SYP22 and VAMP727 regulate BRI1 plasma membrane targeting to control plant growth in Arabidopsis

Brassinosteroids (BRs) are a group of bioactive steroid molecules that act as plant hormones. BR signaling is monitored by the plasma membrane (PM)-bound receptor BRASSINOSTEROID INSENSITIVE1 (BRI1), a leucine-rich repeat (LRR) receptor-like kinase (RLK) (Belkhadir & Jaillais, 2015). Binding of BRs to the extracellular domain of BRI1 initiates an intracellular signal cascade response which is rapidly transmitted to the nucleus ultimately resulting in transcription of BR-responsive genes (Jaillais & Vert, 2016). BR responsive genes ultimately help to control various aspects of plant growth and development, including root development, cell division, vascular patterning, organ boundary development, stress response and pathogen defense (Nakashita et al., 2003; Yamamoto et al., 2007; Gonzalez-Garcia et al., 2011; Gendron et al., 2012; Bekh-Ochir et al., 2013). Signaling of BR is initiated at the PM and expression of BRI1 is critical to controlling a number of processes. Previously, it has been demonstrated that BRI1 undergoes constitutive endocytosis and cycles between the PM and trans-Golgi network/early endosome (TGN/EE) in a ligand-independent manner (Geldner et al., 2007). Disruption of clathrin adaptor protein complex-2 (AP-2) function results in decreased uptake of a fluorescently labeled ligand, decreased endosomal pool of BRI1, and consequently enhanced the BR responses, indicating that clathrin-mediated endocytosis (CME) is essential for BRI1 internalization (Irani et al., 2012; Di Rubbo et al., 2013). In addition, the plant-specific TPLATE adaptor complex (TPC), TML protein, mediates BRI1 endocytosis (Gadeyne et al., 2014). Furthermore, high-resolution imaging has demonstrated that flotillin1 (Flot1)-enriched membrane microdomains mediate cathrin-independent endocytosis (CIE) and act as an additional regulator of BRI1 translocation from the PM (Wang et al., 2015). However, the mechanism of coordinated regulation of BRI1 by CME and CIE requires further elucidation. Endocytosed BRI1 is translocated to vacuoles for degradation through a process which is regulated by ubiquitination or ESCRT sorting machinery into intraluminal vesicles (ILVs) of late endosome/multivesicular body (LE/MVB) (Cardona-Lopez et al., 2015; Zhou et al., 2018). Recently, it has been demonstrated that expression of the VHA-C/DEETIOLATED3 (DET3) gene is related to the acidification and motility of TGN/EE, which aides secretion and recycling of BRI1 to PM in addition to maintaining the BRI1 PM pool (Luo et al., 2015). However, the molecular mechanisms of BRI1 regulation on the PM are not fully understood. Soluble N-ethylmaleimide sensitive factor attachment protein receptors (SNAREs) control shuttling of cargo proteins to their final destinations and are highly conserved in eukaryotic cells (Cai et al., 2007). By use of conserved amino acids of SNARE motifs with coiled-coil helices, SNARE proteins can be classified into Qa-, Qb-, Qc-, SNAP25-like (containing Qb and Qc tandem SNARE motifs), and R-SNAREs. A functional SNARE complex consists of one set of Qa-, Qb-, Qc- (localized to target membrane) and R-SNARE (localized to vesicle membrane) with a four-helical bundle assembled by four SNARE motifs (Fasshauer et al., 1998; Hong, 2005). Fifteen R-SNARE members, including two Sec22-like, two Ykt6-like and seven VAMPs (vesicle-associated membrane proteins) have been identified in Arabidopsis (Lipka et al., 2007). The VAMPs in higher plants can be classified into two major groups: VAMP71 and VAMP72 groups. The VAMP72 group is likely specific to the green plant lineage and represents R-SNARE components involved in secretion (Sanderfoot, 2007). A previous membrane interactome study showed that VAMP727 and SYP22 interact with BRI1 in yeast (Jones et al., 2014). VAMP727 