RAB GTPases are key regulators of membrane traffic. Among them, RAB11, a widely conserved sub-group, has evolved in a unique way in plants; plant RAB11 members show notable diversity, whereas yeast and animals have only a few RAB11 members. Fifty-seven RAB GTPases are encoded in the Arabidopsis thaliana genome, 26 of which are classified in the RAB11 group (further divided into RABA1–RABA6 sub-groups). Although several plant RAB11 members have been shown to play pivotal roles in plant-unique developmental processes, including cytokinesis and tip growth, molecular and physiological functions of the majority of RAB11 members remain unknown. To reveal precise functions of plant RAB11, we investigated the subcellular localization and dynamics of the largest sub-group of Arabidopsis RAB11, RABA1, which has nine members. RABA1 members reside on mobile punctate structures adjacent to the trans-Golgi network and co-localized with VAMP721/722, R-SNARE proteins that operate in the secretory pathway. In addition, the constitutive-active mutant of RABA1b, RABA1bQ72L , was present on the plasma membrane. The RABA1b -containing membrane structures showed actin-dependent dynamic motion . Vesicles labeled by GFP–RABA1b moved dynamically, forming queues along actin filaments. Interestingly, Arabidopsis plants whose four major RABA1 members were knocked out, and those expressing the dominant-negative mutant of RABA1B, exhibited hypersensitivity to salinity stress. Altogether, these results indicate that RABA1 members mediate transport between the trans-Golgi network and the plasma membrane, and are required for salinity stress tolerance.
Membrane traffic is essential for a variety of cellular phenomena from housekeeping to development, and is coordinately regulated by elaborate molecular machineries. RAB GTPases, molecular switches that act through the conformational change between the GTP- and GDP-bound forms, control docking, fusion and, in some cases, fission events during vesicle trafficking (Miserey-Lenkei et al., 2010; Valente et al., 2010). The RAB family has a large number of members: for example, 60 in Homo sapiens and 57 in Arabidopsis thaliana. In order to maintain proper traffic, each RAB member is considered to regulate a specific step in the complicated network of membrane traffic.
In plants, RAB members consist of eight groups: RABA, RABB, RABC, RABD, RABE, RABF, RABG and RABH, which correspond to animal RAB11, RAB2, RAB18, RAB1, RAB8, RAB5, RAB7 and RAB6, respectively (Rutherford and Moore, 2002). Despite of the lack of several RAB groups conserved in animals, some plant RAB proteins have undergone unique diversification. In particular, the RABA (RAB11) group exhibits remarkable expansion in higher plants. In A. thaliana, the RABA group comprises as many as 26 members out of the total of 57 RABs. However, animals and yeast possess only a few corresponding members. In non-polarized mammalian cells, RAB11a localizes to the membrane of recycling endosomes (Ullrich et al., 1996; Green et al., 1997), and regulates transport from the sorting endosomes to the recycling compartments (Ullrich et al., 1996). In polarized epithelial cells, RAB11a, RAB11b and RAB25 localize to the apical recycling endosomes (Casanova et al., 1999). The RAB11 and RAB25 sub-classes appear to co-localize but are suggested to control distinct transport routes between the recycling endosome and the Golgi or the plasma membrane (Casanova et al., 1999; Somsel Rodman and Wandinger-Ness, 2000). The RAB11 counterpart in Saccharomyces cerevisiae, Ypt31/32, has been implicated in export from the late Golgi compartment to the pre-vacuolar/endosomal compartment and the plasma membrane (Benli et al., 1996; Segev, 2001; Ortiz et al., 2002).
