Expip is a cargo adaptor for Sec24p mediated
export the plasma membrane H+ ATPase from the
ER in S. cerevisiae
ARCHNdES
IMASSACHUSETTS IN517Y
By
OF TECHNOLOGY
Darcy L. Morse
MAY 3 1 2013
B.S., Biology (Genetics & Development)
Cornell University, 2001
LiBRARIES
Submitted to the Department of Biology in partial fulfillment of the requirements for the
degree of Doctor of Philosophy at the Massachusetts Institute of Technology
June 2013
©2013 Darcy L. Morse
All Rights Reserved
The author hereby grants to MIT permission to reproduce and to distribute publicly paper
and electronic copies of this thesis document in whole or in part in any medium now
known or hereafter created.
Signature of Author:<
Darcy L. Morse
Department of Biology
June 2013
Certified by:
Chris A. Kaiser
Professor of Biology
Thesis Supervisor
/1
Accepted by:
E. Keating
Associate Professor of Biology
Chairperson, Committee on Graduate Studies
-Amy
1
E
Expip is a cargo adaptor for Sec24p mediated
export the plasma membrane H* ATPase from the
ER in S. cerevisiae
By Darcy L. Morse
Submitted to the Department of Biology in May, 2013 in partial fulfillment of the
requirements for the degree of Doctor of Philosophy
Abstract
The secretory pathway in S. cerevisiae is responsible for the folding, modification
and delivery of plasma membrane and secreted proteins. The secretory pathway consists
of an ordered series of organelles including the endoplasmic reticulum (ER), Golgi,
endosome and vacuole. In the ER, properly folded proteins are separated from misfolded
and ER resident proteins and packaged into COPII coated vesicles for transport to the
Golgi network and ultimately to the cell surface. How cargo proteins are distinguished
from ER resident proteins and loaded into vesicles remains unclear for many cargo
proteins.
To facilitate our study of the mechanism of protein sorting in the ER, we selected
the plasma membrane H+ ATPase PMAJ as a model cargo protein. Pmalp is essential for
creating an electrochemical gradient across the plasma membrane required for nutrient
uptake and maintenance of cytosolic pH and accounts for -25% of the total plasma
membrane proteins. We previously reported that Pmalp requires the Sec24p paralog
Lstlp for efficient export from the ER. Here we report the identification of a novel gene,
EXP1 (ER export of Pmalp), which when overexpressed can restore localization of
Pmalp to the plasma membrane in an ist1A strain. Strains carrying a deletion of EXP1
grow normally, but an ist1A exp1A double mutant is inviable, displaying a severe Pmalp
trafficking defect. EXP1 encodes a 17 kDa integral membrane protein which cycles
between the ER and Golgi and physically interacts with both Sec24p and Pmalp Taken
together our data suggests that Pmalp exits the ER via two independent pathways. Lstlp
mediates the primary route while ExpIp is required as a cargo adaptor for the secondary
Sec24p mediated route.
2
We also present data from a screen of S. cerevisiae essential genes under the
control of a doxycycline regulatable promoter for genes required for trafficking of the
cell wall
mannoprotein Ccwl4p. One gene identified, PGA2,
shares physical
characteristics with EXP1 raising the possibility that Pga2p may function as a cargo
adaptor for ER export. We show that Pga2p is an ER localized integral membrane protein
required for ER export of Ccwl4p, CPY and invertase.
Thesis Supervisor: Chris A. Kaiser
Title: Professor of Biology
3
Acknowledgements
I would like to thank my advisor, Chris Kaiser, for his guidance and support during my
graduate work.
I would also like to thank the members of my committee for their invaluable advice and
support: Frank Solomon, Thomas Schwartz, Troy Littleton, and Daniel Oprian.
I am grateful to members of the Kaiser lab past and present for making the lab an
enjoyable place to work: Michelle O'Malley, Natalie Cain, Carolyn Sevier, Eric Spear,
April Risinger, Jeni Sideris, Raissa Eluere, Yaohua Zhong, Roymarie Ballester, Hwani
Kim, Rahel Siegenthaler, Minggeng Gao, Andrea Vala, Marta Rubio, Hongging Qu,
Barbara Karampalas, and Polly McGahan. I would especially like to thank Barbara
Karampalas and Polly McGahan for keeping the lab running smoothly; Eric Spear,
Natalie Cain and April Risinger for their advice and support; Michelle O'Malley for her
immeasurable guidance and my Starbucks addiction and Carolyn Sevier for acting as a
second mentor always willing to offer guidance and expertise.
I thank all my friends both at MIT and elsewhere for their support throughout my
graduate career. I especially thank Brent Cezairliyan for scientific and non-scientific
guidance and friendship.
My family has been a constant source of love and encouragement. I thank my parents,
Rose Morrow and Robert Morse for always supporting me and pushing me to succeed. I
thank: my brothers Alan Morse, Brian Morse and Bruce Morse for their support and for
always being there when I need advice or a laugh; my nieces and nephews for always
making me smile, and my grandparents Harold and Arlene Shattuck for their love and
support. I would like to especially thank my sister, Kimberly Corder, for always being
there for me when I need advice or just someone to talk to.
4
Table of Contents
TITLE PAGE
...........................................................................................................................................
ABSTRACT
1
................................................................................................................................
ACKNOW
T LEDGEM
.N R ENTS
U
2
S...................................................................................................................4
TABLE OF CONTENTS.......................................................................................................................................
S
CHAPTER 1: INTRODUCTIONS ..T W............................................................................................................17
INTRODUCTION .
N.THE.SECRETRYY........................................
8
....................................................................
OVERVIEW OF THE SECRETORY PATHWAY............................................................................................
........ 10
VESICLE TRANSPORT IN THE SECRETORY
....................................................................................
PATHWAY
16
ER TO GOLGI TRAFFICKING..............................................................................................................................
18
CARGO CAPTURE BY SEC24 ................................................................................................
23
GOLGi RETRIEVAL .........................................................................................................
28
DISEASES ASSOCIATED WITH TRAFFICKING DEFECTS...............................................................................................31
THE PLASMA MEMBRANE ATPASE PMA1P ......................................................................................................
33
REFERENCES .................................................................................................................................................
37
FIGU RES ......................................................................................................................................................
51
CHAPTER 2: EXP1P IS A CARGO ADAPTOR FOR SEC24P MEDIATED TRANSPORT OF THE PLASMA
M EM BRANE ATPASE PM A1P IN S. CEREVISIAE .......................................................................................
55
PREFACE......................................................................................................................................................
56
ABSTRACT ....................................................................................................................................................
57
INTRODUCTION..............................................................................................................................................58
M ATERIALS AND METHODS.............................................................................................................................61
RESULTS ......................................................................................................................................................
67
DISCUSSION..................................................................................................................................................82
REFERENCES .................................................................................................................................................
87
T ABLES ........................................................................................................................................................
91
FIGURES ......................................................................................................................................................
96
CHAPTER 3: PROSPECTUS.....................................................................................................................118
IN VITRO ANALYSIS OF SEC24P MEDIATED EXPORT OF PMA1P...............................................................................120
HOW DOES ExP1P INTERACT WITH SEC24P ......................................................................................................
121
HOw IS PMA1P RECOGNIZED BY LST1P............................................................................................................121
5
W HAT ISTHE CARGO SPECIFICITY OF LST1P AND SEC24P .....................................................................................
122
W HAT ISTHE ROLE OF LST1P AND ExP1P IN LOADING OF LARGE CARGO .................................................................
122
DOES ExP1P REGULATE PMA1P ACTIVITY.........................................................................................................123
IDENTIFICATION OF ADDITIONAL CARGO ADAPTORS.............................................................................................126
REFERENCES ...............................................................................................................................................
127
FIGURES ....................................................................................................................................................
129
APPENDIX
I: CCW 14 SCREEN.................................................................................................................133
ABSTRACTLSAN
........ .MET............................................................................................................................
134
INTRODUCTIONN .F TU EN............
135
............................................................................................................-
MATERIALS AND METHODS...........................................................................................................................
139
RESULTS ....................................................................................................................................................
141
DISCUSSION AND FUTURE DIRECTIONS.............................................................................................................
144
REFERENCES .A..................
149
FIGURES
..........................................................................................................
..................................................................................................................................................
TABLES .......
............................................................................................................................................
APPENDIX II: ANALYSIS OF PGA
..........................................................................................................
152
1 2.164
170
ABSTRACT ..................................................................................................................................................
171
INTRODUCTION...........................................................................................................
172
MATERIALS AND METHODS ................................................................................................
174
RESULTS....................................................................................................................
179
DISCUSSION AND FUTURE DIRECTIONS .............................................................................................................
185
REFERENCES ...............................................................................................................................................
189
FIGURES ....................................................................................................................................................
191
TABLES ......................................................................................................................................................
201
APPENDIX III: EXPLORATION OF ESSENTIAL GENE FUNCTIONS VIA TITRATABLE PROMOTER ALLELES ..202
6
Chapter 1: Introductions
7
Introduction
In eukaryotic cells, proteins destined for the plasma membrane or secretion are
transported through a series of endomembrane organelles that together constitute the
secretory pathway. Approximately 30 percent of the eukaryotic genome enters into the
secretory pathway including proteins that are resident to the organelles of the pathway as
well as all secreted or cell surface proteins (Kanapin et al., 2003). The endoplasmic
reticulum or ER serves as the entry point to the secretory pathway. From here properly
folded proteins segregate from misfolded and ER resident proteins for transport to the
Golgi. At the Golgi, proteins sort to either the plasma membrane/cell surface or to the
endosome/vacuole. An important aspect of secretory pathway function is the tight
regulation of lumenal pH of the organelles. The ER lumen is maintained at a pH of -7.2,
and the pH of subsequent organelles becomes more acidic ending with the pH of the
vacuole at -4.0 (Paroutis et al., 2004). This pH difference is important for modulation of
proper protein interactions allowing binding in one organelle and release in another.
There is also a system of retrograde transport that allows proteins involved in trafficking
or escaped resident proteins to be returned to earlier organelles, also relying on
differences in organelle pH to regulate proper protein interactions (Scheel and Pelham,
1996).
Biochemical, genetic, morphological and structural studies over the past 30 years
have provided detailed information about mechanisms of protein folding and transport
through the secretory pathway. The first secretory (SEC) genes were identified through
an elegant screen that took advantage of the idea that an inability to traffic proteins would
lead to an accumulation of these proteins internally, therefore increasing cell density.
8
Mutagenesis of S. cerevisiae cells identified 23 temperature sensitive mutants with
internal accumulation of secretory proteins, 10 of which are involved in ER to Golgi
transport (Novick et al., 1980). Subsequent
morphological analysis using electron
microscopy showed that when grown at the restrictive temperature sec17, sec18 and
sec22 accumulate vesicles and are likely to be involved in vesicle fusion at the Golgi
whereas sec12, sec13, sec16, and sec23 do not accumulate vesicles and are likely
involved in vesicle formation (Kaiser and Schekman, 1990). Genetic and biochemical
analysis of these vesicle formation mutants led to the discovery of SEC24 and SEC31,
which along with SEC13 and SEC23 are core components of the ER to Golgi vesicle
coat (Hicke et al., 1992; Salama et al., 1997). Further analysis of these SEC proteins and
identification of additional proteins has led to a greater understanding of the steps
involved in transport of proteins from the time of translation through delivery to their
final cellular destination. Although the main components have been identified there are
still many questions remaining about the regulation of this critical pathway.
The work presented here focuses on the movement of cargo proteins through the
early secretory pathway, specifically how the plasma membrane H* ATPase PMA1 is
recognized for loading into vesicles at the ER. Previous work in the lab showed that
efficient export of Pmalp from the ER depends on the Sec24p paralog Lstlp (Roberg et
al., 1999). Here we show that in the absence of Lstlp, the small ER membrane protein
Explp is required as a cargo adaptor for efficient packaging of Pmalp into Sec24p
containing vesicles (Chapter 2). We also present evidence that another protein, Pga2p
may serve as a cargo adaptor for as yet unknown cargo proteins (Appendix II).
9
Overview of the secretory pathway
Translocation into the ER
Cargo proteins of the secretory pathway enter the pathway at the ER, the surface
of which is coated with ribosomes that allow coupling of protein translation with
translocation into the ER. As newly synthesized proteins emerge from the ribosome, they
can interact directly with the translocation machinery. There are three distinct
mechanisms of translocation depending on the location of the hydrophobic signal
sequence (or transmembrane spanning sequence) and the ultimate orientation of the N
and C termini of the protein relative to the membrane. In co-translational translocation
proteins interact with translocation machinery as the nascent chain is forming on the
ribosome. This mechanism is used for proteins with a signal sequence near the N
terminus, such that it is translated first. The signal sequence is recognized by the signal
recognition particle (SRP) and SRP receptor (SR) as it emerges from the ribosome (Ogg
et al., 1998). The SRP/SR complex slows translation and transfers the nascent
protein/ribosome complex to the Sec61 translocon allowing translation to continue as the
protein is translocated into the ER (Wild et al., 2004).
Post-translational translocation occurs when the completely synthesized protein is
released from the ribosome prior to its interaction with the translocation machinery,
requiring the binding of cytosolic chaperone proteins to the nascent chain in order to keep
the peptide unfolded and therefore competent for translocation. These chaperones are
members of the Hsp70 ATPase and Hsp40/DnaJ families (Chirico et al., 1988). In posttranslational translocation, the N-terminal signal sequence is recognized by the Sec63
complex leading to the transfer of the nascent chain to the Sec6l translocon where
10
association with the lumenal Hsp70 protein Kar2p drives translocation into the ER
through an ATP dependent ratcheting mechanism (Rapoport, 2007).
A third mechanism is employed by a small group of proteins which have Cterminal transmembrane tail anchors. As the C terminus is the last to be translated, the
transmembrane region remains blocked by the ribosome until translation is completed,
preventing SRP dependent targeting (Yabal et al., 2003). This tail anchor pathway is also
independent of the Sec6l and Sec63 complexes (Steel et al., 2002; Yabal et al., 2003).
Instead these tail anchor (TA) proteins use the GET (guided entry of TA proteins)
pathway, consisting of Getl-5p and Sgt2p. Once released from the ribosome a nascent
TA protein binds the chaperone Sgt2p, which is in a complex with Get4p/Get5p. The TA
protein is transferred to the ATPase Get3p, which targets the TA proteins to the
membrane, where the GET3p-TA protein complex interacts with the ER membrane
proteins Getlp and Get2p. The cytosolic domain of Getlp disrupts binding of the TA
protein to Get3p allowing it to be inserted into the lipid bilayer (Wang et al., 2011; Hegde
and Keenan, 2011). How this insertion occurs is still unclear.
Protein folding and quality control in the ER
Once a nascent protein enters the ER, it needs to be folded in order to continue on
in the secretory pathway. This folding and maturation process may include cleavage of a
signal sequence, disulfide bond formation, addition of a glycophospatidylinositol (GPI)
anchor, or N- and O-linked glycosylation. Once properly folded, proteins enter into
vesicles for transport to the Golgi (Figure 1, step 1). Misfolded proteins, incompletely
assembled oligomers and ER resident proteins are segregated away from properly folded
cargo proteins through either direct retention in the ER or retrieval from the Golgi. The
11
ER contains a number of chaperone proteins, including Kar2p (BiP) and PDI, that assist
protein folding and monitor the folding state to prevent premature sorting (reviewed in
(Benyair et al., 2011; Araki and Nagata, 2011).
To maintain protein folding homeostasis the cell must carefully monitor the
amount of unfolded protein and adjust the production of protein folding machinery
accordingly. An increase in unfolded or misfolded proteins in the ER can lead to a
condition of ER stress that can have deleterious consequences for a cell. To deal with this
problem, eukaryotic cells have evolved the unfolded protein response (UPR). When
activated by ER stress the UPR coordinates the transcriptional increase of machinery
needed for protein folding, lipid biosynthesis, and ER associated degradation (ERAD)
while also decreasing production of other proteins to reduce the folding load on the cell.
Three parallel UPR pathways exist in metazoans, each regulated by a different ER
resident integral membrane protein: Irel, PERK, or ATF6 (reviewed in (Gardner et al.,
2013). In S. cerevisiae, the UPR pathway is regulated by Irelp. In the presence of ER
stress Ire 1p forms a higher order oligomer resulting in the activation of its Rnase domain
(Sidrauski and Walter, 1997). Activated Ire 1p cleaves the mRNA HAC 1 to create a bZIP
transcription factor that specifically upregulates ER quality control proteins (Travers et
al., 2000). Activation of the UPR by Irelp is tightly regulated. In the monomeric inactive
form, the ER chaperone Kar2p binds Irelp. As the level of unfolded proteins increases,
Kar2p releases Ire 1p and binds to these unfolded proteins (Gardner and Walter, 2011). As
the level of unfolded proteins increases beyond the capacity of ER chaperones to bind
them, the unfolded proteins can bind directly to Irelp, leading to Irelp oligomerization
and activation (Credle et al., 2005; Zhou et al., 2006; Gardner and Walter, 2011). The
12
multiple levels of regulation prevent activation of the UPR when the increase in unfolded
proteins is small and/or transient.
The ER also has a mechanism for recognizing irreversibly misfolded proteins
referred to as ERAD (ER associated degradation) that results in removal of these proteins
from the ER via retrotranslocation, ubiquitination and subsequent degradation by the
cytosolic proteasome. There are two ERAD complexes in S. cerevisiae named for the E3
ubiquitin ligase each contains: HRD1 or DOA10. Misfolded proteins can be recognized
through several mechanisms. The Hrdlp complex recognizes lumenal misfolding signals
such as a specific modification of the N-linked glycosylation mannose chain that can
occur when a protein remains in the ER for an extended period of time (Gauss et al.,
2011). Evidence suggests that an O-linked modification can also trigger ERAD but the
mechanism for this is still unclear (Goder and Melero, 2011). The Hrdlp complex is also
able to recognize defects within the transmembrane region of proteins (Carvalho et al.,
2006). The DoalOp complex recognizes defects on the cytosolic side of ER membrane
proteins (Gnann et al., 2004; Vashist and Ng, 2004). Once identified and ubiquitinated by
either the HrdIp or Doa Op complexes, the misfolded proteins retrotranslocate out of the
ER by an as yet unknown mechanism, which may involve the Sec61p translocon or
Hrdlp, and are degraded by the cytosolic proteasome (Plemper et al., 1999; Carvalho et
al., 2010).
The Golgi apparatus
The Golgi apparatus consists of sequential cis, medial, and trans-cisternae. In
mammalian cells these cisternae form tight stacks with the cis-cisternae closest to the ER.
In S. cerevisiae, the cisternae are diffused throughout the cytosol rather than tightly
13
grouped (Wooding and Pelham, 1998). As secretory proteins traverse the Golgi cisternae,
further proteolytic cleavage or glycosylation can occur. The Golgi cisternae can be
classified by the distinct glycosyltransferases required at each step in maturation
(Hashimoto et al., 2002; Velasco et al., 1993). Though there has been some controversy
about how different cisternae arise, current evidence based on detailed microscopy
supports the idea of cisternal maturation (Papanikou and Glick, 2009; Suda and Nakano,
2012). Newly formed cis-cisternae progressively move towards the trans side of the
Golgi as they are remodeled by loss of early acting Golgi enzymes and acquisition of late
acting enzymes. Secretory cargo remains within a given cisterna while the Golgi enzymes
are trafficked to and from the cisterna by retrograde vesicle transport (Figure 1, step 2).
When the secretory cargo reach the trans-Golgi network, they sort to the plasma
membrane, are secreted, or enter into the endosomal pathway (Figure 1, steps 3 and 4).
At this stage, proteins that are misfolded or endocytosed from the cell surface can be
targeted through the endosomal system for degradation in the vacuole (or lysosome in
mammalian cells) (Figure 1, steps 5 and 6). Other misfolded proteins as well as escaped
ER resident proteins and components of the anterograde trafficking machinery return to
the ER where they can make another attempt at folding or be reused (Figure 1, step 1).
ERGIC in mammalian cells
Although cargo transport in mammalian cells seems to share machinery with that
of S. cerevisiae there are some differences. Mammalian cells have a unique
subcompartment between the ER and Golgi called the ER Golgi intermediate
compartment or ERGIC (reviewed in (Appenzeller-Herzog and Hauri, 2006; LorenteRodriguez and Barlowe, 2011). Vesicles that bud from the ER fuse with the ERGIC in
14
mammalian cells rather than the Golgi as they do in yeast. The ERGIC resides near the
ER membrane and has different properties than either the ER or Golgi (Ben-Tekaya et al.,
2005). The ERGIC has two main functions: to concentrate cargo that leaves the ER and
to sort cargo for anterograde or retrograde transport. These functions are accomplished by
the ER and cis-Golgi in yeast. This introduction will focus on the secretory pathway in S.
cerevisiae rather than mammalian cells, but the mechanisms and many of the proteins
required for anterograde and retrograde trafficking are the same in both systems.
Late Secretory Pathway
At the trans-Golgi network proteins are sorted for delivery to their final cellular
locations. Plasma membrane and secreted proteins are packaged into vesicles that fuse
with the plasma membrane (Figure 1, step 3). Other proteins enter into the endosomal
pathway (Figure 1, step 4). Once a protein enters the endosome it can be trafficked to the
vacuole for use or degradation or can be recycled back to the Golgi for transport to the
plasma membrane (Figure 1, steps 4 and 5). One example of endosomal recycling is the
general amino acid permease (Gaplp), which is sorted to the plasma membrane or
endosome depending upon nutrient availability, and can be returned from the endosome
to the plasma membrane under low nutrient conditions (Chen and Kaiser, 2002; RubioTexeira and Kaiser, 2006).
Another form of vesicle mediated regulation in the cell that is not directly part of
the secretory pathway is endocytosis (Figure 1, step 6) (reviewed in (Conibear, 2010;
Weinberg and Drubin, 2012). This enables the cell to remove proteins from the cell
surface and in doing so also brings soluble components of the extracellular media into the
cell. Endocytosis is also used to internalize damaged proteins for degradation and to
15
regulate protein levels at the cell surface. The target of endocytic vesicles is the
endosome, which means that these internalized proteins can traffic to the vacuole or can
recycle back to the cell surface depending on the cells needs. This system therefore
serves as a method for the cell to regulate trafficking of cell surface proteins based on the
nutrient availability and cell environment at any given time (Chen and Kaiser, 2002).
Vesicle transport in the secretory pathway
Anterograde and retrograde trafficking within the secretory pathway depends on
tightly regulated vesicle formation and delivery. There are three distinct types of coat
complexes that are responsible for the formation of vesicles from specific membranes.
The coat protein complex II (COPII) is required for ER to Golgi anterograde trafficking.
The coat protein complex I (COPI) is involved in Golgi to ER retrograde trafficking as
well as trafficking between the Golgi cisternae. The clathrin coat (CCV) is involved in
vesicle trafficking from the trans-Golgi network and within the endosomal system as well
as endocytosis from the plasma membrane. In order to form a transport vesicle the coat
proteins must: recognize and load cargo, induce membrane bending to form a coated bud,
coordinate membrane scission to release a vesicle, and disassemble to allow fusion of the
vesicle with the target membrane (Faini et al., 2013). Although the components of these
coats are different they have some structural and mechanistic similarities. They all have
an inner coat or adaptor complex that is responsible for initial membrane recruitment and
cargo recognition, with recruitment to specific membranes often regulated by small
GTPases (Sarlp, Arflp or dynamin). The outer coat or cage complex of each vesicle is
responsible for the bulk of the membrane deformation and vesicle formation. Each outer
coat is composed of polymerized protein complexes that contain N-terminal n-propeller
16
WD40 and C-terminal
-solenoid motifs, but each vesicle type has a different cage
architecture (McMahon and Mills, 2004; Stagg et al., 2007). Each coat complex also
relies on additional cargo adaptors to aid in recruitment of diverse cargo. As each type of
vesicle forms, different GTPase activating proteins (GAPs) act to enhance GTP
hydrolysis allowing coat release and vesicle fusion with specific target membranes. As
we are interested specifically in the early trafficking steps of Pmalp this introduction will
focus on trafficking involving COPII and COPI coats.
Clathrin Vesicles
The first coated vesicle discovered was the clathrin coated vesicle (CCV), named
for the protein that forms the outer coat. CCVs are comprised of two distinct complexes:
an adaptor complex and a cage complex. CCVs function in multiple steps in protein
trafficking and there are different adaptor complexes associated with different
intracellular membranes. These adaptor complexes include the large heterotetrameric
adaptors AP1-AP5 and the monomeric GGA (Golgi-localizing, y-adaptin ear containing,
ARF-binding) proteins (McMahon and Mills, 2004). Adaptors are recruited to the
membrane by GTPases or, in the case of AP2, by the membrane phospholipid PIP2. AP 1
mediates trafficking between the trans-Golgi network (TGN) and endosomes, AP2
localizes at the plasma membrane and mediates endocytosis and the GGAs mediate
endosomal sorting. Clathrin heavy and light chains form the outer cage complex. The
heavy chain has an N-terminal n-propeller domain that contains binding sites for adaptor
proteins and a stretch of a helices that assemble into a long a-solenoid that forms a
triskelion cage structure. (reviewed in (Faini et al., 2013). Clathrin vesicles use a wide
17
range of accessory proteins to carry out trafficking of a diverse range of cargo (reviewed
in (McMahon and Mills, 2004).