has previously been demonstrated to be localized to the PM and endosomes, whereas SYP22 is localized in early endosomes and tonoplast and is involved in the regulation of protein storage (Ebine et al., 2008, 2011). However, the role of VAMP727 and SYP22 in BR signaling has not been reported. In the presented work we demonstrate that Q-SNARE SYP22 and R-SNARE VAMP727 mediate the BRI trafficking to PM. SYP22 and VAMP727 physically interact with BRI1, and syp22/vamp727−/+ or SYP22ND mutants exhibit dwarfism and insensitivity to BR treatment. Furthermore, we investigate SYP22- and VAMP727-regulated BRI1 intracellular trafficking by use of genetic, biochemical, and cellular bioassays. Results of this study provide useful information for studying BR receptor BRI1 PM targeting and will contribute to further investigations of BR signaling in plants. The SYP22 ORF and truncated form of SYP22 (SYP22ND) was recombined into the modified pEarlygate104 GW binary destination vector via a Gateway LR reaction, and Agrobacterium-mediated transformation into Arabidopsis thaliana Col-0 was used to generate SYP22ND overexpressing transgenic plants. The primers used to clone SYP22 and SYP22ND regions are listed in Supporting Information Table S1. Surface sterilized seeds were sown on half-strength Murashige and Skoog (MS) medium with 1.2% agar (MS medium) or MS medium supplied with BRs (Sigma, St Louis, MO, USA). After 2 d of stratification at 4°C, seedlings were incubated at 22°C in a 24 h low light chamber (~80 μE m−2) for 7 d. For BL-dependent hypocotyl growth assay, Col-0, SYP22ND #3, and #5 seeds were plated on 0.5× MS media containing 0, 10, 100 and 200 nM BL and hypocotyl length was measured after 7 d. Seedling plates were photographed and hypocotyl lengths were measured using ImageJ software. The N-proximal half of yellow fluorescent protein (nYFP) and C-proximal half of cyan fluorescent protein (cCFP) sequences were fused to the N-terminal sequences of SYP22 or SYP22ND and N-terminal sequences of VAMP727 or C-terminal sequences of BRI1 in PNXGW and PCXGW or PXCGW vectors, respectively. For transient expression, all binary constructs were introduced into Agrobacterium tumefaciens strains (GV3101). Agrobacterium cells containing split YFP fusion constructs were grown in liquid yeast extract peptone (YEP) medium supplemented with antibiotics (spectinomycin 50 μg ml−1 and Rifampicin 50 μg ml−1). Cultured cells were centrifuged at 2500 g for 5 min at room temperature, and then the cell pellet was resuspended in infiltration buffer (10 mM MES, pH 5.6, 10 mM MgCl2, 200 μM acetosyringone) after removing supernatant. Cell density was adjusted with infiltration buffer to give an OD600 of c. 0.5–0.6. Aliquots (0.5 ml) of Agrobacterium cells carrying a split YFP fusion constructs were mixed, and then a syringe was used to infiltrate the mixture into the lower surface of Nicotiana benthamiana leaves. Plants were incubated in a growth chamber for 36 to 48 h (Kim J. G. et al., 2009). Total cellular RNA was isolated with TRIzol (Takara, Dalian, Liaoning, China) and subsequently treated with DNase (Promega, Madison, WI, USA) to eliminate genomic DNA contamination. For cDNA synthesis, GoScriopt™ Reverse Transcription kit was used following the manufacture's instructions (Promega, Madison, WI, USA). Quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis used gene specific primers for BR6OX2, CPD, DWF4 and ACS5 as described (Oh et al., 2012). Actin8 was used as control (Chen et al., 2010). All primers including gene specific primers for SYP22 are listed in Table S1. The laser-scanning confocal images were acquired with Leica TCS SP5 inverted microscope fitted with a ×60 water-immersion objective. For brefeldin A (BFA) treatment, four images of root epidermal cells with a 2-μm Z axis distance were taken. Seedlings expressing BRI1-GFP (green fluorescent protein) mounted