MAG2 and three MAG2-INTERACTING PROTEINs form an ER-localized complex to facilitate storage protein transport in Arabidopsis thaliana

Authors

  • Lixin Li,

    1. Alkali Soil Natural Environmental Science Center, Key Laboratory of Saline–Alkali Vegetation Ecology Restoration in Oil Field (SAVER), Ministry of Education, Northeast Forestry University, Harbin, China
    2. Department of Botany, Graduate School of Science, Kyoto University, Kyoto, Japan
    3. College of Life Science, Northeast Forestry University, Harbin, China
    Search for more papers by this author
    • These authors contributed equally to this work.
  • Tomoo Shimada,

    1. Department of Botany, Graduate School of Science, Kyoto University, Kyoto, Japan
    Search for more papers by this author
    • These authors contributed equally to this work.
  • Hideyuki Takahashi,

    1. Department of Botany, Graduate School of Science, Kyoto University, Kyoto, Japan
    Current affiliation:
    1. School of Environmental Science and Engineering, Kochi University of Technology, Tosayamada, Kochi, Japan
    Search for more papers by this author
  • Yasuko Koumoto,

    1. Department of Botany, Graduate School of Science, Kyoto University, Kyoto, Japan
    Search for more papers by this author
  • Makoto Shirakawa,

    1. Department of Botany, Graduate School of Science, Kyoto University, Kyoto, Japan
    Search for more papers by this author
  • Junpei Takagi,

    1. Department of Botany, Graduate School of Science, Kyoto University, Kyoto, Japan
    Search for more papers by this author
  • Xiaonan Zhao,

    1. Alkali Soil Natural Environmental Science Center, Key Laboratory of Saline–Alkali Vegetation Ecology Restoration in Oil Field (SAVER), Ministry of Education, Northeast Forestry University, Harbin, China
    2. College of Life Science, Northeast Forestry University, Harbin, China
    Search for more papers by this author
  • Baoyu Tu,

    1. Alkali Soil Natural Environmental Science Center, Key Laboratory of Saline–Alkali Vegetation Ecology Restoration in Oil Field (SAVER), Ministry of Education, Northeast Forestry University, Harbin, China
    2. College of Life Science, Northeast Forestry University, Harbin, China
    Search for more papers by this author
  • Hongmin Jin,

    1. Alkali Soil Natural Environmental Science Center, Key Laboratory of Saline–Alkali Vegetation Ecology Restoration in Oil Field (SAVER), Ministry of Education, Northeast Forestry University, Harbin, China
    2. College of Life Science, Northeast Forestry University, Harbin, China
    Search for more papers by this author
  • Zhe Shen,

    1. Alkali Soil Natural Environmental Science Center, Key Laboratory of Saline–Alkali Vegetation Ecology Restoration in Oil Field (SAVER), Ministry of Education, Northeast Forestry University, Harbin, China
    Search for more papers by this author
  • Baoda Han,

    1. Alkali Soil Natural Environmental Science Center, Key Laboratory of Saline–Alkali Vegetation Ecology Restoration in Oil Field (SAVER), Ministry of Education, Northeast Forestry University, Harbin, China
    2. College of Life Science, Northeast Forestry University, Harbin, China
    Search for more papers by this author
  • Meihui Jia,

    1. Alkali Soil Natural Environmental Science Center, Key Laboratory of Saline–Alkali Vegetation Ecology Restoration in Oil Field (SAVER), Ministry of Education, Northeast Forestry University, Harbin, China
    2. College of Life Science, Northeast Forestry University, Harbin, China
    Search for more papers by this author
  • Maki Kondo,

    1. Department of Cell Biology, National Institute for Basic Biology, Okazaki, Japan
    Search for more papers by this author
  • Mikio Nishimura,

    1. Department of Cell Biology, National Institute for Basic Biology, Okazaki, Japan
    Search for more papers by this author
  • Ikuko Hara-Nishimura

    Corresponding author
    1. Department of Botany, Graduate School of Science, Kyoto University, Kyoto, Japan
    Search for more papers by this author

Summary

In Arabidopsis thaliana, MAIGO 2 (MAG2) is involved in protein transport between the endoplasmic reticulum (ER) and the Golgi apparatus via its association with the ER-localized t-SNARE components SYP81/AtUfe1 and SEC20. To characterize the molecular machinery of MAG2-mediated protein transport, we explored MAG2-interacting proteins using transgenic A. thaliana plants expressing TAP-tagged MAG2. We identified three proteins, which were designated as MAG2-INTERACTING PROTEIN 1–3 [MIP1 (At2g32900), MIP2 (At5g24350) and MIP3 (At2g42700)]. Both MIP1 and MAG2 localized to the ER membrane. All of the mag2, mip1, mip2 and mip3 mutants exhibited a defect in storage protein maturation, and developed abnormal storage protein body (MAG body) structures in the ER of seed cells. These observations suggest that MIPs are closely associated with MAG2 and function in protein transport between the ER and Golgi apparatus. MIP1 and MIP2 contain a Zeste–White 10 (ZW10) domain and a Sec39 domain, respectively, but have low sequence identities (21% and 23%) with respective human orthologs. These results suggest that the plant MAG2–MIP1–MIP2 complex is a counterpart of the triple-subunit tethering complexes in yeast (Tip20p–Dsl1p–Sec39p) and humans (RINT1–ZW10–NAG). Surprisingly, the plant complex also contained a fourth member (MIP3) with a Sec1 domain. There have been no previous reports showing that a Sec1-containing protein is a subunit of ER-localized tethering complexes. Our results suggest that MAG2 and the three MIP proteins form a unique complex on the ER that is responsible for efficient transport of seed storage proteins.