In plants, the RABA group is further divided into six sub-classes (RABA1–RABA6). Previous studies have shown roles for RABA members in a wide range of cellular events. NtRAB11b, a RABA1 member in Nicotiana tabacum, has been shown to play a crucial role in pollen tube growth (de Graaf et al., 2005). Arabidopsis RABA2 and RABA3, which show high similarity to mammalian RAB11 and yeast Ypt31/32, have been reported to localize on a novel post-Golgi membrane domain, that partly overlaps with the trans-Golgi network (TGN) in root tip cells. During cytokinesis, RABA2 and RABA3 proteins localize precisely to the growing margins of the cell plate, suggesting important roles in cytokinesis via regulation of polarized secretion (Chow et al., 2008). The involvement of RABA1a in auxin signaling has also been reported (Koh et al., 2009). Preuss et al. (2004, 2006) have shown that RABA4b functions in polarized secretion in root hair cells, in cooperation with its effector phosphatidylinositol 4-OH kinase beta 1 (PI-4Kβ1). Szumlanski and Nielsen (2009) reported that RABA4d is expressed in a pollen-specific manner and is required for proper development of pollen tubes.
In A. thaliana, RABA1 is the largest sub-group of RABA, with nine members (RABA1a–RABA1i). However, the physiological roles of this sub-group remain unexploited. In this study, we have focused on the RABA1 sub-group, particularly RABA1a, RABA1b, RABA1c and RABA1d, whose expression is abundant in the whole plant body. We have shown that these RABA1 proteins localize on a compartment adjacent to the TGN. Several lines of evidence further indicate that RABA1 regulates transport between the TGN and the plasma membrane. Moreover, we have found that RABA1 members redundantly function in salinity stress tolerance.
Expression patterns of RABA1 members
In A. thaliana, the RABA1 sub-group has nine members (RABA1a–RABA1i). First, we compared the expression profiles of these members from the Arabidopsis Information Resource database (http://www.arabidopsis.org/, ExpressionSet:1006710873), using DART to access and organize the microarray data (http://dandelion.liveholonics.com/dart/index.php). RABA1a, RABA1b, RABA1c and RABA1d are abundant in the whole plant body, whereas RABA1e is mainly expressed in roots. The other four members (RABA1f, RABA1g, RABA1h and RABA1i) are mainly expressed in flowers (Figure S1). Next, we generated transgenic plants expressing GFP/Venus-tagged RABA1a, RABA1b, RABA1c, RABA1d, RABA1e and RABA1f proteins under the control of their native regulatory elements and examined their expression patterns. RABA1a, RABA1b, RABA1c and RABA1d were observed in all parts of the plant body, but their expression patterns were slightly different. RABA1a, RABA1b and RABA1c were abundant in the division zones of shoots and roots. Expression of RABA1d was high in the stele and low in other areas. RABA1e was detected only in roots (Figure 1a), especially in root hair cells (Figure 1b, arrowheads). Consistent with the TAIR microarray data, RABA1f was detected only in pollen tubes (Figure 1c). Thus, RABA1a, RABA1b, RABA1c and RABA1d are major members of the RABA1 sub-group that are expressed throughout all tissues, although the TAIR data predicted modest expression of RABA1f and RABA1g in vegetative parts as well (Figure S1).
Detailed observation of growing root hairs showed characteristic accumulation of GFP–RABA1b, GFP–RABA1c and GFP–RABA1e in the tip region of root hairs (Figures 1a and 2a). The pollen-specific member RABA1f also accumulated in the tip region of growing pollen tubes (Figure 1a). The amino acid sequence similarities among RABA1 proteins are considerably high, approximately 70–80%. These results imply that the five RABA1 (RABA1a, RABA1b, RABA1c, RABA1d and RABA1e) members, and perhaps other members as well, are involved in related transport processes, even though they are expressed in different tissues. In this report, we analyzed RABA1b as the representative of the RABA1 sub-group, because it is the most abundant in major tissues (Figure 1a).
