The SNARE Protein SYP71 Expressed in Vascular Tissues is Involved in Symbiotic Nitrogen Fixation in Lotus japonicus Nodules 1

SNARE (soluble N-ethylmaleimide sensitive factor attachment protein receptor) proteins are crucial for signal transduction and development in plants. Here, we investigate a Lotus japonicus symbiotic mutant defective in one of the SNARE proteins. When in symbiosis with rhizobia, the growth of the mutant was retarded compared with that of the wild-type plant. Although the mutant formed nodules, these exhibited lower nitrogen fixation activity than wild-type. The rhizobia were able to invade nodule cells, but enlarged symbiosomes were observed in the infected cells. The causal gene, designated LjSYP71 , was identified by map-based cloning and shown to encode a Qc-SNARE protein homologous to Arabidopsis thaliana SYP71. LjSYP71 was expressed ubiquitously in shoot, roots, and nodules, and transcripts were detected in the vascular tissues. In the mutant, no other visible defects in plant morphology were observed. Furthermore, in the presence of combined nitrogen, the mutant plant grew almost as well as the wild-type. These results suggest that the vascular tissues expressing LjSYP71 play a pivotal role in symbiotic nitrogen fixation in L. japonicus nodules.


INTRODUCTION
Intracellular membrane fusion in eukaryotic cells involves membrane-associated SNARE (soluble N-ethylmaleimide sensitive factor attachment protein receptor) proteins that contain a conserved coiled-coil domain and are anchored to the membrane by either a C-terminal transmembrane domain or a post-translational addition of lipids Parlati et al., 2000). Transport vesicles carrying cargo fuse with the target endomembrane or the plasma membrane and their contents are delivered to different organelles or secreted to the extracellular space, respectively. A SNARE protein localized on the donor vesicle forms a tetrameric bundle of coiled helices with complementary SNARE proteins in the target membrane, which drives the fusion of transport vesicles with the proper target membrane.
These SNARE proteins are divided into v-SNAREs and t-SNAREs depending on their localization on the trafficking transport vesicle or the target membrane, respectively. However, this nomenclature does not apply to homotypic membrane fusion. Thus the SNARE proteins have been reclassified as R-SNAREs and Q-SNAREs based on the conserved arginine or glutamine residue in the center of the SNARE motif (Fasshauer et al., 1998). Furthermore, Q-SNAREs are subdivided into three classes designated Qa-, Qb-and Qc-SNARE depending on their SNARE motif domains. These Q-SNAREs create a three-helix bundle complex (t-SNARE) that interacts with R-SNARE (v-SNARE) in membrane fusion (Bassham and Blatt, 2008).
Syntaxins are one of the components of the t-SNARE complex. In the Arabidopsis thaliana genome, 24 genes were annotated to encode members of the syntaxin family, and were designated syntaxin of plants (SYP) (Sanderfoot et al., 2000). Based on sequence homology, www.plantphysiol.org on August 30, 2017 -Published by Downloaded from Copyright © 2012 American Society of Plant Biologists. All rights reserved. 6 these 24 genes are primarily divided into eight groups named SYP1 to SYP8, each of which contains one to nine members. Further sequence analysis revealed that the SYP1 to SYP4 groups belong to the Qa-SNAREs subfamily while the SYP5 to SYP8 groups belong to the Qc-SNAREs subfamily (Uemura et al., 2004). All groups play essential roles in various aspects of signaling and development in plants by cooperating with other SNARE proteins (Lipka et al., 2007;Pfeffer, 2007;Bassham and Blatt, 2008).
In legumes, rhizobia generally attach to host plant root hairs, invade the root cortical cells through an infection thread, and inhabit the specialized organ formed on the host plant root, the root nodule. Rhizobia then differentiate into bacteroids that are enclosed by a plant-derived symbiosome membrane, and which reduce atmospheric dinitrogen to ammonia in the nodule cells. Recently, syntaxin SYP132 of the model legume Medicago truncatula was shown to be localized not only to the plasma membrane surrounding the infection thread but also abundantly to the symbiosome membrane, suggesting that MtSYP132 is involved in symbiosome formation (Catalano et al., 2007;Limpens et al., 2009). Intriguingly, this contrasts with the plasma membrane syntaxin SYP132 of Nicotiana benthamiana which is thought to be involved in plant resistance against pathogenic bacteria (Kalde et al., 2007). In another model legume, Lotus japonicus, syntaxin SYP32-1 was shown to be required for normal differentiation of nodule tissues (Mai et al., 2006). These results indicate that SNARE proteins are likely to be crucial for the maintenance of legume-Rhizobium symbiosis. However, with the exception of MtSYP132 and LjSYP32-1, little evidence is available regarding the contributions of SNARE proteins to the symbiotic association.
We have been studying L. japonicus Fixmutants that exhibit lower or no nitrogen fixation activity in order to identify plant genes required for the establishment and maintenance of 8 During plant development, growth of the Ljsyp71 mutant plant was always retarded under symbiotic conditions with the compatible rhizobium Mesorhizobium loti ( Fig. 2A). The Ljsyp71 mutant formed smaller and pale white nodules (Fig. 1), of a similar number to that on the wild-type plant throughout plant development (Fig. 2B). However, the growth of Ljsyp71 nodules was arrested after 4 weeks whereas wild-type plant nodules continued to increase in size ( Fig. 2C). The nitrogen fixation activity of Ljsyp71 nodules was delayed compared to that of wild-type nodules, and the activity level remained lower ( Fig. 2D). At 4 weeks, the activity of Ljsyp71 nodules was approximately 14% that of wild-type nodules (Fig. 2E). After potassium nitrate was supplied to the Ljsyp71 mutant, plant growth almost recovered to wild-type plant levels (Fig. 3). In addition, under both symbiotic and non-symbiotic conditions, no visible morphological changes in vegetative or reproductive organs of the Ljsyp71 plant were observed, apart from the nodulation defects.

