Vacuolar/pre-vacuolar compartment Qa-SNAREs VAM3/SYP22 and PEP12/SYP21 have interchangeable functions in Arabidopsis

Authors


For correspondence (fax +81 3 5841 7613; e-mail tuemura@biol.s.u-tokyo.ac.jp).

Summary

SNAREs (soluble N-ethylmaleimide sensitive factor attachment protein receptors) mediate specific membrane fusion between transport vesicles or organelles and target membranes. VAM3/SYP22 and PEP12/SYP21 are Qa-SNAREs that act in the vacuolar transport pathway of Arabidopsis thaliana, and are localized predominantly on the vacuolar membrane and the pre-vacuolar compartment (PVC), respectively. Previous studies have shown that loss-of-function mutants of VAM3/SYP22 or PEP12/SYP21 showed male gametophytic lethality, suggesting that VAM3/SYP22 and PEP12/SYP21 possess different, non-redundant functions. We have re-evaluated the effects of mutations in these genes using T-DNA insertion mutants in the Columbia accession. We found that a mutation in VAM3/SYP22 (vam3-1) caused pleiotropic abnormalities, including semi-dwarfism and wavy leaves. In contrast, a loss-of-function mutant of PEP12/SYP21 (pep12) showed no apparent abnormal phenotype. We also found that the double vam3-1 pep12 mutant had severely reduced fertilization competence, although male and female gametophytes (vam3-1pep12) maintained the ability to fertilize. Moreover, promoter swapping analysis revealed that expression of a GFP–PEP12/SYP21 fusion under the control of the VAM3/SYP22 promoter suppressed all phenotypes of the vam3-1 mutant. These results indicate that the functions of VAM3/SYP22 and PEP12/SYP21 were redundant and interchangeable.

Introduction

Membrane trafficking is an essential part of various cellular activities in all eukaryotic cells, including environmental responses and maintenance of cellular homeostasis. One group of key regulators of membrane trafficking is the SNARE family (soluble N-ethylmaleimide sensitive factor attachment protein receptors). SNAREs mediate membrane fusion between transport vesicles and target organelles (Chen and Scheller, 2001; Jahn et al., 2003; Jurgens, 2004; Ebine and Ueda, 2009; Saito and Ueda, 2009). The Qa-SNARE group is located on the target membrane and forms a SNARE complex with Qb-, Qc- and R-SNAREs by interacting with each coiled-coil SNARE motif. Only correct combinations of cognate SNAREs generate functional SNARE complexes and perform accurate membrane fusions (Bock et al., 2001; Kloepper et al., 2007; Sanderfoot, 2007).

In yeast, two Qa-SNAREs, Vam3p and Pep12p, are localized on the vacuolar membrane and the pre-vacuolar compartment (PVC), respectively. A vam3 mutant was characterized by the accumulation of numerous small vesicles instead of one large vacuole, and defective maturation of vacuolar proteins, including carboxypeptidase Y (CPY) and alkaline phosphatase. This suggests that Vam3p is involved in homotypic vacuolar fusion (Darsow et al., 1997; Wada et al., 1997). On the other hand, pep12 mutants primarily had a single large vacuole, but severely defective CPY delivery; this indicates that Pep12p plays a role in transport between the PVC and the vacuole (Becherer et al., 1996; Cowles et al., 1997). These results indicated that Vam3p and Pep12p are required at different steps of the vacuolar transport pathway.

In Arabidopsis, the Qa-SNAREs VAM3/SYP22 and PEP12/SYP21 were initially isolated as proteins that could complement the yeast vam3 and pep12 mutants, respectively. Immunoelectron microscopy showed that VAM3/SYP22 is localized on vacuolar membranes in shoot apical meristem cells, and on the PVC in root cells (Bassham et al., 1995; Sato et al., 1997; da Silva Conceicao et al., 1997; Sanderfoot et al., 1999). This dual localization was confirmed in living cells of transgenic plants expressing VAM3/SYP22 tagged with either green fluorescent protein (GFP) or monomeric red fluorescent protein (mRFP) under the control of the VAM3/SYP22 promoter (Ebine et al., 2008; Hamaji et al., 2009). VAM3/SYP22 was also identified as the gene responsible for the sgr3 (shoot gravitropism 3) and ssm (short stem and midrib) mutants, which showed a defective shoot gravitropic response and hampered shoot elongation, respectively (Yano et al., 2003; Ohtomo et al., 2005). These mutations were caused by mis-sense substitution of an amino acid (sgr3) and insertion of a short peptide (ssm) in the VAM3/SYP22 sequence. The VAM3/SYP22 protein forms a SNARE complex with VTI11 (Qb-SNARE), SYP5 (Qc-SNARE) and VAMP727 (R-SNARE); this complex plays critical roles in vacuolar traffic and vacuole morphogenesis (Sanderfoot et al., 2001a; Ebine et al., 2008).