Characterization of the RABA1b compartment
In all the cells we observed, RABA1b was localized on small punctate structures of various size that show dynamic motion (Figure 2a and Movie S1). In the division zone of roots, many dot-like RABA1b-containing compartments were seen in the cytoplasm, especially near the plasma membrane (Figure 2a). In the differentiation zone, fluorescent signals near the plasma membrane were reduced, and distinct signals appeared along lines (Figure 2a). These aligned signals are the fluorescent trajectories of high-speed movement of RABA1b-containing vesicles along the actin cytoskeleton (see Figure 7). RABA1b also accumulated on cell plates in dividing cells and in the tip region of growing root hair cells (Figure 2a, arrow). Similar localization patterns were observed in cells expressing fluorescence-tagged RABA1a and RABA1c (Figure 2b). Transgenic plants co-expressing Venus–RABA1a and mRFP–RABA1b, GFP–RABA1b and mRFP–RABA1b, and GFP–RABA1c and mRFP–RABA1b revealed that each pair of proteins co-localize in the division zone of roots (Figure 2b). The GFP–RABA1b and mRFP–RABA1b pair was used as a positive control of co-localization. We also performed quantitative analysis of the co-localization using a macro written in Metamorph software, which semi-automatically measures the distance from the center of each RFP signal to the center of the nearest GFP or Venus signal in acquired images (Figure S4A) (Ito et al., 2012). GFP–RABA1d and GFP–RABA1e were not detected in root tips, because their expression is spatially restricted to the stele and root hair cells, respectively. Hereafter, we refer to the compartment labeled by GFP–RABA1b as the RABA1b compartment.
Next, the RABA1b compartment was tested for the presence of known organelle markers of post-Golgi traffic by high-resolution confocal laser microscopy. As shown in Figure 3(a–e), the RABA1 compartment showed partial association with organelles labeled by mRFP-tagged syntaxin of plants 43 (mRFP–SYP43, TGN marker) (Uemura et al., 2004) and mRFP-tagged vacuolar proton ATPase a1 (VHA-a1–mRFP, TGN marker) (Dettmer et al., 2006), and less with sialyl transferase-fused mRFP (ST–mRFP, Golgi marker) (Munro, 1995; Boevink et al., 1998), mRFP-tagged arabidopsis rab GTPase homolog 7 (mRFP–ARA7, endosome marker) (Ueda et al., 2001) or mRFP-tagged arabidopsis rab GTPase homolog 6 (ARA6–mRFP, endosome marker) (Ueda et al., 2001). However, the co-localization with VAMP721, R- soluble N-ethyl-maleimide sensitive factor attachment protein receptor (R-SNARE), which is another key regulator of membrane traffic, was remarkable (Figure 3f). The comparison of the RABA1b compartment with the TGN was interesting, because they seemed to be frequently adjacent and sometimes associated (Figure 3a,b). However, this association appeared to be not as tight and stable as that of the Golgi and the TGN (Figure 3g). We sometimes observed the RABA1b compartment adjacent to the ARA6 compartment, but rarely adjacent to the ARA7 compartment. Quantitative analysis of the co-localization clarified the tendency of the markers to co-localize with or be adjacent to the RABA1b compartment (Figure S4B).
To characterize the RABA1b compartment in more detail, we examined whether inhibitors of post-Golgi traffic affect the RABA1b fluorescence pattern. Brefeldin A (BFA) inhibits budding of vesicles and aggregates Golgi and post-Golgi compartments, such as the TGN and endosomes (Grebe et al., 2003; Dettmer et al., 2006). The RABA1b compartment also aggregated upon BFA treatment, although some GFP–RABA1b signals remained on large bead-like structures (Figure 4a, arrowheads, and Figure S5). Wortmannin, an inhibitor of phosphatidylinositol 3-kinase (PI3K), exaggerates the different characteristics of the RABA1b compartment and the TGN. Wortmannin caused no morphological change to the TGN labeled by mRFP–SYP43. However, GFP–RABA1b signals increased on the plasma membrane upon wortmannin treatment, while dot-like signals of GFP–RABA1b in the cytoplasm decreased (Figure 4b and Figure S5). Treatment with dimethylsulfoxide (DMSO) had no effect on these compartments (Figure S2). These results indicate that the RABA1b compartment has a property that is different from the TGN represented by SYP43, and that shows a tendency to associate with the plasma membrane upon wortmannin treatment.
