Nitrogen source interacts with ROP signalling in root hair tip-growth


S. Yalovsky. Fax: +972-3-640-6933; e-mail:


Root hairs elongate in a highly polarized manner known as tip growth. Overexpression of constitutively active Rho of Plant (ROP)/RAC GTPases mutants induces swelling of root hairs. Here, we demonstrate that Atrop11CA-induced swelling of root hairs depends on the composition of the growth medium. Depletion of ammonium allowed normal root hair elongation in Atrop11CA plants, induced the development of longer root hairs in wild-type plants and suppressed the effect of Atrop11CA expression on actin organization and reactive oxygen species distribution, whereas membrane localization of the protein was not affected. Ammonium at concentrations higher than 1 mM and the presence of nitrate were required for induction of swelling. Oscillations in wall and cytoplasmic pH are known to accompany tip growth in root hairs, and buffering of the growth medium decreased Atrop11CA-induced swelling. Fluorescence ratio imaging experiments revealed that in wild-type root hairs, the addition of NH4NO3 to the growth medium induced an increase in the amplitude of extracellular and intracellular pH oscillations and an overall decrease in cytoplasmic pH at the cell apex. Based on these results, we suggest a model in which ROP GTPases and nitrogen-dependent pH oscillations function in parallel pathways, creating a positive feedback loop during root hair growth.


Root hairs serve in water and nutrient uptake and significantly increase the root surface area. Root hairs grow out at the basal end (closer to root tip) of specialized root epidermal cells and, like pollen tubes, expand exclusively at the apical end in a strictly polarized manner by a process known as tip growth. Tip growing cells elongate in an oscillating manner and many cellular events oscillate with the same periodicity as the growth rate (Evans, McAinsh & Hetherington 2001; Parton et al. 2001; Hwang et al. 2005; Monshausen et al. 2007). The highly polarized secretion at the apex is maintained by the spatial organization of the actin cytoskeleton, tip-focused ion gradients and by small G proteins of the Rab, Arf and Rho of Plant (ROP)/RAC families (Kost 2008; Lee & Yang 2008; Yalovsky et al. 2008). Longitudinally oriented actin cables in the shank of the hair are required for myosin-mediated organelle transport through the cytoplasm, whereas fine F-actin structures at the tip promote the transport of secretory vesicles to sites of their fusion with the plasma membrane (Fu, Wu & Yang 2001; Kost 2008). The polar growth of root hairs also requires an oscillatory tip-focused Ca2+ gradient (Monshausen, Messerli & Gilroy 2008) and tip-localized ROS (Foreman et al. 2003). Reactive oxygen species (ROS) produced by NADPH oxidase (RHD2/AtrbohC) are important for the establishment of a tip-focused Ca2+ gradient through the stimulation of plasma membrane Ca2+ channels (Foreman et al. 2003). Elevated Ca2+ levels at the root hair apex were recently demonstrated to stimulate NADPH oxidase activity, creating a positive feedback loop that is likely to contribute to the maintenance of the root hair polarity (Takeda et al. 2008). Tip growth is also associated with oscillatory fluxes of H+ at the apex that correlate with the periodicity of growth (Messerli, Danuser & Robinson 1999; Monshausen et al. 2007). In root hairs, oscillations in extracellular pH and ROS have been shown to modulate tip growth and were predicted to act in a coordinated and complementary mode to regulate root hair elongation. Growth accelerates following reduction of apoplastic pH and slows upon pH increase. The periodicity of growth and pH changes are associated with the spatial distribution of ROS in the apoplast. In growing root hairs ROS accumulates at the shank, away from the tip and after growth ceases it accumulates at the tip as well (Monshausen et al. 2007). It was suggested that such ROS distribution is required for maintaining cell wall firmness at the shank, while allowing expansion at tip, thus leading to polarized unidirectional rather than isoformic growth (Monshausen et al. 2007).

Small GTPases belonging to the ROP family, also called RACs (for the sake of clarity the ROP nomenclature will be used throughout this work), play a central role in establishment of cell polarity, including tip growth of root hairs and pollen tubes (Kost 2008). ROPs localize to the plasma membrane at the apex of growing pollen tubes and root hairs, where they are suspected to activate a range of downstream pathways required for tip growth (Kost et al. 1999; Li et al. 1999; Molendijk et al. 2001; Jones et al. 2002). ROP activity is regulated by cycling between GTP-bound, active and GDP-bound, inactive states. ROP regulatory proteins belonging to the GTPase-activating protein (GAP) and guanyl nucleotide dissociation inhibitor (GDI) families play an important role in the regulation of tip growth. (Carol et al. 2005; Klahre et al. 2006; Klahre & Kost 2006; Hwang et al. 2008). In agreement, earlier work has demonstrated that overexpression of constitutively active mutants of ROPs (conserved dominant mutations that abolish the GTPase activity) depolarizes growth of both root hairs and pollen tubes (Kost et al. 1999; Li et al. 1999; Fu et al. 2001; Molendijk et al. 2001; Jones et al. 2002; Chen, Cheung & Wu 2003; Bloch et al. 2005).

