the identification and characterization of proteins required for endocytosis in the budding yeast saccharomyces cerevisiae

Trang 1

THE IDENTIFICATION AND CHARACTERIZATION OF PROTEINS REQUIRED FOR ENDOCYTOSIS IN THE BUDDING YEAST

SACCHAROMYCES CEREVISIAE

by

Jonathan D Shaw

Trang 2

UMI Number: 3197229 Copyright 2005 by Shaw, Jonathan D All rights reserved INFORMATION TO USERS

The quality of this reproduction is dependent upon the quality of the copy submitted Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed-through, substandard margins, and improper alignment can adversely affect reproduction

In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted Also, if unauthorized copyright material had to be removed, a note will indicate the deletion ® UMI UMI Microform 3197229 Copyright 2006 by ProQuest Information and Learning Company

All rights reserved This microform edition is protected against unauthorized copying under Title 17, United States Code

ProQuest Information and Learning Company 300 North Zeeb Road

Trang 3

ABSTRACT

All eukaryotic cells are surrounded by a plasma membrane that segregates the interior of the cell from the extracellular environment Transport to and from this membrane regulates homeostasis of the membrane itself, allows the import of material from the extracellular space, and provides a mechanism for communication with the outer environment The process of endocytosis describes the internalization and subsequent transport of plasma membrane components and extracellular material Endocytosis requires the coordinated activity of many protein factors, and the specific mechanisms employed by the cell for internalization remain only partially understood

The studies presented in this work focus on identifying novel proteins involved in endocytosis and characterizing their role in membrane trafficking An in vitro

endocytosis assay was adapted for use with yeast cytosol, in order to provide a

Trang 5

ACKNOWLEDGMENTS

First and foremost, I thank my thesis advisor, Beverly Wendland Beverly has been a source of unwavering support throughout my graduate career, always encouraging when times were good and optimistic when times were bad She was integral to every aspect of this thesis and helped develop the insight and direction of each of my projects Without Beverly, this thesis, and my graduate career in general, would have been a far less rewarding and less successful experience

Many of the faculty here at Johns Hopkins have been an immense help in guiding this work and my graduate and professional career Trina Schroer and Doug Koshland have played critical roles in determining the direction of my studies, helping see this thesis through from beginning to end, and providing advice on a variety of subjects I’d like to say thanks to Allen Shearn, Andy Fire and Mark Van Doren for their mentorship Also, thank you to the Howard Hughes Medical Institute for generous funding and support

I would also like to thank all of the colleagues I’ve been fortunate enough to work with over the years In particular, Jin Packard, Claudio Aguilar, and Lymarie

Maldonado-Baez each made crucial contributions to the projects described in this thesis Furthermore, Jennifer Baggett, Bill McCormick, and Sarah Barker have become great friends and succeeded in helping to make life in the Wendland Lab fun and entertaining

Trang 6

TABLE OF CONTENTS

ABSTRACT ch HH KT Ki TK kg tk cu ii

Acknowledgmenfs ch HH KĐT TK nen nh nh nà nà và iv

Table UÁ6/ - nh <a V

Index of Figures and Tables CS HH nh nhe y Vii CHAPTER 1: Introducfion ng nh nh heu 1 CHAPTER 2: Developing an in vitro endocytosis assay for use with budding yeast 20 ˆ uc ồớỐ 21 INtrOductiONn 0 ccc cece cece cece eee nent eee ee ee ee ee ee eens ease este ease eeeaeneneaeaegs 21 1A ese 24 IMESO) (0i PPaẳẳđiiềiiáaidiaaaaaiiảảảäảäậậaả 48 AcknowledgmentS cu H HH KH TH KT nn n Tá nà kh nàn 51 lÚb 91 h8 (/(00i099 TH 51 CHAPTER 3: PtdIns(3,5)P, is required for delivery of endocytic cargo into the

Trang 7

CHAPTER 4: An interaction between the yeast epsin Entl and GTPase activating proteins for Cdc42 is required for cell growth and polarity ADSUIACt ƠƠƠỎồồ enna eee enna eens Eee e EEE ED EE ED EES EEE EE DEES REESE DE EERE E EES IntrOdUCtiON - Ăn ng ng HT nh kh KH ni Kon ni và |C 110 aDŨDŨDIỊIidiiiaiiiiiddẳdiiiiiÀiiúi 100 II T0ìL`)11:HHaẳầđầaẳaiiiiidiiiẳi 127 AcknowledgmenS -.- con HH HH HH nh nh n nu n nh hà 134

Supplementary Data eẮe 135

Materials and Methods - ch nh xu 140 CHAPTER Š: Concluding remarks -.-.- cà sec 146

G3." =a 157

Trang 8

Figure 1-1 Figure 1-2 Figure 2-1 Figure 2-2 Figure 2-3 Figure 2-4 Figure 2-5 Figure 2-6 Figure 2-7 Figure 2-8 Figure 3-1 Figure 3-2 Figure 3-3 Figure 3-4 Figure 3-5 Figure 3-6 Figure 4-1 Figure 4-2