RABA1 members are involved in transport between the TGN and the plasma membrane
VAMP721 and VAMP722 are R-SNARE proteins that operate redundantly in the secretory pathway (Kwon et al., 2008). Double staining of RABA1b and VAMP721 revealed that the two proteins largely co-localize on punctate structures (Figure 3f). Most mRFP–VAMP721 signals were accompanied by GFP–RABA1b, but some GFP–RABA1b signals were not associated with mRFP–VAMP721. GFP–RABA1b also co-localized with mRFP–VAMP722 (Figure S2). mRFP–VAMP722 showed similar responses to GFP–RABA1b upon treatment with wortmannin and BFA. Thus the RABA1b compartment appeared to be closely related to the compartment to which VAMP721 and VAMP722 localize, suggesting a role for RABA1b in transport toward the plasma membrane.
To examine the involvement of RABA1b in transport between the TGN and the plasma membrane, we generated transgenic plants expressing a mutant form of RABA1b, either GTP-fixed (GFP–RABA1bQ72L) or GDP-fixed (GFP–RABA1bS27N). GTP-fixed RAB GTPases are constitutively active, and are thought to accumulate on the target membrane. However, GDP-fixed RAB GTPases act in a dominant-negative way by trapping the guanine nucleotide exchange factor (Burstein et al., 1992; Olkkonen and Stenmark, 1997; Feig, 1999). Significant localization of GFP–RABA1bQ72L to the plasma membrane was observed in root tip cells (Figure 5a, upper panel, arrowhead, and Figure S6) in addition to the punctate structures (arrow). Such relocation was not evident in differentiated cells (Figure 5a, lower panel). The punctate structures of GFP–RABA1bQ72L co-localized with mRFP–VAMP721 almost completely (Figure 5d and Figure S7), and also overlapped well with the TGN labeled by mRFP–SYP43 and VHA-a1–mRFP (Figure 5b,c and Figure S7). RABA1bQ72L showed increased cytosolic labeling compared to the wild-type. In contrast to RABA1bQ72L, GFP–RABA1bS27N was localized to larger punctate structures (Figure 5a and Figure S6). These structures co-localized well with ARA6–mRFP (Figure 5e and Figure S7), suggesting recolation of RABA1bS27N to multi-vesicular endosomes (Ebine et al., 2011). Interestingly, the size of the ARA6–mRFP compartments appeared larger than in the control (Figure 3e) when co-expressed with GFP–RABA1bS27N.
The co-localization of RABA1b and VAMP721/722 and the localization of GFP–RABA1bQ72L on the plasma membrane support the idea that RABA1b may function in transport from the TGN to the plasma membrane. To test whether RABA1b plays any role in endocytosis, we examined the effect of GFP–RABA1bQ72L or GFP–RABA1bS27N expression on the uptake of FM4-64, a lipophilic dye used to trace endocytosis. As shown in Figure 6, no significant difference was observed in the kinetics of FM4-64 internalization.
Dynamic motion of RABA1b vesicles is actin-dependent
As shown in Figure 2, which was taken with a long exposure time (1 sec), localization of GFP–RABA1b appeared to occur along distinct lines. To increase the time resolution, we applied variable incidence angle fluorescent microscopy (VIAFM), a technique that is related to total internal reflection fluorescence microscopy (TIRFM), which has a high signal-to-noise ratio and can take pictures at 10 frames per sec. High-speed movies clearly showed that these aligned signals were the fluorescent trajectories of GFP–RABA1b-containing vesicles moving along filamentous structures (Movie S2). To reveal the nature of these structures, we studied transgenic plants co-expressing Lifeact–Venus (actin marker, Era et al., 2009) and GFP–RABA1b. The results indicated that vesicles labeled by GFP–RABA1b form queues along actin filaments (Figure 7a). Such aligned signals were dispersed when actin filaments were depolymerized by latrunculin B. No apparent change was observed in response to treatment with oryzalin, a microtubule-depolymerization agent (Figure 7b and Movie S3). Thus the dynamic motion of RABA1b vesicles is actin-dependent.
