![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 18, Issue 12, 4872-4884, December 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
*Departamento de Inmunología Clínica y Reumatología, Facultad de Medicina, and Centro de Regulación Celular y Patología, Facultad de Ciencias Biológicas, Pontificia Universidad Católica de Chile, 6510260 Santiago, Chile; Millennium Institute for Fundamental and Applied Biology, 7780344 Santiago, Chile; Institute of Biochemistry Faculty of Medicine, Humboldt University, 10117 Berlin, Germany; and Dyson Vision Research Institute, Weill Medical College of Cornell University, New York, NY 10021
Submitted June 13, 2007;
Accepted September 11, 2007
Monitoring Editor: Keith Mostov
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
45 min. Our experiments conclusively demonstrate that 1) AP1B functions exclusively at RE; 2) TGN-to-RE transport is very fast and selective and is mediated by adaptors different from AP1B; and 3) the TGN and AP1B-containing RE cooperate in biosynthetic basolateral sorting. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Mostov et al., 2003; Rodriguez-Boulan et al., 2005). Early studies demonstrated that this polarity was achieved by sorting of newly synthesized proteins at the level of the trans-Golgi network (TGN; Rindler et al., 1984; Fuller et al., 1985; Griffiths and Simons, 1986) and of endocytosed proteins at the level of recycling endosomes (RE; Matter and Mellman, 1994; Mostov and Cardone, 1995; Odorizzi and Trowbridge, 1997). Trafficking in both biosynthetic and recycling routes is controlled by apical sorting signals (e.g., N- and O-glycans, lipid anchors, and protein domains with affinity for lipid rafts), by microtubule motor determinants (Rodriguez-Boulan and Gonzalez, 1999; Schuck and Simons, 2004; Rodriguez-Boulan et al., 2005) and by basolateral sorting signals (e.g., tyrosine, dileucine, and monoleucine motifs, some similar to endocytic determinants, and noncanonic motifs not yet matching any consensus sequence; Matter et al., 1994; Rodriguez-Boulan et al., 2005). Tyrosine-based signals are recognized by a family of organelle-specific tetrameric AP (adaptor protein) adaptors via their µ subunit, whereas dileucine motifs may be recognized via large (
and 
) and small (
1, 3) subunits of AP adaptors acting together (Bonifacino and Lippincott-Schwartz, 2003; Janvier et al., 2003).
The early paradigm of two separate sorting sites for proteins in biosynthetic or recycling pathways has, however, progressively shifted over the past decade to a different paradigm in which both TGN and RE share a sorting role in the biosynthetic route. This new paradigm emerged from the observations by several laboratories showing that newly synthesized PM proteins leaving the Golgi apparatus may traverse the endosomal compartment before arrival at the cell surface (Futter et al., 1995; Leitinger et al., 1995; Brachet et al., 1999; Orzech et al., 2000; Ang et al., 2003; Lock and Stow, 2005). These early observations, however, did not elucidate the magnitude of the transendosomal route nor the individual contribution of TGN or RE to the biosynthetic sorting of individual proteins. The molecular sorting machinery operating in each compartment is also poorly understood.
Some insight into these important questions was provided by the discovery of the basolateral sorting adaptors AP1B (Ohno et al., 1999) and AP4 (Simmen et al., 2002). The role of AP4 in basolateral sorting remains poorly understood. By contrast, considerable advances have been made in our understanding of AP1B function (Folsch, 2005; Ellis et al., 2006). AP1B is a variety of AP1 that displays an epithelial-specific subunit, µ1B, instead of the more ubiquitous µ1A subunit found in AP1A (Ohno et al., 1999). Reconstitution of µ1B into LLC-PK1 epithelial cells, which lack this protein, showed that AP1B is required for the correct sorting of a group of basolateral proteins (Folsch et al., 1999; Gan et al., 2002; Sugimoto et al., 2002; Folsch et al., 2003). Early electron microscopy (EM) studies suggested that AP1B sorts basolateral proteins in the TGN (Folsch et al., 2001). However, more detailed functional and immunofluorescence experiments showed that AP1B sorts transferrin receptor (TfR) and LDL receptor (LDLR) postendocytically in LLC-PK1 cells, probably at the level of RE (Gan et al., 2002; Folsch et al., 2003).
More recently, however, µ1B knockdown experiments in Madin-Darby canine kidney (MDCK) cells demonstrated for the first time that AP1B sorts two basolateral proteins, VSVG and TfR, in the biosynthetic route (Gravotta et al., 2007). Interestingly, for TfR this was only true in recently polarized (1–2 d confluent) MDCK cells, as in 4–5 d confluent monolayers TfR followed an AP1B-independent biosynthetic route, suggesting that as monolayers develop their polarity, they establish a direct route from TGN to PM (Gravotta et al., 2007). Cell fractionation data demonstrated enrichment of endogenous AP1B in endosomal fractions in both 1- and 4.5-d polarized monolayers, suggesting that basolateral sorting in the biosynthetic route might be carried out in RE. However, the experiments did not quantitatively demonstrate the transport of basolateral proteins between TGN and RE or directly show that the sorting role of AP1B took place at RE. Similarly, another report (Ang et al., 2004) demonstrated that ablation of RE with horseradish peroxidase (HRP)-peroxidase blocked transport of VSVG in the biosynthetic route, but did not demonstrate that under those conditions a significant mass of VSVG had left the TGN and quantitatively moved to RE, and more precisely, to RE containing AP1B.