LIST OF FIGURES AND TABLES

Membrane trafficking pathWwayS HH nh nh nu 4 Clathrin triskelions and ÏaffC€S - co c2 SH nh ưa 8 Establishing the ¿1 Vif70 ASSâY dc on TH HH KH nh nh nà nh ng 26 Identifying components required for internalization in the assay 30 Optimizing the aSSây c cà KH nà nh kg 33 Testing the parameters Of the ASSây c HH SH nha 34 Dynamin In the aSSây HH HH nh nh 37 Yeast cytosol causes a reduction in ATP levels in the assay 40 Combinations of yeast and mammalian cytosol demonstrate an

inhibitory effecf Of yeaSf CYfOSỌ HH HH ng HH nh vi, 43 Non-protein factors support the internalization measured with yeast

À4i0U) 0 H ((ÁCaẦẢẦẢẦẦẢẢ 46 The fab/-20 allele contains a mutation within the lipid kinase domain 65 PtdIns(3,5)P, levels are severely reduced, but detectable, in fab/-20 67 Vacuolar morphology and sorting defects in ƒ#b7-20 - 70 Vacuolar degradation of Ste3 Is delayed in ƒ@ư7-20 73 FM4-64 transport from endosomes to the vacuole membrane is delayed in #ab1-20 and ƒab1A -.- SH SH HH KH nh KH tk cự 76 Ste3 is mislocalized to the vacuolar membrane in fab/ mutants 79

tác Ắ= 97

The ENTH domain ¡s required for cell viability -. -«- 101

Trang 9

Figure 4-3 The Y100R and T104A mutations define an essential face within the ENTẨH domain - co non SH HH ki kh Figure 4-4 The ENTH domain binds to Cdc42 GA PS c se sẰ Figure 4-5 The ENTH domain does not affect GAP activity of the Cdc42 GAPs Figure 4-6 Budding and mating projection formation are defective in ent/-/ Figure 4-7 Cell integrity is compromised In ¿wfÏ-Ï cv Figure 4-8 The actin cytoskeleton is disrupted in ewfÏ-Ï sex Figure 4-9 Endocytosis defeCtS I1 @ffÏ-Ï HH HH HH ke kh km se Figure 4-S1 Abpl and Bni1 localize normally in enf]-Ï e2

Figure 4-S2_ Entl protein levels are normal in ent]-J and ent] °° ss

Figure 4-S3 GFP-ent]-/ labeling ccecccscecneneeeeenea seen ee eeeeenenenenensaenes Figure 4-S4 Cdc42 overexpression in @HfÏ-Ï cv ng nh kh nh rà

Trang 10

CHAPTER 1

Trang 11

One of the first crucial steps in the evolution of life was a primordial organism’s ability to establish a defined domain, chemically distinct from its surrounding

environment Although the details of the origins of life remain a mystery, the

proliferation of single-celled organisms was mediated by their ability to create a barrier between themselves and the outer world This barrier was provided by a plasma membrane, made up of lipids and proteins, which prevented diffusion of complex

molecules into and out of the cell As these organisms evolved, and eukaryotes diverged from other early bacteria, they developed intracellular membrane-bound compartments, surrounded by the same lipid bilayer as found on the surface of the cell While these membrane-bound organelles gave them a great deal of flexibility in partitioning cellular functions, cells also needed to maintain and regulate the organelles in order to survive and divide Specifically, cells needed to add to and retrieve components from these membranes, since proteins and lipids were manufactured only in certain specific regions of the cell and required transportation to the organelles and the plasma membrane to replenish and replace older components They solved this problem by developing mechanisms to transport small membrane-encapsulated vesicles throughout the cell Over eons of evolution, as subcellular structures specialized and diversified, membrane trafficking processes similarly grew in type and complexity

Trang 12

providing the first direct observation of the secretory pathway Subsequent studies recognized additional pathways, together responsible for transporting lipids, proteins, and cargo between intracellular organelles, and into and out of the cell via the plasma

membrane (reviewed in (Mellman and Warren 2000)) As the tools and techniques of biology evolved, so did the study of membrane trafficking While the more classical fields of biochemistry and genetics continued to contribute to our understanding of how proteins and lipids are sorted and trafficked between myriad intracellular locales, advances in electron and light microscopy provided higher resolution imaging of subcellular structures, ultracentrifugation allowed enzymatic characterization of distinct membrane-bound organelles, and novel molecular biology techniques led to the

identification of genes and proteins responsible for functions that had heretofore only been correlated with biochemical activities or genetic phenotypes (Schekman 2004) The field of membrane trafficking now touches on almost every aspect of cell biology and encompasses a wide variety of disciplines

Membrane Trafficking Pathways

A number of distinct routes of membrane traffic transport components to their final destination (Fig 1-1) The secretory pathway represents the starting point for most membrane trafficking cargo, transporting proteins and lipds from their sites of synthesis in the endoplasmic reticulum (ER), through the Golgi network, and out to the plasma membrane ER and Golgi resident proteins are identified and selectively retained within their appropriate organelles In the late Golgi and the trans-Golgi network (TGN), proteins bound for the lysosome are diverted from the secretory pathway and instead are

Trang 13

Plasma Membrane <— ý O ers @ e e ` Multivesicular Body Gal Early mndosomes (MVB) ah ⁄

