DRP1A Is Responsible for Vascular Continuity Synergistically Working with VAN3 in Arabidopsis

DRP1A Associates with VAN3

To isolate VAN3 interacting proteins, we performed a yeast two hybrid screening with VAN3 as bait. From the screening of an Arabidopsis cDNA library, we isolated DRP1A as a VAN3 binding protein. DRP1A is a known member of the dynamin family that constitutes a structurally related, but functionally diverse, family of large GTP binding proteins that share a conserved N-terminal GTPase domain and a C-terminal GTPase effector domain (Danino and Hinshaw, 2001). To confirm the association between DRP1A and VAN3, a swapping test was conducted. VAN3 was fused with the DNA activation domain in pACT2, and DRP1A was fused with the DNA binding domain of pAS2 in the yeast two hybrid system. When pACT2-VAN3 and PAS2-DRP1A were cotransformed in the yeast strain Y190, the yeast could survive on a selection agar plate without His. This indicated an interaction between VAN3 and DRP1A in yeast (Fig. 1A).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

VAN3 interaction with DRP1A. A, Yeast two hybrid analysis. The Y190 yeast strain was transformed with pAS2-VAN3 and pACT2-DRP1A. The yeast harboring pAS2-DRP1A and pACT2-VAN3 were grown on agar plates without His. The Y190 yeast strain was transformed with pAS2-VAN3 and the pACT2 vector or pAS2-DRP1A and the pACT2 vector are shown as negative controls. B, Protein-protein interaction assay with GST pull down. VAN3-GAL4 DNA binding domain pull down with purified GST and GST-DRP1A used as bait. The VAN3 protein was detected with a western-blot analysis. C, Time course of DRP1A GTPase activity. Recombinant DRP1A protein was incubated with 5 μm [α -³²P] GTP. GTP hydrolysis was observed (*) in the left section. GTP hydrolysis was quantitatively analyzed with a BIO-imaging analyzer in the right section.

To validate the protein-protein interaction revealed by the yeast two hybrid, a glutathione S-transferase (GST) pull-down assay was performed with a recombinant protein of GST full-length DRP1A as bait and a recombinant VAN3-GAL4 DNA binding domain protein as predator. Results showed that a GST-DRP1A recombinant protein pulled down the VAN3-GAL4 DNA binding domain recombinant protein band. However, GST itself did not (Fig. 1B).

DRP1A Has GTPase Activity

DRP1A is known to have a GTPase domain and a GTPase effecter domain. To examine the GTPase activity of the DRP1A protein, recombinant GST-tagged DRP1A protein was purified, and the hydrolysis of GTP by the recombinant DRP1A was measured. Radionucleotide-labeled GTP bound to DRP1A formed the substrate DRP1A-GTP. Further incubation under a condition that minimized interference by nucleotide exchange or contaminating nucleotidases resulted in GTP hydrolysis on the DRP1A in a time dependent manner (Fig. 1C). This result suggested that the DRP1A protein functions as a GTPase.

Vascular Discontinuity of drp1a

Kang et al. (2003) showed that the drp1a/adl1a mutant had defects in the anisotropic expansion of stigmatic papillae and in the development of leaf trichomes. These results suggested that DRP1A might be responsible for expansion during single cell differentiation. They also discussed the possibility of DRP1A function in vascular formation (Kang et al., 2003).

To examine the DRP1A function in vascular formation, we cleared cotyledons of the drp1a mutants and observed vascular tissues. The cotyledon of Arabidopsis has a very simple continuous vein pattern: one midvein and three or four lateral veins (Fig. 2A). Approximately 10% of drp1a mutants showed disconnected vascular network (19 abnormal plants in 200 mutants; Fig. 2B), and the leaf size and shape is not so affected. This disconnected venation pattern of the drp1a mutant was similar to that of the van3 mutant, although the discontinuity in drp1a was weaker than that of the van3 mutant (Fig. 2C). In contrast, we could not observe disconnection in the wild type (n = 200).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Genetic interaction of the drp1a mutant. A, Wild type. B, About 10% of drp1a mutants show a disconnected vasculature. Arrows show the vascular disconnection. C, van3 mutants show a discontinuous vasculature. D, drp1a van3 double mutants showed a severe defect in vascular development. Arrows show the restricted maturation of the tracheary element. E, gnom mutants show a concentrated vasculature. F, gnom van3 double mutants show a highly fragmented vasculature. G, gnom drp1a double mutants show a fragmented vasculature similar to the gnom van3 double mutant. Bars = 500 μm.

