A Receptor-Like Kinase Mediates Ammonium Homeostasis and Is Important for the Polar Growth of Root Hairs in Arabidopsis

Ammonium (NH₄⁺) is an important nutrient and signal, but can be toxic in excess. Thus, cytoplasmic NH₄⁺ levels must be intricately sensed and tightly controlled. This work identifies a modulator of NH₄⁺ in the tonoplast, CAP1, a receptor-like kinase that mediates NH₄⁺ homeostasis and regulates the polar growth of root hairs by maintaining tip-focused cytoplasmic Ca²⁺ gradients.

Abstract

Ammonium (NH₄⁺) is both a necessary nutrient and an important signal in plants, but can be toxic in excess. Ammonium sensing and regulatory mechanisms in plant cells have not been fully elucidated. To decipher the complex network of NH₄⁺ signaling, we analyzed [Ca²⁺]_cyt-associated protein kinase (CAP) genes, which encode signaling components that undergo marked changes in transcription levels in response to various stressors. We demonstrated that CAP1, a tonoplast-localized receptor-like kinase, regulates root hair tip growth by maintaining cytoplasmic Ca²⁺ gradients. A CAP1 knockout mutant (cap1-1) produced elevated levels of cytoplasmic NH₄⁺. Furthermore, root hair growth of cap1-1 was inhibited on Murashige and Skoog medium, but NH₄⁺ depletion reestablished the Ca²⁺ gradient necessary for normal growth. The lower net NH₄⁺ influx across the vacuolar membrane and relatively alkaline cytosolic pH of cap1-1 root hairs implied that mutation of CAP1 increased NH₄⁺ accumulation in the cytoplasm. Furthermore, CAP1 functionally complemented the npr1 (nitrogen permease reactivator protein) kinase yeast mutant, which is defective in high-affinity NH₄⁺ uptake via MEP2 (methylammonium permease 2), distinguishing CAP1 as a cytosolic modulator of NH₄⁺ levels that participates in NH₄⁺ homeostasis-regulated root hair growth by modulating tip-focused cytoplasmic Ca²⁺ gradients.

INTRODUCTION

To adapt to their limited mobility, plants have evolved complex sensory mechanisms that regulate internal growth, metabolism, and gene expression based on fluctuations in external nutrient concentrations (Forde and Lorenzo, 2001; López-Bucio et al., 2003). For example, root hairs possess intricate regulatory networks that govern cell elongation and employ sophisticated sensory mechanisms that respond to different environment signals. Therefore, root hair cells are an excellent model for investigating the convergence between root hair growth and development and signal sensing (Libault et al., 2010).

The plastic development of root hairs is environmentally regulated, leading to variations in length and density (López-Bucio et al., 2003). Many of the genes encoding transcription factors, protein kinases, and small G proteins involved in root hair initiation and development have been identified (Dolan et al., 1993; Bibikova and Gilroy, 2003; Grierson and Schiefelbein, 2009; Benfey et al., 2010). For example, the transcription factors WEREWOLF (Lee and Schiefelbein, 1999), MYB23 (Kang et al., 2009), and GLABRA3 (Bernhardt et al., 2003) and a Leu-rich repeat receptor-like kinase (RLK) SCRAMBLED (Kwak et al., 2005) function in the differentiation of the root hair cell, where they form a regulatory network that determines the fate of the root epidermal cells (Libault et al., 2010). ROOT HAIRLESS1 is required to initiate root hair development (Schneider et al. 1998). The NADPH oxidase ROOT HAIR DEFECTIVE2 (RHD2/AtrbohC) and the AGC kinase OXIDATIVE SIGNAL-INDUCIBLE1 are essential for establishing the normal morphology of root hairs (Foreman et al., 2003; Rentel et al., 2004).

As in typical polar growth, tip-focused cytoplasmic Ca²⁺ gradients, vesicle/membrane trafficking, cytoskeletal reorganization, and increased oscillations in surface pH and reactive oxygen species (ROS) are involved in root hair elongation (Carol and Dolan, 2002; Monshausen et al., 2007, 2008). The Ca²⁺ gradient localizes newly formed plasma membrane and cell wall to the hair apex, driving cell growth (Carol and Dolan, 2002). Establishing this gradient requires ROS, which are generated by the Respiratory Burst Oxidase Homolog C NADPH oxidase (RHD2/AtrbohC) in Arabidopsis thaliana (Foreman et al., 2003). Other ROS generators may assume RHD2's role later in root hair elongation (Monshausen et al., 2007). The Arabidopsis loss-of-function mutant for annexin1 (ann1) was found to lack root hairs and epidermal OH˙-activated Ca²⁺- and K⁺-permeable conductance (Laohavisit et al., 2012).

In plants, many signals are mediated by RLKs. Three of the 17 members of the RLK subfamily CrRLK1L (Catharanthus roseus receptor-like-kinase-1-like) mediate cell elongation; knockout or reduction of their expression reduced cell elongation (Guo et al., 2009). Rho-related GTPases of plants (ROPs) also regulate root hair initiation and tip growth by stimulating ROS accumulation (Jones et al., 2002, 2007). Ectopic expression of constitutively active Arabidopsis ROP11 (rop11^CA) depolarizes root hair growth, leading to the formation of swollen root hairs (Bloch et al., 2005). High external levels of ammonium (NH₄⁺) are essential for the induction of depolarized root hair growth and the activation of downstream pathways by rop11^CA, suggesting that the coregulation of root hair tip growth by ROP GTPase and by the nitrogen source modulated pH fluctuations (Bloch et al., 2011). A range of transporters in root hairs transport nitrogen, calcium (Ca), potassium (K), and sulfate (Libault et al., 2010).

Nitrogen is an essential macronutrient for plant growth and development, and nitrate (NO₃⁻) and NH₄⁺ are the dominant forms of inorganic nitrogen in soil. Ammonium is taken up when scarce, but toxic in excess (Loqué and von Wirén, 2004). Therefore, cytoplasmic levels must be tightly controlled through regulated compartmentalization, assimilation, and precise and concerted absorption and excretion using ammonium transporters (AMTs). In the toxic range, NH₄⁺ uptake is mediated by a low-affinity transport system. Thus, the rescue of root hair growth in Atrop11^CA plants by higher NH₄NO₃ levels occurs because the activation of ROP proteins alters NH₄⁺ fluxes across the plasma membrane (Bloch et al., 2011). Ammonium uptake was linked to Ca²⁺ release in canola (Brassica napus) root cells (Babourina et al., 2007).

The Arabidopsis genome encodes six AMTs: AMT1.1 to AMT1.5 and AMT2.1, four of which (AMT1.1, AMT1.2, AMT1.3, and AMT1.5) are involved in NH₄⁺ uptake (Yuan et al., 2007). Transcriptional and posttranscriptional regulation of AMTs prevents excess NH₄⁺ accumulation and toxicity (Gazzarrini et al., 1999; Loqué et al., 2006, 2009). AMT1.1 expression is activated by nitrogen deficiency, and phosphorylation of a specific threonine residue exhibits an allosteric effect on AMT and prevents excess NH₄⁺ accumulation. This feedback mechanism modulates NH₄⁺ uptake in roots over a wide range of nutrient levels (Lanquar et al., 2009), suggesting that roots use either the AMT itself or another extracellular sensor to measure NH₄⁺ concentrations in the rhizosphere.

