AAKI, a Novel Kinase that Regulates Cargo Selection for Clatherin-Mediated Endocytosis

Sandra Louise Schmid, Ph.D.

The Scripps Research Institute

My lab has been interested, since its beginnings, in trying to understand the process of receptor-mediated endocytosis. It involves high-affinity binding of macromolecules, like LDL receptors or even virus particles, to specific receptors on the cell surface. These receptors contain sorting motifs, different flavors of sorting motifs, as you will see later in the talk, which are recognized by coat proteins, that form the basis for an endocytic machinery. The coat proteins assemble into a polygonal lattice, which forms curved structures that help to deform and invaginate the membrane forming coated pits. Finally, membrane fission, which is regulated by the GTPase dynamin that assembles into a collar-like structure encircling the neck of deeply invaginated coated pits, releases a transport vesicle that carries concentrated cargo receptor and ligand complexes into the cell.

The major coat proteins have been identified for some time now. They are clathrin triskelions and adapter protein-2 complexes (AP2s). You can think of clathrin as the “brauns” of this machinery. It self-assembles into a polygonal lattice and by so doing drives the membrane invagination. The AP2s are the “brains” of the machinery. They recognize cargo molecules, and trigger clathrin assembly thereby coordinating vesicle formation and cargo recruitment.

In order to study how coat assembly is regulated in the cell and to identify what other factors might be involved in mediating this process of receptor-mediated endocytosis, I began, as a postdoctoral fellow with the



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 90
The Markey Scholars Conference: Proceedings AAKI, a Novel Kinase that Regulates Cargo Selection for Clatherin-Mediated Endocytosis Sandra Louise Schmid, Ph.D. The Scripps Research Institute My lab has been interested, since its beginnings, in trying to understand the process of receptor-mediated endocytosis. It involves high-affinity binding of macromolecules, like LDL receptors or even virus particles, to specific receptors on the cell surface. These receptors contain sorting motifs, different flavors of sorting motifs, as you will see later in the talk, which are recognized by coat proteins, that form the basis for an endocytic machinery. The coat proteins assemble into a polygonal lattice, which forms curved structures that help to deform and invaginate the membrane forming coated pits. Finally, membrane fission, which is regulated by the GTPase dynamin that assembles into a collar-like structure encircling the neck of deeply invaginated coated pits, releases a transport vesicle that carries concentrated cargo receptor and ligand complexes into the cell. The major coat proteins have been identified for some time now. They are clathrin triskelions and adapter protein-2 complexes (AP2s). You can think of clathrin as the “brauns” of this machinery. It self-assembles into a polygonal lattice and by so doing drives the membrane invagination. The AP2s are the “brains” of the machinery. They recognize cargo molecules, and trigger clathrin assembly thereby coordinating vesicle formation and cargo recruitment. In order to study how coat assembly is regulated in the cell and to identify what other factors might be involved in mediating this process of receptor-mediated endocytosis, I began, as a postdoctoral fellow with the