Finally, a recent study (Fields et al., 2007) reported the interaction of various basolateral proteins with different AP adaptors by a yeast two-hybrid approach, but the data obtained provide no information on specific routes in which these adaptors perform their function or on the role of RE in biosynthetic transport. Thus, the role of RE in biosynthetic transport and the sorting role of AP1B in RE is currently supported only by circumstantial evidence.
Here, we utilized a function-blocking antibody against µ1B in Fischer rat epithelial (FRT) cells, a model epithelial cell line with similar sorting characteristics as MDCK cells (Zurzolo et al., 1993), and quantitative live imaging to study these important questions in detail. Because the experiments were carried out in recently confluent FRT cells, to facilitate live imaging, our data can be compared with the results obtained by Gravotta et al. (2007) in recently (1–2 d) confluent MDCK cells. Our experiments demonstrate that two basolateral proteins, VSVG and TfR, move quantitatively with very fast kinetics from TGN to RE (t1/2 3–5 min) and are blocked by AP1B antibody at the level of RE in the biosynthetic route. By contrast, LDLR moves from TGN to RE with slow kinetics (45 min), reflecting that it traffics first from PM to the basolateral surface, before being stopped by the AP1B antibody at the level of RE. Our data demonstrate directly and quantitatively, for the first time, that some newly synthesized basolateral proteins move obligatorily through AP1B-containing RE and that AP1B performs its basolateral sorting role in RE.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
95/295; Benmerah et al., 1999) provided by A. Benmerah (Institut Cochin, U567 INSERM/UMR8104 CNRS, Paris France), protein kinase D2 (pME-Py-GST-PKD2-KD; Yeaman et al., 2004) provided by V. Malhotra (University of California, San Diego). The VSVG3-GFP plasmid (Toomre et al., 1999) encoding an apical version of tsO4-VSVG-GFP (Ang et al., 2004; Yeaman et al., 2004) was provided by K. Simons (Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany). The vector for Rab11-GFP was a gift of M. Zerial (Max Planck Institute of Molecular Cell Biology and Genetics).
Antibodies
µ1B- and µ1A/B-specific antibodies were produced by immunizing rabbits with synthetic peptides conjugated with mollusk Concholepas concholepas hemocyanin (Blue Carrier, Biosonda Biotechnology, Santiago, Chile). The peptides correspond to N-terminal 44-56 residues (K-GALAPLLSHGQVH) and C-terminal 366-378 residues (CK-EKEEVEGRPPIGV) of human µ1B (Folsch et al., 1999) and 261-271 residues (CK-VLFDNTGRGKS) of µ1A. These antibodies were affinity-purified. Immunoglobulin G (IgG) from preimmune sera (PI-IgG) was purified with Econo-Pac Serum IgG Purification Kit (Bio-Rad, Richmond, CA). Secondary antibodies: Alexa488/555-tagged goat-anti-mouse and goat-anti-rabbit (Molecular Probes, Eugene, OR). Monoclonal antibodies were as follows: anti-TGN38 and anti-EEA-1 (Affinity BioReagents, Golden, CO), anti--adaptin clone 100/3 (Sigma, St. Louis, MO), anti--adaptin (BD Transduction Laboratories, Lexington, KY), anti-TfR (Zymed, South San Francisco, CA) and anti-ZO-1 (Chemicon, Temecula, CA). Polyclonal antibodies were as follows: anti-Rab11 antibody (Affinity BioReagents), HRP-conjugated secondary anti-rabbit (Chemicon), and anti-mouse antibodies (Rockland, Gilbertville, PA).
Glutathione S-Transferase-Fusion Protein Purification
Recombinant human µ1A or µ1B glutathione S-transferase (GST)-fusion protein were produced in transformed Escherichia coli cells induced with 0.2 mM isopropyl-1-thio-β-d-galactopyranoside (Invitrogen, Carlsbad, CA) for 3 h. GST-fusion proteins were purified by affinity chromatography on glutathione-Sepharose as described by the manufacturer.
Cell Culture and Microinjection
LLC-PK1 and LLC-PK1-µ1B cells (Gan et al., 2002) were maintained in DMEM (Invitrogen-BRL) supplemented with 5% FBS (Invitrogen), L-glutamine and nonessential amino acids (GeminiBio-Products, Woodland, CA). FRT cells were maintained in F-12 Coon's modification (Sigma) supplemented with 10% fetal bovine serum (FBS). The cells were plated at subclonfluent levels on glass coverslips and microinjected 1 d after reaching confluence. The expression plasmids (25 µg·ml–1) and antibodies (40 µg·ml–1), were prepared in HKCl microinjection buffer (Kreitzer et al., 2000; 10 mM HEPES, 140 mM potassium chloride, pH 7.4). The mix was centrifuged at 14,000 rpm in an Eppendorf-refrigerated centrifuge (Fremont, CA) for 15 min at 4°C. Microinjections were made in the nucleus (plasmids) and cytosol (antibodies) using back-loaded glass capillaries and an Eppendorf Transjector 5246 system mounted on a Zeiss Axiovert S100 inverted microscope (Thornwood, NY), keeping the cells in bicarbonate-free DMEM supplemented with 5% FBS, 20 mM HEPES. After incubating the cells for 1 h at 37°C, the medium was changed to serum-free medium supplemented with 100 µg·ml–1 cycloheximide for the rest of the experiment. Control experiments for µ1Bn antibody were made by preincubating the antibody for 30 min at 37°C in HKM buffer (25 mM HEPES-KOH, pH 7.6; 50 mM KCl; 5 mM MgCl2) with either the N-terminal (µ1Bnp) or C-terminal (µ1Bcp) peptides.