Unlike the NH₄⁺ uptake system, the mechanisms by which cytoplasmic NH₄⁺ concentrations ([NH₄⁺]_cyt) are regulated to prevent excess accumulation are poorly documented. Indeed, reported [NH₄⁺]_cyt obtained by physiological experiments are controversial. Millimolar values of [NH₄⁺]_cyt have been measured using NMR, efflux analysis, ion-specific microelectrodes, and tissue fractionation (Fentem et al., 1983; Lee and Ratcliffe, 1991; Wang et al., 1993; Wells and Miller, 2000). However, because of the potential toxicity of NH₄⁺, [NH₄⁺]_cyt is assumed to be maintained at very low (submillimolar) levels via the high activity and high affinity for NH₄⁺ of Gln synthetase (Pearson and Stewart, 1993; Gerendás et al., 1997). Ammonium uptake alkalinizes the cytoplasmic medium, although its assimilation per se is proton neutral (Britto and Kronzucker, 2005). The observations that millimolar NH₄⁺ concentrations inside the vacuole largely exceeded those of the cytoplasm in Chara corallina cells (Wells and Miller, 2000) and that rape plants showed a large increase in NH₄⁺ concentrations in xylem sap when supplied NH₄⁺ as a major nitrogen source (Finnemann and Schjoerring, 1999) suggest that NH₄⁺ is also exported from the cytoplasm into the apoplasm or the vacuole. In fact, two Arabidopsis genes, TIP2;1 and TIP2;3, which encode aquaporins of the tonoplast intrinsic protein (TIP) subfamily, conferred tolerance to the toxic NH₄⁺ analog methylammonium in yeast and facilitated NH₃ transport into the vacuole (Loqué et al., 2005). However, much remains to be discovered about how cytoplasmic NH₄⁺ levels are sensed and regulated in plant cells.

Using the Arabidopsis root hair system, we show that a [Ca²⁺]_cyt-associated protein kinase (CAP1) in the CrRLK1L family, functioning as a component of NH₄⁺ signaling, can detect internal NH₄⁺ status and downregulate the activity of the tonoplast transporter to avoid toxic accumulation of NH₄⁺ by sequestering it into the vacuole. CAP1 regulates root hair tip growth by maintaining cytoplasmic Ca²⁺ gradients in the root hair cells. Knockouts of CAP1 did not produce normal root hairs on Murashige and Skoog (MS) medium unless NH₄⁺ was depleted. The depletion of environmental NH₄⁺ could reestablish the normal cytoplasmic Ca²⁺ gradient, suggesting that NH₄⁺ functions in the polar growth of root hairs. This study identified a novel regulator in the tonoplast for NH₄⁺ homeostasis that is distinct from the plasma membrane transceptor for NH₄⁺ acquisition (Lanquar et al., 2009).

RESULTS

CAP1 Deficiency Impairs Root Hair Growth

One of the earliest events in a plant's response to hormones, development, or environmental stresses is a rapid, transient elevation of the free cytosolic concentration ([Ca²⁺]_cyt) of the intracellular second messenger calcium (Hetherington and Brownlee, 2004; DeFalco et al., 2010). Many protein kinases are involved in regulating [Ca²⁺]_cyt in early signaling. We performed a survey for protein kinase gene expression from the Microarray data using the AtGenExpress Visualization Tool (http://jsp.weigelworld.org/expviz/expviz.jsp) (Kilian et al., 2007). Fifty protein kinase genes that were significantly induced (more than 5-fold) by different stressors (e.g., cold, drought, and salt) were randomly selected. T-DNA insertion mutants of these kinase genes were obtained from the ABRC (Alonso et al., 2003). To identify their role in sensing and regulating responses, we generated transgenic Arabidopsis plants that expressed cytosolic aequorin in the T-DNA insertion lines (Knight et al., 1991). Protein kinase mutants containing calcium indicators combined with Ca²⁺ imaging techniques represent a powerful tool to monitor spatial and temporal [Ca²⁺]_cyt to investigate the molecular basis of cell sensing. Mutants with abnormal resting [Ca²⁺]_cyt and/or stimulated increase in [Ca²⁺]_cyt in responses to a specific stressor were selected for further analysis (McAinsh et al., 1992). One of the T-DNA insertion mutants, designated cap1-1 ([Ca²⁺]_cyt-associated protein kinase) (Supplemental Figure 1A), was chosen for detailed characterization because of its low level of resting [Ca²⁺]_cyt as indicated by luminescence levels in seedlings (Supplemental Figure 1C).

CAP1, a member of the CrRLK1L family, has no intron. The homozygous Salk_083442 T-DNA insertion line was first tested by PCR. T-DNA was inserted 1047 bp from the start codon ATG (Supplemental Figure 1A). In an RT-PCR analysis using total RNA extracted from roots, the transcript size (∼650 bp) detected on a 1% agarose gel was almost identical for cap1-1 and the wild type (Supplemental Figure 1B, left panel). However, sequencing of the RT-PCR products showed that 28 bp (from 1422 to 1449) in the mutant cDNA were deleted (Supplemental Figure 1D), resulting in a defective CAP1. To discriminate between the different transcripts produced in the wild type and cap1-1, smaller fragments (∼250 bp) across this deletion region were amplified and separated on 4% agarose gel. Wild-type transcripts displayed relatively slower mobility (Supplemental Figure 1B, right panel).

Interestingly, we found that root hairs of cap1 mutants do not elongate (Figure 1). Root hair growth can be subdivided into three major stages: root hair initiation, transition (slow growth phase), and tip growth (fast growth phase) (Dolan et al., 1994). In each stage, significant differences in root hair length and morphology were found between the wild type and mutant (Figure 1). Compared with the wild type, 7-d-old seedlings of cap1-1 exhibited 6-fold shorter root hairs (cap1-1, 83 ± 8 μm; wild type, 566 ± 34 μm; P < 0.01, Dunnett's test; Figure 1B), slightly fewer root hairs (cap1-1, 64 ± 1.6; wild type, 80 ± 1.4; P < 0.01; Figure 1B), and swollen bases during root hair initiation and abnormal shapes during the fast growth phase (Figure 1C). The mutation of CAP1 seemed to specifically affect tip elongation and root hair initiation but not cell viability. No other significant differences were observed in adult plants. These results suggested that CAP1 functions in the tip growth of root hairs.

Figure 1.

Mutation of CAP1 Impairs Root Hair Growth in Arabidopsis.

(A) Root hair growth of the wild type (WT), mutant (cap1-1), and complementation lines (#1 and #2). Seedlings were grown on vertical 1.2% agar MS medium for 7 d.

(B) Comparison of root hair lengths and numbers between the wild type, cap1-1, and complementation lines. Root hair lengths were measured in 7-d-old seedlings 5 mm from the primary root tips. Data bars represent means ± se of root hair lengths from triplicate experiments (wild type and cap1-1, n = 100; complementation lines, n = 60).

(C) Scanning electron micrographs of wild-type and cap1-1 root hairs. Bar = 200 µm.

To confirm this conclusion, wild-type CAP1 was introduced into the mutant under control of the 35S promoter. Transgenic seedlings (#1 and #2 in Figure 1A) displayed complementary root hair numbers, lengths, and morphology to the wild type. RT-PCR of these transgenic lines (Supplemental Figure 1B) showed that these seedlings harbored the wild-type coding sequence (Supplemental Figure 1E). The root hair phenotype was also complemented by the expression of wild-type CAP1 driven by its own promoter (Com-1 and Com-2; Supplemental Figures 2A and 2B). To further ascertain the function of CAP1, we established a set of CAP1-suppression transgenic lines using RNA interference (RNAi). The independent transgenic plants (CAP1-RNAi1, 2, or 3 lines) had distinctly fewer root hairs and abnormally swollen bases compared with the wild type (Supplemental Figure 2C), confirming that deficient expression of functional CAP1 was responsible for the mutant phenotype.