in half-strength MS liquid were excited with a 488 nm laser line and detected between 500 and 550 nm. For intensity quantification, fixed regions of interest (ROIs) were selected to measure the fluorescent intensity of both PM and intracellular space with ImageJ (version 1.48). The relative PM fluorescence level was calculated by dividing the PM intensity by the intracellular intensity. For fluorescence detection in tobacco leaves, YFP was excited with a 515-nm laser line and detected between 530 and 600 nm. mCherry-tagged proteins were excited with a 561-nm laser and detected with a 560- to 640-nm emission filter. The images were edited using Adobe Photoshop CS2. Image acquisition from transgenic plants was performed with a variable-angle total internal reflection fluorescence microscope (VA-TIRFM) based on an inverted microscope (Olympus IX-71) equipped with a TIRF illumination module (Olympus IX2-RFAEVA-2) and a ×100 oil-immersion objective (Olympus, numerical aperture = 1.45). GFP-labeled fusion proteins were excited with 488-nm laser line, from a diode laser (Changchun New Industries Optoelectronics Technology, China) and their emission fluorescence was obtained with a filter BA510IF (525/50). Time-lapse series images of single particles of BRI1-GFP were taken with up to 100 images per sequence and acquired with 100-ms exposure time. For the analysis of protein intensity, the first 10 frames of one serial image were used to measure the number of spots (n1) and the range of ROI (the number of pixels, n2) using Matlab software. The intensity was calculated using the following formula: Intensity = n1/(n2 × 0.162) (Lv et al., 2017). For mating-based split ubiquitin assays, VAM727 and SYP22 were cloned into the mating-based split ubiquitin Nub vector pXN25_GW while BRI1 was cloned into Cub vector pMETYC_GW. Assays were performed as described in a previous study (Lalonde et al., 2010). For transient expression of mCherry-SYP22 and mCherry-SYP22ND fusion proteins, SYP22 ORF and SYP22ND were cloned into the modified pEarly-GW 104 destination plasmid, followed by transient expression in N. benthamiana leaves by using the Agrobacterium-mediated transient expression method (Kim J. G. et al., 2009). The VAMP727-Myc or SYP22-Myc together with BRI1-GFP were coexpressed into N. benthamiana leaves by using the Agrobacterium-mediated transient expression, and protein extraction and further co-immunoprecipitation (co-IP) assays were performed as described previously (Kim T. W. et al., 2009). The microsomal fraction was prepared by centrifugation at 20 000 g and 100 000 g for 1 h. The resulting pellet was resuspended in extraction buffer containing 1% Triton X-100. After centrifugation at 100 000 g for 10 min, solubilized proteins were incubated with anti-GFP antibody or IgG for 2 h, and further bound with protein A beads for 2 h. The beads were washed four times with extraction buffer containing 0.1% Triton X-100 and eluted with sodium dodecyl sulfate (SDS) sample buffer. To test VAMP727-Myc, SYP22-Myc, and BRI1-GFP levels in co-IP experiments, the proteins from whole cell lysate were analyzed. To analyze BRI1 levels, whole cell lysate, microsomal fraction, and PM fraction (Takahashi et al., 2014) were performed using BRI1-GFP and BRI1-GFP/SYP22ND #5 seedlings. From each sample, 20 μg of protein was separated on SDS-PAGE and electrotransferred onto Immobilon-P Transfer Membranes (Millipore, Tokyo, Japan). The membranes were incubated in TBS containing 5% skim milk and 0.05% Tween-20 for 60 min and blotted with primary antibodies at 4°C for 2 h. An anti-Myc antibody (1 : 2000; Abcam, Cambridge, MA, USA) and anti-GFP antibody (1 : 2000; Abcam) were used as primary antibodies. The membranes were incubated for 1 h with an anti-mouse or anti-rabbit horseradish peroxidase (HRP)-conjugated secondary antibody (1 : 2000; Cell Signaling