Indirect Immunofluorescence
Indirect immunofluorescence was performed as described (Soza et al., 2004). Briefly, cells grown on coverslips were washed in phosphate-buffered saline (PBS) supplemented with 0.1 mM CaCl2 and 1 mM MgCl2 (PBS-CM) and fixed in freshly prepared PBS-CM 4% paraformaldehyde (Merck, Rahway, NJ) for 30 min. Cells were permeabilized in PBS-CM with 0.1% Triton X-100 for 10 min and blocked with PBS-CM 0.2% gelatin (Sigma type Bloom G2500, PBS-CM-G) for 10 min at room temperature. Incubation with specific primary antibodies against µ1B (1:200), TGN38 (1:250), TfR (1:500), Early Endosomal Antigen-1 (EEA-1) (1:50), -adaptin (1:100), GST (1:500), and Rab-11 (1:50) was done during 30 min a 37°C in PBS-CM-G buffer. Cells were subsequently labeled with appropriate fluorescent-conjugated secondary antibodies Alexa 488/555-tagged goat-anti-mouse and goat-anti-rabbit (1:2500). Experiments with µ1Bn antibody blocked with either µ1Bnp or µ1Bcp peptides included 1% DMSO in the incubation buffer.
Permanent Silencing of µ1B in FRT Cells
Stable knockdown clones were generated by transfecting FRT cells with a retroviral vector carrying a µ1B-targeting sequence. A 60-mer sense (5' GATCCCCGAACGAGGTCTTCATTGATTTCAAGAGAATCAATGAAGACCTCGTTCTTTTTC-3') and antisense (5'-TCGAGAAAAAGAACGAGGTCTTCATTGATTCTCTTGAAATCAATGAAGACCTCGTTCGGG-3'). Oligonucleotide encoding the 19-m34 µ1B-targeted sequence with a BglII/XhoI restriction sites were annealed and cloned into a pSuper.retro.puro vector (OligoEngine, Seattle, WA) following the manufacture's protocol. Cells were selected for puromycin resistance and characterized by Western blotting.
TGN Block and Time-Lapse Fluorescence Microscopy
Microinjected cells expressing GFP-tagged proteins for 1 h at 37°C were cooled at 4°C and then placed in a 20°C incubator for 2 h in the presence of 100 µg·ml–1 cycloheximide. Cooling was necessary to achieve a better block of the proteins at the TGN with minimal spill into the µ1B compartment. After the 20°C block, the cells were cooled again on ice before reestablishing the transport at 37°C or 32°C for VSVG-GFP. This was done in the absence of FBS to improve the detection of the proteins at the cell surface. The cells were either processed for immunofluorescence or mounted on a POC-R thermal-controlled chamber (Zeiss) for time-lapse imaging in vivo. Images were collected at 10-min intervals.
Image Acquisition and Analysis
Digital fluorescence images were acquired on a Zeiss Axiophot microscope with a Plan-APOCHROMAT 63x/1.4 oil immersion objective and the 14-bit Zeiss Axiocam camera, transferred to a computer workstation running Axiovision imaging software (Zeiss), processed, and analyzed with MetaMorph imaging software (Universal Imaging, West Chester, PA). For quantification, all images from a single experiment were acquired under identical settings (14 bits; 1300 x 1030 pixels, and the same exposure times, avoiding signal saturation). Their integrated fluorescence intensities (equivalent to the sum of all grayscale values for every pixel in the region) were analyzed after 2D deconvolution and threshold adjustment to select the perinuclear region. The percentage of colocalization, measured as integrated pixel intensity in the regions of overlap of LDLR-GFP, VSVG-GFP¤ and TfR-GFP with either µ1B or TGN38 (detected by immunofluorescence) was calculated for each individual cell (n = 20–30). Laser scanning microscopy was used to assess the polarity of FRT cells under the same conditions of the microinjection experiments. Images were acquired on a LSM 5 Zeiss microscope using 488/543 laser and a 63x/1.4 oil immersion objective.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2001, 2003; Gan et al., 2002). None of the anti-µ1B subunit antibodies generated to date (Folsch et al., 1999; Gravotta et al., 2007) have turned out to be suitable for immunofluorescence. We raised rabbit antibodies against described N- (µ1Bn) and C-terminal (µ1Bc) antigenic peptides of human µ1B (Folsch et al., 1999) and a peptide of µ1A (Gravotta et al., 2007; Figure 1A). Both µ1B-specific antibodies (µ1Bn-Ab and µ1Bc-Ab) distinguished in immunoblots human µ1B from the 80% identical µ1A (Ohno et al., 1999) produced in E. coli (Figure 1B). Surprisingly, µ1Bn-Ab turned out to be suitable for immunofluorescence. This antibody decorated human µ1B expressed by microinjection of its cDNA in LLC-PK1 cells, whereas nonmicroinjected cells showed no staining (Figure 1C). µ1Bn-Ab decorated as well endogenous µ1B expressed by polarized FRT cells (Figure 1, D and F), but not endogenous µ1B expressed by MDCK cells (data not shown). In FRT cells, µ1Bn-Ab decorated mainly a perinuclear compartment (Figure 1D). An FRT cell line in which µ1B was permanently knocked down by RNAi (µ1B-KD) exhibited decreased levels of µ1B, as detected by immunofluorescence with µ1Bn-Ab and by immunoblot with µ1Bc-Ab (Figure 1, D and E). The immunofluorescence staining was specifically competed by the µ1Bn peptide but not by the µ1Bc peptide (Figure 1F). These observations indicate that the µ1Bn-Ab specifically recognizes endogenous µ1B in FRT cells. It seems unlikely that the lack of reaction with canine µ1B reflects a difference in the sequence between canine and rat/human µ1B peptide used as immunogen (Figure 1G). Rather, it must reflect a different organization of the AP1B adaptor in different cells (see Discussion).