Ca²⁺ Gradient and Flux Disappeared in the Root Hair Tips of cap1-1

The tip-focused cytoplasmic Ca²⁺ gradient in root hairs regulates root hair growth (Schiefelbein et al., 1992). The Ca²⁺ concentration in Arabidopsis root hair tips increased from 200 nM to at least 1 μM (Wymer et al., 1997). We found that [Ca²⁺]_cyt in cap1-1 seedlings was lower than in the wild type (Supplemental Figure 1C). To understand the regulatory mechanism shaping mutant root hairs, cytoplasmic Ca²⁺ was detected by fluorescence resonance energy transfer based on YC3.6 (Yellow Cameleon 3.6) transformed plants (Monshausen et al., 2008). Consistent with the above results obtained by examining luminescence levels in seedlings, root hair cells and their adjacent epidermal cells in cap1 mutants showed lower levels of [Ca²⁺]_cyt than did the wild type (Supplemental Figure 3A). Furthermore, we found that growing wild-type root hairs displayed the typical tip-focused Ca²⁺ gradient in all three major growth stages (Figures 2A and 2B), with oscillating Ca²⁺ levels at the tip (Supplemental Figure 3B). In contrast, no such gradients were seen in cap1-1 mutants (Figures 2A and 2B; Supplemental Figure 3C), suggesting that the absence of a tip-directed Ca²⁺ gradient induced the formation of abnormal root hairs. Therefore, CAP1 plays an important role in establishing a calcium gradient in root hairs.

Figure 2.

Tip-Focused Ca²⁺ Gradients and Net Fluxes of Ca²⁺ in Root Hair Tips Disappeared in cap1-1 Mutants.

(A) Tip-focused Ca²⁺ gradients in the cytoplasm of Arabidopsis wild-type and cap1-1 root hairs expressing the Ca²⁺ sensor YC3.6 at different developmental stages (a, initiation phase; b, transition phase; and c, tip growth phase). Bright-field and fluorescence ratio images of Ca²⁺ in root hairs of the wild type and cap1-1 were obtained as described by Monshausen et al. (2008). Cytosolic Ca²⁺ levels were pseudo-color-coded according to the scale. Representative images of more than 10 measurements from three separate experiments per genotype are presented. Bars = 10 µm.

(B) Quantitative analysis of cytosolic Ca²⁺ levels in a representative growing root hair. Relative Ca²⁺ concentrations were measured in 10-µm² regions of interest along lengths of the root hairs in (A), as indicated with circles in the bright-field images. An increase in the fluorescence resonance energy transfer/cyan fluorescent protein ratio based on the Ca²⁺ sensor YC3.6 reflects an increase in cytoplasmic Ca²⁺ level.

(C) Ca²⁺ flux profiles of root hairs in the wild type, cap1-1, and cap1-1/CAP1. Ion-selective vibration microelectrode recordings of Ca²⁺ fluxes at the surfaces of root hairs of 7-d-old seedlings were made. Graphs show data from positions corresponding to the base (left) and tip (right) of root hairs (Supplemental Figure 4). Trace is a recording of a typical experimental plot illustrating Ca²⁺ influx and efflux in a root hair (inwards, negative; outwards, positive).

(D) Mean fluxes of Ca²⁺ in root hairs. Bars represent means ± se (n = 5).

Calcium ions may be imported across the plasma membrane of root hair tips via channels (Véry and Davies, 2000). Root hairs can import Ca²⁺ from the surrounding solution. We monitored the Ca²⁺ fluxes by the noninvasive micro-test technique (NMT) (Cárdenas et al., 1999) at two positions: the root hair apex (tip) and below the hair's midpoint (base) (Supplemental Figure 4). Inward Ca²⁺ fluxes of around −368 pmol cm⁻² s⁻¹ were recorded at the tip of the wild type, whereas small outward fluxes (60 pmol cm⁻² s⁻¹) were observed at the base. In contrast, mutant root hairs displayed similar flux values at both the tip and base (51 and 55 pmol cm⁻² s⁻¹, respectively; Figures 2C and 2D), suggesting that Ca²⁺ levels did not differ in the mutant root hair as a result of calcium channel influx. These results agree with those in YC3.6 transgenic mutant root hairs. The transgenic line (cap1-1/CAP1, #1) displayed results similar to the wild type (Figures 2C and 2D; −280 and 67 pmol cm⁻² s⁻¹ inward and outward, respectively). This complementation further indicated that CAP1 regulates root hair growth by affecting Ca²⁺ gradient formation.

Deficiency of NH₄⁺ Restores Root Hair Growth in cap1-1

To pinpoint the causes of root hair growth cessation, we attempted to grow cap1 mutant plants on MS media with different pH values, containing candidate chemicals such as ethylene and indole-3-acetic acid (both required for root hair elongation) (Pitts et al., 1998), or lacking essential ions such as iron, K, and phosphate. Surprisingly, cap1-1 formed wild-type-like root hairs when the MS medium lacked NH₄⁺ (wild type, 762 ± 14 µm; cap1-1, 725 ± 15 µm; P > 0.05, Figures 3A and 3B), whereas other chemicals had no significant effects (Supplemental Figure 5). The cap1-1 root hair morphogenesis (number, length, and cell pattern) resembled the wild type. To exclude the possibility of nitrogen deprivation, we performed the experiment without NO₃⁻. The wild type and cap1 mutants both had distinctly short root hairs (Figures 3A and 3B). This differential response to nitrogen source has two implications. First, in addition to being an essential nutrient, NH₄⁺ functions as a signaling molecule regulating root hair growth. Second, CAP1 serves as an important modulator for NH₄⁺ presence/absence in cells or the external environment.

Figure 3.

NH₄⁺ Deprivation Restored cap1-1 Root Hair Growth and Tip-Focused Ca²⁺ Gradient.

(A) The absence of NH₄⁺, but not NO₃⁻, restored root hair growth in cap1-1. Seedlings of the wild type and cap1-1 grown for 7 d on MS, MS lacking NH₄⁺ (MS-NH₄⁺), and MS lacking NO₃⁻ (MS-NO₃⁻). Root hairs of cap1-1 were only restored in NH₄⁺-free medium.

(B) Comparison of root hair lengths and numbers between the wild type and cap1-1 on MS, MS-NH₄⁺, and MS-NO₃⁻ media. Data were collected from seedlings grown on MS (n = 100), NH₄⁺-free (n = 90), and NO₃⁻ deprivation (n = 60) media from three separate experiments as described in Figure 1B. Error bars represent means ± se.

(C) The tip-focused Ca²⁺ gradient was reestablished in YC3.6-transformed cap1-1 on NH₄⁺-free but not on NO₃⁻ deprivation media. The [Ca²⁺]_cyt of root hairs was measured as described in Figure 2A (a, initiation phase; b, transition phase; and c, tip growth phase). Bars = 10 µm.

To elucidate the mechanism of this recovery effect, we monitored root hair Ca²⁺ gradients in seedlings grown in NH₄⁺ or NO₃⁻-deficient media. Interestingly, the root-tip Ca²⁺ gradient and oscillation were reconstituted in the mutant grown in the NH₄⁺-deficient medium but not in NO₃⁺-deficient medium (Figure 3C; Supplemental Figure 3D), indicating that NH₄⁺ affects Ca²⁺ distribution in mutant root hairs, further confirming that CAP1 is involved in an NH₄⁺-regulated root hair growth pathway.

CAP1 Is Localized in the Tonoplast and Expressed in Root Hairs

To determine CAP1 localization, mesophyll protoplasts that were transiently transformed with green fluorescent protein (GFP)–tagged CAP1 were observed. Compared with the control transformed with the empty pHBT-GFP-NOS vector, cells expressing the CAP1-GFP fusion protein exhibited GFP fluorescence that was distinct from the plasma membrane, strongly suggesting that CAP1-GFP is associated with intracellular membranes. To confirm this, protoplasts were cotransformed with the tonoplast marker γ-TIP1 (Hunter et al., 2007). The merged images of cotransformed protoplasts illustrated the tonoplast localization of the CAP1-GFP fusion protein (Figure 4A). CAP1-GFP fusion protein and GFP empty control were also introduced into onion epidermal cells. Transient expression showed that CAP1 localized in the onion tonoplast, while the epidermal strips transformed with the GFP empty vector fluoresced in a cytosol-rich region (Figure 4B). To confirm the location of CAP1, intact vacuoles were released from protoplasts in hypotonic solutions (Peiter et al., 2005). Fluorescence rapidly dispersed after disruption in protoplasts expressing only GFP (Figure 4C), while it remained associated with the vacuolar membrane in protoplasts expressing CAP1-GFP (Figure 4C), further demonstrating the tonoplast localization of CAP1.