PEP12/SYP21 has also been reported to interact with VTI11 and SYP5 SNAREs (Sanderfoot et al., 2001a). Over-expression of PEP12/SYP21 caused PVC trapping of a sorting receptor, a soluble cargo protein, and a membrane cargo protein destined for lytic vacuoles. Over-expression of VAM3/SYP22 did not result in a distinct phenotype. This suggests that PEP12/SYP21 plays a role in transport between the PVC and lytic vacuoles in plants (Foresti et al., 2006). Sanderfoot et al. (2001b) further reported that gene disruption of either PEP12/SYP21 or VAM3/SYP22 caused male gametophytic lethality in Arabidopsis. These results implied that VAM3/SYP22 and PEP12/SYP21 had essentially non-redundant functions in the vacuolar transport pathway in plants, similar to findings in yeast.

Recently, T-DNA insertion mutants of VAM3/SYP22 were identified in the Colombia (Col-0) accession. They were viable and showed pleiotropic phenotypes, including semi-dwarfism, accumulation of thioglucoside glucohydrolases (also called myrosinase) 1 and 2 (TGG1 and TGG2), a poorly developed vascular network in the leaves, and an abnormal distribution of auxin maxima (possibly be caused by non-polarized distribution of the auxin efflux carrier, PIN-FORMED 1) (Ohtomo et al., 2005; Ueda et al., 2006; Shirakawa et al., 2009). Phylogenetic analyses revealed that VAM3/SYP22 and PEP12/SYP21 are grouped together with the single green algal sequence as the immediate outgroup. This suggests that VAM3/SYP22 and PEP12/SYP21 may be derived from lineage-specific duplication. If this is the case, the specific complementing abilities of VAM3/SYP22 and PEP12/SYP21 for vam3 and pep12 yeast mutants probably result from functional convergence (Dacks et al., 2008). Thus, experimental results are inconsistent concerning the functional redundancy and essentiality of VAM3/SYP22 and PEP12/SYP21. Clarification of these inconsistencies is required in order to understand the precise functions of these important genes.

Here we have re-evaluated the effects of mutations in the VAM3/SYP22 and PEP12/SYP21 genes in Arabidopsis. We isolated a T-DNA insertion mutant for PEP12/SYP21 (pep12) that showed no visible abnormality. Homozygous double mutant plants (vam3-1−/−pep12−/−) could not be isolated. Moreover, promoter swapping analysis showed that expression of a GFP–PEP12/SYP21 fusion under the control of the VAM3/SYP22 promoter complemented all the phenotypes of the vam3-1 mutant completely. These results indicate that the functions of VAM3/SYP22 and PEP12/SYP21 are redundant and interchangeable.

Results

A double mutant of vam3-1 and pep12 shows impaired fertilization competence

A mutant of VAM3/SYP22 (vam3-1) in the Col-0 accession that contained a T-DNA insertion in the 5th exon has been described previously (Figure 1a). This mutant showed multiple abnormal phenotypes, including serrated wavy leaves, semi-dwarfism, resistance to salt stress, and late flowering (Ohtomo et al., 2005; Ueda et al., 2006; Hamaji et al., 2009; Shirakawa et al., 2009). cDNA from the vam3-1 mutant contained transcripts for VAM3/SYP22 that were smaller than wild-type transcripts. The mutant transcripts encoded a non-functional VAM3/SYP22 that lacked 23 amino acids in the SNARE motif (Figure 1b). We also used a Col-0 T-DNA mutant for PEP12/SYP21 (pep12) to comprehensively analyze post-Golgi SNARE functions. We obtained a mutant of PEP12/SYP21 (pep12) in the Col-0 accession with a T-DNA insertion in the 5th exon, and no PEP12/SYP21 transcripts were detectable in this mutant by RT-PCR (Figure 1a,b). The lack of PEP12/SYP21 transcripts was confirmed using two other sets of primers (data not shown). At the seedling stage, vam3-1 and pep12 mutant plants did not exhibit any visible abnormal phenotypes and grew normally (Figure 1c). By 35 days after germination (DAG), vam3-1 exhibited the abnormal phenotypes described above, while pep12 showed no detectable abnormal phenotypes (Figure 1d). These results indicate that neither VAM3/SYP22 nor PEP12/SYP21 is essential for viability of the Col-0 accession. Thus, we were able to perform further analysis of the functions of VAM3/SYP22 and PEP12/SYP21.

Figure 1.

 Phenotypes of vam3-1 and pep12 single mutants.
(a) Schematic structures of VAM3/SYP22 and PEP12/SYP21 genes and positions of T-DNA insertions in vam3-1 and pep12 mutants.
(b) RT-PCR analysis of expression of VAM3/SYP22 and PEP12/SYP21 genes in the vam3-1 and pep12 mutants. The VAM3/SYP22 transcript detected in the vam3-1 mutant was smaller than the wild-type transcript. Expression of PEP12/SYP21 was not detected in the pep12 mutant.
(c) Phenotypes of vam3-1 and pep12 mutants at the seedling stage. Fifteen-day-old wild-type and mutant plants grown on MS plates are shown. DAG, days after germination; MS, Murashige and Skoog medium.
(d) Phenotypes of vam3-1 and pep12 mutants at 35 days old. At this stage, vam3-1 shows serrated wavy leaves and a late-flowering phenotype. There were no visible abnormal phenotypes in the pep12 mutant.