Cellular structure of Ljsyp71 mutant nodules
The cellular structure of Ljsyp71 mutant nodules was examined by light and transmission electron microscopy. Twenty-four days after inoculation, toluidine blue-stained infected cells were observed in both Ljsyp71 and wild-type nodules ( Fig. 4A and 4B, respectively). However, some of the infected cells in the Ljsyp71 nodules stained less densely, and a large vacuole observed in the infected cells of wild-type nodules was absent from most of the infected cells of Ljsyp71 nodules. The infected cells of wild-type nodules were packed with endosymbionts enclosed by a symbiosome membrane (Fig. 4C). The symbiosomes of Ljsyp71 nodules were 9 generally enlarged and, consequently, the symbiosome space between the symbiosome membrane and the bacteroids was increased compared to that of wild-type nodules ( Fig. 4C and   4D).

Map-based cloning of the LjSYP71 gene
Ljsyp71-1 was crossed with ecotype Gifu B-129 and linkage analysis was performed using 972 homozygous F2 mutant individuals to enable map-based cloning of the responsible LjSYP71 gene. This gene was primarily delimited between the simple sequence repeat (SSR) markers TM0696 and TM1077 on the upper part of chromosome 5 (Supplemental Fig. S1A). At the time of this experiment, the genome sequence of this region had not been fully covered, but comparison of the L. japonicus genome  with that of Glycine max (Schmutz et al., 2010) predicted that a transformation-competent artificial chromosome (TAC) clone, LjT03L03, would be positioned between the two markers. Direct sequencing of six genes predicted on the TAC clone revealed a single nucleotide mutation at the third exon/intron boundary that caused an error in RNA splicing in one of these genes in the Ljsyp71-1 mutant (Supplemental Fig. S1B). In the LjSYP71-2 mutant, it appeared to carry a non-autonomous DNA transposon composed of approximately 500 base pairs inserted in the fifth exon of the same gene (Supplemental Fig. S1B).
To confirm that the gene in question was responsible for the Ljsyp71 mutant phenotype, we introduced the wild-type complementary DNA (cDNA), prepared from the EST clone as shoot growth in the transformed Ljsyp71-1 mutant (Fig. 5A). Nitrogen fixation (acetylene reduction) activity of the transformed nodules was also recovered to a level comparable to that of wild-type nodules (Fig. 5B). From these results, we concluded that this gene was indeed mutated in the mutant.