Figure 4.

CAP1 Localizes to the Tonoplast and Is Expressed in Root Hairs.

(A) to (C) Confocal analysis of the subcellular localization of CAP1. Bars = 10 µm.

(A) CAP1-fused GFP was transiently coexpressed with the vacuolar membrane marker γ-TIP1 in protoplasts. Images showed colocalization of CAP1 and γ-TIP1 (bottom panels). Empty vector pHBT-GFP-NOS expressed in protoplasts was a control (top panels).

(B) Transient expression in onion epidermal cells shows fluorescence of CAP1 in the vacuolar membrane. Arrows point to the vacuole.

(C) Fluorescence images of GFP-transformed protoplast (top panels) and GFP-fused, CAP1-transformed protoplast (bottom panels). The same protoplast before (first image) and immediately after bursting (second two images) is shown.

(D) to (G) CAP1 expression pattern. GUS staining of transgenic plants shows CAP1 expressed in roots ([D] to [F]), leaves (E), and flowers (G). Inset image shows root hairs in (F).

(H) RT-PCR indicated that CAP1 was expressed in roots (R), flowers (F), inflorescence stems (S), and young leaves (L).

The CAP1 expression pattern was also tested in transgenic plants expressing β-glucuronidase (GUS) under control of the CAP1 promoter. GUS staining showed, and RT-PCR further confirmed, that CAP1 was universally expressed in most vegetative tissues (Figures 4D to 4H). In roots, strong GUS staining was observed in the maturation and elongation zones, while moderate staining was seen in meristematic zones (Figure 4F). Strong GUS activity was also observed in root hairs and stamens (Figures 4F and 4G).

CAP1 Modulates Ammonium Transport from the Cytoplasm to the Vacuole

Because CAP1 is localized in the vacuolar membrane, it may sense cytoplasmic NH₄⁺ levels and modulate the transport system to sequester excess NH₄⁺ in the vacuole for detoxification. We used NMT to test net NH₄⁺ fluxes of the vacuoles. Vacuoles from root hairs of 7-d-old seedlings were stabilized in a bathing solution for ∼10 min. Figure 5A shows the net influx of NH₄⁺ uptake (negative net flux) for the wild-type vacuoles (average I_in ≈ −97 ± 16 pmol·cm⁻²·s⁻¹, n = 17), demonstrating that there was an effective NH₄⁺ influx system in the tonoplast. In cap1-1 vacuoles, net fluxes of NH₄⁺ were significantly lower (average −19 ± 10 pmol·cm⁻²·s⁻¹, n = 15; P < 0.01, Dunnett's test) than in the wild type (Figure 5A). This implies either decreased influx or increased efflux activity in the vacuole of cap1-1.

Figure 5.

CAP1 Modulates Transportation of Ammonium from the Cytoplasm to the Vacuole.

(A) Changes in NH₄⁺ net fluxes of root hair vacuoles (inwards, negative; outwards, positive). Points are data collected every 6 s. Typical net flux traces are shown in the top panel. The NH₄⁺ net fluxes are averaged from MS (n = 17 for wild type, n = 15 for cap1) and MS-NH₄⁺ media (n = 7 for wild type, n = 9 for cap1) and plotted in the bottom panel. Error bars represent means ± se.

(B) Visualizing pHc in root hairs in stable transgenic wild-type and cap1-1 plants with phGFP using ratiometric pH-sensitive GFP. phGFP accumulates in the peripheral cytoplasm regions near the plasma membrane in cap1-1 seedling’s root hair cells grown on MS medium. pH levels were pseudo-color-coded according to the calibrated 410-nm/470-nm ratio image of the same root hair (Supplemental Figure 6). The different phases of root hair development for the wild type and cap1-1 are shown: a, the initial phase; b, transition phase; c, tip growth phase. Bars = 10 µm.

(C) Whole-cell voltage-activated currents in root hair protoplasts of the wild type and cap1-1 mutants. Typical time-dependent currents recorded in root hair cell protoplasts of the wild type (left) and cap1-1 (middle) and NH₄⁺ current-voltage relationships for the wild type (n = 4) and cap1-1 (n = 6) (I-V curve, right) are shown. I-V curves show means ± se.

(D) Effects of NH₄⁺ on whole-cell currents of wild-type and cap1-1 root hairs. The data of NH₄⁺ application to root hair cells by the addition of 150 mM NH₄⁺ to a pipette solution were recorded in the whole-cell patch-clamp configuration (wild type, n = 7; cap1-1, n = 4).

To determine whether NH₄⁺ affected net flux across the tonoplast, vacuoles were obtained from wild-type and cap1-1 seedlings grown on the NH₄⁺-free medium for 7 d. Wild-type vacuoles exhibited an increased NH₄⁺ release (positive net flux, average −31 ± 19 pmol·cm⁻²·s⁻¹, n = 7) (Figure 5A). The net flux values of vacuoles in cap1 plants under NH₄⁺-free conditions (23 ± 17 pmol·cm⁻²·s⁻¹, n = 9) were similar to those in the wild type. These flux dynamics showed the ability of wild-type vacuoles to efficiently manipulate cytoplasmic NH₄⁺ levels. These data indicated that CAP1 is required to sense NH₄⁺ homeostasis and to activate an ion transport system in the tonoplast to sequester excess NH₄⁺ in the vacuole.

Because no information was available on the intracellular localization of these changes in NH₄⁺ concentration (Speer et al., 1994), two ion-based cytoplasmic measurements were used to estimate the cytoplasmic concentrations of NH₄⁺ in the wild type and cap1-1. Since extracellular application of NH₄⁺ is a standard technique to elevate intracellular pH in animal cells (Boron and De Weer, 1976; Grinstein et al., 1994), root hair cytosolic pH may be alkalinized by equimolar application of NH₄⁺ in MS medium. In the first set of experiments, we visualized cytosolic pH (pHc) in root hairs of stable transgenic wild-type and cap1-1 plants using ratiometric pH-sensitive GFP (Moseyko and Feldman, 2001). Ratio imaging of the pHc indicated a distinct gradient in different root hair developmental stages in the wild type, whose tip zones had relatively alkaline pHc (Figure 5B). As expected, intracellular pH during either the swelling formation or tip growth stages increased significantly after application of 100 mM NH₄⁺ for 20 min (Supplemental Figure 7). However, the whole cytoplasm of cap1-1 root hairs in MS medium had higher pHc (≈8.0) and lacked a pHc gradient. Interestingly, the pHc gradient of mutants was restored in NH₄⁺-deficient medium (Figure 5B). These results are consistent with the cap1 mutant's accumulating more NH₄⁺ in the cytoplasm, which disturbed the [Ca²⁺]_cyt and pHc gradients in root hairs.

To further confirm whether the cap1-1 cytoplasm accumulated high levels of NH₄⁺, we first tried to record the current across the tonoplast using the whole-cell patch clamp technique. However, we could not obtain the NH₄⁺ current from the root hair vacuole under our conditions. Ammonium concentration gradients and membrane potentials have been reported to be able to drive the flow of NH₄⁺ across plasma membranes (Ludewig et al., 2002, 2003). Therefore, if more NH₄⁺ accumulated in the cytoplasm, less should flow across the plasma membrane to the cytosol.