Downstream pathways activated by ROP GTPases include regulation of cytoskeletal dynamics and vesicular trafficking, production of ROS, intracellular Ca2+ gradients and accumulation of signalling lipids, features all related to the regulation of apical growth (Yalovsky et al. 2008; Yang 2008). In rice, ROP GTPases have been reported to directly interact with and activate NADPH oxidase on the plasma membrane (Wong et al. 2007). In the scn1 mutant, ROS production from root hairs is enhanced and depolarized, suggesting that NADPH oxidase activity required for root hair formation is under the control of ROP GTPases (Carol et al. 2005). Moreover, ROS production was also shown to stimulate the expression of a ROP·GAP during plant response to oxygen deprivation, suggesting a negative feedback loop between ROP activity and ROS levels (Baxter-Burrell et al. 2002).

Previously, it was established that overexpression of constitutively active ROP11 (Atrop11CA) depolarizes root hair growth, leading to the formation of swollen root hairs. The swollen root hair phenotype was associated with altered actin organization and inhibition of endocytosis (Bloch et al. 2005). In this work, we demonstrate that depolarized growth of root hairs in Atrop11CA plants was also associated with abolishment of the tip-focused ROS gradient. Moreover, we show that the ability of Atrop11CA to induce root hair swelling strongly depends on growth medium composition, particularly the presence of ammonium ions at millimolar concentration. These observations lead to a model in which environmental factors function in parallel to ROP activity to maintain root hair tip growth.


Atrop11CA-induced root hair swelling depends on the presence of ammonium nitrate (NH4NO3) in the growth medium

Typically, Arabidopsis seedlings are grown on 0.5X Murashige and Skoog (MS) medium (Table 1). Transgenic plants overexpressing activated AtROP11 fused to a fluorescent marker (GFP-Atrop11CA/ mRFP-Atrop11CA) that were grown on MS medium developed swollen root hairs (Figs 1b,d & 2a,b). However, root hair growth is known to be highly plastic in response to a range of environmental stresses and to be highly responsive to nutrient levels. For example, phosphate starvation affects both root hair length and density (Lopez-Bucio, Cruz-Ramirez & Herrera-Estrella 2003; Perry, Linke & Schmidt 2007). Arabidopsis seedlings can germinate and grow for a short period on distilled water, allowing us to ask whether nutrient status could play a role in the phenotype elicited through activated AtROP11. Interestingly, we detected that when Atrop11CA seedlings were germinated on distilled water, their root hairs were not swollen and elongated in a similar fashion to wild-type plants (Fig. 1a). To identify which component/s in the MS medium were required for induction of the root hair swelling, Atrop11CA seedlings were grown on a series of partial MS media, each lacking one component (Table 1). This approach revealed that only in the media lacking NH4NO3 were Atrop11CA root hairs not swollen (Fig. 1c) (Table 1). When Atrop11CA seedlings were grown on complete MS medium, approximately 80–90% of root hairs were swollen, whereas on medium lacking NH4NO3, almost all the root hairs elongated normally and the proportion showing swelling was not significantly different from wild-type root hairs (P ≥ 0.87, Fig. 1d). Comparison of wild-type root hair lengths revealed that they were significantly shorter (P ≤ 0.001) when grown on complete MS in comparison to MS lacking NH4NO3 (Fig. 1b,c,e). These results indicated that NH4NO3 affects root hair growth and Atrop11CA-induced swelling. Importantly, GFP-Atrop11CA expression in root hairs and its plasma membrane localization, as revealed by plasmolysis experiments, were not affected by the depletion of NH4NO3 from the growth medium (Fig. 2a–f and supplemental data Fig. S1). On medium containing NH4NO3 and KNO3 alone GFP-rop11CA was localized on the membrane were an additional growth tip was formed (Fig. 2g–i). Because constitutively active ROP mutants are suspected to continuously activate signalling cascades that regulate tip growth, these results suggested that NH4NO3 affected signalling/growth regulation acting either parallel to or downstream of ROPs. To test whether Atrop11CA signalling was also affected by the depletion of NH4NO3, we next examined the organization of actin and distribution of ROS in root hairs, components likely to be acting downstream of Atrop11CA.

Table 1.  The effect of nutrient composition on the formation of swollen root hairs in Atrop11CA seedlings Thumbnail image of
Figure 1.

Depletion of ammonium nitrate suppresses the root hair swelling phenotype in Atrop11CA and affects root hair elongation in wild-type plants. (a–c) Root hair phenotype of wild type and Atrop11CA transgenic plants grown on medium supplemented with (a) ddH2O; (b) MS; or (c) MS lacking NH4NO3 (MS − NH4NO3). Scale bars correspond to 50 µm, (d) Percent of swollen root hairs on wild-type and Atrop11CA roots grown on MS and Atrop11CA grown on MS lacking NH4NO3 with or without addition of MES buffer. n = 12 roots; 400–500 root hairs, ***P ≤ 0.001 compared to wild type grown on 0.5X MS. P ≤ 0.85 for wild type on MS medium and Atrop11CA grown on MS − NH4NO3. P ≤ 0.26 for Atrop11CA grown on MS and MS − NH4NO3 with addition of MES. (e) Average length of wild-type root hairs grown on MS and MS − NH4NO3.***P ≤ 0.001. Error bars represent standard error; statistical analyses were carried out with the Student t-test.

