The Golgi-Localized Arabidopsis Endomembrane Protein12Contains Both Endoplasmic Reticulum Export and GolgiRetention Signals at Its C Terminus C W

Caiji Gao, Christine K.Y. Yu, Song Qu, Melody Wan Yan San, Kwun Yee Li, Sze Wan Lo, and Liwen Jiang1

School of Life Sciences, Centre for Cell and Developmental Biology, The Chinese University of Hong Kong, Shatin, New Territories,Hong Kong, China

Endomembrane proteins (EMPs), belonging to the evolutionarily conserved transmembrane nine superfamily in yeast andmammalian cells, are characterized by the presence of a large lumenal N terminus, nine transmembrane domains, and a shortcytoplasmic tail. The Arabidopsis thaliana genome contains 12 EMPmembers (EMP1 to EMP12), but little is known about theirprotein subcellular localization and function. Here, we studied the subcellular localization and targeting mechanism of EMP12in Arabidopsis and demonstrated that (1) both endogenous EMP12 (detected by EMP12 antibodies) and green fluorescentprotein (GFP)-EMP12 fusion localized to the Golgi apparatus in transgenic Arabidopsis plants; (2) GFP fusion at the C terminusof EMP12 caused mislocalization of EMP12-GFP to reach post-Golgi compartments and vacuoles for degradation inArabidopsis cells; (3) the EMP12 cytoplasmic tail contained dual sorting signals (i.e., an endoplasmic reticulum export motifand a Golgi retention signal that interacted with COPII and COPI subunits, respectively); and (4) the Golgi retention motif ofEMP12 retained several post-Golgi membrane proteins within the Golgi apparatus in gain-of-function analysis. These sortingsignals are highly conserved in all plant EMP isoforms and, thus, likely represent a general mechanism for EMP targeting inplant cells.

INTRODUCTION

In the endomembrane system of eukaryotic cells, secretoryproteins start their journey from the endoplasmic reticulum (ER)prior to reaching the Golgi apparatus for further sorting to post-Golgi compartments, such as the trans-Golgi network (TGN) andprevacuolar compartment (PVC) (Barlowe et al., 1994). Proteintraffic between the ER and Golgi apparatus is mediated by twodistinct coated vesicles (i.e., coat protein complex II (COPII) andCOPI), which mediate anterograde and retrograde transport,respectively (Cosson and Letourneur, 1997; Kuehn et al., 1998;Rabouille and Klumperman, 2005; Donohoe et al., 2007). Integralmembrane proteins that traffic between the ER and Golgi usuallycontain specific sorting signals that are essential for selectivepackage into COPII vesicles for ER export or into COPI vesiclesfor ER retrieval and Golgi retention (Barlowe, 2003; Beck et al.,2009).

In mammals and yeast, various ER export signals have beenidentified, including the diacidic motif (DXE) of vesicular stomatitisvirus glycoprotein, the dihydrophobic (LL) motif of ERGIC-53, andthe diaromatic motif (FF, YY) of 46-kD Endomembrane ProteinPrecursor (Emp46p) (Nishimura and Balch, 1997; Nufer et al.,

2002; Sato and Nakano, 2002). All of these typical ER exportsignals are found to reside on the cytosolic regions of trans-membrane proteins that interact with COPII vesicles (Barlowe,2003). Similarly, the dilysine motif, KKXX, is one of the best knownsorting signals required for retrograde Golgi-to-ER transport ofseveral type I membrane proteins that interact with COPI vesicles(Cosson and Letourneur, 1994; Schröder et al., 1995; Harter andWieland, 1998; Gomez et al., 2000). Recently, the semiconservedPhe-Leu-Ser-like motifs were identified as Golgi retention signalsin the cytoplasmic tails (CTs) of glycosyltransferases, a group ofGolgi-resident integral membrane proteins in yeast (Tu et al.,2008). These motifs were shown to interact with COPI vesicles viaVps74p to maintain the steady state Golgi localization of theglycosyltransferases, thus providing the best known molecularmechanism of Golgi retention of membrane proteins in eukaryoticcells (Tu et al., 2008).Several studies have been performed to analyze the targeting

mechanisms of integral membrane proteins in plant cells. Thetransmembrane domain (TMD) and CT of binding protein 80 kD(BP-80), a type I integral membrane protein belonging to thevacuolar sorting receptor (VSR) family of proteins (Paris et al.,1997; Li et al., 2002), were shown to be essential and sufficientfor its correct targeting to the PVC, which has been identified asa multivesicular body (MVB) in plant cells (Jiang and Rogers,1998; Tse et al., 2004, 2006; Robinson et al., 2008). In addition,the length of the TMD may affect the subcellular localization ofan integral membrane protein, as green fluorescent protein(GFP) fusions with different TMD sequences of 17, 20, or 23amino acids in length showed distinct subcellular localization tothe ER, Golgi, and plasma membrane (PM), respectively(Brandizzi et al., 2002). Similarly, the TMD length and cytosolic

1 Address correspondence to ljiang@cuhk.edu.hk.The author responsible for distribution of materials integral to the findingspresented in this article in accordance with the policy described in theInstructions for Authors (www.plantcell.org) is: Liwen Jiang (ljiang@cuhk.edu.hk).C Some figures in this article are displayed in color online but in black andwhite in the print edition.WOnline version contains Web-only data.www.plantcell.org/cgi/doi/10.1105/tpc.112.096057

The Plant Cell, Vol. 24: 2086–2104, May 2012, www.plantcell.org ã 2012 American Society of Plant Biologists. All rights reserved.

tail have been shown to be important for proper Golgi localiza-tion of glycosyltransferases, a group of type II integral mem-brane proteins in plant cells (Saint-Jore-Dupas et al., 2006).Similar to yeast and mammalian membrane proteins, various ERexport signals, such as the diacidic DXE motif and dibasic motif,have been identified in different classes of transmembraneproteins in plant cells (Hanton et al., 2005; Yuasa et al., 2005;Mikosch et al., 2006; Schoberer et al., 2009; Zelazny et al.,2009). In addition, the dilysine motif present in the Arabidopsisthaliana p24 (At p24) homolog has also been shown to bindwith the COPI subunit to maintain the ER localization of p24(Contreras et al., 2004b; Langhans et al., 2008). More recently,the rice (Oryza sativa) SECRETORY CARRIER MEMBRANE PRO-TEIN1 (SCAMP1), a polytopic PM protein with four TMDs (Lamet al., 2007a, 2007b, 2008; Law et al., 2012), was shown to reachthe PM via an ER-Golgi-TGN-PM pathway in tobacco (Nicotianatabacum) Bright Yellow-2 cells (Cai et al., 2011). Interestingly, bothits cytosolic sequences and TMDs regulated the proper traffickingsteps from ER to PM owing to the presence of various signalswithin the SCAMP1, including the ER export signal in the cytosolicN terminus, the Golgi export signal within the TMD2-TMD3, and theTGN-PM targeting signal in TMD1 (Cai et al., 2011). However, theunderlying mechanisms remain elusive.

Endomembrane proteins (EMPs) belong to a family of integralmembrane proteins with nine TMDs, a large lumenal N terminus,and a short (10 to 17 amino acids) C terminus located in thecytosol. Three EMP members are found in Dictyostelium dis-coideum (Phg1A to Phg1C) (Cornillon et al., 2000) and Sac-charomyces cerevisiae (Tmn1 to Tmn3) (Froquet et al., 2008) andfour members in human (TM9SF1 to TM9SF4) (Lozupone et al.,2009). Loss of Phg1A function in slime mold D. discoideum ledto a defect in cellular adhesion and inefficient phagocytosis(Cornillon et al., 2000); similarly, Phg1b played a synergistic butnot redundant role in controlling cellular adhesion and phago-cytosis (Benghezal et al., 2003). EMPs in yeast were also foundto be required for cell adhesion and filamentous growth undernitrogen starvation (Froquet et al., 2008). Interestingly, the hu-man EMP protein, TM9SF4, was highly expressed in humanmalignant melanoma cells from metastatic lesions but was un-detectable in healthy human tissues and cells, thus serving asa marker for malignancy (Lozupone et al., 2009). In addition,studies on the subcellular localization of EMPs in yeast andhuman cells yielded different conclusions. For example, myc-tagged human EMP p76 was found predominantly in theendosomes (Schimmöller et al., 1998), and an N-terminal GFP-tagged human EMP (GFP-TM9SF4) was also shown to colo-calize with the Rab5 protein in the early endosome (Lozuponeet al., 2009). By contrast, yeast EMP Yer113c was shown tolocalize to the Golgi apparatus (Huh et al., 2003), whereasa C-terminal GFP-tagged yeast EMP (TMN2-GFP) was recentlyshown to localize to the endosome and vacuole (Aguilar et al.,2010).

The Arabidopsis genome contains 12 EMP isoforms (termedEMP1 to EMP12 in this study), but little is known about theirprotein subcellular localization and function in plants. As a firststep to study the biology of EMPs in plants, we performedstudies on the subcellular localization and targeting mecha-nisms of EMP12 in Arabidopsis. We demonstrated that both

endogenous EMP12 and GFP-EMP12 fusion (as detected byEMP12 and GFP antibodies, respectively) were localized to theGolgi apparatus in wild-type and transgenic Arabidopsis plantsand in transiently expressed Arabidopsis protoplasts. We furthershowed that the cytosolic C terminus of EMP12 contained bothER export and Golgi retention signals that interact with COPIIand COPI subunits, respectively. In addition, this Golgi retentionsignal of EMP12 trapped several post-Golgi membrane proteinswithin the Golgi apparatus in a gain-of-function analysis usingtransiently expressed Arabidopsis protoplasts. It seems thatthese targeting features of EMP12 are conserved among theplant EMPs for their correct targeting to the Golgi apparatus inplant cells.