Next, we monitored whole-cell voltage-activated NH₄⁺ currents across plasma membranes under standard experimental conditions to indicate the cytoplasmic NH₄⁺ levels. Voltage pulses between −190 and 110 mV, in 20-mV increments, were applied to the plasma membrane of a root hair cell protoplast (Figure 5C). An inward current was activated by hyperpolarization of the plasma membrane and an outward current by depolarization. Because Ba²⁺ permeates many Ca²⁺ channels (Gelli and Blumwald, 1997; Klusener and Weiler, 1999), 10 mM Ba²⁺ was supplied in the bathing solution as the charge-carrying ion. In the wild-type root hair protoplast, no significant changes in current were observed (Supplemental Figure 8A). Additional experiments showed that the whole-cell current was not blocked by application of 1 mM La³⁺ (a calcium channel inhibitor) (Supplemental Figure 8B). The data demonstrated that NH₄⁺ carried the current, not calcium ions. The I-V curves (current-voltage) in Figure 5C (right panel) summarize the data recorded in wild-type and cap1-1 root hair protoplasts. At a membrane potential of −190 mV, the average inward current of the wild type (−185.7 ± 24.9 pA; n = 4) and cap1-1 (−82.6 ± 18.8 pA; n = 6) was significantly different (P < 0.05). The lower inward NH₄⁺ current of cap1-1 may imply that mutation of CAP1 could result in more NH₄⁺ accumulating in the cytoplasm.

To confirm this notion, we tried to imitate changes in the cytoplasmic NH₄⁺ concentration by replacing 100 mM K-glutamate in the pipette solution with 150 mM NH₄⁺-glutamate, as previously described (Murata et al., 2001; Warth et al., 2004). A large reduction of inward current was observed in the wild type. On average, currents of −190 mV decreased to −115.3 ± 35.2 pA, and those of 110 mV increased to 458.3 ± 25.2 pA (n = 7). However, currents remained essentially unchanged in cap1-1 (Figure 5D): Currents at 110 mV were the same (n = 4), while those at −190 mV increased by a statistically insignificant amount (from −82.6 ± 18.8 pA to −111.5 ± 51.3 pA; P > 0.05).

In the cap1 mutant, the cytosolic NH₄⁺ concentration increased under NH₄⁺ nutrition. Thus, the cap1 mutant is expected to be more sensitive to NH₄⁺ toxicity. We therefore analyzed the effects of media with various pH values and high amounts of NH₄⁺ on seedling growth. The seeds of the wild type, cap1-1, and two complementary lines (com-1, 2) were sown on nutrient media with different pH values. The cap1 mutant was sensitive to either pH 7.0 or pH 5.5 with higher ammonium (NH₄Cl) levels (Supplemental Figure 9), as indicated by severely inhibited seedling growth. Cotyledons did not green on MS medium with a 6-fold concentration of NH₄⁺ until the 11th day. Thus, the greater sensitivity to NH₄⁺ toxicity of cap1-1 further supported the conclusion that high levels of NH₄⁺ accumulated in its cytoplasm.

CAP1 Can Be Autophosphorylated and Functionally Complement the Activity of Npr1 Kinase in Yeast

To examine the protein kinase activity of CAP1, we constructed glutathione S-transferase (GST)–tagged CAP1. CAP1 could autophosphorylate, but boiled CAP1 (the control) could not (Figure 6A). Phosphoproteomics data indicated that CAP1 is a major phosphorylated protein in plants exposed to rapid changes in NO₃⁻ or NH₄⁺ concentrations (Engelsberger and Schulze, 2012). The peptide INIGGDLI(pS)PK of CAP1 corresponds to the conserved malectin domain in the RLK family that controls glucose binding.

Figure 6.

CAP1 Can Be Autophosphorylated and Complements the Activity of NPR1 Kinase in Yeast.

(A) In vitro phosphorylation of CAP1 kinase activity. Recombinant protein fused with GST was used in the kinase assay. Boiled CAP1 protein (GST-CAP1*) was a negative control. Autophosphorylation was detected after protein gel electrophoresis and phosphor imaging.

(B) Growth of yeast strains on solid minimal medium containing different concentrations of NH₄⁺ as the sole nitrogen source. Strains 23344c (ura3, wild type) and 21994b (npr1-1, ura3) transformed with plasmid pFL38 and pFL38-CAP1 were spotted at 10⁰-, 10⁻¹-, and 10⁻²-fold dilutions on YNB medium supplemented with 0.2 and 3 mM NH₄Cl and incubated for 4 d at 29°C.

(C) In vitro kinase assay for MEP2 phosphorylation by CAP1. Recombinant C-terminal fragments of CAP1 (CT-CAP1-GST) and MEP2 (CT-MEP2-GST) were cultured in kinase buffer, and phosphorylated MEP2 exhibited slower mobility (arrows). MEP2 treated with CIAP (alkaline phosphatase) was the control. The phosphorylation reactions were stopped by boiling for 5 min, and the proteins were separated by SDS-PAGE.

In yeast, three Mep-type NH₄⁺ transport systems (Mep1, Mep2, and Mep3) are involved in NH₄⁺ uptake (Marini et al., 1994; Boeckstaens et al., 2007). MEP2 is a specific NH₄⁺ sensor that stimulates pseudohyphal growth during NH₄⁺ limitation (Lorenz and Heitman, 1998); its optimal NH₄⁺ transport activity requires Npr1 (nitrogen permease reactivator protein) kinase (Boeckstaens et al., 2007), a potential target of rapamycin signaling. Yeast growth in low-NH₄⁺ medium is similarly affected in npr1 mutant cells and in cells lacking the three MEP genes (Feller et al., 2006). The finding that CAP1 functionally perceives NH₄⁺ homeostasis in the cytosol raises the possibility that CAP1 resembles MEP2 and Npr1 in yeast and could modulate NH₄⁺ signaling to regulate root hair growth. To examine whether CAP1 was activated upon NH₄⁺ deprivation, we transferred CAP1 to yeast strains 23344c (ura3) and npr1 mutant cells 21994b (npr1-1 ura3) (Boeckstaens et al., 2007). As shown in Figure 6B, growth of npr1 was greatly impaired on low-NH₄⁺ medium but was rescued by CAP1. To investigate whether MEP2 is a substrate of CAP1, a phosphorylation assay was performed with the recombinant C-terminal tail of MEP2 containing all phosphorylation sites (Van Zeebroeck et al., 2011). As expected, CAP1 could phosphorylate MEP2, while no phosphorylated band was observed in the control protein (Figure 6C). Thus, CAP1, like Npr1 kinase, is required for the optimal uptake activity of NH₄⁺ transport systems, indicating that CAP1 is involved in NH₄⁺ sensing and transport system activation in plants.

DISCUSSION

Ammonium, both as a necessary nutrient and an important signal in plants, can trigger and regulate a number of critical functions to balance metabolism and uptake at varying nutrient concentrations. The mechanism of NH₄⁺ uptake in root hairs has already been established, with AMT family proteins identified as the sole transporters (Lauter et al., 1996; Gazzarrini et al., 1999). However, the precise mechanism by which plants deal with NH₄⁺ toxicity remains unknown. Therefore, the exploration of NH₄⁺ homeostasis sensing and regulation is an essential and interesting scientific endeavor. Our data identified an important component of the NH₄⁺ signaling pathway, CAP1, and provided clues to its precise role in sensing and modulating cytoplasmic NH₄⁺ by regulating compartmentation into vacuoles. Indeed, in cap1 mutants, the presence of NH₄⁺ terminated the growth of root hairs (Figure 1), while growth restored in its absence (Figure 3). In addition, the pHc was alkalized in cap1 mutants (Figure 5B). These data suggested that CAP1 can assess cytoplasmic NH₄⁺ levels and phosphorylate an unknown target, such as TIP (Loqué et al., 2005), resulting in compartmentalization. Excess NH₄⁺ causes toxicity in plant and animal cells. Vacuole compartmentalization of NH₄⁺, which doubles the NH₄⁺ concentration in the vacuole compared with the cytoplasm (Loqué and von Wirén, 2004), is important in protecting cells from excess NH₄⁺.