ER to Golgi trafficking
COPH vesicle formation
Exit of properly folded proteins from the ER occurs via specialized COPII coated
vesicles that are responsible for vesicle formation and cargo loading. The COPII coat
consists of a cytosolic GTPase, Sarlp, and 2 heteromeric subcomplexes Sec23p/Sec24p
and Secl3p/Sec3lp, that act sequentially in the formation of vesicles from the ER
membrane
(Figure 2) (Barlowe et al., 1994). Vesicle formation occurs at specialized
regions of the ER membrane called ER exit sites, or ERES, which
are devoid of
ribosomes and often located in regions of the ER membrane closest to the Golgi (Orci et
al., 1991; Bannykh et al., 1996). Vesicle formation initiates at the ER membrane when
Sec 12p, a guanine nucleotide exchange protein (GEF), catalyzes the nucleotide exchange
of Sarlp-GDP to Sarlp-GTP (Figure 2 step 1) (Lee et al., 2005). Sarlp-GTP becomes
membrane associated and recruits the Sec23p/Sec24p heterodimeric subcomplex (Figure
2, step 1 and 2). Together, Sarlp-GTP and Sec23p/Sec24p form the "pre-budding"
complex which is sufficient for cargo recognition and binding (Belden and Barlowe,
1996).
The Sec23p/Sec24p
inner coat subcomplex recruits the Sec13p/Sec31p
heterotetramer which polymerizes into a lattice-like structure deforming the membrane
and leading to vesicle formation (Figure 2 step 3) (Stagg et al., 2006). The vesicles travel
to the Golgi, where the COPII coat releases from the vesicle during tethering to allow
vesicle fusion with the Golgi membrane (Antonny et al., 2001). In yeast there are three
isoforms of the Sec24p subunit (Sec24p, Lstlp and Isslp) (Roberg et al., 1999; Shimoni
18
et al., 2000; Kurihara et al., 2000). In mammals there are two Sarlp, two Sec23p, four
Sec24p and two Sec31p paralogs suggesting that there is probably some cell and tissue
specificity in COPII formation and cargo sorting (Wendeler et al., 2007).
Sarip
Initiation of COPII vesicle formation depends on the activation and membrane
binding of the small GTPase Sarlp (Figure 2 step 1). The GEF Sec12p catalyzes the
exchange of GDP for GTP at the ER membrane forming Sarlp-GTP (Yoshihisa et al.,
1993). The Sec12p mediated activation of Sarlp is restricted to the ER membrane
ensuring specificity of COPII vesicle formation. Binding of GTP to Sarlp leads to a
conformational change releasing an amphipathic x-helix, which inserts shallowly into the
ER membrane (Lee et al., 2005). This insertion acts as the first step in the membrane
deformation needed for vesicle formation by expanding the cytosolic leaflet of the lipid
bilayer relative to the lumenal leaflet (Lee et al., 2005; Bielli et al., 2005). Like most
RAS family GTPases Sarlp has intrinsically weak GTP hydrolysis activity. Instead, GTP
hydrolysis of Sarlp is a tightly regulated two step process that involves the COPI
subunits Sec23p and Sec3 1p, which will be discussed below.
Membrane deformation and GTP hydrolysis are critical for scission of the
vesicles from the membrane, as replacement of the a-helix with a histidine chain that
allows membrane attachment but not deformation or addition of the nonhydrolyzable
GTP analog GTPyS results in vesicles that are able to form but not release from the
membrane (Lee et al., 2005; Bielli et al., 2005; Bacia et al., 2011). The scission may be a
result of an alteration of membrane curvature upon insertion and removal of the
amphipathic o-helix of Sarlp from the phospholipid bilayer, similar to the mechanism
19
employed by dynamin in clathrin coated vesicle scission (Bacia et al., 2011; Long et al.,
2010).
Sec23p/Sec24p
Once activated and attached to the ER membrane, Sarlp-GTP recruits the
Sec23p/Sec24p subcomplex to form the inner coat of the vesicle (Figure 2 step 2). The
Sec23p/Sec24p subcomplex has a bowtie like structure with a concave surface lined with
basic amino acids that may enhance the interaction of Sec23p/Sec24p with the acidic
phospholipid membrane of the ER (Bi et al., 2002). Although the Sec23p and Sec24p
subunits have similar structures, they do not have the same function. The interaction of
the subcomplex with Sarlp occurs through the Sec23p subunit (Bi et al., 2002; Matsuoka
et al., 1998). Sec23p also acts as a GTPase activating protein (GAP) for the hydrolysis of
Sarlp-GTP by inserting an arginine side chain into the active site of Sarlp, a mechanism
used by many GAPs of RAS family GTPases (Bi et al., 2002). This GAP activity is
further enhanced upon interaction with the Sec31p subunit (Antonny et al., 2001; Bi et
al., 2007).
In addition to its role in regulation of Sarlp activity, Sec23p also plays a role in
positioning of Sec24p in the forming vesicle. Cryo-EM studies have demonstrated that
Sec23p orients the Sec24p subunit such that it is positioned in the open faces of the cages
formed by Sec13p/Sec31p (Bhattacharya et al., 2012). This may be important for
allowing Sec24p to interact with larger cargo proteins.
The Sec24p subunit is responsible for recognition and loading of cargo proteins
into vesicles. Genetic and structural analysis has identified at least three distinct binding
sites for cargo, each interacting with a specific subset of cargo proteins (Miller et al.,
20
2003; Mossessova et al., 2003). S. cerevisiae has two Sec24p paralogs, Lstlp and IssIp,
which may enable a greater diversity of cargo binding (Roberg et al., 1999; Shimoni et
al., 2000; Kurihara et al., 2000). Humans have four isoforms of Sec24p (Wendeler et al.,
2007). This diversity of cargo binding is further increased by the use of cargo adaptor
proteins that link lumenal cargo as well as some membrane proteins to the COPII coats.
Recruitment of Sec23p/Sec24p to the ER membrane by Sarlp, creates a "prebudding" complex which is sufficient for cargo recruitment (Bi et al., 2002). Cargo
binding of Sec23p/Sec24p may play a role in stabilizing the subcomplex on the ER
membrane.
Experiments with reconstituted
membranes
have demonstrated
that
Sec23p/Sec24p remains associated with cargo and therefore with the ER membrane
through multiple rounds of GTP hydrolysis and release of Sarlp, suggesting that cargo
can stabilize Sec23p/Sec24p on the membrane even after Sarlp release (Forster et al.,
2006). This is in contrast to in vitro budding assays in the absence of cargo proteins,
which show that GTP hydrolysis/Sarlp release results in simultaneous release of
Sec23p/Sec24p from the membrane (Forster et al., 2006). Stabilization by cargo content
is supported by fluorescence recovery after photobleaching (FRAP) experiments in vivo
which show that Sec23p and Sec24p have a longer half-life at ERES than Sarlp. This
stabilization of Sec23p and Sec24p is reduced when the amount of transport competent
cargo is decreased (Sato and Nakano, 2005). This process may increase the fidelity and
efficiency of cargo recruitment by releasing pre-budding complexes that are not cargo
associated, allowing them another chance at cargo capture and therefore increasing cargo
concentration into vesicles.
Secl3p/Sec3lp
21
The Sec23p/Sec24p subcomplex recruits the heterotetrameric Sec13p/Sec31p
outer coat subcomplex to the ER membrane (Figure 2 step 3).
This heterotetramer
consists of two molecules each of Sec13p and Sec31p, each of which possess a WD40
domain at their interaction sites (Fath et al., 2007). This subcomplex polymerizes into a
lattice-like network, which deforms the membrane to form vesicles. The Sec13p/Sec31p
complex can self assemble in vitro to form a cage like structure with a cuboctrahedron
architecture (Stagg et al., 2006). Recent studies have shown that Sec31p is the driving
force for vesicle formation, as it is sufficient to form vesicles in the absence of Sec 13p in
vitro. In vivo, Sec13p provides rigidity to forming vesicles to offset the resistance to
membrane curvature added by cargo proteins (Copic et al., 2012). The Sec3lp subunit
also enhances the Sec23p stimulated GTP hydrolysis of Sarlp 10-fold by optimizing the
positioning of Sec23p residues in the catalytic site of Sarlp increasing the rate of Sarlp
release from the ER membrane. This two step process for regulation of GTP hydrolysis
may be important to modulate timing of COPII coat formation and release from vesicles,
ensuring that vesicles only form when cargo is bound and that the coat releases to allow
fusion at the Golgi (Antonny et al., 2001).
Secl6p
Sec 16p was identified in the original screen for secretory genes and has long been
known to play an essential role in vesicle formation. Characterization of Sec 16p has been
hindered by the difficulties in working with such a large protein (-240kD). Secl6p
interacts with all the COPII coat proteins and was initially thought to act as a scaffold for
coat assembly (Gimeno et al., 1996; Shaywitz et al., 1997). Association with Sec 16p also
inhibits GTP hydrolysis of Sarlp by hindering the ability of Sec31p to stimulate Sec23p
22
GAP activity and may therefore regulate the timing of COPII coat disassembly
(Yorimitsu and Sato, 2012). More recent work suggests that Secl6p may play a role in
regulating ERES assembly and therefore control the site of COPII recruitment. How
Secl6p is recruited to the ER is still unclear but may involve direct binding to specific
membrane regions based on variations in phospholipid content (Supek et al., 2002).
Vesicle tethering and fusion
Once released from the ER membrane, COPII coated vesicles travel to the Golgi
membrane or in mammalian cells the ERGIC. When in the vicinity of the Golgi, COPII
vesicles are recognized by the tethering machinery. The Rab GTPase Yptlp (Rab1 in
mammals) coordinates tethering and fusion of COPII coated vesicles, with the TRAPPI
tethering complex acting as a GEF for Yptlp (Cai et al., 2008). TRAPPI binds to Sec23p
on COPII coated vesicles after Sarlp-GTP has been hydrolyzed and released (Cai et al.,
2007). Upon binding to Sec23p, TRAPPI activates Yptlp resulting in the recruitment of
the Usolp (p 115) tethering factor which links vesicles to the Golgi membrane (Cao et al.,
1998; Behnia et al., 2007). Usolp brings the vesicle into close proximity of the Golgi
membrane where the Golgi serine/threonine kinase Hrr25 displaces the TRAPPI complex
on Sec23p. Hrr25p then phosphorylates Sec23p leading to coat disassembly (Lord et al.,
2011). Once the coat is removed the SNARE proteins in the vesicle can pair with SNARE
proteins on the Golgi membrane catalyzing membrane fusion (Lin and Scheller, 1997;
Sutton et al., 1998).
Cargo capture by Sec24
Sec24p cargo binding sites and motifs
23
Biophysical and structural studies of Sec24p have greatly enhanced the
understanding of cargo recognition and loading. The Sec24p subunit has at least three
distinct sites for cargo binding, each interacting with a specific subset of cargo proteins.
The A-site forms a hydrophobic pocket that recognizes a YxxxNPF motif on the SNARE
protein Sed5p (Mossessova et al., 2003).This recognition of Sed5 seems to be yeast
specific as homologs of Sed5 in higher eukaryotes lack this sequence. Later studies
revealed a novel site in the mammalian Sec24C and Sec24D isoforms that recognizes an
IxM motif on the Sed5p homolog syntaxin 5 and membrin (Mancias and Goldberg,
2008).The B-site forms a zinc finger domain, similar to the arginine finger domain on
Sec23p that regulates Sarlp-GTP activity. The B-site has been implicated in binding a
variety of cargo proteins, through multiple ER export motifs including a diacidic
(DxD/E) motif found in Syslp and Gaplp and the LxxL/ME motif of SNARE proteins
Betlp and Sed5p (Mossessova et al., 2003; Miller et al., 2003). The C-site was identified
through mutational analysis and is defective in the capture of the SNARE protein Sec22p
(Miller et al., 2003). Sec22p packaging into COPII coated vesicles is impaired by
mutations in both the B and C sites, and Sed5p is affected by mutations in both the A and
C sites further adding to the complexity of cargo recognition and binding.
Mutational analysis of individual residues within the binding sites of Sec24p has
provided insight into the mechanism and specificity of cargo interaction. Mutations in the
B-site (Sec24R230A, Sec24R235A or Sec24L616W) all decrease the packaging
efficiency of BetIp and Bos 1p while having no effect on packaging of the soluble protein
pro-oc-factor (Miller et al., 2003). Interestingly the B site mutant Sec24L616W disrupted
binding of some diacidic (DxD/E) motif containing proteins, while others were
24
unaffected despite evidence that they are recognized by the B-site (Miller et al., 2003;
Mossessova et al., 2003). This suggests that the diacidic motifs are not all recognized in
the same way. The B-site is conserved among the Sec24p paralogs Lstlp and Isslp.
Similar mutational analysis of the B-site of Lstlp specifically disrupted packaging of
Erplp and Emp24p. The fact that Erplp packaging but not Emp24p packaging is
disrupted by mutations in either Sec24p or Lstlp suggests that there is some overlap in
proteins recognized by the paralogs (Miller et al., 2003).
Some proteins are able to regulate their loading into COPII vesicles via
oligomerization state. ERGIC-53 contains a cytoplasmic di-phenylalanine sorting signal
for binding to COPII, but is only efficiently transported when in a homooligomer (Nufer
et al., 2003). This has also been shown to be the case for the yeast ERGIC-53 homologs
Emp46p and Emp47p (Sato & Nakano 2003). Another example of a requirement of
conformation is the SNARE protein Sed5p, where the Sec24p binding motif becomes
accessible only when Sed5p is in complex with the SNARE proteins Boslp and Sec22p
(Mossessova et al., 2003). This ensures that only SNARE proteins in a complex and
therefore competent for fusion exit the ER.
Cargo Adaptors
While some cargo proteins are able to interact directly with the COPII coat,
soluble proteins and some membrane proteins require the assistance of accessory
proteins. These accessory proteins fall into two categories based on their mechanism of
loading cargo: ER exit chaperones mediate a cargo protein/COPII interaction without
being incorporated into vesicles whereas cargo adaptors mediate the interaction between
cargo and COPII coats by entering into the vesicle themselves.
25
ER exit chaperones play a role in loading of large cargo such as procollagen and
prechylomicrons into COPII vesicles. TANGO 1 is an ER membrane protein that has been
shown to interact with both procollagen and Sec23p/Sec24p and is required for loading of
procollagen into COPII vesicles. TANGO1 along with its binding partner cTAGE5 are
thought to act by binding to Sec23p preventing the recruitment of Sec31p (Saito et al.,
2011). This would slow the hydrolysis of Sarlp enabling a larger coat to form around the
procollagen fibers. Once the vesicle completely engulfs procollagen, TANGO1
dissociates from Sec23p allowing Sec31p binding and thereby vesicle scission. The
release of the COPII coat before vesicle scission means that TANGO 1 remains in the ER
rather than being trafficked to the Golgi (Saito et al., 2011, 2009).
Lumenal and GPI-anchored cargo proteins lacking cytosolic regions are unable to
interact directly with Sec24p and instead rely on the assistance of cargo adaptors which
act as a link between the cargo proteins and Sec24p. The ERV family of cargo adaptors
are multi-spanning transmembrane proteins that recognize a variety of cargo proteins. In
yeast, the ER membrane protein Erv29p interacts with pro-a-factor and is required for it
to be loaded into COPII coats (Belden and Barlowe, 1996). A deletion of ERV29 results
in a significant decrease in pro-a-factor transport to levels similar to bulk flow (Belden
and Barlowe, 1996; Malkus et al., 2002). Additional studies showed that Erv29p is also
involved in loading of other soluble cargo proteins including
vacuolar hydrolases,
carboxypeptidase Y (CPY) and proteinase A, but not all soluble cargo since invertase is
not affected but a deletion of ERV29 (Caldwell et al., 2001). In mammalian cells,
ERGIC-53 and MCFD2 are required together for the trafficking of the coagulation factors
26
V and VIII as well as cathepsin and other glycosylated proteins (Appenzeller et al.,
1999).
The exit of GPI-anchored proteins from the ER relies on members of the p2 4
family of cargo receptors. In yeast, the p24 protein Emp24p is required for loading of the
GPI anchored protein GasIp into COPII vesicles (Mufiiz et al., 2000). Interestingly, it has
been shown that GPI anchored proteins are loaded into vesicles that are distinct from
vesicles containing other cargo proteins and that they can be concentrated into ERES
independently of their interaction with the COPII coat (Mufiiz et al., 2001; Castillon et
al., 2009). These vesicles were shown to contain the Sec24p paralog Lstlp. Members of
the p24 family play a role in trafficking of non-GPI anchored proteins as well, as deletion
of EMP24 results in a delay in transport of the soluble protein invertase from the ER
(Schimm6ller et al., 1995).
Some integral membrane proteins also require the assistance of cargo adaptors for
loading into COPII coats. The yeast bud site selection protein Axl2p cannot be
incorporated into vesicles in the absence of Ervl4p (Powers and Barlowe, 2002). In
Drosophila, the Ervl4p homolog comichon is required for the ER export of Gurkin, a
signaling molecule required to establish polarity during oogenesis (Powers and Barlowe,
2002, 1998). Ervl4p is also required for trafficking of the prospore membrane protein
Sma2p during sporulation (Nakanishi et al., 2007). Erv26p is required for transport of
soluble protein alkaline phosphatase and the transmembrane Ktr3p mannosyltransferase
Ktr3p (Bue and Barlowe, 2009). There are many cargo proteins for which ER exit
methods have yet to be identified. Analysis of cargo proteins could lead to the discovery
of both new exit motifs and new cargo adaptors.
27
Golgi Retrieval
Retrograde trafficking
Movement of cargo proteins through the secretory pathway is countered by a
retrograde trafficking pathway that allows for the retrieval of proteins from the Golgi
back to the ER. Escaped ER resident proteins, misfolded proteins, and proteins involved
in ER to Golgi anterograde transport can all be retrieved from the Golgi by targeting to
Coat Protein Complex I (COPI) coated vesicles. The COPI coat also operates within the
Golgi cisternae transporting enzymes from the late to early cisternae to ensure continual
creation of cis-cistemae and maturation of trans-cistemae. COPI vesicle formation is
initiated by the activation and subsequent Golgi membrane binding of small GTPases of
the ARF family (Popoff et al., 201lb; Yu et al., 2012). The ARF family of GTPases are
activated by a family of guanine nucleotide exchange factors (GEFs) characterized by a
conserved Sec7 domain with ARF1 regulating Golgi to ER retrieval (Jackson and
Casanova, 2000). Specific members of this family reside on various Golgi cisternae and
add a level of specificity to vesicle formation. Membrane associated ARF 1p-GTP recruits
the COPI complex which polymerizes to deform the Golgi membrane forming vesicles
(Kreis et al., 1995). COPI consists of seven subunits: x (RETI),
P (SEC26),
P' (SEC27),
y(SEC21), 8 (RET2), E (SEC28) and ( (RET3) (Waters et al., 1991). Unlike clathrin and
COPII, COPI assembles in the cytosol and binds to Arflp at the membrane en bloc.
Although the coat binds as one complex, there is still an inner and outer coat. The F
subcomplex consisting of the P, 6,y, and ( subunits is structurally and functionally similar
to the clathrin adaptor complexes and is responsible for recruitment to Arflp and cargo
recognition (Yu et al., 2012). The x, P', E subunits form the B subcomplex, with subunits
28
x and P', containing N-terminal WD-40 (f-propeller) repeats and C-terminal a-solenoid
domains similar to the outer coat of clathrin and Sec13p/Sec31p of COPII, which is
responsible for membrane deformation and vesicle formation (Lee and Goldberg, 2010).
Once vesicles are released from the Golgi membrane they are targeted to either the ER or
earlier Golgi cisternae.
Tethering of vesicles to the ER membrane is regulated by the Dsllp protein
complex, composed of the soluble proteins Dsllp, Dsl3p/Sec39p and Tip20p. Dsllp
localized to the ER through interaction with the ER SNARE proteins Ufelp, USElp and
Sec20p and also interacts with the a and E subunits of the COPI coat (Meiringer et al.,
2011). Tethering within the Golgi is dependent on a variety of mechanisms that largely
rely on coiled coil proteins such as p115 and Golgins, which all interact with the COG
complex (reviewed in (Popoff et al., 201la)). In all cases fusion ultimately occurs as a
result of specific SNARE protein bundle assembly (Lin and Scheller, 1997; Sutton et al.,
1998).
Retrieval motifs
Several mechanisms have been identified for COPI mediated exit from the Golgi.
The first retrieval signals identified, HDEL and KKXX, both need to be in the extreme C
terminus of a cargo protein in order to be recognized (Cosson and Letoumeur, 1994;
Majoul et al., 2001). HDEL (or KDEL in mammals) is used by soluble ER resident
proteins such as Kar2p and PDI as well as the type II membrane protein Sec20p
(reviewed in (Majoul et al., 2001; Szul and Sztul, 2011). Type I membrane proteins,
including members of p24 family and Emp47 contain a di-lysine KKXX motif which
enables direct interaction with the COPI coat (Schrdder-K5hne et al., 1998). More
29
recently a di-arginine motif (RR, RXR or RXXR) has been identified. Unlike the HDEL
and KKXX motifs, the RXR motif has been identified in the N terminus, C terminus and
cytosolic loops of proteins (Michelsen et al., 2005). This di-arginine motif interacts
directly with the
P subunit of the COPI coat to enable packaging into vesicles.
Cargo adaptors
Although some proteins contain retrieval signals that allow direct interaction with
subunits of the COPI coat, many require cargo adaptors. The HDEL sequence on the Cterminus of soluble proteins is recognized by the Erd2p receptor, which has a KKXX
motif enabling it to interact directly with the COPI coat (Majoul et al., 2001; Semenza et
al., 1990).
Golgi cisternal maturation requires glycosyltransferases to recycle back to earlier
cisternae. However, they do not have any of the canonical retrieval motifs. The discovery
that the medial Golgi glycosyltransferase Kre2p binds to Vps74p and that this interaction
is required for exit from the Golgi provided insight into a new mechanism for retrograde
trafficking (Schmitz et al., 2008).Vps74p binds to a semi-conserved F/L-L/V/I-X-X-RK
motif in the cytoplasmic N terminus of many Golgi glycosyltransferases (Tu et al., 2008).
Although in theory transmembrane proteins can interact directly with the COPI
coat, not all do. The Golgi membrane protein Rerlp facilitates the retrieval of escaped ER
resident proteins such as Sec12p and Sec71p as well as misfolded or misassembled
proteins (Sato et al., 2003, 2004). The interaction between Rerlp and its cargo occurs
within the transmembrane domains. Interestingly, mutations of Rerlp have been shown
to affect only a subset of its known cargo suggesting that there are multiple recognition
sites within the Rerlp transmembrane region (Sato et al., 2003). Like the HDEL receptor
30
Erd2p, Rerlp contains a di-lysine KKXX motif enabling direct interaction with COPI
(Sato et al., 2003).
Diseases associated with trafficking defects
The secretory pathway mediates the secretion and localization of thousands of
proteins including hormones, growth factors, antibodies, and digestive enzymes. In
humans, this pathway machinery is comprised of more than 2000 proteins (Gilchrist et
al., 2006). Defects in any step of this pathway could have dire effects. Depending on the
mutation, this effect could be restricted to a single or small subset of cargo, or it could
disrupt the entire trafficking system. Over the last few decades there has been a
significant increase in the identification of mutations in protein trafficking machinery as
the cause for many human diseases (reviewed in (Khoriaty et al., 2012; De Matteis and
Luini, 2011)). It is likely that many other diseases will be identified as well. Given the
essential nature of this pathway, it is not surprising that most of the defects identified are
selective to a small subset of cargo proteins as a mutation causing widespread trafficking
defects is likely to be lethal. In addition to defects in trafficking machinery, other diseases
are caused by mutations in specific cargo proteins that interfere with their association
with the trafficking machinery and therefore lead to their mislocalization. These diseases
can result from the lack of a given protein in its proper cellular location, a novel gain of
function in the location it is mislocalized to, or UPR activation due to ER accumulation
as is the case for the dominant form of Charcot-Marie-Tooth disease (Lin and Popko,
2009).
Humans have multiple isoforms of many of the COPII components, and these
isoforms often show cell type and cargo specificity. As a result disruption of one isoform
31
can have cell type specific effects. A missense point mutation in one of the two SEC23
genes, Sec23A is associated with cranio-lenti-sutural dysplasia (CLSD). This Sec23A
mutant is no longer able to recruit Sec 13p/Sec3 1p efficiently and therefore decreases the
GAP activity of Sec23A. Interestingly this is specific to Sar1B, since when SarlA is the
only isoform present the severity of the phenotypes is decreased as Sec13p/Sec31p is
more efficiently recruited (Fromme et al., 2007). Another mutant in Sec23A results in
over activation of SarlB GTPase activity, leading to premature vesicle release from the
ER (Kim et al., 2012). This results in a specific defect in transport of collagen, as vesicles
are released before collagen can be loaded. Mutations in the other Sec23p isoform,
Sec23B, have been implicated in congenital dyserythropoietic anemia type II, supporting
the role of different isoforms in different tissues (Schwarz et al., 2009; Bianchi et al.,
2009).
Mutations in the TRAPPC2 tethering complex leads to the progressive skeletal
disorder, spondyloepiyseal dyslasia tarda (SEDL) (Gedeon et al., 1999). Anderson
disease, also known as chylomicron retention disorder (CMRD), characterized by
malabsorption in lipids from the diet and accumulation of chylomicrons in enterocytes,
has been linked to mutations in Sar1B (Jones et al., 2003).
Defects in trafficking of specific cargo proteins due to mutations in cargo adaptors
have also been identified. Mutations of ERGIC-53 cause bleeding disorders due to a
deficiency in coagulation factor V and VIII in the plasma (Nichols et al., 1998). Over 30
distinct mutations in the LMAN1 gene that encodes ERGIC-53 have been identified
which reduce the coagulation factor levels to 50-95% of normal (Nichols et al., 1998;
Zhang et al., 2003). Additional mutations in the ERGIC-53 binding partner MCFD2 have
32
also been identified in patients with deficiencies in coagulation factor V and VIII (Zhang
et al., 2003; Seligsohn and Ginsburg, 2006). Since the genetic basis of most diseases is
still unknown, it is likely that the secretory pathway is involved in others as well.