OCR for page 90
The Markey Scholars Conference: Proceedings support of the Markey Fellowship, to develop new cell free assays, using perforated A431 cells that faithfully reconstitute these events. We have also developed stage specific biochemical assays that allow us to detect sequential intermediates in endocytic-coated vesicle formation. These assays follow the receptor-mediated endocytosis of transferrin, which is biotinylated with a cleavable disulfide bond (referred to as BSSTfn). The “sequestration” of receptor-bound BSSTfn into an intermediate in vesicle formation, called a constricted coated pit, can be detected when the ligand becomes inaccessible to avidin. However, because constricted coated pits remain accessible to small molecules through a narrow opening at their necks that connects them to the plasma membrane, the BSSTfn sequestered in them can be cleaved by the small membrane-impermeant, reducing agent, MesNa. Thus, late events in vesicle formation, namely the release of a sealed vesicle from the plasma membrane are detected when the BSSTfn becomes resistant to MesNa. Finally, we can focus on measuring the earliest events in vesicle formation, by supplementing assays performed in the presence of limiting concentrations of cytosol with purified coat proteins. The coat protein-stimulated sequestration of BSSTfn into constricted coated pits reflects de novo coat assembly. Many years of work helped us to biochemically characterize these stages of vesicle formation. We found that while the coat proteins, clathrin and AP2, are necessary for coated vesicle formation, they are not sufficient. For example, intermediate and late events in this process are regulated by the GTPase dynamin. Most of my work in the past 6 years has focused on understanding the function in vivo and in vitro of this GTPase, although it is not subject of the talk today. These late events also require ATP hydrolysis, by an as-yet-unidentified ATPase, as well as numerous accessory proteins that have been implicated by virtue of their interaction with dynamin and with the coat proteins. Early events in CCV formation require the lipid, phosphatidylinositol-4,5 -bisphosphate (PI-4,5-P2), which provides at least part of the targeting signal for recruitment of AP2 complexes and other endocytic accessory proteins, including dynamin to the plasma membrane. Early events also require ATP hydrolysis and again, we have not identified the ATPase. One candidate, however, will be the subject of my talk today. The role of numerous accessory proteins that interact with adapters and clathrin are not known. I will focus on the regulation of endocytic vesicle formation. An extremely talented postdoc, Sean Conner, did all of the work I am going to tell you about today. Because the adaptors are a key factor in initiating coat assembly—“the brains of the operation”—we focused our attention on these. The AP2 complex consists of four subunits, called alpha, beta,

OCR for page 90
The Markey Scholars Conference: Proceedings sigma, and mu adaptins. The alpha adaptin is involved in targeting to the plasma membrane, through interactions with both proteins and PI-4,5-P2. The beta adaptin, particularly the hinge region, binds and triggers clathrin assembly. The N-terminal appendage domains of both the alpha and beta adaptins (also called “ear” domains as the AP2 complex looks like a Mickey mouse head with 2 alpha and beta appendage domain “ears”) also bind a number of accessory molecules whose exact function in endocytosis has yet to be established. The mu subunit recognizes specific sorting motifs on cargo molecules, and in particular has been shown to recognize tyrosine-based sorting motifs such as those found in the transferrin receptor. To identify new factors that might regulate AP2 complex function we decided to “go fishing” using the appendage domain of the alpha adaptin fused to GST as “bait.” We chose the phage-display approach and cast our immobilized GST-alpha ear “fishing pole” into a phage pool. This approach allowed us to control the quality of our bait and the biochemical conditions in which we measure the interaction. Through multiple rounds of binding, elution and phage amplification we could increase the signal:noise ratio to identify specific interactions and identify less abundant molecules. All of the proteins known to interact with clathrin adaptors are very abundant-scaffolding proteins, identified from cell lysates through protein pull-down approaches. As we were interested in regulatory molecules, we took this approach to look for proteins that might be less abundant. In addition to all the known interacting proteins, we found that 30 percent of the clones in our selected phage samples encoded a C-terminal fragment from a novel member of the serine/threonine kinase family. This multidomain protein encodes an N-terminal serine/threonine kinase domain, a central glutamine, proline and alanine-rich domain, and the C-terminal, alpha interacting domain (AID). The AID has several DPF and NPF motifs that are known to bind to the alpha ear. We were able to verify these protein interactions in vitro showing that the C-terminal region of what we now call adapter-associated kinase one or AAK1 indeed interacts directly with the alpha appendage domain. We expressed the intact AAK1 using the baculovirus expression system and confirmed that it, like the AID alone, is able to pull down the alpha adapter from rat brain cytosol. We also made antibodies to AAK1 and showed that the protein copurifies with clathrin-coated vesicles from bovine brain and from rat liver. AAK1, together with the coat proteins, can be extracted from the membrane by Tris-buffer. When the extracted coat proteins are subsequently purified by gel filtration, we find that AAK1 copurifies with APs and not with clathrin. These studies provided the functional basis for the name, AAK1.