We first aimed to identify a double mutant for vam3-1 and pep12. However, we did not find any double homozygous mutants (vam3-1−/−pep12−/−) in progeny derived from self-pollinated plants that were heterozygous for vam3-1 and homozygous for pep12 (vam3-1+/−pep12−/−). The observed segregation ratio (vam3-1+/+pep12−/−: vam3-1+/−pep12−/−: vam3-1−/−pep12−/−) of 83:11:0 suggested that the double mutant was not embryonically lethal, but was impaired in fertilization competence (Table 1). The VAM3/SYP22 and PEP12/SYP21 genes are both located on the 5th chromosome. In order to perform a precise cross-pollination analysis between wild-type and double mutant plants, we generated a double heterozygous mutant (F1vam3-1+/−pep12+/−) with both vam3-1 and pep12 mutations on the same chromosome; this was generated by cross-pollination between vam3-1+/−pep12−/− mutant and wild-type plants. Determination of the genotype of the progeny derived from cross-pollination between F1vam3-1+/−pep12+/− (female) and wild-type (male) indicated that the female gametophytes (vam3-1pep12) were not lethal, but the efficiency of vam3-1pep12 fertilization was lower than that of wild-type (Table 2). On the other hand, cross-pollination of wild-type (female) and F1vam3-1+/−pep12+/− (male) showed that the fertilization competence of the male gametophytes (vam3-1pep12) was severely impaired (Table 2). Next, we investigated whether the single mutation (vam3-1 or pep12) also impeded fertility. Cross-pollination analysis showed that the fertilization competence of male and female gametophytes was also reduced in the vam3-1 mutant, but both male and female gametophytes showed normal fertility in the pep12 mutant (Table 3). These results imply that VAM3/SYP22 plays a more important role than PEP12/SYP21 in the fertilization process, and that VAM3/SYP22 and PEP12/SYP21 have redundant functions.

Table 1.   Segregation of the vam3-1 mutation progeny from heterozygous vam3-1 and homozygous pep12 mutants (vam3-1+/−pep12−/−)
GenotypeDeduced genotype of gametophyteNumber
  1. Progeny derived from the self-pollination of heterozygous vam3-1, homozygous pep12 mutant plants (vam3-1+/−pep12−/−) (VAM3/vam3-1 pep12/pep12 female ×VAM3/vam3-1 pep12/pep12 male) were genotyped by PCR. The number of progeny of each genotype is indicated.

VAM3/VAM3 pep12/pep12VAM3 pep1283
VAM3/vam3-1 pep12/pep12VAM3 pep12 or vam3-1 pep1211
vam3-1/vam3-1 pep12/pep12vam3-1 pep120
Table 2.   Cross-pollination analysis between wild-type and double mutant plants
GenotypeDeduced genotype of gametophyteNumber
  1. Progeny derived from reciprocal crosses between wild-type Col-0 and F1vam3-1+/−pep12+/− in which the vam3-1 and pep12 mutations are located on the same chromosome. Genotyping was performed by PCR.

F1vam3-1+/−pep12+/− female (F1 progeny from VAM3/vam3-1pep12/pep12× Col-0) × Col-0 male
 VAM3/VAM3 PEP12/pep12VAM3 pep1238
 VAM3/vam3-1 PEP12/PEP12vam3-1 PEP1237
 VAM3/vam3-1 PEP12/pep12vam3-1 pep1230
 VAM3/VAM3 PEP12/PEP12VAM3 PEP1248
Col-0 female × F1vam3-1+/−pep12+/− male (F1 progeny from VAM3/vam3-1pep12/pep12× Col-0)
 VAM3/VAM3 PEP12/pep12VAM3 pep1247
 VAM3/vam3-1 PEP12/PEP12vam3-1 PEP1231
 VAM3/vam3-1 PEP12/pep12vam3-1 pep122
 VAM3/VAM3 PEP12/PEP12VAM3 PEP1270
Table 3.   Cross-pollination analysis between wild-type and vam3-1 or pep12 single mutant plants
GenotypeDeduced genotype of egg/spermNumber
  1. Progeny derived from reciprocal crosses between wild-type Col-0 and vam3-1 or pep12 single mutants were genotyped by PCR. The number of progeny of each genotype is indicated.