RESULTS

EMP12 Is a Golgi-Localized Protein with Multiple TMDs

All 12 isoforms of Arabidopsis EMP proteins (termed EMP1 toEMP12 in this study; see Supplemental Figure 1 online) sharehigh similarity at the amino acid level and in predicted topology.All At EMPs are predicted to have a large lumenal N-terminaldomain, followed by nine TMDs and a short CT (Figure 1A).Eleven At EMP members (excluding EMP1) are predicted tohave a signal peptide (SP) in their N terminus. As a first step tounderstanding the biology of EMPs in plants, we analyzed thesubcellular localization and trafficking route of EMP12 in Arab-idopsis.We first used a GFP fusion approach by making a GFP-

EMP12 construct in which a SP-GFP was used to replace theSP of EMP12 (Figure 1B). To test if this new fusion maintainsthe identical topology as EMP12, we performed a proteaseprotection assay using microsomes isolated from Arabidopsiscells expressing the GFP-EMP12 fusion, followed by immuno-blot analysis using GFP antibody (Figure 1C). The N terminus ofGFP-EMP12 was found to be resistant to protease digestionbecause the 47-kD GFP-TMD1 fragment remained intact in thepresence of trypsin, whereas the 75-kD full-length GFP-EMP12fusion did not (Figure 1C, lanes 1 and 2), indicating the lumenallocation of the GFP tag as predicted (Figure 1A). The reliability ofsuch a protease protection assay was further verified by anidentical experiment using a GFP fusion of the type I integralmembrane protein GFP-VSR2 as a control, in which the lumenalGFP remained intact upon protease digestion (Figure 1C, lanes4 and 5; Cai et al., 2011).To study the subcellular localization of EMP12 and GFP-

EMP12, we next generated EMP12 antibodies and transgenicArabidopsis plants expressing GFP-EMP12 for immunoblotanalysis. As shown in Figure 2A, GFP antibody detected a singleprotein band around 75 kD, likely representing GFP-EMP12 inthe cell membrane (CM) fraction but not the cell-soluble (CS)fraction of the transgenic Arabidopsis seedlings expressingGFP-EMP12, whereas no such protein band was detected inwild-type plants. By contrast, anti-EMP12 antibodies detectedthe putative 50-kD endogenous EMP protein band in the CMfraction of both wild-type and GFP-EMP12 plants (Figure 2A,lanes 6 and 8, asterisk) as well as the GFP-EMP12 fusion (lane 8,

Targeting and Subcellular Localization of EMP12 2087

arrowhead). A subcellular fractionation study using a continuoussucrose gradient and wild-type Arabidopsis cells showed thatanti-EMP12 had an identical distribution pattern as the cis-Golgimarker anti-Man1 (Tse et al., 2004; Lam et al., 2007b), thus in-dicating a possible Golgi localization of EMP12 (Figure 2B). Tofurther prove this, confocal immunofluorescence labeling wasnext performed in root cells of transgenic Arabidopsis plantsexpressing GFP-EMP12. As shown in Figure 2C, the punctateGFP signals of GFP-EMP12 were fully colocalized with anti-EMP12. In addition, anti-EMP12 was found to fully colocalizewith the cis-Golgi marker yellow fluorescent protein (YFP)-SYP32 but was largely separated from the PVC marker YFP-ARA7 (Geldner et al., 2009) in these transgenic Arabidopsis rootcells (Figure 2C), indicating that both endogenous EMP12 andGFP-EMP12 fusion are localized to the Golgi apparatus inArabidopsis cells. Thus, GFP-EMP12 can be used to study thesubcellular localization and trafficking of EMP12 in plant cells.

To further confirm the Golgi localization of EMP12 and GFP-EMP12 proteins from the above confocal immunofluorescencestudy, we next performed immunogold electron microscopy(EM) using either EMP12 or GFP antibodies on ultrathin sections

prepared from high-pressure frozen/freeze-substituted root tipsof wild-type or transgenic GFP-EMP12 Arabidopsis plants. Asshown in Figure 3, in wild-type Arabidopsis root cells, anti-EMP12 labeled the Golgi apparatus in which the majority of thegold particles (>80%) were found in the cis- and medial-Golgi(Figure 3A). Similar patterns of Golgi localization were also ob-tained in GFP-EMP12–transformed Arabidopsis root cells usinganti-GFP antibodies (Figure 3B). In addition, Golgi labeling withEMP12 or GFP antibodies was specific, as little labeling wasseen in other organelles, including TGN, MVB, and mitochondriain these experiments (Figure 3, Table 1). Taken together, theseconfocal immunofluorescence and immunogold EM studiesdemonstrated that EMP12 and GFP-EMP12 localized to theGolgi apparatus with preferential localization to the cis- andmedial-Golgi in Arabidopsis cells.We next wanted to test if the established transient expression

system of Arabidopsis protoplasts (Miao and Jiang, 2007) couldbe used to study the trafficking of GFP-EMP12. Consistent withthe above immunofluorescence and immunogold EM studies intransgenic Arabidopsis root cells (Figures 2 and 3), GFP-EMP12was found to be fully colocalized with the cis-Golgi marker

Figure 1. Topology Analysis of EMP12.

(A) Predicated topology of EMP12 using TMHMM Server 2.0.(B) Construction of GFP-EMP12 and GFP-VSR2 fusions.(C) Protease protection assay. Microsomes were isolated from the Arabidopsis (At) protoplasts expressing GFP-EMP12 or GFP-VSR2 for trypsindigestion, followed by protein extraction and separation via SDS-PAGE for immunoblot analysis using GFP antibody. M, molecular weight marker inkilodaltons.[See online article for color version of this figure.]

2088 The Plant Cell

Man1-mRFP (for monomeric red fluorescent protein; Tse et al.,2004; Lam et al., 2007b) when they were transiently coex-pressed together in Arabidopsis protoplasts (Figure 4A). Bycontrast, GFP-EMP12 was largely separated from the TGNmarker mRFP-SYP61 and the PVC marker mRFP-VSR2 inArabidopsis protoplasts (Figures 4B and 4C). In addition, whenthe ER-to-Golgi traffic was blocked by overexpression ofSec12p, a Sar1p-specific guanosine nucleotide exchange factorfor COPII vesicle formation (Pimpl et al., 2003; Cai et al., 2011),both GFP-EMP12 and Man1-mRFP were trapped and colo-calized within the ER (Figure 4D), indicating that GFP-EMP12reached the Golgi via the COPII-dependent ER-to-Golgi sortingroute, a result consistent with the Golgi localization of Man1-mRFP (Langhans et al., 2008).

The Cytosolically Exposed C-Terminal Region ContainsEssential ER Export Signals for the Trafficking of EMP12from the ER to the Golgi

Since GFP-EMP12 is fully colocalized with the endogenousEMP12 (as detected by anti-EMP12) to the Golgi apparatus inArabidopsis cells (Figure 2) or with the Golgi marker in Arabi-dopsis protoplasts (Figure 4), we thus used Arabidopsis proto-plasts to study the targeting of GFP-EMP12. Since little isknown about the Golgi targeting of EMPs with nine TMDs, wefirst tested whether the large lumenal N terminus or the shortcytosolic C terminus are essential for Golgi localization of GFP-EMP12 by generating two truncated constructs, GFP-EMP12(DN) and GFP-EMP12(DC), in which either the lumenal N terminus

Figure 2. Golgi Localization of Endogenous EMP12 and GFP-EMP12 Fusion in Arabidopsis.

(A) Characterization of EMP12 antibody. Proteins were isolated from wild-type (WT) or transgenic Arabidopsis plants expressing GFP-EMP12, followedby SDS-PAGE and immunoblot analysis using GFP or EMP12 antibodies. Both anti-GFP and anti-EMP12 recognized the full-length GFP-EMP12 fusion(lanes 4 and 8), whereas anti-EMP12 also detected the endogenous EMP12 protein in both wild-type and transgenic Arabidopsis plants (lanes 6 and 8).Arrowhead and asterisk indicate GFP-EMP12 fusion and endogenous EMP12, respectively.(B) Subcellular fractionation analysis of endogenous EMP12. The microsomes prepared from wild-type Arabidopsis (At) cells were fractionated byultracentrifugation on a continuous (25 to 50%) Suc gradient. Proteins extracted from individual fractions were probed with anti-EMP12, anti-Man1(Golgi marker), and anti-VSR (PVC marker) as indicated.(C) Golgi localization of EMP12 and GFP-EMP12 fusion in transgenic Arabidopsis plants. Confocal immunofluorescence labeling showing full colo-calization of anti-EMP12 with GFP-EMP12 (panel 1) or Golgi marker YFP-SYP32 (panel 2) in root cells of transgenic Arabidopsis plants expressing thesetwo XFP fusions but distinct localization from the PVC marker YFP-ARA7 in root cells of a transgenic Arabidopsis plant (panel 3). The insets representa three times enlargement of a randomly selected area. DIC, differential interference contrast. Bars = 10 mm.

Targeting and Subcellular Localization of EMP12 2089

or the cytosolic CT was deleted, respectively. As shown inSupplemental Figure 2 online, upon transient expression inArabidopsis protoplasts, GFP-EMP12(DN) remained in the Golgias it colocalized with the Golgi marker, indicating that the Nterminus of EMP12 was not essential for its Golgi localization.By contrast, CT deletion resulted in ER localization of EMP12because GFP-EMP12(DC) fully colocalized with the ER markerCalnexin-mRFP upon their coexpression, indicating the possibleexistence of an ER export signal at the CT of EMP12.

So far, various ER export signals have been identified at thecytoplasmic part of integral membrane proteins. Best known

examples are the diacidic motif DXE within the vesicular sto-matitis virus glycoprotein cytosolic tail, the dihydrophobic motifs(FF, YY, LL, or FY) found in the p24 family proteins, and theTyro-based signals identified in the tail sequence of ERGIC53/Emp47p family proteins (Nishimura and Balch, 1997; Nufer et al.,2002; Sato and Nakano, 2002; Langhans et al., 2008). Alignmentanalysis of CTs of Arabidopsis EMPs indicates the existence ofa putative dihydrophobic motif and a Tyr residue in all At EMPs(see Supplemental Figure 1 online). To find out if the putativedihydrophobic motif and Tyr residue in the EMP12 CT(SNLFVRRIYRNIKCD) are important for the ER export of EMP12,

Figure 3. Immunogold EM Localization of EMP12 and GFP-EMP12.