Structure of the LjSYP71 protein
The EST clone MR020b12, corresponding to the LjSYP71 gene, was found in the L.
japonicus EST database (http://est.kazusa.or.jp/en/plant/lotus/EST/index.html). The clone contained a nearly full-length cDNA that encoded a polypeptide composed of 265 amino acids with a molecular mass of 29.8 kDa. LjSYP71 appeared to be homologous to syntaxin SYP71 (At3g09740) of A. thaliana (Fig. 6). The AtSYP7 family, unique to plants, has three members, SYP71, SYP72 and SYP73, with largely unknown functions (Sanderfooot et al., 2000).
Sequence analysis of the 54 A. thaliana SNARE genes revealed that the AtSYP7 family belongs to Qc-SNAREs (Uemura et al., 2004). In the deduced LjSYP71 amino acid sequence (Supplemental Fig. S2), the SOSUI domain prediction program (Hirokawa et al., 1998) predicted a trans-membrane domain in the C-terminal region. In addition, the MOTIF program (http://www.genome.jp/tools/motif) detected a SNARE coiled-coil motif centered on a glutamine (Q) residue adjacent to the C-terminal trans-membrane domain. These structural features matched those of Q-SNARE proteins (Fasshauer et al., 1998;Sanderfooot et al., 2000).
Consequently, we designated the gene LjSYP71. Determination of the nucleotide sequence of the EST clone MPD004c03 showed that the deduced amino acid sequence is 86% identical to LjSYP71 (Supplemental Fig. S3). Sequence comparison revealed that the MPD004c03 also belongs to the AtSYP71 protein family (Fig. 6).

Expression of the LjSYP71 gene
LjSYP71 gene expression was analyzed by quantitative real-time PCR. RNA transcripts were detected in all vegetative tissues examined, including stems, leaves, roots and nodules (Fig. 7) as well as in reproductive tissues such as flowers and pods (Supplemental Fig. S3). The highest expression was observed in young roots, though the difference in levels of expression was no more than twice as high as that in other tissues. After inoculation with M. loti, expression levels decreased in roots; furthermore, the level of expression in nodules was usually maintained during their development. These results are consistent with those obtained from expression analysis using macro-array (http://est.kazusa.or.jp/en/plant/lotus/EST/cDNA.html; Kouchi et al., 2004). In situ-hybridization further revealed that the LjSYP71 gene was expressed in vascular tissues in both roots and nodules ( Fig. 8A and 8B). This was also confirmed by promoter-ß-glucuronidase (GUS) fusion experiments. The development of a blue color was observed in the vascular tissues of the hairy roots transformed with the LjSYP71 promoter-GUS fusion gene as well as in the vascular tissues of the nodules borne on these hairy roots ( Fig. 8C and 8E). By contrast, GUS activities were undetectable in the hairy roots transformed with an empty vector and in the nodules formed on these hairy roots ( Fig. 8D and 8F).