Genetic Interaction of DRP1A and VAN3

To observe the genetic interaction between the van3 and drp1a mutants, we produced van3 drp1a double mutants. The van3 drp1a double mutant had a defect in germination, although its embryo development was anatomically indistinguishable to that of the wild type (data not shown). In the F₂ progeny of a cross between van3 and drp1a, mutants from each combination segregated at ratios of about 9:3:3:1 (wild type:van3:drp1a:van3 drp1a = 176:64:56:18; x² 0.002 < P < 0.45). Genotypes of the 18 van3 drp1a double mutants were confirmed by PCR methods (Sawa et al., 1997). Seventeen of the 18 mutants failed to germinate, and only one mutant line showed seedling lethality without producing rosette leaves. This van3 drp1a double mutant had severely fragmented and poorly maturated vasculature (Fig. 2D). In repeated experiments, about 5% of van3 drp1a double mutants could germinate (9/200), and all double mutants showed a similar phenotype. Thus, the drp1a mutation enhanced the van3 phenotype, suggesting that the DRP1A gene may be responsible for synergistic vascular formation with the VAN3 gene.

Genetic Interaction of DRP1A and GNOM

The gnom/van7 mutant is known to show an abnormal venation pattern (Koizumi et al., 2000). About one-half of the gnom seedlings (10/23) produced fused cotyledons, and the gnom cotyledon produced excess vascular tissues (Fig. 2E). Although the gnom mutant sometimes produces partially fragmented vascular tissues, the fragmentation is not as obvious, and the fragmentation was never observed in the primary veins. GNOM is known to encode an ARF-GEF, which is needed in PIN1 recycling between the plasma membrane and endosomes. On the other hand, VAN3 is shown to encode an ARF-GAP (Koizumi et al., 2005). In seedlings of the van3 gnom double mutants, about one-half produced fused cotyledons (21/43). The cotyledon of the van3 gnom double mutants contained one or two midveins and highly fragmented lateral veins (Fig. 2F). Because we revealed that DRP1A could bind with VAN3, genetic interactions between the drp1a mutant and the gnom mutant were also examined. In seedlings of the gnom drp1a double mutant, about one-half produced fused cotyledons and a highly fragmented vasculature (Fig. 2G). Interestingly, the fragmentation was also observed even in the primary veins (Fig. 2G). This was similar to that of the van3 gnom double mutant. Thus, the drp1a mutation appears to have a similar effect to that of the van3 mutation, which enhances the fragmentation of vasculature in the gnom mutants. These results suggested that the DRP1A gene might be responsible for vascular formation.

Overlapping Expression of the DRP1A and VAN3 Genes

To examine the temporal and spatial interrelationship between DRP1A and VAN3 during leaf development, we analyzed the expression of DRP1A∷GUS (Kang et al., 2003) and VAN3∷GUS transgenic plants. The VAN3 promoter sequence used in this β-glucuronidase (GUS) fusion constructs includes 1.1 kb upstream from the putative transcription start site, and it is sufficient to complement the van3 mutant phenotype in the complementation test (Koizumi et al., 2005). The progeny from three independent VAN3-GUS transgenic plant lines were used for a histochemical analysis of GUS activity.

DRP1A∷GUS and VAN3∷GUS transgenic plants showed overlapping reporter gene expression profiles during leaf development. GUS staining was ubiquitously observed in the cotyledons of both the DRP1A∷GUS and VAN3∷GUS transgenic seedlings 2 d after germination (DAG; Fig. 3, A and D). Strong GUS staining was characteristically observed in vascular tissues at 10 DAG (Fig. 3, B and E). At 21 DAG, GUS staining was observed at the margin of the cotyledons (Fig. 3, C and F). A similar expression pattern was observed in the rosette leaves, while a ubiquitous expression of the two genes was observed at early stages of rosette leaf development (Fig. 3, G and K). GUS staining was also observed in developing trichomes (Fig. 3, G, H, K, and L). As leaves developed, the ubiquitous expression was weakened (Fig. 3, H–J, L–N) and restricted to vascular tissues. At this stage, GUS staining was not obvious in the trichomes. This overlapping spatial expression pattern of DRP1A and VAN3 strongly suggests that DRP1A and VAN3 work in the same tissues and cells.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Temporal and spatial expression pattern of DRP1A and VAN3 during leaf development. Histochemical localization of VAN3∷GUS (A–C, G–J) and the DRP1A∷GUS (D–F, K–N) reporter gene expression in the cotyledon (A–F) and in rosette leaf (G–N). Samples were fixed and incubated in a reaction buffer including 2 mg/mL of X-Gluc for 10 min. Bars = 500 μm (A–F), 1 mm (G–N).