The [Ca²⁺]_cyt gradient in root hair cells may be essential for root hair growth (Monshausen et al., 2008; Schiefelbein et al., 1992; Bibikova et al., 1997). Inhibition of this gradient halts growth. In this study, mutant cap1 had short root hairs and lacked an apparent Ca²⁺ gradient (Figures 1 and 2; Supplemental Figure 3). Consistent with this, calcium fluxes in cap1-mutant root hair tips were significantly smaller than in the wild type (Figure 2). The dwarf phenotypes in herk1 the1 double mutants and FER knockdown plants suggested that HERK1, THE1, and FER, members of the CrRLK1L subfamily participate in cell elongation in hypocotyls and petioles (Guo et al., 2009). ANX1 and ANX2, also in the subfamily, specifically regulate the growth of pollen tubes, but not of vegetative tissues (Boisson-Dernier et al., 2009). Another member of the subfamily, CAP1, is expressed in root hairs, and serves a similar role in regulating tissue-specific cell elongation, i.e., root hair tip growth.

Our data demonstrated that CAP1 regulates the establishment and maintenance of a Ca²⁺ gradient at the root hair tip, maintaining root hair growth. Analysis of the cap1 mutant also supported the idea that NH₄⁺ is an important factor for maintaining and reestablishing [Ca²⁺]_cyt and pHc gradients in root hairs for polar growth (Figure 3). The tip-focused Ca²⁺ gradient in root hairs is sustained by voltage-gated Ca²⁺ channels (Véry and Davies, 2000) and by ROS (Foreman et al., 2003). Cytoplasmic Ca²⁺ has been shown to decrease in canola root cells when they are taking up NH₄⁺ (Babourina et al., 2007). Our results, similar to the observation of Babourina et al. (2007), demonstrated a link between NH₄⁺ and Ca²⁺ gradient maintenance through CAP1 (Figure 3C; Supplemental Figure 3D). Depriving plants of individual nutrients or essential elements proved unsuccessful in restoring tip growth of cap1 root hairs, with the exception of NH₄⁺, indicating that NH₄⁺ may serve as a signal for CAP1. Interestingly, on MS medium (which has 20.6 mM NH₄⁺), cap1-1 root hairs could not elongate, but the mutants formed normal root hairs and had normal Ca²⁺ localization when the medium lacked NH₄⁺ (Figure 3). The accumulation of NH₄⁺ in the cytoplasm may lead to plasma membrane depolarization, resulting in channel activation/deactivation and the consequent formation of a [Ca²⁺]_cyt gradient in root hairs. Therefore, we propose a mechanism in which root hair growth defects in cap1-1 cells arise through the accumulation of NH₄⁺ and impair homeostasis sensing in the cytoplasm. Thus, excess NH₄⁺ in the cytoplasm of cap1-1 mutants will disturb homeostasis by disrupting the establishment of the normal [Ca²⁺]_cyt gradient, halting root hair growth. However, this distinct response to extracellular NH₄⁺ by the cap1 mutant indicates that understanding the signaling pathway of NH₄⁺-related CAP1-regulated tip growth may shed light on the mechanism of tip growth.

How CAP1 senses NH₄⁺ homeostasis for detoxification is unclear at present. However, our results showed that CAP1 could functionally complement a yeast mutant defective in high-affinity NH₄⁺ uptake by npr1 kinase and showed kinase activity that regulated NH₄⁺ transport between the cytoplasm and vacuole (Figures 5 and 6). The amplitude of extracellular and intracellular pH oscillations increases after the addition of NH₄NO₃ to the growth medium, whereas an overall decrease in cytoplasmic pH at the cell apex is induced (Bloch et al., 2011). In animal cells, in contrast, extracellular application of NH₄⁺ alkalinizes the cytoplasmic medium (Boron and De Weer, 1976; Grinstein et al., 1994). These findings are consistent with our observation that suddenly imposing a high level of NH₄⁺ in MS medium on root hairs elevated intracellular pH (Supplemental Figure 7). In cap1-1, a lack of NH₄⁺ in the medium reconstituted the pH gradient and restored tip growth of root hairs. Thus, when CAP1 is defective, the greater accumulation of NH₄⁺ and the effects of pH changes arrested root hair growth, which manifested as swollen root hairs. In agreement with a possible effect of CAP1 on regulating NH₄⁺ homeostasis in root hairs, we observed that cap1 mutants were more sensitive than the wild type to higher NH₄⁺ concentrations in MS medium (Supplemental Figure 9). One possible mechanism for this modulation is that NH₄⁺ accumulation induces the activation of CAP1 for phosphorylation of unknown targets that compartmentalize NH₄⁺ into the vacuole.

Based on previous reports and the data presented here, we propose a working model for NH₄⁺ detoxification in which CAP1 senses the cytoplasmic NH₄⁺ concentration and regulates its uptake and relocation (Figure 7). In the wild type, CAP1 in the vacuolar membrane could sense the NH₄⁺ concentration in the cytoplasm and trigger the transporter and/or relocation system in the tonoplast, resulting in the accumulation of NH₄⁺ in the vacuole and the maintenance of NH₄⁺ homeostasis. When CAP1 was rendered deficient in the mutant, excess NH₄⁺ collected in the cytoplasm, causing a loss of calcium gradient and eventual cessation of root hair growth because of toxicity. Our data elucidated an important feature, CAP1, of the NH₄⁺ signaling pathway and provided clues to its precise role in sensing and transmitting signals in NH₄⁺ homeostasis. In summary, CAP1 is an important NH₄⁺ modulator that participates in the pathway of environmental NH₄⁺-regulated root hair growth by regulating tip-focused cytoplasmic Ca²⁺ gradients.

Figure 7.

Model Showing the Putative Regulation Pathway of NH₄⁺ Homeostasis Mediated by CAP1 in Root Hairs.

In the wild type (in blue), CAP1 in the vacuolar membrane senses cytosolic NH₄⁺ levels and phosphorylates an unknown target, such as tonoplast intrinsic protein, resulting in compartmentalization of NH₄⁺ in the vacuole and the maintenance of normal NH₄⁺ levels in the cytoplasm. Normal NH₄⁺ homeostasis is necessary for the establishment of the [Ca²⁺]_cyt gradient in the polar growth of root hairs. When CAP1 was rendered deficient in the mutant (in gray), inward NH₄⁺ flux across the vacuole membrane significantly decreased and excess NH₄⁺ accumulated in the cytoplasm, causing a loss of calcium gradient (black line with block end) and eventual cessation of root hair growth. The cytosolic Ca²⁺ gradient in the wild type is indicated by the scale below. No gradient was established in cap1 mutants. Red dots represent ammonium ions. Red lines with arrows indicate ion influx.

METHODS

Plant Materials and Growth Conditions

All Arabidopsis thaliana plants used were in the Columbia background. Surface-sterilized seeds were sowed on MS medium and kept for 3 d at 4°C in the dark to break dormancy. To analyze root hair growth, different hormones or ions were added or removed from the medium. The plates were then transferred to a culture room at 22°C with a 16-h-light/8-h-dark photoperiod and grown vertically. The CAP1 T-DNA insertion line Salk_083442 was obtained from the ABRC.

