Rho-of-plant activated root hair formation requires Arabidopsis YIP4a/b gene function

ABSTRACT Root hairs are protrusions from root epidermal cells with crucial roles in plant soil interactions. Although much is known about patterning, polarity and tip growth of root hairs, contributions of membrane trafficking to hair initiation remain poorly understood. Here, we demonstrate that the trans-Golgi network-localized YPT-INTERACTING PROTEIN 4a and YPT-INTERACTING PROTEIN 4b (YIP4a/b) contribute to activation and plasma membrane accumulation of Rho-of-plant (ROP) small GTPases during hair initiation, identifying YIP4a/b as central trafficking components in ROP-dependent root hair formation.

Detection of ROP in SYP61-positive vesicles. SYP61 vesicles were isolated employing the immunopurification procedure described previously (Wattelet-Boyer et al., 2016). In brief, Arabidopsis thaliana seedlings were grown in liquid culture for eight days and then ground in vesicle extraction buffer (HEPES 50 mM pH 7.5, 0.45 M sucrose, 5 mM MgCl2, 1 mM DTT, 0.5% (w/v) PVP (Sigma), 1 mM PMSF). Intact pools of light vesicles were collected at the 33/8% interface of a 38/33/8% sucrose step gradient after overnight centrifugation at 150 000 g at 4°C. This membrane fraction was then resuspended in the resuspension buffer (50 mM HEPES pH 7.4, 0.25 M sucrose, 1.5 mM MgCl2, 150 mM NaCl, 1 mM PMSF, protease inhibitor cocktail) and the total membrane fraction was used as input for immuno-precipitation (IP). IP was performed with magnetic Dynabeads coupled to protein A (Invitrogen) according to the manufacturer's instructions. For each IP, 75 µl of beads were incubated with 7 µl of rabbit anti-GFP antibodies (Invitrogen) for 1 h with shaking at 4°C, washed with PBS-Tween (137 nM NaCl, 2.7 nM KCl, 10 nM Na2HPO4, 1.8 nM KH2PO4,, equilibrated in the resuspension buffer for 10 min on ice and incubated with 1 ml of purified total membrane extract for 1 h with shaking at 4°C. Several washes were performed with 1 ml of resuspension buffer for 5 min with shaking at 4°C. Step gradientpurified total membrane fractions (IP input) and bead-immuno-purified fractions (IP output) were loaded at equal quantity on an SDS-PAGE gel and subjected to western-blotting. Purified rabbit anti-ROP antibody at 1:200 dilution was used to detect ROPs as described, previously (Kiefer et al., 2015).
Label-free proteomics was used to identify ROP2, ROP4 and ROP6 protein peptides in the SYP61 IP fraction. In brief, IP samples were treated with 25 µl 1% (w/v) SDS for 30 min at 37°C, 0.3 µl DTT 2 M was then added with subsequent incubation for 30 min at 37°C, 2.3 µl iodoacetamide 1 M was added followed by 30 min incubation at 37°C and finally 7 µl 5x Laemmli loading buffer was added followed by incubation for 30 min at 37°C. Samples were loaded and subjected to SDS-PAGE electrophoresis, extracted from the gel and injected into an LC-MS/MS Q-Exactive with a gradient time of 120 min. Labelfree quantitative data analysis was performed on raw LC-MS/MS data imported in Progenesis QI for Proteomics 2.0 (Nonlinear Dynamics Ltd). Data was processed by volume integration for 2-6 chargestate ions and calculation of protein abundance (sum of the volume of corresponding peptides).
Quantitative data was considered for proteins quantified by a minimum of two peptides. Development: doi:10.1242/dev.168559: Supplementary information    For ROP2 activity assays in WT, yip4a yip4b and spk1-4, total proteins were extracted from five-day-old seedlings grown on 1/2x MS agar medium. Twenty micrograms of MBP-RIC1-conjugated agarose beads were added to the protein extracts and incubated at 4°C for 2 h on a rocking table. The beads were washed four times in wash buffer (25 mM HEPES, pH 7.4, 1 mM EDTA, 5 mM MgCl2, 1 mM dithiothreitol, and 0.5% (v/v) Triton X-100) at 4°C. GTP-bound ROP proteins that were associated with the MBP-RIC1 beads were boiled and used for analysis by western blotting with a ROP2-specific antibody. ROP2

Development • Supplementary information
polyclonal antibodies was generated against the peptide QFFIDHPGAVPITTNQG (Abicode). Prior to the pull-down assay, a fraction of total proteins was analyzed by immunoblot assay to determine total ROP2 (GDP bound and GTP bound, blot on the left). The amount of the GTP-bound active form of ROP2 (blot on the right) was normalized to that of total ROP2.  (A) Wild-type roots displaying a functional secretory system contain numerous ROP patches (red) at incipient hair initiation sites which persist at the tip of the growing hair. (B) When YIP4s (and their complex partner ECHIDNA) are lacking, the secretion of some cargos en route to the PM is impaired (Boutte et al., 2013;Gendre et al., 2013), likely due to a defect in secretory vesicle formation (Boutte et al., 2013).
Mutants then lack hairs and show a drastic reduction in number and intensity of ROP patches. (C) Type-I ROP (ROP1-8) proteins synthesized at free polysomes in the cytosol require prenylation and the posttranslational modification of the C-termini at the surface of the endoplasmic reticulum (ER) for their membrane anchoring (Wright and Philips, 2006;Bracha-Drori et al., 2008;Feiguelman et al., 2018).
Furthermore, the presence of ROP-GEF SPK1 both at the ER exit site (Zhang et al., 2010) and on punctae associated with the PM (Yanagisawa et al., 2018) reinforces the importance of the ER. Studies in yeast and mammals suggest that that Rho/Rac/Cdc42 could exit directly from the ER via a complex with RhoGDI after having been extracted from the organelle membrane (route 2, fast) or via vesicles (route 1, slow) (Slaughter et al., 2009;Garcia-Mata et al., 2011;Watson et al., 2014). By analogy, ROP may reach the PM by two different means in plants. ROPs are found in SYP61-vesicles, EYFP-ROP2 recovery after photobleaching is impaired at the plasma membrane and intensity of ROP patches is decreased in the secretion-defective yip4a yip4b mutant, suggesting that a part of ROP is taking the secretory pathway on its way to the PM. It remains open whether secretion of other factors crucial to Development: doi:10.1242/dev.168559: Supplementary information activate ROP at the PM, like ROP-GEF, are also impaired in yip4a yip4b resulting in the decrease in ROP patches. In addition, tight regulation of the ROP activation/inactivation cycle keeps ROP polarized. Data from yeast indicates that the initial polar aggregation of Cdc42 seems highly dependent on the GTPase cycles and the cytosolic pool of GTPase and only partially dependent of secretion via actin filaments (Wedlich-Soldner et al., 2004). Hence, a contribution by both the secretory pathway and the cytosolic ROP pool is feasible.