Introduction

The endomembrane system comprises several organelles, including the endoplasmic reticulum (ER), Golgi apparatus, endosomes and vacuoles/lysosomes. Protein transport between these organelles is mediated by transport vesicles or carriers. COPII vesicles are involved in anterograde transport of newly synthesized proteins from the ER to the Golgi apparatus (Lee et al., 2004). COPI vesicles function in retrograde protein transport from the Golgi apparatus to the ER during retrieval of ER-resident proteins and/or factors that are necessary for anterograde transport (Lee et al., 2004). Vesicle transport may be divided into several essential steps: budding, transport, tethering, docking and fusion (Bonifacino and Glick, 2004). After budding from the membrane, vesicles may be transported to their destination. Vesicle tethering, the initial interaction between the vesicle and the target membrane, is mediated by tethering factors or multi-subunit tethering complexes (Yu and Hughson, 2010). Tethering factors are typically long or extended molecules, such as p115, GM130 and Uso1p. The multi-subunit tethering complexes that have been reported in yeast and mammals include COG (conserved oligomeric Golgi), GARP (Golgi-associated retrograde protein), TRAPP (transport protein particle), HOPS (homotypic fusion and vacuole protein sorting), the exocyst complex, and the Dsl1p complex (Whyte and Munro, 2002).

Several multi-subunit tethering complexes of A. thaliana have been described. The exocyst complex that mediates vesicle tethering during exocytosis plays a role in morphogenesis, such as pollen and root hair tip growth, hypocotyl elongation and tracheary element development (Cole et al., 2005; Synek et al., 2006; Li et al., 2013). The HOPS complex, which localizes on the vacuole and pre-vacuolar compartment, is required for vacuolar biogenesis, and plays a role in the proper function of post-Golgi trafficking (Rojo et al., 2001, 2003). The GARP complex is implicated in pollen tip growth (Lobstein et al., 2004; Guermonprez et al., 2008) and in responses to heat and osmotic stress (Lee et al., 2006). The TRAPPII subunits AtTRS130 and AtTRS120 are required for cell-plate biogenesis (Jaber et al., 2010; Thellmann et al., 2010). The COG complex is required for appropriate cell expansion and meristem organization (Ishikawa et al., 2008). Recently, a putative Dsl1 complex was reported to be involved in the abscisic acid-mediated response to abiotic stresses (Zhao et al., 2013).

Maturing seeds synthesize large quantities of storage protein precursors on the rough ER, and actively transport these to protein storage vacuoles (PSVs) for further processing into mature forms. Consequently, storage proteins serve as ideal markers to investigate the trafficking machinery for cargo proteins. Multiple mechanisms have been proposed for vacuolar sorting of seed storage proteins (Hara-Nishimura et al., 2004; Jolliffe et al., 2005; Vitale and Hinz, 2005). One mechanism is the receptor-mediated sorting pathway. In this pathway, storage proteins are delivered to the Golgi apparatus where receptors sort them into the PSVs. The Arabidopsis vacuolar sorting receptor 1 (VSR1) delivers storage proteins from the Golgi apparatus to the pre-vacuolar compartments (Shimada et al., 2003a; Fuji et al., 2007). Mutant vsr1 plants accumulate storage proteins precursors in the extracellular space. Another mechanism, termed aggregation sorting, enables the efficient transport of large quantities of proteins. Precursor-accumulating vesicles, which contain abundant amounts of storage proteins, and dense vesicles, which contain highly condensed storage globulins, are reported to be responsible for aggregation sorting. ER-derived precursor-accumulating vesicles are observed in maturing seeds of pumpkin (Cucurbita maxima) (Hara-Nishimura et al., 1998), soybean (Glycine max) (Mori et al., 2004), and rice (Oryza sativa) (Takahashi et al., 2005), whereas Golgi-derived dense vesicles are observed in maturing seeds of common bean (Phaseolus vulgaris) (Chrispeels, 1983) and pea (Pisum sativum) (Hinz et al., 1999; Hillmer et al., 2001).

To identify the factors responsible for the transport of seed storage proteins, we previously isolated Arabidopsis mutants with a defect in the transport of storage proteins and named them maigo (mag) mutants (Shimada et al., 2006). One of the mag mutants, mag2, abnormally accumulates precursors of two major storage proteins, 2S albumin and 12S globulin, in seeds, and develops numerous precursor-accumulating structures (termed MAG bodies) within the ER lumen (Li et al., 2006). These results suggest that storage proteins are prevented from exiting the ER. The morphological structure of MAG bodies resembles that of precursor-accumulating vesicles; for instance, both of them are derived from the rough ER and contain a highly electron-dense core. The core comprises the precursor forms of 2S albumin. The storage 12S globulins are segregated from 2S albumin and are localized in the matrix region of the MAG bodies together with the ER chaperones BiP and PDI. These observations suggest that MAG2 is required for efficient exit of storage protein precursors from the ER (Li et al., 2006). MAG2 exhibits low sequence identity to yeast Tip20p and mammalian RINT1 (Li et al., 2006). Tip20p forms a complex with Dsl1p and Sec39p/Dsl3p, and is responsible for retrograde transport from the Golgi apparatus to the ER (Koumandou et al., 2007). RINT1 was originally reported to function in a cell-cycle checkpoint during mitosis in animal cells (Xiao et al., 2001). It forms a complex with ZW10 and NAG, and regulates membrane trafficking from the Golgi apparatus to the ER during interphase (Hirose et al., 2004; Arasaki et al., 2006, 2007; Varma et al., 2006; Aoki et al., 2009). Mammalian ZW10 and yeast Dsl1p are considered to be orthologs (Hirose et al., 2004; Arasaki et al., 2006), despite their low homology.