Figure 2.

Ammonium depletion does not compromise membrane localization of GFP-Atrop11CA in root hairs. GFP-Atrop11CA localization in root hairs grown on standard MS medium (a, b) or MS medium lacking NH4NO3 (d, e). Plasmolysis revealed that on MS (c) and MS − NH4NO3 (f) medium the protein was localized to the plasma membrane, as can be seen by its detachment from the cell wall (arrows). (g–i) On NH4NO3 (10.3 mM) + KNO3 (9.4 mM) medium GFP-Atrop11CA was localized to the plasma membrane, however projection of stacks of optical sections through the whole root hair cell revealed that the protein was more concentrated on the membrane where an additional growth tip was formed (arrowhead). (a, d, g) Nomarsky DIC images; (b, c, e, f) single confocal scan images. (h, i) Maximum projection stacks images of multiple confocal scans; (i) pseudocolour of GFP fluorescence shown in (h). Scale bars correspond to 20 µm.

NH4NO3 depletion suppresses activation of downstream pathways by Atrop11CA

As was reported previously, GFP-Atrop11CA expression induced alterations in actin organization in swollen root hairs grown under standard growth conditions, as revealed by the YFP-Talin (YFP-Tn) marker (Fig. 3b) (Bloch et al. 2005). In such root hairs, actin cables were denser, frequently arranged at different angles and reached the extreme apex, which is normally occupied by diffuse fluorescence, usually referred to as a fine actin meshwork, in the wild type (Fig. 3a,b). However, when Atrop11CA seedlings were grown on media lacking NH4NO3, the organization of actin was similar to the wild-type root hairs (Fig. 3c). Under these growth conditions, a fine F-actin meshwork was detected at the tip of Atrop11CA root hairs, while the extensive actin cables typical for root hairs grown on complete MS media were no longer visible.

Figure 3.

NH4NO3 depletion affects organization of the actin cytoskeleton in GFP-Atrop11CA root hairs. (a) Wild-type root hair expressing YFP-Tn marker grown on MS medium. Arrowhead indicates F-actin mesh at the growing tip. (b–c) YFP-Tn/GFP-Atrop11CA double transgenic root hairs grown on MS medium (b) or MS − NH4NO3 (c). Arrowhead in (c) shows appearance of actin mesh at the tip. All images are maximum projection of multiple confocal scans. Scale bars correspond to 10 µM.

As an additional indication for downstream pathways activated by Atrop11CA expression, we looked at the localization of reactive oxygen species in growing root hairs. To this end, root hairs of wild-type and mRFP-Atrop11CA plants were loaded with 5-(and 6)-chloromethyl-2′, 7′- dichlorodihydro fluorescein (CM-H2DCF), a cell-permeable ROS-sensitive dye. Previous studies showed that ROS production by ATRBOHC was required for root hair tip growth, and that intracellular ROS concentration increased after the root hair underwent the transition to tip growth (Foreman et al. 2003). In agreement with these previous observations, in wild-type root hairs grown on MS medium, the ROS signal was localized at the bulge in the epidermal cell indicative of root hair initiation and then focused around the apex of the hair during tip growth (Fig. 4a; Supporting Information Video Clip S1 and S2). On the contrary, this ROS signal was dispersed throughout the root hair cell with no detectable gradient in mRFP-Atrop11CA root hairs grown under the same conditions (Fig. 4b; Supporting Information Video Clip S3). Depletion of NH4NO3 resulted in accumulation of ROS signal at the tip during the root hair elongation in the mRFP-Atrop11CA root hairs (Fig. 4c; Supporting Information Video Clip S4), similar to the distribution in wild type. Under the experimental conditions used, in complete MS medium, the rate of root hair elongation in wild type was approximately 1 µm per minute, whereas expression of Atrop11CA reduced the elongation rate to ∼0.2 µm per minute. Rescue of tip growth by removal of NH4NO3 restored root hair elongation rate of the Atrop11CA plants to approximately 1 µm per minute. NH4NO3 is the sole source of ammonium in the MS medium, whereas there are other sources of nitrate (KNO3). Therefore, we reasoned that the effect of NH4NO3 depletion was most likely through loss of a source of NH4+ in these experiments. Therefore, the effect of NH4+ concentration on Atrop11CA-induced swelling was examined.

Figure 4.

NH4NO3 depletion suppresses the effect of mRFP-rop11CA on ROS distribution in growing root hairs. (a) Distribution of ROS in wild-type root hair growing on MS medium. Note that during the growth, CM-H2DCF fluorescence accumulates first in the bulge and then remains concentrated at the apex during the root hair elongation. (b) mRFP-Atrop11CA root hair grown on MS medium. Note that no tip-focused gradient was detected. (c) Tip-focused gradient of ROS was detected in mRFP-Atrop11CA root hairs grown on MS medium lacking NH4NO3.The CM-H2DCF dye intensity shown as pseudocolour, from high (yellow) to low (blue). Scale bars correspond to 10 µm. See also Supporting Information Video Clips S1–S4.