Overlapped Localization of the DRP1A and VAN3 Proteins in Subcellular Compartment

To understand the functional interaction between DRP1A and VAN3, we examined the localization of DRP1A and VAN3. Red fluorescent protein (RFP)-tagged DRP1A was introduced into Arabidopsis suspension-cultured cells, and their subcellular locations were observed with a confocal laser scanning microscope (Fig. 4A). DRP1A localized to a fiber like structure. In addition to the fiber like structure, a DRP1A positive signal was obvious in some punctate structures (Fig. 4A). It has been suggested that these fiber like structures are microtubules (Kang et al., 2001; Hong et al., 2003b). To clarify that this is the case, we used green fluorescent protein (GFP)-tagged MAP4 as a microtubule marker (Fig. 4B). When RFP-tagged DRP1A and the GFP-tagged MAP4 were cointroduced, the DRP1A positive signal overlapped to that of MAP4. This indicated that DRP1A was localized to the microtubules under our conditions (Fig. 4, D–F).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Subcellular localization of DRP1A and VAN3. 35S-promoter-driven RFP-tagged DRP1A. Arrows show punctate structures. A, GFP-tagged MAP4 as a microtubule marker (B) and VENUS-tagged VAN3 (C) were introduced into protoplasts of Arabidopsis suspension-cultured cells. RFP-DRP1A and GFP-MAP4 (D–F), and RFP-DRP1A and VAN3-Venus (G–I) were cointroduced. Localization of RFP-DRP1A (D), MAP4-GFP (E), and merged image (F) are shown. The location of RFP-DRP1A (G), VAN3-VENUS (H), and merged image of RFP-DRP1A and VAN3-VENUS (I) are shown. Fluorescence was observed by confocal laser microscopy (LSM510 META, Carl Zeiss).

The VAN3 protein is known to be located at the TGN (Koizumi et al., 2005). We confirmed this by VENUS (Nagai et al., 2002)-tagged VAN3 (Fig. 4C). When VAN3-VENUS and RFP-tagged DRP1A were cointroduced, VAN3-VENUS-positive compartments overlapped the punctate structures of the RFP-DRP1A positive signal (Fig. 4, G–I). Similar overlapping figures were obtained in with tobacco (Nicotiana tabacum) BY2 suspension-cultured cells (data not shown). These results suggest that the VAN3 protein can colocalize with DRP1A at the TGN, which might be associated with microtubules.

DRP1A and VAN3 Interaction in Vascular Formation

Here, we isolated a plant dynamin, DRP1A, as a VAN3 interaction protein using yeast two hybrid screening, and their interaction was confirmed by swapping test and a biochemical pull-down assay. We analyzed the interaction between DRP1A and VAN3 in situ with different types of experiments. A histochemical analysis of the drp1 mutant revealed disconnected vascular formation, which is weaker, but significantly resembled that of the van3 phenotype. The phenotype of the drp1a van3 double mutants indicated a genetic interaction between the two mutations with regard to vascular formation. Furthermore, the drp1a mutation enhanced the vascular fragmentation of the gnom mutants in the same manner as the van3 mutation. A promoter analysis of VAN3 and DRP1A indicated spatially and temporally overlapped expression patterns of these two genes. Taken together with the finding of subcellular colocalization, these results strongly support the in vivo interaction of DRP1A and VAN3 in controlling vascular formation.

Vesicle Trafficking Mechanisms Regulated by DRP1A and VAN3 in Vascular Formation

Formation of transport vesicles involves the assembly of distinct coat complexes that induces membrane budding and the selection of cargo proteins. This process is regulated by small GTPases such as ARFs, which are required for the formation of TGN-derived clathrin coated vesicles (Bednarek and Falbel, 2002). In mammalian cells, the release of clathrin-coated TGN-derived vesicles involves the action of another GTPase dynamin (Jones et al., 1998). Animal dynamin was originally isolated from bovine brain extracts as a microtubule-binding protein (Shpetner and Vallee, 1989), and its GTPase activity was shown to be activated by microtubules (Shpetner and Vallee, 1992). It has been suggested that dynamin GTPase may be involved in the transport of Golgi-derived vesicles along the microtubules, from the minus end to the plus end (Fullerton et al., 1998). Using an antibody against Arabidopsis DRP1A, Kang et al. (2001) observed significant immunolabeling of punctate structures with the antibody in Arabidopsis suspension cultured cells, although they did not determine the nature of the structures. Furthermore, Murphy et al. (2002) indicated that β-adaptin, a subunit of the clathrin adaptor protein complex, and the AtSEC14 protein were cofractionated with DRP1A by affinity chromatography on immobilized-naphtylphthalamic acid. These results suggest that dynamins may play a role in vesicle transport from the TGN (Murphy et al., 2002). We have indicated that VAN3 is located on the TGN (Koizumi et al., 2005). In this paper, we revealed that DRP1A is colocalized with VAN3 and also with microtubule marker, MAP4. Hong et al. (2003b) also revealed the colocalization of DRP1A with microtubules in tobacco cultured cells. All of these findings strongly suggest that the DRP1A is responsible for the vesicle trafficking from the TGN, which is associated with microtubules, synergistically working with VAN3.