To characterize the molecular machinery of the MAG2-dependent transport pathway of storage proteins, pull-down assays using transgenic plants expressing tandem-affinity purification (TAP)-tagged MAG2 were performed. This work resulted in identification of three proteins, which were designated as MAG2-INTERACTING PROTEINs (MIPs). Mutant mip plants exhibited a defect in vacuolar transport of seed storage proteins and developed MAG bodies in seed cells, similar to those observed in the mag2 mutant. Our results suggest that MAG2 and MIP proteins form a complex on the ER, and that this complex is responsible for efficient transport of seed storage proteins.

Results

Identification of a MAG2 complex that comprises MAG2 and three MIPs

To identify MAG2-interacting proteins, transgenic Arabidopsis plants expressing TAP-tagged MAG2 were generated and subjected to pull-down analysis using an anti-myc antibody that recognized the TAP tag used. The pull-down sample produced four protein bands of comparable intensity on a silver-stained SDS–PAGE gel (Figure 1a, TAP-MAG2): >210 kDa (m1), approximately 130 kDa (m2), approximately 90 kDa (m3), and approximately 80 kDa (m4). These bands were not detected in the sample from wild-type plants having no tagged MAG2 (Figure 1a, WT). This result suggests that these four proteins are components of the MAG2 complex.

Figure 1.

Identification of Arabidopsis MAG2-interacting proteins. (a) Extracts of Arabidopsis seedlings expressing TAP-tagged MAG2 (TAP-MAG2) or TAP-tagged MIP1 (TAP-MIP1) were subjected to a pull-down assay with anti-c-myc antibody and subsequent SDS–PAGE followed by silver staining. The positive bands (m1, m2, m3, m4, p1, p2 and p3) were analyzed with by MALDI-TOF/MS. Col-0 (WT) was included as a control. (b) The identified proteins that correspond with each band and their AGI (Arabidopsis Genome Initiative) codes. (c) Domain structures of the identified MIP proteins. MIP1, MIP2 and MIP3 contain ZW10, Sec39 and Sec1 domains, respectively. The solid line below MIP1 shows the polypeptide region used for antibody production. (d) The pull-down samples from seedlings expressing TAP-MAG2 or TAP-MIP1 were subjected to immunoblot analysis using antibodies against MIP1 or MAG2.

Using mass spectrometric analysis, the approximately 130 kDa protein (m2) was found to be TAP-tagged MAG2 (Figure 1b). This protein has a molecular mass consistent with the sum of the predicted masses of MAG2 (90.2 kDa) and TAP (40 kDa). Mass spectrometry also identified the >210 kDa (m1), approximately 90 kDa (m3) and approximately 80 kDa (m4) proteins as the gene products of At5g24350, At2g42700 and At2g32900, respectively (Figure 1b) with high sequence coverage (Figures S1–S3). The approximately 80 kDa, >210 kDa and approximately 90 kDa proteins were designated as MAG2-INTERACTING PROTEIN 1–3 (MIP1–3), respectively.

MIP1 (At2g32900.1) is predicted to be an 83.9 kDa protein containing a conserved Zeste–White 10 (ZW10) domain (IPR009361), as shown in Figure 1(c) (MIP1), although it exhibits low sequence identity (21%) with human ZW10 (Figure S4). ZW10 was reported to be a subunit of the mammalian tethering complex RINT1–ZW10–NAG (Hirose et al., 2004; Aoki et al., 2009). MIP2 (At5g24350.1) is predicted to be a 266.8 kDa protein that contains a Sec39 domain (IPR013244) (Figure 1c, MIP2). Thus, MIP2 resembles yeast Sec39p/Dsl3p, although the primary structures and molecular weights of MIP2 and Sec39p are markedly different. Sec39p was reported to be a subunit of the yeast tethering complex Tip20p–Dsl1p–Sec39p (Kraynack et al., 2005). These results suggest that MAG2, MIP1 and MIP2 function as subunits of a tethering complex in plants.

MIP3 (At2g42700.1) is predicted to be a 91.7 kDa protein that contains a conserved Sec1 domain (IPR001619) at its C-terminus (Figure 1c, MIP3). MIP3 exhibits low sequence identity (21%) to the human SEC1 FAMILY DOMAIN-CONTAINING 2 (SCFD2) (Figure S5). SCFD2 was identified as a p53 target gene (Krieg et al., 2006), although its function is currently unknown. The current study provides evidence for the function of MIP3/SCFD2 family proteins (discussed below).

To confirm the association of these MIP proteins with MAG2, transgenic Arabidopsis plants expressing TAP-tagged MIP1 were generated and subjected to pull-down analysis. The pull-down sample produced three bands of comparable intensity on a silver-stained SDS–PAGE gel (Figure 1a, TAP-MIP1), indicating that the >210 kDa (p1), approximately 125 kDa (p2) and approximately 90 kDa (p3) proteins specifically co-purified with TAP-tagged MIP1. Mass spectrometry identified the approximately 125 kDa protein (p2) as TAP-tagged MIP1, with a molecular mass consistent with the sum of the predicted masses of MIP1 (83.9 kDa) and TAP (40 kDa). The >210 kDa band (p1) corresponded to MIP2 (266.8 kDa). The approximately 90 kDa band (p3) was found to contain both 91.7 kDa MIP3 and 90.2 kDa MAG2 (Figure 1b) with high sequence coverage (Figure S6).

Immunoblot analysis further confirmed the association between MAG2 and MIP1 (Figure 1d). The pull-down sample from transgenic plants expressing either TAP-MAG2 or TAP-MIP1 produced positive signals on an immunoblot with anti-MIP1 antibody and anti-MAG2 antibody. By contrast, the pull-down sample from wild-type plants did not produce a signal. Taken together, these results indicate that MAG2 forms a stable complex with three proteins: MIP1, MIP2 and MIP3.