|
; Folsch et al., 2003). In agreement with these observations, the perinuclear compartment of FRT cells that contained endogenous µ1B was enriched in the recycling endosomal marker TfR and was depleted of the TGN marker TGN38 (Figure 2A, fluorescence panels). Quantitative image analysis demonstrated 72.3 ± 1.9% colocalization between µ1B and TfR and 23.7 ± 0.9% colocalization between µ1B and TGN38 (Figure 2A, graph). Consistently, -adaptin, a subunit shared by both AP1B and AP1A adaptors, showed an intermediate extent of colocalization with µ1B (35%), consistent with the fact that a substantial amount of this protein localizes to TGN as a component of AP1A (Figure 2A, graph). EEA1, a marker of sorting endosomes (the earliest of the early endosomes) colocalized very poorly with µ1B (Figure 2A). Rab11, which in polarized MDCK cells localizes to apical recycling endosomes (ARE) rather than to Tf-rich common RE (Brown et al., 2000), did not colocalize either with TfR in FRT cells (Figure 2A). Although our rabbit µ1B antibody precluded colocalization studies with the polyclonal antibody to Rab11, we observed that Rab11-GFP expressed by microinjection did not colocalize with µ1B (not shown), thus confirming that the µ1B compartment is distinct from ARE and providing additional evidence for its identification as the equivalent of common RE in MDCK cells (Brown et al., 2000). The association of AP1A with TGN requires ARF GTPase activity (Stamnes and Rothman, 1993; Traub et al., 1993). In FRT cells, the perinuclear staining of both µ1B and -adaptin vanished upon BFA treatment, indicating that recruitment of AP1B to RE also depends on an ARF-like GTPase activity (Figure 2B).
|
reported that protein kinase D (PKD) is involved in the generation of basolateral transport vesicles from the TGN; expression of dominant-negative PKD-2 in MDCK cells leads to TGN tubulation and apical missorting of model basolateral markers (Yeaman et al., 2004). Expression of the PKD-2-KD protein by microinjection of its cDNA into FRT cells and subsequent incubation for 1 h resulted in the formation of tubules rich in TGN38 without affecting the µ1B/TfR compartment (Figure 3). PKD-2-KD colocalized with TGN-derived tubules rich in TGN38 but not with µ1B or TfR. These results further strengthen the notion that µ1B resides in a TfR-rich compartment distinct from the TGN, both spatially and functionally.
|
We have reported that AP1B participates in postendocytic basolateral recycling of TfR and LDLR in LLC-PK1 cells (Gan et al., 2002); hence, we first used the µ1Bn-Ab to interfere with the recycling of these receptors in FRT cells. To facilitate surface detection of TfR returning from RE, we coinjected the antibody together with an expression plasmid encoding a dominant negative mutant of Eps15 protein, Eps15-EH29-GFP, that inhibits the endocytosis of TfR (Benmerah et al., 1998). Expression of Eps15-EH29-GFP increased the percentage of cells displaying detectable TfR staining at the cell surface, from <10% to nearly 80% (Figure 4), as expected for PM accumulation of receptors returning from RE. Under these conditions, comicroinjection of the µ1Bn-Ab, but not of the corresponding PI-IgG fraction, inhibited the appearance of TfR at the cell surface. The specificity of this effect was further demonstrated by its abrogation with the µ1B N-terminal peptide (µ1Bnp) but not by the µ1Bc peptide (µ1Bcp; Figure 4). The simplest interpretation of these data are that µ1Bn-Ab arrested recycling of TfR in RE by blocking the function of AP1B at RE.
|
). To make more evident the blocking effect of the µ1B antibody, we adjusted the conditions of the experiment (expression time and serum concentration) to maximize the PM localization and minimize the endosomal localization of LDLR-GFP, Incubating the cells in serum-free medium was sufficient to cause such an effect (Supplementary Figure S1), suggesting that LDLR spends less time in the endocytic pathway in the absence of ligand. Under these conditions, control FRT cells comicroinjected with PI-IgG displayed LDLR-GFP mainly at the basolateral PM (Figure 5A, top three panels). By contrast, when µ1Bn-Ab was comicroinjected with the cDNA, LDLR-GFP moved normally to the PM and was internalized, but accumulated in a perinuclear compartment (Figure 5A, middle three panels). This accumulation was much slower with an endocytosis-deficient mutant of LDLR (LDLRY18A-GFP; Figure 5A, bottom three panels), further supporting the view that the antibody blocks LDLR trafficking postendocytically. Immunofluorescence experiments showed that the compartment that accumulated LDLR-GFP was enriched in both endogenous µ1B and TfR, but not in TGN38 (Figure 5, B and C). These results and the TfR experiments previously described (Figure 4) demonstrate that µ1Bn-Ab blocks the recycling of LDLR and TfR to the basolateral surface, promoting their accumulation in RE closely apposed to the TGN. To our knowledge, this is the first report of an antibody against a subunit of an AP adaptor that has trafficking-blocking properties.
|
) showed striking results (Figure 6A). The µ1Bn-Ab did not alter the kinetics of LDLR-GFP release from the Golgi area, in agreement with the results described above (Figure 5). Instead, it drastically inhibited the release of VSVG protein and TfR from the perinuclear region (Figure 6A). Images with higher exposure times, which saturate the perinuclear staining but are necessary to visualize the weaker surface GFP fluorescence, demonstrated that VSVG-GFP and TfR-GFP reached the basolateral membrane in cells microinjected with control PI-IgG but were blocked intracellularly in the presence of µ1Bn-Ab (Figure 6B).
|
2 d. Under our experimental conditions, FRT cells exhibited a relatively flat phenotype that facilitated our microinjection and quantitative live imaging assays; however, they were sufficiently polarized to localize TfR, VSVG, and LDLR predominantly to the basolateral membrane and the apical variant of tsO45-VSVG-GFP (VSVG3-GFP; Keller et al., 2001), to the apical surface, relative to the tight junction marker protein ZO-1 (Supplementary Figure S2).