Ammonium in the millimolar range is required for Atrop11CA- induced root hair swelling

Transport systems that mediate ion fluxes across the plasma membrane of root cells are divided into two categories: high affinity transport systems (HATS) that mediate uptake from relatively dilute solutions at relatively low rates and low affinity transport systems (LATS) that operate at high rates and higher external concentrations (Britto & Kronzucker 2006). For NH4+ ions, the HATS are plasma membrane localized NH4+-specific transporters (AMTs) that are most likely proton-coupled, and their expression and function are repressed at external ammonium concentrations of 1 mM or higher (Williams & Miller 2001; Loque & von Wiren 2004; Lanquar et al. 2009). In contrast, ammonium uptake by LATS is believed to take place through non-specific cation channels (Williams & Miller 2001; Ludewig, Neuhauser & Dynowski 2007). The NH4+ concentration in the 0.5X MS medium is 10.3 mM, exceeding by an order of magnitude the concentration at which the high affinity NH4+ uptake system is repressed. Thus, to distinguish between the effects of high- and low-affinity NH4+ uptake, Atrop11CA seedlings were grown on MS media containing decreasing concentrations of ammonium. To eliminate the possible effect of NO3- on root hair swelling, NH4NO3 was replaced with NH4Cl. It was found that significant reduction in the percentage of swollen root hairs was observed at concentration of 0.5 mM NH4Cl, whereas at 0.1 mM, almost all root hairs elongated normally (Fig. 5). These data showed that the high rate ammonium uptake by LATS and/or suppression of the HATS were required for induction of swelling in Atrop11CA plants. The next set of experiments was performed in order to verify whether nitrate also contributes to the swelling of Atrop11CA root hairs.

Figure 5.

Effect of NH4+ concentration on formation of Atrop11CA-induced root hair swelling. Percent of swollen root hairs on Atrop11CA roots grown on MS medium supplemented with decreasing concentrations of NH4Cl. The concentration of NH4Cl in the medium is indicated under each bar. Error bars represent standard error, n ≥ 12 seedlings. ***P ≤ 0.001 compared to 10.3 mM NH4Cl. Statistical analyses were performed with the Student t-test.

Both NH4+ and NO3- are required for Atrop11CA-induced root hair swelling

Plants can absorb and use various forms of nitrogen from soils, primarily the inorganic ions ammonium and nitrate (Britto & Kronzucker 2001). In the MS medium, ammonium ions are supplemented in the form of NH4NO3, whereas both NH4NO3 and KNO3 serve as sources for the nitrate. Most higher plants, including Arabidopsis, develop severe toxicity symptoms when exposed to millimolar concentration of ammonium in the growth medium, especially when it is the sole nitrogen source (Britto & Kronzucker 2001). In agreement, once seedlings were grown on medium containing 10.3 mM NH4NO3 alone, roots became necrotic and failed to develop in both wild-type and Atrop11CA seedlings (Fig. 6a). Interestingly, when seedlings were germinated on NH4NO3 supplemented with five essential micronutrients (CoCl2, CuSO4, FeSO4-EDTA, MnSO4 and ZnSO4), at the same concentrations as in the 0.5X MS medium, wild-type roots failed to develop, whereas root growth in Atrop11CA plants was rescued (Supporting Information Fig. S2). This result strongly suggested that Atrop11CA plants have an altered response to external concentration of ammonium not only in root hairs, but also in the whole root.

Figure 6.

The effect of NH4+ and NO3- on formation of swollen root hairs in Atrop11CA plants. (a) Inhibition of primary root elongation in wild-type and Atrop11CA transgenic plants grown on plates supplemented with NH4NO3 alone. (b–d) Root hair phenotype in wild-type and Atrop11CA plants grown on combination of indicated nutrients: (b) NH4NO3 + KNO3; (c) NH4Cl + KCl; (d) NH4Cl + KNO3. Scale bars are 50 µm. Arrowheads point to multiple root hair growth tips.

It has been demonstrated that potassium can suppress NH4+ toxicity and enable root growth in wild-type Arabidopsis (Cao, Glass & Crawford 1993). Recently, it was also shown that in barley, rapid futile cycling of NH4+ across the plasma membrane, which is responsible for the toxicity symptoms, is alleviated by elevated K+ supply (Szczerba et al. 2008). Thus, to overcome the NH4+ inhibition of root development, wild-type seedlings were grown on medium containing NH4NO3 (10.3 mM) + KNO3 (9.4 mM). The addition of KNO3 suppressed the inhibition of root growth, enabling roots to develop to maturation. Roots of seedlings grown on NH4NO3 + NaNO3 did not develop to maturation, similar to roots of seedlings that were grown on NH4NO3 alone (data not shown), suggesting that, indeed, potassium suppressed NH4+ toxicity and enabled root growth. Wild-type plants grown on NH4NO3 + KNO3 also developed normal looking root hairs (Fig. 6b). In contrast, under these growth conditions, the root hairs of Atrop11CA plants were swollen either becoming balloon-like or adopting different shapes due to the establishment of two or more growing tips (Fig. 6b, arrowheads).