Constructions and Plant Lines

A gene-specific fragment of CAP1 cDNA was amplified by PCR. For the CAP1 overexpression construct, the primers 5′-TTACCCGGGATGGGAGGAGATTTTCGT-3′ and 5′-TCCGAGCTCTCACGGTATTGAATGCGA-3′ were used. The amplified product was cloned into the pCAMBIAsuper-1300 vector cut with SmaII and SacI. For the GFP reporter vector, the full-length CAP1 coding sequence was amplified with primers 5′-AAAGGTACCATGGGAGGAGATTTTCGTCAT-3′ and 5′-TTTCCCGGGAACGGTATTGAATGCGACG-3′ and then cloned into pHBT-GFP vector cut with KpnI and SmaII. To generate the CAP1-GUS construct, a 619-bp CAP1 promoter fragment was amplified using primers 5′-GCCGAATTCTCGCTTTGAGGTCATTTT-3′ and 5′-GAACTGCAGAATATCCGGCGAGGTTTTGAAG-3′, and the PCR product was digested with EcoRI and PstI and cloned into pCAMBIA1381 vector. The constructs were introduced into Agrobacterium tumefaciens strain GV3101 and transformed by floral infiltration into the wild type (for GUS staining assays) and cap1-1 mutant (for gene complementation) plants.

To generate the gene silencing vector, 151-bp fragments corresponding to the 5′ coding sequence of CAP1 were amplified by PCR with the primers CAP1-XhoI-S-F (5′-GAACTCGAGCGCTTAACGACCTTATCTT-3′) and CAP1-NcoI-S-R (5′-GAACCATGGGTGATCGTAGACGCATTAA-3′) and the primers CAP1-SmaI-S-F (5′-TTACCCGGGCGCTTAACGACCTTATCTT-3′) and CAP1-BamHI-S-R (5′-TAAGGATCCGGTGATCGTAGACGCATTA-3′). PCR products were digested with XhoI-NcoI and SmaI-BamHI to produce hairpin RNA for CAP1. The two fragments were sequentially ligated to the vector pFGC5941-35S-RNAi. The resulting construct was transformed into Arabidopsis Columbia-0 (Col-0) mediated by Agrobacterium GV3101 via the standard floral dip method.

For phosphorylation assays of CAP1 and MEP2, GST-fused plasmid constructions were made. The C termini of CAP1 (1549 to 2529 bp) and MEP2 (1251 to 1500 bp) were amplified with the following primer pairs: forward CT-CAP1-GST (5′-ATAGGTACCGAACTACAGACCGCGACACAA-3′) and reverse CT-CAP1-GST (5′-TTTGAATTCTCACGGTATTGAATGCGACGG-3′), and forward CT-MEP2-GST (5′-ACGGTCGACTTCCATTTTTAAAACTAAGA-3′) and reverse CT-MEP2-GST (5′-GCGAAGCTTCTTATACTATATGGTCAGTGT-3′), respectively. The PCR fragments were cloned as EcoRI-KpnI (for CT-CAP1-GST) and SalI-HindIII (for CT-MEP2-GST) restriction fragments in the pGEX-2T Escherichia coli expression vector for expressing GST-tagged C-terminal proteins in E. coli strain BL21.

Measurement of Root Hair Length

Root hairs were observed in 7-d-old seedlings grown on MS agar plates. Root hair lengths were determined by measuring the five longest root hairs within 5 mm of the root tip of each line. Root hairs were photographed using a FV1000 confocal microscope (Olympus). Root hair lengths were measured using ImageJ (http://rsbweb.nih.gov/ij/).

Calcium Imaging

The 35S promoter-driven YC3.6 vector (in pEarleyGate100 vector) was kindly provided by Simon Gilroy. Ratio images were performed essentially as described previously (Monshausen et al., 2008) using root hairs of wild-type and cap1-1 transgenic seedlings and the FV1000 confocal microscope. The process of root hair elongation takes ∼300 min (Dolan et al., 1994), and focusing on the same root cells through all the growth stages is difficult. Therefore, different root hair cells at different stages were scanned along the root tip.

Subcellular Localization and Histochemical Detection of GUS Activity

For GFP analysis, the final construct pHBT-GFP-AtCAP1 and empty vector pHBT-GFP were transiently expressed in mesophyll protoplasts (Yoo et al., 2007) and in onion epidermal cells using a particle gun–mediated system (Li et al., 2010). For transformation with protoplasts, GFP fluorescence was scanned after vacuoles were released from protoplasts (Peiter et al., 2005). Vacuolar membrane–localized pHBT-γTIP1-DsRED was cotransformed with pHBT-GFP-AtCAP1, and the empty vector pHBT-DsRED was cotransformed with pHBT-GFP for further analysis of CAP1 location. Transgenic cells were examined under the FV1000 confocal microscope. For the GUS assays, excised tissues from 15 independent transgenic lines containing the CAP1 promoter–GUS construct were tested according to Song et al. (2005).

Net Ca²⁺ and NH₄⁺ Flux Measurements with NMT

Net fluxes were obtained with a BIO-IM Series NMT system (YoungerUSA) at the Xuyue Beijing NMT Research Service Center, China.

Seven-day-old Arabidopsis roots cultured in MS media were placed in Petri dishes containing 10 mL of liquid medium (0.1 mM CaCl₂, 0.1 mM KCl, 0.1 mM MgCl₂, 0.5 mM NaCl, 0.2 mM Na₂SO₄, 0.3 mM MES, and 1% sucrose, pH 6.0), and treated for 20 to 30 min before the Ca²⁺ flux experiments. The same medium was used for net flux measurements throughout our experiments. After primary scans along the root hairs, the root hair tips and middle regions were selected for the measurement of net Ca²⁺ fluxes.

Microelectrodes selective for Ca²⁺ were freshly fabricated prior to the NMT tests. Silanized glass microelectrodes (inner diameter 4 ± 1 μm; YG-IS-ME02; YoungerUSA) were first backfilled with Ca²⁺ solution (100 mM CaCl₂) to a length of ∼1 cm and then front-filled with 25-μm columns of selective Liquid Ion eXchange (LIX; YG-LIX-Ca01; YoungerUSA). Before and after each flux measurement, the microelectrodes were calibrated with the culture medium with 1, 0.1, and 0.01 mM Ca²⁺. Only electrodes with a Nernstian slope >26 mV/decade for Ca²⁺ were used in our study.

Seven-day-old seedlings were used for net NH₄⁺ flux measurements. Root epidermal cell protoplasts were first isolated from the root hair zones by digesting the zones for 2 to 3 h in 1.5% cellulose RS (Yakult) and 0.075% pectolyase Y-23 (Yakult), and then vacuoles were released by washing these protoplasts with a solution containing 10 mM EDTA and 10 mM EGTA, pH 7.8, with an osmolarity of 120 mOsM adjusted with mannitol. Vacuoles adhered to poly-Lys pretreated cover slides that were presettled in a Petri dish. Vacuoles were gently rinsed with bathing solution (0.1 mM NH₄Cl, 0.1 mM KCl, 0.5 mM NaCl, 0.1 mM CaCl₂, 120 mM mannitol, and 0.05 mM MES, pH 7.8) and stabilized in this solution for ∼10 min before measurement. Net flux measurements were performed in this solution. Microelectrodes selective for NH₄⁺ were freshly made prior to the NMT tests. Silanized glass microelectrodes (diameter 1.5 ± 0.5 μm; XY-DJ-02; YoungerUSA) were first backfilled with NH₄⁺ solution (100 mM NH₄Cl) to a length of ∼1 cm and then front-filled with 15- to 50-μm columns of selective liquid ion-exchange cocktails (LIX; XY-SJ-NH₄; YoungerUSA). Before measurements, microelectrodes selective for NH₄⁺ were first calibrated with bathing solution with 0.05 and 0.5 mM NH₄⁺. Only electrodes with a Nernstian slope >53 mV/decade for NH₄⁺ were used in our study.

The net ion fluxes were calculated based on Fick’s law of diffusion: J = −D₀ (dc/dx), where J is the ion flux (picomoles/cm²/s), D₀ is the ionic diffusion coefficient of a specific ion in a given medium, dc is the ion concentration difference based on microvolt differences, and dx is the distance the microelectrode moved from one point to another perpendicular to the root hair or vacuole surfaces. The microelectrodes were positioned 1 ± 0.5 μm away from the samples by the computer-controlled NMT system. Net fluxes were calculated using JCal 1.0 (a free MS Excel spreadsheet, http://youngerusa.com/jcal or http://ifluxes.com/jcal).