Newly Synthesized VSVG Protein and TfR Reach RE Directly from the TGN, Whereas LDLR Reaches RE Postendocytically
The µ1B-Ab blocking effects allowed us to investigate a second major question regarding the biology of AP1B, i.e., does the adaptor work in TGN or in RE intersecting the biosynthetic pathway? The experiments described above did not elucidate the nature of the perinuclear compartment, where the AP1B antibody stopped the biosynthetic route of basolateral proteins. To identify this compartment, we accumulated GFP-tagged LDLR, TfR, and VSVG proteins in the TGN using microinjection and the 20°C block as in Figure 5 and quantified with MetaMorph software the extent of colocalization of GFP with TGN38 and µ1B before and 10 min after releasing the temperature block (Figure 7). As expected, more than 75% of all three proteins colocalized with TGN38 before release of the 20°C block and exited the TGN quickly upon release of the block (Figure 7, A–C). Remarkably, different cargo proteins showed strikingly different rates of transport from TGN to the AP1B-positive compartment. LDLR-GFP did not colocalize significantly with µ1B 10 min after release of the 20°C block (Figure 7A), whereas the bulk (80%) of TfR-GFP and VSVG-GFP moved to the AP1B compartment within that time (Figure 7, B and C), and these proteins were blocked from exiting this compartment by µ1Bn-Ab. Time-course analysis (Figure 7D) demonstrated that TfR and VSVG achieved 75% colocalization with µ1B in just 5 min and reached a plateau at 80% colocalization. By contrast, LDLR only started to colocalize with µ1B after a long lag period of 60 min, likely reflecting its passage through the basolateral PM, internalization, and endosomal trafficking before accumulation at the AP1B-positive compartment (Figure 7D, see also Figure 5). These experiments clearly demonstrate that the major site of function of AP1B in the biosynthetic route is not the TGN, but closely apposed RE (see model in Figure 9).
|
) showed that µ1Bn-Ab elicited the perinuclear accumulation of all three GFP-tagged reporters only in cells expressing µ1B (Figure 8A). Furthermore, the apical variant of tsO45-VSVG-GFP (VSVG3-GFP; Keller et al., 2001), bearing a basolateral sorting motif masked by GFP (Ang et al., 2003, 2004; Folsch et al., 2003; Yeaman et al., 2004), was not perinuclearly arrested by µ1Bn-Ab (Figure 8B). Together with the results described before (Figure 5), these experiments demonstrate that the µ1Bn-Ab inhibits basolateral transport in cargo- and compartment-specific manners: it blocks biosynthetic delivery of VSVG protein and TfR and the postendocytic recycling of LDLR and TfR to the basolateral membrane.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
3–5 min) from the TGN to AP1B-containing RE. These results provide solid experimental support for the emerging paradigm that RE and TGN have a cooperative sorting role in the biosynthetic route.
The ability of the antibody to react with µ1B in live or fixed FRT cells was unexpected because crystallography and modeling data (Heldwein et al., 2004) indicate that the N-terminal epitope of µ1B used as the antigen is likely buried within the fully assembled AP1 complex. Furthermore, the antibody did not react with MDCK cells even though the N-terminal sequence of canine µ1B is highly homologous to both rat and human µ1B. However, the following controls confirmed that the antibody reacted specifically with rat and human µ1B: 1) the peptide antigen abrogated both the immunostaining and the transport-blocking effects of the antibody; 2) µ1B immunostaining was observed upon transfection of µ1B into LLCPK1 cells (which normally lack this protein) and was decreased after partial RNAi-mediated knockdown of µ1B expression in FRT cells; and 3) the antibody displayed traffic-blocking effects in LLCPK1 cells that were only observed upon µ1B transfection. Our experiments suggest that AP1B may be present not only as a fully assembled adaptor, but also in incompletely assembled configurations that are essential for its biological function and that can vary in different cells. This latter property could explain the higher reactivity of the antibody in FRT cells than in MDCK cells. Indeed, published evidence suggests the presence of 
- and β-µ dimers in cells (Deneka et al., 2003) and indicates that µ2, the medium subunit of AP2 (Liu et al., 1994) and µ1B (Folsch et al., 1999; Sugimoto et al., 2002) may exist unincorporated into tetrameric AP complexes. Other experiments suggest that conformational changes leading to exposure of the µ2 subunit in AP2 (Haucke and De Camilli, 1999; Collins et al., 2002) and µ1A in AP1A (Ghosh and Kornfeld, 2003) might be triggered during the functional cycle of AP adaptors. It is also possible that the attachment of the adaptor to its binding site (in this manuscript shown to involve ARF activity) may promote an open configuration different from that observed by crystallography. Unfortunately, our µ1B antibody was unsuitable for immunoprecipitation, which prevented us from carrying out additional experiments to discern among these possibilities.