Ultrathin sections were prepared from high-pressure frozen/freeze-substituted samples of wild-type (WT) or transgenic GFP-EMP12 Arabidopsis roottips, followed by immunogold labeling using either EMP12 antibodies (A) or GFP antibodies (B), respectively. The relative percentage distribution ofgold particles among cis-,medial-, and trans-Golgi was also quantified. Arrows indicate examples of gold particles present in the Golgi. CW, cell wall; G,Golgi; M, MVB; V, vacuole. Bars = 100 nm.

Table 1. Distribution of Immunogold Particles with EMP12 and GFP Antibodies in Wild-Type and Transgenic GFP-EMP12 Arabidopsis Root Cells

Organelle No. of Organelles No. of GPs GPs per Organelle

Wild type (Anti-EMP12)Golgi 30 94 3.03**TGN 30 7 0.23**MVB 30 5 0.17Mitochondrion 30 6 0.20

GFP-EMP12 (anti-GFP)Golgi 30 128 4.27**TGN 30 8 0.27**MVB 30 4 0.13Mitochondrion 30 5 0.17

Significant difference between the distribution in the Golgi and TGN was analyzed using the two-tailed paired t test (**P < 0.001). Data were collectedand analyzed from three independent labeling experiments. GP, immunogold particles.

2090 The Plant Cell

we performed mutagenesis analysis on the dihydrophobic or Tyrmotifs of GFP-EMP12 CT (Figure 5A). Mutations of dihy-drophobic combined with Tyr motifs totally block the ER exportof GFP-EMP12, which was fully colocalized with the ER markerCalnexin-mRFP (Figure 5B). Interestingly, the solo Tyr mutationonly partially inhibited the ER export of GFP-EMP12 (Figures 5Cand 5D). However, the combined mutations of Tyr with onehydrophobic amino acid, Phe or Val, effectively blocked the ERexport of GFP-EMP12 (Figures 6A and 6B), indicating that thedihydrophobic residue synergistically combined with Tyr to drivethe ER export of EMP12. This conclusion was further confirmedby double mutation of the dihydrophobic motif, Phe and Val, thatonly partially blocked the ER export of EMP12 as they colo-calized with both Golgi and ER markers (Figures 6C and 6D);however, single mutation on one hydrophobic amino acid, Pheor Val, had no effect on the ER export of EMP12 (Figures 5E and6E). Mutations of other amino acids besides the dihydrophobicor Tyr residues did not affect the ER export of GFP-EMP12 (seeSupplemental Figure 3 online). Taken together, these resultssupport the conclusion that both the hydrophobic motif and Tyrresidue in the cytosolic tail play essential roles in the ER exportof EMP12.

Transmembrane cargos usually possess a binding affinity withSar1–Sec23–Sec24 prebudding complexes through their ERexport signals in order to be included in COPII vesicles for ERexport (Bi et al., 2002; Barlowe, 2003). We next wanted to findout if COPII subunits interact with the ER export motifs in theEMP12 CT using in vitro peptide binding experiments (Contreraset al., 2004b). The soluble fractions isolated from Arabidopsisprotoplasts expressing YFP-Sec24 were used as a source ofCOPII subunits (Hanton et al., 2009) for binding with columnsconjugated with synthetic peptides of EMP12 CT or its mutationon the dihydrophobic/Tyr residues. As shown in Figure 7, theexpressed YFP-Sec24 was retained in the column with theEMP12 CT peptides (lane 3), but such binding was dramaticallyreduced in the mutated column (lane 4), indicating that the di-hydrophobic /Tyr residues of EMP12 CT preferably bound to theCOPII Sec24 subunit.

The C-Terminal-Fused GFP Tag Causes Mislocalization ofthe EMP12-GFP Fusion to the TGN, PVC, and Vacuole

Interestingly, when GFP was fused at the C terminus of EMP12(Figure 8A), the resulting EMP12-GFP was found to localize to

Figure 4. Golgi Localization of GFP-EMP12 in Arabidopsis Protoplasts.

GFP-EMP12 was coexpressed with various known organelle markers as indicated ([A] to [C]) in Arabidopsis protoplasts, followed by confocal imagingat 12 to 14 h after transfection, whereas overexpression of Sec12p was used to inhibit the ER export of GFP-EMP12 and Man1-mRFP (D). Bars = 10 mm.

Targeting and Subcellular Localization of EMP12 2091

both the vacuole and punctate structures distinct from the Golgimarker Man1-mRFP (Figure 8B), but partially colocalized withthe TGN marker mRFP-SYP61 and the PVC marker mRFP-VSR2(Figures 8C and 8D) in Arabidopsis protoplasts, indicating thatGFP fusion at the C terminus of EMP12 caused it to move fromthe Golgi to the vacuole via TGN/PVC. In addition, EMP12-GFPalso passed through the ER-Golgi sorting route via the COPIIvesicle en route to the vacuole because overexpression ofSec12p trapped it within the ER (Figure 8E), indicating that theC-terminal GFP fusion did not affect the ER-to-Golgi transport

but caused mislocalization of EMP12-GFP to post-Golgi com-partments, including TGN, PVC, and the vacuole.

The EMP12 CT Contains a Novel KXD/E Motif forGolgi Retention

To further find out if the EMP12 CT is essential for its Golgiretention, we used a gain-of-function approach and added anadditional EMP12 CT sequence to the C terminus of EMP12-GFP (Figure 9A). Indeed, the resulting EMP12-GFP-CT was

Figure 5. The EMP12 C Terminus Contained Multiple ER Export Signals.

(A) The GFP-EMP12 C terminus mutation constructs and their subcellular localization in Arabidopsis protoplasts. CT, C terminus; NT, N terminus; T,TMD; WT, the wild type.(B) to (E) Typical subcellular localization patterns of four selective GFP-EMP12 C terminus mutation fusions upon their coexpression with knownmarkers (as indicated) in Arabidopsis protoplasts. Confocal images were collected from cells at 12 to 14 h after transfection. Bars = 10 mm.

2092 The Plant Cell

found to remain in the Golgi as it colocalized with the Golgimarker Man1-mRFP but largely separated from the TGN andPVC markers upon their transient coexpression in Arabidopsisprotoplasts (see Supplemental Figure 4 online). Further muta-genesis (deletion) analysis of the EMP12 CT demonstrated thatthe last six amino acids RNIKCD (EMP12-GFP-RNIKCD; Figure9B) or even the last three residues KCD (EMP12-GFP-KCD; seeSupplemental Figure 4 online) were sufficient to retain theEMP12-GFP within the Golgi or efficiently retrieve EMP12-GFPback to the Golgi apparatus.

Interestingly, sequence alignment analysis of the CTs of EMPhomologs from Arabidopsis, rice, and maize (Zea mays) showedthe presence of a highly conserved Lys in the –3 position and anacidic residue (Asp or Glu) in the –1 position (see Supplemental

Figure 1D online). To test if this conserved Lys-based motif KXD/E was essential for Golgi retention, we performed point mutationanalysis on this motif via substitution with Ala residues. Doublepoint mutations on Lys and Asp (EMP12-GFP-RNIACA; seeSupplemental Figure 5 online) or single mutation on Lys(EMP12-GFP-RNIACD; Figure 9C) or Asp (EMP12-GFP-RNIK-CA; Figure 9D) all resulted in mislocalization of these GFP fu-sions to vacuole and post-Golgi PVC organelles that werelargely distinct from the Golgi apparatus (see SupplementalFigures 5A to 5G online). Thus, it seems that both K and D/E inthe KXD/E motif are essential for Golgi localization or retention ofEMP12. In addition, the KXD/E motif must be presented at theC-terminal end to function as a Golgi retention signal, as the additionof three Ala residues at the C terminus of the Golgi-localized

Figure 6. Subcellular Localization of GFP-EMP12 Harboring Mutations on the ER Export Signals.

These various GFP fusions were coexpressed with known markers (as indicated) in Arabidopsis protoplasts. Confocal images were collected from cellsat 12 to 14 h after transfection. Bars = 10 mm.

Targeting and Subcellular Localization of EMP12 2093

EMP12-GFP-RNIKCD caused its mislocalization to the vacuoleand PVC (Figure 9E; see Supplemental Figures 5H and 5I online).The PVC-mediated degradation of these mislocalized EMP12-GFP fusions was further proved by coexpression with a consti-tutively active Rab5 mutant, ARA7(Q69L), which can induce PVCenlargement and trap the degradative cargo proteins within itslumen (Kotzer et al., 2004). Indeed, both EMP12-GFP andEMP12-GFP-RNIKCDAAA occurred as large round dots withinthe lumen of mRFP-ARA7(Q69L)–labeled ring-like structures,distinct from the punctate patterns of Golgi-localized GFP-EMP12and EMP12-GFP-RNIKCD (see Supplemental Figure 6 online).

The detection of GFP signals in the vacuole in cells expressingGFP-tagged transmembrane proteins usually indicates the occur-rence of protein degradation (Kleine-Vehn et al., 2008; Langhanset al., 2008; Foresti et al., 2010; Cai et al., 2012). To provide furtherevidence that these various EMP12-GFP-RNIKCD mutation fusionswere degraded in vacuoles upon their expression in Arabidopsisprotoplasts (e.g., Figure 9), we performed immunoblot analysis usingGFP antibodies. As shown in Figure 10, upon expression in Arabi-dopsis protoplasts, the Golgi-localized EMP12-GFP-RNIKCD largely(>90%) remained as the full-length protein in the CM fraction, while<10% of the signal was detected as the GFP core (Foresti et al.,2010) in the CS fraction. By contrast, in cells expressing the threeother constructs with a mutation on the KXD/E motif (EMP12-GFP-RNIACD and -RNIKCA) or C-terminal addition (–RNIKCDAAA),>50% (50 to 70%) of the detected signals were found as the GFPcore in the CS fraction (Figure 8C), indicating that their turnover rateis much higher than that of EMP12-GFP-RNIKCD. Quantificationanalysis of these GFP signals in the CM versus CS fractions ofindividual proteins further confirmed this conclusion (Figure 10C).Thus, both confocal data (Figure 9) and immunoblot analysis (Figure10) point to the same conclusion that the C-terminal KXD/E motiffunctions as a Golgi retention signal for EMP12.