MIP1, MIP2 and MIP3 are required for proper maturation of seed storage proteins

To identify the function of MIP proteins, T-DNA-tagged lines of each MIP gene were isolated. The MIP1 gene contains 14 exons and 13 introns (Figure 2a, MIP1). The mip1-1 mutant (SALK_017317) contains a T-DNA insertion in the last exon (Figure 2a, MIP1), and normal MIP1 mRNA was not detected in the mip1-1 mutant (Figure 2b). The MIP1 deficiency resulted in abnormal accumulation of the precursors of both 12S globulin and 2S albumin in dry seeds (Figure 3a, mip1-1). This suggested that the mip1-1 mutant has a defect in either the transport pathway from the ER to the PSVs or conversion of precursors into mature forms in PSVs. Previously, we reported that a defect in the transport pathway from the ER to the PSVs results in a 17 kDa precursor of 2S albumin (Shimada et al., 2003a), whereas the latter defect results in 15 and 16 kDa precursors (Shimada et al., 2003b). Figure 3(a) shows that the mip1-1 mutant accumulates the 17 kDa precursor, but neither the 15 nor 16 kDa precursors. This result suggests that the mip1-1 mutant has a defect in the transport machinery (Figure 3a). Molecular complementation analysis supported this suggestion. Introduction of a genomic fragment of the MIP1 gene (MIP1g) into the mip1-1 mutant rescued the defect of mip1 seeds (Figure 3c).

Figure 2.

MIP genes and T-DNA-tagged mutants. (a) Structures of the MIP genes and the T-DNA insertion sites of each mutant. The locations of primers used for RT-PCR analyses are indicated by arrows. Black boxes indicate exons; solid lines indicate introns; gray boxes indicate untranslated regions. (b,c) Expression levels of MIP genes in the T-DNA-tagged mutants. RT-PCR was performed using primer sets as indicated in (a).

Figure 3.

Defects of mip1-1, mip2-1 and mip3-1 mutants in storage protein trafficking. (a) Immunoblot (right) and Coomassie brilliant blue staining of SDS–PAGE (left) of dry seeds of wild-type (WT), mip1-1, mip2-1 and mip3-1. Mutant mip2-1 seeds were a mixture of wild-type, hemizygous and homozygous seeds. p12S, pro12S globulin; p2S, pro2S albumin; 12S-a, 12S globulin subunits; 2S, 2S albumin subunits. (b) Coomassie brilliant blue staining of an SDS–PAGE gel (left) for individual grains of mip2-1 hemizygous and homozygous seeds. Only mip2-1 homozygous seeds abnormally accumulated precursors of storage proteins (red arrowheads), indicating a defect of protein trafficking. (c,d) Complementation of mip1-1 and mip3-1 mutants. Immunoblot of seeds of wild-type (WT), mip1-1, mip3-1, three independent transgenic lines of mip1-1 expressing a genomic fragment of MIP1 (MIP1 g), and three independent transgenic lines of mip3-1 expressing a genomic fragment of MIP3 (MIP3 g).

The MIP2 gene contains 12 exons and 11 introns, and T-DNA is inserted within the 5′ UTR region of the gene in the mip2-1 mutant (SALK_147329) (Figure 2a, MIP2). Genotyping revealed that homozygous mip2-1 seedlings were not obtained, although homozygous dry seeds were present. The mip2-1 mutation may cause a defect of seed germination. The seed harvested from a mip2-1 hemizygote displayed abnormal accumulation of storage protein precursors including 17 kDa pro2S albumin [Figure 3a, mip2-1 (mix)]. Previously, we established a single-grain method to determine both the genotype and phenotype of one seed grain (Takahashi et al., 2010). Individual grains of seeds from a mip2-1 hemizygote were subjected to this method. Figure 3(b) shows that each grain of homozygous mip2-1 seeds accumulated storage protein precursors (mip2-1/mip2-1), whereas each grain of hemizygous mip2-1 seed did not (mip2-1/+). These results suggest that the mip2-1 mutant also has a defect in the transport machinery (as for mip1-1).

The MIP3 gene contains seven exons and six introns (Figure 2a, MIP3). Two T-DNA insertion mutants of MIP3, mip3-1 (SALK_141975) and mip3-2 (GABI_110G11), were isolated (Figure 2a, MIP3). MIP3 mRNA was not detected in mip3-1 or mip3-2 (Figure 2c), suggesting that these are null mutants. Both mip3-1 (Figure 3a) and mip3-2 (Figure S7) seeds abnormally accumulated storage protein precursors, including 17 kDa pro2S albumin, indicating that MIP3 is involved in the maturation of seed storage proteins. This conclusion was supported by the result that introduction of a genomic fragment of the MIP3 gene (MIP3g) into the mip3-1 mutant rescued the defect of mip3 seeds (Figure 3d).

Mutant mip1, mip2 and mip3 plants develop MAG bodies in the seed cells, similar to the mag2 mutant

Electron microscopy revealed that the mip mutants mip1-1, mip2-1 and mip3-1 developed abnormal structures in the seed cells (Figure 4a, red arrowheads). Enlarged images show that these structures contain electron-dense cores (Figure 4b). Immunogold analysis showed that storage proteins were localized to PSVs and to the abnormal structures. Interestingly, the localization of 12S globulin and 2S albumin differed: 12S globulin was localized to the matrix region of the abnormal structures and 2S albumin was localized to the electron-dense core (Figure 4b). In addition, the structures were frequently surrounded by ribosomes (Figure 4c), suggesting that the precursors of storage proteins accumulated in the ER lumen. The PSV localization of storage proteins in the mip mutants is consistent with the accumulation of mature storage proteins in these mutant seeds (Figure 3a). By contrast, wild-type seeds showed limited localization of storage proteins to PSVs (Figure 4, WT).

Figure 4.