FIGURE 1-1 Membrane trafficking pathways

Proteins synthesized in the ER are transported through the ER-Golgi intermediate compartment (ERGIC) ER resident proteins (dark blue) are mostly maintained within the ER, while secretory (purple) and lysosomal (orange) proteins continue through the Golgi to the trans-Golgi network (TGN) Secretory proteins are delivered to the plasma membrane, where endocytic cargo (red and light blue) are internalized via vesicles to early endosomes Some endocytic cargo can be recycled to the plasma membrane, but the majority is transported with lysosomal proteins to the late endosome or multivesicular body (MVB) The MVB forms interior vesicles which are delivered to the lysosome upon fusion of the limiting membranes, where endocytic cargo are degraded and resident lysosomal proteins are activated for proteolysis

Trang 14

transported to endosomes Early endosomes mature into late endosomes, which internalize their limiting membrane to form multi-vesicular bodies (MVBs) These MVBs then fuse with the lysosome, delivering their cargo into the lumen of this

proteolytic compartment for degradation Resident lysosomal proteins are either retained on the outer membrane of the MVB or delivered to the lumen, where they require

maturation by proteolysis in order to be activated Together, these membrane trafficking pathways deliver biosynthetic cargo to their sites of activity within the cell However, membrane trafficking pathways are not solely utilized by biosynthetic cargo

Retrograde trafficking pathways provide mechanisms to maintain the normal subcellular localization of proteins and connect most trafficking destinations Cells must recognize and retrieve incorrectly targeted proteins, such as ER resident proteins that escape and get transported to the Golgi Further, as biosynthetic cargo is delivered to various sites within the cell, older components must be retrieved and turned over One of the most important retrograde membrane trafficking pathways provides recovery of plasma membrane components Endocytosis describes the process of internalization from the plasma membrane, including the import of extracellular material Typically, endocytosed cargo is delivered to early endosomes, and from there sorted to its final destination While some components are recycled back to the plasma membrane or targeted to specific intracellular destinations, most internalized material is transported through late endosomes and delivered to the lysosome for degradation

Endocytosis can be further divided into specific types of internalization of extracellular material Phagocytosis, or “cell eating”, describes the uptake of large particles via filopodia-like membranous extensions that engulf and internalize material

Trang 15

that can be almost as large as the cell itself This type of internalization is most often observed in specialized cells dedicated to the uptake and removal of large particles or invading organisms, such as macrophages and neutrophils in the immune system (Watts and Marsh 1992) Pinocytosis, or “cell drinking”, involves the uptake of extracellular solutes and signaling molecules While the term “endocytosis” technically refers to all types of internalization, conventionally it is used to describe pinocytosis Examples of this type of uptake include caveolae-dependent internalization and macropinocytosis (Conner and Schmid 2003; Nichols 2003) However, by far the most-studied and best- understood form of endocytosis is clathrin-dependent internalization

Clathrin-dependent Endocytosis

The importance of clathrin-dependent endocytosis was first elucidated through studies of patients with the inherited disease familial hypercholesterolemia (FH), which causes high levels of cholesterol in the blood and early heart attacks in adults

Trang 16

Subsequent studies established clathrin as the prototypical coat protein involved in membrane traffic (reviewed in (Kirchhausen 2000; Mousavi et al 2004)) The clathrin assembly unit is an oligomeric complex consisting of three heavy chains and three light chains The clathrin heavy chain can be divided into three distinct domains, the N- terminal globular region, a central curved “knee”, and the C-terminal hub region Extended areas known as proximal and distal legs connect these regions Light chains interact with heavy chains at the proximal leg, and are thought to be important for subsequent assembly into higher order lattices (Ybe et al 1998) Three of these clathrin heavy/light chain dimers further interact through their hub domains (Liu et al 1995), forming a three-legged oligomer that is called a triskelion for its distinctive shape (Fig 1- 2A) Clathrin triskelions assemble into higher-order lattice structures, which can be found in either planar or spherical forms (Fig1-2B) Planar structures are made up of hexagonal arrays, while the spherical coats contain both pentagonal and hexagonal arrays (Heuser 1980) These spherical lattices coat nascent vesicles as they bud from the membrane Soon after the budding process, as vesicles are transported away from the membrane, the clathrin coat is disassembled and vesicles are moved on to fuse with their target

membrane While most studies have analyzed clathrin-dependent vesicle formation and internalization at the plasma membrane, clathrin coats are also important in vesicle budding from the TGN

Clathrin-mediated endocytosis plays an important role in maintaining normal cell physiology, and defects in endocytic processes can lead to health problems and disease The internalization of extracellular material such as iron, which is transported in the bloodstream bound to the protein transferrin, provides a mechanism for cells to import

Trang 17

A Globular domain 4 Light chain Proximal leg

FIGURE 1-2 Clathrin triskelions and lattices

(A) A clathrin triskelion is made up of three legs, each consisting of a heavy chain and a light chain Heavy chains contain three domains (hub, knee, and globular), each

Trang 18

biomolecules that they otherwise would be unable to produce (Ponka and Lok 1999) Loss or reduction of LDL uptake as described above is a major cause of inherited high cholesterol levels, leading to atherosclerosis In addition to the uptake of extracellular nutrients, endocytosis also helps to moderate extracellular signals from molecules such as hormones For example, the epidermal growth factor (EGF) receptor binds to EGF and initiates a signal cascade that induces cell proliferation This signal is turned off by the internalization and degradation of the activated EGF receptor via endocytosis In addition to the failure to internalize important biomolecules, endocytosis defects can also lead to overactive receptor signaling Unmitigated EGF signaling can lead to overproliferation of epidermal cells, and has been linked to the onset of cancer (Floyd and De Camilli