The KXD/E Golgi Retention Motif Can Interact withCOPI Vesicles

To further explore the molecular mechanism of KXD/E motiffunction in Golgi retention, we next used the synthetic peptides

corresponding to the EMP12 CT as bait to isolate possible in-teracting proteins from wild-type Arabidopsis proteins in pull-down experiments, followed by protein elution, separation viaSDS-PAGE, and tandem mass spectrometry (MS/MS) analysisfor protein identification. In duplicate experiments, several pro-teins that were specifically pulled out by the EMP12 CT peptidecolumn were identified as the COPI coat subunits (seeSupplemental Figure 7 online). The COPI vesicles are the pre-dominant means by which the membrane proteins recycle backto the ER or are retained in the Golgi (Cosson and Letourneur,1994; Tu et al., 2008). We thus further tested whether COPIcoatomers could bind with the KXD/E motif using in vitro peptidebinding experiments as described previously (Contreras et al.,2004b). The soluble fraction isolated from Arabidopsis proto-plasts expressing Sec21-Myc, an Arabidopsis g-COP protein(Movafeghi et al., 1999), was used as a source of COPI coatprotein in the binding experiment using the synthetic peptide ofEMP12 CT or its mutants of the KXD/E motif as binding baits. Asshown in Figure 11, Sec21-Myc specifically binds to the EMP12CT as detected by anti-Myc antibody (Figure 11, lane 3), butsuch binding was completely abolished when the KXD motif inthe CT was mutated to either ACA or ACD (Figure 11, lanes4 and 5), while the binding ability was reduced significantly forthe KCA peptide (Figure 11, lane 6), indicating that the Lys in theKXD motif played a more critical role in COPI binding. Thesebinding results are consistent with those obtained with confocalsubcellular localization and immunoblot analysis of EMP12-GFP-RNIKCA (Figures 9 and 10). Therefore, binding with theCOPI coatomer was a means by which the KXD/E motif achievedthe Golgi retention function.

RNIKCD Functions as a Golgi Retention Motif for Post-GolgiMembrane Proteins

To find out if the Golgi retention signal RNIKCD identified in theEMP12 CT would have a similar function in other membraneproteins, we next used a gain-of-function approach by addingthis motif to the C terminus of GFP fusions of two selective post-Golgi membrane proteins: SCAMP1-GFP and TPK1-GFP (Figure12). GFP fusions with the SCAMP1 (SCAMP1-GFP or GFP-SCAMP1), an integral membrane protein with four TMDs, werelocalized to both the TGN and PM via an ER-Golgi-TGN-PMpathway (Lam et al., 2007b; Cai et al., 2011). Indeed, the originalTGN/PM localization of SCAMP1-GFP was altered by the ad-dition of the RNIKCD motif to the C terminus of SCAMP1-GFPbecause SCAMP1-GFP-RNIKCD was largely colocalized withthe Golgi marker Man1-mRFP (Figure 12A, panels 1 and 2) inArabidopsis protoplasts. A similar result was obtained when theRNIKCD motif was fused to the C terminus of GFP-SCAMP1because the resulting GFP-SCAMP1-RNIKCD fusion was alsofound to be trapped in the Golgi, colocalizing with the Golgimarker Man1-mRFP (Figure 12A, panel 3), indicating that theRNIKCD motif showed similar Golgi retention function evenwithout direct fusion with GFP.The Arabidopsis two-pore potassium channel (TPK1) protein

is a tonoplast-localized potassium channel with four TMDs anda cytoplasmic N terminus and C terminus (Gobert et al., 2007;Latz et al., 2007; Dunkel et al., 2008; Isayenkov et al., 2011). As

Figure 7. The ER Export Signal at the C Terminus of EMP12 Binding toCOPII Subunit in a Sequence-Specific Manner.

Soluble proteins extracted from Arabidopsis protoplasts expressingYFP-Sec24 were incubated with Sepharose beads conjugated withsynthetic peptides corresponding to the wild type (WT; lane 3) or mu-tated (mut; lane 4) C terminus of EMP12 for a pull-down assay, followedby immunoblot detection of bound proteins using a GFP antibody. M,molecular weight marker in kilodaltons.

2094 The Plant Cell

expected, TPK1-GFP showed a typical tonoplast localizationpattern that was distinct from the Golgi marker Man1-mRFP inArabidopsis protoplasts (Figure 12B). However, when RNIKCDwas added to its C terminus, the resulting TPK1-GFP- RNIKCDwas found to localize to both Golgi (colocalized with the Golgimarker Man1-mRFP) and tonoplast (Figure 12B), indicating thatRNIKCD partially trapped the tonoplast-localized TPK1-GFPwithin the Golgi. Such dual localization of TPK1-GFP-RNIKCD inthe Golgi and tonoplast may be caused by the competitionbetween the KXD/E Golgi retention signal and the tonoplasttargeting signals found in TPK1 (Dunkel et al., 2008; Isayenkovet al., 2011). Thus, it will be interesting to investigate the com-binational effect of these distinct targeting signals in determiningtargeting of these proteins. Taken together, the C-terminal ad-dition of RNIKCD caused Golgi retention of two post-Golgimembrane proteins, SCAMP1-GFP and TPK1-GFP, which wereoriginally localized in the TGN/PM and tonoplast, respectively. In

addition, since most of these targeting studies were performedusing the Arabidopsis protoplast transient expression system, itwill be of interest to generate transgenic tobacco Bright Yellow-2 or Arabidopsis cell lines and transgenic Arabidopsis plantsexpressing some of the key fusion constructs (e.g., EMP12-GFP-CT, SCAMP1-GFP-RNIKCD, and TPK1-GFP-RNIKCD) foruse in future studies.

DISCUSSION

The Position of the GFP Tag Affects the Proper GolgiLocalization of EMP12

Reporter fusion proteins have been a useful tool for studyingthe targeting and subcellular localization of integral mem-brane proteins in plant cells. Because of the complex natureof membrane protein topology, care must be taken when

Figure 8. EMP12-GFP Localized to the TGN, PVC, and Vacuole.

(A) EMP12-GFP construct with GFP fused at the C terminus of EMP12. CT, C terminus; NT, N terminus; T, TMD.(B) to (D) Subcellular localization of EMP12-GFP upon its coexpression with known organelle markers (as indicated) inArabidopsis protoplasts.(E) Overexpression of Sec12p was used to inhibit the ER export of GFP-EMP12 and Man1-mRFP in cells coexpressingthese two fusion proteins. Confocal images were collected from cells at 12 to14 h after transfection. Bars = 10 mm.

Targeting and Subcellular Localization of EMP12 2095

a reporter (e.g., GFP) is fused to an integral membrane proteinto ensure that the fusion protein’s topology remains identicaland that the targeting signals are not affected. For example,a SP-GFP was used to replace the N terminus of BP-80,a type I integral membrane protein belonging to the VSRfamily, to generate the GFP-BP-80 reporter that was colo-calized with the endogenous VSRs to PVC/MVB (Tse et al.,2004), demonstrating the correct targeting of GFP-BP-80 andvarious GFP-VSR fusions in plant cells (Jiang and Rogers,1998; Tse et al., 2004; Miao et al., 2006). In this study, weadopted a similar approach and used SP-GFP to replace theN terminus of EMP12, whereas the generated GFP-EMP12fusion was shown to colocalize with the endogenous EMP12

proteins to the Golgi apparatus in transgenic Arabidopsisplants and protoplasts (Figures 2 to 4). Similar Golgi localizationwas shown for GFP fusions to three other EMPs (EMP2, EMP3,and EMP8) in Arabidopsis protoplasts (see Supplemental Figure8 online). Such Golgi localization of EMPs in Arabidopsis isconsistent with previous results obtained from an organelleproteome study using Arabidopsis culture cells (Dunkley et al.,2006). By contrast, N-terminal GFP fusion with the human EMPprotein (i.e., GFP-TM9SF4) was shown to localize to early en-dosomes in mammalian cells (Lozupone et al., 2009). Suchvariation in EMP localization (i.e., Golgi versus endosome lo-calization in plant versus mammalian cells) suggests that EMPsevolved distinct functions in different organisms.

Figure 9. The C Terminus of EMP12 Contained a Golgi Retention Signal.

(A) EMP12-GFP-CT mutation constructs and their subcellular localization in Arabidopsis protoplasts. CT, C terminus; NT, N terminus; T, TMD.(B) to (E) Subcellular localization patterns of four selective EMP12-GFP-CT mutation fusions with or without Golgi retention signals at their CT upon theircoexpression with Golgi or PVC marker in Arabidopsis protoplasts. Confocal images were collected from cells at 12 to 14 h after transfection. Bars = 10 mm.

2096 The Plant Cell

Interestingly, C-terminal GFP fusion to EMP12 caused mis-targeting of EMP12-GFP to post-Golgi compartments and, ul-timately, degradation in vacuoles (Figures 8 and 13); a traffickingpathway previously demonstrated to be mediated by PVCs (Tseet al., 2004; Miao et al., 2008). Indeed, EMP12-GFP was foundto internalize into the ARA7(Q69L)-induced enlarged PVCs(Kotzer et al., 2004) prior to reaching the vacuole. The mis-localization of EMP12-GFP is likely caused by masking of theKCD Golgi retention signal within the short CT of EMP12 be-cause C-terminal fusions of both GFP (Figure 8) and short triplepeptide sequences (AAA) (Figure 9) resulted in its mislocalizationto the post-Golgi compartment. Such mislocalization uponC-terminal fusion was also true for other tested EMP members,including EMP2, EMP3, and EMP8 (see Supplemental Figure 8online), indicating a general targeting mechanism for the EMPsin Arabidopsis plants. These results support the idea that thevacuole rather than the PM is the default destination for mis-sorted or misfolded integral membrane proteins in plant cells(Höfte and Chrispeels, 1992). In addition, our study might ex-plain the different subcellular localizations of two EMPs in yeast:While the localization of Yer113c to the Golgi (Huh et al., 2003) isconsistent with that of plant EMPs, the localization of a C-terminalGFP-tagged yeast EMP (TMN2-GFP) to the endosome and vac-uole (Aguilar et al., 2010), rather than the Golgi, was likely causedby the C-terminal GFP fusion masking the Golgi retention signalas described in this study. It would thus be interesting to find out ifGFP-TMN2 would localize to the Golgi apparatus in yeast infuture experiments.