The plasma membrane ATPase Pmalp
To facilitate our study of cargo sorting in the ER, we have chosen to study the
plasma membrane protein Pmalp in S. cerevisiae. Pmalp accounts for more than 25% of
the total protein at the plasma membrane and as such is a major cargo of the secretory
pathway (Ghaemmaghami et al., 2003). Pmalp is an H+ ATPase that is a member of the
P-type ATPases, characterized by the formation of an aspartyl-phosphate intermediate as
part of their reaction mechanism (Serrano et al., 1986)
P-type ATPases are a large family of integral membrane transporters that use the
energy of ATP hydrolysis to transport a variety of cations, heavy metals and lipids across
membrane. These ATPases are found in organisms ranging from bacteria to humans and
regulate membrane potential in a range of tissues and organs. They can be subdivided
into 5 classes (P1-P5) based on transport specificity (Axelsen and Palmgren, 1998).
Pmalp is a member of the P3-type ATPase subfamily. All P-type ATPases share a
similar structure with four distinct functional domains. The transmembrane region (M)
has a central core of six a-helices, with varying numbers of additional a-helices (Rice
and MacLennan, 1996). Pmalp has a total of 10 transmembrane spanning regions with
M4, M5, M6 and M8 forming the proton pump (Morsomme et al., 2000; Petrov, 2010).
The cytosolic portion of Pmalp is responsible for ATP hydrolysis and can be subdivided
into 3 functional domains: the nucleotide-binding (N), phosphorylation (P) and actuator
(A) domains (Chourasia and Sastry, 2012). P-type ATPases cycle between two
33
conformational states: the El state is associated with autophosphorylation by ATP and
has
a
high
affinity
for cations
whereas
the
E2
state
is
associated
with
autodephopsporylation and has a low affinity for cations. Binding of cations at the M
domain and ATP at the N domain leads to autophosphorylation of the aspartate residue
within the conserved DKTGT motif of the P domain by transfer of the y-phosphate of
ATP forming an aspartyl phosphoanhydride intermediate. This is followed by a
conformational change in the M domain allowing release of cations on the extracellular
or lumenal face of the membrane. This conformation change also brings the glutamate
of the conserved TGE motif of the A domain into close proximity of the P domain
leading to hydrolysis of the aspartyl phosphate (reviewed in (Chourasia and Sastry, 2012;
Palmgren and Nissen, 2011).
At the plasma membrane, Pmalp couples ATP hydrolysis to proton transport
generating an electrochemical gradient that is required for nutrient uptake as well as
regulation of cellular pH homeostasis and consumes as much as 20% of the total cellular
ATP (Serrano et al., 1986). Pmalp is essential, with at least 30% wild type activity being
required for cell viability (Ambesi et al., 1997). A reduction in Pmalp levels at the cell
surface results in resistance to the antibiotic hygromycin B, sensitivity to growth in low
pH media, and an aberrant multi-budded morphology (Cid et al., 1987; Perlin et al.,
1988).
Pmalp is a 100 kDa protein that forms hexamers and dodecamers within the ER
membrane, and exists as a detergent resistant 1.8 mDa oligomer that associates
specifically with specialized regions of the ER enriched in sphingolipids and ergosterols
known as lipid rafts (Bagnat et al., 2001; Toulmay and Schneiter, 2007). Inhibition of
34
sphingolipid biosynthesis disrupts raft association and oligomerization of Pmalp, both of
which are required for delivery to the cell surface although both monomeric and
oligomeric Pmalp are able to exit the ER (Bagnat et al., 2001, 2000; Toulmay and
Schneiter, 2007).
Proper folding of monomeric Pmalp and oligomer assembly is regulated by the
ER quality control machinery. A number of the mutants created by site directed
mutagenesis result in accumulation in the ER due to misfolding (Morsomme et al., 2000).
Surprisingly, many of these ER blocked mutants display a dominant negative phenotype,
with the most studied being mutations in the P-domain Asp378 residue (Portillo, 1997;
Nakamoto et al., 1998; DeWitt et al., 1998). When expressed with wild type Pmalp, an
D378A mutant causes aggregation and accumulation of both the mutant and wild type
Pmalp protein in the ER (DeWitt et al., 1998; Nakamoto et al., 1998). This aggregation
results from the heterooligomerization of mutant and wild type Pmalp monomers, as a
second site suppressor of the dominant negative D378T mutant resulting in the loss of
transmembrane domain 10 prevents oligomerization and is able to rescue the dominant
lethal phenotype (Eraso et al., 2010).Terminally misfolded Pmalp mutants in the ER are
targeted for degradation by the ERAD pathway, and also trigger the osmostress HOG1
MAPK pathway (Eraso et al., 2011, 2010). Other point mutants such as Pmal-7 and
Pma 1-10 are able to exit the ER, but are targeted for degradation at the Golgi and plasma
membrane respectively, illustrating the multiple checkpoints used to recognize and
degrade misfolded proteins (Gong and Chang, 2001; Chang and Fink, 1995).
Efficient export of Pmalp from the ER requires the Sec24p paralog Lstlp, as cells
lacking Lstlp accumulate Pmalp in the ER and display a severe growth defect on low pH
35
medium (Roberg et al., 1999). Optimal packaging of Pmalp into ER derived vesicles
occurs when both Sec24p and Lstlp are present (Shimoni et al., 2000). In the absence of
Lstlp a 10-fold increase in Sec24p expression is required to restore wild type levels of
Pmalp transport (Shimoni et al., 2000). The mechanism by which Pmalp is packaged
into vesicles is unclear. Truncation of the entire cytosolic C terminus of Pmalp prevents
ER exit, whereas truncation of just the last 18 residues enables transport to the plasma
membrane. However, the Pmalp that reaches the plasma membrane is only partially
active for proton transport and is rapidly targeted for endocytosis (Mason et al., 2006).
The 20 residues just after TM 10 that appear to be required for ER exit do not contain any
of the known ER exit motifs. The work presented here suggests that Pmalp requires a
cargo adaptor, Explp, for loading into Sec24p vesicles, although whether this is true for
Lstlp vesicles as well remains unclear.
36
References
Ambesi, A., K.E. Allen, and C.W. Slayman. 1997. Isolation of transport-competent
secretory vesicles from Saccharomyces cerevisiae. Analytical biochemistry. 251:127-9.
doi: 10.1006/abio. 1997.2257.
Antonny, B., D. Madden, S. Hamamoto, L. Orci, and R. Schekman. 2001. Dynamics of
the COPII coat with GTP and stable analogues. Nat Cell Biol. 3:531-537.
Appenzeller, C., H. Andersson, F. Kappeler, and H.P. Hauri. 1999. The lectin ERGIC-53
is a cargo transport receptor for glycoproteins. Nature cell biology. 1:330-4.
doi: 10. 1038/14020.
Appenzeller-Herzog, C., and H.-P. Hauri. 2006. The ER-Golgi intermediate compartment
(ERGIC): in search of its identity and function. Journal of cell science. 119:2173-83.
doi: 10.1242/jcs.03019.
Araki, K., and K. Nagata. 2011. Protein folding and quality control in the ER. Cold
Spring Harbor perspectives in biology. 3:a007526. doi:10. 101/cshperspect.a007526.
Axelsen, K.B., and M.G. Palmgren. 1998. Evolution of substrate specificities in the Ptype ATPase superfamily. Journal of molecular evolution. 46:84-101.
Bacia, K., E. Futai, S. Prinz, A. Meister, S. Daum, D. Glatte, J.A.G. Briggs, and R.
Schekman. 2011. Multibudded tubules formed by COPII on artificial liposomes.
Scientific reports. 1:17. doi: 10.103 8/srepooo 17.
Bagnat, M., A. Chang, and K. Simons. 2001. Plasma membrane proton ATPase Pmalp
requires raft association for surface delivery in yeast. Mol Biol Cell. 12:4129-4138.
Bagnat, M., S. Kerinen, A. Shevchenko, and K. Simons. 2000. Lipid rafts function in
biosynthetic delivery of proteins to the cell surface in yeast. Proceedings of the National
Academy of Sciences of the United States of America. 97:3254-9.
doi: 10.1073/pnas.060034697.
Bannykh, S.I., T. Rowe, and W.E. Balch. 1996. The organization of endoplasmic
reticulum export complexes. The Journal of cell biology. 135:19-35.
Barlowe, C., L. Orci, T. Yeung, M. Hosobuchi, S. Hamamoto, N. Salama, M.F. Rexach,
M. Ravazzola, M. Amherdt, and R. Schekman. 1994. COPII: a membrane coat formed by
Sec proteins that drive vesicle budding from the endoplasmic reticulum. Cell. 77:895907.
37
Behnia, R., F.A. Barr, J.J. Flanagan, C. Barlowe, and S. Munro. 2007. The yeast
orthologue of GRASP65 forms a complex with a coiled-coil protein that contributes to
ER to Golgi traffic. The Journal of cell biology. 176:255-61. doi:10.1083/jcb.200607151.
Belden, W.J., and C. Barlowe. 1996. Erv25p, a component of COPII-coated vesicles,
forms a complex with Emp24p that is required for efficient endoplasmic reticulum to
Golgi transport. J Biol Chem. 271:26939-26946.
Ben-Tekaya, H., K. Miura, R. Pepperkok, and H.-P. Hauri. 2005. Live imaging of
bidirectional traffic from the ERGIC. Journal of cell science. 118:357-67.
doi:10.1242/jcs.01615.
Benyair, R., E. Ron, and G.Z. Lederkremer. 2011. Protein quality control, retention, and
degradation at the endoplasmic reticulum. International review of cell and molecular
biology. 292:197-280. doi:10.1016/B978-0-12-386033-0.00005-0.
Bhattacharya, N., J. 0 Donnell, and S.M. Stagg. 2012. The structure of the Secl3/31
COPII cage bound to Sec23. Journal of molecular biology. 420:324-34.
doi: 10. 1016/j.jmb.2012.04.024.
Bi, X., R.A. Corpina, and J. Goldberg. 2002. Structure of the Sec23/24-Sarl pre-budding
complex of the COPII vesicle coat. Nature. 419:271-7. doi: 10.103 8/nature0 1040.
Bi, X., J.D. Mancias, and J. Goldberg. 2007. Insights into COPII coat nucleation from the
structure of Sec23.Sarl complexed with the active fragment of Sec31. Developmental
cell. 13:635-45. doi:10.1016/j.devcel.2007.10.006.
Bianchi, P., E. Fermo, C. Vercellati, C. Boschetti, W. Barcellini, A. lurlo, A.P. Marcello,
P.G. Righetti, and A. Zanella. 2009. Congenital dyserythropoietic anemia type II
(CDAII) is caused by mutations in the SEC23B gene. Human mutation. 30:1292-8.
doi: 10.1002/humu.21077.
Bielli, A., C.J. Haney, G. Gabreski, S.C. Watkins, S.I. Bannykh, and M. Aridor. 2005.
Regulation of Sarl NH2 terminus by GTP binding and hydrolysis promotes membrane
deformation to control COPII vesicle fission. The Journal of cell biology. 171:919-24.
doi: 10. 1083/jcb.200509095.
Bue, C.A., and C. Barlowe. 2009. Molecular dissection of Erv26p identifies separable
cargo binding and coat protein sorting activities. The Journal of biological chemistry.
284:24049-60. doi:10.1074/jbc.M109.022590.
Cai, H., S. Yu, S. Menon, Y. Cai, D. Lazarova, C. Fu, K. Reinisch, J.C. Hay, and S.
Ferro-Novick. 2007. TRAPPI tethers COPII vesicles by binding the coat subunit Sec23.
Nature. 445:941-4. doi:10.1038/nature05527.
38
Cai, Y., H.F. Chin, D. Lazarova, S. Menon, C. Fu, H. Cai, A. Sclafani, D.W. Rodgers,
E.M. De La Cruz, S. Ferro-Novick, and K.M. Reinisch. 2008. The structural basis for
activation of the Rab Yptlp by the TRAPP membrane-tethering complexes. Cell.
133:1202-13. doi:10.1016/j.cell.2008.04.049.
Caldwell, S.R., K.J. Hill, and A.A. Cooper. 2001. Degradation of endoplasmic reticulum
(ER) quality control substrates requires transport between the ER and Golgi. The Journal
of biological chemistry. 276:23296-303. doi: 10.1074/jbc.M 102962200.
Cao, X., N. Ballew, and C. Barlowe. 1998. Initial docking of ER-derived vesicles
requires Usolp and Yptlp but is independent of SNARE proteins. The EMBO journal.
17:2156-65. doi:10.1093/emboj/17.8.2156.
Carvalho, P., V. Goder, and T.A. Rapoport. 2006. Distinct ubiquitin-ligase complexes
define convergent pathways for the degradation of ER proteins. Cell. 126:361-73.
doi: 10.10 16/j.cell.2006.05.043.
Carvalho, P., A.M. Stanley, and T.A. Rapoport. 2010. Retrotranslocation of a misfolded
luminal ER protein by the ubiquitin-ligase Hrdlp. Cell.
143:579-91.
doi: 10.1016/j.cell.2010.10.028.
Castillon, G.A., R. Watanabe, M. Taylor, T.M.E. Schwabe, and H. Riezman. 2009.
Concentration of GPI-anchored proteins upon ER exit in yeast. Traffic (Copenhagen,
Denmark). 10:186-200. doi:10. 1111/j.1600-0854.2008.00857.x.
Chang, A., and G.R. Fink. 1995. Targeting of the yeast plasma membrane [H+]ATPase: a
novel gene AST1 prevents mislocalization of mutant ATPase to the vacuole. The Journal
of cell biology. 128:39-49.
Chen, E.J., and C.A. Kaiser. 2002. Amino acids regulate the intracellular trafficking of
the general amino acid permease of Saccharomyces cerevisiae. Proceedings of the
National Academy of Sciences of the United States of America. 99:14837-42.
doi: 10.1073/pnas.232591899.
Chirico, W.J., M.G. Waters, and G. Blobel. 1988. 70K heat shock related proteins
stimulate
protein
translocation
into
microsomes.
Nature.
332:805-10.
doi: 10.1038/332805a0.
Chourasia, M., and G.N. Sastry. 2012. The nucleotide, inhibitor, and cation binding sites
of P-type II ATPases. Chemical biology & drug design. 79:617-27. doi:10. 1111/j.17470285.2012.01334.x.
Cid, A., R. Perona, and R. Serrano. 1987. Replacement of the promoter of the yeast
plasma membrane ATPase gene by a galactose-dependent promoter and its physiological
consequences. Curr Genet. 12:105-110.
39
Conibear, E. 2010. Converging views of endocytosis in yeast and mammals. Current
opinion in cell biology. 22:513-8. doi: 10.101 6/j.ceb.2010.05.009.
Copic, A., C.F. Latham, M.A. Horlbeck, J.G. D'Arcangelo, and E.A. Miller. 2012. ER
cargo properties specify a requirement for COPII coat rigidity mediated by Sec13p.
Science (New York, N.Y.). 335:1359-62. doi:10.1 126/science.1215909.
Cosson, P., and F. Letourneur. 1994. Coatomer interaction with di-lysine endoplasmic
reticulum retention motifs. Science (New York, N.Y.). 263:1629-31.
Credle, J.J., J.S. Finer-Moore, F.R. Papa, R.M. Stroud, and P. Walter. 2005. On the
mechanism of sensing unfolded protein in the endoplasmic reticulum. Proceedings of the
National Academy of Sciences of the United States of America. 102:18773-84.
doi: 10.1073/pnas.0509487102.
DeWitt, N.D., C.F. dos Santos, K.E. Allen, and C.W. Slayman. 1998. Phosphorylation
region of the yeast plasma-membrane H+-ATPase. Role in protein folding and
biogenesis. The Journal of biological chemistry. 273:21744-51.
Eraso, P., M.J. Maz6n, and F. Portillo. 2010. A dominant negative mutant of PMA1
interferes with the folding of the wild type enzyme. Traffic (Copenhagen, Denmark).
11:37-47. doi:10.1111/j.1600-0854.2009.01005.x.
Eraso, P., M.J. Maz6n, F. Posas, and F. Portillo. 2011. Gene expression profiling of
yeasts overexpressing wild type or misfolded Pmal variants reveals activation of the
Hogi MAPK pathway. Molecular microbiology. 79:1339-52. doi:10.1111/j.13652958.2010.07528.x.
Faini, M., R. Beck, F.T. Wieland, and J.A.G. Briggs. 2013. Vesicle coats: structure,
function, and general principles of assembly. Trends in cell biology.
doi:10.1016/j.tcb.2013.01.005.
Fath, S., J.D. Mancias, X. Bi, and J. Goldberg. 2007. Structure and organization of coat
proteins in the COPII cage. Cell. 129:1325-36. doi:10.1016/j.cell.2007.05.036.
Fromme, J.C., M. Ravazzola, S. Hamamoto, M. Al-Balwi, W. Eyaid, S.A. Boyadjiev, P.
Cosson, R. Schekman, and L. Orci. 2007. The genetic basis of a craniofacial disease
provides insight into COPII coat assembly. Developmental cell. 13:623-34.
doi: 10.1016/j.devcel.2007.10.005.
Gardner, B.M., D. Pincus, K. Gotthardt, C.M. Gallagher, and P. Walter. 2013.
Endoplasmic Reticulum Stress Sensing in the Unfolded Protein Response. Cold Spring
Harbor perspectives in biology. 5. doi: 10.110 1/cshperspect.aO13169.
40
Gardner, B.M., and P. Walter. 2011. Unfolded proteins are Ire 1-activating ligands that
directly induce the unfolded protein response. Science (New York, N.Y.). 333:1891-4.
doi: 10.1126/science. 1209126.
Gauss, R., K. Kanehara, P. Carvalho, D.T.W. Ng, and M. Aebi. 2011. A complex of
Pdi Ip and the mannosidase Html p initiates clearance of unfolded glycoproteins from the
endoplasmic reticulum. Molecular cell. 42:782-93. doi: 10.101 6/j.molcel.2011.04.027.
Gedeon, A.K., A. Colley, R. Jamieson, E.M. Thompson, J. Rogers, D. Sillence, G.E.
Tiller, J.C. Mulley, and J. Gecz. 1999. Identification of the gene (SEDL) causing X22:400-4.
Nature
genetics.
dysplasia
tarda.
linked
spondyloepiphyseal
doi:10.1038/11976.
Ghaemmaghami, S., W.-K. Huh, K. Bower, R.W. Howson, A. Belle, N. Dephoure, E.K.
O'Shea, and J.S. Weissman. 2003. Global analysis of protein expression in yeast. Nature.
425:737-41. doi:10.1038/nature02046.
Gilchrist, A., C.E. Au, J. Hiding, A.W. Bell, J. Fernandez-Rodriguez, S. Lesimple, H.
Nagaya, L. Roy, S.J.C. Gosline, M. Hallett, J. Paiement, R.E. Kearney, T. Nilsson, and
J.J.M. Bergeron. 2006. Quantitative proteomics analysis of the secretory pathway. Cell.
127:1265-81. doi:10.1016/j.cell.2006.10.036.
Gimeno, R.E., P. Espenshade, and C.A. Kaiser. 1996. COPII coat subunit interactions:
Sec24p and Sec23p bind to adjacent regions of Sec16p. Mol Biol Cell. 7:1815-1823.
Gnann, A., J.R. Riordan, and D.H. Wolf. 2004. Cystic fibrosis transmembrane
conductance regulator degradation depends on the lectins Htmlp/EDEM and the Cdc48
protein complex in yeast. Molecular biology of the cell. 15:4125-35.
doi: 10.1091 /mbc.E04-01-0024.
Goder, V., and A. Melero. 2011. Protein O-mannosyltransferases participate in ER
protein quality control. Journal of cell science. 124:144-53. doi: 10.1242/jcs.072181.
Gong, X., and A. Chang. 2001. A mutant plasma membrane ATPase, Pmal-10, is
defective in stability at the yeast cell surface. Proceedings of the National Academy of
Sciences of the United States of America. 98:9104-9. doi:10.1073/pnas.161282998.
Hashimoto, H., M. Abe, A. Hirata, Y. Noda, H. Adachi, and K. Yoda. 2002. Progression
of the stacked Golgi compartments in the yeast Saccharomyces cerevisiae by
overproduction of GDP-mannose transporter. Yeast (Chichester, England). 19:1413-24.
doi: 10.1002/yea.925.
Hegde, R.S., and R.J. Keenan. 2011. Tail-anchored membrane protein insertion into the
endoplasmic reticulum. Nature reviews. Molecular cell biology. 12:787-98.
doi:10.1038/nrm3226.
41
Hicke, L., T. Yoshihisa, and R. Schekman. 1992. Sec23p and a novel 105-kDa protein
function as a multimeric complex to promote vesicle budding and protein transport from
the endoplasmic reticulum. Molecular biology of the cell. 3:667-76.
Jackson, C.L., and J.E. Casanova. 2000. Turning on ARF: the Sec7 family of guaninenucleotide-exchange factors. Trends in cell biology. 10:60-7.
Jones, B., E.L. Jones, S.A. Bonney, H.N. Patel, A.R. Mensenkamp, S. EichenbaumVoline, M. Rudling, U. Myrdal, G. Annesi, S. Naik, N. Meadows, A. Quattrone, S.A.
Islam, R.P. Naoumova, B. Angelin, R. Infante, E. Levy, C.C. Roy, P.S. Freemont, J.
Scott, and C.C. Shoulders. 2003. Mutations in a Sarl GTPase of COPII vesicles are
genetics.
34:29-31.
absorption disorders.
Nature
associated
with lipid
doi:10.1038/ng 1145.
Kaiser, C.A., and R. Schekman. 1990. Distinct sets of SEC genes govern transport vesicle
formation and fusion early in the secretory pathway. Cell. 61:723-33.
Kanapin, A., S. Batalov, M.J. Davis, J. Gough, S. Grimmond, H. Kawaji, M. Magrane, H.
Matsuda, C. Schdnbach, R.D. Teasdale, and Z. Yuan. 2003. Mouse proteome analysis.
Genome research. 13:1335-44. doi:10.1101/gr.978703.
Khoriaty, R., M.P. Vasievich, and D. Ginsburg. 2012. The COPII pathway and
hematologic disease. Blood. 120:31-8. doi: 10. 1182/blood-2012-01-292086.
Kim, S.-D., K.B. Pahuja, M. Ravazzola, J. Yoon, S.A. Boyadjiev, S. Hammamoto, R.
Schekman, L. Orci, and J. Kim. 2012. The [corrected] SEC23-SEC31 [corrected]
interface plays critical role for export of procollagen from the endoplasmic reticulum.
The Journal of biological chemistry. 287:10134-44. doi:10.1074/jbc.M 111.283382.
Kreis, T.E., M. Lowe, and R. Pepperkok. 1995. COPs regulating membrane traffic.
Annual
review
of
cell
and
developmental
biology.
11:677-706.
doi:10.1146/annurev.cb.11.110195.003333.
Kurihara, T., S. Hamamoto, R.E. Gimeno, C.A. Kaiser, R. Schekman, and T. Yoshihisa.
2000. Sec24p and Issip function interchangeably in transport vesicle formation from the
endoplasmic reticulum in Saccharomyces cerevisiae. Molecular biology of the cell.
11:983-98.
Lee, C., and J. Goldberg. 2010. Structure of coatomer cage proteins and the relationship
among
COPI, COPII,
and clathrin vesicle
coats.
Cell.
142:123-32.
doi: 10. 1016/j.cell.2010.05.030.
Lee, M.C.S., L. Orci, S. Hamamoto, E. Futai, M. Ravazzola, and R. Schekman. 2005.
Sarlp N-terminal helix initiates membrane curvature and completes the fission of a
COPII vesicle. Cell. 122:605-17. doi: 10.101 6/j.cell.2005.07.025.
42
Lin, R.C., and R.H. Scheller. 1997. Structural organization of the synaptic exocytosis
core complex. Neuron. 19:1087-94.
Lin, W., and B. Popko. 2009. Endoplasmic reticulum stress in disorders of myelinating
cells. Nature neuroscience. 12:379-85. doi:10.1038/nn.2273.
Long, K.R., Y. Yamamoto, A.L. Baker, S.C. Watkins, C.B. Coyne, J.F. Conway, and M.
Aridor. 2010. Sarl assembly regulates membrane constriction and ER export. The
Journal of cell biology. 190:115-28. doi:10.1083/jcb.201004132.
Lord, C., D. Bhandari, S. Menon, M. Ghassemian, D. Nycz, J. Hay, P. Ghosh, and S.
Ferro-Novick. 2011. Sequential interactions with Sec23 control the direction of vesicle
traffic. Nature. 473:181-6. doi:10.1038/nature09969.
Lorente-Rodriguez, A., and C. Barlowe. 2011. Entry and exit mechanisms at the cis-face
of the Golgi complex. Cold Spring Harbor perspectives in biology. 3.
doi: 10. 1101/cshperspect.a005207.
Majoul, I., M. Straub, S.W. Hell, R. Duden, and H.D. S61ing. 2001. KDEL-cargo
regulates interactions between proteins involved in COPI vesicle traffic: measurements in
living cells using FRET. Developmental cell. 1:139-53.
Malkus, P., F. Jiang, and R. Schekman. 2002. Concentrative sorting of secretory cargo
proteins into COPI-coated vesicles. J Cell Biol. 159:915-921.
Mancias, J.D., and J. Goldberg. 2008. Structural basis of cargo membrane protein
discrimination by the human COPII coat machinery. The EMBO journal. 27:2918-28.
doi: 10.1038/emboj.2008.208.
