 2002).

Increased apoplastic ROS in water-stressed roots – potential roles in spatial growth regulation

In contrast to the oxidative damage that can occur if ROS levels are excessive, ROS may also play positive roles in regulating the response of root growth to water stress. Thus, it is of particular interest that the maize cell wall proteomic study revealed that in region 1 of water-stressed roots, several proteins which increased in abundance were related to ROS generation (Fig. 4; Zhu et al. 2007). These included two putative oxalate oxidase/germin proteins and a superoxide dismutase [Cu-Zn], which contribute to H2O2 production (Fig. 4). Notably, an oxalate oxidase was also the second-highest up-regulated transcript in region 1 of water-stressed maize roots (Spollen et al. 2008), and histochemical analysis indicated that oxalate oxidase activity increases particularly in the apical region under water-stressed conditions (J. Zhu & R.E. Sharp, unpublished data). Oxalate, the substrate of oxalate oxidase, might be generated from ascorbate or glycolate (Davies & Asker 1983; Green & Fry 2005). Consistent with these results, increased apoplastic ROS levels were demonstrated in region 1 of water-stressed roots by quantification of H2O2 content in apoplastic fluid and by in situ imaging (Fig. 4, inset). The increased ROS levels were sustained during growth under steady conditions (as described in Fig. 1) rather than a transient event following stress imposition, suggesting that the response may be associated with a continuing process of stress adaptation. Clearly, given the maintenance of elongation under water stress in region 1 (Fig. 2), the increase in apoplastic ROS in this region seems likely to be associated with positive rather than negative consequences. Oxalate oxidase proteins also increased in abundance in regions 2 and 3 of water-stressed roots, again predicting increased production of H2O2; however, this could not be confirmed because of technical limitations of the assay procedures in these regions.

To test if enhanced oxalate oxidase activity and apoplastic ROS affect root elongation, a transgenic maize line constitutively expressing a wheat oxalate oxidase (Ramputh et al. 2002) is being evaluated. Initial results show that under well-watered conditions, primary root elongation rate increased by as much as 30% compared with the segregated transgene null line (J. Zhu, J. Simmonds & R.E. Sharp, unpublished data). Interestingly, oxalate oxidase activity was increased particularly in the apical region of the elongation zone, simulating the effect of water stress. Taken together, these results support the hypothesis that the increase in oxalate oxidase activity and apoplastic ROS in the apical region of water-stressed roots is positively associated with the maintenance of elongation in this region. Experiments to determine the mechanisms of root growth regulation by oxalate oxidase/apoplastic ROS are in progress.

At least two possible mechanisms can be postulated for promotion of root elongation by apoplastic ROS (Fig. 4). Firstly, generation of hydroxyl radicals from H2O2, by either the Fenton reaction or peroxidase activity in the presence of superoxide and/or reductant (e.g. NADH), can play a direct role in cell wall loosening via polysaccharide cleavage (Fry 1998; Liszkay, Kenk & Schopfer 2003; Passardi, Penel & Dunand 2004; Schopfer & Liszkay 2006; Kukavica et al. 2009; Müller et al. 2009). This possibility is of particular interest because of the above-described increase in longitudinal cell wall extensibility in region 1 of the maize primary root under water-stressed conditions. Importantly, there is evidence for this mechanism of wall loosening in the elongation zone of well-watered maize primary roots (Liszkay, van der Zalm & Schopfer 2004), although the possible modification of this activity in the response of root growth to water stress has not been investigated. Secondly, ROS have been proposed to act as signalling molecules in various processes (Moller et al. 2007); for example, apoplastic hydroxyl radicals have been shown to activate Ca channels that are necessary for growth of root hairs (Foreman et al. 2003).

There is evidence that apoplastic superoxide is localized preferentially in the apical region of the maize primary root under well-watered and osmotically stressed conditions (Liszkay et al. 2004; Bustos et al. 2008). In addition to being a potential source of the increase in H2O2 production in region 1 of water-stressed roots (together with the increase in oxalate oxidase activity), superoxide may also interact with peroxidases to convert the H2O2 to hydroxyl radicals, as noted above. Superoxide can also reduce Fe3+ back to Fe2+ to sustain the Fenton reaction, thereby increasing hydroxyl radical generation. Accordingly, the apical localization of superoxide may be an important factor in determining the specificity of growth maintenance in the apical region of water-stressed roots. Plasma membrane NADPH oxidase is an important source of apoplastic superoxide (Foreman et al. 2003), and up-regulation of NADPH oxidase by both water stress and ABA treatment was reported in maize leaves (Jiang & Zhang 2002). However, four NADPH oxidase-related sequences that were included in the maize transcriptomic analysis were not differentially expressed in either region 1 or 2 of water-stressed maize roots (Spollen et al. 2008).