1998; Yarden 2001) A clearer understanding of the mechanisms underlying endocytosis may help to overcome health problems that arise due to defective internalization of plasma membrane proteins and lipids

Endocytosis in Yeast

The budding yeast Saccharomyces cerevisiae is a single-celled eukaryotic organism that utilizes many of the same membrane trafficking processes as mammalian cells Yeast and mammalian endocytosis are remarkably similar, employing many of the same pathways and components (Geli and Riezman 1998) Additionally, yeast cells are highly amenable to both classical and molecular genetics, allowing for experimental approaches otherwise unavailable or impractical in mammalian cells Studies in yeast have provided many novel insights into the mechanisms of membrane trafficking in general, and endocytosis in particular (Munn 2001; Shaw et al 2001)

Trang 19

Endocytosed cargo in yeast follows a very similar itinerary to that of mammalian cells Internalized proteins, lipids, and extracellular material are sorted in early

endosomes, transported to late endosomes, and delivered to the vacuole (the yeast lysosome) for degradation Many of the components that perform the underlying

transport functions are conserved between yeast and mammals, and studies in yeast have identified many new proteins and interactions that have subsequently been verified in mammalian cells For example, the role of the actin cytoskeleton in endocytosis was rapidly recognized in yeast, but was not appreciated in mammalian cells until more recently (Wendland et al 1998; Yarar et al 2005) However, many questions regarding endocytosis remain unanswered, and yeast continues to be an excellent system with which to approach these problems Before addressing current research, I first review what is known about the endocytic pathway in yeast

Cargo selection and internalization signals

Trang 20

appended to target proteins through lysines in their amino acid sequence Multiple copies of ubiquitin (polyubiquitination) often target a protein for degradation in the proteasome (Ciechanover 1998) However, monoubiquitination of proteins like the a-factor receptor Ste2 can trigger internalization without targeting the protein for proteolysis in the

proteasome Experiments establishing both the necessity and sufficiency of this ubiquitin-dependent internalization signal demonstrated that when ubiquitinated lysine residues in Ste2 were deleted the receptor was not internalized, but a Ste2-ubiquitin fusion protein was competent for endocytosis (Shih et al 2000) Ubiquitination of plasma membrane proteins is mediated, at least in part, by the ubiquitin ligase Rsp5 (Rotin et al 2000) Together, these signals provide recognition motifs for endocytic machinery to aid in sequestering cargo into sites of endocytosis

Coated pit formation

Coated pits are sites of vesicle formation, in which clathrin and endocytic cargo assemble in preparation for internalization Aside from cargo and clathrin, adaptor proteins are key components of coated pits These proteins link together the various components required for endocytosis For example, the yeast epsins Entl and Ent2 bind to ubiquitinated cargo through their ubiquitin interaction motifs and also to clathrin through a C-terminal clathrin-binding motif (Wendland et al 1999; Aguilar et al 2003) Additional putative adaptors with similar domain organization also function in

endocytosis at the plasma membrane (Baggett et al 2003), and still others function in membrane trafficking at other subcellular locales (Zhdankina et al 2001; Friant et al 2003) This large diversity of adaptor proteins might be attributable to a cargo-sorting

Trang 21

mechanism, in which specific adaptors are utilized for the trafficking of a subset cargo proteins However, many receptors recognize multiple internalization signals, and hence it has been hypothesized that individual adaptors may mediate more general

physiological processes, such as stress response, by triggering the internalization of a particular spectrum of cargo (Wendland 2002)

Although not as tightly associated with coated vesicles as adaptors, accessory proteins also aid in the assembly and formation of coated pits These proteins have also been labeled endocytic scaffolding proteins, due to their proposed function of regulating and assembling endocytic complexes Panl and Edel are two examples of accessory proteins in yeast (Gagny et al 2000; Aguilar et al 2003; Miliaras et al 2004) Both are large proteins containing a diverse set of protein-protein interaction domains, and both have been implicated in endocytosis Accessory/scaffolding proteins are also targets for regulation of the endocytic pathway (see below) Although additional factors are

required for efficient vesicle formation, cargo, clathrin, adaptors, and accessory proteins are perhaps the components most intimately involved in the initial assembly of coated pits

Endosomal sorting and vesicle trafficking

Trang 22

and delivered to the vacuole lumen for degradation However, selected cargo can be recycled to the plasma membrane or transported to other destinations within the cell

Once vesicles have budded from donor membranes, such as the plasma membrane or early endosomes, how do they direct themselves to the proper target membrane? SNARE proteins and Rab GTPases together supply membrane targeting information to transport vesicles (Bonifacino and Glick 2004) SNAREs (SNAP Receptors) are extended, membrane-bound proteins that help to define membrane identity v-SNAREs are included in vesicles and bind to cognate t-SNAREs present in the target membranes These interactions provide two functions: first, to supply targeting information to a budded vesicle, so that it only fuses to membranes at its intended destination, and second, to support the membrane fusion process once the vesicle and target membranes dock This fusion occurs through a zipper-like mechanism involving the interaction of the cytoplasmic domains of the SNAREs (Bonifacino and Glick 2004) The accessory proteins a-SNAP and NSF (Sec17 and Sec18 in yeast, respectively (Wilson et al 1989; Griff et al 1992)) bind and dissociate the v-SNARE/t-SNARE complexes, allowing these components to be reutilized for subsequent rounds of membrane traffic