Mason, A.B., K.E. Allen, and C.W. Slayman. 2006. Effects of C-terminal truncations on
trafficking of the yeast plasma membrane H+-ATPase. The Journal of biological
chemistry. 281:23887-98. doi:10.1074/jbc.M601818200.
Matsuoka, K., L. Orci, M. Amherdt, S.Y. Bednarek, S. Hamamoto, R. Schekman, and T.
Yeung. 1998. COPII-coated vesicle formation reconstituted with purified coat proteins
and chemically defined liposomes. Cell. 93:263-75.
De Matteis, M.A., and A. Luini. 2011. Mendelian disorders of membrane trafficking. The
New England journal of medicine. 365:927-38. doi: 10.1056/NEJMraO910494.
McMahon, H.T., and I.G. Mills. 2004. COP and clathrin-coated vesicle budding:
different pathways, common approaches. Current opinion in cell biology. 16:379-91.
doi: 10.101 6/j.ceb.2004.06.009.
Meiringer, C.T.A., R. Rethmeier, K. Auffarth, J. Wilson, A. Perz, C. Barlowe, H.D.
Schmitt, and C. Ungermann. 2011. The Dsll protein tethering complex is a resident
43
endoplasmic
reticulum
complex, which interacts
with
five
soluble NSF
(N-
ethylmaleimide-sensitive factor) attachment protein receptors (SNAREs): implications
for fusion and fusion regulation. The Journal of biological chemistry. 286:25039-46.
doi:10.1074/jbc.M1 10.215327.
Michelsen, K., H. Yuan, and B. Schwappach. 2005. Hide and run. Arginine-based
endoplasmic-reticulum-sorting motifs in the assembly of heteromultimeric membrane
proteins. EMBO reports. 6:717-22. doi: 10.103 8/sj.embor.7400480.
Miller, E.A., T.H. Beilharz, P.N. Malkus, M.C. Lee, S. Hamamoto, L. Orci, and R.
Schekman. 2003. Multiple cargo binding sites on the COPII subunit Sec24p ensure
capture of diverse membrane proteins into transport vesicles. Cell. 114:497-509.
Morsomme, P., C.W. Slayman, and A. Goffeau. 2000. Mutagenic study of the structure,
function and biogenesis of the yeast plasma membrane H(+)-ATPase. Biochimica et
biophysica acta. 1469:133-57.
Mossessova, E., L.C. Bickford, and J. Goldberg. 2003. SNARE selectivity of the COPII
coat. Cell. 114:483-495.
Munfiz, M., P. Morsomme, and H. Riezman. 2001. Protein sorting upon exit from the
endoplasmic reticulum. Cell. 104:313-20.
Muiiz, M., C. Nuoffer, H.P. Hauri, and H. Riezman. 2000. The Emp24 complex recruits
a specific cargo molecule into endoplasmic reticulum-derived vesicles. The Journal of
cell biology. 148:925-30.
Nakamoto, R.K., S. Verjovski-Almeida, K.E. Allen, A. Ambesi, R. Rao, and C.W.
Slayman. 1998. Substitutions of aspartate 378 in the phosphorylation domain of the yeast
PMA1 H+-ATPase disrupt protein folding and biogenesis. The Journal of biological
chemistry. 273:7338-44.
Nakanishi, H., Y. Suda, and A.M. Neiman. 2007. Ervl4 family cargo receptors are
necessary for ER exit during sporulation in Saccharomyces cerevisiae. Journal of cell
science. 120:908-16. doi:10.1242/jcs.03405.
Nichols, W.C., U. Seligsohn, A. Zivelin, V.H. Terry, C.E. Hertel, M.A. Wheatley, M.J.
Moussalli, H.P. Hauri, N. Ciavarella, R.J. KaufmTan, and D. Ginsburg. 1998. Mutations in
the ER-Golgi intermediate compartment protein ERGIC-53 cause combined deficiency of
coagulation factors V and VIII. Cell. 93:61-70.
Novick, P., C. Field, and R. Schekman. 1980. Identification of 23 complementation
groups required for post-translational events in the yeast secretory pathway. Cell.
21:205-15.
44
Nufer, 0., F. Kappeler, S. Guldbrandsen, and H.-P. Hauri. 2003. ER export of ERGIC-53
is controlled by cooperation of targeting determinants in all three of its domains. Journal
of cell science. 116:4429-40. doi:10.1242/jcs.00759.
Ogg, S.C., W.P. Barz, and P. Walter. 1998. A functional GTPase domain, but not its
transmembrane domain, is required for function of the SRP receptor beta-subunit. The
Journal of cell biology. 142:341-54.
Orci, L., M. Ravazzola, P. Meda, C. Holcomb, H.P. Moore, L. Hicke, and R. Schekman.
1991. Mammalian Sec23p homologue is restricted to the endoplasmic reticulum
transitional cytoplasm. Proceedings of the National Academy of Sciences of the United
States of America. 88:8611-5.
Palmgren, M.G., and P. Nissen. 2011. P-type ATPases. Annual review of biophysics.
40:243-66. doi:10.1 146/annurev.biophys.093008.131331.
Papanikou, E., and B.S. Glick. 2009. The yeast Golgi apparatus: insights and mysteries.
FEBS letters. 583:3746-51. doi:10.1016/j.febslet.2009.10.072.
Paroutis, P., N. Touret, and S. Grinstein. 2004. The pH of the secretory pathway:
measurement, determinants, and regulation. Physiology (Bethesda, Md.). 19:207-15.
doi: 10.11 52/physiol.00005.2004.
Perlin, D.S., C.L. Brown, and J.E. Haber. 1988. Membrane potential defect in
hygromycin B-resistant pmal mutants of Saccharomyces cerevisiae. The Journal of
biological chemistry. 263:18118-22.
Petrov, V. V. 2010. Point mutations in Pmal H+-ATPase of Saccharomyces cerevisiae:
influence on its expression and activity. Biochemistry. Biokhimiia. 75:1055-63.
Plemper, R.K., J. Bordallo, P.M. Deak, C. Taxis, R. Hitt, and D.H. Wolf. 1999. Genetic
interactions of Hrd3p and Der3p/Hrdlp with Sec6lp suggest a retro-translocation
complex mediating protein transport for ER degradation. Journal of cell science. 112 ( Pt
2:4123-34.
Popoff, V., F. Adolf, B. Briigger, and F. Wieland. 201 1a. COPI budding within the Golgi
stack.
Cold
Spring
Harbor
perspectives
in
biology.
3:a00523 1.
doi: 10.11 01/cshperspect.a00523 1.
Popoff, V., J.D. Langer, I. Reckmann, A. Hellwig, R.A. Kahn, B. Briigger, and F.T.
Wieland. 2011 b. Several ADP-ribosylation factor (Arf) isoforms support COPI vesicle
formation.
The
Journal
of
biological
chemistry.
286:35634-42.
doi:10.1074/jbc.M 111.261800.
Portillo, F. 1997. Characterization of dominant lethal mutations in the yeast plasma
membrane H+-ATPase gene. FEBS letters. 402:136-40.
45
Powers, J., and C. Barlowe. 1998. Transport of axl2p depends on ervl4p, an ER-vesicle
protein related to the Drosophila cornichon gene product. J Cell Biol. 142:1209-1222.
Powers, J., and C. Barlowe. 2002. Ervl4p directs a transmembrane secretory protein into
COPII-coated transport vesicles. Mol Biol Cell. 13:880-891.
Rapoport, T.A. 2007. Protein translocation across the eukaryotic endoplasmic reticulum
and bacterial plasma membranes. Nature. 450:663-9. doi:10. 1038/nature06384.
Rice, W.J., and D.H. MacLennan. 1996. Scanning mutagenesis reveals a similar pattern
of mutation sensitivity in transmembrane sequences M4, M5, and M6, but not in M8, of
the Ca2+-ATPase of sarcoplasmic reticulum (SERCAla). The Journal of biological
chemistry. 271:31412-9.
Roberg, K.J., M. Crotwell, P. Espenshade, R. Gimeno, and C.A. Kaiser. 1999. LST1 is a
SEC24 homologue used for selective export of the plasma membrane ATPase from the
endoplasmic reticulum. J Cell Biol. 145:659-672.
Rubio-Texeira, M., and C.A. Kaiser. 2006. Amino acids regulate retrieval of the yeast
general amino acid permease from the vacuolar targeting pathway. Molecular biology of
the cell. 17:3031-50. doi:10.1091/mbc.E05-07-0669.
Saito, K., M. Chen, F. Bard, S. Chen, H. Zhou, D. Woodley, R. Polischuk, R. Schekman,
and V. Malhotra. 2009. TANGO1 facilitates cargo loading at endoplasmic reticulum exit
sites. Cell. 136:891-902. doi:10.1016/j.cell.2008.12.025.
Saito, K., K. Yamashiro, Y. Ichikawa, P. Erlmann, K. Kontani, V. Malhotra, and T.
Katada. 2011. cTAGE5 mediates collagen secretion through interaction with TANGO 1 at
endoplasmic reticulum exit sites. Molecular biology of the cell. 22:2301-8.
doi:10.1091/mbc.E 1-02-0143.
Salama, N.R., J.S. Chuang, and R.W. Schekman. 1997. Sec3l encodes an essential
component of the COPII coat required for transport vesicle budding from the
endoplasmic reticulum. Molecular biology of the cell. 8:205-17.
Sato, K., M. Sato, and A. Nakano. 2003. Rerlp, a retrieval receptor for ER membrane
proteins, recognizes transmembrane domains in multiple modes. Molecular biology of the
cell. 14:3605-16. doi:10.1091/mbc.E02-12-0777.
Sato, M., K. Sato, and A. Nakano. 2004. Endoplasmic reticulum quality control of
unassembled iron transporter depends on Rerlp-mediated retrieval from the golgi.
Molecular biology of the cell. 15:1417-24. doi: 10.109 1/mbc.E03-10-0765.
Scheel, A.A., and H.R. Pelham. 1996. Purification and characterization of the human
KDEL receptor. Biochemistry. 35:10203-9. doi:10.1021 /bi960807x.
46
Schimmller, F., B. Singer-KrUger, S. Schrdder, U. KrUger, C. Barlowe, and H. Riezman.
1995. The absence of Emp24p, a component of ER-derived COPI-coated vesicles,
causes a defect in transport of selected proteins to the Golgi. The EMBO journal.
14:1329-39.
Schmitz, K.R., J. Liu, S. Li, T.G. Setty, C.S. Wood, C.G. Burd, and K.M. Ferguson.
2008. Golgi localization of glycosyltransferases requires a Vps74p oligomer.
Developmental cell. 14:523-34. doi: 10.101 6/j.devcel.2008.02.016.
Schrdder-K6hne, S., F. Letourneur, and H. Riezman. 1998. Alpha-COP can discriminate
between distinct, functional di-lysine signals in vitro and regulates access into retrograde
transport. Journal of cell science. 111 ( Pt 2:3459-70.
Schwarz, K., A. Iolascon, F. Verissimo, N.S. Trede, W. Horsley, W. Chen, B.H. Paw, K.P. Hopfner, K. Holzmann, R. Russo, M.R. Esposito, D. Spano, L. De Falco, K. Heinrich,
B. Joggerst, M.T. Rojewski, S. Perrotta, J. Denecke, U. Pannicke, J. Delaunay, R.
Pepperkok, and H. Heimpel. 2009. Mutations affecting the secretory COPII coat
component SEC23B cause congenital dyserythropoietic anemia type II. Nature genetics.
41:936-40. doi: 10.103 8/ng.405.
Seligsohn, U., and D. Ginsburg. 2006. Deciphering the mystery of combined factor V and
factor VIII deficiency. Journal of thrombosis and haemostasis: JTH. 4:927-31.
doi:10.1111/j.1538-7836.2006.01939.x.
Semenza, J.C., K.G. Hardwick, N. Dean, and H.R. Pelham. 1990. ERD2, a yeast gene
required for the receptor-mediated retrieval of luminal ER proteins from the secretory
pathway. Cell. 61:1349-57.
Serrano, R., M.C. Kielland-Brandt, and G.R. Fink. 1986. Yeast plasma membrane
ATPase is essential for growth and has homology with (Na+ + K+), K+- and Ca2+ATPases. Nature. 319:689-693.
Shaywitz, D.A., P.J. Espenshade, R.E. Gimeno, and C.A. Kaiser. 1997. COPII subunit
interactions in the assembly of the vesicle coat. The Journal of biological chemistry.
272:25413-6.
Shimoni, Y., T. Kurihara, M. Ravazzola, M. Amherdt, L. Orci, and R. Schekman. 2000.
Lstlp and Sec24p cooperate in sorting of the plasma membrane ATPase into COPII
vesicles in Saccharomyces cerevisiae. The Journal of cell biology. 151:973-84.
Sidrauski, C., and P. Walter. 1997. The transmembrane kinase Irelp is a site-specific
endonuclease that initiates mRNA splicing in the unfolded protein response. Cell.
90:1031-9.
47
Stagg, S.M., C. Giirkan, D.M. Fowler, P. LaPointe, T.R. Foss, C.S. Potter, B. Carragher,
and W.E. Balch. 2006. Structure of the Secl3/31 COPII coat cage. Nature. 439:234-8.
doi: 10.1038/nature04339.
Stagg, S.M., P. LaPointe, and W.E. Balch. 2007. Structural design of cage and coat
scaffolds that direct membrane traffic. Current opinion in structural biology. 17:221-8.
doi: 10.1016/j.sbi.2007.03.010.
Steel, G.J., J. Brownsword, and C.J. Stirling. 2002. Tail-anchored protein insertion into
yeast ER requires a novel posttranslational mechanism which is independent of the SEC
machinery. Biochemistry. 41:11914-20.
Suda, Y., and A. Nakano. 2012. The yeast Golgi apparatus. Traffic (Copenhagen,
Denmark). 13:505-10. doi:10.1111/j.1600-0854.2011.01316.x.
Supek, F., D.T. Madden, S. Hamamoto, L. Orci, and R. Schekman. 2002. Secl6p
potentiates the action of COPII proteins to bud transport vesicles. The Journal of cell
biology. 158:1029-38. doi:10.1083/jcb.200207053.
Sutton, R.B., D. Fasshauer, R. Jahn, and A.T. Brunger. 1998. Crystal structure of a
SNARE complex involved in synaptic exocytosis at 2.4 A resolution. Nature. 395:34753. doi:10.1038/26412.
Szul, T., and E. Sztul. 2011. COPII and COPI traffic at the ER-Golgi interface.
Physiology (Bethesda, Md.). 26:348-64. doi: 10.1152/physiol.000 17.2011.
Toulmay, A., and R. Schneiter. 2007. Lipid-dependent surface transport of the proton
pumping ATPase: a model to study plasma membrane biogenesis in yeast. Biochimie.
89:249-54. doi: 10.101 6/j.biochi.2006.07.020.
Travers, K.J., C.K. Patil, L. Wodicka, D.J. Lockhart, J.S. Weissman, and P. Walter. 2000.
Functional and genomic analyses reveal an essential coordination between the unfolded
protein response and ER-associated degradation. Cell. 101:249-58.
Tu, L., W.C.S. Tai, L. Chen, and D.K. Banfield. 2008. Signal-mediated dynamic
retention of glycosyltransferases in the Golgi. Science (New York, N.Y.). 321:404-7.
doi:10.1 126/science.1159411.
Vashist, S., and D.T.W. Ng. 2004. Misfolded proteins are sorted by a sequential
checkpoint mechanism of ER quality control. The Journal of cell biology. 165:41-52.
doi: 10.1083/jcb.200309132.
Velasco, A., L. Hendricks, K.W. Moremen, D.R. Tulsiani, 0. Touster, and M.G.
Farquhar. 1993. Cell type-dependent variations in the subcellular distribution of alphamannosidase I and II. The Journal of cell biology. 122:39-51.
48
Wang, F., A. Whynot, M. Tung, and V. Denic. 2011. The mechanism of tail-anchored
protein insertion
into the ER membrane.
Molecular
cell. 43:738-50.
doi:10.1016/j.molcel.2011.07.020.
Waters, M.G., T. Serafini, and J.E. Rothman. 1991. "Coatomer": a cytosolic protein
complex containing subunits of non-clathrin-coated Golgi transport vesicles. Nature.
349:248-51. doi:10.1038/349248a0.
Weinberg, J., and D.G. Drubin. 2012. Clathrin-mediated endocytosis in budding yeast.
Trends in cell biology. 22:1-13. doi:10.1016/j.tcb.2011.09.001.
Wendeler, M.W., J.-P. Paccaud, and H.-P. Hauri. 2007. Role of Sec24 isoforms in
selective export of membrane proteins from the endoplasmic reticulum. EMBO reports.
8:258-64. doi:10.1038/sj.embor.7400893.
Wild, K., M. Halic, I. Sinning, and R. Beckmann. 2004. SRP meets the ribosome. Nature
structural & molecular biology. 11:1049-53. doi:10.1038/nsmb853.
Wooding, S., and H.R. Pelham. 1998. The dynamics of golgi protein traffic visualized in
living yeast cells. Molecular biology of the cell. 9:2667-80.
Yabal, M., S. Brambillasca, P. Soffientini, E. Pedrazzini, N. Borgese, and M. Makarow.
2003. Translocation of the C terminus of a tail-anchored protein across the endoplasmic
reticulum membrane in yeast mutants defective in signal peptide-driven translocation.
The Journal of biological chemistry. 278:3489-96. doi: 10.1074/jbc.M210253200.
Yorimitsu, T., and K. Sato. 2012. Insights into structural and regulatory roles of Sec 16 in
COPII vesicle formation at ER exit sites. Molecular biology of the cell. 23:2930-42.
doi: 10.109 1/mbc.E12-05-0356.
Yoshihisa, T., C. Barlowe, and R. Schekman. 1993. Requirement for a GTPase-activating
protein in vesicle budding from the endoplasmic reticulum. Science (New York, N.Y.).
259:1466-8.
Yu, X., M. Breitman, and J. Goldberg. 2012. A structure-based mechanism for ArflCell.
148:530-42.
membranes.
to
of coatomer
dependent
recruitment
doi: 10.1016/j.cell.2012.01.015.
Zhang, B. 2009. Recent developments in the understanding of the combined deficiency of
FV and FVIII. British journal of haematology. 145:15-23. doi:10.11ll1/j.13652141.2008.07559.x.
Zhang, B., M.A. Cunningham, W.C. Nichols, J.A. Bernat, U. Seligsohn, S.W. Pipe, J.H.
McVey, U. Schulte-Overberg, N.B. de Bosch, A. Ruiz-Saez, G.C. White, E.G.D.
Tuddenham, R.J. Kaufman, and D. Ginsburg. 2003. Bleeding due to disruption of a
49
ER-to-Golgi
cargo-specific
doi:10.1038/ng1153.
transport
complex.
Nature
genetics.
34:220-5.
Zhou, J., C.Y. Liu, S.H. Back, R.L. Clark, D. Peisach, Z. Xu, and R.J. Kaufman. 2006.
The crystal structure of human IREl luminal domain reveals a conserved dimerization
interface required for activation of the unfolded protein response. Proceedings of the
National Academy of Sciences of the United States of America. 103:14343-8.
doi: 10.1073/pnas.0606480103.
50
Figures
Figure 1 Overview of the secretory pathway. Proteins enter the secretory pathway at the
endoplasmic reticulum (ER). Once properly folded and modified, proteins are transported
to the Golgi via COPII coated vesicles (1). Cargo proteins undergo further modifications
as they traverse the Golgi cisternae. Once at the trans-Golgi network proteins can be
delivered to either the plasma membrane (3) or the endosome (4). Retrograde trafficking
via COPI coated vesicles enables recycling of proteins for reuse at prior steps (1, 2, and
4). Proteins can also undergo internalization from the plasma membrane to the endosome
via clathrin coated vesicles (6). At the endosome proteins can either be degraded by the
vacuole (5) or recycled back to the Golgi (4).
51
52
Figure 2 COPII coated vesicle formation. (1) Vesicle formation is initiated by Secl2p
mediated guanine nucleotide exchange on Sarlp. GTP binding to Sarlp causes a
conformation change releasing an amphipathic at-helix, which inserts into the ER
membrane.
(2) Once membrane associated, Sarlp-GTP recruits Sec23p/Sec24p by
interaction with Sec23p. Sec24p recognizes and binds to cargo proteins in the ER
membrane. (3) Sec23p/Sec24p recruits Sec13p/Sec31p, which deforms the membrane
leading to vesicle formation.
53
COPIl Coat Formation
DP
2
1W-
~~~.1+
~
3
'I
rp
u~numi
Legend
Secl2p
*
Sarlp
Wf
Sec23p
%%
Sec3lp
Cargo
Sec24p
......
.....
Chapter 2: Expip is a cargo adaptor
for Sec24p mediated transport of the
plasma membrane ATPase Pmalp in
S. cerevisiae
55
Preface
This chapter represents primarily my own work. Kevin Roberg initially isolated ExpIp as
a low copy suppressor of the lstJ-1 sec13-1 lethality and performed preliminary genetic
analysis of exp1A. Michelle Crotwell-Kirtley performed initial characterization of the
ist1A explA strain. The Pmalp localization microscopy in figures 1B and 2D as well as
the Sec24 pull-downs in figure 7 represent her work.
56
Abstract
The plasma membrane H+ ATPase PMAI in S. cerevisiae is essential for creating
an electrochemical gradient across the plasma membrane required for nutrient uptake and
maintenance of the cellular pH. We previously reported that Pmalp requires the Sec24p
paralog Lstlp for efficient export from the ER. Here we report the identification of a
novel gene, EXP1 (_R
export of Pmalp) which when overexpressed can restore
localization of Pmalp to the plasma membrane in an ist1A strain. Strains carrying a
deletion of EXP1 grow normally, but an ist1A explA double mutant is inviable,
displaying a severe Pmalp trafficking defect. EXP1 encodes a 17 kDa integral membrane
protein which cycles between the ER and Golgi. Recombinant Explp binds to the
Sec23p/Sec24p COPII subcomplex and Explp can be co-immunoprecipitated with
Pmalp. Mutagenic analysis identified specific regions of the cytosolic C terminus of
EXPI that fail to complement the lethality of an ist1A exp1A double mutant, prevent ER
exit and eliminate binding of Sec23p/Sec24p and that the transmembrane region is
required for interaction with Pmalp. Taken together, our data suggest that Pmalp exits
the ER via two independent pathways. Lstlp containing vesicles mediate the primary
route while Explp is required as a cargo adaptor for the secondary Sec24p mediated
route.
57
Introduction
In eukaryotic cells, the secretory pathway is responsible for the folding,
modification and delivery of plasma membrane and secreted proteins. The secretory
pathway is an ordered series of organelles that includes the endoplasmic reticulum (ER),
Golgi, endosome and vacuole (Palade, 1975). Approximately 30% of all proteins traffic
through the secretory pathway (Huh et al. 2003). Vesicle budding from the ER requires
the COPII protein coat complex, composed of the heteromeric
subcomplexes,
Sec23p/Sec24p and Secl3p/Sec3lp, and a cytosolic GTPase Sarlp (Barlowe et al.,
1994). The budding process initiates at specific sites on the ER membrane called ER exit
sites (ERES), when the ER localized guanine nucleotide exchange protein Secl2p
catalyzes exchange of the guanine nucleotide bound to Sarlp to generate Sarlp-GTP.
Sarlp-GTP binds to the ER membrane via an amphipathic x-helix that is exposed upon
GTP binding (Bielli et al., 2005; Lee et al., 2005). Once Sarlp-GTP is membrane
associated it recruits the Sec23p/Sec24p protein complex, which then recruits the
Sec13p/Sec31p protein complex. The Sec13p/Sec31 subcomplex polymerizes into a
lattice like network promoting curvature of the lipid membrane to allow vesicle formation
(Stagg et al., 2006). Hydrolysis of Sarlp-GTP by Sec23p/Sec31p upon tethering at the
Golgi aids in coat disassembly allowing vesicle fusion with the Golgi membrane
(Antonny et al., 2001; Bi et al., 2007).
Secretory cargo proteins concentrate into COPII vesicles at the earliest stages of
vesicle formation associating with the Sec23p/Sec24p/Sarlp-GTP pre-budding complex
(Belden and Barlowe, 1996). The Sec24p subunit of the COPII coat is responsible for
recognition and binding of properly folded cargo proteins in the ER via exposed ER
58
export motifs (Miller et al., 2002). Mutational studies of SEC24 combined with structural
analysis of Sec23p/Sec24p bound to peptides derived from cargo proteins show that
Sec24p binds to multiple ER export motifs through at least three spatially distinct sites
(Miller et al., 2003; Mossessova et al., 2003). S. cerevisiaehas two additional paralogs of
SEC24, ISS1
and LST1 which may allow for binding of an even more diverse range of
cargo proteins as the different paralogs appear to transport distinct but overlapping cargo
(Roberg et al., 1999; Shimoni et al., 2000; Kurihara et al., 2000).
Although integral membrane proteins may interact directly with the COPII coat
proteins, evidence suggests that some integral membrane proteins as well as soluble
proteins require accessory proteins called cargo adaptors in order to be incorporated into
the budding vesicle. Cargo adaptors act as a linker between Sec24p and the cargo protein
and are incorporated into vesicles while actively recruiting integral membrane proteins
and cytosolic proteins into COPII vesicles. Examples of this type of cargo adaptor are
Ervl4p, which interacts specifically with the polytopic bud selection protein Axl2p and
the COPII coat escorting Axl2p into the budding vesicle and the p24 family of cargo
adaptors that are required for loading of GPI anchored proteins into COPII vesicles
(Munfiz et al., 2000; Castillon et al., 2009; Powers and Barlowe, 2002, 1998). Despite the
discovery of a number of cargo adaptor proteins the method by which most secretory
proteins exit the ER remains unclear.