Apoplastic ROS can also have wall tightening effects, involving oxidative cross-linking of cell wall phenolics (Fig. 5). As noted above, apoplastic ROS generation probably also increased in region 2 of water-stressed maize roots (in this regard, the transcriptomic analysis suggested that glycolate oxidase, which was up-regulated in this region, might contribute to production of oxalate, the substrate of oxalate oxidase). Notably, transcripts for both O-methyltransferase and extensin-like protein were up-regulated in region 2 under water-stressed conditions (Spollen et al. 2008); O-methyltransferase is probably involved in the synthesis of wall phenolics (Riboulet et al. 2009). Peroxidase catalyses the polymerization of phenolic compounds as well as the tyrosine residues of extensin using hydrogen peroxide (Passardi et al. 2004), thereby giving mechanical strength to cell walls. Consistent with these results, increased levels of wall phenolics were observed in the basal region of the elongation zone in water-stressed maize roots (Fig. 5, inset), as shown previously by Fan et al. (2006). This response may play an important role in decreasing longitudinal wall extensibility in region 2 under water-stressed conditions. Apoplastic ROS signalling may also occur in region 2, but for clarity, this is not shown in Fig. 5.

Taken together, the cell wall proteomic and transcriptomic results suggest that regulation of apoplastic ROS has critical roles in determining both the maintenance of elongation in the apical region and the inhibition of elongation in the basal region of water-stressed maize roots.

Phenylpropanoid metabolism – region-specific accumulation of (iso)flavonoids and wall phenolics

The transcriptomic and proteomic analyses indicated that the synthesis of phenylpropanoids including (iso)flavonoids and wall phenolics in the root elongation zone is regulated in a complex and region-specific manner in response to water stress. The total protein analysis of water-stressed soybean roots (Yamaguchi et al. 2009) showed that chalcone synthase 7 increased in abundance in region 1 and chalcone reductase and two isoflavone reductases increased in both regions 1 and 2 (Fig. 6). These three enzymes are involved in isoflavonoid biosynthesis; in particular, chalcone synthase 7 has been reported to play a critical role in isoflavonoid synthesis in soybean seeds (Dhaubhadel et al. 2007). Similarly, the maize transcriptome analysis showed that enzymes involved in flavonoid synthesis were up-regulated under water stress; putative flavone hydroxylase and chalcone isomerase were up-regulated in regions 1 and 2, respectively (Fig. 6). Accordingly, the results suggested that (iso)flavonoids accumulate in the root elongation zone under water stress, perhaps especially in region 1 in the case of soybean. As already mentioned, this prediction was confirmed by analysing the spatial profile of isoflavone composition in the elongation zone of soybean roots, which showed that isoflavonoids accumulated towards the root apex in roots growing at low water potentials (Fig. 6, inset).

As discussed above, both flavonoids and isoflavonoids are potent scavengers of ROS and, therefore, the preferential accumulation of these compounds in the apical region of water-stressed roots is likely to help in protection from oxidative damage. In addition, these compounds have other functions in plant cells that can be postulated to have roles in water stress tolerance mechanisms. Of particular interest is evidence of their ability to influence auxin transport by inhibiting PIN (PIN-FORMED) proteins, thereby causing the accumulation of auxin in nearby tissues (Mathesius et al. 1998; Subramanian, Stacey & Yu 2006). Interestingly, water stress has been shown to cause auxin accumulation in the elongation zone of maize primary roots (Ribaut & Pilet 1994), and several auxin-inducible genes/proteins were up-regulated in the elongation zone of both maize and soybean roots in response to water stress (Poroyko et al. 2007; Spollen et al. 2008; Yamaguchi et al. 2009). Furthermore, applied auxin was shown to cause shortening of the elongation zone in well-watered roots of several species in a manner similar to the effect of water stress in maize and soybean (Fig. 2; Hejnovicz 1961; Goodwin 1972; Ishikawa & Evans 1993). Taken together, these observations suggest that auxin, and its regulation by (iso)flavonoids, may be important for controlling the growth response of primary roots to water stress, although this possibility has not yet been assessed. In this regard, it is interesting that, as mentioned above, ABA has been reported to regulate flavonoid synthesis (Ithal & Reddy 2004); accordingly, ABA might play a role in modulating auxin localization in the elongation zone of water-stressed roots.