OCR for page 90
The Markey Scholars Conference: Proceedings Immunolocalization studies in neurons established that AAK1 is enriched at sites of endocytosis, which are identified by the uptake of rhodamine-labelled dextran into activated hippocampal neurons. AAK1 colocalizes with these endocytic hot spots together with AP2 and dynamin. Endogenous AAK1 also colocalizes with the AP2 adapters on the plasma membrane in non-neuronal cells. Based on these results, we felt very comfortable that the kinase we had identified was relevant to the process of endocytosis. What, then, is the function of AAK1? We first approached this question by asking, what are its substrates? We found in in vitro assays that although AAK1 binds the alpha adaptin, it specifically phosphorylates the mu subunit of the AP2 complex. In collaboration with Stefan Höning and Doris Ricotta (University of Göttingen, Germany) we identified threonine 156 as the AAK1 phosphorylation site on the mu subunit. A former postdoc of mine, Elizabeth Smythe, now leading her own lab in Shefield, had recently shown that phosphorylation of threonine 156 on the mu subunit was essential for endocytosis both in vivo and in vitro (Olusanya, O. et al. 2001. Curr Biol. 11:896-900), providing further evidence that AAK1 was indeed a functionally relevant kinase for endocytosis. What are the consequences of this phosphorylation of mu? As I mentioned, the mu subunit recognizes cargo-sorting motifs on cargo molecules. Our collaborators, Doris and Stefan, therefore, looked at the affinity of AP2s for these sorting motif-containing peptides, using a BiaCore to measure surface plasmon resonance. AP2 complexes, isolated from bovine brain extracts, bind peptides containing these tyrosine-based sorting motifs with a certain affinity (116 nM). If the AP2s, which copurify with an endogenous kinase, are first treated with a phosphatase, then the binding affinity is decreased approximately five-fold. Conversely, if we incubate this preparation with ATP and allow the endogenous kinase to phophorylate mu, we see an approximate two-fold increase in binding affinity. Strikingly, incubation with purified AAK1 and ATP enhances the binding affinity of AP2 for tyrosine motif-containing peptides by 25-fold compared to phosphatase treated AP2s. From this we conclude that AAK1 phosphorylates mu on Thr156 to regulate its affinity for cargo molecules containing tyrosine-based sorting motifs. To examine the effects of AAK1 on endocytic coated vesicle formation, we took advantage of our perforated cell system and our ability to selectively measure early events in coated-vesicle formation. In an assay in which we supplement incubations performed under limiting concentrations of cytosol with purified AP2s, adding increasing concentrations of AP2 leads to increased sequestration of BSSTfn into constricted coated pits—a measure of early events in endocytosis. When these reactions are supplemented with AAK1, the AP2-stimulated early events are signifi-

OCR for page 90
The Markey Scholars Conference: Proceedings cantly inhibited. If we first inactivate the AAK1 by treating it with an irreversible kinase inhibitor, FSBA, it becomes a much less potent inhibitor of the AP2-stimulated signal. Thus, excess AAK1 appears to inhibit the sequestration of receptor-bound BSSTfn into newly formed constricted coated pits. I had previously described the work of Liz Smythe who had shown that phosphorylation of mu was necessary for endocytosis. This might seem paradoxical with our finding that adding excess AAK1 inhibits endocytosis. I will try to resolve this paradox with a model at the end of my talk. First, however, we wanted to see whether there were any other substrates that we could detect in these more complex cell extracts. When AAK1 is incubated with either cell membranes or cytosol, the principal product, other than autophosphorylation of AAK1, is the mu subunit. Thus, although we cannot rule out that there are other substrates involved, we believe that the effects on endocytosis are due primarily to the phosphorylation of mu. Given the apparent paradox of needing AAK1-mediated phosphorylation of mu for early events in endocytosis in a perforated system, yet finding that excess active AAK1 inhibits endocytosis, we decided to look in vivo to see what the effects of AAK1 overexpression might be. To this end we constructed adenovirus expressing wild-type and mutant AAK1 kinases under the control of a tetracycline regulatable promoter so that we could control levels of expression. When either the wild type or kinase-inactive AAK1 were overexpressed using this adenoviral system, Tfn-receptor-mediated endocytosis was inhibited. The kinase-dead mutant was a less potent inhibitor than the WT AAK1, thus these in vivo data closely matched the results we saw in our in vitro, perforated cell system. Inhibition was not due to sequestration of some binding partner of AAK1, or to rampant kinase activity, because when we overexpressed truncation products of AAK1 lacking the AID or the AID alone, these constructs do not inhibit Tfn endocytosis. Further characterization revealed that overexpressing either the wildtype or kinase-inactive AAK1 mutants, displaced AP2s from the coated pits. Instead of the normal punctate distribution of AP2 in coated pits, as detected by immunolocalization, the AP2 staining becomes diffuse. Sub-cellular fractionation studies established that although the AP2 complexes were no longer clustered into coated pits, they remained associated with the plasma membrane and the relative distribution between the membrane pool and the cytosolic pool was unchanged by overexpressing either the wild-type or mutant kinase. Surprisingly, while AP2 distribution in coated pits was disrupted, the distribution of clathrin was not affected. This finding goes against that expected by the dogma that AP2s are responsible for triggering clathrin