To facilitate our study of the mechanisms of protein sorting, we have selected
Pmalp, the plasma membrane H+ ATPase of S. cerevisiae, as a model cargo protein.
Pmalp is a member of the P-type ATPase family which generates and maintain electrochemical gradients across cellular membranes by translocating cations, heavy metals and
59
lipids (Serrano et al., 1986; Catty et al., 1997). Pmalp is a polytopic plasma membrane
protein that uses energy derived from ATP hydrolysis to pump protons across the plasma
membrane, a process that establishes the proton gradient necessary for the uptake of
nutrients from the extracellular medium as well as maintenance of intracellular pH
homeostasis (Ambesi et al., 2000). Cells partially defective for Pmalp grow poorly on
acidic medium, presumably because protons generated by metabolism are not efficiently
expelled from the cytosol. This results in the acidification of the cell, which is detrimental
to the organism.
Previously, we reported that efficient trafficking of Pmalp from the ER requires
the Sec24p paralog Lstlp (Roberg et al., 1999). Strains with a deletion of LST1 are
viable, but display phenotypes attributable to a partial defect in Pmalp localization to the
plasma membrane, including sensitivity to low pH and decreased proton pumping
activity.
In vitro analysis confirmed that both Sec24p and Lstlp are required to
efficiently package Pmalp into COPII vesicles as in the absence of Lstlp, a 20-fold
excess of Sec24p is required to attain wild type levels of Pmalp vesicle incorporation
(Shimoni et al., 2000). During the initial isolation of the LST1 gene, we identified a
second gene not linked to LST1, that can efficiently suppress some of the growth defects
of an isti-1 temperature sensitive mutant (Roberg et al., 1999). Here we show that this
novel suppressor gene, named EXP1 (ER-export ofPmalp), participates in the selective
export of Pmalp from the ER and may act as a cargo adaptor enhancing Pmalp entry into
COPII vesicles through specific interaction with the Sec23p/Sec24p subcomplex.
60
Materials and Methods
Media, Strains,and Plasmids
S. cerevisiae strains used in this study are listed in Table I. Cells were grown in
either rich medium (YPD) or supplemented minimal medium (SMM), as described in
(Kaiser et al., 1994). For growth under acidic conditions, YPD or SMM media was
adjusted to pH 3.0 with HCl. pMC18, which contains EXP1 expressed from the GAL10
promoter, was constructed by ligating a PCR amplified EXP1 fragment from pKR10 into
pCD43 (HIS3). pDM12 was made by ligating a PCR amplified EXP1 fragment from
CKY8 (wild type) genomic DNA into pRS316 (URA CEN).
Immunoblotting
Proteins were solubilized in sample buffer (2% SDS, 10 % glycerol, 80 mM TrisHCl pH 6.8, 0.1 mg/ml bromophenol blue, 100 mM DTT) and resolved by SDS-PAGE
according to standard protocols. For transfer of Gaslp methanol was omitted from the
transfer buffer. Rabbit anti-Pmalp (a gift of A. Chang, University of Michigan) was used
at a dilution of 1:500. Mouse anti-Pmalp was used at a dilution of 1:1000 (Abcam).
Mouse monoclonal anti-Dpmlp (Invitrogen) was used at a dilution of 1:250. Rabbit antiSec23p (a gift of L. Hicke, Northwestern University) and anti-Sec24p (gift of T.
Yoshihisa, Nagoya University, Japan) were used at 1:500. HA-Sec3lp and HA-Pmalp
was detected using mouse monoclonal 12CA5 anti-HA at 1:1000. Rabbit anti-Explp,
anti-Gas 1, anti-Pdi l p and anti-Ero 1p were used at a dilution of 1:1000. HRP conjugated
anti-rabbit and anti-mouse were used at 1:10,000 (GE Healthcare).
Microscopy
61
Digital photomicrographs were taken with a Hamamatsu Digital camera attached
to a Nikon Eclipse E800 microscope and visualized with OpenLab software
(Improvision). Cell morphology was examined using DIC microscopy of cells treated
with 3.7% formaldehyde. Aberrant, multi-budded morphology was quantitated by
counting the percentage of cells with three or more buds. 600 cells were examined for
each strain and the percentages shown represent the average of two separate experiments.
Localization of Pmalp by immunofluorescence was performed as described in
Roberg, et al., 1999. Anti-Pmalp was affinity-purified as described in Roberg, et al, 1999
and used at a 1:5 dilution. Alexa-conjugated anti-rabbit IgG (Molecular Probes) was used
at a dilution of 1:200. Mounting medium was supplemented with 4', 6-diamidino-2phenylindole (DAPI).
Cellfractionation
Subcellular distribution of Explp was examined as follows. Wild-type cells (CKY
8) were grown to exponential phase in YPD, converted to spheroplasts, and gently lysed
by agitation with glass beads in lysis buffer (50 mM Tris-HCl pH7.5, 1mM EDTA, 200
mM sorbitol) containing protease inhibitors. The cell extract was sequentially centrifuged
at 500 x g for 10 minutes, 13,000 x g for 10 minutes, and 100,000 x g for 30 minutes.
Proteins solubilized in sample buffer were separated with SDS-PAGE and analyzed by
immunoblotting with antiserum to Explp, PGK, or Dpmlp.
Release of Exp 1p from the particulate/membrane bound fraction was performed
as described in Kaiser, et al., 2002 with the following modifications. CKY8 cells growing
logarithmically in YPD were harvested and converted to spheroplasts. Spheroplasts were
lysed on ice by douncing in lysis buffer plus protease inhibitors. After a clearing spin (5
62
minutes at 500 x g), cell extracts were treated with either 100 mM Na 2CO 3 pH 11.5; 500
mM NaCl; 2.5 M urea; 1% Triton; or buffer alone and incubated on ice for 1 hour.
Treated extracts were centrifuged at 100,000 x g for 75 minutes to separate membrane
bound and soluble proteins. Samples were solubilized in sample buffer and analyzed by
immunoblotting with antiserum to ExpIp or Dpmlp.
Sucrose density gradient fractionation was performed as described in (Kaiser et
al., 2002). The presence of Pmalp, Gasllp, Explp, Erolp and Dpmlp in each fraction
was detected by SDS-PAGE followed by immunoblotting. The relative amounts of each
protein in cell fractions were determined using an Image Station 440CF (Kodak Digital
Sciences) and 1D Image Analysis Software (Kodak Digital Sciences). The presence of
Golgi GDPase activity was detected enzymatically as described in (Kaiser et al., 2002).
Protease accessibilityof Expip
Wild type (CKY8) cells were converted to spheroplasts, lysed in lysis buffer (250
mM sorbitol, 150 mM potassium acetate, 20 mM HEPES pH 6.8, 1 mM magnesium
acetate) by Dounce homogenization on ice to create microsomes, and cell debris was
removed by centrifugation at 500 x g for 5 minutes. Microsomal membranes were
collected by centrifugation of cleared cell extracts for 15 minutes at 13,000 x g and
treated with 0.75 mg/ml Trypsin +/- 1% Triton X-100. Proteolysis was terminated with
25 mM Pefabloc (Roche) after 5, 15, and 30 minutes. Proteins were separated by SDSPAGE and analyzed by immunoblotting with antiserum to ExpIp or PDI.
EXPJ Mutagenesis and Chimera Construction
Deletion mutations in EXP1 were created by PCR amplification of the flanking
regions on the 5' and 3' sides of the deletion in separate reactions introducing restriction
63
sites into the 3' end of the 5' fragment and the 5' end of the 3' fragment. For generation
of the deletion mutation in pMC3 (GST-EXP1) an EcoRI site was added and the two
PCR fragments were ligated into pGEX5-3. A similar strategy was used to generate the
same deletion mutation in the yeast vector pRS316 (EXP1 CEN).
The two PCR
fragments were ligated together using a BamHI site, and the resulting deletion was
introduced into pDM12 by subcloning an EcoRI/MscI fragment from the deletion
construct into pDM12. Presence of the deletion was confirmed by sequence analysis.
Alanine mutations were generated using the Stratagene Quik-Change PCR mutagenesis
kit with pDM 12 as the template DNA.
Chimeric proteins were constructed using a two step overlapping PCR reaction. A
pair of reverse complement primers were designed at the fusion site such that each primer
contained DNA sequence flanking each side of the fusion site. In the first round of PCR
each of these primers was used with a primer to the upstream and downstream UTR of
the corresponding genes creating two PCR fragments which each contained the fusion
site and flanking sequence. The two PCR products were mixed and PCR amplified in the
absence of primers for 5 cycles, followed by 30 cycles with the 5' or 3' UTR primers to
amplify the full length chimeric DNA sequence. Chimera constructs were confirmed by
sequence analysis.
Binding of Explp to Sec23p/Sec24p
EXP1 was fused to the glutathione-S-transferase gene under the control of the lac
promoter in the E. coli expression vector pGEX2-T (Pharmacia). Purification was
performed as described in the GE Healthcare GST Gene Fusion System Handbook batch
purification protocol. Wild type and mutant Explp fusion proteins were purified from
64
BL21 (DE3).
Fusion proteins were induced with 1 mM IPTG for 5 hours at 37'C,
suspended in PBS, and frozen at -80'C. Bacterial cells were thawed, lysed by sonication,
and membranes were solubilized by addition of 1%Triton. Cell debris was collected by
centrifugation at 12,000 x g for 10 minutes, and the supernatant was added to glutathionesepharose beads (Pharmacia). After incubation at 24'C for 1 hour, beads were washed
three times in PBS and resuspended in PBS. The quantity of protein bound to
glutathione-sepharose beads was estimated by BCA protein assays (Thermo Scientific),
and the purity of the protein preparation was assessed by SDS-PAGE followed by
Coomassie staining.
For protein-binding assays, yeast cytosolic extracts were prepared as described in
(Gimeno et al., 1996). Exponentially growing cells from pep4A or SEC31-HA strains
were harvested and lysed in immunoprecipitation buffer (20 mM HEPES pH 6.8, 80 mM
potassium acetate, 200 mM NaCl, 0.02% Triton X-100, 5 mM DTT, 1 mM EDTA)
containing protease inhibitors. Cell debris and intracellular membranes were removed by
centrifugation at 13,000 x g and 100,000 x g. Total protein concentration was determined
by Bradford assay. Purified GST-Explp fusion proteins bound to sepharose beads were
added to the cell extracts and incubated for 1 hour at 4'C. The beads were washed four
times in immunoprecipitation buffer and resuspended in 30 pl of sample buffer for
analysis by SDS-PAGE. COPII proteins were detected by immunoblotting with anti-HA
12CA5 antibody or antiserum to Sec23p or Sec24p.
Phenotypic analysis of istlA explA strains
IstIA and explA strains expressing either EXPJ from the GAL1O promoter or
EXP1 in pRS316 (URA CEN) were mated, and diploids were sporulated and dissected on
65
galactose-containing
medium
or SMM-Ura
medium. For growth
analysis and
microscopy, PGALJ-EXPJ istlA explA strains grown in galactose medium were
transferred to raffinose containing medium for 8 hours prior to the addition of glucose.
To show synthetic lethality of ist1A exp1i, the ist1A exp1A plus pDM12 (EXP1 URA
CEN) cells were spotted onto SMM medium containing 5-fluorooratic acid (5-FOA) and
grown for 3 days at 24'C.
PmaJ Immunoprecipitation
Ist1A cells containing pXZ33 (Gall::HA-PMAl-LEU2) (a gift of James Haber,
Brandeis University), and Expl mutants in pRS426 (URA 2g) were grown in raffinose
containing media and shifted to galactose for 3 hours at 24'C to induce HA-Pmalp.
Harvested cells were converted to spheroplasts, and washed in B88 (20 mM HEPES pH
6.8, 150 mM KOAc, 250 mM sorbitol, 5 mM Mg [OAc]
2).
Spheroplasts were collected
by spinning 5 minutes at 2500 x g and resuspended in IP buffer (10 mM Tris-HCl pH 7.6,
150 mM NaCl, 2 mM EDTA, 1% Triton) plus protease inhibitors. Membranes were
solubilized 1 hr at 24'C. Unsolubilzed membranes were removed by spinning at 13,000 x
g for 5 minutes. Anti-HA 3F10 was added to solubilized membranes at 4'C overnight.
Samples were incubated for 1 hour at 24'C after addition of protein G sepharose beads,
washed twice in IP buffer, once in detergent free wash buffer (150 mM NaCl, 50 mM
Tris-HCl pH 7.4), and resuspended in sample buffer. Samples were solubilized at 37'C
for 1 hour, and analyzed by immunoblotting with antiserum to Exp lp or HA (12CA5).
66
Results
Isolation of EXPJ
LST1 (lethal with sec thirteen) was identified in a genetic screen for mutations
that were lethal in combination with the COPII temperature sensitive mutation sec]31(Roberg et al., 1999). The LST] gene was isolated by screening for clones that could
restore viability to a sec13-1 1st1-1 strain grown at the permissive temperature. We also
identified a second complementing locus unrelated to LST1. The genomic region from
clone p21-31, a representative of this second complementing locus, was inserted into an
integrating vector (pRS306), linearized, and integrated at its chromosomal locus into
CKY472, which contains wild-type copies of SEC13 and LST1. The resulting integrant
was crossed to a sec13-1 lst]-1 strain, and the diploids were sporulated and dissected.
Tetrad analysis showed that the locus containing the inserted plasmid was not linked to
LSTJ and, therefore, encoded an extragenic suppressor of the lethality displayed by
sec13-1 lst]-1 strains at the permissive temperature (data not shown). We sequenced the
p21 -31 clone and isolated the suppressing locus as yDL121c, which we have named
EXP1 (ER-export of_malp).
Overexpression of EXP1 restores plasma membrane localization of Pmalp in an
istlA mutant
In the absence of LSTJ, a decreased delivery of Pmalp to the cell surface causes
yeast cells to grow poorly on acidic medium (pH_<4.0), particularly at high temperatures
(Roberg et al., 1999). Ectopic expression of EXP] from a centromeric plasmid restored
the ability of lst1A strains to grow at pH < 4.0 (Fig. IA). To test whether overexpression
of EXP1 can also suppress the Pmalp trafficking defect of ist1A strains, localization of
67
Pmalp was examined by immunofluorescence microscopy. In ist1A cells, Pmalp staining
is observed in the ER at the nuclear periphery as well as at the cell surface. As shown in
Figure 1B, overexpression of EXP1 completely restored cell surface localization of
Pmalp in an ist1A genetic background.
As an independent test for Pmalp localization, extracts from ist1A strains
ectopically expressing EXP1 were fractionated on a linear sucrose density gradient under
conditions that separate the ER and Golgi membrane from the denser plasma membrane.
In wild type cells all of the Pmalp is located in the denser plasma membrane fractions,
whereas in ist1A cells more than half of the total Pmalp fractionates with the ER marker
Dpmlp, indicative of a delay in export from the ER (Fig. 1C). In ist1A strains ectopically
expressing EXP1from a centromeric plasmid, all the detectable Pmalp cofractionated
with the cell surface marker Gaslp (Fig. IC). On the basis of these localization
experiments, we conclude that EXP1 ectopic overexpression suppresses the low pH
growth sensitivity of an ist1A strain by overcoming the requirement for Lstlp in the
export of Pmalp from the ER.
ist1A exp1A double mutants exhibit a severe defect in Pmalp trafficking
Unlike ist1A strains, exp1A strains grew normally on acidic medium (pH < 4.0)
and efficiently exported Pmalp to the cell surface as demonstrated by cell fractionation
and immunofluorescence experiments (Fig. 2D and data not shown). However, ist1A
explA double mutants are inviable at all temperatures, as they are unable to grow in the
absence of an EXPI containing plasmid demonstrated by the inability of ist1A exp1A to
grow on medium containing 5-FOA (Fig. 2A). To determine the severity of the Pmalp
trafficking defect in the ist1A explA double mutant strain, we conditionally expressed
68
EXP1 under the control of the GAL10 promoter in ist1A exp1A cells and examined the
localization of Pmalp in cells depleted of Explp. Within six hours after addition of
glucose, which represses expression of PGALIO-EXP1, Explp could no longer be detected
in PGALJ-EXP1 ist1A exp1A strains as determined by immunoblotting with anti-Explp
(data not shown). After growth in glucose for 14 hours, a PGALIO-EXP1 ist1A exp1A strain
exhibited a significant reduction in growth rate indicating that the cells were no longer
growing exponentially (Fig. 2B). The time lag between full depletion of Explp from
PGALI-EXP1 ist1A
exp1A cells and the cessation of exponential growth corresponded to
about 8 hours. The estimated half life of Pmalp is about 11 hours, so the lag between loss
of Explp and cessation of exponential growth is likely the time it takes for the level of
Pmalp at the cell surface to fall below the 30% of wild type needed for viability (Bagnat
et al., 2001). These observations suggest that upon depletion of Exp Ip, newly synthesized
Pmalp cannot be delivered to the cell surface, and cells cease dividing as pre-existing
Pmalp at the cell surface is endocytosed and degraded.
A striking phenotype of cells defective in Pmalp activity at the cell surface is the
presence of cells exhibiting an aberrant, multi-budded morphology (Fig. 2C) (Cid et al.,
1987). In cells depleted of Pmalp at the cell surface the mother cell is able to continue
budding until the Pmalp already at the cell surface is endocytosed. But the daughter cells
are unable to traffic Pmalp to the cell surface, and therefore are inviable and remain
associated with the mother cell presumably because of a defect in cell wall formation or
cell separation. Examination of PGAL1-EXPJ ist1A exp1A strains by phase contrast
microscopy after incubation in glucose minimal medium for 14 hours revealed that more
than 30% of these cells displayed a multi-budded morphology compared to only 3% of
69
ist1A cells and 1% of wild-type cells grown under the same conditions (Table II).
Immunofluorescence microscopy was used to examine the location of Pmalp in
PGALIO-
EXP1 ist1A exp1A mutants. After 14 hours of growth in glucose, Pmalp staining could
not be detected in the nascent buds, and in mother cells Pmalp was visible only in ER at
the nuclear periphery, not the plasma membrane (Fig. 2D). (Because the buds of the
multi-budded cells have completed cytokinesis, the dramatic multi-budded morphology
of these strains was not preserved after digestion of the cell wall in preparation for
microscopy (Roberg et al., 1999).) Together, these results indicate that ist1A exp1A
double mutants are severely defective for Pmalp trafficking out of the ER. The residual
slow growth of PGAL-EXP1 1stlA explIA cells depleted of ExpIp can be explained by the
presence of Pmalp that had been delivered to the plasma membrane prior to depletion of
Explp.
Expip is a type Ib integral membrane protein
EXP1 encodes a 149 amino acid protein with a sequence of hydrophobic amino
acids at the N terminus, predicted to serve either as a transmembrane domain or a signal
sequence, and a highly charged C terminus. As an experimental test of the intracellular
distribution of Exp1p, lysates from wild-type cells were converted to spheroplasts and
subjected to a series of centrifugation steps designed to separate soluble proteins from
those that are membrane bound. At both the 13,000 x g and 100,000 x g centrifugation
steps, Explp was located in the pellet along with the membrane bound marker Dpmlp,
indicating that Explp is tightly associated with cellular membranes (Fig. 3A).
To confirm that Explp is an integral membrane protein, we subjected
spheroplasted cell lysates to chemical treatments prior to centrifugation at 100,000 x g.
70
Incubation of cell lysates with high salt (500 mM NaCi), high pH (100 mM Na 2CO 3 pH
11.5), or 2.5 M urea, treatments known to perturb the interaction of peripheral membrane
proteins with cell membranes, had no effect on the association of Exp 1p with the 100,000
x g insoluble fraction. In contrast, treatment with 1% Triton was sufficient to release
ExpIp into the soluble fraction (Fig. 3B). Thus, ExpIp is an integral membrane protein.
The TopPred algorithm for membrane topology prediction, which relies on the
"positive-inside" rule for charged residues flanking the transmembrane domain, predicted
that Explp is a type lb membrane protein whose C terminus localizes to the cytosol
(Claros and Von Heijne, 1994; von Heijne, 1992). The orientation of Explp was
experimentally determined by testing the accessibility of Explp in microsomes to
digestion with Trypsin. Explp in intact microsomal membranes was efficiently degraded
by Trypsin, suggesting that the C terminus of Explp is accessible to protease and,
therefore, is oriented to the cytosolic face of the membrane (Fig. 3C). As a control for the
integrity of the microsomes under these conditions, we also examined the protease
accessibility of the lumenal ER chaperone PDI and found that PDI remained stable and
not accessible to protease unless the membranes had first been solubilized by detergent
(Fig. 3C). Together, these results demonstrate that Explp is a type lb integral membrane
protein, anchored to the membrane through its N-terminal transmembrane domain with
the C terminus facing the cytosol.
Expip cycles between the ER and Golgi
To determine the subcellular localization of Explp we fractionated cell
membranes on sucrose density gradients. In the presence of magnesium, ribosomes
remain associated with ER membranes, causing ER membranes to sediment with the
71
dense fractions near the bottom of the gradient. In contrast, in the presence of EDTA
magnesium is chelated and ribosomes are released from the ER, such that the ER
membranes equilibrate at a relatively low density near the top of the gradient (Sanderson
and Meyer, 1991). More than 75% of Explp co-fractionated with the ER marker protein
Dpmlp in magnesium and in EDTA containing buffers, indicating that most of ExpIp is
localized to the ER (Fig. 4A). Explp also partially co-fractionated with Golgi marker
proteins, suggesting that at steady state a small portion of Explp may localize to the
Golgi.
Colocalization of a small proportion of Explp with Golgi marker proteins
suggested that Explp might be actively cycling between the ER and the Golgi. As an
experimental test of the cycling of Explp, we compared the subcellular localization of
Explp in wild type strains and in strains expressing a mutant version of the COPI
component Sec21p. In cells expressing sec21-1 at the semi-permissive temperature
(30'C), retrograde transport from the Golgi to the ER is blocked while anterograde
transport from the ER to the Golgi remains unaffected (Gaynor and Emr, 1997). Under
these conditions, proteins that are normally recycled from the Golgi back to the ER
accumulate in the Golgi, in later secretory compartments such as the vacuole, or are
secreted from the cell. Cell extracts from wild type and sec21-1 strains grown at a semipermissive temperature (30'C) were fractionated on a linear sucrose gradient in buffer
containing magnesium, separating Golgi and vacuolar membranes (low density) from the
ER and plasma membrane (high density). In sec21-1 cells, there is a large shift of Explp
to co-fractionate with the lower density fractions along with the Golgi marker GDPase,
whereas in wild type cells, most Explp co-fractionated with the higher density ER
72
marker (Fig. 4B). Thus, steady-state localization of ExpIp in the ER requires ExpIp to be
recycled from the Golgi in a COPI dependent manner.
The cytosolic C terminus is required for Expip function
Explp is homologous to predicted proteins in several related yeast species (Fig.
5). In order to determine regions of EXP 1 required for function, we deleted the regions of
Expip most highly conserved among the Explp orthologs, creating a series of five short
deletions together spanning most of the cytosolic domain (Fig. 5). The functionality of
each deletion mutation was tested by its ability to support growth of a PGAL/-EXP1 lst1
explA strain on glucose medium. Deletion alleles explA86-102, expJA107-119 and
exp1A121-149 did not suppress the lst1A exp1z lethality despite stable expression of their
protein products, while exp1A49-58 and explA61-68 suppressed ist1A exp1A lethality as
well as wild-type EXP1 (Fig. 6A).
We examined the effect of each of the EXP1 deletion alleles on the prevalence of
the multi-budded rosette phenotype in PGALOEXP1 lst1A exp1A strains depleted of
Explp. 20-26% of cells expressing the deletion alleles exp1A86-102, explA121-149, and
explA107-119 exhibited the multi-budded morphology compared to only 2-5% of cells
expressing exp1A49-58, exp1A61-68, or wild-type EXP1, indicating that the C-terminal
domain of Exp Ip is required for efficient export of Pmal p from the ER (Table II).
One possible reason for the inability of the deletion constructs to rescue lethality
is that the protein either folds incorrectly or is unable to insert into the membrane, both of
which could result in degradation of the protein. We confirmed that each construct was
expressed and inserted into the membrane by differential centrifugation. All mutants were
recovered in the 13,000 and 100,000 x g membrane pellets along with the ER membrane
73
marker Dpmlp (data not shown). We next sought to determine whether the deletion
mutants interfered with the intracellular trafficking of Explp by comparing the
intracellular distribution of each of the mutants in wild type and sec21-1 strains. As
expected, Explp-A49-58 and Explp-A61-68 displayed a significant shift in distribution in
the presence of sec2l-1 and were indistinguishable from wild type Explp (Figure 6B and
data not shown). In contrast, Explp-A107-119 and Explp-A121-149 fractionated with the
ER marker protein even when Golgi-to-ER traffic was blocked with the sec21-1 allele,
suggesting that these deletion mutants were unable to exit the ER (Fig. 6B). Surprisingly,
Explp-A86-102 fractionated with the Golgi marker protein GDPase in both sec21-1 and
wild type strains, indicating that this deletion mutant may be defective in Golgi to ER
retrieval (Fig. 6B). Together, these observations establish a strong correlation between
the role of Explp in Pmalp trafficking, as measured by ability to rescue lst1
explA
lethality and the ability of ExpIp to traffic between the ER and Golgi.