VAM3/vam3-1 female × Col-0 male
 VAM3/vam3-1vam3-158
 VAM3/VAM3VAM3104
PEP12/pep12 female × Col-0 male
 PEP12/pep12pep1291
 PEP12/PEP12PEP1293
Col-0 female ×VAM3/vam3-1 male
 VAM3/vam3-1vam3-144
 VAM3/VAM3VAM3137
Col-0 female ×PEP12/pep12 male
 PEP12/pep12pep1222
 PEP12/PEP12PEP1218

VAM3/SYP22 and PEP12/SYP21 are interchangeable

To analyze the functions and subcellular locations of VAM3/SYP22 and PEP12/SYP21, we constructed two vectors that encoded fusion proteins comprising GFP and the genomic sequence of either VAM3/SYP22 or PEP12/SYP21. These are referred to as proVAM3:GFP-VAM3 and proPEP12:GFP-PEP12, respectively. We introduced these constructs into the vam3-1 mutant (Figure 2a). All macroscopic phenotypes of vam3-1 were rescued by expression of proVAM3:GFP-VAM3/SYP22 (Figure 2b). In these complemented plants, GFP–VAM3/SYP22 was localized on the vacuolar membrane and occasionally on small dot-like structures (Figure 2c). On the other hand, expression of proPEP12:GFP-PEP12/SYP21 did not rescue the phenotypes of the vam3-1 mutant. Fluorescence analysis showed that GFP–PEP12/SYP21 was localized on bright dot-like structures, assumed to be PVCs, and also on the vacuolar membrane (Figure 2d). These results suggest that twofold over-expression of PEP12/SYP21 had no effect on the vam3-1 mutation, although PEP12/SYP21 was localized on the vacuolar membrane.

Figure 2.

 Over-expression of GFP–PEP12/SYP21 under the control of a native promoter did not complement the vam3-1 phenotypes.
(a) Schematic structures of GFP-fused VAM3/SYP22 and PEP12/SYP21 genes under the control of their respective promoters.
(b) Phenotypes of vam3-1 mutants with or without complementation by GFP-tagged VAM3/SYP22 or PEP12/SYP21. Top panel: 30-day-old wild-type and vam3-1 mutants expressing either GFP–VAM3/SYP22 (lines 3-1 and 11-1) or GFP–PEP12/SYP21 (lines 2-1 and 16-1). Bottom panel: rosette leaves from individual plants. Multiple phenotypes of the vam3-1 mutant were rescued by GFP-tagged VAM3/SYP22, indicating that this chimeric protein is functional.
(c) GFP–VAM3/SYP22 expressed under the control of its own promoter in a vam3-1 mutant was predominantly localized on the vacuolar membrane in root epidermal cells.
(d) Fluorescent images of GFP–PEP12/SYP21 expressed under the control of own promoter in a vam3-1 mutant. The protein was located predominantly on punctate PVC structures in root epidermal cells, but was also observed on the vacuolar membrane.
Insets in (c) and (d) show magnifications of selected regions. Scale bars = 10 μm.

The dual localization of PEP12/SYP21 to PVCs and the vacuolar membrane was similar to that of VAM3/SYP22. To test whether the non-coding regions affected the complementation/suppression ability of the vam3-1 mutation, we replaced the promoter and terminator regions of VAM3/SYP22 with those of PEP12/SYP21, and those of PEP12/SYP21 with those of VAM3/SYP22. These constructs are referred to as proPEP12:GFP-VAM3/SYP22 and proVAM3:GFP-PEP12/SYP21, respectively (Figure 3a). Interestingly, expression of proVAM3:GFP-PEP12/SYP21 completely rescued all macroscopic phenotypes of vam3-1, including the wavy serrated leaves, the semi-dwarfism, and the late flowering (Figure 3b). Fluorescence analysis showed that GFP–PEP12/SYP21 was localized to the dot-like structures and the vacuolar membrane; thus, the VAM3/SYP22 promoter and terminator regions did not affect localization (Figure 3c). On the other hand, expression of proPEP12:GFP-VAM3/SYP22 did not rescue any of the visible phenotypes of vam3-1 (Figure 3b,d). Fluorescence analysis showed that GFP–VAM3/SYP22 was predominantly localized on the vacuolar membrane. Thus, the promoter and terminator regions of VAM3/SYP22 are required to enable both VAM3/SYP22 and PEP12/SYP21 to complement the vam3-1 mutation.

Figure 3.