To examine whether NO3- in addition to NH4+ affected root hair swelling, wild-type and GFP-Atrop11CA seedlings were germinated on agar plates supplemented with either 10.3 mM NH4Cl + 9.4 mM KCl or 10.3 mM NH4Cl + 9.4 mM KNO3. Wild-type seedlings developed normal-looking root hairs in both media (Fig. 6c,d). GFP-Atrop11CA seedlings grown on NH4Cl + KNO3 developed swollen root hairs, as well as some split root hairs with more than one growing tip (Fig. 6d, arrowheads). However, wild-type looking root hairs developed in Atrop11CA seedlings grown on NH4Cl + KCl (Fig. 6c), indicating that, in addition to ammonium, nitrate was also required for Atrop11CA-induced root hair swelling. Swollen root hairs developed when GFP-rop11CA seedlings were grown on MS media with NO3 concentrations of 1 mM or higher. However, reducing NO3 concentrations below 1 mM induced toxicity, resulting in complete inhibition of root hair development (data not shown).

NH4+/NO3- enhance the amplitude of pH oscillations at the apex of wild type root hairs

As was mentioned earlier, tip growth of Arabidopsis root hairs is associated with cyclic changes of pH at the apex of the cells that correlate with the periodicity of growth. These pH changes are thought to reflect cyclical stabilization of the cell wall by high pH and relaxation of the wall at low pH. The combination of these two pH-dependent phases has been proposed to be needed to support tip growth (Monshausen et al. 2007). The ammonium ion acts as a weak acid when assimilated (NH4+inline image NH3 + H+), and its uptake by the LATS is accompanied by concurrent active efflux of NH4+ to the rooting medium by a cation/H+ antiport mechanism (Britto & Kronzucker 2006). In yeast, NH4+ was predicted to function as a protonophore, facilitating equilibration of the pH across the plasma and vacuolar membranes (Plant et al. 1999). We thus predicted that high ammonium concentrations in the growth medium may affect the tight balance of pH oscillations at the apex of root hairs, and so provide a possible mechanism through which NH4+ could interact with the ROP11-dependent growth machinery to alter tip growth.

2-(N-morpholino)ethanesulfonic acid (MES) serves as an effective buffer to stabilize the pH of solutions containing NH4 and NO3 during plant germination (Ewing & Robson 1991). When GFP-rop11CA seedlings were grown on MS media that contained MES, only 20% of the root hairs were swollen compared to 80–90% in seedlings that were grown on MS medium without MES. Similarly, around 20% swollen root hairs developed on GFP-rop11CA seedlings grown on MS medium that contained MES but lacked NH4NO3 (Fig. 1d). Collectively, these data suggested that GFP-rop11CA-induced root hair swelling may be associated with NH4O3-induced pH fluctuations.

Cytosolic and apoplastic pH oscillations were measured in wild-type root hairs prior to and after application of 10 mM NH4NO3 to the growth medium in order to determine how these ions affected growth-related pH fluctuations or proton fluxes in the hair. To visualize oscillations of intracellular and extracellular pH simultaneously, root hairs expressing a cytosolic pH-sensitive GFP variant were immersed in pH-sensitive fluorescein dextran, according to the previously described procedure (Monshausen et al. 2007). Addition of NH4NO3 to the medium significantly increased the amplitude of cytoplasmic pH oscillations and induced an overall decrease in intracellular pH (Fig. 7 and Supporting Information Table S1). Before adding NH4NO3, cytoplasmic pH oscillations ranged from about 7.4 to 7.7. Following application of NH4NO3 the pH oscillated from about 7.0 to 7.6 (Fig. 7). Calculation of the average relative amplitude range of cytoplasmic pH showed that before application of NH4NO3, oscillations had an average amplitude of 0.1512 ± 0.0102, whereas within minutes after application of NH4NO3, the oscillation amplitude increased to 0.2604 ± 0.0294 (P < 0.005, n = 10). The frequencies of oscillations remained relatively constant before or after adding NH4NO3 (Fig. 7 and Supporting Information Table S1 and Video Clip S5). Extracellular pH oscillations also showed an increase in the amplitudes after application of 10 mM NH4NO3. Prior to application of NH4NO3, the extracellular pH ranged from about 5.6 to 6.1. After adding NH4NO3, the pH ranged from about 5.6 to 6.4 (Fig. 7). Calculation of the average relative change of the amplitude range showed that before application of NH4NO3, the range was 0.5462 ± 0.0219 and following addition of NH4NO3, 0.6314 ± 0.0516 (P = 0.0538, n = 7) (Fig. 7; Supporting Information Table S1 and Video Clip S5). Addition of ammonium or nitrate alone in the form of NH4Cl or NaNO3/KNO3 had no significant effect on the amplitude of pH oscillations (data not shown). Collectively, these results show that consistent with a possible role in modulating the tip growth machinery, NH4+ and NO3- affected both tip growth-related pH oscillations and the overall cytoplasmic pH at the extreme apex of growing root hairs.

Figure 7.

Application NH4NO3 to the growth medium affects the amplitude of intracellular and extracellular pH oscillations at the apex of wild-type root hairs. pH oscillations were monitored in root hairs expressing a soluble pH-sensitive GFP and immersed in the extracellular pH sensor fluorescein-dextran for simultaneous measurement of cytosolic (pHcyt) and surface (pHex) pH prior to and after addition of 10 mM NH4NO3. pH values were quantified from the ratio of fluorescence emission upon excitation by 458 nm and 488 nm light. Measurements from one representative root hair of 10 independent experiments are shown.