Expip binds to Sec23p/Sec24p in vitro
Genetic, biochemical, and structural studies have established a central role for
Sec24p and its paralogs in cargo recognition and selection into budding vesicles (Miller
et al., 2002). To assay interactions between Explp and components of the COPII coat, we
performed a pull-down assay with the cytoplasmic domain of ExpIp fused to GST as the
bait. Glutathione beads with bound GST-Explp were incubated with cytosolic extracts
from yeast, and proteins that remained bound to the GST-Explp beads after stringent
washing were analyzed by SDS-PAGE and detected by immunoblotting with antibodies
to components of the secretory pathway (Fig. 7A). Sec23p and Sec24p bound to beads
carrying GST-Explp but not to beads carrying GST alone (Fig. 7A and data not shown).
74
This interaction appeared to be specific for the Sec23p/Sec24p subcomplex since the
GST-Explp beads failed to associate with the COPII subunit Sec31p (Fig. 7A).
If Explp exits the ER through direct interaction with Sec24p, then the mutants
that fail to exit the ER should be unable to interact with Sec23p/Sec24p in the in vitro
binding assay. To test this prediction we replaced the transmembrane domain of each
deletion mutant with GST and looked for binding to Sec24p. As predicted, GST-ExplpA107-119 and GST-Explp-A121-149, failed to interact significantly with Sec23p/Sec24p
in this assay, while GST-Explp-A49-58, GST-Explp-A61-68, and GST-Explp-A86-102
bound as much Sec23p/Sec24p as GST-Explp (Fig. 7B). These tests of the deletion
mutants for binding to Sec23p/Sec24p verified the significance of this interaction for
function in vivo as the mutant alleles that did not complement an ist1A exp1A double
mutant and localized to the ER by sucrose gradients also exhibited a defect in binding to
the Sec23p/Sec24p subcomplex.
Since Explp does not require the Sec24p paralog Lstlp for its function, we
hypothesized that Explp may interact selectively with the Sec23p/Sec24p subcomplex
but not with Sec23p/Lstlp subcomplex. Contrary to this expectation, Lstlp-HA
interacted with GST-Explp as efficiently as Sec24p, and similar amounts of Lstlp-HA
were precipitated by each of the Explp deletion mutants (data not shown).
Three distinct regions of the C terminus are required for Expip function
The inability of exp1A107-119 and expIA121-149 to exit the ER or bind Sec24p
suggests that these regions may encode Sec24p binding motifs. Specific motifs that are
required for Sec24p binding and ER exit have been identified for some known cargo
proteins, including diacidic (DxD/E) and dihydrophobic motifs (Votsmeier and Gallwitz,
75
2001; Nishimura et al., 1999; Nufer et al., 2002). Interestingly, ExpI has several putative
DxD motifs within the exp1A107-119 and a potential dihydrophobic motif (FV) at the
end of the exp1A121-149 region. exp1A86-102, has a putative di-arginine motif that may
be necessary for Golgi retrieval (Michelsen et al., 2005). In order to determine if these
motifs or other specific residues within the deletion regions are required for Explp
function we performed site directed alanine mutagenesis. For each deletion region,
groups of 3-5 residues were mutated to alanine and the functionality of each construct
was tested by its ability to support growth of a PGAL1-EXPI lst1Aexp1A strain on glucose
medium (Table III). For the region covered by exp1A86-102, three mutants (expiIDSK90-93, expl-TGLK94-97 and exp1-RRVIG98-102) were unable to rescue an ist1A
exp1A strain, whereas expl-RKFE86-89 was indistinguishable from wild type (Fig. 8B).
For exp1A121-149 we found three mutations unable to grow on glucose medium (expiKYNG136-139, expl-YEG145-147 and expl-FV]48-149) (Fig. 8B). For exp1A107-119
we determined that residues 111-115 are required for Explp function, as mutants expIFDF111-113 and expl-FDIJ13-115 are unable to grow on glucose medium (Fig. 8B).
Interestingly, expl-DFD112-114was indistinguishable from wild type (Figure 8B).
The region spanned by expl-A107-119 (DPNDFDFDIDADDLI)
contains
multiple possible DxD motifs. To test if these may be essential for function we mutated
D 110, 112, 114 and 116 to A or E singly and in combination and tested each for ability to
rescue lethality of the ist1A exp1A strain. Surprising, all of these constructs were able to
rescue lethality, suggesting that there is not an absolute requirement for a DxD motif in
this region (Table III). All of the alanine mutants within a given deletion region that were
unable to rescue lethality of an ist1A explA showed similar results in all further
76
experiments and therefore a single representative of each region will be shown for the
remainder of the paper. These representatives are: expl-RRVIG98-102, expl-FDF111113, and expl-FV148-149.
We examined the effects of each alanine mutant of EXP1 on the prevalence of the
multi-budded rosette phenotype in PGALJ-EXP1 ist1A exp1A strains depleted of Explp by
growth in glucose. Similar to what was observed for the deletion mutants there was an
increased prevalence of multi-budded cells in the alanine mutants (25-35% vs. 2-5% in
WT or ist1A) (Table II).
We also examined the ability of each alanine mutant to traffic between the ER and
Golgi using the sec21-1 gradient assay. As expected, Explp-FDFl 11-113 fractionated
with the ER marker protein even in a sec21-1 block suggesting a defect in ER export
(Fig. 8C). Explp-FV148-149 shows a slight shift in sec21-1 suggesting that there is a
delay in trafficking from the ER rather than a complete block (Fig. 8C). As seen for
Explp-A86-102, Exp 1p-RRVIG98- 102 fractionated with the Golgi in both wild type and
sec21-1 strains (Fig. 8C). This suggests that the di-arginine motif may be important for
Golgi retrieval. To further examine the role of the alanine mutants we replaced the
transmembrane regions with GST and tested interaction with Sec24p using our pull-down
assay. GST-RRVIG98-102 interacts with Sec24p, as would be expected for a Golgi
retrieval mutation (Fig. 8D). GST-FDFl 11-113 and GST-FV148-149 do not interact with
Sec24p, as would be expected for a protein unable to exit the ER (Fig. 8D). Together,
these results suggest that there is not a single linear ER export motif since mutations in
multiple regions of the C terminus can interfere with ER export.
77
Genetic interactions between explA and secretion mutants
Mutations in COPII component genes are synthetically lethal with other COPII
mutant alleles, but not with mutant alleles of genes involved in vesicle tethering or fusion
to the Golgi, or COPI mediated retrieval. To further investigate the relationship between
Explp and the COPII coat proteins, we combined exp1A and thermosensitive alleles of
known secretory genes. We found that exp1A exacerbates the temperature sensitivity of
the COPII alleles sec13-1, sec23-1, and sec24-2, but not sec12-4, sec16-1, sec31-1, or the
COPI allele sec21-1 (Table IV). Powers and Barlowe (2002) observed a similar pattern of
genetic interaction between the cargo adaptor allele erv14A and COPII mutant alleles
(Powers and Barlowe, 2002). We also tested the effect of exp1A on the sensitivity of
COPII mutant strains to growth on low pH media at the permissive temperature.
Interestingly, sec23-land sec24-2 strains grew poorly on acidic medium (pH 3.0) at the
permissive temperature, and the presence of the exp1A allele further inhibited growth of
these COPII mutant strains (Table IV). Similarly sec13-1, which grew normally on acidic
medium, grew poorly on acidic medium when combined with explA. An exp1A mutation
did not cause increased sensitivity to growth on acidic medium of the COPII alleles
sec12-4, sec16-1, sec31-1, or the COPI allele sec21-1. The selective genetic interaction
between the exp1A allele and mutant alleles of the early acting COPII components
suggests that Explp may act in the earliest stages of vesicle budding. This raises the
possibility that Explp may be acting as a cargo adaptor enhancing incorporation of
Pmalp into COPII vesicles by linking Pmalp to Sec24p.
Expip physically interacts with Pmalp
78
If Explp is a cargo adaptor linking Pmalp to Sec24p, it would need to physically
interact with both proteins. We showed above that Explp physically interacts with
Sec24p and that mutation of specific regions of Explp disrupts its interaction with
Sec24p (Fig. 7 and Fig. 8D). To assay the interaction between Explp and Pmalp, we
ectopically expressed HA-Pmalp under the control of a GAL10 promoter in an ist1A
strain to transiently increase the amount of Pmalp present in the ER. An ist1A strain was
used to further increase the amount of HA-Pmalp present in the ER where the interaction
with Explp would occur. After a 3 hour induction of HA-Pmalp, cells were
spheroplasted and HA-Pmalp was immunoprecipitated. Interaction of Pmalp and Explp
was examined by immunoblotting with antibodies to HA and Explp. Wild type Explp
coimmunoprecipitated with HA-Pmalp (Figure 9). To try to determine what region of
Explp is required for interaction with Pmalp we tested the alanine mutants by coimmunoprecipitation. All of the alanine mutants tested were able to interact with HAPmalp, suggesting that the C terminus of Explp is not required for Pmalp interaction
(Figure 9).
Role of transmembrane region and N terminus
Since mutational analysis of the C terminus failed to identify mutants that were
unable to interact with Pmalp we wanted to investigate the possibility of this interaction
occurring in either the N terminus or the transmembrane region of Explp. To examine
this we first performed alanine mutagenesis on conserved residues of the N terminus and
the transmembrane region and tested their ability to rescue lethality of PGALIO-EXP1
ist1A exp1A grown on glucose. All mutations tested were able to rescue lethality and
were indistinguishable from wild type (Table III). Since mutation of single residues did
79
not interfere with Explp function we investigated the effect of changing the entire
transmembrane and/or N terminus. To do this we designed chimeric proteins by domain
swapping with Pga2p (Fig 10A).
Pga2p is another small ER localized single
transmembrane spanning protein that like Explp has a highly charged cytosolic C
terminus and a small lumenal N terminus (Appendix II). Replacing the short N terminus
of Explp with the slightly longer N terminus of Pga2p (PEE) did not affect Explp
function (Fig. 1OB). Swapping the C terminus of the proteins (EEP) resulted in a protein
no longer able to rescue lethality of PGAL1-EXP1 Ist1A exp1A (Fig. 1OB). Swapping the
transmembrane region of Explp with that of Pga2p (EPE) resulted in a protein that was
partially able to rescue lethality of PGAL1-EXP1 ist1A exp1A (Fig lOB).
One
possible
reason for the inability of EPE and EEP to rescue lethality is that the chimeras could be
misfolded and therefore either targeted for degradation or unable to correctly insert into
the ER membrane. To investigate this we used differential centrifugation to look at
membrane association. PEE, EPE and EEP were all expressed at levels comparable to
wild type (EEE) and found in the 13,000 and 100,000 x g pellets along with the ER
membrane maker Dpmlp (data not shown). Another possibility is that trafficking
between the ER and Golgi is altered. To investigate this we used our sec21-1 gradient
recycling assay. Both PEE and EPE show a shift in distribution in sec21-1 similar to wild
type Explp consistent with cycling between the ER and Golgi (Fig. 10C and data not
shown). We know from our GST pull-down experiments that the C terminus physically
interacts with Sec24p and is therefore likely to be required for Explp trafficking.
Gradient recycling assays confirmed that the C terminus contains information necessary
80
for exit from the ER as EEP does not show a shift in distribution in sec21-1 but instead
remains in the ER, colocalizing with ER marker (Fig. 1OC).
Both the EEP and EPE constructs have an increased prevalence of multi-budded
cells (Table II). Since our Sec24p interaction assay fuses GST to the C terminus of
Explp, it is the equivalent of a transmembrane replacement. The fact that this GSTExplp construct interacts with Sec24p suggests that the specific sequence of the
transmembrane region of Exp 1p is not essential for Sec24p interaction.
All of our C terminal Explp mutants were able to interact with Pmalp, suggesting
that this region of the protein does not contain a Pmalp binding site. Since the
transmembrane region is not required for ER exit, we wanted to investigate whether the
inability of the EPE fusion to fully rescue ist1A exp1A lethality was due to an inability to
interact with Pmalp. To test this we used our HA-Pmalp co-immunoprecipitation assay.
The EPE chimera was not detectable in the HA-Pmalp immunoprecipitation, suggesting
that the transmembrane domain contains information required for the interaction of
ExpIp with Pmalp.
81
Discussion
We previously reported the identification of Lstlp as a Sec24p paralog involved
in the selective export of Pmalp from the ER (Roberg et al., 1999). During the cloning of
LST1, we isolated an unlinked suppressor of 1st1-1, which we have named EXP1 (ER
export of Pmalp). Here we show that ectopic overexpression of EXP1 completely
suppresses the Pmalp trafficking defects of lst1A strains, whereas deletion of EXP1 in an
ist1A genetic background is lethal. As diagrammed in Figure 11, the simplest
interpretation of these genetic interactions is that Explp and Lstlp function in distinct
parallel pathways to export Pmal p from the ER. A role for Exp 1p in the export of Pma 1p
from the ER is further supported by examination of the 1stIA exp1A double mutant,
which shows a more severe Pmalp trafficking defect than an Ist1A single mutant. When
compared to ist1A strains, PGALI-EXP1 ist1A exp1A strains depleted of Explp exhibit a
ten-fold increase in the occurrence of the multi-budded cells (Table II), display a
significant decrease in growth rate that closely resembles the behavior of PGAL-PMA1
pmalA strains depleted of Pmalp (Fig. 2B), and show an accumulation of Pmalp in the
ER of the mother cell and an absence of Pmalp in new cell buds (Fig. 2C). The severity
of the Pmalp-trafficking defects exhibited by lst1A exp1A strains can be explained if the
combined absence of LSTJ and EXP1 cause a complete or nearly complete block in
Pmalp-trafficking out of the ER. Under these conditions, mother cells that contain a
stable pool of Pmalp in their plasma membranes continue to divide, while the daughter
cells are inviable due to a failure to deliver newly synthesized Pmalp to the emerging
bud.
82
How does Exp Ip facilitate the export of Pma Ip from the ER? Evaluation of the
EXP1 gene product has led to the following findings, which provide evidence that Exp Ip
acts as a cargo adaptor to promote the selective export of Pmal p from the ER: (i) Exp lp
is an integral membrane protein primarily located in the ER membrane (ii) ExpIp cycles
between the ER and Golgi (iii) Exp lp binds to the COPII subunit Sec24p and regions of
Explp shown to be required for its function and trafficking out of the ER are also
necessary for Sec24p binding (iv) Explp interacts directly with Pmalp likely through its
transmembrane region (v) exp1A shows synthetic interactions with components of the
COPII coat, similar to those seen for the cargo adaptor Ervl4p (Powers and Barlowe,
2002). Previously we showed that overexpression of SEC24 can suppress the Pmalp
trafficking defects of an ist1A strain (Roberg et al., 1999), but in similar tests we have
found that overexpression of SEC24 cannot suppress lethality of istlA explA (data not
shown). Thus, we propose that Exp lp facilitates Pmalp export from the ER by enhancing
Sec24p mediated capture of Pmalp into COPII vesicles, probably through a direct
interaction between the cytosolic portion of Explp and the Sec24p subunit of the COPII
vesicle.
Although the residues involved in the possible Pmalp-Lstlp or Pmalp-Sec24p
interactions have not yet been identified, Miller et al. (2003) have hypothesized that
Pmalp may interact with Lstlp and Sec24p at a common, conserved site but that Pmalp
has a greater affinity for the binding site on Lstlp than for the corresponding site on
Sec24p (Miller et al., 2003). We propose that Explp facilitates Pmalp export from the
ER by enhancing Sec24p mediated capture of Pmalp into COPII vesicles and that the
binding of Explp to Sec24p may modulate the affinity of Pmalp for a coat complex
83
containing Sec23p/Sec24p. We show that Explp interacts with Pmalp, likely through its
transmembrane region. It is unclear whether this interaction is required for loading of
Pmalp into Sec24p containing COPII vesicles or serves another purpose.
Like LST1, EXP1 is non-essential, and the two proteins appear to act in redundant,
parallel pathways to export Pmalp from the ER. Since Pmalp is one of the most
abundant secretory proteins, and since its function is essential for fundamental
physiological processes, redundant pathways for Pmalp export from the ER may have
evolved to ensure adequate delivery of Pmalp to the cell surface (McCusker et al., 1987;
Serrano
et
al.,
1986;
Vallejo
and
Serrano,
1989).
Sec23p/Lstlp
and
Sec23p/Sec24p/Explp may be involved in transporting distinct but overlapping sets of
cargo molecules including Pmalp. This is supported by evidence that the transport of GPI
anchored proteins is primarily through the interaction of the p24 family of cargo adaptors
specifically with Lstlp (Castillon et al., 2011).
Why do yeast cells require two proteins, ExpIp and Sec24p, to perform a function
that Sec24p could theoretically perform on its own? One possibility is ExpIp is part of a
family of adaptor proteins, and various combinations of Sec24p paralogs with these
adaptor proteins broadens the range of cargo proteins selectively exported by the COPII
coat. Such a combinatorial model for cargo selection has already been suggested as an
explanation for the inclusion of different Sec24p paralogs, each of which preferentially
packages a distinct subset of cargo proteins into vesicles (Barlowe, 2003). Alternately,
Explp may be acting directly through Sec24p to alter the structure of the COPII coat.
Pmalp assembles into a 1.8 mDa complex in the ER membrane which is specifically
associated with lipid rafts. As such it may be too large to be loaded into a normal Sec24p
84
vesicle (Bagnat et al., 2001; Toulmay and Schneiter, 2007). It has been shown that Lstlp
containing vesicles are larger in size than vesicles containing Sec24p alone (Shimoni et
al., 2000). This idea could be tested by examining vesicle size in Pmalp containing
vesicles made using in vitro budding assays with Sec24p/Explp or Lstlp compared to
vesicles that do not contain Pmalp.
Pmalp belongs to a large family of P-type ATPases. Recently, several of these
proteins have been shown to be functionally regulated by small accessory proteins.
SERCA, a Ca2+ ATPase of the sarcoplasmic reticulum is regulated by phopholamban
(PLN) and sarcolipin (SLN) in skeletal and atrial muscle (Gorski et al., 2013; Traaseth et
al., 2008). Members of the FXYD family regulate activity of Na*/ K- channels in a
variety of different tissues (Garty and Karlish, 2006; Cheung et al., 2013; Shattock,
2009). Like FXYD, PLN and SLN, Explp is a small single membrane spanning protein
that appears to associate with a P-type ATPase through its transmembrane region. This
raises the possibility that Exp lp could play a regulatory role, perhaps by keeping Pmalp
inactive until it reaches the plasma membrane. Pmalp is essential for regulation of
cytosolic pH, pumping protons out of the cytosol. When Lstlp is present, Pmalp is
efficiently transported from the ER and therefore there is very little Pmalp in the ER
membrane at any given time. When Lstlp is absent, Pmalp is packaged into Sec24p
containing COPII vesicles but less efficiently than when Lstlp is present. This leads to an
accumulation of Pmalp in the ER membrane. If the proton pump function of Pmalp in
the ER membrane is active, protons from the cytosol could be pumped into the ER
lumen. This would result in acidification of the ER lumen, which would be detrimental to
85
the cell. This idea could be tested by measuring the pH of the ER lumen under conditions
that cause Pmalp to accumulate in the ER in the presence or absence of ExpIp.
86
References
Ambesi, A., M. Miranda, V. V Petrov, and C.W. Slayman. 2000. Biogenesis and function
of the yeast plasma-membrane H(+)-ATPase. The Journal of experimental biology.
203:155-60.
Antonny, B., D. Madden, S. Hamamoto, L. Orci, and R. Schekman. 2001. Dynamics of
the COPII coat with GTP and stable analogues. Nat Cell Biol. 3:531-537.
Bagnat, M., A. Chang, and K. Simons. 2001. Plasma membrane proton ATPase Pmalp
requires raft association for surface delivery in yeast. Mol Biol Cell. 12:4129-4138.
Barlowe, C. 2003. Signals for COPI-dependent export from the ER: what's the ticket
out? Trends Cell Biol. 13:295-300.
Barlowe, C., L. Orci, T. Yeung, M. Hosobuchi, S. Hamamoto, N. Salama, M.F. Rexach,
M. Ravazzola, M. Amherdt, and R. Schekman. 1994. COPII: a membrane coat formed by
Sec proteins that drive vesicle budding from the endoplasmic reticulum. Cell. 77:895907.
Belden, W.J., and C. Barlowe. 1996. Erv25p, a component of COPII-coated vesicles,
forms a complex with Emp24p that is required for efficient endoplasmic reticulum to
Golgi transport. JBiol Chem. 271:26939-26946.
Bi, X., J.D. Mancias, and J. Goldberg. 2007. Insights into COPII coat nucleation from the
structure of Sec23.Sarl complexed with the active fragment of Sec31. Developmental
cell. 13:635-45. doi:10.1016/j.devcel.2007.10.006.
Bielli, A., C.J. Haney, G. Gabreski, S.C. Watkins, S.I. Bannykh, and M. Aridor. 2005.
Regulation of Sarl NH2 terminus by GTP binding and hydrolysis promotes membrane
deformation to control COPII vesicle fission. The Journal of cell biology. 171:919-24.
doi: 10.1 083/jcb.200509095.
Castillon, G.A., A. Aguilera-Romero, J. Manzano-Lopez, S. Epstein, K. Kajiwara, K.
Funato, R. Watanabe, H. Riezman, and M. Muiiz. 2011. The yeast p24 complex
regulates GPI-anchored protein transport and quality control by monitoring anchor
remodeling. Molecular biology of the cell. 22:2924-36. doi:10.1091/mbc.El 1-04-0294.
Castillon, G.A., R. Watanabe, M. Taylor, T.M.E. Schwabe, and H. Riezman. 2009.
Concentration of GPI-anchored proteins upon ER exit in yeast. Traffic (Copenhagen,
Denmark). 10:186-200. doi:10.111 1/j.1600-0854.2008.00857.x.
Catty, P., A. de Kerchove d'Exaerde, and A. Goffeau. 1997. The complete inventory of
the yeast Saccharomyces cerevisiae P-type transport ATPases. FEBS letters. 409:325-32.
87
Cheung, J.Y., X.-Q. Zhang, J. Song, E. Gao, T.O. Chan, J.E. Rabinowitz, W.J. Koch,
A.M. Feldman, and J. Wang. 2013. Coordinated regulation of cardiac Na(+)/Ca (2+)
exchanger and Na (+)-K (+)-ATPase by phospholemman (FXYD1). Advances in
experimental medicine and biology. 961:175-90. doi:10.1007/978-1-4614-4756-6_15.
Cid, A., F. Vara, and R. Serrano. 1987. Inhibition of the proton pumping ATPases of
yeast and oat root plasma membranes by dicyclohexylcarbodiimide. Archives of
biochemistry and biophysics. 252:496-500.
Claros, M.G., and G. von Heijne. 1994. TopPred II: an improved software for membrane
protein structure predictions. Comput Appl Biosci. 10:685-686.
Cliften, P., P. Sudarsanam, A. Desikan, L. Fulton, B. Fulton, J. Majors, R. Waterston,
B.A. Cohen, and M. Johnston. 2003. Finding functional features in Saccharomyces
genomes by phylogenetic footprinting. Science. 301:71-76.
Garty, H., and S.J.D. Karlish. 2006. Role of FXYD proteins in ion transport. Annual
review of physiology. 68:431-59. doi: 10.11 46/annurev.physiol.68.040104.131852.
Gaynor, E.C., and S.D. Emr. 1997. COPI-independent anterograde transport: cargoselective ER to Golgi protein transport in yeast COPI mutants. J Cell Biol. 136:789-802.
Gilstring, C.F., M. Melin-Larsson, and P.O. Ljungdahl. 1999. Shr3p mediates specific
COPII coatomer-cargo interactions required for the packaging of amino acid permeases
into ER-derived transport vesicles. Mol Biol Cell. 10:3549-3565.
Gorski, P.A., J.P. Glaves, P. Vangheluwe, and H.S. Young. 2013. Sarco(endo)plasmic
Reticulum Calcium ATPase (SERCA) Inhibition by Sarcolipin Is Encoded in Its Luminal
Tail. The Journal of biological chemistry. 288:8456-67. doi:10.1074/jbc.M 112.446161.
Von Heijne, G. 1992. Membrane protein structure prediction. Hydrophobicity analysis
and the positive-inside rule. JMol Biol. 225:487-494.
Kaiser, C., E.J. Chen, and S. Losko. 2002. Subcellular fractionation of secretory
organelles. Methods in enzymology. 351:325-38.
Kaiser, C., S. Michaelis, A. Mitchell, and C.S.H. Laboratory. 1994. Methods in yeast
genetics: a Cold Spring Harbor Laboratory course manual. Cold Spring Harbor
Laboratory Press.
Kellis, M., N. Patterson, M. Endrizzi, B. Birren, and E.S. Lander. 2003. Sequencing and
comparison of yeast species to identify genes and regulatory elements. Nature. 423:241254.
Kuehn, M.J., J.M. Herrmann, and R. Schekman. 1998. COPII-cargo interactions direct
protein sorting into ER-derived transport vesicles. Nature. 391:187-190.
88
Lee, M.C.S., L. Orci, S. Hamamoto, E. Futai, M. Ravazzola, and R. Schekman. 2005.
Sarlp N-terminal helix initiates membrane curvature and completes the fission of a
COPII vesicle. Cell. 122:605-17. doi: 10. 1016/j.cell.2005.07.025.
McCusker, J.H., D.S. Perlin, and J.E. Haber. 1987. Pleiotropic plasma membrane ATPase
mutations of Saccharomyces cerevisiae. Mol Cell Biol. 7:4082-4088.
Michelsen, K., H. Yuan, and B. Schwappach. 2005. Hide and run. Arginine-based
endoplasmic-reticulum-sorting motifs in the assembly of heteromultimeric membrane
proteins. EMBO reports. 6:717-22. doi:10.1038/sj.embor.7400480.
Miller, E., B. Antonny, S. Hamamoto, and R. Schekman. 2002. Cargo selection into
COPII vesicles is driven by the Sec24p subunit. Embo J. 21:6105-6113.