DISCUSSION
While studying the Fixmutants of L. japonicus, four genes were identified: SST1, IGN1, FEN1 and SEN1 (Krusell et al., 2005;Kumagai et al., 2007;Hakoyama et al., 2009;Hakoyama et al., 2012). Two further genes, encoding the EFD transcription factor and DNF1, were shown to be essential for symbiotic nitrogen fixation during the investigation of the M. truncatula Fixmutants (Vernié et al., 2008;Wang et al., 2010). All of these genes regulate rhizobial symbiotic nitrogen fixation in various aspects of the developmental process. With the exception of IGN1, all genes are expressed in a nodule-specific manner, and their products function in nodule-infected cells. IGN1, by contrast, is expressed constitutively in all organs, although it is uncertain whether its product plays an important role in nodule-infected cells.
In the present study, we identified another L. japonicus Fixmutant and showed that the mutated gene encodes a Qc-SNARE protein homologous to A. thaliana SYP71. In symbiotic association with M. loti, the Ljsyp71 mutant plant exhibits symptoms of nitrogen deficiency because of the lower nitrogen-fixing activity of the nodules. Inside the Ljsyp71 nodules, enlarged symbiosomes are observed in infected cells like other Fixmutants (Suganuma et al., 2003;Krusell et al., 2005;Hossain et al., 2006;Kumagai et al., 2007). However, in contrast to LjSST1, LjFEN1, LjSEN1, MtEFD and MtDNF1, but similar to LjIGN1, LjSYP71 is expressed in all vegetative tissues. In addition, no obvious morphological defects of the Ljsyp71 mutant were observed under both symbiotic and non-symbiotic conditions, similar to the Ljign1 mutant. This provides further evidence that regulation of rhizobial symbiotic nitrogen fixation requires both nodule-specific genes and genes expressed throughout the plant.
Members of the SYP7 family, inculding LjSYP71, have not been identified in yeast or mammalian SNAREs (Sanderfoot et al., 2000), so are likely to play specific roles in plant development, one of which may be related to their expression in vascular tissues. In A. thaliana, SYP71 is predominantly localized to the plasma membrane where it is thought to function in the 13 secretion process (Alexandersson et al., 2004;Marmagne et al., 2004;Tyrrell et al., 2007;Suwastika et al., 2008). The discovery of the LjSYP71-defective Fixmutant suggests the existence of a systemic regulation of the nitrogen-fixing activity mediated by vascular tissues. A substance in the shoot that affects nitrogen-fixing activity might be secreted via vesicle trafficking into the phloem and transported to infected nodule cells. Alternatively, a substance produced from nitrogen fixation in the nodules might be exported to the plant shoot through the xylem. The nitrogen-fixing activity of the nodules could be impaired by an interruption in the translocation of either substance, which is caused by the LjSYP71 mutation. Guinel (2009) focused on the legume nodule cortex surrounding the central infected zone and highlighted the importance of vascular tissues in the development and regulation of nodules in the legume-Rhizobium symbiosis. Our results present further evidence for the pivotal role played by the vasculature in symbiotic nitrogen fixation.
In A. thaliana, AtSYP71 is expressed in vegetative tissues and its expression is detected in the vascular tissues of the roots (Suwastika et al., 2008). This pattern is similar to that of LjSYP71 in L. japonicus. However, it should be noted that no homozygous T-DNA insertion mutant of A. thaliana SYP71 was isolated from progenies of the SYP71/syp71 heterozygote, suggesting that AtSYP71 is essential for the development of A. thaliana (Suwastika et al., 2008). By contrast, LjSYP71 does not appear to be essential for the development of the L. japonicus plant because under non-symbiotic conditions but supplied with combined nitrogen, the Ljsyp71 mutant grows similarly to the wild-type plant. The cDNA clone, MPD004c03, from the L. japonicus EST database was shown to be homologous to LjSYP71, and the predicted protein belongs to the AtSYP71 family. A previous cDNA array experiment (Kouchi et al., 2004)found that MPD004c03 is also expressed in shoot, roots and nodules. It is likely that LjSYP71 evolved to 14 fulfill an important role in symbiotic nitrogen fixation and that the paralogous gene expressed in parallel with LjSYP71 is required for the correct development of L. japonicus. In nodules, the paralogous gene may have a partially overlapping fucntion, resulting in basal levels of nitrogen fixation.
Some SYP proteins have already been linked to nodulation. In M. truncatula, SYP132 of the Qa-SNARE protein family is located on symbiosome membranes, and is thought to be involved in symbiosome formation (Catalano et al., 2007;Limpens et al., 2009). Another member of the Qa-SNARE protein family, SYP32-1, is involved in symbiotic nitrogen fixation as well as nodule formation of L. japonicus (Mai et al., 2006). However, LjSYP32-1-suppressed transformants display significantly retarded plant growth even when the nutrient medium contains nitrogen. LjSYP32-1 is expressed not only in the vascular tissues of roots and nodules but also in the inner cortical cell layer surrounding the infected zone of nodules and in the meristematic area of developing lateral roots. By contrast, the expression of LjSYP71 is detected in the vascular tissues of roots and nodules, but not in the infected cells of nodules. Thus, it is suggested that the role of LjSYP71 in symbiotic nitrogen fixation differs from those of MtSYP132 and LjSYP32-1.
In conclusion, this study showed that LjSYP71, a component of the SNARE complex that is expressed in vascular tissues, plays an essential role in symbiotic nitrogen fixation in L. japonicus nodules. Further analysis of the role of LjSYP71 will provide novel insights into legume and Rhizobium symbiosis.

Plant and Bacterial Materials
Seeds of L. japonicus ecotype Miyakojima MG-20 were used as wild type. The Ljsyp71-1 mutant was obtained by EMS mutagenesis. Approximately 5,000 seeds were immersed in 0.4% EMS overnight and allowed to germinate. M1 plants were grown in a green-house and M2 seeds collected. A screen for Fixmutants that form nodules but exhibit retarded plant growth was performed in approximately 25,000 M2 plants. A single individual of the M3 self-progeny from the M2 candidate mutant was backcrossed with the parent MG-20. The Ljsyp71-2 mutant was produced by carbon-ion beams as described previously (Tanaka et al., 1997;Magori et al., 2009). Seeds were surface-sterilized and sown in sterilized vermiculite with Mesorhizobium loti MAFF303099, which had been cultured on yeast-mannitol-agar plates for 7 days. The plants were grown in a nitrogen-free nutrient solution in the greenhouse under natural daylight or in a controlled chamber on a 16-h day/8-h night cycle at 26°C as described (Imaizumi-Anraku et al., 1997). Under non-symbiotic conditions, potassium nitrate (10 mM) was added to the nutrient solution.