It is well known that root hair development is highly plastic and regulated by environmental signals. Length, frequency and position of the root hairs are affected by the availability of soluble nutrients such as phosphate (P), iron (Fe) and manganese (Mn) [reviewed in Lopez-Bucio et al. (2003) and Perry et al. (2007)]. Numerous Arabidopsis mutants have been identified in genes involved in the regulation of the cell polarization machinery and are affected in either initiation of root hair growth or in tip growth itself (Carol & Dolan 2002; Kost 2008). Despite the known function of ROP GTPases and their regulatory proteins in root hair development, there is no data in the literature describing the relationship between ROP signalling and environmental factors in this process. Here, we demonstrated that the well-known induction of root hair swelling by constitutively active ROP11 mutants occurs only under specific growth conditions, indicating the there is an interplay between ROP activity and the external environment, particularly nitrogen supply. Our results indicate that high external concentrations of ammonium are essential for the induction of depolarized root hair growth and activation of downstream pathways by constitutively active AtROP11. Depletion of ammonium did not affect the membrane localization and expression of GFP-Atrop11CA, implying that NH4+ was required in addition to the ROP activity to cause root hair swelling. In agreement with this idea, normal actin organization and ROS localization were detected in Atrop11CA root hairs when NH4+ was depleted, suggesting that ammonium functions downstream of, or in parallel to, ROP signalling.

Plants can absorb and use various forms of nitrogen from soils, primarily the inorganic ions ammonium and nitrate. The concentrations of these ions vary across several orders of magnitude among different soils and as a result of seasonal changes (Glass et al. 2002). Thus, in order to maintain normal N nutrition, plants need to regulate their fluxes accurately. In several plant species, a high and/or exclusive supply of ammonium induces toxicity symptoms (Britto & Kronzucker 2001). In the toxic range, NH4+ uptake is mediated by a low-affinity transport system. Thus, the rescue of root growth in Atrop11CA plants we observed when grown on 10.3 mM NH4NO3 supplemented with five micronutrients (CoCl2, CuSO4, FeSO4-EDTA, MnSO4 and ZnSO4) suggests that the activated ROP protein altered NH4+ fluxes across the plasma membrane. Interestingly, the rescue of root growth at high ammonium concentrations was previously reported in the auxin influx (aux1) and response (axr1 and axr2) mutants (Cao et al. 1993). Auxin has been implicated in ROP activation, and ROPs were shown to be involved in the regulation of auxin-induced gene expression (Tao, Cheung & Wu 2002; Tao et al. 2005). Root development is regulated by directional auxin transport that depends on constitutive endocytic recycling of PIN family auxin efflux transporters (Kleine-Vehn et al. 2008). Endocytic vesicle recycling at the plasma membrane is known to be inhibited by Atrop11CA (Bloch et al. 2005), possibly affecting polar auxin transport. Moreover, overexpression of an activated AtROP2 mutant altered the auxin response in the root (Li et al. 2001). Thus, the observed rescue of root growth in Atrop11CA plants could have been a consequence of altered auxin response or transport in this mutant.

High rates of LATS-mediated NH4+ influx are concurrent with a nearly equivalent rate of efflux (Britto & Kronzucker 2001, 2006). Current models suggest that cation influx in the LATS range occurs passively through channels following the electrochemical gradient across the plasma membrane. Cation efflux, on the other hand, depends on a proton electrochemical potential across the plasma membrane by engaging a cation/H+ exchange antiport mechanisms (Britto & Kronzucker 2006). The plasma membrane H+-ATPase is centrally implicated in re-establishment of membrane potential deviations caused by all electrogenic fluxes, and in re-establishing the proton gradient across the plasma membrane (Palmgren 2001). The root hair swelling in Atrop11CA plants occurred primarily at external ammonium concentrations greater than 1 mM, and thus was associated with an uptake by the LATS. Simultaneous ratio fluorescence imaging of internal and external pH revealed that application of 10 mM NH4NO3 enhanced the amplitude of pH oscillations at the extreme apex of wild-type root hairs. These oscillations are thought to modulate tip growth through altering the extensibility of the wall (Monshausen et al. 2008). Thus, one possible explanation for the observed swelling of the root hair apex in Atrop11CA-expressing plants in media containing NH4NO3 is that Atrop11CA root hairs are affected in their ability to re-establish the normal proton gradient across the plasma membrane in response to ammonium transport. The altered proton gradient would then prevent the normal localized oscillatory changes in pH-dependent wall properties required to localize expansion to the very tip of the elongating root hair.