Diverse DRP1A Function in Vesicle Trafficking and in Plant Development

drp1a showed pleiotropic abnormalities in plant development including normal embryogenesis, seedling development, and reproduction (Kang et al., 2003). We also observed the fragmented vascular tissue formation. DRP1A appears to be localized at different subcellular compartments such as punctate organelles and microtubules, probably depending on cell stages and cell types, and therefore, DRP1A may have pleiotropic functions in plant development. DRP1A is categorized to be a member of five phragmoplastin-like DRPs in Arabidopsis (Hong et al., 2003a). DRP1A and DRP1E are suggested to have overlapped functions in embryo development because of the embryo lethality in the drp1a drp1e double mutant (Kang et al., 2003). Although it is still unclear about the physiological function of other three members of DRP1 homologs, these genes may have diverse physiological functions in plants.

VAN3 is an ACAP-type ARF-GAP, which have BIN/amphiphysin/RVS, pleckstrin homology, ARF-GAP, and ankyrin repeats domains, and have been shown to localize at subpopulation of TGN (Koizumi et al., 2005). In Arabidopsis, there are three other VAN3 homologs, which possess all the four domains. The domain structures strongly suggest that the VAN3 homologs may also be responsible for vesicle trafficking. The preliminary examination with their promoters fused to the GUS gene suggested spatial expression patterns of the VAN3 and three VAN3 homologs are different from each other; i.e. some show distinct tissue specificity and others show ubiquitous expression (data not shown). This result suggests that VAN3 and the VAN3 homologs function in vesicle trafficking, responsible for diverse physiological events even in plants. Here, we indicated the interaction of DRP1A with VAN3, which complexes locate on TGN (Koizumi et al., 2005). Therefore, a unique set of a dynamin and a VAN3 homolog might play a specific role in vesicle trafficking at distinctive TGN during various developmental processes in plants. Detailed analysis of subcellular localization and functions of the dynamin and VAN3 family genes in various cell types and cell stages may cast a new light on the molecular mechanisms of vesicle trafficking underlying various plant developmental events.

Plant Materials

The van3 mutant was isolated from Arabidopsis (Arabidopsis thaliana) Landsberg erecta by ethylmethane sulfonate mutagenesis (Koizumi et al., 2000). Seeds were sown on the surface of vermiculite in small pots and incubated at 4°C for 3 d. Plants were grown in a laboratory under continuous illumination of 50 to 100 μE/m2 per second at 22°C. For histochemical analysis, GUS staining was performed as described by Koizumi et al. (2000), except that samples were incubated in the GUS substrate solution for 10 min. All of the drp1a plants were grown in the Murashige and Skoog medium containing 100 mm Suc to suppress the developmental defects (Kang et al., 2001).

Yeast Two Hybrid Screening

Full-length cDNA of the VAN3 was ligated to the plasmid pAS-2, and the resulting pAS-VAN3 was used as bait to screen an Arabidopsis cDNA library constructed in the pACT2 vector. The cDNA library constructed in pACT2 was transformed to yeast strain Y190 containing pAS-VAN3. Positive colonies were selected on SC-Trp-Leu-His medium. After the confirmation by the retransformation test and X-gal test, the inserts were sequenced.