MAG bodies were observed in mip mutant seed cells. (a) Electron micrographs of seed cells of wild-type (WT), mip1-1, mip2-1 and mip3-1. MAG bodies (red arrowheads) were observed in the mutants. Scale bars = 5 μm. (b) Immunoelectron micrographs of WT, mip1-1, mip2-1 and mip3-1 with anti-12S and anti-2S antibodies. Scale bars = 0.5 μm. (c) Magnified images of MAG bodies in mip1-1, mip2-1 and mip3-1. MAG bodies are surrounded by ribosomes (red arrows). Scale bars = 0.5 μm. PSV, protein storage vacuole; CW, cell wall.

The structures identified in all mip mutants were similar to the MAG bodies that were previously identified in the seeds of the mag2 (Li et al., 2006) and mag4 (Takahashi et al., 2010) mutants, which are defective in protein transport between the ER and the Golgi apparatus. The structures comprise storage proteins and formed in the ER lumen. These results suggest that the mip1-1, mip2-1 and mip3-1 mutants have defects in the export of newly synthesized storage protein precursors from the ER.

MIP1 is associated with the ER membrane

Previously, we reported that MAG2 functions in membrane trafficking on the ER by associating with the ER-localized t-SNARE proteins SYP81/AtUfe1 and SEC20 (Li et al., 2006). To determine the localization of MIP1, subcellular fractionation of wild-type seedlings was performed. MIP1 was detected in a microsomal P100 fraction (Figure 5a). The P100 fraction was subsequently subjected to sucrose density gradient centrifugation in the presence of MgCl2 or EDTA. MIP1 showed a magnesium-dependent density shift on the gradients in parallel with the ER marker BiP (Figure 5b), suggesting that MIP1 localized to the ER membrane. The ER localization of MIP1 was confirmed by transient expression analysis with tobacco BY-2 suspension-cultured cells. The network structure labeled by GFP–MIP1 was also labeled by mRFP–SEC20, an ER membrane marker, although a proportion of GFP–MIP1 remained in the cytosol (Figure 5c).

Figure 5.

MIP1 is associated with the ER membrane. (a) The 100 000 g pellet (P100) and supernatant (S100) from wild-type seedlings were subjected to immunoblot analysis with anti-MIP1 antibodies. (b) Magnesium-dependent density shift of the MIP1-containing compartment in the sucrose density gradient, showing the ER localization of MIP1. A microsomal fraction (P100) was prepared from wild-type seedlings in the presence of MgCl2 or EDTA and then subjected to sucrose density gradient centrifugation (15–45% w/v)in the presence of MgCl2 or EDTA. Each fraction was subjected to immunoblot analysis with anti-MIP1 and anti-BiP antibodies. BiP is an ER marker. (c) Co-localization of GFP–MIP1 with mRFP–SEC20, an ER marker, in protoplasts of tobacco BY-2 suspension-cultured cells. The cells were inspected using a confocal laser scanning microscope. Scale bar = 20 μm.

Protein transport and dynamics of the ER and Golgi apparatus in vegetative cells

In animal and yeast cells, defects of ER-localized tethering complexes have been shown to result in morphological changes of the ER or Golgi, and in the defect in transport between them (Sweet and Pelham, 1993; Arasaki et al., 2006). To examine a defect in the vegetative tissues, we performed phenotypic analysis of mip and mag2 mutant seedlings. GFP fused N-terminally with a signal peptide and C-terminally with an ER-retention signal sequence (SP-GFP-HDEL) (Figure S8) and sialyltransferase–GFP (Figure S9) were stably introduced into the mutant plants. These GFP fusion proteins are markers for the ER and Golgi, respectively, and are often used to monitor organelle morphology, distribution and/or ER-Golgi transport (Nakano et al., 2012). However, there were no significant defects in GFP transport and organelle dynamics in the root cells of mip1, mip3 and mag2 seedlings (Figures S8 and S9).

Discussion

Arabidopsis MAG2 complex is a possible tethering complex that is evolutionarily conserved among plants, mammals and yeasts

The three MAG2-interacting proteins identified in this study are MIP1 with a ZW10 domain, MIP2 with a Sec39 domain, and MIP3 with a Sec1 domain. The mag2, mip1, mip2 and mip3 mutant phenotypes included abnormal transport of storage proteins and abnormal development of MAG bodies in the ER of seed cells (Li et al., 2006) (Figures 3 and 4), suggesting that MAG2 and MIPs are involved in protein trafficking between the ER and Golgi apparatus.

ZW10 was originally identified as a centromere/kinetochore-associated protein essential for the spindle checkpoint in Drosophila (Smith et al., 1985; Williams et al., 1992). In addition to its role in mitosis, ZW10 has been shown to play a role in membrane trafficking between the ER and Golgi during interphase (Hirose et al., 2004). Sec39p (710 amino acids) was originally reported to be involved in protein secretion by genome-wide analysis (Mnaimneh et al., 2004), and was subsequently reported to function in retrograde transport from the Golgi to the ER in yeast (Kraynack et al., 2005). The mammalian equivalent of yeast Sec39p, a neuroblastoma amplified gene (NAG) product, was reported to interact with the ER t-SNARE complex that regulates Golgi-to-ER transport in humans (Aoki et al., 2009). NAG is a large protein (2374 amino acids) that contains a Sec39 domain in its middle region. This feature is similar to MIP2 (2376 amino acids) (Figure 1c), although MIP2 exhibits low sequence identity (23%) to NAG (Figure S10). Together with the fact that MAG2 is an ortholog of mammalian RINT1 and yeast Tip20p, these results suggest that a MAG2 complex consisting of MAG2, MIP1 and MIP2 functions as a multi-subunit tethering complex in plants.