Yeast Expression

Yeast strains 23344c (ura3, wild type) and 21994b (npr1-1 ura3) transformed with plasmids pFL38 and pFL38-CAP1 were grown in liquid YPD (yeast extract peptone dextrose) medium at 29°C for ∼1 d, diluted 10⁻¹- to 10⁻²-fold, and dropped on solid Yeast Nitrogen Base W/O amino acids medium (Difco) supplemented with 0.2 and 3 mM NH₄Cl. Yeast cells were incubated for 4 d at 29°C.

Kinase Activity Assay

The recombinant GST-tagged fusion protein was affinity-purified from bacterial BL21 lysate using glutathione sepharose 4B (GE Healthcare). The kinase activities of the fusion proteins were then measured according to the method of Peck (2006).

According to Van Zeebroeck et al. (2011), the GST-tagged C-terminal tail of MEP2 was purified from E. coli strain BL21. For the phosphorylation assay, C-terminal proteins of CAP1 and MEP2 were cultured in the kinase buffer (500 mM HEPES, pH 7.4, 100 mM MgCl₂, and 20 mM MnCl₂) with 10 mM ATP for 30 and 60 min. MEP2 treated with CIAP (alkaline phosphatase; TaKaRa) for 30 min was used as the control. The reactions were stopped by boiling for 5 min, and the proteins were separated by SDS-PAGE.

Whole-Cell Patch Clamping and Data Acquisition

Arabidopsis root hair cell protoplasts were isolated as previously described (Ivashikina et al., 2001). Highly vacuolated root hair cell protoplasts were patch clamped (Véry and Davies, 2000). The bathing solution included 10 mM NH₄-glutamate, 2 mM CaCl₂, 4 mM MgCl₂, and 5 mM MES, pH 5.6, with osmolality adjusted to 300 mOsM with sorbitol. The standard pipette solution contained 100 mM K-glutamate, 2 mM MgCl₂, 2 mM EGTA, 10 mM HEPES, and 2 mM Mg-ATP, pH 7.2, with adjusted osmolality of 420 mOsM. Recording pipettes were made from borosilicate glass capillaries (Kimax-51; Kimble Glass) using a vertical two-stage puller (model PC-10; Narishige) and fire-polished with a microforge (model MF-90; Narishige) before use. Whole-cell NH₄⁺-current recordings were performed essentially as described previously (Bei and Luan, 1998; Ivashikina et al., 2001). Data were acquired 15 min after the formation of the whole-cell configuration. Whole-cell currents were measured in response to 3-s voltage pulses from −190 to 110 mV in 20-mV steps using an EPC-9 patch-clamp amplifier (HEKA Elektronik). Whole-cell data were analyzed with the software PLUSE and PLUSEFIT (version 8.3) as described by Zhang et al. (2001).

pHc Imaging

Stable Pt–GFP transgenic wild-type line N9561 (background: Col-0) was obtained from the Nottingham Arabidopsis Stock Centre (http://Arabidopsis.info/StockInfo?NASC_id=9561), and cap1-1 plants containing pH-sensitive GFP were obtained by crossing with Pt-GFP lines. Root hairs of 4- to 5-d-old seedlings were used to monitor intracellular pH. GFP fluorescence was monitored using a FV1000 confocal microscope (exciters, 410 nm and 470 nm; emitter, 525 nm); the F410-to-F470 ratio was used as a measure of pH. pH levels were pseudo-color-coded according to the calibrated 410-nm/470-nm ratio image of the same root hair. The pH titrations were performed in situ as described by Moseyko and Feldman (2001).

Generation of pMAQ2 Transgenic Plants and Measurements of Ca²⁺ Luminescence

The pMAQ2 plasmid was transformed into Arabidopsis Col-0 and the homozygous T-DNA insertion lines mediated by Agrobacterium GV3101 via the standard floral dip method. Stable transgenic Arabidopsis plants expressing cytosolic apoaequorin were used. For [Ca²⁺]_cyt measurements, aequorin was reconstituted in vivo essentially as described previously (Knight et al., 1991) by incubating seedlings in water containing 2.5 μM native coelenterazine (Promega) overnight in the dark at room temperature. One seedling was placed in a transparent plastic cuvette without liquid. The cuvette was placed inside a TD20/20n digital luminometer (Turner Biosystems). Luminescence was recorded every 0.2 s. At the end of each experiment, the remaining aequorin was discharged by adding an equal volume of 2 M CaCl₂ and 20% ethanol. Luminescence values were converted to calcium concentrations as described previously (Knight et al., 1996).

CAP1 Expression Analyzed by RT-PCR

RNA was isolated from 100 mg roots using TRIzol reagent (Invitrogen). First-strand cDNA was synthesized using M-MLV reverse transcriptase (Promega). Primers used for CAP1 were: 5′-TTGATGGGAAATACAAAGGAC-3′ and 5′-GGACAAGGTAGGAATAGGGTTA-3′. Another pair of primers for obtaining smaller fragments across the deletion region is: 5′-CGCTAACCGCTTTCTTAGGGGTTGT-3′ and 5′-TCGGTAAAGGGAAAGTACCGACCTA-3′. The cDNA yield was determined by measuring the amount of product that was amplified from an internal standard, the housekeeping gene Actin2. Primers used for Actin2 were: 5′-TTCCTCATGCCATCCTCCGTCTT-3′ and 5′-CAGCGATACCTGAGAACATAGTGG-3′. All PCR reactions were performed in triplicate.

Statistical Analysis

To determine significant differences among different lines or different treatments, all the data were analyzed by Dunnett's test using SPSS16.0 software.

Accession Numbers

Sequence data from this article can be found in the National Center for Biotechnology Information database under the following accession numbers: Saccharomyces cerevisiae MEP2 (NM_001182980) and NPR1 (Z71459); Arabidopsis gene CAP1 (At5g61350).

Supplemental Data

The following materials are available in the online version of this article.

• Supplemental Figure 1. The CAP1 T-DNA Insertion Mutant Has a Lower Cytoplasmic Calcium Concentration and BLAST Results of cap1-1 and Overexpression Transcripts.

• Supplemental Figure 2. Root Hair Phenotype Was Complemented by CAP1 Driven by the Wild-Type Promoter in Mutants and Suppression of CAP1 Expression in Arabidopsis RNAi Knockdown Lines.

• Supplemental Figure 3. FRET-Based YC3.6 Shows Levels of [Ca²⁺]_cyt and Oscillation of Tip-Focused Ca²⁺ Gradients Only in Wild-Type Root Hairs.

• Supplemental Figure 4. Ion-Selective Vibration Microelectrode Recording of Ca²⁺ Fluxes at Root Hair Surfaces in 7-d-Old Seedlings.

• Supplemental Figure 5. Effects of Auxin (IAA), Ethylene (ETH), Various Nutrients, and pH on the Growth of Root Hairs in Wild-Type and cap1-1 Plants.

• Supplemental Figure 6. Standard Curve of the Calibrated 410 nm/470 nm Ratio in Different pH Solutions.

• Supplemental Figure 7. High Levels of NH₄⁺ Enhances Root Hair pHc.

• Supplemental Figure 8. Characterization of Ammonium Current Recording on the Whole-Cell Model Using Ba²⁺ and La³⁺.

• Supplemental Figure 9. cap1-1 Was More Sensitive to High Levels of Ammonium Than Wild-Type and the Complementary Lines.

Glossary

RLK

receptor-like kinase

ROS

reactive oxygen species

ROP

Rho-related GTPases of plant

AMT

ammonium transporter

MS

Murashige and Skoog

[Ca²⁺]_cyt

free cytosolic concentration

YC3.6

Yellow Cameleon 3.6

NMT

noninvasive micro-test technique

pHc

cytosolic pH

Col-0

Columbia-0