RABA1 members function in salinity stress tolerance
All pieces of evidence obtained so far suggest that the major function of RABA1b is in the secretory pathway, at the transport step from the TGN to the plasma membrane. To further understand possible physiological roles of RABA1 members, we obtained four T-DNA-inserted mutant lines, which were completely knocked out for expression of RABA1A, RABA1B, RABA1C and RABA1D (Figure 8a). Single mutants showed no macroscopically abnormal phenotypes. We therefore constructed multiple mutated plants by cross-pollination. The quadruple mutant raba1a raba1b raba1c raba1d exhibited only a marginal phenotype under ordinary growth conditions, with slightly shorter roots than the wild-type (Figure 8b,c). Transgenic plants expressing the dominant-negative RABA1bS27N mutant showed a more prominent growth defect than the quadruple mutant (Figure 8d), suggesting that other RABA1 members support loss of RABA1a–RABA1d but are dominantly impaired by expression of RABA1bS27N. Next, we focused on stress tolerance and investigated the phenotypes of the quadruple mutants and the plants expressing RABA1bS27N under biotic or abiotic stress. We found that these plants exhibited prominent phenotypes when subjected to salinity stress. Interestingly, the quadruple mutant showed a very severe growth defect under salinity stress conditions (Figure 8c). On one-tenth strength MS medium containing 15 mm NaCl, the majority of the quadruple mutants died during the cotyledon stage (Figure 8c). The fresh weights of the quadruple mutant were fivefold lower than those of the wild-type plants on average (Figure 8e). This salt hypersensitivity of the quadruple mutant was complemented by Venus–RABA1a, GFP–RABA1b or GFP–RABA1c (Figure 8c). The plants expressing RABA1bS27N were also hypersensitive to salinity stress (Figure 8d,e). Neither the quadruple knockout mutation nor expression of RABA1bS27N affected growth sensitivity to 30 mm sorbitol (Figure S3), indicating that the hypersensitivity was not due to high osmolarity but to the high salt concentration. Taken together, these results show that at least four members of the RABA1 group, RABA1a, RABA1b, RABA1c and RABA1d, act redundantly in tolerance to salinity stress.
Characteristics of the RABA1 compartment
In this study, we analyzed the subcellular localization of RABA1b as a representative of the RABA1 group, and show that its compartment overlaps with the TGN or lies in its vicinity. The RABA1b compartment shows dynamic motion in an actin-dependent fashion. Members of other RABA sub-groups, RABA2 and RABA3, have been reported to partially co-localize with structures labeled by VHA-a1–GFP and FM4-64, but not with the Golgi apparatus or the pre-vacuolar compartment labeled by ST–GFP and GFP–BP-80 (Chow et al., 2008). The authors concluded that the RABA2/3 compartment is an early endosomal TGN-associated membrane domain. In a collaborative study, Feraru et al. (2012) showed that RABA1b and RABA2a co-localize. Whether these sub-classes of RABA share membrane traffic events around the TGN remains to be pursued.
RABA1b regulates the transport between the TGN and the plasma membrane
RAB GTPases generally act as molecular switches through conformational change between GTP- and GDP-bound forms in order to control docking and fusion events. GTP-bound Rab GTPases are considered to localize on the target organelle (Chavrier and Goud, 1999). The fact that GTP-fixed RABA1b largely localized to the plasma membrane is likely to reflect the function of the RABA1b, suggesting a role in the exocytic event. Dettmer et al. (2006) indicates that the TGN functions as the early endosome in the endocytic pathway. However, neither the GTP-fixed nor the GDP-fixed mutant of RABA1b affected endocytosis of the FM4-64 dye. A similar result was obtained in an independent study on epidermal leaf cells of Nicotiana benthamiana (S. Choi, Department of Biological Sciences, Graduate School of Science, The University of Tokyo, T. Tamaki, Department of Biological Sciences, Graduate School of Science, The University of Tokyo, K. Ebine, Department of Biological Sciences, Graduate School of Science, The University of Tokyo, T. Uemura, T. Ueda and A. Nakano, unpublished results). A similar conclusion was also drawn on the basis of another independent study on recycling of PIN proteins (Feraru et al., 2012). Thus, we suggest that RABA1b, and probably other RABA1 members as well, act on the transport process from the TGN-associated compartment to the plasma membrane.