A newly defined family (MIP3/SCFD2) in the Sec1/Munc18 superfamily

The molecular composition of the tethering complex is partly conserved between the yeast Tip20p–Dsl1p–Sec39p complex, the mammalian RINT1–ZW10–NAG complex and the Arabidopsis MAG2–MIP1–MIP2 complex as described above. However, the MAG2 complex differs from the tethering complexes that have been reported to date as it contains a fourth subunit harboring a Sec1 domain, MIP3. Consistent with the absence of a Sec1 domain-containing protein in the yeast Dsl1p tethering complex (Tip20p–Dsl1p–Sec39p), both a BLAST sequence search (Saccharomyces Genome Database, http://www.yeastgenome.org) and a conserved domain search (InterPro, http://www.ebi.ac.uk/interpro/) failed to identify a yeast homolog of MIP3. MIP3 shows low sequence similarity to human SCFD2, whose function is yet unknown (Krieg et al., 2006). Recently, systematic analysis of human protein complexes has suggested the possibility of an interaction between ZW10 and SCFD2 (Hutchins et al., 2010). These results suggest that SCFD2 may potentially be a fourth subunit of the mammalian tethering complex, which is known to consist of RINT1, ZW10 and NAG.

Sec1 domain (IPR001619)-containing proteins form a Sec1/Munc18 (SM) protein superfamily. SM proteins interact with a syntaxin protein and play a prominent role in vesicle trafficking (Carr and Rizo, 2010). Seven Sec1 domain-containing proteins are present in Arabidopsis, eight are present in human, and four are present in yeast. A BLAST search of the National Center for Biotechnology Information database (http://blast.ncbi.nlm.nih.gov/Blast.cgi) and clustering using the MAFFT program at GenomeNet (http://www.genome.jp/ja/) revealed that Sec1 domain-containing proteins from Arabidopsis, human and yeast may be classified into five families (Figure 6). Four of them (the Vps33 family, the Vps45 family, the Sec1 family and the Sly1 family) are well conserved in eukaryotes (Koumandou et al., 2007). Arabidopsis VPS33 is a subunit of the HOPS complex that is involved in membrane fusion on both pre-vacuolar compartments and the tonoplast (Rojo et al., 2003). AtVPS45 is localized on the trans-Golgi network and may be involved in membrane fusion of retrograde vesicles recycling vacuolar trafficking machinery (Bassham et al., 2000; Zouhar et al., 2009). KEULE is a Sec1 protein and binds the cytokinesis-specific syntaxin KNOLLE (Assaad et al., 2001; Park et al., 2012).

Figure 6.

Identification of MIP3/SCFD2 family in the Sec1/Munc18 (SM) protein superfamily. Sec1 domain (IPR001619)-containing proteins form a Sec1/Munc18 (SM) protein superfamily that includes five families: the Vps33 family, the Vps45 family, the Sec1 family, the Sly1 family and the MIP3/SCFD2 family. The MIP3/SCFD2 family was identified in this study, and includes Arabidopsis MIP3, human SCFD2 and several orthologs from plants and animals; it has no yeast members. Protein sequences were aligned using the MAFFT program at GenomeNet (http://www.genome.jp/ja/), and an unrooted phylogenetic tree using the neighbor-joining algorithm was drawn. Accession numbers are as follows: Arabidopsis thaliana MIP3, At2g42700.2; Vitis vinifera MIP3, XP_002281104; Glycine max MIP3, XP_003527341; Oryza sativa MIP3, AAX93012; Homo sapiens SCFD2, Q8WU76; Strongylocentrotus purpuratus, XP_783509; Daphnia pulex, EFX74459; Apis mellifera, XP_625059; Dictyostelium discoideum, XP_642748; A. thaliana Vps33, At3g54860.2; H. sapiens Vps33A, Q96AX1; H. sapiens Vps33B, Q9H267; Saccharomyces cerevisiae Vps33, YLR396C; A. thaliana Sec1A, At1g02010.1; A. thaliana KEULE, At1g12360.1; A. thaliana Sec1B, At4g12120.1; H. sapiens STXBP1, P61764; H. sapiens STXBP2, Q15833; H. sapiens STXBP3, O00186; S. cerevisiae Sec1, YDR164C; A. thaliana Vps45, At1g77140.1; H. sapiens Vps45, Q9NRW7; S. cerevisiae Vps45, YGL095C; A. thaliana Sly1, At2g17980.1; H. sapiens SCFD1, Q8WVM8; S. cerevisiae Sly1, YDR189W.

In this study, the fifth family is defined by Arabidopsis MIP3 and human SCDF2, and includes orthologs from plants (rice, grape and soybean) and animals (slime mold, honey bee, Daphnia and purple sea urchin) (Figure 6). We designated the family the MIP3/SCDF2 family. To date, HOPS is the only tethering complex that contains an SM protein, Vps33p, which binds to the vacuolar t-SNAREs (Seals et al., 2000). The HOPS associates with SNARE complexes, proofreads the correct trans-SNARE complex conformation and regulates its capacity for membrane fusion (Starai et al., 2008). It is possible that members of the MIP3/SCFD2 family play a role in tethering and/or fusion of Golgi-derived vesicles by interacting with an ER-localized syntaxin in plant and animal cells.

The MAG2 complex is responsible for efficient exit of storage protein precursors from the ER

In this study, we showed that the MAG2 complex is required for efficient exit of storage protein precursors from the ER. MAG2, together with three MIP proteins, forms a stable complex on the ER membrane (Figures 1 and 5). In addition, MAG2 was previously shown to interact with the ER-localized t-SNAREs SYP81/AtUfe1 and SEC20 (Li et al., 2006). Because t-SNAREs are considered to function in the step in which the vesicle membrane fuses to the target membrane, the MAG2 complex may function in retrograde transport from the Golgi apparatus to the ER. This is consistent with the fact that the yeast Dsl1p complex and the mammalian counterpart function as tethering complexes in retrograde transport from the Golgi apparatus to the ER.