Multiple Sorting Signals and Proper COPII Vesicle FunctionAre Involved in ER Export of EMP12

In the secretory pathway of eukaryotic cells, soluble and mem-brane cargo proteins are packed into ER-derived COPII vesiclesfor Golgi and post-Golgi transport. The COPII coat consists of

Figure 10. Immunoblot Analysis of Protoplasts Expressing the Four EMP12-GFP Fusions with the Golgi Retention Motif or Its Mutation.

(A) Arabidopsis protoplasts expressing the four EMP12-GFP fusions (as described in Figures 7B to 7E) were subjected to protein isolation into CS andCM fractions, followed by protein separation via SDS-PAGE and protein gel blot analysis using GFP antibody. Arrowhead and arrow indicated the full-length EMP12-GFP fusion and its degradative GFP core, respectively. M, molecular weight marker in kilodaltons.(B) Coomassie blue–stained gel was used as control for showing equal protein loading.(C) The relative grayscale intensity (scale 0 to 1.0) of the two protein bands in the CM (full-length fusion) and in the CS (GFP core) fractions of individualEMP12-GFP fusions was quantified using ImageJ software, in which the combinational intensity of CM plus CS was expressed as 1.0 (or 100%).

Figure 11. The Cytoplasmic C Terminus of EMP12 Binds to COPISubunit in a Sequence-Specific Manner.

Soluble proteins extracted from Arabidopsis protoplasts expressingSec21-Myc were incubated with Sepharose beads conjugated withsynthetic peptides corresponding to the wild-type (WT) C terminus ofEMP12 (lane 3) or its various mutations (lanes 4 to 6; K587A/D589A,K587A, and D589A) for a pull-down assay, followed by immunoblotdetection of bound proteins using anti-Myc antibodies. Blank Sepharosebeads without peptide conjugation were used as a control. M, molecularweight marker in kilodaltons.

Targeting and Subcellular Localization of EMP12 2097

Figure 12. The Golgi Retention Motif of EMP12, RNIKCD, Retained Post-Golgi Membrane Proteins within the Golgi.

(A) C-terminal fusion of RNIKCD to SCAMP1-GFP or GFP-SCAMP1 caused their relocalization from their original PM/TGN (panel 1) to the Golgiapparatus (panels 2 and 3). Bars = 10 mm.(B) C-terminal fusion of RNIKCD to the tonoplast-localized TPK1-GFP (panel 4) partially trapped TPK1-GFP within the Golgi (panel 5). Bars = 10 mm.

Sar1, Sec23-Sec24, and Sec13-Sec31 complexes that are se-quentially recruited to the ER membrane (Barlowe, 2003).Transmembrane cargo proteins usually contain ER export sig-nals in their cytoplasmic region that interact with the COPII coatcomplex. Typical ER export signals identified in membranecargo proteins in yeast, mammalian, and plant cells include thediacidic motif (DXE), dihydrophobic motif (LL), diaromatic motif(FF, YY), dibasic motif (RR, RK), and Tyr-based motif YXXF(where X is any amino acid, and F is a bulky hydrophobic group)(Sato and Nakano, 2002; Barlowe, 2003; Hanton et al., 2005).

In this study, using both loss-of-function and gain-of-functionapproaches in transiently transformed Arabidopsis protoplasts,we also identified the dihydrophobic and Tyr-based motif, FV/Y,at the short cytosolic C terminus of EMP12 to be an ER exportsignal for EMPs in plants. FV and Y in EMP12 may functionsynergically as the ER export signals to facilitate its ER exportand Golgi localization, as triple mutations of FV/Y resulted incomplete ER localization, whereas double mutations of FV ora single mutation of Y resulted in dual ER and Golgi localization.Furthermore, an in vitro peptide binding assay demonstratedthat triple mutations of FV/Y in the C terminus of EMP12 causeda dramatic reduction in the original interaction between the

COPII coat protein Sec24 and EMP12 CT. Since the FV/Y motifis highly conserved in the C terminus of all plant EMPs, includingArabidopsis, rice, and maize, this motif may represent a generalmechanism that regulates the ER export of EMPs in plants. Inaddition, when all EMP12 fusions and their mutants were ex-pressed alone, they showed identical patterns as those coex-pressed with other markers (see Supplemental Figure 9 online),thus verifying that coexpression with the organelle markers didnot affect the localization of EMP12 fusions and its mutants inthis study.

The KXD/E Motif and COPI Vesicle Mediated GolgiLocalization of EMP12

In eukaryotic cells, the best-known signal-mediated Golgi re-tention mechanism for membrane proteins is related to COPIfunction. In yeast, the Golgi-localized glycosyltransferases con-tain the semiconserved motif (F/L)-(L/V)-(S/T) that directly inter-acts with the Golgi resident protein Vps74p (Schmitz et al., 2008;Tu et al., 2008). Vps74 was shown to interact with multiplesubunits of the COPI coat, and Vps74 thus likely served as anadapter in mediating cargo packaging into the retrograde COPI

Figure 13. Working Model of EMP12 Trafficking in Plant Cells.

(A) Structure and topology of EMP12, GFP-EMP12, and EMP12-GFP fusions. Both EMP12 and GFP-EMP12 reached Golgi via ER. The ER exportmotifs FV/Y of EMP12 interacted with the COPII subunit for its transport via COPII vesicles to the Golgi apparatus for residency because of the presenceof the Golgi retention motif KXD and its interaction with the COPI subunit.(B) Pathways and signals. EMP12-GFP reached the Golgi via the ER. Since its Golgi retention motif KXD was masked by C-terminal GFP fusion,EMP12-GFP reached the vacuole via the TGN and PVC.[See online article for color version of this figure.]

Targeting and Subcellular Localization of EMP12 2099

vesicles for retrograde transport in maintaining the steady stateGolgi localization of glycosyltransferases (Tu et al., 2008). Sim-ilarly, the localization of Emp46p family proteins to the Golgi inyeast was achieved via direct interaction of their dilysine signalswith the COPI coat (Sato and Nakano, 2002). Indeed, thedilysine motifs (KKXX or KXKXX), which are usually located inthe cytoplasmic C terminus of membrane proteins, have beenshown to bind with the COPI subunit, especially the g-subunit ofcoatomer, in yeast, mammalian, and plant cells (Cosson andLetourneur, 1994; Harter and Wieland, 1998; Contreras et al.,2004a).

In this study, we demonstrated that the KXD/E motif, whichoccurs within the last six amino acids of the C terminus ofEMP12, functioned as a Golgi retention signal that regulates theGolgi localization of EMP12 via its interaction with the COPIvesicle. This conclusion is supported by several lines of exper-imental evidence. First, mutation of the KCD residues causedmistargeting of EMP12. Second, blockage of this cytosolicsignal by C-terminal GFP or triple peptide AAA fusion resulted inmislocalization of the fusion protein. Third, this motif interactedspecifically with the COPI subunit. Last, when attached to otherpost-Golgi membrane proteins, the motif functioned as a Golgiretention signal that retained these proteins largely within theGolgi. Similar to the ER export signal, this KXD/E Golgi retentionmotif is also highly conserved within the cytosolic C terminus ofall plant EMPs, indicating a general mechanism of Golgi re-tention for EMPs in plants. Thus, this KXD/E motif representsa newly identified Golgi retention signal for multiple TMD pro-teins in eukaryotes.

Interestingly, although the KXD/E motif in EMPs is differentfrom the dilysine motif KKXX or KXKXX identified in single TMDproteins, the Lys in the –3 position of these motifs is highlyconserved, indicating the COPI binding ability of the KXD/Emotif because the Lys in the –3 position of KKXX or KXKXX hasbeen shown to play a more vital role in COPI binding (Harter andWieland, 1998; Hardt and Bause, 2002). Indeed, results from thein vitro peptide binding assay and confocal study demonstratedthat the Lys in the –3 position of the KXD/E motif played a morecritical role in COPI binding and Golgi retention of EMP12,whereas the last acidic residue played a minor role, indicatingthat the Lys-based motif can also function in Golgi retention ofmultiple TMD proteins in plant cells.

Both p24 (Contreras et al., 2004b) and EMP (this study) areshown to interact with the COPI vesicle. However, p24 is re-cycled back to the ER (Langhans et al., 2008), whereas EMP istrapped in the Golgi, as demonstrated by both confocal immu-nofluorescence and immunogold EM in this study, in which bothendogenous EMP12 and the GFP-EMP12 fusion were found tolocalize mainly to themedial- and cis-Golgi cisternae. A previousEM tomography study (Donohoe et al., 2007) suggests the ex-istence of distinct COPI populations in plant cells: The COPIavesicles that bud exclusively from cis-Golgi cisternae may me-diate the recycling transport to the ER, whereas the COPIbvesicles that bud exclusively from medial- and trans-Golgi cis-ternae may mediate the retrograde transport of Golgi residentproteins (Donohoe et al., 2007). Since EMP12 proteins showmainlymedial- and cis-Golgi localization in this study, we are notsure if EMP12 would achieve its Golgi localization by interacting

with COPIb vesicles in Arabidopsis cells. It would therefore beinteresting to test the possible distinct functions of KKXX andKXD/E motifs in Golgi-ER recycling (for p24) or Golgi retention (forEMP12) by swapping the two COPI interaction motifs of p24 andEMP for subsequent dynamics and subcellular localization studies.

METHODS

Plasmid Construction

The full-length cDNA of EMP12 was amplified and cloned into the pBI121backbone for construction of the GFP fusion and generation of transgenicplants using floral dip (Clough andBent, 1998). All EMP12mutant constructsused for transient expression in protoplasts were PCR amplified and clonedinto the pBI221 backbone containing the cauliflower mosaic virus 35Spromoter, GFP coding sequence, and the nopaline synthase terminator(Miao et al., 2008). The SP sequence for making the SP-GFP-EMP12–linkedconstructs was derived from barley (Hordeum vulgare) proaleurain (Jiangand Rogers, 1998) or from the native signal sequence of EMP12 based ontheSignalP 3.0 Server prediction (http://www.cbs.dtu.dk/services/SignalP/).The protein topology was predicted using TMHMM Server 2.0 (http://www.cbs.dtu.dk/services/TMHMM/). The primers used to generate variousEMP12 constructs are listed in Supplemental Table 1 online. All constructswere confirmed by restriction mapping and DNA sequencing.