Acetylene Reduction Assay
The nitrogenase activity of mutant nodules was measured by an acetylene reduction assay, in a closed system with nodulated roots detached from freshly harvested intact plants. Nodulated roots were placed in 20-mL vials and incubated at 25ºC (Suganuma et al., 2003). After 30 min, the amount of ethylene produced was determined by gas chromatography equipped with a flame ionization detector and a column of Porapak N (Waters, Milford, MA).

Light and Electron Microscopy
The cellular structure of the Ljsyp71-1 nodules was observed by light and transmission electon microscopy, and observations were carried out essentially as described previously (Suganuma et al., 2003). Nodules on root segments were fixed in FAA containing 5% (v/v) formaldehyde, 5% (v/v) acetic acid, and 63% (v/v) ethanol. After dehydration, the samples were embedded in Paraplast Plus. Serial microtome sections (10 µm) were stained with 1% (w/v) toluidine blue in 0.5% (w/v) sodium tetraborate (pH 9.0). For transmission electron microscopy, nodules were fixed in 2% glutaraldehyde and post-fixed in 2% osmium tetroxide in 0.1 M sodium phosphate buffer (pH 7.2). After being dehydrated, the samples were embedded in an epoxy resin (Quetol-812, Nisshin EM, Tokyo, Japan). Ultrathin sections were stained with uranium acetate and lead citrate, and were observed under electron microscopy.

Map-based Cloning
The causal gene of the Ljsyp71 mutants was identified by map-based cloning as described previously (Suganuma et al., 2003). Total DNA was extracted using the DNeasy Plant Mini kit (QIAGEN, Hilden, Germany) from leaves of F 2 homozygous mutant plants generated by crossing the mutant with L. japonicus ecotype Gifu B-128, and PCR was carried out with SSR markers . For fine mapping, additional PCR markers were developed on the basis of sequence differences between the two parents. The PCR product was resolved on a non-denaturing 15% polyacrylamide gel and stained with SYBR Green I (TaKaRa, Shiga,

Sequence Analysis
EST clones, MR020b12 and MPD004c03 (Asamizu et al., 2000;Sato et al., 2001;Asamizu et al., 2004), were obtained from The National BioResource Project (L. japonicus and G. max) Office, Department of Agriculture, Miyazaki University, Japan. Plasmides containing of each cDNA clone were isolated and the nucleotide sequence of each clone was determined.

Complementation Test
Wild-type cDNA was introduced into the Ljsyp71-1 mutant: the coding region of the  RNAs were reverse-transcribed using a ReverTra Ace qPCR RT Kit (TOYOBO, Tokyo, Japan).
Quantitative real-time PCR was performed using an iQ SYBR Green Super mix (Bio-Rad, Hercules, CA).

In situ Hybridization
Spatial expression of LjSYP71 in nodules was determined by in situ-hybridization, which was carried out using the method described by Kouchi and Hata (1993). Nodules were fixed in 4% (w/v) paraformaldehyde and 0.25% (w/v) glutaradehyde in 0.1 M sodium phosphate buffer (pH7.4). The sections were hybridized with RNA probes prepared from linearized plasmids with digoxigenin-UTP (Roche Diagnostics) containing the entire coding region of the LjSYP71 gene.

Promoter GUS Analysis
Spatial expression patterns of LjSYP71 in roots and nodules were also studied by promoter GUS analysis. The LjSYP71 promoter and terminator fragments were amplified by PCR from genomic DNA with proF04-F primer: 5'-GGTACCGTACCCACTAACAAACAAGG-3', proF04-R primer:, 5'-GGATCCGGTGTGTGGTGGTGGAGTCA-3', terF04-F primer:  Nodules on transgenic hairy roots were sliced longitudinally with a double-edged razor blade and immersed in the staining solution. After incubation for 2 h at 37 °C in the dark, the stained materials were observed with a light microscope.

Phylogenetic Analysis
Amino acid sequences of A. thaliana syntaxins and LjSYP71 were aligned using the CLUSTALW program (http://www.ddbj.nig.ac.jp). The phylogenetic tree was drawn using the program TreeView32 (Page, 1996).
Sequence data from this article can be found in the GenBank/EMBL/DDBJ data libraries under accession number AB704757.        wild-type Miyakojima seeds were germinated and the seedlings were transferred to vermiculite 6 days after sowing. After transfer, Mesorhizobium loti was inoculated 4 days later. Stems and leaves were harvested at 8 days after inoculation. Roots were harvested 0, 4, and 8 days after inoculation. Nodules were collected from the roots 12, 17, and 22 days after inoculation. The housekeeping gene ubiquitin was used to assess the relative expression of the LjSYP71 gene. All values are means of three biological independent determinations and vertical bars indicate standard errors.