The phenotypes of the mag2, mip1, mip2 and mip3 mutants overlap with those described previously for mag4 seed cells (Takahashi et al., 2010). However, MAG4 and the MAG2 complex function in different steps during protein transport between the ER and Golgi apparatus. MAG4 localizes to the cis-Golgi stacks and functions as a tethering factor to facilitate anterograde transport from the ER to the Golgi apparatus (Takahashi et al., 2010). By contrast, the MAG2 complex localizes to the ER and may function as a tethering complex to facilitate retrograde transport from the Golgi apparatus to the ER. Taken together, the results of this study suggest that an equilibrium may exist between anterograde and retrograde transport that is important for efficient exit of proteins from the ER. The reduction in retrograde transport may reduce anterograde transport in the mag2 and mip mutants, in which active synthesis of storage proteins continued. Thus, the abnormal aggregate structures that were observed in the seed cells of these mutants may result from increased levels of storage proteins in the ER lumen of these mutants. Similar defects in anterograde transport were previously reported in yeast tip20 cells (Sweet and Pelham, 1993) and RINT1-silenced HeLa cells (Arasaki et al., 2006).

In contrast to the defect in maturing seed cells as described above, we did not detect visible abnormalities within vegetative cells of mag2 and mip mutants (Figures S8 and S9). This suggests that subunits of the MAG2 complex have a minor impact in vegetative cells under normal growth conditions. However, a recent study reported that the MAG2 complex is involved in abscisic acid-mediated response to abiotic stresses (Zhao et al., 2013). It is possible that mag2 and mip mutant cells exhibit abnormal phenotypes under such stressed conditions.

Experimental procedures

Plant materials and growth conditions

The wild-type plants used in this study were Arabidopsis thaliana ecotype Col-0. The mag2 mutants have been described previously (Li et al., 2006). All of mip1-1 (SALK_017317), mip2-1 (SALK_147329), mip3-1 (SALK_141975) and mip3-2 (GABI_110G11) were provided by the Arabidopsis Biological Resource Center (ABRC) at Ohio State University. Information about these mutants was obtained from the Salk Institute Genomic Analysis Laboratory website (http://signal.salk.edu). The T-DNA insertion sites of the mutants were sequenced and determined to be located 3730 bp from the transcription initiation site of the MIP1 gene in mip1-1, 299 bp from the transcription initiation site of the MIP2 gene in mip2-1, 560 bp from the transcription initiation site of the MIP3 gene in mip3-1, and 2783 bp from the transcription initiation site of the MIP3 gene in mip3-2.

Transgenic Arabidopsis plants (Col-0) expressing TAP-tagged MAG2 or TAP-tagged MIP1 were generated using a modified pN-TAPa vector (Rubio et al., 2005). Arabidopsis seeds were surface-sterilized and sown either on soil or 0.5% gellan gum (Wako, http://www.wako-chem.co.jp/english/) with Murashige and Skoog (MS) salts (Wako) supplemented with 1% w/v sucrose. Plants were grown at 22°C under continuous light. Protoplasts were prepared from tobacco BY-2 suspension-cultured cells after sub-culture, and were incubated in medium containing MS salts, B5 vitamins, 1% w/v sucrose and 0.4 m mannitol as described previously (Tamura et al., 2003).

Complementation analysis

We amplified the genomic fragments of MIP1 and MIP3 using the primer pairs InfusionMIP1FNotI + InfusionMIP1RNotI and MIP3-F+MIP3-R, respectively (Table S1). We cloned the genomic fragment of MIP1 into the pENTRTM 1A plasmid (Invitrogen, http://www.lifetechnologies.com/us/en/home/brands/invitrogen.html) using an In-Fusion™ Advantage PCR cloning kit (Takara Bio Inc., http://www.takara-bio.com/), and cloned the genomic fragment of MIP3 into the pENTR/D-TOPO plasmid (Invitrogen) by the TOPO reaction. We introduced the plasmids into a binary vector, FAST-G01 (Shimada et al., 2010), by LR reaction to generate proMIP1:MIP1 and proMIP3:MIP3. Agrobacterium tumefaciens (strain GV3101) was transformed with these constructs. mip1-1 and mip3-1 plants were infected with the bacteria by the floral-dip method (Clough and Bent, 1998). T1 seeds were selected in medium containing 25–50 mg/l hygromycin B.

Pull-down analysis

Pull-down assays were essentially performed as described previously (Tamura et al., 2010) using a μMACS epitope tag protein isolation kit (anti-c-myc, Miltenyi Biotec, https://www.miltenyibiotec.com/en/). Two-week-old seedlings of Arabidopsis plants (approximately 0.5 g) were homogenized on ice in 2 ml lysis buffer [50 mm Tris/HCI, pH 8.0, 150 mm NaCl, 1 mm CaCl2, 1 mm MgCl2, 1% CHAPS and EDTA-free protease inhibitor cocktail (Roche, http://www.roche.com/index.htm)]. Lysates were transferred to 2 ml tubes and mixed well. After incubation on ice for 30 min, the homogenates were filtered through a Falcon cell strainer (70 μm, BD Biosciences, http://www.bdbiosciences.com/) by centrifugation at 200 g. A 2 ml aliquot of the filtrate was centrifuged at 10 000 g and 4°C for 10 min. The supernatants were incubated with 50 μl anti-c-myc MicroBeads (Miltenyi Biotec, https://www.miltenyibiotec.com/en/) for 30 min on ice. Cell lysates were applied to μ Columns that were set on the μMACS separator. The μ Columns were rinsed five times with 200 μl wash buffer 2 (50 mm Tris/HCI, pH 8.0, 150 mm NaCl, 1 mm CaCl2, 1 mm MgCl2). Twenty microliters of pre-heated (95°C) elution buffer were applied to each μ Column, and incubated for 5 min at room temperature. Finally, 50 μl of pre-heated (95°C) elution buffer was applied to each μ Column to elute the samples. The eluates were collected for further analysis.