 Expression of GFP–PEP12/SYP21 under the control of the VAM3/SYP22 promoter completely complements the vam3-1 phenotypes.
(a) Schematic structures of GFP-fused VAM3/SYP22 and PEP12/SYP21 genes used in the promoter swapping analysis. The VAM3/SYP22 and PEP12/SYP21 promoters and 3′ UTR regions were swapped to create proVAM3:GFP-PEP12/SYP21 (left) and proPEP12:GFP-VAM3/SYP22 (right).
(b) Phenotypes of vam3-1 mutants with or without complementation by proPEP12:GFP-VAM3/SYP22 or proVAM3:GFP-PEP12/SYP21. Top panel: 30-day-old wild-type and vam3-1 mutants expressing either proPEP12:GFP-VAM3/SYP22 (lines 4-2, 6-4 and 10-3) or proVAM3:GFP-PEP12/SYP21 (lines 3-5, 10-4 and 13-1). Bottom panel: rosette leaves from individual plants. Multiple phenotypes of the vam3-1 mutant were completely rescued by proVAM3:GFP-tagged PEP12/SYP21.
(c) GFP–PEP12/SYP21, expressed under the control of the VAM3/SYP22 promoter in the vam3-1 mutant, was localized on the vacuolar membrane and PVC-like structures in root epidermal cells.
(d) GFP–VAM3/SYP22, expressed under the control of the PEP12/SYP21 promoter in the vam3-1 mutant, was located on PVC-like punctate structures and on the vacuolar membrane in root epidermal cells.
Insets in (c) and (d) show magnifications of selected regions. Scale bars = 10 μm.

It has been reported that abnormally high levels of the myrosinases TGG1 and TGG2 in vam3 mutants was caused by excessive differentiation of myrosin cells (Ueda et al., 2006). To investigate whether expression of proVAM3:GFP-PEP12/SYP21 could also complement the abnormal myrosinase accumulation in the vam3-1 mutant, we analyzed the amount of myrosinase in rosette leaves by Western blotting. As shown in Figure 4, TGG1 accumulation in vam3-1 was rescued by expression of proVAM3:GFP-VAM3/SYP22 and proVAM3:GFP-PEP12/SYP21, but was not, or was only partially, rescued with expression of proPEP12:GFP-VAM3/SYP22 or proPEP12:GFP-PEP12/SYP21 (Figure 4). These results also showed that the PEP12/SYP21 and VAM3/SYP22 proteins are interchangeable. However, the promoter/terminator of VAM3/SYP22 was essential for conferring function on PVC/vacuolar Qa-SNAREs.

Figure 4.

 Myrosinase (TGG) accumulation in vam3-1 mutants was complemented by expression of proVAM3:GFP-PEP12/SYP21.
Immunoblot analyses with anti-TGG1 and anti-TGG2 antibodies were performed on lysates from 30-day-old plants with the indicated genotypes. The sample numbers correspond to plants shown in Figures 2(b) and 3(b).

VAM3/SYP22 and PEP12/SYP21 are expressed throughout the Arabidopsis plant

To compare the activity of the VAM3/SYP22 and PEP12/SYP21 promoters, we performed RT-PCR on cDNA derived from RNA extracted from leaves, and examined the expression levels of GFP–VAM3/SYP22 and GFP–PEP12/SYP21 under the control of the VAM3/SYP22 and PEP12/SYP21 promoters and terminators. We did not find any difference in the expression of mRNA between the two promoters (Figure 5a). Next, we examined the accumulation of GFP-tagged proteins in seedling and leaf extracts by immunoblotting with an anti-GFP antibody. Curiously, expression of GFP–PEP12/SYP21 under the control of either the PEP12/SYP21 or VAM3/SYP22 promoters and terminators was almost undetectable in seedlings (Figure 5b), and completely undetectable in leaves (Figure 5c). Nevertheless, expression of proVAM3:GFP-PEP12/SYP21 was able to complement the vam3-1 phenotypes, and GFP fluorescence was observed in root and leaf tissues (Figures 3 and 7). These results imply that tissue- or stage-specific expression under the control of the VAM3/SYP22 promoter is sufficient to complement the vam3-1 mutation. In order to obtain insight into the tissue specificity of the VAM3/SYP22 promoter, we created a transgenic Arabidopsis plant that expressed a fusion protein of β-glucuronidase (GUS) and the 1st exon of either VAM3/SYP22 or PEP12/SYP21 under the control of its own promoters and without a terminator. GUS staining patterns were compared between these transgenic plants. We found that both genes were expressed abundantly and ubiquitously in all tissues examined, including embryos in late developmental stages, seedlings, leaves and flowers (Figure 6). Moreover, both genes were expressed in male and female gamete cells (Figure 6f,l). The level of expression of VAM3/SYP22 was slightly higher than that of PEP12/SYP21. These expression patterns were also confirmed in transgenic Arabidopsis expressing GFP-tagged VAM3/SYP22 and PEP12/SYP21 under the control of their own promoters. Thus, the VAM3/SYP22 and PEP12/SYP21 genes are expressed in the same tissues, based on our observations. The multiple abnormal phenotypes of the vam3-1 mutant might be explained by specific cells or developmental stages that show only VAM3/SYP22 expression, but these have not been identified.

Figure 5.

 The VAM3/SYP22 promoter does not cause over-expression of GFP–PEP12/SYP21.
(a) Expression of GFP-fused genes in leaves was analyzed by RT-PCR of transcripts from 30-day-old plants with the indicated genotypes. TUB, tubulin transcript (loading control).
(b, c) The amount of GFP-tagged protein in leaves was analyzed by SDS-PAGE and subsequent immunoblotting with anti-GFP monoclonal antibodies (αGFP) using lysates from 10-day-old seedlings (b) or leaves of 30-day-old plants (c) with the indicated genotypes.
The sample numbers correspond to plants shown in Figures 2(b) and 3(b).