We also observed that RABA1b shows partial association with ARA6, which is more evident when RABA1b is locked in the GDP form. ARA6 is a plant-unique RAB5 member that mediates transport from the multi-vesicular endosomes to the plasma membrane (Ebine et al., 2011). Expression of the RABA1b GDP mutant increased the size of ARA6 endosomes. These results may provide an indication of the molecular mechanisms of the complicated post-Golgi traffic system in plants.
Evolution of RABA1 members – redundancy and diversity
Here, we have demonstrated that RABA1 members function redundantly. It is also possible that some members of other RABA sub-groups share part of functions with RABA1. Many members of the RABA family show similar localization patterns, especially in relation to the TGN. Furthermore, RABA4b actively accumulates in the tip region of root hairs (Preuss et al., 2004), as for RABA1b and RABA1e (Figures 1a and 2a). RABA4d, which has a crucial role in the formation of pollen tubes (Szumlanski and Nielsen, 2009), accumulates in the tip of pollen tubes, as does RABA1f (Figure 1c). On the other hand, some RABA members have been implicated in endocytic events, unlike RABA1b (our unpublished results). Further extensive analysis is necessary to understand the whole spectrum of the RABA GTPase functions, which have shown multiplication and diversification during evolution.
Analysis of the quadruple raba1a raba1b raba1c raba1d mutant has revealed that RABA1a–d proteins are required for tolerance to salinity stress. This implies that these RABA1 proteins regulate the localization of cell-surface proteins, such as pumps and channels. The fact that the GDP-fixed mutant of RABA1b causes even more severe salt sensitivity indicates that other members of the RABA1 sub-family are also involved. For land plants, it is imperative to adapt quickly and flexibly to a change in environment. The expansion and diversification of RABA members may be advantageous to sense the emergence of stress and respond to it. Determination of the cargo transported by RABA1 members and regulators of RABA1 is a very important subject for future studies.
Plant materials and growth conditions
T-DNA insertion lines were derived from the Col-0 accession of A. thaliana. Both lines were backcrossed three times into the Col-0 accession. Seeds for raba1a (SALK77747), raba1c (SALK145363) and raba1d (SALK88879) were obtained from the Arabidopsis Biological Resource Center. Seeds for raba1b (SAIL875F08) were obtained from the Nottingham Arabidopsis Stock Centre.
Arabidopsis seeds were sterilized and planted on 0.3% agar plates containing half-strength Murashige and Skoog medium, 1% w/v sucrose and vitamins (pH 5.7). Plants were grown in a climate chamber at 23°C under the continuous light.
Total RNA was extracted from rosette leaves of 10-day-old wild-type and mutant plants using RNAqueous-4PCR (Life Technologies, http://www.lifetechnologies.com/), and cDNA was synthesized using SuperScript III reverse transcriptase (Life Technologies). PCR conditions were as follows: 95°C for 2 min, then 30 cycles of 95°C for 30 s, 55°C for 30 sec and 72°C for 1 min, then 72°C for 5 min.
Plasmids and transformation of plants
By fluorescent tagging of full-length proteins (Tian et al., 2004), fragments of each gene, including cDNA encoding GFP, Venus or mRFP in front of the start codon of each gene, were generated. The RABA1A fragment contains 3.3 kb of 5′ and 1.1 kb of 3′ flanking sequences. The RABA1B fragment contains 1.7 kb of 5′ and 1.0 kb of 3′ flanking sequences. The RABA1C fragment contains 1.9 kb of 5′ and 1.2 kb of 3′ flanking sequences. The RABA1D fragment contains 3.2 kb of 5′ and 0.9 kb of 3′ flanking sequences. The RABA1E fragment contains 2.4 kb of 5′ and 0.7 kb of 3′ flanking sequences. The RABA1F fragment contains 1.6 kb of 5′ and 0.8 kb of 3′ flanking sequences. Amplified chimeric fragments were sub-cloned into the binary vector pGWB1, a kind gift from Dr. T. Nakagawa (Center for Integrated Research in Science, Shimane University, Japan), which were used for transforming Arabidopsis plants. Transformation of Arabidopsis plants was performed by floral dipping (Clough and Bent, 1998) using Agrobacterium tumefaciens strain GV3101::pMP90.
