Root hair defective4 encodes a phosphatidylinositol-4-phosphate phosphatase required for proper root hair development in Arabidopsis thaliana

Abstract

Polarized expansion of root hair cells in Arabidopsis thaliana is improperly controlled in root hair-defective rhd4-1 mutant plants, resulting in root hairs that are shorter and randomly form bulges along their length. Using time-lapse fluorescence microscopy in rhd4-1 root hairs, we analyzed membrane dynamics after labeling with RabA4b, a marker for polarized membrane trafficking in root hairs. This revealed stochastic loss and recovery of the RabA4b compartment in the tips of growing root hairs, consistent with a role for the RHD4 protein in regulation of polarized membrane trafficking in these cells. The wild-type RHD4 gene was identified by map-based cloning and was found to encode a Sac1p-like phosphoinositide phosphatase. RHD4 displayed a preference for phosphatidylinositol-4-phosphate [PI(4)P] in vitro, and rhd4-1 roots accumulated higher levels of PI(4)P in vivo. In wild-type root hairs, PI(4)P accumulated primarily in a tip-localized plasma membrane domain, but in rhd4-1 mutants, significant levels of PI(4)P were detected associated with internal membranes. A fluorescent RHD4 fusion protein localized to membranes at the tips of growing root hairs. We propose that RHD4 is selectively recruited to RabA4b-labeled membranes that are involved in polarized expansion of root hair cells and that, in conjunction with the phosphoinositide kinase PI-4Kbeta1, RHD4 regulates the accumulation of PI(4)P on membrane compartments at the tips of growing root hairs.

Figure 1.

Spatial and Temporal Distribution of RabA4b Is Altered in rhd4-1 Mutants. Wild-type and rhd4-1 Arabidopsis seedlings expressing EYFP-RabA4b were visualized by time-lapse fluorescence microscopy, and frames from a 20-min period are shown, with a reference line for root hair length at time 0. Relative length was measured from time 0 as 0 μm. The fluorescent signal located within the proximal 7% of the length of the root hair was defined as tip fluorescence, and this was presented as a percentage of the fluorescence detected within the entire root hair. (A) EYFP-RabA4b was localized to the tip of a growing wild-type root hair over a time period of 20 min. (B) Quantification of EYFP-RabA4b localization in a wild-type root hair. Black line, relative length of the root hair in micrometers; gray line, percentage of tip localization of the fluorescent signal. (C) EYFP-RabA4b localization is altered in a growing rhd4-1 root hair. Arrows indicate time points at which the root hair is bulging, and EYFP-A4b localization is lost. (D) Quantification of EYFP-RabA4b localization in an rhd4-1 root hair. Root hair tip growth correlates with increased EYFP-RabA4b tip fluorescence (shaded regions). Arrows indicate time points at which the root hair is bulging, and EYFP-A4b localization is lost. Black line, relative length of the root hair in micrometers; gray line, percentage of tip localization of the fluorescent signal.

Figure 2.

RHD4 Encodes a Sac1p-Like Phosphoinositide Phosphatase. (A) Map-based cloning of the rhd4 mutation. RHD4 is located on the right arm of chromosome 3, At3g51460. For spatial reference, described genetic markers are shown with their corresponding map position in centimorgans (cM). The rhd4-1 allele had a G→A base pair substitution at the first base of the first intron. The rhd4-2 allele had a mutation in exon 16 causing an amino acid substitution of Gly to Ser, G449S, in the protein. Two SALK T-DNA insertional lines were obtained, designated rhd4-3 and rhd4-4, and these had insertions in the second and fourth introns, respectively. (B) Amino acid alignment of the SAC domains of RHD4 and other SAC domain proteins from yeast and animals. The SAC domain sequences of yeast Sac1 (Sac1p), human Sac2 (hSAC2), rat SAC1 (rSAC1), human synaptojanin (Synapto), and RHD4 were aligned using ClustalW. Identical amino acids are shaded black, and similar amino acids are shaded gray. The seven conserved motifs of the SAC domain (Hughes et al., 2000) are marked by solid lines above the sequences. The location of the rhd4-2 G449S amino acid substitution in motif VII is marked by an asterisk.