Plant Materials and Transient Expression in Protoplasts

Procedures for generating transgenic Arabidopsis thaliana plants ex-pressing XFP fusion proteins and maintaining Arabidopsis suspensioncultured cells were described previously (Clough and Bent, 1998; Jiangand Rogers, 1998; Tse et al., 2004; H. Wang et al., 2010; J. Wang et al.,2010). Transient expression using protoplasts derived from Arabidopsis sus-pension culture cells was essentially performed according to our establishedprotocol (Miao and Jiang, 2007; Lam et al., 2009; Wang and Jiang, 2011).

Immunofluorescence Study and Confocal Microscopy

Fixation and preparation of Arabidopsis roots for immunofluorescentlabeling and confocal analysis were performed as described (Sauer et al.,2006). Fixed roots were incubated with anti-EMP12 antibody at 4 mg/mLovernight at 4°C for immunolabeling and were then probed with Alexa 568goat anti-rabbit IgG (Invitrogen) secondary antibody for confocal obser-vation. Expression of the fluorescent fusion proteins andobservation of theirsubcellular localization in protoplasts were usually performed at 12 to 14 hafter transient expression. For each experiment or construct, more than 50individual cells were observed for confocal imaging that represented >80%of the cells showing similar expression levels andpatterns. Images collectedfrom 10 individual cells were used for the data process and analysis. Allconfocal images were captured using the Olympus FV1000 laser scanningconfocal system (http://www.olympusconfocal.com/). The GFP signal wasvisualized with excitation at 488 nm and emission at 500 to 550 nm (usinga band-pass filter). mRFP or Alexa 568 signals were visualized in anotherdetection channel with excitation at 559 nm and emission at 570 to 630 nm(band-pass filter). The line sequential scanning mode was always selectedin dual-channel observations to avoid possible crosstalk between twofluorophores. All images were prepared using Adobe Photoshop as de-scribed previously (Jiang and Rogers, 1998; Tse et al., 2004; H. Wang et al.,2010; J. Wang et al., 2010).

EM Study

The general procedures for transmission EM sample preparation, thinsectioning, and immunogold labeling were performed essentially as

2100 The Plant Cell

described previously (Tse et al., 2004; Lam et al., 2007b; Miao et al., 2011;Shen et al., 2011; Wang et al., 2011; Wang et al., 2012). Root tips of 7-d-old wild-type and transgenic GFP-EMP12 Arabidopsis plants were cutand immediately frozen in a high-pressure freezer (EM PACT2; Leica),followed by subsequent freeze substitution in dry acetone containing0.1% uranyl acetate at –85°C in an AFS freeze substitution unit (Leica).Infiltration with HM20, embedding, and UV polymerization were per-formed stepwise at –35°C. Immunogold labeling was performed with GFPor EMP12 antibodies at 40 mg/mL and gold-coupled secondary antibodyat a 1:50 dilution. Transmission EM examination was done with a HitachiH-7650 transmission electron microscope with a charge-coupled devicecamera (Hitachi High-Technologies) operating at 80 kV.

Antibodies, Protein Preparation, and Immunoblot Analysis

Synthetic peptide (GeneScript), CLRNDYAKYAREDDD, corresponding tothe cytosolic region between the TMD1 and TMD2 of EMP12 was con-jugated with keyhole limpet hemocyanin and used to immunize rabbits atthe Laboratory Animal Services Center of the Chinese University of HongKong. Antibodies were affinity purified using cyanogen bromide-activatedSepharose 4B column (Sigma-Aldrich) conjugated with the peptides. Mycantibody was purchased from Santa Cruz. Alexa 568 goat anti-rabbit IgGwas purchased from Invitrogen.

For the subcellular fractionation study, microsomal membrane fractionwas first isolated from the wild-type Arabidopsis suspension culture cellsvia fractionation on a 25 to 50% continuous Suc gradient according to thepublished procedure (Tse et al., 2004). These fractions were then probedwith anti-VSR, anti-Man1, and anti-EMP12 antibodies. To prepare cellextracts from protoplasts, transformed protoplasts were first dilutedthreefold with 250 mM NaCl and then harvested by centrifugation at 100gfor 10min, followed by resuspension in lysis buffer containing 25mMTris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM phenylmethylsulphonylfluoride, and 25 mg/mL leupeptin. The protoplasts were further lysed bypassing through a 1-mL syringe with needle and then spun at 600g for3min to remove intact cells and large cellular debris. The supernatant totalcell extracts were then centrifuged at 100,000g for 30 min at 4°C; thesupernatant and pellet were assigned as soluble andmembrane fractions,respectively. Proteins were separated by SDS-PAGE and analyzed byimmunoblotting. Quantification of the relative gray-scale intensity inimmunoblot analysis was done with ImageJ software v1.45 according tothe procedure (http://lukemiller.org/index.php/2010/11/analyzing-gels-and-western-blots-with-image-j/).

In Vitro Binding Assay of COPI and COPII Coat Proteins toSorting Motifs

In vitro pulldown and peptide binding assay were performed according topreviously published procedures (Contreras et al., 2004b; Suen et al., 2010).Synthetic peptides (GeneScript) corresponding to the C-terminal cytosolictail of EMP12 were coupled, via their N-terminal NH2 group, to cyanogenbromide-activated Sepharose 4B (3 mg peptides per mL of beads) ac-cording to the standard procedure. The coupling reactionwas quenched byincubation with 1 M Gly, pH 8.0, at room temperature for 2 h. Peptidecoupling efficiencywasmonitored bymeasuring the absorbance at 280 nm.

For the peptide binding assay, the soluble fractions, which wereextracted from the wild-type protoplasts, the protoplasts transfected withthe COPI coat-Myc fusion (Sec21-Myc), or the COPII coat-GFP fusion(Sec24-GFP), were diluted with 23 lysis buffer (50 mM Tris-HCl, pH 7.5,300 mM NaCl, 0.5% Triton X-100, 2 mM phenylmethylsulphonyl fluoride,and 50 mg/mL leupeptin) and were then incubated with Sepharose beadscoupled with the peptides for 4 h at 4°C in a top to end rotator. Afterincubation, the beads were washed five times with 13 lysis buffer andthen eluted by boiling in reducing SDS sample buffer. Proteins wereseparated by SDS-PAGE and subjected to liquid chromatography-MS/MS analysis or immunoblotting using appropriate antibodies.

Sample preparation and MS were performed as previously described(Wu et al., 2009). Briefly, the gel was silver stained after SDS-PAGE sep-aration, and the interesting protein bands were cut out for in-gel trypsindigestion. The peptides were extracted from the trypsin-digested gel forMSanalysis with amatrix-assisted laser desorption ionization-time of flight/timeof flight tandemmass spectrometer ABI 4700 proteomics analyzer (AppliedBiosystems). Mass data acquisitions were piloted by 4000 Series ExplorerSoftware v3.0. TheMS andMS/MS data were loaded into the GPS Explorersoftware v3.5 (Applied Biosystems) and searched against the NationalCenter for Biotechnology Information nonredundant database (released onMarch 5, 2010) with species restriction to Arabidopsis by Mascot searchengine v2.1 (Matrix Science).

Topology Analysis and Protease Protection Assay

Microsome isolation and protein topology analysis were performed asdescribed with some modifications (Abas and Luschnig, 2010; Cai et al.,2011). Basically, the protoplasts expressing GFP-EMP12 or GFP-VSR2were diluted with 250 mM NaCl and harvested by centrifugation. Theharvested protoplasts were then lysed in the extraction buffer (40 mMHEPES-KOH, pH 7.5, 1 mM EDTA, 10 mMKCl, and 0.4 M Suc) by passingthrough a 1-mL syringe with a needle several times. After centrifugation at600g for 3 min to remove intact cells and large cellular debris, the su-pernatants were further centrifuged at 100,000g for 30 min at 4°C. Thepellet was assigned as the microsomal-enriched membranes and wasresuspended in the extraction buffer to be divided into three parts withequal volume for protease protection assay with three distinct conditions:without trypsin, 100 mg/mL trypsin and 100 mg/mL trypsin with 1% TritonX-100. After incubation at 37°C for 30 min, the samples were pelleted forprotein extraction and immunoblot analysis using GFP antibody.

Accession Numbers

The Arabidopsis Genome Initiative locus identifiers for the genes men-tioned in this article are EMP2 (At5g25100), EMP3 (At2g24170), EMP8(At1g14670), EMP12 (At1g10950), Sec21 (At4g34450), and Sec24(At3g07100).

Supplemental Data

The following materials are available in the online version of this article.

Supplemental Figure 1. Plant EMP Proteins.

Supplemental Figure 2. The C Terminus of EMP12 Was Essential forER Export.

Supplemental Figure 3. Subcellular Localization of GFP-EMP12 withTriple Mutations at Its C Terminus.

Supplemental Figure 4. Subcellular Localization of EMP12-GFP withthe Golgi Retention Motif Fused to the C-Terminal End.

Supplemental Figure 5. Subcellular Localization of EMP12-GFP withC-Terminal Fusion of Mutated Golgi Retention Motif.

Supplemental Figure 6. EMP12-GFP Reached the Vacuole via thePVC for Degradation.

Supplemental Figure 7. EMP12 C Terminus Interacted with COPISubunits.

Supplemental Figure 8. Subcellular Localization of N-Terminal orC-Terminal GFP Fusions with EMP2, EMP3, and EMP8 in ArabidopsisProtoplasts.

Supplemental Figure 9. Localization Patterns of Singly ExpressedVarious EMP12 Fusions and Their Mutants in Arabidopsis Protoplasts.

Supplemental Table 1. Primers Used in This Study.

Targeting and Subcellular Localization of EMP12 2101

ACKNOWLEDGMENTS

We thank Akihiko Nakano and Takashi Ueda (University of Tokyo, Japan)for providing the cDNA encoding ARA7(Q69L). This work was supportedby grants from the Research Grants Council of Hong Kong (CUHK466309,CUHK466610, CUHK466011, CUHK2/CRF/11G, and HKBU1/CRF/10)and Chinese University of Hong Kong Schemes A/B/C to L.J.