OCR for page 90
The Markey Scholars Conference: Proceedings assembly. In cells overexpressing AAK1, clathrin appears to assemble in the absence of AP2 clustering. But are these coated pits still functional? The mu subunit recognizes tyrosine-based sorting motifs and phosphorylation by AAK1 enhances the affinity for those sorting motifs 25-fold, therefore we decided to look at another cargo molecule, which has a different kind of sorting motif. Endocytosis of EGF receptors is known to occur through clathrin-coated pits. Indeed, double label immunoelectron microscopy shows that EGF receptors are found in the same coated pits as transferrin receptors. However, in contrast to the Tfn receptor, endocytosis of EGF receptors is ligand dependent and not affected by mutations in mu that disrupt their ability to bind tyrosine-based sorting motifs. We therefore looked at EGF uptake and found that it was unaffected by AAK1 overexpression. As a positive control, we overexpressed a dominant negative dynamin mutant that potently inhibits both EGF and Tfn receptor uptake. These findings were confirmed microscopically using fluorescently-labeled EGF or Tfn as endocytic tracers, in which cells overexpressing AAK1 fail to efficiently take up fluorescent-Tfn molecules, but continue to take up fluoresent-EGF molecules. Together these data suggest that AAK1 can selectively regulate clathrin-mediated endocytosis of different cargo molecules. We heard from Dan Madison yesterday another striking example of cargo selective regulation of endocytosis as he described the regulation of postsynaptic NMDA receptor endocytic trafficking by calcium-dependent phosphatases and kinases that function to control the levels of AMPA receptor expression and responses to neurotransmitters. Thus, a new concept emerging from these studies is the existence of cargo-selected regulation of uptake into the cell. The AAK1 kinase is one of the first mechanistically described examples of this regulation. Importantly, AAK1 appears to regulate endocytosis of transferrin receptors, which are nutrient receptors classically defined as being constitutively endocytosed. Thus, the “constitutive” endocytic pathway is, in fact, subject to regulation. Our current working model to explain all of the observations we have, not all of which I have described to you, is that the effective concentration of cargo molecules into coated pits requires cycles of phosphorylation and dephosphorylation of the mu subunit, mediated by AAK1. The recently published three-dimensional structure of the AP2 core (Collins, B.M., et al. 2002. Cell 109:523-35) suggests a mechanism by which phosphorylation at Thr156 in a linker region of mu subunit induces a conformational change to expose the otherwise buried tyrosine-motif binding site in the molecule. In this way, phosphorylation is thought to be required for high-affinity binding to cargo molecules at the surface. Once bound to cargo molecules, subsequent dephosphorylation of AP2 com-

OCR for page 90
The Markey Scholars Conference: Proceedings plexes may be required to allow the AP2s to self-associate and/or to trigger clathrin assembly and vesicle formation. We propose then that the kinase inactive mutant blocks cargo recruitment at an early stage in coated vesicle formation, and that overexpression of wildtype AAK1 blocks at a subsequent step. Our future efforts will be directed towards testing this model and elucidating the role of AAK1 in regulating clathrin-mediated endocytosis. NOTE: The following papers report some of the studies described here, Conner, S. D., and S. L. Schmid. 2002. J Cell Biol. 156:921-9; and Ricotta, D. et al. 2002. J Cell Biol. 156:791-5. Conner, S. D. and S. L. Schmid. 2003. J Cell Biol. 162:773-80.