Figure 3.

RHD4 Alleles and Complementation. (A) RT-PCR of RHD4 expression in wild-type seedlings and in mutant seedlings containing each of the different rhd4 alleles. RNA was extracted from seedlings for cDNA synthesis and RT-PCR. In the rhd4-1 sample, a larger band was observed (asterisk), and this band corresponds to the RHD4 transcript including the first intron due to the G→A base pair substitution, which prevents proper splicing. β-tubulin is used as an internal loading control. (B) Phenotypes caused by the rhd4 mutant alleles. Bars = 100 μm. (C) Complementation of the rhd4-1 mutation. rhd4-1 mutant Arabidopsis plants were transformed by Agrobacterium-mediated transformation with an At3g51460 native promoter-At3g51460 cDNA construct (left panel) and a native promoter-EYFP-cDNA construct observed by light microscopy (middle panel) and fluorescence microscopy (right panel). Bars = 100 μm. (D) Quantification of root hair length. Root hairs of wild-type, rhd4 alleles, and complemented seedlings were measured (n = 314, 311, 272, 268, 308, 464, and 398 root hairs, respectively). Error bars represent 1 sd.

Figure 4.

RHD4 Is a PI(4)P Phosphatase. (A) GST-RHD4 has high levels of PI(4)P phosphatase activity in vitro. GST, GST-RHD4, and GST-rhd4-2 were expressed and purified from E. coli. Error bars represent the sd of two biological replicates. (B) GST-RHD4 has low levels of activity against all other phosphoinositide substrates tested. Values are calculated with PI(4)P activity as 100%. For each substrate, phosphatase release was measured for GST and GST-RHD4; the GST values were then subtracted from GST-RHD4 values to eliminate background. Error bars represent the sd of two replicates. (C) rhd4-1 roots accumulate PI(4)P. Two-week-old wild-type and rhd4-1 seedlings were transferred to 0.25× Murashige and Skoog (MS) media supplemented with ³H-myoinositol for 40 h. Radiolabeled phospholipids were isolated from root tissue, deacylated, and fractionated with HPLC. PI(4)P levels were calculated as a percentage of total PI. Error bars represent the sd from the average of nine samples. (D) Representative traces from the HPLC of ³H-labeled phospholipids from rhd4-1 and wild-type roots. The glycerophosphoinositol peaks (gPI) for gPI(3)P and gPI(4)P are shown. rhd4-1 (black line) has elevated levels of PI(4)P compared with the wild type (gray line), while PI(3)P levels are comparable between the wild type and rhd4-1.

Figure 5.

EYFP-FAPP1 Is Localized to Subcellular Membrane Compartments in rhd4-1 Root Hairs. (A) GST, GST-2XFYVE, and GST-FAPP1 lipid binding assays. (B) GFP-FYVE, a PI(3)P binding protein, is localized to the endosomal compartment in wild-type root hairs. A z-stack projection is shown. Bar = 10 μm. (C) GFP-FYVE localization in the rhd4-1 background. A z-stack projection is shown. Bar = 10 μm. (D) EYFP-FAPP1 is localized to the tips of growing wild-type root hairs. Arrowheads show localization to some subcellular membrane compartments. A z-stack projection of a wild-type root hair is shown. Bar = 10 μm. (E) EYFP-FAPP1 is enriched on the plasma membrane of wild-type root hairs. Cross sections were made through the z-stack projection at various positions in the tip region as represented by solid white lines. Cross sections were then used to quantify the amounts of EYFP-FAPP1 along a line, represented by the dashed white line. (F) Quantification of EYFP-FAPP in wild-type cross sections. ImageJ was used to quantify the pixel intensity of EYFP-FAPP1 along a vertical line through the center of each cross section (dashed white line in [E]). Each pixel intensity plot corresponds to the cross section above it in (E). (G) EYFP-FAPP1 is accumulated on subcellular membrane compartments in rhd4-1 root hairs. A z-stack projection of a wild-type root hair is shown. Bar = 10 μm. (H) EYFP-FAPP1 is found on subcellular membrane compartments throughout rhd4-1 root hairs (arrowheads). Cross sections were made through the z-stack projection at various positions in the tip region as represented by solid white lines. Cross sections were then used to quantify the amounts of EYFP-FAPP1 along a line, represented by the dashed white line. (I) Quantification of EYFP-FAPP1 in rhd4-1 cross sections. ImageJ was used to quantify the pixel intensity of EYFP-FAPP1 along a vertical line through the center of each cross section (dashed white line in [H]). Each pixel intensity plot corresponds to the cross section above it in (H).