Experiments using this antibody localized AP1B to a perinuclear compartment that, although physically very close to the TGN, lacked typical TGN markers. This compartment was enriched in TfR and did not contain EEA1, a marker of early sorting endosomes (Wilson et al., 2000), or Rab11, a marker of AREs in MDCK cells (Casanova et al., 1999; Wang et al., 2001; Figure 2A). All of these characteristics are typical of a compartment denominated common RE, involved in the recycling of TfR and LDLR to the basolateral membrane and in the transcytosis of Polymeric IgA Receptor to the apical membrane via ARE (Futter et al., 1998; Brown et al., 2000; Wang et al., 2000). Additional experiments showed that the compartment decorated by µ1Bn-Ab, unlike the TGN, was insensitive to tubulation induced by a kinase-dead form of PKD (Figure 3), a protein recently implicated in the regulation of the trafficking of basolateral proteins (Yeaman et al., 2004). Taken together, these experiments provide strong support to our original suggestion that AP1B localizes predominantly to RE (Gan et al., 2002).
Surprisingly, the µ1Bn antibody blocked the recycling of TfR and LDLR to the basolateral membrane, promoting the accumulation of these fast-recycling receptors in perinuclear RE (Figures 4 and 5). These results suggest that RE is a site of function of µ1B, congruently with previous evidence (Gan et al., 2002), but cannot discard an additional sorting role of AP1B at the TGN. The possibility that AP1B might function at the TGN, originally proposed on the basis of EM immunogold data (Folsch et al., 1999, 2001), gained further traction based on our recent observations in siRNA knockdown MDCK cells, which demonstrated µ1B-dependent basolateral sorting both during biosynthetic and recycling traffic (Gravotta et al., 2007). Although AP1B was mainly detected in an endosomal fraction, it cannot be disregarded that a small, highly dynamic pool of this adaptor actually functions at the TGN. We took advantage of the trafficking-arresting properties of µ1Bn-Ab to provide a definitive answer to this question. The µ1Bn-Ab blocked biosynthetic trafficking of VSVG and TfR not in the TGN but in a closely apposed µ1B-enriched RE that they reached within 5 min after leaving the TGN (Figure 7). Importantly, LDLR only started to accumulate in RE 45 min after leaving the TGN, reflecting the time required for this protein to be exocytosed, internalized, and transported to perinuclear RE (Figure 7). These data conclusively demonstrate that the sorting role of AP1B in the biosynthetic route takes place not at both TGN and RE, but exclusively in RE located physically and kinetically very close to the TGN. Our results are also the first to show that AP1B-containing RE are an obligatory step in the biosynthetic route of some but not all basolateral proteins and that the TGN discriminates among distinct tyrosine-dependent basolateral motifs.
The mechanisms utilized by AP1B to recognize different basolateral sorting signals have not yet been fully elucidated. AP1B's µ1B subunit shares with AP2's µ2 subunit a pocket lined with a cluster of conserved amino acid residues that recognize Yxx
motifs (Owen and Evans, 1998; Bonifacino and Dell'Angelica, 1999). This pocket is required for the sorting of basolateral proteins with typical Yxx motifs, e.g., VSV G protein, but not for basolateral sorting of AP1B- dependent TfR and LDLR, suggesting that AP1B utilizes nonpocket regions to interact with these proteins' atypical basolateral signals (Sugimoto et al., 2002). In TfR, the basolateral sorting signal is a GDNS sequence downstream of the endocytic motif YTRF, which lacks basolateral signaling ability (Dargemont et al., 1993; Odorizzi and Trowbridge, 1997). LDLR displays a "weak" membrane-proximal basolateral signal constituted by its endocytic motif NPVY and a more distal "strong" basolateral signal constituted by an atypical tyrosine motif, which depends on Tyr 35, the adjacent Gly34, and a downstream acidic cluster (EDD; Matter et al., 1992, 1993; Koivisto et al., 2001). It is conceivable that AP1B may interact with these signals via subunits other than µ1B, as shown for the interaction between AP1 or AP3 with (DE)XXXL(LI) motifs (Janvier et al., 2003).
The observation that the µ1Bn antibody blocked transport out of RE, rather than promoting diversion of the basolateral proteins into an apical route, as typically observed by many groups when basolateral sorting signals are mutated, suggests that the antibody trapped basolateral proteins into a complex that prevented any type of transport. This possibility is supported by the finding that the transport of the three basolateral markers in LLC-PK1 cells, which lack µ1B expression, as well as the transport of VSVG3-GFP variant with its basolateral sorting signal inactivated (Keller et al., 2001), was not blocked by the µ1Bn-Ab (Figure 8). Only upon expression of exogenous µ1B did LLC-PK1 cells become sensitive to the blocking effects of µ1Bn-Ab. Thus, our antibody seems to arrest basolateral proteins into a complex that prevented transport out of RE in a µ1B- and basolateral motif–dependent manner. Other experiments also showed that BFA promoted dissociation of µ1B from RE (Figure 2B), thus indicating for the first time that BFA-sensitive ARF GTPase activity is crucially involved in AP1B recruitment to RE membranes. This result provides a mechanistic explanation for previous published data showing that BFA promotes missorting of recycling basolateral proteins (Matter et al., 1993; Futter et al., 1998; Wang et al., 2001).
Early reports using cell fractionation detected TfR (Futter et al., 1995) and other basolateral proteins (Leitinger et al., 1995; Orzech et al., 2000) in endosomes while en route from the TGN to the PM. The resolution of these experiments was insufficient to determine the magnitude of the TGN-endosome-PM trafficking or to identify the compartments involved. A more recent study (Ang et al., 2004), used videomicroscopy and biochemical assays in MDCK cells to reveal transient associations of VSVG-GFP protein with transferrin-positive RE minutes after release from the TGN block. These authors, utilizing an endosomal ablation procedure based on HRP-peroxidase, suggested that 85% of newly synthesized VSVG traveled through the endosomal compartment before reaching the cell surface. However, these experiments did not identify the site of the transport block as either RE containing Rab11 or AP1B, or other endocytic compartments (early endosomes), nor excluded the possibility that the traffic arrest might have take place at the TGN itself rather than at RE, because of indirect effects of the ablation procedure on TGN function (e.g., blocking of a recycling route essential for TGN function). Our results clearly go beyond this and other previous studies by demonstrating that two newly synthesized basolateral proteins, TfR and VSVG, must obligatorily transit AP1B-containing RE and that the bulk of these cargo proteins reach the AP1B-RE with very fast kinetics after exiting the TGN.
Our results also provide new mechanistic insights about the sorting machinery of the TGN and RE. Rab11-containing endosomes have been found to intersect the biosynthetic pathway of E-cadherin in HeLa cells, and the evidence suggested that Rab11 is required for dileucine-mediated basolateral sorting in MDCK cells (Lock and Stow, 2005). However, as mentioned, Rab11 associates with ARE but not with common RE involved in basolateral recycling of TfR or LDLR (Brown et al., 2000; Wang et al., 2000). In agreement with these reports, we did not observe an overlap between Rab11 and µ1B-containing endosomes in FRT cells (Figure 2A) and E-cadherin was only marginally affected in µ1B-knockdown MDCK cells (Gravotta et al., 2007). Therefore, different basolateral proteins might traverse different sets of RE in their biosynthetic routes. In addition, the evidence also indicates that basolateral proteins use different sorting elements and transport pathways to exit the TGN. We have shown in acute trafficking-blocking experiments with membrane-permeant peptides that the TGN discriminates between basolateral cargo bearing tyrosine-based or nonconventional sorting motifs (Soza et al., 2004). Our present results imply that the TGN also distinguishes basolateral proteins that possess distinct classes of tyrosine-based motifs (VSVG and LDLR) and sorts them either to AP1B-containing RE (VSVG) or to the PM bypassing this compartment (LDLR). Thus, our results are the most clear in predicting that different adaptors, distinct from AP1B, mediate sorting from TGN into routes to either RE or the PM. On the basis of previous reports (Nishimura et al., 2002; Simmen et al., 2002), we suggest that these adaptors might be AP3 and AP4 (Figure 9).
|
). These experiments suggested that TfR might switch to a direct route from TGN to PM in more established MDCK monolayers; however, given the slow installation of the RNAi inhibition of AP1B expression, we were not able to eliminate other possibilities, e.g., the activation of a different basolateral adaptor at RE in more mature MDCK monolayers. Because the experiments reported here were carried out only in recently polarized FRT cells (2 d of confluency), we cannot discard the possibility that these cells, like MDCK cells, might develop an AP1B-independent biosynthetic route for TfR at later confluency times. In summary, our results provide definitive evidence for the functional sorting role of AP1B in RE, and for the hypothesis, so far supported only by circumstantial data, that RE carry out an important sorting role in both the biosynthetic and recycling basolateral routes. Future work must complete the mapping of the basolateral routes and the adaptors involved in transport between TGN and RE or the PM and determine the role of RE in sorting events previously thought to occur in the TGN.
|
ACKNOWLEDGMENTS |
|---|
|
Footnotes |
|---|
The online version of this article contains supplemental material at MBC Online (http://www.molbiolcell.org). 
Address correspondence to: Alfonso González (agonzara{at}med.puc.cl).
Abbreviations used: AP, adaptor protein (complex); ARE, apical recycling endosomes; FRT, Fisher rat thyroid; LDLR, low-density lipoprotein receptor; RE, recycling endosomes; TfR, transferrin receptor; VSVG, vesicular stomatitis virus glycoprotein G.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Ang, A. L., Taguchi, T., Francis, S., Folsch, H., Murrels, J. L., Pypaert, M., Warren, G., and Mellman, I. (2004). Recycling endosomes can serve as intermediates during transport from the Golgi to the plasma membrane of MDCK cells. J. Cell Biol 167, 531–543.
Benmerah, A., Bayrou, M., Cerf-Bensussan, N., and Dautry-Varsat, A. (1999). Inhibition of clathrin-coated pit assembly by an Eps15 mutant. J. Cell Sci 112, (Pt 9), 1303–1311.[Abstract]
Benmerah, A., Lamaze, C., Begue, B., Schmid, S. L., Dautry-Varsat, A., and Cerf-Bensussan, N. (1998). AP-2/Eps15 interaction is required for receptor-mediated endocytosis. J. Cell Biol 140, 1055–1062.
Bonifacino, J. S., and Dell'Angelica, E. C. (1999). Molecular bases for the recognition of tyrosine-based sorting signals. J. Cell Biol 145, 923–926.
Bonifacino, J. S., and Lippincott-Schwartz, J. (2003). Coat proteins: shaping membrane transport. Nat. Rev. Mol. Cell Biol 4, 409–414.[CrossRef][Medline]
Brachet, V., Pehau-Arnaudet, G., Desaymard, C., Raposo, G., and Amigorena, S. (1999). Early endosomes are required for major histocompatiblity complex class II transport to peptide-loading compartments. Mol. Biol. Cell 10, 2891–2904.