Technology, Danvers, MA, USA). Reaction products were visualized by detecting chemiluminescence by using an ECL Western Blotting Detection System (GE Healthcare, Piscataway, NJ, USA). Band density was analyzed by ImageJ (v.1.48). Statistical calculations were performed using Prism 5 (GraphPad, San Diego, CA, USA). Comparisons between two groups were performed with student t-test, while multi groups comparisons were performed using one-way analysis of variance (ANOVA), followed by Bonferroni's multiple comparison test. The P values of < 0.05 were considered statistically significant. The R-SNARE VAMP727 and the Qa-SNARE SYP22 have been shown to interact with BRI1 in yeast and tobacco, respectively (Jones et al., 2014). To confirm VAMP727 and SYP22 interaction with BRI1, we performed split-ubiquitin yeast two hybrid and co-IP assays. The results indicated that VAMP727 and SYP22 can interact with BRI1 in yeast (Fig. 1a). Proteins from the microsomal fraction were immunoprecipitated using either IgG or GFP antisera, and the immunoprecipitated proteins were detected using an anti-Myc antibody. The results indicated that BRI1 interacts with VAMP727 and SYP22 in plants (Fig. 1b). A previous study has demonstrated that expression of the cytosolic region of Q-SNARE member SYP121 acts as a negative dominant mutant to disrupt its function (Tyrrell et al., 2007). In the present study, the negative dominant form of SYP22 (SYP22ND) was localized in the cytosol, whereas SYP22 was localized in the cytosol and vesicle-like compartments in tobacco leaves (Fig. S1a). Further, overexpression of SYP22ND severely inhibited plant growth (Fig. S1b), and the morphology of a line showing severe growth inhibition (SYP22ND #5) was similar to that of syp22/vamp727−/+ double-mutant plants (Fig. S1c). The expression levels of SYP22ND were correlated with a plant dwarf phenotype (Fig. S1d). Furthermore, unlike vamp727 and syp22 single mutants, syp22/vamp727−/+ mutant plants were insensitive to exogenously applied BR (Jones et al., 2014). A BR-dependent hypocotyl elongation assay showed that, compared with wild-type plants, SYP22ND #3, #5, and bri1-5 seedlings are less sensitive to exogenously applied BR (Fig. 1c–e). Given that VAMP727 and SYP22 interact with BRI1, and that syp22/vamp727−/+ and SYP22ND mutants are less sensitive to BR treatment, we examined the expression of genes related to BR signaling in bri1-5, syp22/vamp727−/+, and SYP22ND #5 and the corresponding wild-type plants. We found that three BR biosynthesis genes, BR6OX2, CPD, and DWF4, were more highly expressed in bri1-5, syp22/vamp727−/+, and SYP22ND #5 than in the wild-type plants. By contrast, the expression of ACS5 was highly suppressed in bri1-5, syp22/vamp727−/+, and SYP22ND #5 compared with the wild-type plants (Fig. S2). To investigate whether growth inhibition in the SYP22ND mutant is caused by the suppression of BR signaling, SYP22ND #5 plants were crossed with bzr1-D plants that constitutively activate BR signaling and suppress BR-deficient and BRI1 mutant phenotypes (Wang et al., 2002). We found that bzr1-D partially rescued the SYP22ND #5 phenotype characterized by an enlarged size of plants (Fig. 1f), and the progeny developed a relatively longer petiole than SYP22ND #5 plants (Fig. 1g). In addition, BR6OX2 and CPD levels were lower in bzr1-D/SYP22ND #5 than in SYP22ND #5 (Fig. 1h). To examine how SYP22ND affects BRI1 function, we performed a split-GFP assay to analyze the interaction between SYP22ND and VAMP727 or BRI1. The interaction between SYP22 and VAMP727 was detected in the cytosol and the vesicle-like compartments. However, SYP22ND interacted only with VAMP727 in the vesicle-like compartments and showed no interaction with BRI1 (Fig. S3). Further, BRI1 levels in the PM were investigated using pBRI1-BRI1-GFP and SYP22ND #5 crossed plants using VA-TIRFM (Wang et al., 2015). The results