Plants expressing GFP-, Venus- or mRFP-tagged proteins were observed under an LSM710 confocal microscope equipped with a META device (Carl Zeiss, http://corporate.zeiss.com/) for multi-color observations, and an Olympus (http://www.olympus-global.com/) BX51 microscope equipped with a confocal scanner unit (model CSU10; Yokogawa Electric, http://www.yokogawa.com/) for single-color observations, with a cooled CCD camera (model ORCA-AG; Hamamatsu Photonics, http://www.hamamatsu.com/index.html). The co-localization analysis was performed as described previously (Ito et al., 2012). Images were captured using an LSM710 confocal microscope with oil immersion lens (× 63, numerical aperture = 1.40). The line scan analysis was performed using Metamorph software (Molecular Devices, http://www.moleculardevices.com/).
Roots of plants after 5 days of culture were observed. For observation of root hairs, seeds were germinated on thin agar medium (compositionally the same as described above) in a glass-bottomed dish and observed without any preparation after 10 days of culture.
Variable incidence angle fluorescent microscopy
Roots of transgenic A. thaliana seedlings were placed on glass slides (76 × 26 mm; Matsunami, http://www.matsunami-glass.co.jp/), covered with a 0.12–0.17 mm thick cover slip (24 × 60 mm; Matsunami), and root epidermal cells were observed under a fluorescence microscope (Nikon Eclipse TE2000-E and a CFI Apo TIRF × 100 H/1.49 numerical aperture objective, http://www.nikon.com/) equipped with a Nikon TIRF2 system. GFP was excited using an 488 nm laser. All images were acquired using an Andor iXonEM electron multiplying charge coupled device (EMCCD) camera (http://www.andor.com/).
To visualize the process of endocytosis, seedlings were mounted in half-strength MS liquid with 16.5 μm FM4-64 (Invitrogen/Molecular Probes; T13320; diluted from a 16.5 mm stock in water) on ice, then washed out and incubated at 23°C. For BFA treatment, seedlings were incubated in half-strength MS liquid containing 50 μm BFA diluted from a 50 mm stock in DMSO, and then mounted on the slides in the presence of BFA. Wortmannin was used at 33 μm diluted from a 10 mm stock in DMSO, and the treatment was performed in a similar way as BFA treatment. Latrunculin B and oryzalin treatments were also performed in similar ways. Control treatment was performed with DMSO in place of each chemical.
Salinity stress assay
The salinity stress assay was performed as described by Krebs et al. (2010). Surface-sterilized seeds were sown on 1% agar plates containing one-tenth strength MS medium, 0.5% w/v sucrose and 10 mm MES/KOH (pH5.8) with or without 15 mm sodium chloride or 30 mm sorbitol (as an osmotic control). Fresh weight was measured after 14 days of culture.
We thank the Arabidopsis Biological Resource Center and Nottingham Arabidopsis Stock Centre for providing T-DNA insertion mutants of Arabidopsis. We also thank T. Nakagawa (Center for Integrated Research in Science, Shimane University, Japan) and M. Sato (Graduate School of Biostudies, Kyoto Prefectural University, Japan) for sharing materials, N. Tsutsumi (Graduate School of Agricultural and Life Sciences, University of Tokyo, Japan) for variable incidence angle fluorescent microscopy and K. Ebine (National Institute of Infectious Diseases, Tokyo, Japan) for valuable advice. This work was supported by a Grant-in-Aid for Specially Promoted Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan, and by funds from the Extreme Photonics and Cellular Systems Biology Projects of RIKEN. R.A. is the recipient of a Research Fellowship for Young Scientists from the Japan Society for the Promotion of Science.