Rab GTPases are low molecular weight GTPases that also contribute to the regulation of membrane specificity in membrane trafficking These signaling GTPases localize to the intracellular face of membranes, and cycle between GTP- and GDP-bound forms In their active, GTP-bound form, Rab proteins signal to a diverse set of

downstream effectors Although the specific functions of these effectors are variable, these signaling processes support vesicle fusion to target membranes (Pfeffer 2001)

Trang 23

Together, SNARE proteins and Rab GTPases help to guide endocytic cargo through endosomes and sort internalized proteins, delivering them to their proper destination

Late endosomes and multivesicular bodies

Most endocytic cargo are transported through early endosomes and delivered to late endosomes, which have also been labeled pre-vacuolar compartments or

multivesicular bodies (MVBs) The term multivesicular body is derived from the distinctive morphology of late endosomes as they prepare to fuse with the vacuole and deliver cargo for degradation MVBs are formed through the inward budding of the limiting membrane of the late endosome, forming vesicles within the endosomal lumen (Stahl and Barbieri 2002) These lumenal vesicles are delivered to the vacuole interior upon fusion of the MVB with the vacuole, thus providing a mechanism for the

degradation of membrane proteins and lipids while preventing hydrolysis of the vacuolar limiting membrane itself and maintaining vacuolar integrity The late endosome

represents an intersection of the endocytic and biosynthetic trafficking pathways, as vacuolar resident proteins are delivered to endosomes from the TGN (Wendland et al

1998)

Trang 24

formation of the intralumenal vesicles themselves (Babst 2005) The initial step in sorting is the recognition of ubiquitinated cargo by Vps27, which is localized to endosomal membranes through its phosphatidylinositol (3) phosphate- (PtdIns(3)P) binding domain, and binds ubiquitin through its UIM domains (Katzmann et al 2003) Vps27 subsequently binds the ESCRT-I complex, consisting of Vps37, Vps28, and Vps23, and initiates MVB sorting Ubiquitinated cargo are transferred from Vps27 to ESCRT-I (Katzmann et al 2001), which activates the ESCRT-II complex (Vps22, Vps25, Vps36), which in turn recruits ESCRT-III (Vps2, Vps20, Vps24, Snf7) through an interaction with Vps20 (Babst et al 2002) ESCRT-III completes the sorting process and supports the formation of MVB vesicles (Babst et al 2002) Finally, the

deubiquitinating enzyme Doa4, together with its accessory partner Brol, removes the ubiquitin tag from MVB vesicular cargo, and the AAA-ATPase Vps4 triggers

dissociation of the ESCRT complexes from the endosomal membrane (Babst et al 1998; Luhtala and Odorizzi 2004) Upon fusion of the late endosome and vacuolar limiting

membranes, invaginated MVB vesicles are delivered to the lumen of the vacuole, where biosynthetic cargo are activated and most endocytic cargo are ultimately degraded

The Actin Cytoskeleton

The yeast actin cytoskeleton displays a dynamic but characteristic distribution pattern throughout the cell cycle (Engqvist-Goldstein and Drubin 2003) F-actin is typically found in three forms in yeast, cortical patches, actin cables, and the actomyosin ring During cell division, cortical patches polarize into the daughter bud and actin cables run through the mother along the mother-bud axis Actin also contributes to cytokinesis

Trang 25

through the actomyosin contractile ring that forms at the mother-bud neck Once cell division has been completed, actin patches depolymerize before repolarizing to the nascent bud site once the next round of cell division begins

Links between actin and endocytosis in yeast have been recognized for over a decade (Kubler and Riezman 1993) Cortical actin patches are hypothesized to be concentrated sites of endocytosis, and patch size and polarity is disrupted in many endocytosis mutants (Munn 2001; Kaksonen et al 2003) Furthermore, many

temperature-sensitive mutations in actin, actin-related, and actin-associated proteins show endocytosis defects immediately after a shift to non-permissive temperatures, and drugs that disrupt actin dynamics also induce rapid inhibition of internalization (Ayscough et al

1997; Ayscough 2000; Whitacre et al 2001)

The Arp2/3 complex is an actin-nucleating complex that plays a primary role in endocytosis Upon activation, the seven subunits of the Arp2/3 complex support the formation of branched networks of filamentous actin (Machesky and Gould 1999)

Arp2/3-dependent actin nucleation is tightly regulated, and requires the activity of Arp2/3 activators such as Las17, Abp1, and the endocytic scaffolding protein Pan1 (Winter et al

1999; Duncan et al 2001; Goode et al 2001) Many of these Arp2/3 activators are, in turn, closely linked to other endocytic factors, often through direct protein-protein interactions

Regulation of Endocytosis

Trang 26

polymerization (Wendland et al 1998; Engqvist-Goldstein and Drubin 2003) For example, the endocytic scaffolding protein Pan1 is regulated by the actin-regulating kinase Prk1 Panl is an Arp2/3 complex activator, and this actin nucleation activity may be primarily utilized upon Pan1’s interaction with other endocytic components and subsequent recruitment to active sites of endocytosis (Kaksonen et al 2003)