Brown, P. S., Wang, E., Aroeti, B., Chapin, S. J., Mostov, K. E., and Dunn, K. W. (2000). Definition of distinct compartments in polarized Madin-Darby canine kidney (MDCK) cells for membrane-volume sorting, polarized sorting and apical recycling. Traffic 1, 124–140.[CrossRef][Medline]
Casanova, J. E., Wang, X., Kumar, R., Bhartur, S. G., Navarre, J., Woodrum, J. E., Altschuler, Y., Ray, G. S., and Goldenring, J. R. (1999). Association of Rab25 and Rab11a with the apical recycling system of polarized Madin-Darby canine kidney cells. Mol. Biol. Cell 10, 47–61.
Collins, B. M., McCoy, A. J., Kent, H. M., Evans, P. R., and Owen, D. J. (2002). Molecular architecture and functional model of the endocytic AP2 complex. Cell 109, 523–535.[CrossRef][Medline]
Dargemont, C., Le Bivic, A., Rothenberger, S., Iacopetta, B., and Kuhn, L. C. (1993). The internalization signal and the phosphorylation site of transferrin receptor are distinct from the main basolateral sorting information. EMBO J 12, 1713–1721.[Medline]
Deneka, M., Neeft, M., Popa, I., van Oort, M., Sprong, H., Oorschot, V., Klumperman, J., Schu, P., and van der Sluijs, P. (2003). Rabaptin-5alpha/rabaptin-4 serves as a linker between rab4 and gamma(1)-adaptin in membrane recycling from endosomes. EMBO J 22, 2645–2657.[CrossRef][Medline]
Ellis, M. A., Potter, B. A., Cresawn, K. O., and Weisz, O. A. (2006). Polarized biosynthetic traffic in renal epithelial cells: sorting, sorting, everywhere. Am. J. Physiol. Renal Physiol 291, F707–F713.
Fields, I. C., Shteyn, E., Pypaert, M., Proux-Gillardeaux, V., Kang, R. S., Galli, T., and Folsch, H. (2007). v-SNARE cellubrevin is required for basolateral sorting of AP-1B-dependent cargo in polarized epithelial cells. J. Cell Biol 177, 477–488.
Folsch, H. (2005). The building blocks for basolateral vesicles in polarized epithelial cells. Trends Cell Biol 15, 222–228.[CrossRef][Medline]
Folsch, H., Ohno, H., Bonifacino, J. S., and Mellman, I. (1999). A novel clathrin adaptor complex mediates basolateral targeting in polarized epithelial cells. Cell 99, 189–198.[CrossRef][Medline]
Folsch, H., Pypaert, M., Maday, S., Pelletier, L., and Mellman, I. (2003). The AP-1A and AP-1B clathrin adaptor complexes define biochemically and functionally distinct membrane domains. J. Cell Biol 163, 351–362.
Folsch, H., Pypaert, M., Schu, P., and Mellman, I. (2001). Distribution and function of AP-1 clathrin adaptor complexes in polarized epithelial cells. J. Cell Biol 152, 595–606.
Fuller, S. D., Bravo, R., and Simons, K. (1985). An enzymatic assay reveals that proteins destined for the apical or basolateral domains of an epithelial cell line share the same late Golgi compartments. EMBO J 4, 297–307.[Medline]
Futter, C. E., Connolly, C. N., Cutler, D. F., and Hopkins, C. R. (1995). Newly synthesized transferrin receptors can be detected in the endosome before they appear on the cell surface. J. Biol. Chem 270, 10999–11003.
Futter, C. E., Gibson, A., Allchin, E. H., Maxwell, S., Ruddock, L. J., Odorizzi, G., Domingo, D., Trowbridge, I. S., and Hopkins, C. R. (1998). In polarized MDCK cells basolateral vesicles arise from clathrin-gamma- adaptin-coated domains on endosomal tubules. J. Cell Biol 141, 611–623.
Gan, Y., McGraw, T. E., and Rodriguez-Boulan, E. (2002). The epithelial-specific adaptor AP1B mediates post-endocytic recycling to the basolateral membrane. Nat. Cell Biol 4, 605–609.[Medline]
Ghosh, P., and Kornfeld, S. (2003). AP-1 binding to sorting signals and release from clathrin-coated vesicles is regulated by phosphorylation. J. Cell Biol 160, 699–708.
Gravotta, D., Deora, A., Perret, E., Oyanadel, C., Soza, A., Schreiner, R., Gonzalez, A., and Rodriguez-Boulan, E. (2007). AP1B sorts basolateral proteins in recycling and biosynthetic routes of MDCK cells. Proc. Natl. Acad. Sci. USA 104, 1564–1569.
Griffiths, G., and Simons, K. (1986). The trans Golgi network: sorting at the exit site of the Golgi complex. Science 234, 438–443.
Haucke, V., and De Camilli, P. (1999). AP-2 recruitment to synaptotagmin stimulated by tyrosine-based endocytic motifs. Science 285, 1268–1271.
Heldwein, E. E., Macia, E., Wang, J., Yin, H. L., Kirchhausen, T., and Harrison, S. C. (2004). Crystal structure of the clathrin adaptor protein 1 core. Proc. Natl. Acad. Sci. USA 101, 14108–14113.
Janvier, K., Kato, Y., Boehm, M., Rose, J. R., Martina, J. A., Kim, B. Y., Venkatesan, S., and Bonifacino, J. S. (2003). Recognition o