AUTHOR CONTRIBUTIONS

L.J. supervised the project. C.G. and L.J. designed the experiments.C.G., C.K.Y.Y., S.Q., M.W.Y.S., K.Y.L., and S.W.L. performed specificexperiments and analyzed data. C.G. wrote the article draft, and L.J.revised the article.

Received January 19, 2012; revised March 12, 2012; accepted April 18,2012; published May 8, 2012.

REFERENCES

Abas, L., and Luschnig, C. (2010). Maximum yields of microsomal-type membranes from small amounts of plant material without re-quiring ultracentrifugation. Anal. Biochem. 401: 217–227.

Aguilar, P.S. et al. (2010). A plasma-membrane E-MAP reveals linksof the eisosome with sphingolipid metabolism and endosomaltrafficking. Nat. Struct. Mol. Biol. 17: 901–908.

Barlowe, C. (2003). Signals for COPII-dependent export from the ER:what’s the ticket out? Trends Cell Biol. 13: 295–300.

Barlowe, C., Orci, L., Yeung, T., Hosobuchi, M., Hamamoto, S.,Salama, N., Rexach, M.F., Ravazzola, M., Amherdt, M., andSchekman, R. (1994). COPII: A membrane coat formed by Secproteins that drive vesicle budding from the endoplasmic reticulum.Cell 77: 895–907.

Beck, R., Rawet, M., Wieland, F.T., and Cassel, D. (2009). The COPIsystem: Molecular mechanisms and function. FEBS Lett. 583:2701–2709. Erratum. FEBS Lett. 583: 3541.

Benghezal, M., Cornillon, S., Gebbie, L., Alibaud, L., Brückert, F.,Letourneur, F., and Cosson, P. (2003). Synergistic control of cellularadhesion by transmembrane 9 proteins. Mol. Biol. Cell 14: 2890–2899.

Bi, X.P., Corpina, R.A., and Goldberg, J. (2002). Structure of theSec23/24-Sar1 pre-budding complex of the COPII vesicle coat.Nature 419: 271–277.

Brandizzi, F., Frangne, N., Marc-Martin, S., Hawes, C., Neuhaus,J.M., and Paris, N. (2002). The destination for single-pass mem-brane proteins is influenced markedly by the length of the hydro-phobic domain. Plant Cell 14: 1077–1092.

Cai, Y., Jia, T., Lam, S.K., Ding, Y., Gao, C., San, M.W., Pimpl, P.,and Jiang, L. (2011). Multiple cytosolic and transmembrane de-terminants are required for the trafficking of SCAMP1 via an ER-Golgi-TGN-PM pathway. Plant J. 65: 882–896.

Cai, Y., Zhuang, X., Wang, J., Wang, H., Lam, S.K., Gao, C., Wang,X., and Jiang, L. (April 4, 2012). Vacuolar degradation of two in-tegral plasma membrane proteins, AtLRR84A and OsSCAMP1, iscargo ubiquitination-independent and prevacuolar compartment-mediated in plant cells. Traffic http://dx.doi.org.

Clough, S.J., and Bent, A.F. (1998). Floral dip: A simplified method forAgrobacterium-mediated transformation of Arabidopsis thaliana.Plant J. 16: 735–743.

Contreras, I., Ortiz-Zapater, E., and Aniento, F. (2004a). Sortingsignals in the cytosolic tail of membrane proteins involved in theinteraction with plant ARF1 and coatomer. Plant J. 38: 685–698.

Contreras, I., Yang, Y.D., Robinson, D.G., and Aniento, F. (2004b).Sorting signals in the cytosolic tail of plant p24 proteins involved in theinteraction with the COPII coat. Plant Cell Physiol. 45: 1779–1786.

Cornillon, S., Pech, E., Benghezal, M., Ravanel, K., Gaynor, E.,Letourneur, F., Brückert, F., and Cosson, P. (2000). Phg1p isa nine-transmembrane protein superfamily member involved indictyostelium adhesion and phagocytosis. J. Biol. Chem. 275:34287–34292.

Cosson, P., and Letourneur, F. (1994). Coatomer interaction with di-lysine endoplasmic reticulum retention motifs. Science 263: 1629–1631.

Cosson, P., and Letourneur, F. (1997). Coatomer (COPI)-coatedvesicles: Role in intracellular transport and protein sorting. Curr.Opin. Cell Biol. 9: 484–487.

Donohoe, B.S., Kang, B.H., and Staehelin, L.A. (2007). Identificationand characterization of COPIa- and COPIb-type vesicle classesassociated with plant and algal Golgi. Proc. Natl. Acad. Sci. USA104: 163–168.

Dunkel, M., Latz, A., Schumacher, K., Müller, T., Becker, D., andHedrich, R. (2008). Targeting of vacuolar membrane localizedmembers of the TPK channel family. Mol. Plant 1: 938–949.

Dunkley, T.P. et al. (2006). Mapping the Arabidopsis organelle pro-teome. Proc. Natl. Acad. Sci. USA 103: 6518–6523.

Foresti, O., Gershlick, D.C., Bottanelli, F., Hummel, E., Hawes, C.,and Denecke, J. (2010). A recycling-defective vacuolar sortingreceptor reveals an intermediate compartment situated betweenprevacuoles and vacuoles in tobacco. Plant Cell 22: 3992–4008.

Froquet, R., Cherix, N., Birke, R., Benghezal, M., Cameroni, E.,Letourneur, F., Mösch, H.U., De Virgilio, C., and Cosson, P.(2008). Control of cellular physiology by TM9 proteins in yeast andDictyostelium. J. Biol. Chem. 283: 6764–6772.

Geldner, N., Dénervaud-Tendon, V., Hyman, D.L., Mayer, U.,Stierhof, Y.D., and Chory, J. (2009). Rapid, combinatorial analysisof membrane compartments in intact plants with a multicolormarker set. Plant J. 59: 169–178.

Gobert, A., Isayenkov, S., Voelker, C., Czempinski, K., andMaathuis, F.J.M. (2007). The two-pore channel TPK1 gene en-codes the vacuolar K+ conductance and plays a role in K+ ho-meostasis. Proc. Natl. Acad. Sci. USA 104: 10726–10731.

Gomez, M., Scales, S.J., Kreis, T.E., and Perez, F. (2000). Mem-brane recruitment of coatomer and binding to dilysine signals areseparate events. J. Biol. Chem. 275: 29162–29169.

Hanton, S.L., Matheson, L.A., Chatre, L., and Brandizzi, F. (2009).Dynamic organization of COPII coat proteins at endoplasmic re-ticulum export sites in plant cells. Plant J. 57: 963–974.

Hanton, S.L., Renna, L., Bortolotti, L.E., Chatre, L., Stefano, G., andBrandizzi, F. (2005). Diacidic motifs influence the export of trans-membrane proteins from the endoplasmic reticulum in plant cells.Plant Cell 17: 3081–3093.

Hardt, B., and Bause, E. (2002). Lysine can be replaced by histidinebut not by arginine as the ER retrieval motif for type I membraneproteins. Biochem. Biophys. Res. Commun. 291: 751–757.

Harter, C., and Wieland, F.T. (1998). A single binding site for dilysineretrieval motifs and p23 within the gamma subunit of coatomer.Proc. Natl. Acad. Sci. USA 95: 11649–11654.

Höfte, H., and Chrispeels, M.J. (1992). Protein sorting to the vacuolarmembrane. Plant Cell 4: 995–1004.

Huh, W.K., Falvo, J.V., Gerke, L.C., Carroll, A.S., Howson, R.W.,Weissman, J.S., and O’Shea, E.K. (2003). Global analysis of pro-tein localization in budding yeast. Nature 425: 686–691.

Isayenkov, S., Isner, J.C., and Maathuis, F.J. (2011). Rice two-poreK+ channels are expressed in different types of vacuoles. Plant Cell23: 756–768.

2102 The Plant Cell

Jiang, L., and Rogers, J.C. (1998). Integral membrane protein sortingto vacuoles in plant cells: Evidence for two pathways. J. Cell Biol.143: 1183–1199.

Kleine-Vehn, J., Leitner, J., Zwiewka, M., Sauer, M., Abas, L.,Luschnig, C., and Friml, J. (2008). Differential degradation of PIN2auxin efflux carrier by retromer-dependent vacuolar targeting. Proc.Natl. Acad. Sci. USA 105: 17812–17817.

Kotzer, A.M., Brandizzi, F., Neumann, U., Paris, N., Moore, I., andHawes, C. (2004). AtRabF2b (Ara7) acts on the vacuolar traffickingpathway in tobacco leaf epidermal cells. J. Cell Sci. 117: 6377–6389.

Kuehn, M.J., Herrmann, J.M., and Schekman, R. (1998). COPII-cargo interactions direct protein sorting into ER-derived transportvesicles. Nature 391: 187–190.

Lam, S.K., Cai, Y., Hillmer, S., Robinson, D.G., and Jiang, L. (2008).SCAMPs highlight the developing cell plate during cytokinesis intobacco BY-2 cells. Plant Physiol. 147: 1637–1645.

Lam, S.K., Cai, Y., Tse, Y.C., Wang, J., Law, A.H., Pimpl, P., Chan,H.Y., Xia, J., and Jiang, L. (2009). BFA-induced compartmentsfrom the Golgi apparatus and trans-Golgi network/early endosomeare distinct in plant cells. Plant J. 60: 865–881.

Lam, S.K., Siu, C.L., Hillmer, S., Jang, S., An, G., Robinson, D.G.,and Jiang, L. (2007b). Rice SCAMP1 defines clathrin-coated, trans-Golgi-located tubular-vesicular structures as an early endosome intobacco BY-2 cells. Plant Cell 19: 296–319.

Lam, S.K., Tse, Y.C., Robinson, D.G., and Jiang, L. (2007a).Tracking down the elusive early endosome. Trends Plant Sci. 12:497–505.

Langhans, M., Marcote, M.J., Pimpl, P., Virgili-López, G.,Robinson, D.G., and Aniento, F. (2008). In vivo trafficking and lo-calization of p24 proteins in plant cells. Traffic 9: 770–785.