Phosphorylation of Panl by Prk1 inhibits both the formation of a complex between Pan1 and its endocytic partners End3 and Slal (Zeng et al 2001; Miliaras et al 2004), and Panl-dependent Arp2/3 actin nucleation (Toshima et al 2005) Prk1 also regulates additional endocytic components through phosphorylation, including the adaptors Ent1 and Ent2 (Watson et al 2001; Zeng and Cai 2005)

Ubiquitination of the endocytic machinery itself may be another post-translational modification regulating endocytosis The ubiquitin ligase Rsp5, in addition to its role in cargo selection, regulates endocytosis directly Experiments demonstrating reduced rates of internalization of a Ste2-ubiquitin fusion protein in rsp5 temperature sensitive mutants established a role for Rsp5 beyond tagging cargo for endocytosis (Dunn and Hicke 2001) Although the specific Rsp5 substrates responsible for this endocytic regulation have yet to be identified (Dupre et al 2004), the adaptor protein epsin and the scaffold Eps15 have both been shown to be ubiquitinated in mammalian cells (Klapisz et al 2002; Oldham et al 2002)

The GTPase dynamin is a major target for regulation of endocytosis in

mammalian cells Dynamin oligomerizes in a ring or spiral formation at the necks of coated pits (Hinshaw 1999) Upon GTP hydrolysis, dynamin rings promote fission of nascent vesicles from the plasma membrane (Chen et al 2004) Dynamin has a diverse

Trang 27

set of interacting proteins, including upstream and possibly downstream regulators of endocytosis (Schafer 2004) Although the majority of the endocytic machinery is conserved between yeast and mammalian cells, dynamin is the primary example of differences in internalization between the two species The yeast genome encodes three dynamin homologs, Dnm1, Vps1, and Mgm1, but none of them have been implicated in endocytosis (Danino and Hinshaw 2001) While the possibility remains that a dynamin- like protein does indeed function in yeast endocytosis (Yu and Cai 2004), current models propose that a combination of membrane lipid modification and actin polymerization provide the scission mechanism in yeast (Stefan et al 2005) Although the precise mechanism of fission may differ between yeast and mammalian cells, many of the regulatory molecules, including those that interact with mammalian dynamin or the proposed yeast fission machinery, are conserved between the two species, highlighting the important role of endocytic regulatory factors in internalization and cell physiology

Identification and Characterization of Endocytic Proteins

The following body of work aims to identify and characterize protein factors required for endocytosis in the baker’s yeast Saccharomyces cerevisiae The goal of Chapter 2 is to develop novel tools for the identification of endocytic proteins Specifically, attempts are made to adapt a biochemical endocytosis assay used in

Trang 28

Chapter 3 describes the identification and characterization of a gene found ina screen for mutants with defective uptake of a lipophilic dye The dim4-I] allele contains a mutation in the FAB/ gene, which encodes a lipid kinase responsible for the production of PtdIns(3,5)P, This lipid functions at the late endosome to sort biosynthetic cargo into MVB vesicles The characterization of the dim4-1/fab1-20 allele identifies novel defects in flux through the endocytic pathway and in sorting the endocytic cargo protein Ste3 at the MVB, demonstrating a universal role for PtdIns(3,5)P, in sorting all types of cargo into MVB vesicles and in regulating endocytosis

Chapter 4 characterizes a novel function for the endocytic adaptor protein Ent] The analysis of a temperature sensitive allele, ent/-7, establishes interactions between Entl and regulators of polarity Defects found in ent/-/ lead to loss of the interaction with Cdc42 GTPase activating proteins, as well as dramatic defects in both polarity- dependent processes and endocytosis These studies describe one of the first direct connections between components of endocytosis and polarity signaling pathways

Finally, in Chapter 5, I review five current areas of research that hold particular promise in extending our knowledge of membrane trafficking: the order of events

involved in assembling protein complexes for endocytosis, the emerging field of clathrin- independent endocytosis, roles for phosphoinositides in membrane trafficking, cross- regulation of membrane trafficking with other cellular processes, and the increasingly common identification of pathogens and other cellular invaders that utilize membrane trafficking machinery for their own purposes

Trang 29

CHAPTER 2

Developing an in vitro endocytosis assay for use with budding yeast

In this chapter, I attempt to develop a biochemical endocytosis assay for the identification of novel proteins involved in endocytosis, which could also represent a new approach for determining specific stages at which endocytic proteins function in yeast The data presented here review the development and optimization of this assay, which utilizes mammalian membranes in combination with yeast cytosol I found that the measurable activity of the assay was due to non-protein factors in yeast cytosol, rendering the assay ineffective for the identification and characterization of endocytic proteins These results indicate that although individual internalization factors are conserved between yeast and mammals, yeast cytosol was unable to support the internalization of mammalian

Trang 30

ABSTRACT

We have attempted to adapt a mammalian stage-specific in vitro endocytosis assay for use with cytosol from budding yeast This assay utilizes cultured A431 cells that are permeabilized and washed free of their cytosol Internalization is reconstituted through the addition of cytosol and ATP, and assayed through the internalization of biotinylated transferrin into coated pits or sealed vesicles This assay has been used extensively within mammalian systems as an effective tool for staging early endocytic events and identifying cytosolic endocytosis factors Our initial results indicated that low levels of yeast cytosol in combination with mammalian membranes supported

internalization of transferrin comparably to mammalian cytosol However, increased levels of yeast cytosol were inhibitory to the assay Optimization attempts demonstrated a requirement for ATP and GTP in the assay, and dialysis of yeast cytosol demonstrated that non-protein small molecules were responsible for the low levels of endocytosis detected in the assay