Concurrent absorption of NH4+ and NO3- maintains the cation-anion balance within both the rooting medium and the root, and thus potentially has an important function in maintaining intracellular and extracellular pH (Stitt 1999; Britto & Kronzucker 2002). Simultaneous application of ammonium and nitrate were required for induction of swelling. Moreover, concurrent application of these ions affected the amplitude of pH oscillations, whereas addition of NH4+ or NO3- alone had no significant effect on oscillations. In agreement with this idea, addition of MES buffer suppressed root hair swelling. In rice, NH4+ fluxes across the plasma membrane and its accumulation in the cytosol were enhanced by the presence of NO3- (Kronzucker et al. 1999). These data strongly suggest that NH4+-dependent root hair swelling in the activated ROP plants resulted from physiological changes in ion balance rather than a direct effect of this ammonium on enzymatic activities required for root hair growth. In addition to root hair swelling, simultaneous application of NH4+ and NO3-, in the absence of other ions, induced formation of additional growth tips, in which the membrane localized GFP-Atrop11CA was concentrated. This observation suggests that interplay between the regulation of ROP localization and activity and the regulation of nitrogen fluxes has an important function in maintenance of unidirectional growth during the root hair elongation.

It has been previously reported that root hair elongation is coupled to spatially distinct regulation of extracellular pH oscillations and ATRBOHC-mediated ROS production, suggesting that feedback between these two systems is required to support normal tip growth (Monshausen et al. 2007). Because high concentrations of NH4+ and NO3- strongly affected intracellular and, to a lesser extent, the extracellular pH oscillations at the root hair apex, it seems likely that there is a mechanism that can adjust the fluxes of nitrogen ions relative to these proton fluxes in order to maintain the oscillations in pH such that polarized growth would be continued. One possible mechanism for this coordination is through the highly localized ROP cycling between active and inactive states that has an important role in the spatial activation of cell polarization machinery (Carol et al. 2005; Hwang et al. 2005, 2008; Klahre et al. 2006; Klahre & Kost 2006). Due to the function of ROP GTPases in vesicle trafficking, actin organization and maintenance of ROS and Ca2+ gradients (Li et al. 1999; Fu et al. 2001; Molendijk et al. 2001; Baxter-Burrell et al. 2002; Fu, Li & Yang 2002; Jones et al. 2002, 2007; Carol et al. 2005; Gu et al. 2005; Hwang et al. 2005; Wong et al. 2007), expression of activated AtROP11 may indirectly influence cell wall properties by altering the localization and/or recycling of cation and anion transporters/channels or plasma membrane H+-ATPases and, in this way, to affect the maintenance of the proton gradients. Our results indicate that ROP activity and environmental factors that affect pH oscillations function in parallel pathways during the tip growth of root hairs. Thus, we propose a model in which spatial regulation of ROP activity creates a positive feedback loop with pH oscillations around the growing apex of root hairs. According to this model, ROP cycling between active and inactive states spatially and temporally activates the downstream signalling cascades essential for the tip growth of root hairs, as previous results suggested. At the same time, localization of membrane proteins involved in maintenance of normal nitrogen fluxes across the plasma membrane is indirectly affected by ROP signalling. NH4+/NO3- fluxes increase the amplitude of pH oscillations at the root hair apex. These oscillations, in turn, may affect cell-wall properties. Thus, when the ROP activity is up-regulated by dominant mutations, the synergistic effects of pH changes and constant activation of ROP downstream effectors result in the uncontrolled cell expansion. Previous studies suggested that feedback between oscillatory pH change and ROS distribution is required to support tip growth (Monshausen et al. 2007). However, the factors that may coordinate these processes are unknown. Our results suggest that spatial regulation of ROP activity in response to changing environments is one of the key elements that may coordinate the pH and ROS oscillations during the root hair tip growth.

Collectively, these findings shed light on the interaction between ROP signalling and environmental factors, such as nutrient composition during the root hair tip growth.


Plasmid construction and plant transformation

The pDsRed-Express-N1 (Clontech, Palo Alto, CA, USA) plasmid was used as a template for PCR amplification of the monomeric red fluorescence protein (mRFP) sequence with SYP546-AAACCATGGCCGCCTCCTCCGAGGAC and SYP545-AAAGAGCTCGGCGCCGGTGGAGTGG. The PCR product was subcloned into pGEM-T vector (Promega, Madison, WI, USA) to create pSY525 and sequenced. pSY525 was digested with NcoI and SacI. The resulting fragment containing mRFP was subcloned into NcoI, SacI digested pMRC-GFP vector (Rodríguez-Concepción et al.) replacing GFP to obtain pSY526 (pmRFP). To create mRFP-Atrop11CA fusion, the pSY507 plasmid was digested with SacI and Atrop11CA CDS was subcloned into the pSY526 digested with SacI, to obtain pSY530. For expression in plants, pSY530 was digested with HindIII to isolate a cassette comprised of the 35S promoter of CaMV, mRFP-Atrop11CA and NOS transcriptional terminator. This cassette was subcloned into pCAMBIA3300 (CAMBIA) to obtain pSY528.

Wild-type Col-0 Arabidopsis plants were transformed using the floral dip method (Clough & Bent 1998) with Agrobacterium tumerfaciens strain GV3101/pMp90.