Figure 6.

EYFP-FAPP1 Localization in Wild-Type and rhd4-1 Root Hairs. Arabidopsis seedlings expressing EYFP-FAPP1 were imaged by time-lapse fluorescence microscopy. Beginning, middle, and end frames are shown. Bars = 10 μm. (A) Wild-type root hair. These frames are from Supplemental Movie 3 online. (B) rhd4-1 root hair. These frames are from Supplemental Movie 4 online.

Figure 7.

GFP-FABD2 Localization in rhd4-1 Root Hairs and Epidermal Cells. Arabidopsis seedlings expressing GFP-FABD2 (for localization of F-actin filament bundles) were imaged by confocal microscopy, and z-stack projections are shown. Bars = 20 μm. (A) and (B) Localization in expanding wild-type root hairs. F-actin filament bundles are found throughout the root hair. (C) and (D) Localization in expanding rhd4-1 root hairs. F-actin filaments do not appear to be disturbed. (E) Localization in wild-type root epidermal cells. A meshwork of F-actin is observed. (F) Localization in rhd4-1 root epidermal cells. A meshwork of F-actin is observed similar to the wild type; however, more actin patches are observed.

Figure 8.

EYFP-RHD4 Is Localized to the Tips of Growing Root Hairs. (A) to (D) Arabidopsis seedlings expressing various YFP fusion proteins were imaged by confocal microscopy. A medial fluorescence image, the corresponding differential interference contrast (DIC) image, and an overlay of these images are shown (columns, left to right). Bars = 10 μm. (A) EYFP-RHD4 was expressed by its native promoter in rhd4-1 Arabidopsis plants and is localized to the tips of growing root hairs. (B) EYFP-RabA4b is localized to a post-Golgi secretory compartment at the tips of growing root hairs. (C) G-yk (a YFP-Golgi maker) is localized to the Golgi compartment throughout growing root hairs. (D) ER-yk (a YFP-ER marker) is localized to the ER throughout growing root hairs. (E) to (G) Arabidopsis seedlings expressing YFP fusion proteins were imaged by confocal microscopy. Inset: higher magnification. Z-stack projections are shown. Bars = 20 μm. (E) EYFP-RabA4b localization in root epidermal cells. (F) G-yk localization in root epidermal cells. (G) ER-yk localization in root epidermal cells. (H) to (J) Arabidopsis seedlings expressing EYFP-RHD4 were fixed, processed for immunolocalization of anti-RabA4b antibody, and analyzed by confocal microscopy. Medial sections are shown. Bars = 10 μm. (H) and (I) RabA4b colocalizes with EYFP-RHD4 at the tips of root hairs. EYFP-RHD4 (green) and anti-RabA4b (red) are localized to overlapping tip-localized membrane compartments (yellow). (J) Tip localization of anti-RabA4b is specific, as no tip-localized fluorescence was observed when anti-RabA4b primary antibody was left out.