Latz, A., Becker, D., Hekman, M., Müller, T., Beyhl, D., Marten, I., Eing,C., Fischer, A., Dunkel, M., Bertl, A., Rapp, U.R., and Hedrich, R.(2007). TPK1, a Ca(2+)-regulated Arabidopsis vacuole two-pore K(+)channel is activated by 14-3-3 proteins. Plant J. 52: 449–459.

Law, A.H., Chow, C.M., and Jiang, L. (2012). Secretory carriermembrane proteins. Protoplasma 249: 269–283.

Li, Y.B., Rogers, S.W., Tse, Y.C., Lo, S.W., Sun, S.S., Jauh, G.Y.,and Jiang, L. (2002). BP-80 and homologs are concentrated onpost-Golgi, probable lytic prevacuolar compartments. Plant CellPhysiol. 43: 726–742.

Lozupone, F., Perdicchio, M., Brambilla, D., Borghi, M., Meschini,S., Barca, S., Marino, M.L., Logozzi, M., Federici, C., Iessi, E., deMilito, A., and Fais, S. (2009). The human homologue of Dictyo-stelium discoideum phg1A is expressed by human metastaticmelanoma cells. EMBO Rep. 10: 1348–1354.

Miao, Y., and Jiang, L. (2007). Transient expression of fluorescentfusion proteins in protoplasts of suspension cultured cells. Nat.Protoc. 2: 2348–2353.

Miao, Y., Li, K.Y., Li, H.Y., Yao, X., and Jiang, L. (2008). The vacuolartransport of aleurain-GFP and 2S albumin-GFP fusions is mediatedby the same pre-vacuolar compartments in tobacco BY-2 andArabidopsis suspension cultured cells. Plant J. 56: 824–839.

Miao, Y., Yan, P.K., Kim, H., Hwang, I., and Jiang, L. (2006). Localizationof green fluorescent protein fusions with the seven Arabidopsis vacuolarsorting receptors to prevacuolar compartments in tobacco BY-2 cells.Plant Physiol. 142: 945–962.

Miao, Y.S., Li, H.Y., Shen, J.B., Wang, J.Q., and Jiang, L.W. (2011).QUASIMODO 3 (QUA3) is a putative homogalacturonan methyl-transferase regulating cell wall biosynthesis in Arabidopsis sus-pension-cultured cells. J. Exp. Bot. 62: 5063–5078.

Mikosch, M., Hurst, A.C., Hertel, B., and Homann, U. (2006). Di-acidic motif is required for efficient transport of the K+ channelKAT1 to the plasma membrane. Plant Physiol. 142: 923–930.

Movafeghi, A., Happel, N., Pimpl, P., Tai, G.H., and Robinson, D.G.(1999). Arabidopsis Sec21p and Sec23p homologs. Probable coatproteins of plant COP-coated vesicles. Plant Physiol. 119: 1437–1446.

Nishimura, N., and Balch, W.E. (1997). A di-acidic signal required forselective export from the endoplasmic reticulum. Science 277:556–558.

Nufer, O., Guldbrandsen, S., Degen, M., Kappeler, F., Paccaud,J.P., Tani, K., and Hauri, H.P. (2002). Role of cytoplasmic C-terminalamino acids of membrane proteins in ER export. J. Cell Sci. 115:619–628.

Paris, N., Rogers, S.W., Jiang, L.W., Kirsch, T., Beevers, L.,Phillips, T.E., and Rogers, J.C. (1997). Molecular cloning and fur-ther characterization of a probable plant vacuolar sorting receptor.Plant Physiol. 115: 29–39.

Pimpl, P., Hanton, S.L., Taylor, J.P., Pinto-daSilva, L.L., andDenecke, J. (2003). The GTPase ARF1p controls the sequence-specific vacuolar sorting route to the lytic vacuole. Plant Cell 15:1242–1256.

Rabouille, C., and Klumperman, J. (2005). Opinion: The maturingrole of COPI vesicles in intra-Golgi transport. Nat. Rev. Mol. CellBiol. 6: 812–817.

Robinson, D.G., Jiang, L., and Schumacher, K. (2008). The endo-somal system of plants: Charting new and familiar territories. PlantPhysiol. 147: 1482–1492.

Saint-Jore-Dupas, C., Nebenführ, A., Boulaflous, A., Follet-Gueye,M.L., Plasson, C., Hawes, C., Driouich, A., Faye, L., and Gomord,V. (2006). Plant N-glycan processing enzymes employ differenttargeting mechanisms for their spatial arrangement along the se-cretory pathway. Plant Cell 18: 3182–3200.

Sato, K., and Nakano, A. (2002). Emp47p and its close homologEmp46p have a tyrosine-containing endoplasmic reticulum exitsignal and function in glycoprotein secretion in Saccharomycescerevisiae. Mol. Biol. Cell 13: 2518–2532.

Sauer, M., Paciorek, T., Benková, E., and Friml, J. (2006). Immu-nocytochemical techniques for whole-mount in situ protein locali-zation in plants. Nat. Protoc. 1: 98–103.

Schimmöller, F., Díaz, E., Mühlbauer, B., and Pfeffer, S.R. (1998).Characterization of a 76 kDa endosomal, multispanning membraneprotein that is highly conserved throughout evolution. Gene 216:311–318.

Schmitz, K.R., Liu, J., Li, S., Setty, T.G., Wood, C.S., Burd, C.G.,and Ferguson, K.M. (2008). Golgi localization of glycosyl-transferases requires a Vps74p oligomer. Dev. Cell 14: 523–534.

Schoberer, J., Vavra, U., Stadlmann, J., Hawes, C., Mach, L.,Steinkellner, H., and Strasser, R. (2009). Arginine/lysine residuesin the cytoplasmic tail promote ER export of plant glycosylationenzymes. Traffic 10: 101–115.

Schröder, S., Schimmöller, F., Singer-Krüger, B., and Riezman, H.(1995). The Golgi-localization of yeast Emp47p depends on its di-lysine motif but is not affected by the ret1-1 mutation in alpha-COP.J. Cell Biol. 131: 895–912.

Shen, Y., Wang, J., Ding, Y., Lo, S.W., Gouzerh, G., Neuhaus, J.M.,and Jiang, L. (2011). The rice RMR1 associates with a distinctprevacuolar compartment for the protein storage vacuole pathway.Mol. Plant 4: 854–868.

Suen, P.K., Shen, J.B., Sun, S.S.M., and Jiang, L.W. (2010). Ex-pression and characterization of two functional vacuolar sortingreceptor (VSR) proteins, BP-80 and AtVSR4 from culture media oftransgenic tobacco BY-2 cells. Plant Sci. 179: 68–76.

Tse, Y.C., Lo, S.W., Hillmer, S., Dupree, P., and Jiang, L. (2006).Dynamic response of prevacuolar compartments to brefeldin a inplant cells. Plant Physiol. 142: 1442–1459.

Targeting and Subcellular Localization of EMP12 2103

Tse, Y.C., Mo, B., Hillmer, S., Zhao, M., Lo, S.W., Robinson, D.G.,and Jiang, L. (2004). Identification of multivesicular bodies asprevacuolar compartments in Nicotiana tabacum BY-2 cells. PlantCell 16: 672–693.

Tu, L., Tai, W.C.S., Chen, L., and Banfield, D.K. (2008). Signal-mediateddynamic retention of glycosyltransferases in the Golgi. Science 321:404–407.

Wang, H., and Jiang, L. (2011). Transient expression and analysis offluorescent reporter proteins in plant pollen tubes. Nat. Protoc. 6:419–426.

Wang, H., Tse, Y.C., Law, A.H.Y., Sun, S.S.M., Sun, Y.B., Xu, Z.F.,Hillmer, S., Robinson, D.G., and Jiang, L.W. (2010). Vacuolar sortingreceptors (VSRs) and secretory carrier membrane proteins (SCAMPs)are essential for pollen tube growth. Plant J. 61: 826–838.

Wang, H., Zhuang, X.H., Hillmer, S., Robinson, D.G., and Jiang, L.W.(2011). Vacuolar sorting receptor (VSR) proteins reach the plasmamembrane in germinating pollen tubes. Mol. Plant 4: 845–853.

Wang, J., Ding, Y., Wang, J., Hillmer, S., Miao, Y., Lo, S.W., Wang,X., Robinson, D.G., and Jiang, L. (2010). EXPO, an exocyst-

positive organelle distinct from multivesicular endosomes andautophagosomes, mediates cytosol to cell wall exocytosis inArabidopsis and tobacco cells. Plant Cell 22: 4009–4030.

Wang, J., Tse, Y.C., Hinz, G., Robinson, D.G., and Jiang, L. (2012).Storage globulins pass through the Golgi apparatus and multi-vesicular bodies in the absence of dense vesicle formation duringearly stages of cotyledon development in mung bean. J. Exp. Bot.63: 1367–1380.

Wu, T., Yuan, T.Z., Tsai, S.N., Wang, C.M., Sun, S.M., Lam, H.M.,and Ngai, S.M. (2009). Mass spectrometry analysis of the variantsof histone H3 and H4 of soybean and their post-translationalmodifications. BMC Plant Biol. 9: 98.

Yuasa, K., Toyooka, K., Fukuda, H., and Matsuoka, K. (2005).Membrane-anchored prolyl hydroxylase with an export signal fromthe endoplasmic reticulum. Plant J. 41: 81–94.

Zelazny, E., Miecielica, U., Borst, J.W., Hemminga, M.A., andChaumont, F. (2009). An N-terminal diacidic motif is required forthe trafficking of maize aquaporins ZmPIP2;4 and ZmPIP2;5 to theplasma membrane. Plant J. 57: 346–355.

2104 The Plant Cell

DOI 10.1105/tpc.112.096057; originally published online May 8, 2012; 2012;24;2086-2104Plant Cell

JiangCaiji Gao, Christine K.Y. Yu, Song Qu, Melody Wan Yan San, Kwun Yee Li, Sze Wan Lo and Liwen

Reticulum Export and Golgi Retention Signals at Its C Terminus Endomembrane Protein12 Contains Both EndoplasmicArabidopsisThe Golgi-Localized

 This information is current as of January 12, 2021

 

Supplemental Data /content/suppl/2012/04/24/tpc.112.096057.DC1.html

References /content/24/5/2086.full.html#ref-list-1

This article cites 76 articles, 34 of which can be accessed free at:

Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists