Molecular Biology of the Cell track citations

Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
QUICK SEARCH:   [advanced]


     

Originally published as MBC in Press, 10.1091/mbc.E06-12-1162 on May 30, 2007

Vol. 18, Issue 8, 2991-3001, August 2007

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
E06-12-1162v1
18/8/2991    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kametaka, S.
Right arrow Articles by Bonifacino, J. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kametaka, S.
Right arrow Articles by Bonifacino, J. S.

Canonical Interaction of Cyclin G–associated Kinase with Adaptor Protein 1 Regulates Lysosomal Enzyme Sorting

Satoshi Kametaka*, Kengo Moriyama*,{dagger}, Patricia V. Burgos*, Evan Eisenberg{ddagger}, Lois E. Greene{ddagger}, Rafael Mattera*, and Juan S. Bonifacino*

*Cell Biology and Metabolism Branch, National Institute of Child Health and Human Development, and {ddagger}Laboratory of Cell Biology, National Heart Lung and Blood Institute, National Institutes of Health, Bethesda, MD 20892

Submitted December 29, 2006; Revised May 1, 2007; Accepted May 21, 2007
Monitoring Editor: Sandra Lemmon


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The adaptor protein 1 (AP1) complex is a heterotetramer that participates in cargo sorting into clathrin-coated vesicles at the trans-Golgi network (TGN) and endosomes. The {gamma} subunit of AP1 possesses a C-terminal "ear" domain that recruits a cohort of accessory proteins through recognition of a shared canonical motif, {Psi}G[PDE][{Psi}LM] (where {Psi} is an aromatic residue). The physiological relevance of these ear-motif interactions, however, remains to be demonstrated. Here we report that the cyclin G–associated kinase (GAK) has two sequences fitting this motif, FGPL and FGEF, which mediate binding to the AP1-{gamma}-ear domain in vitro. Mutation of both {gamma}-ear–binding sequences or depletion of AP1-{gamma} by RNA interference (RNAi) decreases the association of GAK with the TGN in vivo. Depletion of GAK by RNAi impairs the sorting of the acid hydrolase, cathepsin D, to lysosomes. Importantly, expression of RNAi-resistant GAK restores the lysosomal sorting of cathepsin D in cells depleted of endogenous GAK, whereas expression of a similar construct bearing mutations in both {gamma}-ear–binding sequences fails to correct the sorting defect. Thus, interactions between the {Psi}G[PDE][{Psi}LM]-motif sequences in GAK and the AP1-{gamma}-ear domain are critical for the recruitment of GAK to the TGN and the function of GAK in lysosomal enzyme sorting.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Newly synthesized acid hydrolase precursors are modified with mannose 6-phosphate groups as they traverse the cisternae of the Golgi complex. This modification enables them to bind to transmembrane mannose 6-phosphate receptors (MPRs) upon arrival at the trans-Golgi network (TGN; Ghosh et al., 2003Go). The resulting hydrolase-MPR complexes concentrate within clathrin-coated areas of the TGN by virtue of interactions between the cytosolic tails of the receptors and clathrin adaptors. This is followed by formation of clathrin-coated vesicles or carriers that transport the hydrolase-MPR complexes to endosomes. The low pH of the endosomal lumen dissociates these complexes, after which the hydrolase precursors are delivered to lysosomes while the receptors return to the TGN (Ghosh et al., 2003Go). Two types of clathrin adaptor have been proposed to function in hydrolase-MPR complex sorting at the TGN (and possibly at endosomes as well): members of the GGA family of monomeric proteins and the adaptor protein 1 (AP1) heterotetrameric complex (Ghosh et al., 2003Go). The GGA family comprises three members in mammals (GGA1, GGA2, and GGA3), each of which consists of four domains named VHS, GAT, hinge, and GAE ({gamma}-adaptin ear) (Bonifacino, 2004Go). AP1 is composed of four subunits, three of which occur as two or three distinct isoforms encoded by different genes. The subunits of AP1 and their isoforms (in parentheses) are named {gamma} ({gamma}1 and {gamma}2), beta1, µ1 (µ1A and µ1B), and {varsigma}1 ({varsigma}1A, {varsigma}1B, and {varsigma}1C) (Boehm and Bonifacino, 2002Go; Robinson 2004Go). The amino-terminal portions of {gamma} and beta1, together with µ1 and {varsigma}1, form the "core" of AP1, whereas the carboxy-terminal portions of both {gamma} and beta1 encompass two domains named "hinge" and "ear."

The GAE domains of the GGA proteins and the ear domains of the AP1-{gamma} subunit isoforms have a similar globular fold consisting of an eight-stranded immunoglobulin-like beta-sandwich (Kent et al., 2002Go; Nogi et al., 2002Go; Collins et al., 2003Go; Miller et al., 2003Go). Consistent with this structural similarity, all five domains bind with various affinities to a common set of "accessory proteins," including rabaptin-5 (Hirst et al., 2000Go; Doray and Kornfeld, 2001Go; Shiba et al., 2002Go; Mattera et al., 2003Go), {gamma}-synergin (Page et al., 1999Go; Hirst et al., 2000Go; Takatsu et al., 2000Go), p56 (Collins et al., 2003Go; Lui et al., 2003Go), NECAP1 and NECAP2 (Ritter et al., 2003Go; Mattera et al., 2004Go), aftiphilin (Mattera et al., 2004Go), {gamma}-BAR (Neubrand et al., 2005Go), and a protein known as Clint, enthoprotin, or epsinR (Kalthoff et al., 2002Go; Wasiak et al., 2002Go; Hirst et al., 2003Go; Mills et al., 2003Go). Rabaptin-5 is part of a complex with the Rab5 guanine-nucleotide-exchange factor, rabex-5 (Horiuchi et al., 1997Go; Mattera et al., 2003Go), whereas {gamma}-synergin and aftiphilin form a complex with another protein named p200 (Hirst et al., 2005Go). These accessory proteins share a canonical peptide motif that mediates interactions with the GGA-GAE and {gamma}-ear domains (Duncan et al., 2003Go; Mills et al., 2003Go; Mattera et al., 2003Go, 2004Go). Systematic bioinformatic, mutational, and binding analyses allowed us to define this canonical motif as {Psi}G[PDE][{Psi}LM] (where {Psi} is an aromatic residue; pattern denoted according to PROSITE syntax; Mattera et al., 2004Go; Table 1).


View this table:
[in this window]
[in a new window]

  		Table 1. Accessory proteins that contain sequences fitting the {Psi}G[PDE][{Psi}LM] motif

 
Peptides fitting this motif bind to two contiguous, shallow hydrophobic depressions formed by the side chains of basic groups on the ear domains (Collins et al., 2003Go; Miller et al., 2003Go; see Figure 2B). In addition, some GGA and AP1-{gamma} accessory proteins bind directly to clathrin (e.g., enthoprotin; Kalthoff et al., 2002Go; Wasiak et al., 2002Go; Hirst et al., 2003Go; Mills et al., 2003Go), phosphoinositides (e.g., enthoprotin; Kalthoff et al., 2002Go; Hirst et al., 2003Go; Mills et al., 2003Go), Rab proteins (e.g., rabaptin-5; Stenmark et al., 1995Go; Vitale et al., 1998Go), and NPF-motif–containing proteins (e.g., {gamma}-synergin; Page et al., 1999Go; Fernandez-Chacon et al., 2000Go). Multiple protein interactions may thus contribute to the recruitment of GGA and AP1 accessory proteins to membranes. Despite the detailed understanding of the structural bases of {Psi}G[PDE][{Psi}LM]-ear interactions, their physiological relevance vis-à-vis other possible interactions remains to be demonstrated.

Here we report that the AP1-binding, cyclin G–associated kinase (GAK; also known as auxilin 2; Kanaoka et al., 1997Go; Kimura et al., 1997Go; Greener et al., 2000Go; Umeda et al., 2000Go; Lee et al., 2005Go, 2006Go) specifically interacts with the ear domains of AP1-{gamma}1 and -{gamma}2 through two sequences fitting the {Psi}G[PDE][{Psi}LM] consensus motif. More importantly, we demonstrate that this interaction is required for the association of GAK with the TGN and for the transport of the precursor of the acid hydrolase, cathepsin D, to lysosomes. These findings constitute the first demonstration that a canonical interaction between AP1 and one of its accessory proteins is critical for lysosomal enzyme sorting.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Cell Culture and Transfection
HeLa cells were obtained from the American Type Culture Collection (ATCC, Manassas, VA) and maintained in DMEM (Biosource, Rockville, MD) containing 4.5 g/l glucose, 50 U/ml penicillin, 50 µg/ml streptomycin, 2 mM L-glutamine, and 10% vol/vol fetal bovine serum (FBS). Transient transfection of cells was performed using FuGENE6 (Roche Molecular Biochemicals, Indianapolis, IN) or Lipofectamine 2000 (Invitrogen, Carlsbad, CA), according to the manufacturers' instructions.

Plasmids
The pEGFP-GAK and -auxilin plasmids were described previously (Lee et al., 2006Go). Site-directed mutagenesis was carried out with the QuickChange kit (Stratagene, La Jolla, CA) using pEGFP-GAK as a template to introduce the F961A-L964A, F981A-F984A, and K69A substitutions. EcoRI-KpnI fragments of the wild-type (wt) and the double mutant GAK constructs were subcloned into the corresponding sites of the pPAGFP(A206K) plasmid to generate expression vectors for photoactivatable green fluorescent protein (PAGFP)-tagged GAK proteins. A mammalian expression vector encoding EGFP-AAK1 was kindly provided by Sean Conner (University of Minnesota, Minneapolis, MN). pGEX-mouse {gamma}1-ear was described previously (Mattera et al., 2003Go). Site-directed mutagenesis of the {gamma}1-ear to introduce A753Q, K756Q, and R793Q substitutions was also performed using the QuickChange kit. For simplicity, enhanced GFP will hereafter be referred to as GFP.

Antibodies
Mouse monoclonal antibodies to AP1-{gamma}1, clathrin heavy chain (CHC), and GGA3 were purchased from BD Biosciences PharMingen (San Diego, CA). Mouse monoclonal antibodies to GFP were purchased from Covance (Princeton, NJ) and Roche. Mouse mAb to CHC (clone X22) was obtained from the Developmental Studies Hybridoma Bank (University of Iowa, Iowa City, IA). Sheep polyclonal antibody to TGN46 was purchased from Serotec (Oxford, United Kingdom), and rabbit polyclonal antibody to cathepsin D was from Calbiochem (EMD Biosciences, San Diego, CA). A rabbit polyclonal antibody to GAK was described previously (Greener et al., 2000Go). Donkey Alexa488- or Alexa594-conjugated anti-rabbit or anti-mouse IgG were purchased from Molecular Probes (Eugene, OR). Donkey Cy3-conjugated anti-sheep antibody was obtained from Jackson ImmunoResearch Laboratories (West Grove, PA). Horseradish peroxidase–conjugated anti-mouse or anti-rabbit IgG were purchased from Amersham Biosciences (Piscataway, NJ).

Pulldown Assays
HeLa cells were cultured on 10-cm dishes to 60–80% confluence. Twenty-four hours before the pulldown experiments, cells were transfected with the pEGFP-GAK, -AAK1 or -auxilin constructs. Cells were washed twice with ice-cold phosphate-buffered saline (PBS) and extracted in lysis buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.25% wt/vol Triton X-100, 1x protease inhibitor cocktail [Roche], 1x protein phosphatase inhibitor cocktail I and II [Sigma, St. Louis, MO]) for 30 min at 4°C, followed by centrifugation at 20,000 x g for 15 min. The supernatants were subsequently incubated with glutathione-Sepharose 4B beads (Amersham Biosciences) for 30 min at 4°C, followed by centrifugation at 20,000 x g for 15 min, to yield precleared lysates. Thirty micrograms of purified GST or GST-fusion proteins was incubated with precleared lysates (derived from half a 10-cm dish) in the presence or absence of competing peptides for 2 h at 4°C with gentle mixing. The GST-protein complexes were subsequently pulled down by incubation with 30 µl (bed volume) glutathione-Sepharose 4B beads for 1 h at 4°C. The beads were next washed three times with 1 ml ice-cold lysis buffer, and the bound proteins eluted by boiling in 40 µl Laemmli buffer. Samples were analyzed by SDS-PAGE and immunoblotting.

RNA Interference
RNA interference (RNAi) in HeLa cells was carried out as described previously (Janvier and Bonifacino, 2005Go). Small interfering RNA (siRNA) for CHC was also described previously (Janvier and Bonifacino, 2005Go). SmartPool siRNA for human AP1-{gamma}1 and a custom-designed siRNA for human GAK (5'-AACGAAGGAACAGCUGAUUCA-3') were purchased from Dharmacon, (Chicago, IL). This GAK siRNA targets the 3' untranslated region of the cDNA and, therefore, does not affect the expression of transfected GAK open reading frame (ORF) constructs lacking this region. Cells were analyzed after 72 h of treatment with siRNAs. In the rescue experiments, 48 h after transfection with GAK siRNA, the cells depleted of endogenous GAK were transfected with GFP-GAK ORF constructs using Lipofectamine 2000 and analyzed 24 h later. The transfection efficiency of these rescue plasmids was ~70% as assessed by fluorescent microscopy.

Isothermal Titration Calorimetry
Recombinant GST-{gamma}1-ear was expressed in Escherichia coli and purified with glutathione-Sepharose 4B according to the manufacturer's instructions. This protein as well as GAK peptides (Table 2; synthesized by EZBiolab, Westfield, IN) were dialyzed overnight at 4°C against excess isothermal titration calorimetry (ITC) buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl). All ITC experiments were carried out at 30°C using a VP-ITC instrument (MicroCal LLC, Northampton, MA). Typically, the chamber contained 1.4 ml of 40 µM GST-{gamma}1-ear, and the GAK peptide (~1.4 mM) was added in 32 injections of 7 µl each. Titration curves were analyzed using Origin software (MicroCal). Binding constants were calculated by fitting the curves corresponding to the wt and mutant GAK peptides to two- and one-site models, respectively (Table 3).


View this table:
[in this window]
[in a new window]

  		Table 2. Synthetic peptides used in this study

 


View this table:
[in this window]
[in a new window]

  		Table 3. Binding affinities of GAK peptides for GST-{gamma}1-ear calculated from ITC data

 
Fluorescence Microscopy and Image Analysis of Fixed Cells
Cells grown on cover glasses were fixed with 4% wt/vol paraformaldehyde (Electron Microscopy Sciences, Hatfield, PA) in PBS for 15 min at room temperature or with methanol:acetone (1:1) for 10 min at –20°C. Fixed cells were subsequently permeabilized for 5 min with 0.1% wt/vol Triton X-100 in PBS and blocked for 30 min with 1 mg/ml bovine serum albumin (BSA) at room temperature. After incubation with the primary and fluorescent secondary antibodies, the cells were imaged with a Zeiss 510 confocal microscope (Carl Zeiss, Thornwood, NY). For quantitative analysis of fluorescence intensities, nonsaturated images were captured with a Plan-Neofluar 25x/0.80 objective and a fully open pinhole. Fluorescence quantification was performed as described previously (Kametaka et al., 2005Go). In brief, intensities corresponding to total cellular fluorescence and TGN-associated fluorescence were measured using the NIH Image software after subtraction of background fluorescence outside the cells. Only cells that did not exhibit pixel saturation were used. Identical conditions were used for capturing the images corresponding to wt and mutant GFP-GAK constructs. Thirty-nine cells expressing wt GFP-GAK and 36 cells expressing the mutant GFP-GAK sites 1 and 2 (mut1/2) constructs were analyzed with the software, and the mean ± SD was calculated. The statistical significance of the differences in the distribution of wt and mutant GAK was analyzed using Student's t test.

Live Cell Imaging
Live cell imaging of GFP- or PAGFP-tagged GAK proteins were performed as described previously (Kametaka et al., 2005Go) with slight modifications. In brief, HeLa cells transiently expressing GFP- or PAGFP-GAK fusion proteins grown on glass-bottom culture dishes (MatTek, Ashland, MA) were imaged in buffered medium using a Zeiss 510 confocal microscope equipped with a stage heated to 37°C. Selective photobleaching was performed using a 488-nm laser at full power, and recovery was monitored by time-lapse imaging at 5-s intervals with low-intensity illumination. Photoactivation of PAGFP was carried out using a single hit from a 413-nm laser at full power, and fluorescence decay was monitored at 3-s intervals. Fluorescence quantification was carried out as described previously (Kametaka et al., 2005Go). Each experiment was repeated at least five times under identical conditions, and the mean ± SD was graphically represented.

Metabolic Labeling and Pulse-Chase Analysis of Cathepsin D
Metabolic labeling of HeLa cells was carried out as described (Wasmeier et al., 2005Go). Briefly, cells grown in a six-well plate were pulse-labeled for 2 h at 20°C using 0.1 mCi/ml [35S]methionine-cysteine (Express Protein Label; Perkin Elmer-Cetus, Boston, MA). Cells were subsequently chased for 1–5 h at 37°C in regular culture medium in the presence of 5 mM mannose 6-phosphate and excess cold methionine (0.06 mg/ml) and cysteine (0.1 mg/ml). After chase, cells were rinsed twice with PBS and lysed in 50 mM Tris, pH 7.5, 150 mM NaCl, 10% glycerol, 5 mM EDTA, 1% (vol/vol) Triton X-100, supplemented with 1x complete protease inhibitor cocktail (Roche). The extracts were immunoprecipitated with anti-cathepsin D antibody and analyzed by SDS-PAGE and fluorography. Quantification was performed using a Typhoon 9200 PhosphorImager (Amersham Biosciences) and ImageQuant analysis software.

In Vitro Kinase Assay
Lysates of HeLa cells transiently expressing GFP or GFP-GAK fusion constructs were subjected to immunoprecipitation with anti-GFP mAb. Equivalent amounts of immunoprecipitated GFP proteins were used for in vitro kinase assays using myelin basic protein (MBP) as a substrate (Rubinfeld and Seger, 1998Go). In brief, beads containing immunoprecipitated GFP proteins were incubated at 37°C for 30 min with 2 µg MBP in 15 µl of kinase reaction buffer (25 mM beta-glycerophosphate, pH 7.3, 10 mM MgCl2, 0.3% BSA, 1 mM dithiothreitol, 1 mM EGTA, 0.1 mM sodium orthovanadate, and 33 µM [{gamma}-32P]ATP (~4000 cpm/pmol). After incubation, the kinase reaction was terminated by addition of 5 µl 4x Laemmli buffer and boiling for 5 min. Samples were analyzed by SDS-PAGE. The phosphorylated MBP and autophosphorylated GFP-GAK were detected on a phosphor imager.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
GAK Contains Two Sequences Fitting the {Psi}G[PDE][{Psi}LM] Consensus Motif
A search of protein sequence databases using as a pattern the {gamma}-ear– and GGA-GAE–binding motif, {Psi}G[PDE][{Psi}LM] (Mattera et al., 2004Go), revealed two sequences fitting this motif, FGPL and FGEF, within a protein known as GAK or auxilin 2 (Kanaoka et al., 1997Go; Kimura et al., 1997Go; Table 1 and Figure 1). Both sequences are contained within a relatively unstructured segment (designated clathrin-binding or CB) of the protein that also contains binding motifs for the clathrin terminal domain (i.e., DLL) and for the ear domains of the AP2-{alpha} and -beta2 subunits (i.e., DPF and WAAW; Jha et al., 2004Go; Mishra et al., 2004Go; Figure 1). GAK also has a serine/threonine-kinase domain homologous to that of the mammalian adaptor-associated kinase AAK1 (Conner and Schmid, 2002Go) and the yeast Ark/Prk proteins (Zeng and Cai, 1999Go; Watson et al., 2001Go), and tensin (also known as PTEN-homology) and J domains similar to those of auxilin (Ahle and Ungewickell, 1990Go; Figure 1). Previous studies showed that GAK is enriched in plasma membrane and TGN clathrin coats (Greener et al., 2000Go; Lee et al., 2006Go), where it phosphorylates the AP2-µ2 and AP1-µ1 subunits, respectively (Umeda et al., 2000Go; Korolchuk and Banting, 2002Go). In addition, GAK was shown to promote clathrin uncoating, thus regulating receptor endocytosis at the plasma membrane (Greener et al., 2000Go; Zhao et al., 2001Go; Zhang et al., 2004Go; Lee et al., 2005Go) and sorting of MPRs and cathepsin D at the TGN (Lee et al., 2005Go; Zhang et al., 2005Go). These properties of GAK are consistent with the presence of putative {gamma}-ear– and GGA-GAE–binding motifs within its sequence.


Figure 1
View larger version (16K):
[in this window]
[in a new window]

  		Figure 1. Schematic representation of GAK and related proteins. The different domains of auxilin, AAK1 and GAK/auxilin 2 are indicated. The scheme also shows the putative AP2-binding (DPF and WAAW) and clathrin-binding (DLL) sequences, as well as the two {Psi}G[PDE][{Psi}LM]-type, putative {gamma}-ear–binding sequences (sites 1 and 2; highlighted in red) within the clathrin-binding (CB) domain of GAK.

 
Preferential Binding of GAK to the Ear Domains of AP1-{gamma}1 and -{gamma}2
We performed glutathione S-transferase (GST) pulldown assays to examine the possible interaction of the ear domains of various clathrin adaptors with GAK and related proteins (Figure 2A). To this end, we expressed in HeLa cells GFP-tagged GAK, -auxilin or -AAK1. Lysates from these cells were tested for binding to immobilized GST-ear fusion proteins. In agreement with previous reports (Owen et al., 1999Go; Umeda et al., 2000Go; Conner and Schmid, 2002Go; Praefcke et al., 2004Go), GAK, auxilin, and AAK1 were found to interact with the ear domain of the AP2-{alpha}C subunit isoform (Figure 2A). Importantly, GAK interacted with the ear domains of the AP1-{gamma}1 and -{gamma}2 subunit isoforms but not with the structurally related GGA-GAE domains (Figure 2A). In contrast, auxilin and AAK1 did not detectably bind to the {gamma}1- and {gamma}2-ear domains, and none of the tested proteins bound to the AP3-beta3B-ear domain (used here as a negative control; Figure 2A). As previously shown (Hirst et al., 2000Go; Doray and Kornfeld, 2001Go; Shiba et al., 2002Go; Mattera et al., 2003Go), rabaptin-5 interacted with the AP1-{gamma}1- and -{gamma}2-ear domains as well as the GAE domains of the three GGAs (positive control for GGA-GAE interactions), but not with the AP2 {alpha}C- and AP3 beta3B-ear domains (Figure 2A). These experiments thus demonstrated that GAK binds specifically to the {gamma}1- and {gamma}2-ear domains.


Figure 2
View larger version (43K):
[in this window]
[in a new window]

  		Figure 2. Interaction of GAK with the {gamma}-ear domain demonstrated by GST pulldown experiments. (A) Extracts from HeLa cells expressing GFP-tagged GAK, -auxilin, or -AAK1 were used in pulldown experiments with immobilized GST-ear constructs as described in Materials and Methods. GST-{alpha}C ear was used as a positive control, whereas GST and GST-beta3B-adaptin ear were used as negative controls. Bound proteins were analyzed by SDS-PAGE and immunoblotting (IB) with antibody to GFP. Immunoblots were also probed with antibody to rabaptin-5. A fraction (3%) of the input extract was run on the first lane of the top four gels for comparison. (B) Surface representation of the structure of the {gamma}1-ear domain (Kent et al., 2002Go). Basic residues that are important for the binding of accessory proteins, and the less important Ala753 are indicated in blue and orange, respectively (Nogi et al., 2002Go). (C) Extracts of HeLa cells expressing GFP-GAK were used in pulldown experiments with either wild-type (wt) or the mutant GST-{gamma}1-ear constructs indicated in the figure. Bound GFP-GAK was resolved by SDS-PAGE and detected by immunoblotting with antibody to GFP. Ponceau Red staining of the blotted GST-proteins is shown at the bottom of A and C. The apparent molecular masses of markers (in kDa) are indicated at the left of A and C.

 
GAK Binds to the Previously Defined Site for Accessory Proteins on the {gamma}1-ear Domain
X-ray crystallographic and structure-based mutational analyses (Kent et al., 2002Go; Nogi et al., 2002Go) have identified the residues on the {gamma}1-ear that are involved in interactions with sequences fitting the {Psi}G[PDE][{Psi}LM] motif (Figure 2B). GST pulldown assays showed that replacement of glutamine for either of two such {gamma}1 residues, Lys756 and Arg793, substantially reduced binding to GAK (Figure 2C). A similar replacement for Ala753, which is less important for binding to other accessory proteins (Nogi et al., 2002Go), had a weaker effect on GAK binding (Figure 2C). These results indicate that GAK binds to the previously defined site on {gamma}1 for {Psi}G[PDE][{Psi}LM]-containing accessory proteins.

Binding to the {gamma}1-ear Is Dependent on the GAK Sequences Fitting the {Psi}G[PDE][{Psi}LM] Motif
To assess whether binding to the {gamma}1-ear is indeed dependent on the GAK FGPL and FGEF sequences, we carried out competition pulldown experiments using synthetic peptides encompassing the two tetrapeptide sequences (residues 958–988; Table 2). The wild-type GAK peptide (GAK wt) inhibited the interaction between GAK and the {gamma}1-ear in a concentration-dependent manner (Figure 3A). Likewise, a wild-type enthoprotin peptide spanning residues 368–377 and containing a {Psi}G[PDE][{Psi}LM]-type sequence (FGDW; Table 2), inhibited the GAK–{gamma}1-ear interaction with a potency similar to that of the GAK wt peptide (Figure 3A). Substitutions of key residues in both the FGPL and FGEF sequences in the GAK peptide (mut1/2, Table 2) or in the FGDW sequence in the enthoprotin peptide (mut, Table 2), abrogated the ability of the peptides to compete (Figure 3B), indicating that the competition was mediated by the {Psi}G[PDE][{Psi}LM]-type sequences from both proteins. These observations demonstrated that the interaction with the {gamma}1-ear is dependent on the {Psi}G[PDE][{Psi}LM] motifs in GAK.


Figure 3
View larger version (39K):
[in this window]
[in a new window]

  		Figure 3. Peptide competition of the interaction between GAK and the {gamma}1-ear. (A) Pulldown of GFP-GAK by immobilized GST-{gamma}1-ear was carried out as indicated in the legend to Figure 2 and in Materials and Methods, except that peptides derived from GAK or enthoprotin (wt sequences; Table 2) were added to the incubation mixture at concentrations ranging from 75 to 0.3 µM in 1:3 serial dilutions. Samples corresponding to 5% of the input extract and the pulldown of GFP-GAK by GST-{gamma}1-ear in the absence of peptide (–) were run on the first two lanes for comparison. (B) Pulldown of GFP-GAK by immobilized GST-{gamma}1-ear in the absence (–) or presence of 75 µM of the wild-type (wt) or mutant GAK (mut1/2) peptides, or the wild-type (wt) or mutant (mut) enthoprotin peptides described in Table 2. In both A and B, bound GFP-GAK was detected by immunoblotting (IB) with antibody to GFP. Blots were also probed for GGA3 as a specificity control. Ponceau Red staining of blotted proteins is shown at the bottom of both panels. The apparent molecular masses of markers (in kDa) are indicated on the left of both panels.

 
Predominant Role of the FGEF Sequence in GAK for Interactions with the {gamma}1-ear Domain
To discern the relative contribution of the FGPL and FGEF sequences in the binding of GAK to the {gamma}1-ear, we performed competition pulldown assays using peptides with substitutions in each sequence (mut1 and mut2 peptides, respectively; Table 2). As shown in Figure 4A, although the mut1 peptide was almost as effective as the wt peptide in inhibiting binding of GAK to the {gamma}1-ear, the mut2 peptide was inactive in this assay, indicating that the FGEF sequence is more important than the FGPL sequence for {gamma}1-ear binding.


Figure 4
View larger version (39K):
[in this window]
[in a new window]

  		Figure 4. Relative contributions of two canonical GAK sequences to binding to the {gamma}1-ear. (A) Wild-type GAK peptide (wt) and GAK peptides with substitutions in two {Psi}G[PDE][{Psi}LM]-motif sequences (mut1, mut2, or mut1/2; see Table 2; used at 75 µM) were examined for their ability to inhibit the GAK–{gamma}1-ear interaction in GST pulldown assays similar to those described in the legend to Figure 3 and in Materials and Methods. (B) Extracts of HeLa cells expressing GFP-tagged, full-length GAK (wt or substituted as in the mut1, mut2, or mut1/2 peptides; Table 2) were used in pulldown assays with GST-{gamma}1-ear, -{alpha}C-ear, and -GGA3-GAE as described in the legend to Figure 2 and in Materials and Methods. Ponceau Red staining of blotted proteins is shown at the bottom of A and B. The apparent molecular masses of markers (in kDa) are indicated on the left of both panels.

 
We also mutated the FGPL and/or FGEF sequences in the context of full-length GFP-GAK and performed pulldown assays with GST-fusions to the {gamma}1-ear, {alpha}C-ear, or GGA3-GAE domains (the latter two as positive and negative controls, respectively; Figure 4B). Consistent with the peptide competition experiments (Figure 4A), mutation of FGEF substantially reduced binding to the {gamma}1-ear (mut2, Figure 4B). Mutation of FGPL alone, on the other hand, had no effect on binding (mut1, Figure 4B), but had a synergistic effect when combined with the FGEF mutation (mut1/2, Figure 4B). None of these mutations affected the interaction of GAK with the {alpha}C-ear, underscoring the specificity of the interaction with the {gamma}1-ear.

Finally, we analyzed these interactions by ITC using recombinant GST-{gamma}1-ear and synthetic GAK peptides (Table 2) as ligands. We found that the wt GAT peptide bound to the {gamma}1-ear (Figure 5) with an isotherm that could be best fitted to a two-site model, yielding equilibrium dissociation constants of 47 and 1200 µM for each site (Table 3). Amino acid substitutions in each motif yielded single-site isotherms with very different effects on affinity. Although substitutions in the FGPL motif (mut1) alone had only a slight effect on the interaction, substitutions in the FGEF motif alone (mut2) or in both motifs together (mut1/2) largely abrogated the interaction (Figure 5 and Table 3). Taken together, these experiments indicated that interactions with the {gamma}1-ear are mainly mediated by the FGEF sequence in GAK.


Figure 5
View larger version (24K):
[in this window]
[in a new window]

  		Figure 5. ITC analysis of the binding of GAK peptides (wt, mut1, mut2, and mut1/2; Table 2) to GST-{gamma}1 ear, performed as described in Materials and Methods. The binding isotherms shown are representative of at least three experiments per peptide. The Kd values estimated from these experiments are shown in Table 3. The inset shows the heat change elicited by serial injections of wild-type GAK peptide onto GST-{gamma}1-ear.

 
Association of GAK with the TGN Is Dependent on AP1
We next examined whether the interactions described above are important for the intracellular localization of GAK. In agreement with previous reports (Greener et al., 2000Go; Zhang et al., 2005Go; Lee et al., 2006Go), immunofluorescence microscopy analyses showed that a fraction of endogenous GAK colocalized with TGN46 (Figure 6, A–C), clathrin (Figure 6, D–F), and AP1 (Figure 6, G–I) at the TGN. Treatment with brefeldin A (BFA), which dissociates Arf1-regulated coat proteins such as AP1 from membranes (Robinson and Kreis, 1992Go; Wong and Brodsky, 1992Go), redistributed GAK to the cytosol (Figure 6, J–L). Treatment for 3 d with specific siRNAs showed that depletion of the {gamma}1 subunit of AP1 (Figure 7A) caused dissociation of GAK from the TGN (Figure 7, E–G). Depletion of GAK, on the other hand, did not lead to dissociation of AP1 from membranes, although it did cause some dispersal of AP1-containing structures (Figure 7, B–D). These changes were not due to disruption of the Golgi complex, because the distribution of the Golgi-stack marker GM130 was unaffected by depletion of AP1 subunits or GAK (data not shown). These results thus demonstrated that association of GAK with the TGN is dependent on AP1. Longer periods of siRNA treatment (i.e., 5–7 d) did cause disruption of the Golgi complex in some cells, as detected by staining with antibodies to GM130 and TGN46 (data not shown). This was in agreement with previous observations made by plasmid-based RNAi knockdown, in which the cells were selected with hygromycin B for 5–7 d after transfection with RNAi vectors (Lee et al., 2005Go). Golgi disruption upon prolonged RNAi treatment may be due to more complete depletion of GAK or to the accumulated effects of impaired trafficking. To avoid this effect on Golgi structure, subsequent functional experiments (i.e., see Figures 10 and 11) were performed by depleting cells of GAK for 3 d.


Figure 6
View larger version (91K):
[in this window]
[in a new window]

  		Figure 6. Colocalization of GAK with TGN46, clathrin and AP1 at the TGN, and effect of brefeldin A. HeLa cells incubated in the absence (A–I) or presence of 5 µg/ml brefeldin A for 15 min (J–L) were fluorescently coimmunostained for endogenous GAK (A, D, G, and J) and TGN46 (B), clathrin heavy chain (E). or AP1-{gamma}1 (H and K), and examined by confocal microscopy as described in Materials and Methods. C, F, I, and L are merged images. Bars, 10 µm.

 


Figure 7
View larger version (84K):
[in this window]
[in a new window]

  		Figure 7. AP1-{gamma}1 depletion decreases association of GAK with the TGN. (A) Levels of endogenous GAK, AP1-{gamma}1, and actin in cells from control and GAK-RNAi- and AP1-{gamma}1-RNAi-treated HeLa cells. (B–G) Immunofluorescence microscopy for AP1-{gamma}1 (B–D) and GAK (E–G) in control HeLa cells (B and E) and HeLa cells treated with siRNAs for GAK (C and F) or AP1-{gamma}1 (D and G). Asterisks indicate a cell in which AP1-{gamma}1 expression was not silenced. Bar, 10 µm.

 
The {gamma}-ear–binding Motifs Contribute to the Association of GAK with the TGN
To determine the importance of the {gamma}-ear–binding motifs for the association of GAK with the TGN, we compared the localization of GFP-tagged forms of wild-type GAK (wt) and GAK with mutations in both the FGPL and FGEF sequences (mut1/2) in transiently transfected HeLa cells. GFP-tagged wt GAK showed good localization to the TGN, as determined by costaining with antibody to TGN46 (Figure 8, A–F). Mutation of the FGPL and FGEF sequences, however, caused a shift in the distribution of GAK from the TGN to the cytosol, although a small amount of TGN staining could be seen in some cells (Figure 8, G–I; quantitative analysis shown in Figure 8, J and K). Fluorescence recovery after photobleaching (FRAP) analysis of the TGN showed a faster rate of recovery for mut1/2 GAK relative to wt GAK (Figure 9A, Table 4). Moreover, the dissociation of PAGFP-GAK mut1/2 from the TGN was also faster than that of GAK wt (Figure 9B, Table 4). These results indicate that mutation of the FGPL and FGEF motifs results in faster exchange of GAK between TGN and cytosolic pools, which reflects a looser association of mutant GAK with TGN clathrin/AP1 coats.


Figure 8
View larger version (49K):
[in this window]
[in a new window]

  		Figure 8. The ability of GAK to bind AP1 is important for its association with the TGN. HeLa cells transiently transfected with plasmids encoding wt (A–F) or mut1/2 (G–I) GFP-GAK were fluorescently counterstained with antibody to TGN46 (B, E, and H). A–C and D–F are two examples of the colocalization of GFP-GAK with TGN46 at different magnification. Bars, 10 µm. (J) Quantitative analysis of the relative fluorescence intensity of wt ({circ}; 39 cells) and mut1/2 (bullet; 36 cells) GFP-GAK proteins associated with the TGN. (K) Mean ± SD of datasets in G are represented by bar graphs. The statistical significance of the difference in TGN association was assessed by Student's t test (*p < 0.001).

 


Figure 9
View larger version (34K):
[in this window]
[in a new window]

  		Figure 9. Dynamics of GAK association with the TGN in live cells analyzed by photobleaching and photoactivation. (A) The kinetics of wt and mut1/2 GFP-GAK recruitment to the TGN in transiently transfected HeLa cells were analyzed by fluorescence recovery after photobleaching (FRAP) as described in Materials and Methods and in a previous publication (Kametaka et al., 2005Go). (B) The kinetics of dissociation from the TGN of wt and mut1/2 GAK tagged with photoactivatable GFP (PAGFP) in transiently transfected HeLa cells were analyzed after photoactivation of the TGN-associated population of PAGFP-GAK as described in Materials and Methods. Data points represent the mean ± SD from five experiments.

 


View this table:
[in this window]
[in a new window]

  		Table 4. Kinetics of membrane association and dissociation of GFP-GAK fusion proteins

 
Binding of GAK to AP1 Is Required for its Function in Lysosomal Enzyme Sorting
Depletion of GAK has been recently shown to affect lysosomal sorting both in altering the localization of MPRs (Lee et al., 2005Go) and delaying the maturation of cathepsin D (Zhang et al., 2005Go). This maturation involves successive proteolytic cleavage from precursor (pro-catD) to intermediate (int-catD) and, eventually, to mature cathepsin D (m-catD). Immunoblot analysis showed that depletion of GAK or clathrin increased the proportion of pro-catD and int-catD, while decreasing the proportion of m-catD, at steady state (Figure 10A). Consistent with these results, pulse-chase analysis showed that depletion of GAK or clathrin for 3 d delayed the processing of pro-catD to int-catD and then to m-catD (Figure 10, B and C). Importantly, transfection with RNAi-resistant GFP-GAK (wt) corrected the cathepsin D maturation defect in GAK-depleted cells, whereas RNAi-resistant GFP-GAK with mutations in the FGPL and FGEF sequences (mut1/2) or kinase-deficient GAK K69A (Zhang et al., 2005Go; Figure 11A) did not rescue cathepsin D maturation, as determined by immunoblot analysis (Figure 11B). These results indicated that both the ability of GAK to interact with AP1 and its kinase activity are important for the role of GAK in the lysosomal sorting of cathepsin D.


Figure 10
View larger version (42K):
[in this window]
[in a new window]

  		Figure 10. The GAK-AP1 interaction is important for lysosomal enzyme sorting. (A) Immunoblot (IB) analysis of pro- (pro-catD), intermediate (int-catD), and mature cathepsin D (m-catD) in untreated HeLa cells (control) and HeLa cells treated with siRNAs directed to GAK or clathrin heavy chain (CHC). Cell lysates were resolved by SDS-PAGE followed by immunoblotting with antibodies to cathepsin D, GAK, and CHC. (B) Pulse-chase analysis of cathepsin D maturation in [35S]methionine-cysteine–labeled HeLa cells treated with control, GAK, and CHC siRNAs as described in Materials and Methods. (C) Densitometric quantification of the results from B for the different cathepsin D species. The percentage of each species at each time point was normalized by the corresponding number of methionine and cysteine residues.

 


Figure 11
View larger version (54K):
[in this window]
[in a new window]

  		Figure 11. Both the AP1-{gamma}1-ear–binding site and kinase activity of GAK are required to reverse the defect in cathepsin D maturation in GAK-depleted cells. (A) In vitro kinase assay using GFP alone and GFP fused to wild-type (wt), AP1-{gamma}1-ear–binding site mutant (mut1/2) or kinase-deficient (K69A; Zhang et al., 2005Go) GAK. GFP proteins were expressed by transfection in HeLa cells, immunoprecipitated with anti-GFP antibody, and analyzed using an in vitro kinase assay with myelin basic protein (MBP) as a substrate, as described in Materials and Methods. Autophosphorylated GFP-GAK proteins (asterisk) and phosphorylated MBP labeled with 32P are indicated in the left panel; the amounts of GFP proteins used for the assay, as detected by immunoblotting with anti-GFP antibody, are shown in the right panel. Notice that mutation of the AP1-{gamma}1-ear–binding site has no effect on kinase activity, whereas mutation of the active site K69 residue decreases GAK autophosphorylation and completely abrogates MBP phosphorylation. (B) HeLa cells depleted of GAK by treatment with siRNA to the 3' UTR were transfected with either wt, mut1/2, or K69A GFP-GAK ORF cDNAs, as described in Materials and Methods, and analyzed by SDS-PAGE and immunoblotting with antibody to cathepsin D (top panel; cathepsin D species are denoted as indicated in the legend to Figure 10). Expression of the GFP-GAK constructs was determined by immunoblotting with antibody to GFP (bottom panel). Notice that both mut1/2 and K69A GAK fail to rescue cathepsin D processing in GAK-depleted cells. Moreover, K69A GAK appears to exert a dominant-negative effect, because it causes a buildup of pro-catD. The apparent molecular masses of markers (in kDa) are indicated on the left.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The results presented here add GAK to the growing list of accessory proteins that interact through canonical motifs with the ear domain of the {gamma} subunit of AP1. The ability of GAK to interact in this manner with AP1 is critical for its recruitment to the TGN (Figures 7Go–9) and for lysosomal enzyme sorting (Figure 11). To our knowledge, these findings represent the first demonstration that canonical motif-ear interactions are functionally relevant for protein sorting in vivo.

Interaction of GAK with the Ear Domains of AP1-{gamma}1 and -{gamma}2
GAK has two contiguous sequences, FGPL (residues 961–964) and FGEF (residues 981–984), that fit the consensus motif, {Psi}G[PDE][{Psi}LM], for {gamma}-ear binding (Mattera et al., 2004Go). Both sequences are preceded by acidic residues (Asp at –2 relative to the Phe defined as position 0), which are typical of known {gamma}-ear–binding sequences (Duncan et al., 2003Go; Mattera et al., 2003Go, 2004Go; Mills et al., 2003Go). Moreover, both sequences occur within a segment of GAK that is predicted to be largely unstructured and therefore accessible for interactions with {gamma}-ear domains. Our experimental analyses bear out these predictions by demonstrating that, indeed, both the FGPL and FGEF sequences mediate binding of GAK to {gamma}-ear domains in vitro (Figures 2GoGo–5). The FGPL sequence does so with lower affinity probably due to the proline residue at +2, which might hinder the ability of the tetrapeptide to adopt a favorable conformation for binding. Despite their marked differences in affinity, both sequences contribute to the overall binding avidity of full-length GAK (Figure 4B).

A noteworthy feature of these interactions is the preference of GAK for the ear domains of the {gamma}1 and {gamma}2 subunit isoforms of AP1 over the GAE domains of the three GGA proteins (Figure 2). Other accessory proteins such as {gamma}-synergin, enthoprotin, and NECAP1, which share sequences fitting the {Psi}G[PDE][{Psi}LM] motif (Table 1), exhibit a similar preference for {gamma}-ears (Mills et al., 2003Go; Mattera et al., 2004Go; Ritter et al., 2004Go). This is in contrast to p56 (Collins et al., 2003Go; Lui et al., 2003Go) and rabaptin-5 (Mattera et al., 2003Go), which bind to {gamma}-ear and GGA-GAE domains with comparable avidities. These distinct patterns of interaction are intriguing because {gamma}-ear and GGA-GAE domains share a similar fold and a conserved motif-binding site (Kent et al., 2002Go; Nogi et al., 2002Go; Miller et al., 2003Go). The selectivity of binding must thus be due to the exact identity of residues within each motif and to subtle differences in the properties of the motif-binding site.

Recruitment of GAK to the TGN Requires Interaction with AP1-{gamma}
Consistent with the direct interaction of GAK with AP1 demonstrated here, these two proteins colocalize to a juxtanuclear region of the cytoplasm that contains the TGN and a subset of endosomes (Greener et al., 2000Go; Umeda et al., 2000Go; Lee et al., 2006Go; Figure 6). For simplicity, we refer to this localization as TGN, although our observations are equally applicable to endosomes, where a population of AP1 is known to be located (Futter et al., 1998Go). Using RNAi-mediated silencing, we found that depletion of AP1 causes a substantial decrease in the association of endogenous GAK with the TGN (Figure 7). Moreover, a transgenic GAK construct with substitutions in both the {gamma}-ear binding FGPL and FGEF motifs shows reduced TGN localization at steady state (Figure 8) and faster exchange between the TGN and the cytosol (Figure 9). These observations indicate that the ability of GAK to interact with AP1 is critical for its localization to the TGN.

Notwithstanding the predominant role of GAK-AP1 interactions, a small amount of GAK associates with the TGN independently of its interaction with AP1 (Figures 8 and 9). This suggests that additional interactions might contribute to GAK recruitment to the TGN. GAK has clathrin-binding motifs (Figure 1), but RNAi-mediated depletion of clathrin had no discernible effect on GAK localization to the TGN (our unpublished observations). A more likely contributor to TGN localization is the tensin/PTEN-homology domain (Figure 1), which binds various phosphoinositides and regulates the dynamics of GAK association with clathrin-coated pits at the plasma membrane (Lee et al., 2006Go).

Requirement of the GAK-AP1 Interaction for Lysosomal Enzyme Sorting
Both AP1 (Meyer et al., 2000Go; Hirst et al., 2003Go, 2005Go) and GAK (Zhang et al., 2005Go) are required for the sorting of the precursor of the acidic hydrolase, cathepsin D, to lysosomes. Using an RNAi-rescue approach, we showed that the ability of GAK to bind to AP1 is critical for the lysosomal sorting of cathepsin D (Figure 11). Thus, even a catalytically active GAK cannot function in lysosomal enzyme sorting unless it is recruited to the TGN by virtue of its canonical motif-ear interaction with AP1. These findings suggest that similar canonical interactions with AP1 might contribute to the recruitment of other accessory proteins to the TGN and to their possible function in protein sorting. Because AP1 and GAK are both components of TGN clathrin coats, it is not surprising that RNAi-mediated depletion of clathrin also causes missorting of cathepsin D (Figure 10). This latter finding is in agreement with observations from a previous study in which clathrin expression was reduced by antisense RNA (Iversen et al., 2003Go).

The involvement of AP1 in the sorting of cathepsin D could depend on its role in the packaging of mannose 6-phosphate receptors into clathrin-coated vesicles at the TGN (Doray et al., 2002Go). However, AP1 has also been proposed to be localized to endosomes (Futter et al., 1998Go) and to play a role in the sorting of mannose 6-phosphate receptors from endosomes to the TGN (Meyer et al., 2000Go). These findings raise the possibility that AP1 could mediate bidirectional transport between the TGN and endosomes. GAK could thus play its role in lysosomal enzyme sorting through interaction with AP1 at the TGN, endosomes, or both.

The effects of GAK depletion on cathepsin D sorting were examined in cells treated for 3 d with synthetic siRNAs, which caused 80–90% decrease in the levels of endogenous GAK. Under these conditions, cathepsin D was missorted even though the Golgi/TGN remained intact. Thus, these early effects were likely directly related to decreased GAK function. Longer treatments may lead to a greater decrease of endogenous GAK but result in varying degrees of Golgi/TGN disruption, as previously reported (Lee et al., 2005Go). Prolonged or more complete GAK depletion therefore leads to drastic alteration of the structure of the Golgi complex, which could be an indirect effect of GAK loss.