Miller, E.A., T.H. Beilharz, P.N. Malkus, M.C. Lee, S. Hamamoto, L. Orci, and R.
Schekman. 2003. Multiple cargo binding sites on the COPII subunit Sec24p ensure
capture of diverse membrane proteins into transport vesicles. Cell. 114:497-509.
Mossessova, E., L.C. Bickford, and J. Goldberg. 2003. SNARE selectivity of the COPII
coat. Cell. 114:483-495.
Mufniz, M., C. Nuoffer, H.P. Hauri, and H. Riezman. 2000. The Emp24 complex recruits
a specific cargo molecule into endoplasmic reticulum-derived vesicles. The Journal of
cell biology. 148:925-30.
Nishimura, N., S. Bannykh, S. Slabough, J. Matteson, Y. Altschuler, K. Hahn, and W.E.
Balch. 1999. A di-acidic (DXE) code directs concentration of cargo during export from
the endoplasmic reticulum. The Journalof biological chemistry. 274:15937-46.
Nufer, 0., S. Guldbrandsen, M. Degen, F. Kappeler, J.-P. Paccaud, K. Tani, and H.-P.
Hauri. 2002. Role of cytoplasmic C-terminal amino acids of membrane proteins in ER
export. Journalof cell science. 115:619-28.
Palade, G. 1975. Intracellular aspects of the process of protein synthesis. Science (New
York, N.Y.). 189:347-58.
Powers, J., and C. Barlowe. 1998. Transport of axl2p depends on ervl4p, an ER-vesicle
protein related to the Drosophila cornichon gene product. J Cell Biol. 142:1209-1222.
Powers, J., and C. Barlowe. 2002. Ervl4p directs a transmembrane secretory protein into
COPII-coated transport vesicles. Mol Biol Cell. 13:880-891.
Roberg, K.J., M. Crotwell, P. Espenshade, R. Gimeno, and C.A. Kaiser. 1999. LSTl is a
SEC24 homologue used for selective export of the plasma membrane ATPase from the
endoplasmic reticulum. J Cell Biol. 145:659-672.
89
Sanderson, C.M., and D.I. Meyer. 1991. Purification and functional characterization of
membranes derived from the rough endoplasmic reticulum of Saccharomyces cerevisiae.
JBiol Chem. 266:13423-13430.
Serrano, R., M.C. Kielland-Brandt, and G.R. Fink. 1986. Yeast plasma membrane
ATPase is essential for growth and has homology with (Na+ + K+), K+- and Ca2+ATPases. Nature. 319:689-693.
Shattock, M.J. 2009. Phospholemman: its role in normal cardiac physiology and potential
as a druggable target in disease. Current opinion in pharmacology. 9:160-6.
doi:10.1016/j.coph.2008.12.015.
Shimoni, Y., T. Kurihara, M. Ravazzola, M. Amherdt, L. Orci, and R. Schekman. 2000.
Lstlp and Sec24p cooperate in sorting of the plasma membrane ATPase into COPII
vesicles in Saccharomyces cerevisiae. J Cell Biol. 151:973-984.
Stagg, S.M., C. Girkan, D.M. Fowler, P. LaPointe, T.R. Foss, C.S. Potter, B. Carragher,
and W.E. Balch. 2006. Structure of the Secl3/31 COPII coat cage. Nature. 439:234-8.
doi: 10.1038/nature04339.
Toulmay, A., and R. Schneiter. 2007. Lipid-dependent surface transport of the proton
pumping ATPase: a model to study plasma membrane biogenesis in yeast. Biochimie.
89:249-54. doi: 10.101 6/j.biochi.2006.07.020.
Traaseth, N.J., K.N. Ha, R. Verardi, L. Shi, J.J. Buffy, L.R. Masterson, and G. Veglia.
2008. Structural and dynamic basis of phospholamban and sarcolipin inhibition of
Ca(2+)-ATPase. Biochemistry. 47:3-13. doi: 10.1021 /bi701668v.
Vallejo, C.G., and R. Serrano. 1989. Physiology of mutants with reduced expression of
plasma membrane H+-ATPase. Yeast. 5:307-319.
Votsmeier, C., and D. Gallwitz. 2001. An acidic sequence of a putative yeast Golgi
membrane protein binds COPII and facilitates ER export. Embo J. 20:6742-6750.
90
Tables
Table I Strains Used
Genotype
Strain
CKY8
Source
MATa ura3-52 leu2-3,112
CKY761
MA Ta ura3-52pep4::kanMX6
CKY804
Mata expl.:kanMX6 leu2AO ura3AO metlS4O his3
CKY814
MATa ura3-52 leu2-3,112 expl::URA3 his4-619
EUROSCARF
secl3-1
CKY816
MATa ura3-52 leu2-3,112 expl:: UR A3 sec23-1
CKY818
MA Ta ura3-52 expl::URA3 sec12-1
CKY820
MATaura3-52 sec24-2 expl::URA3
CKY821
MA Ta ura3-52 sec21-1 expl::UR A3
CKY823
MATa ura3-52 sec19-lexp1::URA3
CKY825
MA Ta ura3-52 leu2-3,112 sec16-1 explP:UR A3
CKY827
MATaura3-52 leu2-3,112 sec18-1 expl::URA3
CKY896
MATaura3-52 expl::kanMX6pep4::kanMX6sec21-1
CKY897
MA Ta ura3-52 expl::kanMX6pep4::kanMX6
CKY1171
Mata lstJ::kanMX6 expl::kanMX6 leu2AO ura3AO
his3Al+pMCJ8 (pGALJO-EXPJ HIS3)
CKY1173
Matalst::kanMX6 leu2AO ura3AO lys2AO his3AJ
91
EUROSCARF
Table II Prevalence of the multi-budded morphology in 1st 1D strains depleted of Exp 1p
in the presence of EXP 1 mutants and EXP 1/PGA2 chimeras
Strain/ Plasmid
Percentage of cells with multi budded morphology
Wild Type
1.3
Ist1A
2.8
pRS316
43.2
EXP1 CEN
2.8
EXP1 2p
1.8
exp1A49-58
4.6
exp161-68
3.0
explA86-102
20.0
explA107-119
23.3
explA121-149
26.0
exp1 -RRVIG98-102
21.2
expl-FDF111-113
32.7
expl-FV148-149
28.3
EPE
14.0
EEP
26.5
PEE
2.5
All plasmids are in a pRS316 backbone and transformed into PGALIO-EXP1 ist1A explA.
After growth to exponential phase in SMM with 2% raffinose, cells were transferred to
SMM medium containing 2% raffinose + 2% glucose. After 14 hours, cells were fixed in
1% formaldehyde and transferred to microscope slides. The proportion of cells with three
or more buds was analyzed using phase contrast microscopy. 600 cells were counted for
each strain.
92
Table III Summary of alanine mutagenesis
Mutation
Rescues istlA explA lethality?
RKFE86-89
Yes
IDSK90-93
No
TGLK94-97
NO
RRVIG98-102
No
DPN107-109
Yes
PND108-110
Yes
DFD110-112
Yes
FDF111-113
No
DFD112-114
Yes
FDI113-115
No
DIDD114-117
Yes
DDL116-118
Yes
L1118-119
Yes
EDE121-123
Yes
LDER124-127
Yes
REEE128-131
Yes
KKLK132-135
Yes
KYNG136-139
No
KKNE140-143
Yes
YEG145-147
No
FV148-149
No
93
Table III Summary of alanine mutagenesis continued.
D110A
Yes
D112A
Yes
D114A
Yes
D116A
Yes
D110E
Yes
D112E
Yes
D114E
Yes
D116E
Yes
D110A D114A
Yes
D110E D114E
Yes
D112A D116A
Yes
D112E D116E
Yes
D110AD112AD114AD116A
Yes
D110E D112E D114E D116E
Yes
YGY4-6
Yes
P21 (A G R D I T F or S)
Yes
SG24-25
Yes
TFK28-30
Yes
KP33-34
Yes
Each mutant plasmid was transformed into CKY 1171 (ist1A exp1A + pGAL1O-EXP1).
Rescue of lethality of
ist1A explA was determined by ability to grow on glucose
containing medium.
94
Table IV Genetic interaction between exp1I and sec mutant alleles.
growth at pH 3.0a
growth at pH 3.0a
difference in
of sec mutant
of sec exp1A double mutants
restrictive
temperatureb
secl2-4
++
secl2-4 explA
++
secl3-1
++
secl3-1 explA
+-3
sec31-1
++
sec31-1 explA
+/-
0
sec23-1
+/-
sec23-1 explA
-
-3
sec24-2
±1-
sec24-2 explA
--
3
sec18-1
++
sec18-1 explA
++
0
sec21-1
++
sec21-1 explA
++
0
a The
0
indicated strains were grown on YPD adjusted to pH 3.0 for three days at
24 0 C.
bThe COPII mutant strains and the COPII exp1A double mutant
strains were
grown on YPD at 24'C, 27 0 C, 300 C, 330 C, and 37'C.
95
Figures
Figure 1 Overexpression of EXP1 suppresses the Pmalp related phenotypes of istlA.
(A) ist1A strains with the indicated plasmids were grown on SMM-Leu, pH 3.0 or SMMLeu, pH 4.5 at 30*C for 3 days. (B) ist1A strains containing the indicated plasmids were
fixed and labeled for immunofluorescence with affinity-purified Pmalp antibody
followed by Alexa@488-conjugated goat anti-rabbit secondary antibody. Nuclear DNA
was stained with DAPI, and cell bodies were visualized by DIC microscopy. (C) ist1A
strains containing the indicated plasmids were grown in SMM-Leu at 30'C, and cell
extracts were fractionated on 20-60% sucrose density gradients containing 10 mM
EDTA. Fractions were collected from the top of the gradient. Relative levels of Pmalp,
and marker proteins Gas Ip (plasma membrane), and Dpm 1p (ER) in each fraction were
quantitated by immunoblotting and densitometry. GDPase activity (Golgi) in each
fraction was assayed enzymatically. For simplicity the fraction localization of marker
proteins is indicated by horizontal lines above each graph and labeled as ER, Golgi or
PM.
96
A.
/st1A (LST1)
IstIA (empty vector)
IstlA (EXPI CEN)
pH 4.5
B.
anti-Pmalp
pH 3.0
DAPI
DIC
/stlA
[CEN LST1I
U
U
Ist1A
[CEN EXP1
C.
25
20
E 15
IstlA
IstlA(LST1)
10
-- ,-lst1A(EXP1 CEN)
5
0
1
5
13
9
Fraction Number
97
17
Figure 2 istlA exp1A strains are inviable and accumulate Pmalp in the ER. (A)
CKY1 173 (lst1::KanMX6) was crossed to CKY804 (exp1::KanMX6) expressing EXP1 in
a URA3 marked centromeric plasmid (pDM12). Tetrads were dissected and spores
analyzed for the presence of the URA3 or KanMX markers. ist1A exp1A spores were
analyzed for viability by testing the requirement for maintenance of the EXP1 (URA)
plasmid on 5-FOA medium. As a positive control CKY 1173 (lst1::KanMX6) and
CKY804 (exp1::KanMX6) both expressing pDM12 were also plated on the 5-FOA plates
(B) After growth to exponential phase in SMM with 2% raffinose, cells were transferred
to SMM containing 2% raffinose and 2% glucose (t = 0). The optical density at 660 nm
of each culture was measured at indicated time intervals. (C) Cells were grown as in (B).
After 14 hours of growth in glucose, cells were fixed with 3.7% formaldehyde for 1 hour
and visualized with differential
interference
phase contrast microscopy
(DIC).
Representative examples of cells displaying the multi-budded rosette phenotype are
shown. (D) istlA, explA, and PGALIO-EXPJ Ist1A explA strains were grown and fixed as
described above. Cells were incubated with affinity-purified Pmalp antibody followed by
Alexa@ 488-conjugated anti-rabbit secondary antibody. Nuclear DNA was visualized
with DAPI staining, and cell bodies were visualized by DIC microscopy.
98
B.
A.
3.5
3
*
PGAL10-EKP IstlAexplA
2.5
expl A
IstlA
2
Ist1A exp1AA
01.5
*
IsM1A
0
Ist1Aexp1A8
SMM-Ura
SMM-Ura
5-FOA
0.5
3
4
5 +*2
a
S
~; :.*
0
o
2
4
6
8 10 12 14 16
time (hours)
D.
C.
anti-Pmal p
<uk)
~4.
Ist1A
exp 1A
*AAt
Ist1A exp1A PCA-EXP1
after 14 hours in glucose
IstlA aplA
[PALQ-EXP1]
(After 14 hours
InGlucose)
99
DAPI
DIC
18 20
Figure 3 Explp is a type Ib integral membrane protein. (A) Wild type cells (CKY 8)
were grown to exponential phase in YPD. Cleared cell extracts were sequentially
centrifuged at 500 x g for 10 minutes, 13,000 x g for 10 minutes, and 100,000 x g for 30
minutes. An aliquot of the cleared cell extract was removed prior to centrifugation
(LYSATE). Proteins in the soluble and particulate fractions after each spin were analyzed
by immunoblotting with antisera raised against Explp, soluble PGK, or
membrane
bound Dpmlp. (B) Wild type cell extracts were treated for 1 hour at 4'C with 1% Triton
X-100, 100 mM Na 2 CO 3 pH 11.5, 2.5 M urea, 500 mM NaCl, or buffer alone. Treated
samples were separated into soluble (S) or particulate (P) fractions by centrifugation at
100,000 x g for 75 minutes. Samples were solubilized in sample buffer and analyzed by
immunoblotting with anti-Expip or anti-Dpmlp. (C) Microsomes generated from wild
type cell extracts were digested with 0.75 mg/ml Trypsin in the presence or absence of
1%Triton X-100. Proteins were analyzed by immunoblotting with antiserum to Explp or
PDI, a lumenal ER protein.
100
A.
SK
Explp
Dpmlp
(membrane protein)
PGK
(cytosolic protein)
NaCI
Buffer
B.
pH 11.5
P S P S
Expip Am
Dpmlp
P S
urea
Triton
P S P S
me
._
a
C.
Time (minutes)
5 15 30 5 15 30 5 15 30 5 15 30
Expip
PDI
(lumen protein)
Nw
ara*am W *
101
Figure 4 ExpI cycles between the ER and Golgi. (A) Wild type cell membrane extracts
were fractionated on 20-60% sucrose density gradients containing either 10 mM EDTA
or 2 mM MgCl 2 . Fractions were collected from the top of the gradient. Relative levels of
Pmalp (PM), Explp, and Dpmlp (ER) in each fraction were quantitated by
immunoblotting and densitometry. GDPase activity (Golgi) in each fraction was assayed
enzymatically. The location of each marker protein is shown by lines above the graph.
The Golgi membrane shifts slightly in magnesium compared to EDTA (B) pep4A exp1A
or pep4A exp1A sec21-1 strains expressing ExpIp ectopically on a centromeric plasmid
were grown in YPD at the semi-permissive temperature of 30'C, and cell membrane
extracts were fractionated on 20-60% sucrose density gradients containing 2 mM MgCl 2 .
Relative levels of Pmalp, Explp, and Dpmlp, and GDPase activity in each fraction were
quantitated as described above.
102
EDTA
A.
ER / Golgi
20
PM
16
CL
L12
3 8
--
Expip
13
9
Fraction Number
17
IS
4
0
5
1
Magnesium
ER/PM
20
CL
Golgi
16
C12
0
8
--
4
Explp
0
5
B.
25
13
9
Fraction Number
Golgi
ER
-
-20
17
explA
explAsec2l-1
CL
!15
Ow10
5
01
5
9
Fracdon Number
103
13
17
Figure 5: Alignment of Explp homologs. Expip homologs were identified using BlastP
and aligned using ClustalW with labeling of 50% identity. The transmembrane region
(amino acids 7-28) is indicated by a blue line. The location of each deletion is shown by
numbered green lines.
104
Transmembrane
S.
V.
C.
P.
A.
C.
L.
P.
D.
K.
S.
V.
C.
P.
A.
C.
L.
P.
D.
K.
cerevisiae
polyspora
glabrata
stipitis
gossypii
albicans
elongisporus
guilliermondii
hansenti
lactis
cerevisiae
polyspora
glabrata
stipitis
gossypii
albicans
elongisporus
guilliermondii
hansenii
lactis
KiM
JI
70
80
S
---- EQASSTSSS
DR-------SRNREGDSRAVDVLNA
QDLDDQANQQHQSLKD
S----DSTGTGFKSK
An ------- namnTUrnnRR
NM W------KEDAPVUFV
130 2
..
S.
V.
C.
P.
A.
C.
L.
P.
D.
K.
cerevisiae
polyspora
glabrata
stipitis
gossypii
albicans
elongisporus
guilliermondii
hansenii
lactis
90
100
------ FDLKDTEESLGHDSASASS
-----FNLKDQTKATAFSSURAG----
KK
---
60
50
-----SATAQSATGKLGKREYL
------ NDN------------KK--I
-------AAVSNKKSPDS----RREKL
--- -PKI rDPVAVA---KKHDDYDQ
-------------------------SQBSKP---PQQQQQQY
RSHH-QS
SGPNDGNVS
LKSV
VHS---TSAVSSGT
K
----V
NYLN--SY
IZ L LI ------XTTRRSGD------AJDSEK
-------
1
10
1
50
1
1
52
33
1
10
54
40
500
52
33
56
60
50
51
48
40
30
20
10
l...
....
140
120
---------------------
95
77
-------- TPEA
----------SGSTASGMS
------ AKNEQ
-------- DGT
-- --------QD
99
70
11 5
10 9
94
99
89
------------
50
150
...
S
110
160
53
39
49
51
33
55
59
49
50
47
3 170
I
.. .7v4-
95
77
90
100
70
116
110
95
100
89
105
149
131
144
156
122
172
166
150
156
143
Figure 6 Deletions in the cytoplasmic domain disrupt Explp function. (A) The indicated
deletion mutant alleles were expressed ectopically on centromeric plasmids in
PGALiO-
EXP1 istlA explA strains and assayed for growth on glucose media at 300 C for 3 days.
Serial dilutions are 1:10. (B) exp1A pep4A and exp1A pep4A sec2l-1 strains expressing
the indicated deletion alleles were grown in SMM at 300 C (a semi-permissive
temperature for sec21-1 strains) for 2 hours. Cell membrane lysates were fractionated on
20-60% sucrose density gradients containing 2 mM MgCl 2 to maximize the separation
between ER and Golgi containing fractions. Fractions were collected from the top of the
gradient, and the relative levels of Explp, Dpmlp (ER), and Pmalp (PM) were
quantitated by immunoblotting and densitometry. GDPase activity (Golgi) was assayed
enzymatically. The location of each marker protein is shown by lines above the graphs.
106
A.
empty vector
EXP1
A49-58
A61-68
A86-102
A107-119
A121-149
Glucose
Galactose
B.
A86-102
A49-58
ER
30
25
explAsec21-1
25
20
4O%\
~
-
0 20~-
-210.
10
15R
0
1
5
9
13
17
1
5
FracoNunber
9
17
A121-149
107-119
12
*
OVE
16
GR9
Go6
13
pradian NuMber
12
10
-*10
4-
4
15
913
1
17
13
59
frati
Fradia Nwmber
107
ndmbr
17
Figure 7 Exp Ip interacts with Sec24p via distinct regions in the cytoplasmic domain. (A)
Glutathione-sepharose beads bound to bacterially-purified GST-Explp or GST alone
were added to wild-type yeast extracts or extracts from SEC31-HA strains.
Proteins
bound to the glutathione-sepharose beads after stringent washing were detected by SDSPAGE followed by immunoblotting with anti-Sec23p, anti-Sec24p, or anti-HA (12CA5).
An aliquot of the total cleared lysate (Total) representing 1%of the total number of cell
equivalents used in each binding assay was removed prior to the addition of the proteinbound Glutathione-sepharose beads. (B) Each deletion allele was fused to GST, and the
resulting fusion protein was purified from E. coli with glutathione-sepharose beads.
Equivalent amounts of protein-bound beads were added to wild-type yeast cell extracts
and tested for binding to Sec23p/Sec24p p as described above.
108
A.
Sec24p
Sec3lp
-
B.
N
Sec24p
,"
109
Figure 8 Alanine mutagenesis identified residues required for Explp function. (A)
Schematic of regions of EXPI targeted for alanine mutagenesis. Mutations of the
residues in green affect Golgi to ER retrieval and mutations of the residues in red affect
ER to Golgi transport (B) Alanine mutants were tested for their ability to complement
lethality of ist1A exp1A double mutant strains by assaying growth of PGALIo-EXP1
ist1A exp ]A with each alanine mutant plasmid on glucose medium. Strains were plated
with serial dilutions of 1:10. (C) Trafficking of each alanine mutant was tested by sucrose
density gradients in explA pep4A or explApep4A sec21-1 at the semi-permissive
temperature. Fractions were analyzed as described in Figure 6B. (D) Each alanine mutant
was fused to GST. The resulting fusion proteins were purified from E. coli with
glutathione sepharose beads and analyzed as in Figure 7A.
110
A.
RKFEIDSKTGLKRRVIGI
DNFDFDIDL
B.
DELDERREEEKKKKYNGKKNEAYEGFV
1191 114
102 107
87
C.
empty vector
EXP1
FDF111-113
FD113-115
KYNG1 36-139
YEG145-147
FV1 48-149
FDF111-113
ER
3S
30
Golgi
23
20,
expi2
explZ sec2l-1
10
S
a
empty vector
EXP1
IDSK90-93
" 9
Ft,~e Nmftbet
3
FV148-149
TGLK94-97
RRVIG98-102
17
ER
2s
Golgi
A
20
Galactose
Is
Glucose
!to
S
D.
'N
0
V.,
9
Wn
ub
13
17
is
hQ~
-
0
RRVIG98-102
ER
L~(~
(.~
Golgi
25
20
/
4101
0
111
I
5
9
FRadon
NuuMber
13
17
Figure 9 Explp interacts physically with Pmalp. ist1A cells expressing Gal-HA-PMAl
and the indicated EXP1 plasmids were grown in SMM + 2% galactose for 3 hours,
spheroplasted, treated with DSP crosslinker and immunoprecipitated with aXHA 3F10.
The resulting immunoprecipitants were analyzed by immunoblotting with xHA 12CA5
and aExplp.
112
Qv~
PN
4..
4 q
"I
'4
'ow,
,""WON.
-4-
HA-Pmalp
--
Expip
"Melt
3%Total Lysate
anti-HA 3F1 0 IP
113
Figure 10 Role of the transmembrane region of Explp in Pmalp trafficking. (A)
Schematic of the chimeric proteins created between Explp and Pga2p. (B) Fusion
constructs were tested for their ability to complement lethality of ist1A exp1A double
mutant strains by assaying growth of PGAL1-EXP1 ist1A exp1A with each fusion plasmid
on glucose medium. Serial dilutions are 1:10. (C) Trafficking of each fusion construct
was tested by sucrose density gradients in exp1Apep4A and exp1Apep4A sec21-1 strains
at the semi-permissive temperature. Fractions were analyzed as described in Figure 6B.
114
A-.
B.
Pga2 (PPP)
rE
-
EPE L.
EEP | |1
Exp1 (EEE) |
|
PEE
EPE
EEP
EEE
I
EMMMMMMIMIL-
]
|
Galactose
I
Glucose
C.
EPE
EEP
Golgi
30
GOlgi
A
12
10
25 -
-4-explA
exp1ase21-1
120
15
10
ER
14
ER
35
62
6 explAsec2l-1
5
9
Fraion
13
17
NkMbw
115
5
9
13
17
Figure 11 Model for Pmalp export from the ER. There are two distinct parallel pathways
for Pmalp export from the ER. One pathway is mediated by the Sec24p paralog Lstlp.
The other pathway is mediated by Sec24p and requires Explp. In wild type cells both
pathways are functional, and Pmalp is efficiently incorporated into COPII vesicles. In the
absence of Expip, the Lstlp pathway is fully functional and able to traffic Pmalp to the
cell surface efficiently, so there is no accumulation of Pmalp in the ER membrane. In the
absence of Lstlp, the Sec24p/Explp pathway can package Pmalp into COPII vesicles,
but this is less efficient than when Lstlp is present resulting in an accumulation of Pmalp
in the ER membrane. In the absence of both Lstlp and Expip, Sec24p cannot load
Pmalp into COPII vesicles resulting in a large accumulation of Pmalp in the ER and a
subsequent absence of Pmalp at the plasma membrane leading to cell death.
116
Pathways of Pma Ip Exit from the ER
Lstl p-Dependent Pathway
Sec24p-Dependent pathway
cytosol
Wild Type
I
ER membrane
cytosol
expA
ER membrane
~
cytosol
Ist1A
ofi ifti'r
membrane
I fER
cytosol
IstIA expIA
J
117
}I iER}
membrane
Chapter 3: Prospectus
118
The work presented here focuses on the mechanisms of ER export of the plasma
membrane ATPase Pmalp in S. cerevisiae. Pmalp is an essential protein required to
establish an electrochemical gradient needed for nutrient uptake and maintenance of
cytosolic pH. Previous work in the lab showed that the Sec24p paralog Lstlp is required
for efficient ER export of Pmalp (Roberg et al., 1999). Here we show that in the absence
of Lstlp, Explp is required as a cargo adaptor for Sec24p mediated ER export of Pmalp.
We propose that there are two parallel pathways for ER export of Pmalp mediated by
either Lstlp or Sec24p/Explp. In support of this idea we show that: (i) ectopic expression
of ExpIp rescues defects of an ist1A strain (ii) an 1st1A exp1A strain is inviable due to a
severe Pmalp trafficking defect (iii) Explp is a small ER membrane protein that can
cycle between the ER and Golgi (iv) Explp physically interacts with Pmalp and Sec24p
through distinct regions of the protein and (v) exp1A displays synthetic effects with
conditional alleles of COPII components. Together these results support the idea of
ExpIp acting a cargo adaptor to link Sec24p and Pmalp.