Plant material and growth conditions

Seeds of Arabidopsis wild type (Col-0) and transgenic lines 35S::GFP-Atrop11CA, 35S::YFP-Tn, 35S::GFP-Atrop11CA35S::YFP-Tn (Bloch et al. 2005) and 35S::mRFP-Atrop11CA were surface sterilized and sown on plates supplemented with the indicated nutrients, 1% (w/v) sucrose and 0.8% (w/v) plant agar and titrated to pH 5.5 with KOH/NaOH. Where indicated, MES was added to a final concentration of 2.5 mM, and then the medium was titrated to pH 5.5 with KOH/NaOH. After 48 h of vernalization at 4 °C, the plates were transferred to a growth chamber and grown vertically under long-day conditions (16 h light/8 h dark cycle) at 23 °C for 7 d. The light intensity was 100 µE m−2 s−1.

Analysis of nutrient composition-dependent root hair development

Seeds were germinated on deionized water (ddH2O) non-supplemented agar plates or supplemented with either the full mixture of 0.5X MS medium or partial mixture of nutrients (Table 1). In partial nutrient mixtures, the concentration of each component that was added was identical to its concentration in the 0.5X MS. A total of ≥12 seeds were germinated on each plate and at least three replicates were scored. Results were scored after 7 d in the growth chamber. For calculation of percentage of Atrop11CA swollen root hairs, a constant 1-mm-long section of the root at a distance of ∼1.5 mm proximal to the root tip was analysed. The root hair phenotype was scored using a microscope with dry 10X objective. Typically, ∼50 root hairs were analysed in each root. The degree of significance was calculated using the Student t-test.


To induce plasmolysis in root hairs, roots were submerged in 400 mM mannitol solution for several minutes and then mounted on microscope slides in the same solution.

Detection of ROS in Arabidopsis growing root hairs

The ROS sensitive dye CM-H2DCFDA [5-(and 6)-chloromethyl-2′,7′- dichlorodihydro fluorescein] (Molecular Probes, Carlsbad, CA, USA) was dissolved in DMSO 99.9%, anhydrous (Sigma, St. Louis, MO, USA) to obtain 10 mM stock solution (stored under N2, in aliquots, at −20 °C). CM-H2DCFDA was diluted to a final concentration of 5 µM. A small piece of agar containing the seedling (∼1 × 1 cm) was dissected from the growth plate and transferred to a microscope glass slide. The dye solution was injected into the agar near the root and gently covered with a cover slip. Roots were immediately examined with a Leica Confocal Laser Scanning Microscope (CLSM) microscope using a 20X multi-immersion objective (N.A. = 0.7), as described in further discussion. Time-lapse experiments were performed on emerging root hairs in the root differentiation zone. To follow the root hair growth, images were taken at 30 s intervals.

Light and confocal microscopy

Brightfield imaging was performed with an Axioplan-2 Imaging microscope (Carl Zeiss, Jena, Germany) equipped with an Axio-Cam, cooled charge-coupled device camera by using either 10X dry or 20X dry objectives. CLSM imaging was performed using Leica TCS-SL confocal laser scanning microscope with 20X (N.A. = 0.7) multi-immersion or 63X (N.A. = 1.2) water objectives. GFP and CM-H2DCFDA were visualized by excitation with an argon laser at 488 nm. Emission was detected with a spectral detector set between 505 and 530 nm. YFP was visualized by excitation with an argon laser at 514 nm. Emission was detected with spectral detector set between 525 and 570 nm.

For actin imaging, to obtain maximal signal-to-noise ratios and minimize bleaching, the argon laser was set to 30% and the Acusto Optical Tunable Filter (AOTF) in all scans did not exceed 16%. All scans were carried out at 512 × 512 pixel resolution with repeated scanning of two lines and two frames. Typically, 70–100 (pending on cell size) 0.4-µm-thick sections were taken. Images were produced by maximum intensity projections of multiple confocal scans.

For ROS imaging, to minimize photo-oxidation, the argon laser was set to 30%, and the AOTF in all scans did not exceed 5%. Scans were carried out in an xyt (x, y, time) mode. All scans were carried at 1024 × 1024 pixel resolution with 4× line averaging. The signal intensity was pseudocoloured using the RGB Rainbow application of Image J.

pH measurements

pH measurements were carried out as previously described (Monshausen et al. 2007). Briefly, for cytosolic pH measurements Arabidopsis plants expressing the H148D pH sensitive GFP, driven by the 35S promoter, were visualized using a Zeiss LSM 510 confocal microscope. Ratio images were taken using a 40× water-immersion, 1.2 numerical aperture, C-Apochromat objective. Excitation used the argon ion laser alternated between 458 and 488 nm (switching between wavelengths after each scanned line at <<1 s), with emission collected using a 488 nm dichroic mirror and 505 nm long-pass filter. Extracellular pH was imaged simultaneously by supplementing the growth medium with 150 µg ml−1 of the pH-sensitive fluorescent dye fluorescein, conjugated to 10 kDa dextran (Sigma). pH values were calibrated as previously described (Monshausen et al. 2007). For time-lapse analysis, images were collected every 3 s, with each individual image scan lasting approx 2 s. Data were analysed by using the Zeiss LSM software and Microsoft Excel.


We thank The Manna Institute at the department of Plant Sciences, Tel Aviv University, for technical support. This work was supported by Israel Science Foundation (ISF-312/07), US–Israel Binational Research and Development fund (BARD-IS-4032-07) and the Deutschland–Israel Program (DIP-H.3.1) to SY and NSF (MCB 0641288) to GBM and SG.