How Does GAK Participate in Lysosomal Enzyme Sorting?
Although it is now clear that both the expression of GAK (Lee et al., 2005Go; Zhang et al., 2005Go; this study) and its ability to interact with AP1 (this study) are required for lysosomal enzyme sorting, the exact mechanism by which GAK participates in this process remains to be elucidated. GAK contains a Ser/Thr kinase domain homologous to that of AAK1, an enzyme that phosphorylates the µ2 subunit of AP2 and activates it for recognition of YXXØ-type sorting signals (Conner and Schmid, 2002Go; Ricotta et al., 2002Go; Honing et al., 2005Go). Indeed, GAK has been shown to phosphorylate the µ1 subunit of AP1 in vitro (Umeda et al., 2000Go; however, see also Korolchuk and Banting, 2002Go). By analogy to µ2, GAK-catalyzed phosphorylation of µ1 could activate AP1 for recognition of some sorting signal in the tails of the mannose 6-phosphate receptors (Ghosh et al., 2003Go; Ghosh and Kornfeld, 2004Go). The kinase domains of both GAK and AAK1 are homologous to those of Saccharomyces cerevisiae Ark/Prk proteins. These yeast proteins have been shown to regulate endocytosis and the actin cytoskeleton through phosphorylation of multiple substrates (Smythe and Ayscough, 2003Go). Thus, the role of GAK in lysosomal enzyme sorting could involve phosphorylation of other components of the trafficking machinery or the actin cytoskeleton. Although the key GAK substrate involved in the biosynthetic sorting of cathepsin D remains to be identified, we have shown that that kinase activity of GAK is essential for this process (Figure 11). GAK also contains a J domain homologous to that of auxilin (Greener et al., 2000Go). This domain cooperates with the chaperone Hsc70 to uncoat clathrin-coated vesicles (Greener et al., 2000Go; Umeda et al., 2000Go). It is therefore also possible that GAK functions to remove the coat from TGN- or endosome-derived clathrin-coated vesicles that transport mannose 6-phosphate receptors. The interaction of GAK with AP1 described here thus likely enables the recruitment of both the kinase and uncoating activity of GAK to sites of lysosomal enzyme sorting.


   ACKNOWLEDGMENTS
 
We thank Sean Conner for his kind gift of GFP-AAK1 plasmid, Xiaolin Zhu and Hsin-I Tsai for expert technical assistant, and Wolf Lindwasser for critical review of the manuscript. This research was supported by the Intramural Research Program of the National Institute of Child Health and Human Development, National Institutes of Health.


   Footnotes
 
This article was published online ahead of print in MBC in Press (http://www.molbiolcell.org/cgi/doi/10.1091/mbc.E06-12-1162) on May 30, 2007.

{dagger} Present address: New England Inflammation and Tissue Protection Institute, Bouve College of Health Sciences, 113 Mugar Health Sciences Building, 360 Huntington Avenue, Boston, MA 02115. Back

Address correspondence to: Juan S. Bonifacino (juan{at}helix.nih.gov).


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Ahle, S., and Ungewickell, E. (1990). Auxilin, a newly identified clathrin-associated protein in coated vesicles from bovine brain. J. Cell Biol 111, 19–29.[Abstract/Free Full Text]

Boehm, M., and Bonifacino, J. S. (2002). Genetic analyses of adaptin function from yeast to mammals. Gene 286, 175–186.[CrossRef][Medline]

Bonifacino, J. S. (2004). The GGA proteins: adaptors on the move. Nat. Rev. Mol. Cell Biol 5, 23–32.[CrossRef][Medline]

Collins, B. M., Praefcke, G. J., Robinson, M. S., and Owen, D. J. (2003). Structural basis for binding of accessory proteins by the appendage domain of GGAs. Nat. Struct. Biol 10, 607–613.[CrossRef][Medline]

Conner, S. D., and Schmid, S. L. (2002). Identification of an adaptor-associated kinase, AAK1, as a regulator of clathrin-mediated endocytosis. J. Cell Biol 156, 921–929.[Abstract/Free Full Text]

Doray, B., and Kornfeld, S. (2001). Gamma subunit of the AP-1 adaptor complex binds clathrin: implications for cooperative binding in coated vesicle assembly. Mol. Biol. Cell 12, 1925–1935.[Abstract/Free Full Text]

Doray, B., Bruns, K., Ghosh, P., and Kornfeld, S. (2002). Interaction of the cation-dependent mannose 6-phosphate receptor with GGA proteins. J. Biol. Chem 277, 18477–18482.[Abstract/Free Full Text]

Duncan, M. C., Costaguta, G., and Payne, G. S. (2003). Yeast epsin-related proteins required for Golgi-endosome traffic define a gamma-adaptin ear-binding motif. Nat. Cell Biol 5, 77–81.[CrossRef][Medline]

Fernandez-Chacon, R., Achiriloaie, M., Janz, R., Albanesi, J. P., and Sudhof, T. C. (2000). SCAMP1 function in endocytosis. J. Biol. Chem 275, 12752–12756.[Abstract/Free Full Text]

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.[Abstract/Free Full Text]

Ghosh, P., Dahms, N. M., and Kornfeld, S. (2003). Mannose 6-phosphate receptors: new twists in the tale. Nat. Rev. Mol. Cell Biol 4, 202–212.[CrossRef][Medline]

Ghosh, P., and Kornfeld, S. (2004). The cytoplasmic tail of the cation-independent mannose 6-phosphate receptor contains four binding sites for AP-1. Arch. Biochem. Biophys 426, 225–230.[CrossRef][Medline]

Greener, T., Zhao, X., Nojima, H., Eisenberg, E., and Greene, L. E. (2000). Role of cyclin G-associated kinase in uncoating clathrin-coated vesicles from non-neuronal cells. J. Biol. Chem 275, 1365–1370.[Abstract/Free Full Text]

Hirst, J., Lui, W. W., Bright, N. A., Totty, N., Seaman, M. N., and Robinson, M. S. (2000). A family of proteins with gamma-adaptin and VHS domains that facilitate trafficking between the trans-Golgi network and the vacuole/lysosome. J. Cell Biol 149, 67–80.[Abstract/Free Full Text]

Hirst, J., Motley, A., Harasaki, K., Peak Chew, S. Y., and Robinson, M. S. (2003). EpsinR: an ENTH domain-containing protein that interacts with AP-1. Mol. Biol. Cell 14, 625–641.[Abstract/Free Full Text]

Hirst, J., Borner, G. H., Harbour, M., and Robinson, M. S. (2005). The aftiphilin/p200/gamma-synergin complex. Mol. Biol. Cell 16, 2554–2565.[Abstract/Free Full Text]

Honing, S., Ricotta, D., Krauss, M., Spate, K., Spolaore, B., Motley, A., Robinson, M., Robinson, C., Haucke, V., and Owen, D. J. (2005). Phosphatidylinositol-(4,5)-bisphosphate regulates sorting signal recognition by the clathrin-associated adaptor complex AP2. Mol. Cell 18, 519–531.[CrossRef][Medline]

Horiuchi, H., Lippe, R., McBride, H. M., Rubino, M., Woodman, P., Stenmark, H., Rybin, V., Wilm, M., Ashman, K., Mann, M., and Zerial, M. (1997). A novel Rab5 GDP/GTP exchange factor complexed to Rabaptin-5 links nucleotide exchange to effector recruitment and function. Cell 90, 1149–1159.[CrossRef][Medline]

Iversen, T. G., Skretting, G., van Deurs, B., and Sandvig, K. (2003). Clathrin-coated pits with long, dynamin-wrapped necks upon expression of a clathrin anti