Figure 7.

 mRFP–VAM3/SYP22 co-localized with GFP–PEP12/SYP21 on vacuolar membrane and PVCs.
Confocal laser scanning microscopy was used to capture fluorescence images in plants that expressed mRFP–VAM3/SYP22 (magenta) and GFP–PEP12/SYP21 (green) under the control of native promoters. (a) Root epidermal cells (differentiation zone), (b) root epidermal cells (division zone), (c) 10-day-old rosette leaves, and (d) 30-day-old rosette leaves. In merged images, co-localization is indicated by the absence of color (white). Arrowheads indicate some PVCs in which only GFP–PEP12/SYP21 (green) is detected. Scale bars = 10 μm.

Figure 6.

 Promoter–reporter assay for VAM3/SYP22 and PEP12/SYP21.
The cDNA for β-glucuronidase (GUS) was connected to the coding region of VAM3/SYP22 (top row) or PEP12/SYP21 (bottom row) and introduced into Arabidopsis. Both proteins were expressed throughout the development of Arabidopsis plants, including maturing embryos (a, g), 10-day-old seedlings (b, h), 30-day-old rosette leaves (c, i), 10-day-old roots (d, j), apices of shoots (e, k) and flowers (f, l).

VAM3/SYP22 and PEP12/SYP21 co-localize on PVC and vacuolar membranes in the same cells

In protoplasts, VAM3/SYP22 and PEP12/SYP21 expressed under the control of the 35S promoter were shown to co-localize on PVCs and the vacuolar membrane (Uemura et al., 2004). To minimize mis-localization due to over-expression and to determine the precise subcellular localization in the same cells, we generated transgenic plants that expressed both mRFP–VAM3/SYP22 and GFP–PEP12/SYP21, each under the control of its own promoter and terminator, in the vam3-1 background. These proteins maintained their wild-type functions, as indicated by their ability to complement the vam3-1 mutation. In root cells, GFP–PEP12/SYP21 was located predominantly on mobile punctate PVCs in the cytoplasm and on the vacuolar membrane. On the other hand, mRFP–VAM3/SYP22 was localized predominantly on the vacuolar membrane. Localization of mRFP–VAM3/SYP22 on PVCs was also observed, but this population is quite low. The merged images show that mRFP–VAM3/SYP22 and GFP–PEP12/SYP21 almost completely co-localized on both PVCs and the vacuolar membranes (Figure 7a,b). However, we also observed PVCs on which only GFP–PEP12/SYP21 was present (Figure 7b, arrowheads). These PVCs could potentially possess functions that are distinct from those of PVCs with both proteins. In young and mature rosette leaves, very similar patterns of subcellular co-localization were observed (Figure 7c,d). The considerable co-localization of mRFP–VAM3/SYP22 and GFP–PEP12/SYP21 is consistent with the possibility that these proteins have interchangeable functions in vacuolar transport.

Discussion

SYP2 proteins mediate vacuolar transport in a redundant manner

VAM3/SYP22 and PEP12/SYP21 were originally isolated by screening for genes that complemented yeast vam3 and pep12 mutants, respectively. The subcellular localization of VAM3/SYP22 and PEP12/SYP21 was reminiscent of that of yeast Vam3p and Pep12p: VAM3/SYP22 is located mainly on the vacuolar membrane and occasionally on the PVCs, and PEP12/SYP21 is predominantly located on the PVCs (Sato et al., 1997; da Silva Conceicao et al., 1997; Sanderfoot et al., 1999; Uemura et al., 2002, 2004). Analysis of Arabidopsis disruption mutants for VAM3/SYP22 and PEP12/SYP21 and over-expression of these genes in leaves or protoplasts in Nicotiana tabacum suggested that VAM3/SYP22 and PEP12/SYP21 had essential and distinct functions (Sanderfoot et al., 2001b; Foresti et al., 2006). Here, we have re-evaluated the functions of VAM3/SYP22 and PEP12/SYP21 in planta. Bioimaging analysis clearly indicated that VAM3/SYP22 and PEP12/SYP21 co-localized almost completely on the vacuolar membrane and the PVCs. Moreover, genetic analysis revealed that VAM3/SYP22 and PEP12/SYP21 had interchangeable functions. In yeast, Vam3p and Pep12p were shown to be non-essential, and over-expression of Pep12p suppressed the vam3 mutant phenotypes. Likewise, over-expression of Vam3p suppressed the pep12 mutant phenotypes (Darsow et al., 1997; Gotte and Gallwitz, 1997). These observations led to the suggestion that Arabidopsis VAM3/SYP22 and PEP12/SYP21 may have a relationship similar to that observed for yeast Vam3p and Pep12p. However, recent comparative genomics analysis has suggested that VAM3/SYP22 and PEP12/SYP21 arose by plant lineage-specific duplication. Thus, the ability of Arabidopsis SYP2 proteins to complement vam3 or pep12 yeast mutants may be a result of functional convergence (Dacks et al., 2008).