INTRODUCTION

The plasma membrane comprises the barrier between intracellular and extracellular regions of all cells Besides acting as a barrier, this membrane also facilitates interactions between the cell itself and the extracellular environment One such process is endocytosis, during which extracellular material and plasma membrane components are internalized via membrane bound vesicles into the interior of the cell Endocytosis is a multi-step process, and requires many protein factors to be executed with maximum efficiency and selectivity The identification of endocytic functions for

Trang 31

novel proteins continues to add to our understanding of the mechanisms underlying this complex process

The budding yeast Saccharomyces cerevisiae has proven to be an excellent model organism for the identification and characterization of endocytic factors Yeast genetic techniques have supplied much information on novel components, interactions, and functions for proteins involved in endocytosis For example, the end3 and end4 alleles were identified through a screen for mutants defective in a-factor uptake (Raths et al

1993), and PAN/, a yeast protein related to Eps15 and intersectin, was found to be required for endocytosis of the dye FM4-64 through a FACS-based genetic screen (Wendland et al 1996) Yeast and mammalian endocytosis rely on homologous

components and regulators, and much of the new data gathered utilizing yeast has led to elucidation of similar processes in higher eukaryotes (Geli and Riezman 1998; Engqvist- Goldstein and Drubin 2003; Dupre et al 2004)

Although yeast genetic analysis remains a very powerful tool, complementary approaches offer alternative advantages that can identify functional components that genetics might otherwise miss Jn vitro biochemical assays, in which biological

processes are reconstituted “from scratch”, provide an alternative approach to genetics for the identification and analysis of proteins required for a process such as endocytosis These types of assays can be developed equally effectively using mammalian

Trang 32

genetic data can be a very powerful and effective method of improving our understanding of the mechanisms underlying cellular process such as endocytosis

Biochemical assays have been used to reconstitute other membrane trafficking processes in yeast For example, the laboratory of Randy Schekman has successfully established cell-free assays to reproduce secretory traffic in vitro (Schekman 2002) Initial studies demonstrated that the in vitro assay reproduced in vivo trafficking events, as lysates from sec mutants were also defective for trafficking in the biochemical assay (Baker et al 1988) Subsequent studies utilized this assay to identify intermediate steps in the secretory pathway and the corresponding Sec proteins responsible for each of these steps (Rexach and Schekman 1991; Salama et al 1993) The identification of proteins responsible for secretory traffic through genetic techniques, followed by the assignment of these proteins to specific membrane trafficking functions through in vitro

reconstitution assays, proved to be an incredibly powerful combination that mapped out much of the secretory pathway and continues to influence research into membrane trafficking to this day

In this study, we attempt to develop a biochemical in vitro endocytosis assay utilizing soluble yeast components We have taken an approach in which we adapt a functional in vitro assay established in the mammalian system and attempt to convert it for use with yeast cytosol This assay utilizes mammalian membranes labeled with tagged ligand and follows their internalization upon the addition of cytosolic components (Schmid and Smythe 1991; Smythe et al 1992) This approach is designed to identify novel cytosolic factors required for endocytosis in yeast that may have been overlooked in genetic screens, such as proteins required for viability Factors identified through this

Trang 33

assay will be subsequently analyzed in vivo using standard yeast techniques, including potential genetic manipulation

RESULTS

Developing the in vitro assay

We adapted a mammalian in vitro endocytosis assay to function with yeast components The methodology of the assay was similar to that developed by the Schmid laboratory (Smythe et al 1989; Schmid and Smythe 1991; Smythe et al 1992), utilizing yeast cytosol instead of mammalian (Fig 2-1A) Briefly, mammalian cultured A431 cells were perforated and washed to remove cytosolic components Biotin-labeled transferrin was added to the semi-intact membranes and incubated to allow binding to native transferrin receptors These labeled membranes were incubated at 37°C for 30 minutes with cytosol (either yeast or mammalian) and an ATP regenerating system to support endocytosis After this incubation, excess streptavidin was added to bind and block all exposed biotin-transferrin; biotin sites that were internalized into constricted pits or sealed vesicles (i.e., inaccessible to the streptavidin probe) remained unbound This approach allowed differentiation of internalized versus surface transferrin molecules

Subsequent steps of the assay allowed simple and quantitative detection of the internalized transferrin molecules utilizing a colorimetric assay (Fig 2-1A) First, excess free biocytin was added to soak up any streptavidin molecules that remained unbound to surface transferrin Next, the entire mixture was solubilized and added to ELISA plates coated with transferrin antibodies After incubation, these plates were washed

Trang 34

transferrin molecules that were internalized in the assay contained free biotin, while biotin-transferrin molecules that remained uninternalized were already bound by streptavidin To bind and label the internalized transferrin containing free biotin,

horseradish peroxidase-labeled streptavidin (avidin-POD) was added to the ELISA plates A colorimetric substrate (OPD) was subsequently added to the mixture, allowed to

develop via catalysis by POD in the ELISA plate, and the reaction stopped with an acidic solution The resulting colored solution was then quantified in a spectrophotometer Thus, this assay allowed us to measure, with a straightforward colorimetric readout, internalization of transferrin from perforated membranes under controlled conditions