Whereas the soybean results showed that water stress-induced isoflavonoid accumulation occurs preferentially in the apical region where elongation rates are maintained, the results from both maize and soybean roots indicate that synthesis of wall phenolics including lignin and ferulates, which constitute other major branch pathways in phenylpropanoid metabolism, is enhanced in region 2 where elongation is inhibited (Fig. 6). As already noted, phenolic components can result in cross-linking between wall polymers, and accumulate in the basal region of the elongation zone in water-stressed maize primary roots (Fig. 5, inset). Fan et al. (2006) showed that this response occurred in association with up-regulation of cinnamoyl-CoA reductase, a key enzyme in lignin synthesis. Consistently in soybean, the recent proteomic analysis (Yamaguchi et al. 2009) showed that caffeoyl-CoA O-methyltransferase (CCoAOMT), which catalyses the fourth step in lignin biosynthesis from p-coumaroyl-CoA, was highly up-regulated in region 2 under water stress, and accumulation of lignins was demonstrated in the basal region of the elongation zone (Fig. 6, inset).

Interestingly, in maize, the cell wall proteomic analysis showed that several β-d-glucosidases markedly increased in abundance in regions 2 and 3 of the elongation zone under water-stressed conditions (Fig. 5; Zhu et al. 2007). Apoplastic β-d-glucosidases have been implicated in lignin synthesis, via the cleavage of monolignol glucosides that are synthesized inside the cell and secreted into the cell wall (Whetten, Mackay & Sederoff 1998). Therefore, the increased abundance of β-d-glucosidases may play a role in the accumulation of lignin in the basal region of the elongation zone under water stress. It should be noted that most of these β-d-glucosidases were also highly up-regulated in region 1 of water-stressed roots. Their function in this region is not known; an intriguing possibility, as discussed by Zhu et al. (2007), is that they might function in release of ABA and/or other hormones from conjugated forms.

In maize, additionally, O-methyltransferase (ZRP4) transcript expression was highly down-regulated in region 1 and up-regulated in region 2 (different isoforms) under water stress (Fig. 6). This enzyme is closely associated with production of both monolignols and ferulates (Barrière et al. 2007; Riboulet et al. 2009). These results suggest that ferulate synthesis may be decreased and increased, respectively, in regions 1 and 2 of water-stressed roots. Ferulates are abundant in the cell walls of monocotyledonous plants and have a role in cross-linking wall polysaccharides. Increases in wall-bound ferulates have been shown to correlate closely with decreases in wall extensibility with tissue aging (Kamisaka et al. 1990). Interestingly, in wheat coleoptiles, both osmotic stress and ABA treatment suppressed the increase of ferulate and diferulate in the cell walls which occurs during normal development, and this was associated with suppression of cell wall stiffening (Wakabayashi, Hoson & Kamisaka 1997a,b). These results suggest the possibility that the pronounced accumulation of ABA in the apical region of water-stressed roots (Saab et al. 1992) may restrict ferulate synthesis and, thereby, contribute to the enhancement of wall extensibility and maintenance of elongation in this region.

Importantly, the (iso)flavonoid and wall phenolic branch pathways of phenylpropanoid metabolism share a common precursor, p-coumarate-CoA (Fig. 6). Evidence for substrate competition between lignin synthesis and flavonoid synthesis was provided by Besseau et al. (2007) in a study of Arabidopsis; inhibition of lignin biosynthesis (by gene silencing) induced accumulation of flavonoids in the leaves. Similarly, coordinated balance of metabolite flux between (iso)flavonoid and wall phenolics synthesis could play a critical role in determining the spatial regulation of root elongation under water stress (Yamaguchi et al. 2009).