indicated that the density of BRI1-GFP in the PM of the root cells of wild-type and SYP22ND #5 plants was c. 0.64 ± 0.05 and 0.33 ± 0.06 particles per μm2, respectively (Fig. 1i, j), indicating that BRI1 levels in the PM were reduced by SYP22ND expression. In parallel, a biochemical study was performed to analyze BRI1 levels in the whole-cell lysate (WCL), microsomal fraction (MSF), and plasma membrane fraction (PMF) obtained from the BRI1-GFP and BRI1-GFP/SYP22ND #5 plants. Western blot analysis data showed that BRI1 levels in the WCL and MSF were similar in BRI1-GFP and BRI1-GFP/SYP22ND #5 plants. However, the level of BRI1 was ~two-fold lower in the PMF of BRI1-GFP/SYP22ND #5 than in control plants (Fig. 1k, l). To further assess whether BRI1 recycling to the PM is affected in SYP22ND #5, we performed washout experiments after treatment with 50 μM BFA for 60 min in BRI1-GFP and BRI1-GFP/SYP22 ND #5 roots (Fig. 1m). The percentage of cells containing BFA bodies (Fig. 1n), BFA body size (Fig. 1o), and relative PM fluorescent level (Fig. 1p) were determined at indicated time points after washing. Compared with the controls, a higher number of BFA body-containing cells, larger BFA body size, and lower relative PM fluorescent levels were recorded in BRI1-GFP/SYP22ND #5 plants subjected to BFA treatment, suggesting that more BRI1 is retained in the cytosol of SYP22ND #5 root cells. Following inhibitor removal, more cells containing BFA compartments with larger size were still observed in the roots of BRI1-GFP/SYP22ND #5 plants than in BRI1-GFP plants. Moreover, there was a significantly delayed recovery of the PM BRI1 pool in the SYP22ND #5 line, although relocation of BRI1 to the PM from the BFA body was detected after BFA washout. These data indicate that BRI1 PM targeting and recycling were affected in SYP22ND lines. Recently it has been demonstrated that SNARE molecules play important roles in intracellular trafficking and therefore are essential for the regulation of plant membrane protein activity and signal transduction (Hachez et al., 2014; Zhang et al., 2015). BR signaling helps to control plant growth and development and is modulated by activity of the BRI1 receptor at the PM (Yamagami et al., 2017). Thus, revealing the mechanisms which control BRI1 activity at the PM is important to investigate BR signaling. Our understanding of endocytic regulation of BRI1 has improved recently, however, there lacks an understanding of how BRI1 cycling to the PM is controlled. In this study, we combined biochemical, genetic and bioimaging techniques to investigate the function of SYP22 and VAMP727 SNARE proteins in the PM targeting of BRI1 and BR-related plant growth. In a previous study by our group, the interaction of VAMP727 and SYP22 with BRI1 on the PM was investigated by use of a split-GFP assay (Jones et al., 2014). In the current study, we further confirm that the SNAREs VAMP727 and SYP22 interact with BRI1 in vivo as demonstrated by results of the Co-IP and yeast assays. Although SNARE proteins play an important role in the movement of membrane proteins within the cell, few specific cargo proteins for SNARE-mediated vesicle transport have been identified in plants. Fluorescence imaging indicated that the powdery mildew 8.2 protein (RPW8.2) is delivered to the plant–fungal interface via VAMP721 and VAMP722 vesicles (Kim et al., 2014). In addition, a recent study demonstrated by use of proteomic analysis that a number of cargo proteins are differentially mediated by SYP121 and SYP122 (Waghmare et al., 2018). Therefore, results of the current study suggest that the BRI1 receptor is likely the cargo protein for vesicles regulated by VAMP727 and SYP22. Previously, overexpression of the cytosolic region of Q-SNARE has been used as an approach to investigate the functions of specific Q-SNARE in plant growth and protein trafficking (Geelen et al., 2002; Besserer et al., 2012). Results of our study showed that overexpression of truncated SYP22 (SYP22ND) inhibited plant growth in an abundance-dependent manner. Specifically, SYP22ND #5 phenotypes were similar to that of syp22/vamp727−/+ mutants and exhibited insensitivity to BR treatment, thus suggesting that SYP22 and VAMP727 might regulate BR signaling via interaction with BRI1. Gene expression analysis revealed that three BR biosynthesis genes were highly upregulated in SYP22ND #5, consistent with the expression in bri1-5 and syp22/vamp727−/+ when compared to with wild-type plants. Conversely, expression of BR related genes suggested signaling downstream was highly suppressed. In addition, expression of bzr1-D in SYP22ND #5 plants partially rescued the BR-defective phenotypes, suggesting that SYP22ND #5 affects signal transduction via regulation of BRI1. Therefore, these results might suggest that SYP22 and VAMP727 are involved in the BR signaling via regulation of BRI1 trafficking. We further provide evidence that show how SYP22ND affects BRI1 function. mCherry-tagged SYP22ND was detected in the cytosol and was different from the pattern observed with SYP22. In addition, mCherry-tagged SYP22ND interacted solely with VAMP727 and was restricted to vesicle-like compartments, suggesting that SYP22ND might affect localization of SYP22-VAMP727 complexes, thus inhibiting the intracellular trafficking of BRI1. Furthermore, overexpression of SYP22ND reduced the abundance and density of BRI1 on the PM, suggesting that delivery of BRI1 to PM was impacted. Particularly, BFA treatment revealed recycling defects of BRI1 to the PM in SYP22ND #5, which might account for reduced pool of BRI1 on the PM in SYP22ND mutants. Previous work suggested that PVC-localized VAMP727 at the PM was regulated by ARA6 functioning independently of known secretory pathways (Ebine et al., 2011). Moreover, SYP22 immunoprecipitated with the secretory component of VAMP721 in Arabidopsis despite their different subcellular localizations (Fujiwara et al., 2014), thus suggesting that a SNARE can have differential localization and trafficking, which is related to specific physiological functions. In this scenario, we propose that SNARE complexes containing VAMP727 and SYP22 regulate an unknown secretory pathway for BRI1. Therefore, reduced abundances of BRI1 on the PM in SYP22ND plants is likely the result of altered BRI1 trafficking to the PM, but not the degradation in vacuoles, as similar abundances of BRI1 in whole-cell lysates and microsomal fractions were observed in control and SYP22ND plants. Overall, results of the current study provide insights into the mechanisms by which PM associated BRI1 regulate BR signaling in plant growth. This work was supported by the grants from the National Natural Science Foundation of China (31770197, 31571985 and 31330063), and the Support Plan for Innovative Talents in Colleges and Universities of Liaoning Province (LR2017037). The authors thank Prof. Ueda Takashi (Tokyo University) for providing the mutant seeds of vamp727, spy22, and syp22/vamp727−/+ and Prof. Zhi-yong Wang (Carnegie Institution for Science, Washington, DC, USA) for providing bri1-5, bzr1-D and BRI1-GFP seeds. All authors declare no conflict of interests. LZ and YHX planned and designed the research. LZ, YL and XFZ performed most of experiments. TYL, JHJ, QS, LJC and YXD analyzed data. LZ and YHX wrote the manuscript. LZ, YL and XFZ contributed to this are not for the or of Supporting Information supplied by the than be to the New expression of SYP22 negative dominant overexpression plants. of BR biosynthesis and signaling genes in bri1-5, Col-0, syp22/vamp727−/+, and SYP22ND #5 seedlings. between SYP22ND and VAMP727 or BRI1. Table sequences used in this The is not for the or of information supplied by the than be to the corresponding for the