Although VAM3/SYP22 and PEP12/SYP21 have 65% amino acid identity, there were clear differences in their behaviors. GFP–VAM3/SYP22 was easily detected by microscopy and immunoblotting (Figures 2c, 3d and 5). On the other hand, GFP–PEP12/SYP21 was not detected in leaves by immunoblotting, but was observed by fluorescence microscopy (Figures 2d, 3c and 5). An almost identical result was obtained when we performed immunoblotting using a polyclonal anti-GFP antibody (data not shown). It is not clear why GFP–PEP12/SYP21 was difficult to detect by immunoblotting. Perhaps less GFP–PEP12/SYP21 was required for detection by GFP fluorescence microscopy than by immunoblotting. Alternatively, GFP–PEP12/SYP21 may be unstable and easily degraded during the extraction process used for immunoblotting.

The question remains why the single vam3-1 and pep12 mutations do not cause lethality in the Col-0 accession. This is probably due to the redundant functions of VAM3/SYP22 and PEP12/SYP21, which are both expressed throughout almost all Col-0 tissues. The expression pattern may differ among Arabidopsis accessions, which would explain why mutations of an SYP2 gene caused male sterility in the Arabidopsis accession used previously (Sanderfoot et al., 2001b). In Col-0, male gametes with the double mutation for vam3-1 and pep12 maintained fertility, although their fertilization competence was severely impaired. We could not clarify which stage of pollen development, pollen tube growth or fertilization was impaired in vam3-1 pep12 gametes in this study. A mutant for VCL1 (VACUOLELESS), which encodes a subunit of the tethering complex, homotypic fusion and vacuole protein sorting (HOPS), and functions in the vacuolar transport pathway, showed embryonic lethality, but was intact for gametophyte fertilization competence (Rojo et al., 2001). VCL1 also localizes to the tonoplast and the PVCs, and interacts with VAM3/SYP22 and PEP12/SYP21 (Rojo et al., 2003). These results imply that the vacuolar transport pathway in gametes may not play an essential role in fertilization. However, the male gametes that harbored double mutations for vam3-1 and pep12 did have severely impaired fertilization competence. One explanation could be that vacuolar/PVC Qa-SNAREs might play other roles in addition to their role in membrane fusion. For example, the plasma membrane Qa-SNARE SYP121 interacts with KC1 (a regulatory K+ channel subunit) and AKT1 (a Kv-like K+ channel). This tripartite SNARE–K+ channel complex regulates gating for the K+ channel, an unexpected role for plasma membrane Qa-SNAREs (Honsbein et al., 2009). SYP2 proteins may also interact with ion channels located on the vacuolar membrane and regulate their activities, which might be important for fertilization. Searching for SYP2-interacting proteins and analysis of their functions will be an interesting future project.

SYP2 proteins in Arabidopsis have interchangeable functions

Arabidopsis expresses three members of the SYP2 group (Qa-SNAREs): PEP12/SYP21, VAM3/SYP22 and PLP/SYP23. VAM3/SYP22 has 65% amino acid identity to PEP12/SYP21 and 74% identity to PLP/SYP23. PLP/SYP23 lacks the transmembrane region in the Col-0 accession (Zheng et al., 1999) and is therefore considered non-functional, although it is ubiquitously expressed throughout all tissues. However, in yeast, Pep12p also lacks the transmembrane region, but its over-expression complements the phenotype of the pep12 mutant (Gerrard et al., 2000). Therefore, it is possible that PLP/SYP23 maintains the ability to form a complex with other SNAREs in Col-0 Arabidopsis plants.

The vam3-1 mutant showed no visible abnormality in root development. This may be due to the expression of PEP12/SYP21 and PLP/SYP23 throughout root tissues. On the other hand, there are multiple abnormal phenotypes in the aerial parts of the vam3-1 mutant, despite ubiquitous expression of PEP12/SYP21 and PLP/SYP23. It is not clear why the vam3-1 mutant showed such pleiotropic phenotypes only in the aerial parts of the plant. Our promoter swapping analysis showed that expression of proVAM3:GFP-PEP12/SYP21 completely suppressed all of the vam3-1 mutant phenotypes; this suggested that the difference in phenotypes between the vam3-1 and pep12 mutants may be attributed to differences in the activities of their promoters and/or terminators. Currently, identification of specific tissues or developmental stages that express only VAM3/SYP22 has been unsuccessful. Further analyses of the detailed cellular expression patterns of SYP2 family members will be necessary to unveil how the functions of SYP2 proteins are controlled, and this would also assist in understanding how the vacuolar transport pathway contributes to plant development.