Before testing the assay, we sought to characterize the efficacy of the perforation and washing process To address this, we stained the cells before and after perforation with the membrane impermeable dye trypan blue Prior to perforation, less than 10% of the A431 cells stained positive for trypan blue After the perforation procedure, over 95% of the cells stained blue (data not shown), indicating that the vast majority had lost membrane integrity To test the effectiveness of the cytosol-removal washes, we stained intact and perforated cells for actin filaments using rhodamine-phalloidin, which binds, stabilizes, and labels actin filaments (Fig 2-1B) We saw a dramatic, but not complete, reduction in actin filaments after perforation and washing, consistent with partial removal of cytosolic components Addition of either mammalian or yeast cytosol after perforation partially restored some of the actin staining These data indicate that the perforation and washing processes were reasonably effective, and that some intracellular structures remained within the semi-intact membranes

Trang 35

Figure 2-1 Establishing the in vitro assay

(A) A diagrammatic representation of the in vitro endocytosis assay Biotin- transferrin is added to semi-intact membranes and incubated to allow

internalization Avidin blocks surface biotin sites, leaving internalized transferrin unbound After blocking free avidin and solubilizing, the mixture is added to ELISA plates coated with anti-transferrin antibodies Peroxidase-labeled avidin binds to unblocked, internalized transferrin, and the peroxidase levels are measured through a colorimetric reaction

(B) A431 cultured cells were labeled for filamentous actin with rhodamine-

phalloidin Cells were either intact (left panel) or perforated (three right panels), and mammalian or yeast cytosol was added to the perforated membranes

(C) Percentages of internalized transferrin are shown for the controls typically run in each assay These results represent averages of multiple experiments, with the standard deviation shown as error Assay mixtures were incubated with or without ATP and yeast cytosol at 37°C, or complete mixtures were incubated at 4C

(D) A representative assay showing internalization measured using either yeast or mammalian cytosol across a range of cytosol amounts

Trang 36

e®

"ee a°Y °

tose wt gtite avo TOTS De® S > & 08

Add BxxTfn (®) “avidin (8) AGG Excess Solubilize

si-#— mã —~=——

Add mix to ceoated Wash andadd Add colorimetric Read color

Trang 37

To convert the spectrophotometer results into reliable measurements of

internalized transferrin, each assay included a set of control wells The first control wells represented the total amount of biotin-transferrin, both surface and internalized, present on the membranes (“totals”) For these wells, the “blocking” step, in which excess avidin was added to bind uninternalized transferrin, was omitted, leaving all transferrin

unblocked and accessible to the subsequent streptavidin-POD label The absorbance of these “totals” was defined as 100%, and subsequent controls and conditions were calculated as the fraction of total transferrin that was internalized

Additional controls omitted components of the assay necessary for optimal internalization, such as ATP, cytosol, or temperature (Fig 2-1C) Reactions containing semi-intact membranes alone, with neither ATP nor cytosol added, were considered to represent background levels of absorbance This signal was subtracted from the detected absorbance in all subsequent results The addition of ATP to the membranes supported a low level of endocytosis, consistent with our data indicating that some residual

intracellular components remain associated with the membranes after perforation and washing However, addition of cytosol without ATP, or incubating reaction mixtures at 4°C, resulted in internalization approximately equivalent to background These data indicate that the internalization measured in this assay is temperature- and energy- dependent

Trang 38

However, as the amount of cytosol was increased, yeast cytosol ceased to support further internalization, and even showed a mild negative effect at high levels

Initial analyses using yeast cytosol

We were encouraged to see that yeast and mammalian cytosols had similar stimulatory effects on internalization when used at low levels, and we utilized this phenomenon to begin to analyze alterations in cytosol components First, we ran yeast cytosol over an ion exchange column, and combined the resulting fractions into three bulk pools: the flowthrough, the low salt eluate, and the high salt eluate Each of these were tested in the assay, individually and in combination (Fig 2-2A) Both the

flowthrough and high salt eluate pools showed activity in the assay However, the low salt pool appeared to reduce this activity when combined with either of the other pools When all three were combined, we saw an intermediate effect

In addition to biochemical fractionation, we sought to use this assay to analyze mutants that may have defects in the endocytic pathway Cytosols were prepared from strains containing deletions or mutations of membrane trafficking proteins and assayed for transferrin uptake, with the results normalized against wild type (Fig 2-2B) The mutants end3A and abpIA showed moderate reductions in internalization compared to wild type, consistent with their established roles in endocytosis (Raths et al 1993; Wesp et al 1997) Surprisingly, deletion of the yeast dynamin homolog VPS/ also led toa reduction in internalization (see further analysis below) Other mutations showed even more moderate effects However, due to the limited amount of internalization detectable

Trang 39

Figure 2-2 Identifying components required for internalization in the assay (A) Yeast cytosol was fractionated over an ion exchange column, and the fractions

collected into three batches These batches were run alone or in combination in the assay 6ug of flowthrough, 4.5ug of low salt eluate, and/or lug of high salt eluate (as determined through a Bradford assay) were used in each assay

(B) Mutant cytosols were run in the assay Between 5-30 ug of mutant and wild type cytosol were used, and results for each mutant were normalized to wild type at each cytosol amount The combined averaged data are shown here, with the error bars indicating standard deviation All mutants were deletions, with the exception of panI, which was the pan1-20 allele, and all were compared to their