Glutamate and proline metabolism

As detailed above, previous studies showed that proline accumulates to high concentrations towards the apex of water-stressed maize primary roots (Fig. 7, inset; Voetberg & Sharp 1991), and the radiolabelling flux analysis results of Verslues & Sharp (1999) suggested that this response is primarily attributable to increased proline transport to the root tip rather than local changes in proline metabolism. Nevertheless, the maize transcriptomic analysis provided additional information on changes in amino acid metabolism that may supplement this response (Fig. 7). Firstly, pyrroline-5-carboxylate synthetase, which catalyses the rate-limiting step in proline synthesis from glutamate, was up-regulated in region 1 under water stress, indicating that local proline synthesis may have increased to some extent. Secondly, a putative proline oxidase was down-regulated in region 1. As discussed by Verslues & Sharp (1999), although their results suggested that proline catabolism in the apical region was not suppressed relative to well-watered roots, this does not imply a lack of regulation; it is unlikely that proline could accumulate to such high levels if proline catabolism increases in proportion to proline concentration.

In region 2, pyrroline-5-carboxylate synthetase and proline oxidase were further up- and down-regulated, respectively, in response to water stress relative to their responses in region 1 (Fig. 7). In addition, several other enzymes involved in glutamate biosynthesis were up-regulated, indicating increased production of glutamate through the glutamate synthase cycle and from α-ketoglutarate in the tricarboxylic acid (TCA) cycle. The evidence for activated proline synthesis, taken together with the fact that proline deposition rates decreased in region 2 in water-stressed roots compared with well-watered roots (Voetberg & Sharp 1991), suggests that the basal region of the elongation zone may serve as a source of proline for transport to region 1. Indeed, proline deposition rates in region 2 were negative under water stress in the study of Verslues & Sharp (1999), indicating a net loss of proline from this region. Nevertheless, as analysed in detail in that paper, the amount of proline transported from the basal to the apical region may be limited relative to that from the seed or other parts of the plant. In addition to its roles in osmotic adjustment and protection from excess ROS, the proline produced in region 2 could also be utilized for synthesis of extensin, a hydroxyproline-rich protein; as noted above, extensin-like protein was up-regulated in region 2 under water-stressed conditions.

Glutamate may also be used for γ-aminobutyrate acid (GABA) synthesis (Fig. 7), and glutamate decarboxylase, which catalyses the conversion of glutamate to GABA, was highly up-regulated in region 2 of water-stressed roots. The production of GABA may have at least two roles in water stress tolerance, the GABA shunt and pH regulation. Succinic semialdehyde dehydrogenase, which catalyses the last step in the GABA shunt, was also up-regulated in region 2 under water stress (Fig. 7). The GABA shunt provides an alternate path for the TCA cycle, bypassing mitochondrial enzymes including α-ketoglutarate dehydrogenase that may be degraded during oxidative stress. Therefore, under water-stressed conditions, the shunt may have important roles to prevent uncoupling of mitochondrial electron transport and, thereby, limit the generation of ROS, and to maintain the TCA cycle (Bouché & Fromm 2004). GABA, together with other polyamines, can also help to maintain a stable cytosolic pH because of its H+ buffering properties. Consistently, lysine decarboxylase and arginine decarboxylase, which produce polyamines from lysine and arginine, were also up-regulated in region 2 under water stress (not shown in Fig. 7).

CONCLUSION

The recent transcriptomic and proteomic studies reviewed here have built on previous physiological knowledge to gain novel insights into the coordination of root growth regulation and adaptation under water-stressed conditions. Understanding has been greatly facilitated by taking advantage of a kinematic approach to transcript and protein profiling. Many of the processes involved in root growth adaptation, and the underlying coordination of gene networks, proteins and metabolites, are controlled in a region-specific manner in association with the distinct region specificity of growth regulation. In this review, we have focused on selected processes for which the available information allows a framework of understanding. Clearly, much additional research is needed to more fully explore the complexity of root growth adaptation to water stress; for example, the roles of hormones other than ABA, and the interactions between different hormones, remain poorly understood. Continued progress in understanding of root growth regulation under water stress will lead to novel approaches for improving drought tolerance through genetic and metabolic engineering of root function.

ACKNOWLEDGMENTS

Preparation of this review was supported by a grant from Monsanto to R.E.S. and by the University of Missouri Food for the 21st Century Program. We thank Dr. Dale Blevins (University of Missouri) and Dr. Mel Oliver (USDA, Columbia, Missouri) for useful discussions and helpful comments on the manuscript, and Dr. Mayandi Sivaguru (University of Illinois at Urbana-Champaign) for the diagram in the inset to Fig. 4.

Ancillary