Purification of the DRP1A Protein

To express the GST-tagged DRP1A protein in a bacteria, an expression plasmid was constructed by use of pGEX-5T. The DRP1A gene was amplified by PCR, and the obtained fragment was ligated into pGEX-5T. The sequence of the DRP1A in the expression plasmid was confirmed by cycle sequencing. Expression of the DRP1A protein was induced in Escherichia coli BL21 cells harboring the expression plasmid by addition of isopropylthio-β-galactoside. Cultivation of the E. coli was carried out at 14°C. When the absorbance of the culture at 600 nm reached 0.8, 1 mm isopropylthio-β-galactoside was added to the culture media, and the cultivation was then continued for an additional 2 d. Purification procedure was carried out at 4°C. A 1-litter culture of E. coli cells was suspended in 50 mL of 10 mm Tris-HCl at pH 7.5, 0.1 mm EDTA, 5 mm β-mercaptoethanol, and sonicated on ice for 2 min, and centrifuged at 15,000 rpm (27,000g) for 30 min. The supernatant was applied to a glutathione column (Pharmacia Biotech, Uppsala). After the column had been washed with a phosphate-buffered saline, DRP1A was eluted with a solution containing glutathione.

GTPase Assay

DRP1A GTPase activity was determined using an in vitro assay that measures a single round of GTP hydrolysis on recombinant DRP1A protein (Makler et al., 1995; Damke et al., 2001). A total of 5 μm DRP1A was first loaded with 5 μm [α -³²P] GTP in a buffer (25 mm HEPES, pH 7.5, 1 mm dithiothreitol, 2 mm EDTA, 1 mm MgCl₂, 10 mm ATP, and 0.1% sodium cholate). DRP1A-GTP binding efficiency was calculated by filter assay (Koizumi et al., 2005). GTPase assay was performed at 37°C in 50 mm HEPES, pH 7.5, 4 mm MgCl₂, and 10 mm ATP. The reaction was initiated by the addition of 1 μm of [α-³²P] GTP-loaded DRP1A, and stopped by the addition of 250 mm EDTA, followed by placing on ice. The nucleotides were separated by thin layer chromatography on PEI-cellulose sheets (Merck, Rahway, NJ) developed with 1 m LiCl and 1 m HCOOH. The sheets were dried by air dryer, autoradiographed on an imaging plate, and quantitatively analyzed with a BIO-imaging Analyzer (BAS 2000, Fuji Photo Film, Tokyo). By using the DRP1A-GTP binding efficiency value, we calculated the efficiency of the GTP hydrolysis. In the control experiment, GST protein was incubated in the reaction buffer, and no GTP hydrolysis was observed.

Pull-Down Assay

Yeast strain Y190 carrying pAS-VAN3 was sonicated on ice for 2 min and centrifuged at 15,000 rpm (27,000g) for 30 min. Proteins included in the supernatant were precipitated by ammonium sulfate. Precipitant was dissolved with Tris-HCl, pH 7.5, and it was dialyzed by Tris-HCl, pH 7.5, to remove ammonium sulfate, resulting in partially purified VAN3-GAL4 DNA binding domain recombinant protein. Recombinant GST-tagged DRP1A protein was also dialyzed against 25 mm Tris-HCl, pH 7.5. A total of 15 μg of partially purified VAN3-GAL4 DNA binding domain recombinant protein was mixed with 5 μg of purified GST-DRP1A recombinant protein or purified GST recombinant protein in 100 μL reaction buffer (25 mm Tris-HCl, pH 7.5, 150 mm NaCl, 0.1% Triton X-100). The reaction mixture was shaken at 4°C for 45 min, and 15 μL of glutathione Sepharose 4B bed was then added to the mixture that was shaken for an additional 30 min at 4°C. The glutathione Sepharose 4B bead was then washed five times with 1 mL reaction buffer and boiled in SDS-PAGE sample buffer. After SDS-PAGE, we detected the VAN3 protein by the western-blot experiment, using anti-VAN3 antibody as a probe. Polyclonal anti-VAN3 antibodies were produced against to the synthetic peptide EKMQEYKRQVDRESR by a rabbit. Western-blot experiment was performed as described (Ito and Fukuda, 2002).

Subcellular Localization

Full-length VAN3 cDNA was isolated by reverse transcription-PCR from Columbia ecotype, and an XhoI/NcoI restriction site was introduced at both ends. The fragment was translationally fused to the N terminus of VENUS yellow fluorescent protein. The chimeric gene was subcloned under the control of the cauliflower mosaic virus 35S promoter and the Nos terminator. 35S∷RFP-DRP1A was also produced in the same way to the VAN3-VENUS, except for the DRP1A fusion to the C terminus of RFP protein. 35S∷MAP4-GFP and 35S∷Talin-GFP were used as a intracellular markers of tubulin and actin, respectively. Double transient expression of 35∷VAN3-Venus and of intracellular markers in the protoplasts of cultured Arabidopsis cells were analyzed as described by Ueda et al. (2001, 2004). Fluorescence was observed by confocal laser microscopy (LSM510 META, Carl Zeiss, Jena, Germany).