Mass spectrometry

Mass spectrometric identification was performed as described previously (Sugano et al., 2010). Proteins were separated by SDS–PAGE followed by silver staining using a silver staining kit (GE Healthcare, http://www3.gehealthcare.com/en/Global_Gateway). The bands were individually excised and subjected to in-gel digestion with trypsin (Promega, http://worldwide.promega.com/country.aspx?returnUrl=/) in a buffer containing 50 mm ammonium bicarbonate, pH 8.0, 2% acetonitrile at 37°C overnight. The digests were subjected to MALDI–TOF/MS using an ultra-flex TOF/TOF (Bruker Daltonics, http://www.bruker.com/).

Antibodies and immunoblot analysis

The anti-MIP1 antibody was raised against the MIP1 polypeptide from amino acid 341 to the C-terminus. A cDNA encoding the above region was amplified by PCR and inserted into pET32a (EMD Millipore, http://www.emdmillipore.com/). Escherichia coli was transformed with the plasmid, and the His-tagged protein was purified using a Hi-Trap chelating column (GE Healthcare). The antibodies were prepared previously: anti-12S (Shimada et al., 2003b), anti-2S (Shimada et al., 2003b), anti-2S3P (Li et al., 2006) and anti-MAG2 (Li et al., 2006).

SDS–PAGE and immunoblot analysis were performed as described previously (Shimada et al., 2003a). Antibody dilutions were as follows: anti-MAG2, anti-MIP1 and anti-BiP, 1:1000; anti-12S, 1:20 000, anti-2S, 1:10 000; anti-2S3P, 1:5000. The dilution of horseradish peroxidase-conjugated goat antibodies raised against rabbit IgG (Pierce, http://www.piercenet.com/) was 1:5000. Signals were detected using an enhanced chemiluminescence (ECL) detection system (GE Healthcare).

Electron microscopy

Immunoelectron microscopy with dry seeds was performed as described previously (Shimada et al., 2003a). Samples were treated with antibodies against 12S globulin (dilution 1:50) and 2S albumin (dilution1:50). Sections were examined using a 1200EX transmission electron microscope (JEOL, http://www.jeol.co.jp/en/) at 80 kV.

Subcellular fractionation

The fractionation was performed essentially as described previously (Ueda et al., 2010). Ten-day-old seedlings of wild-type plants (Col-0; 3 g fresh weight) were used. The homogenate was centrifuged at 100 000 g at 4°C for 1 h after sequential centrifugations at 3000 g for 20 min and 10 000 g for 20 min. The pellet (P100) and supernatant (S100) were subjected to immunoblot analysis with anti-MIP1 antibodies. For subcellular fractionation, microsomal fractions were prepared from 11-day-old seedlings of wild-type plants (Col-0; 0.6 g fresh weight) as described previously (Tamura et al., 2005). The microsomal fractions were layered on top of a 16 ml linear sucrose density gradient (15–45% w/v), and centrifuged at 100 000 g for 13 h at 4°C. After centrifugation, 1 ml fractions were collected. Each fraction was concentrated with acetone and subjected to immunoblotting.

RT-PCR analysis

Total RNA was isolated from leaves as described previously (Shimada et al., 2006). The primer sets used are listed in Table S1. The PCR products were separated by 1.0% agarose gel electrophoresis.

Transient and stable expression of GFP fusions in plant cells

For transient expression analysis, an MIP1 cDNA fragment was amplified using the MIP1-specific primers MIP1-TOPO-F and MIP1-R+stop (Table S1). The amplified fragment was inserted into entry vector pENTR/D-TOPO by TOPO cloning according to the manufacturer's instructions (Invitrogen). The MIP1 cDNA was transferred from the entry clone to the destination vector pUGW6 (Nakagawa et al., 2007). An mRFP-SEC20 fragment was amplified by overlap PCR using mRFP and SEC20 cDNAs with specific primers, and was then inserted into the pBI221 vector (Tamura et al., 2003). The vectors for GFP–MIP1 and mRFP–SEC20 were transiently introduced into protoplasts from tobacco BY-2 suspension-cultured cells (Tamura et al., 2003).

For stable expression analysis, transgenic Arabidopsis plants expressing either sialyltransferase–GFP or SP–GFP–HDEL (Nakano et al., 2012) were crossed with mag2-1, mag2-3, mip1-1, mip3-1 and mip3-2 mutant plants. The cells were inspected by confocal laser scanning microscopy and differential interference contrast microscopy (model LSM510 META; Zeiss, http://corporate.zeiss.com/gateway/en_de/home.html).

Acknowledgements

The authors are grateful to K. Okawa (Graduate School of Medicine, Kyoto University, Japan) for excellent mass spectrometric analysis, and T. Kato (Graduate School of Biological Sciences, Nara Institute of Science and Technology, Japan) for kind donation of the modified pN-TAPa vector. The authors also thank the Arabidopsis Biological Resource Center for providing the seeds of Arabidopsis T-DNA insertion mutants. This work was supported by Specially Promoted Research of Grant-in-Aid for Scientific Research to I.H-.N. (grant number 22000014) and a research fellowship to J.T. (grant number 241880) from the Japan Society for the Promotion of Science, and by Grants for Excellent Graduate Schools from the Ministry of Education, Culture, Sports, Science and Technology of Japan. This work was also supported by grants from the National Natural Science Foundation of China (grant number 30970223), the Fundamental Research Funds for the Central Universities (grant number DL09DA02), the China Postdoctoral Foundation (grant number 200902365) and the Specialized Research Fund for the Doctoral Program of Higher Education (grant number 20090062120008).

Ancillary

Advertisement