Taken together, our results show that the SYP2 family proteins in Arabidopsis have redundant and interchangeable functions in vacuolar transport. VAM3/SYP22 is the primary contributor to this pathway, but we have not ruled out the possibility that each SYP2 protein may have additional distinct functions.

Experimental Procedures

Plant materials and plasmids

The vam3-1 (SALK_060946) and pep12 (SAIL_84_H07) mutants were obtained from the Arabidopsis Biological Resource Center (ABRC). They were backcrossed at least three times with wild-type Arabidopsis thaliana (Col-0). cDNAs for fluorescent proteins (GFP and mRFP) were fused to VAM3/SYP22 or PEP12/SYP21 sequences by fluorescence tagging of full-length proteins (Tian et al., 2004). Either the GFP or mRFP cDNA sequence was inserted in front of the start codon of VAM3/SYP22 (including 3.2 kb of the 5′ flanking sequence and 1.2 kb of the 3′ flanking sequence) or PEP12/SYP21 (including 2.8 kb of the 5′ flanking sequence and 1.1 kb of the 3′ flanking sequence). The same promoter and 3′ UTR regions described above were used for promoter swapping and for promoter–β-glucuronidase (GUS) reporter analysis. An amplified chimeric fragment was subcloned into the binary vector pGWB1 (a kind gift from T. Nakagawa, Center for Integrated Research in Science, Shimane University, Japan), which was used for transforming Arabidopsis plants. Transformation of Arabidopsis plants was performed by floral dipping with Agrobacterium tumefaciens (stain GV3101).

RT-PCR

RNAqueous-4PCR (Ambion, http://www.ambion.com/) was used for extracting total RNA from the rosette leaves of 30-day-old plants expressing wild-type VAM3, mutant vam3-1 and mutant pep12. This RNA was reverse-transcribed into cDNA using SuperScript III reverse transcriptase (Invitrogen, http://www.invitrogen.com/). The primer sets used for PCR amplification were: VAM3/SYP22 forward (5′-GAGATTAGAGAAAACTCCGA-3′) and VAM3/SYP22 reverse (5′-TAGTTCCAGTCATTGATGCC-3′); PEP12/SYP21 forward (5′-GGGAAAACTTATAATTCACC-3′) and PEP12/SYP21 reverse (5′-TCCATAGATTCGCTTGATGC-3′); TUA3 forward (5′-GGACAAGCTGGGATCCAGG-3′) and TUA3 reverse (5′-CGTCTCCACCTTCAGCACC-3′); GFP forward (5′-AGCAAGGGCGAGGAGCTGTT-3′) and GFP reverse (5′TCGTCCATGCCGTGAGTGAT-3′). PCR conditions were as follows: 94°C for 2 min, 34 cycles at 94°C for 30 sec, 55°C for 30 sec and 72°C for 1 min, then 72°C for 10 min.

Light microscopy

For fluorescence protein analysis, we used an Olympus BX51 fluorescence microscope (http://www.olympus-global.com/) equipped with a confocal scanner unit (model CSU10, Yokogawa Electric, http://www.yokogawa.co.jp/) and a cooled CCD camera (model ORCA-AG, Hamamatsu Photonics, http://jp.hamamatsu.com/). Images were processed using IPLab software (BD Biosciences, http://www.bdbiosciences.com/). A confocal laser scanning microscope (model LSM710; Zeiss, http://www.zeiss.com/) was used for observation of transgenic plants that expressed both GFP–PEP12/SYP21 and mRFP–VAM3/SYP22.

Immunoblotting procedure and antibodies

Seedlings of 10-day-old plants or leaves of 30-day-old plants were homogenized in a grinding buffer [12 ml of 50 mm HEPES/KOH, pH 7.5, 5 mm MgCl2, 5 mm EGTA, 250 mm sorbitol, 1 mm DTT, 1% polyvinylpyrolidone, 1% ascorbic acid and protease inhibitors (Roche, http://www.roche.com)]. The homogenates were centrifuged at 2000 g. The supernatants were collected and subjected to 12.5% SDS-PAGE for separation, and then transferred onto PVDF membranes for Western blotting. The membranes were treated with antibodies against TGG1 (1:5000, kindly donated by Dr I. Hara-Nishimura, Graduate School of Science, Kyoto University, Japan), TGG2 (1:5000, kindly donated by Dr I. Hara-Nishimura) or GFP (1:500; Nacalai Tesque Inc., http://www.nacalai.co.jp/). Immunoreactive signals were detected using an chemiluminescence detection system (SuperSignal West Femto kit, Pierce Biotechnology, http://www.piercenet.com/).

Acknowledgements

We thank the Salk Institute for providing T-DNA insertion mutants of Arabidopsis. We also thank I. Hara-Nishimura (Graduate School of Science, Kyoto University, Japan) and T. Nakagawa (Center for Integrated Research in Science, Shimane University, Japan) for sharing materials. This work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Ancillary