Biochemical and genetic studies combined with in vitro budding assays have
provided extensive detail into the mechanisms of cargo selectivity of Sec24p paralogs
(Miller et al., 2003; Mossessova et al., 2003). Sec24p and Lstlp package both distinct
and common cargo, suggesting that the presence of both greatly increases the diversity of
cargo proteins recognized. The identification of cargo adaptors expands the range of
cargo that can be recognized by Sec24p and Lstlp. The identification of Explp provides
insight not only into the role of cargo adaptors but also into the partial redundancy cells
have evolved to ensure trafficking of vital proteins. To gain a better understanding of the
roles of Explp, Lstlp and Sec24p in the transport of Pmalp and in trafficking in general,
119
the following questions could be addressed: (1) Is Explp required for Sec24p mediated
vesicle transport in vitro, and is Explp incorporated into vesicles? (2) How does Explp
interact with Sec24p? (3) How is Pmalp recognized and transported by Lstlp and
Sec24p? Is there a conserved Pmalp binding site? (4) What is the cargo specificity of
Lstlp and Sec24p? (5) Is Lstlp required for vesicle transport of large cargo? Does ExpIp
alter vesicle size? (6) Does Expip regulate the proton transport activity of Pmalp in the
ER membrane? (7) Is Exp 1p part of a family of adaptor proteins?
In vitro analysis of Sec24p mediated export of Pmalp
In order to confirm that Explp promotes export of Pmalp through direct
interaction with Sec24p, the requirement of Explp for loading of Pmalp into Sec24p
vesicles can be examined in vitro. The presence of both Lstlp and Sec24p is required for
optimal vesicle incorporation of Pmalp (Shimoni et al., 2000; Miller et al., 2002). In the
absence of Lstlp, a 20-fold excess of Sec24p is required to restore wild type levels of
vesicle incorporation of Pmalp (Shimoni et al., 2000). If our model that Explp acts as a
cargo adaptor is correct, then using membranes from cells overproducing Explp should
decrease the amount of Sec24p needed to reach the level of vesicle incorporation of
Pmalp obtained when both Lstlp and Sec24p are present. Performing in vitro budding
assays using explA membranes would verify the requirement of Explp for Sec24p
mediated incorporation of Pmalp. In the absence of Explp in the membrane our model
predicts that vesicles created using Sec24p would not contain Pmalp, but that vesicles
created with Lstlp would still contain Pmalp. Analysis of these in vitro vesicles will also
confirm that Exp Ip acts as a canonical cargo adaptor, by being incorporated into vesicles
along with its cargo.
120
How does Expip interact with Sec24p
We showed that the C terminus of Exp Ip is important for export from the ER and
most likely interacts with Sec24p. Sec24p has multiple distinct binding sites for cargo,
and mutations in these sites have been generated that interrupt binding of specific cargo
protein (Mossessova et al., 2003; Miller et al., 2003). To further analyze the interaction
between Sec24p and Explp, these Sec24p binding site mutants could be used in
conjunction with our GST-Expl pull-down assay. If ExpIp is binding to a specific known
binding site on Sec24p, this mutant should show a decreased affinity for GST-Explp
when compared to wild type Sec24p. The use of purified Sec24p not bound to Sec23p
would also establish that the interaction we detected with Sec24p in our pull-down assays
is due to direct binding to Sec24p rather than an interaction with Sec23p.
How is Pmalp recognized by Lst1p
Lstlp does not appear to require the assistance of Explp for Pmalp transport, as
Pmalp trafficking is unaffected in an explA. It is possible that Sec24p and Lstlp bind
Pmalp via a conserved binding site, but that Lstlp has a higher binding affinity. An istlA
is viable but shows growth defects when grown on low pH medium, presumably due to a
specific defect in transport of Pmalp. This phenotype would enable screening for mutants
in Lstlp that are no longer able to transport Pmalp. Plasmids containing LSTJ could be
mutagenized using conditions that create error prone PCR. These mutants could be
transformed into an lst1A, plated on media at pH >4.0 to select viable strains, and then
replica plated to media at pH 3.0 to identify colonies with a growth defect at low pH. The
specificity of these mutants for Pmalp trafficking could be examined by confirming
121
synthetic lethality with exp1A and making sure that other cargo normally dependent on
Lstlp, such as the GPI-anchored proteins are not affected. Once mutants affecting Lstlp
interaction with Pmalp have been identified, the corresponding mutations can be made in
Sec24p to see if there is a similar Pmalp binding defect.
What is the cargo specificity of Lst1p and Sec24p
Lstlp and Sec24p have been shown to mediate the trafficking of distinct but
overlapping cargo proteins. In humans there are four isoforms of Sec24 (A-D). Sec24A
and Sec24B show sequence similarity to SEC24 in S. cerevisiae, whereas Sec24C and
Sec24D are more similar to LST1. In support of this Sec24D, like Lstlp is required for
ER export of GPI-anchored proteins (Bonnon et al., 2010; Wendeler et al., 2007). The
identification of isoform specificity of individual cargo proteins has relied on
examination of trafficking of a specific cargo protein in the absence of each isoform.
However, the isoform specificity of most cargo proteins is still unknown. To examine this
issue, the cargo content of vesicles created by in vitro budding assays using different
Sec24p isoforms could be analyzed and compared using mass spectrometry.
What is the role of Lst1p and Expip in loading of large cargo
Although the specificity of most proteins for Sec24p and Lstlp is still unclear,
some interesting observations can be made based on current cargo classifications. In
particular, the cargo proteins that require Lstlp or its mammalian counterparts Sec24C
and Sec24D share similar characteristics:
large size, oligomerization into large
complexes, and/or association with lipid rafts (Wendeler et al., 2007; Bonnon et al., 2010;
Mancias and Goldberg, 2008). Vesicles containing both Sec24p and Lstlp are ~15%
122
larger than vesicles containing only Sec24p (Shimoni et al., 2000). This raises the
possibility that the presence of Lst 1p may alter the COPII coat structure. Sec 13p/Sec3 1p
can polymerize into cages of different sizes and geometries that may allow vesicles to
accommodate larger cargo (Bhattacharya et al., 2012). Lstlp may alter this geometry to
create larger vesicles. In the absence of Lstlp vesicles are smaller and may be unable to
accommodate large cargo. Pmalp forms a 1.8 mDa oligomer in the ER membrane and
associates with lipid rafts (Bagnat et al., 2001; Toulmay and Schneiter, 2007). This may
make Pmalp too large to be efficiently loaded into vesicles containing only Sec24p.
ExpIp binding to Pmalp and Sec24p may cause a conformational change in the structure
of the COPII coat, altering cage geometry to allow loading of larger cargo. To test this
idea, in vitro budding assays could be done with Lst 1p and Sec24p using wild type (Exp 1
containing) and exp1A membranes. Comparison of the size of vesicles produced in the
presence or absence of Explp may provide insight into the possible role of Explp in
COPII cage geometry.
Does Expip Regulate Pmalp activity
Pmalp is a member of a large family of P-type ATPases that are responsible for
the transport of cations, heavy metal and lipids in all organisms from bacteria to humans.
Members of this family have been shown to be regulated by small transmembrane
proteins. The activity of SERCA, a sarcoplasmic reticulum calcium ATPase found in
skeletal and atrial muscles, is regulated by two proteins phospholamban (PLN) and
sarcolipin (SLN). PLN and SLN are small (52 and 31 amino acid) phosphoproteins that
act as Ca2 transport inhibitors when bound to SERCA (Gorski et al., 2013; Traaseth et
al., 2008). The FXYD proteins are a family of 7 small single transmembrane spanning
123
proteins (>100 amino acids) that modulate the activity of Na/K ATPases in a tissue
specific manner (Garty and Karlish, 2006; Cheung et al., 2013; Shattock, 2009).
In an ist1A strain, there is significant accumulation of Pmalp in the ER. This is
presumably due to the inefficient loading of Pmalp into COPII vesicles containing only
Sec24p, as this accumulation can be rescued by a 20-fold increase in Sec24p (Shimoni et
al., 2000). We have shown that this accumulation can also be rescued by additional
Explp. Overexpression of Pmalp in an ist1A is toxic (Figure lA, empty vector). This
could be due to overwhelming the protein folding machinery or due to active Pmalp in
the ER acidifying the ER lumen. Preliminary data shows that overexpression of Explp
can rescue this toxicity (Figure 1A, EXP1). Interestingly, Expl mutants that do not exit
the ER or rescue the lethality of an ist1A explA strain can suppress the toxicity of Pmalp
overexpression (Figure 1A, FDF and RRVIG). One explanation for this is that -while
these Explp mutants lack an ER export signal, they are still able to bind to Pmalp. In
doing so, Explp may either protect accumulating Pmalp from activating a stress
response or may inactivate proton transport by Pmalp preventing lumen acidification.
1st] A strains exhibit a growth defect on low pH medium (Figure 1B, empty vector). We
have shown that this is rescued by ectopic expression of Explp. Surprisingly, ectopic
expression of Explp mutants that are unable to exit the ER can also rescue this pH
sensitivity, suggesting that the presence of Explp is not merely increasing export of
Pmalp to the cell surface (Figure 1B).
Eraso and colleagues (2011) showed that overexpression of wild type or dominant
negative D378T Pmalp does not induce the UPR, but can activate the osmostress Hogl
pathway (Eraso et al., 2011). It would be interesting to see if overexpression of wild type
124
or ER block mutant Explp alters this Hogi pathway activation. They also showed that
upregulation of the UPR has no effect on the rate of Pmalp degradation by ERAD,
making it less likely that the toxicity is due to a stress response.
The idea that too much Pmalp in the ER membrane can result in the acidification
of the ER lumen can be tested by measuring the lumenal pH using a pH sensitive variant
of GFP (pHluorin) (Miesenb6ck et al., 1998). Ratiometric pHluorins has two excitation
wavelengths, the intensity of which are affected by changes in pH is opposite ways. At
the 405 nm excitation wavelength the emission intensity increases with increasing pH
whereas at the 485 nm excitation wavelength the emission intensity decreases with
increasing pH (Miesenb6ck et al., 1998; Brett et al., 2005). To calculate the pH in the
location where pHluorin is being expressed, the emission intensity at each excitation
wavelength is calculated and a ratio of 405 nm/485 nm is compared to a standard curve
generated by collecting data from cells grown in calibration buffers of known pH
(Miesenb6ck et al., 1998). We have created an ER lumen localized construct of a
ratiometric pHluorin (gift of Rajini Rao, Johns Hopkins) by adding a Kar2p signal
sequence at the N terminus and HDEL retrieval sequence to the C terminus (Figure 2A)
(Brett et al., 2005; Merksamer et al., 2008). Figure 2B shows that this construct is
expressed in the ER lumen and detectable at both excitation wavelengths, although
weaker at 405 nm. To examine the effect of Pmalp on ER lumenal pH we will use this
construct to measure the pH under conditions of Pmalp accumulation in the ER such as
in an ist1A under acidic conditions or an Ist1A exp1A depleted of Explp, and examine
how the pH changes in these same strains when they also express either wild type or
mutant Exp 1p.
125
Identification of additional cargo adaptors
Exp Ip may be a member of a family of cargo adaptor that aid in the loading of
cargo proteins into vesicles. Advances in technology have made it possible to easily
screen entire deletion libraries for synthetic interactions with a specific deletion or
conditional mutant. As such, additional cargo adaptors could be identified using synthetic
genetic arrays of the deletion collection with either an ist1A or thermosensitive sec24
alleles. Candidate genes could be analyzed to see if they fit the basic criteria of cargo
adaptors: ER-Golgi cycling, interaction with Sec24p or its paralogs, and required for
trafficking of a specific subset of cargo proteins.
126
References
Bagnat, M., A. Chang, and K. Simons. 2001. Plasma membrane proton ATPase Pmalp
requires raft association for surface delivery in yeast. Molecular biology of the cell.
12:4129-38.
Bhattacharya, N., J. 0 Donnell, and S.M. Stagg. 2012. The structure of the Secl3/31
COPII cage bound to Sec23. Journal of molecular biology. 420:324-34.
doi: 10.1016/j.jmb.2012.04.024.
Bonnon, C., M.W. Wendeler, J.-P. Paccaud, and H.-P. Hauri. 2010. Selective export of
human GPI-anchored proteins from the endoplasmic reticulum. Journal of cell science.
123:1705-15. doi:10.1242/jcs.062950.
Brett, C.L., D.N. Tukaye, S. Mukherjee, and R. Rao. 2005. The yeast endosomal
Na+K+/H+ exchanger Nhx1 regulates cellular pH to control vesicle trafficking.
Molecular biology of the cell. 16:1396-405. doi: 10.1091 /mbc.E04-11-0999.
Cheung, J.Y., X.-Q. Zhang, J. Song, E. Gao, T.O. Chan, J.E. Rabinowitz, W.J. Koch,
A.M. Feldman, and J. Wang. 2013. Coordinated regulation of cardiac Na(+)/Ca (2+)
exchanger and Na (+)-K (+)-ATPase by phospholemman (FXYD1). Advances in
experimental medicine and biology. 961:175-90. doi:10.1007/978-1-4614-4756-6_15.
Eraso, P., M.J. Maz6n, F. Posas, and F. Portillo. 2011. Gene expression profiling of
yeasts overexpressing wild type or misfolded Pmal variants reveals activation of the
Hogi MAPK pathway. Molecular microbiology. 79:1339-52. doi:10.1111/j.13652958.2010.07528.x.
Garty, H., and S.J.D. Karlish. 2006. Role of FXYD proteins in ion transport. Annual
review ofphysiology. 68:431-59. doi:10.1 146/annurev.physiol.68.040104.131852.
Gorski, P.A., J.P. Glaves, P. Vangheluwe, and H.S. Young. 2013. Sarco(endo)plasmic
Reticulum Calcium ATPase (SERCA) Inhibition by Sarcolipin Is Encoded in Its Luminal
Tail. The Journalof biological chemistry. 288:8456-67. doi:10.1074/jbc.M112.446161.
Mancias, J.D., and J. Goldberg. 2008. Structural basis of cargo membrane protein
discrimination by the human COPII coat machinery. The EMBO journal. 27:2918-28.
doi: 10.1038/emboj.2008.208.
Merksamer, P.I., A. Trusina, and F.R. Papa. 2008. Real-time redox measurements during
endoplasmic reticulum stress reveal interlinked protein folding functions. Cell. 135:93347. doi:10.1016/j.cell.2008.10.011.
127
Miesenb6ck, G., D.A. De Angelis, and J.E. Rothman. 1998. Visualizing secretion and
synaptic transmission with pH-sensitive green fluorescent proteins. Nature. 394:192-5.
doi:10.1038/28190.
Miller, E., B. Antonny, S. Hamamoto, and R. Schekman. 2002. Cargo selection into
COPII vesicles is driven by the Sec24p subunit. Embo J. 21:6105-6113.
Miller, E.A., T.H. Beilharz, P.N. Malkus, M.C. Lee, S. Hamamoto, L. Orci, and R.
Schekman. 2003. Multiple cargo binding sites on the COPII subunit Sec24p ensure
capture of diverse membrane proteins into transport vesicles. Cell. 114:497-509.
Mossessova, E., L.C. Bickford, and J. Goldberg. 2003. SNARE selectivity of the COPII
coat. Cell. 114:483-495.
Shattock, M.J. 2009. Phospholemman: its role in normal cardiac physiology and potential
as a druggable target in disease. Current opinion in pharmacology. 9:160-6.
doi:10.1016/j.coph.2008.12.015.
Shimoni, Y., T. Kurihara, M. Ravazzola, M. Amherdt, L. Orci, and R. Schekman. 2000.
Lstlp and Sec24p cooperate in sorting of the plasma membrane ATPase into COPII
vesicles in Saccharomyces cerevisiae. The Journalof cell biology. 151:973-84.
Toulmay, A., and R. Schneiter. 2007. Lipid-dependent surface transport of the proton
pumping ATPase: a model to study plasma membrane biogenesis in yeast. Biochimie.
89:249-54. doi: 10.101 6/j.biochi.2006.07.020.
Traaseth, N.J., K.N. Ha, R. Verardi, L. Shi, J.J. Buffy, L.R. Masterson, and G. Veglia.
2008. Structural and dynamic basis of phospholamban and sarcolipin inhibition of
Ca(2+)-ATPase. Biochemistry. 47:3-13. doi:10.1021 /bi701668v.
Wendeler, M.W., J.-P. Paccaud, and H.-P. Hauri. 2007. Role of Sec24 isoforms in
selective export of membrane proteins from the endoplasmic reticulum. EMBO reports.
8:258-64. doi:10.1038/sj.embor.7400893.
128
Figures
Figure 1 Overexpression of Pmalp lethality can be rescued by Explp. (A) lst1z strains
containing pXZ33 (Gal-HA-PMA1) (a gift of James Haber, Brandeis) + indicated EXP1
constructs in URA3 CEN plasmids were spotted on either 2% raffinose or 2% galactose
media and grown for 3 days at 30'C. (B) ist1A strains expressing indicated Explp
constructs were spotted on SC-Ura plates at pH 4.0 or pH 3.0 and grown for 3 days at
300 C.
129
A.
Empty Vector
EXPI (CEN)
FDF
AAA
A11
RRVIG& 10 2 AAAAA
2% Raffinose
2% Galactose
B.
Empty Vector
WT EXP1 in pRS316
WT EXP1 in pRS426
RRVIG 98 102AAAAA
YEG, 4 147AAA
FV4 1 49AA
FDF 111-113AAA
FDI 113 15AAA
SC-ura pH2.5
SC-ura
130
Figure 2 pHluorin is expressed in the ER. (A) The pH sensitive variant of GFP,
ratiometric pHluorin, was cloned into pPM28 (eroGFP) such that it replaced the redox
GFP. This puts pHluorin behind a constitutive TDH3 promoter and adds the Kar2 signal
sequence at the N terminus. An HDEL sequence was also added to the C terminus to
create ER-pHluorin (pDM127). (B) Wild type CKY263 cells expressing pDM137 (ERpHluorin) were grown at 30'C O/N. Cells were resuspended in RXN buffer (50 mM
HEPES, 50 mM MES, 50 mM KCl, 50 mM NaCl) at pH 6.5 and imaged using a
PerkinElmer Ultraview Spinning Disk Confocal microscope (W. M. Keck microscopy
facility, Whitehead Institute) with excitation at 405 nm and 488 nm, and emission at 510
nm. Images were analyzed with Image J software.
131
HDEL
A.
Kar2 signal
sequence
ER-plorin
TDH3 promoter
URA3 marker
B.
485 nm
405 nm
132
Appendix I: CCW14 Screen
133
Abstract
Ccwl4p is a mannoprotein of the inner cell wall of S. cerevisiae. In the ER,
Ccwl4p is modified by addition of a GPI-anchor and glycosylation. As it traverses
through the Golgi, GPI-Ccwl4p is further glycosylated. At the trans-Golgi GPI-Ccwl4p
is packaged into vesicles and ultimately transported to the inner cell wall, where it
becomes covalently attached to the glucans of the cell wall losing the GPI-anchor.
Ccwl4p is an ideal marker protein for identification of secretory pathway defects. In a
wild type cell, very little Ccwl4p is detectable, as most has reached the cell wall and is
therefore soluble only with treatment with glucanase. If there is even a slight defect in
trafficking of Ccwl4p, it will accumulate in the GPI-anchored form internally and be in
the soluble fraction of cell lysates. The stage of secretory block can be determined by
looking at the Ccwl4p protein on Western blots as the addition of the GPI-anchor and the
amount of glycosylation alter the size of the protein. We used Ccwl4p accumulation to
screen the TET collection of essential genes under the control of a doxycycline
regulatable promoter. A total of 773 strains (~75%
of the essential genes in yeast) were
screened and 172 of these showed an accumulation of Ccwl4p. This included 39 genes
known to function in the secretory pathway as well as 6 genes of unknown function. One
of the genes identified in the screen was PGA2 (yNL149c), which we are currently
characterizing (Appendix II).
134
Introduction
The secretory pathway of the yeast S. cerevisiae is required for proper folding,
modification and targeting of many newly synthesized proteins. This pathway consists of
an ordered set of organelles including the endoplasmic reticulum (ER), Golgi, endosome,
vacuole (lysosome), and plasma membrane (Palade, 1975). A protein enters the secretory
pathway by translocation into the ER, where it undergoes modifications needed for
proper folding, assembly and function. These ER modifications can include signal
sequence cleavage, GPI-anchor addition, 0- and N-linked glycosylation, and disulfide
bound formation (Weerapana and Imperiali, 2006; Gemmill and Trimble, 1999; Fujita
and Kinoshita, 2012). The ER also serves as a site of quality control, preventing unfolded
or incorrectly folded proteins from progressing further along the pathway. From the ER,
properly folded proteins are packaged into vesicles for transport to the Golgi. The Golgi
is a series of compartments that further modify proteins through the addition of specific
sugar residues (Hashimoto et al., 2002). From the Golgi, proteins are packaged into
vesicles for secretion out of the cell, or for transport to their final cellular locations such
as the vacuole or plasma membrane. Proteins necessary for formation of vesicles at each
stage are recycled back to the donor membranes for reuse in vesicle formation
A variety of approaches have been used to identify yeast genes involved in the
secretory pathway. The first secretory genes were identified by taking advantage of the
increase in cell density that occurs when yeast cells are unable to increase their plasma
membrane due to blocks in vesicle transport, but can still synthesize proteins.
This
screen identified 23 temperature sensitive mutants through their increased cell density
(Novick et al., 1980). Many tests for secretory pathway genes in S. cerevisiae have made
135
use of three soluble marker proteins: the secreted pheromone a-factor, the secreted
enzyme invertase, and the vacuolar protease carboxypeptidase Y (CPY) (Wuestehube et
al., 1996; Carlson et al., 1981; Novick et al., 1980). Assays using CPY, invertase and u.factor allow detection of complete disruption of
protein maturation, as would be
expected for major components of the secretory pathway, but these assays are less
efficient at recognizing more subtle disruptions, as would be expected for genes that play
a minor, partially redundant or regulatory role in protein trafficking. Not all mutants are
defective in proper localization of all three of these markers, suggesting that different
cargo proteins can utilize different pathways and components for modification, targeting,
and transport. For example, CPY requires disulfide bond formation for proper maturation
and transport, whereas invertase does not. Another limitation of these markers is that
soluble proteins comprise only a subset of the repertoire of transport proteins; the
majority of proteins transported through the secretory pathway are integral membrane
proteins, which may utilize somewhat different mechanisms of transport from soluble
proteins (Nothwehr and Stevens, 1994). Finally, the marker proteins CPY, invertase and
a-factor are not ideal for large-scale screening since their detection requires laborintensive methods such as pulse chase labeling and immunoprecipitation. These
limitations make using these markers useful for small-scale tests, but not practical for
high throughput screening. More recent systematic studies have focused on genetic
interactions or fluorescence microscopy as a means to identify novel secretory genes
(Schuldiner et al., 2005; Davierwala et al., 2005; Costanzo et al., 2010). These
approaches are useful for large scale analysis, but are not able to detect small alterations
136
in protein secretion, and do not necessarily provide information about where in the
secretory pathway a gene functions.
Our lab has developed a simple immunoblotting assay for detecting defects in
protein trafficking using Ccwl4p, a GPI-anchored inner cell wall protein (Moukadiri et
al., 1997; Mrsa et al., 1999). As shown in figure 1, when the secretory pathway is
functioning correctly, Ccwl4p travels rapidly to the inner cell wall where it is covalently
linked to the cell wall glucans and therefore insoluble unless digested with glucanase. If
there is any slowing in the rate of transport, intracellular forms of Ccw l4p accumulate in
the cell and remain in the detergent soluble fraction of cell lysates. This accumulation is
readily detectable by immunoblotting soluble fractions of cell lysates with Ccwl4p
antibody. Analysis of the size of the Ccw l4p that accumulates also provides information
about the step at which Ccwl4p trafficking is disrupted.
Many of the known secretory genes are essential, which is expected since
secretory function is vital to cell viability. Until the last decade, it had not been possible
to easily screen all the essential genes because temperature sensitive alleles of most
essential genes have not been isolated. We were able to take advantage of a recently
created collection of conditional alleles of essential genes that allows regulation of gene
expression by tetracycline activators (Mnaimneh et al., 2004). This TET collection is
designed such that the endogenous promoter of each essential gene is replaced by a
promoter containing tetracycline response elements (TRE) called the TETO 7 promoter. A
tet activator (tTA*), encoded by a separate construct inserted at the URA3 locus, binds to
the TRE to activate transcription of the downstream gene (Fig. 3A). When doxycycline is
added to the growth media, it enters cells and binds to the tTA* preventing the activator
137
from binding to the TRE, thus blocking transcriptional activation (Figure 3B). This
transcriptional block leads to a reduction in downstream gene product over time based on
the rate of degradation of the already made protein product (Hughes et al., 2000;
Mnaimneh et al., 2004). We were able to monitor the internal levels of Ccwl4p to look
for strains in which a loss of the protein after addition of doxycycline leads to an
accumulation of Ccwl4p.
A total of 773 strains (~75%
Ccwl4p accumulation
of the essential genes in yeast) were screened for
and 172 of these showed an accumulation of Ccwl4p
(summarized in tables II and II). This included 39 genes known to function in the
secretory pathway (Table III, purple) as well as 6 genes of unknown function (Table LII,
green). One of the genes identified in the screen was PGA2 (yNL149c), which we are
currently characterizing (Appendix II).
138
Materials and Methods
Media and strains
S. cerevisiae strains used in this study are listed in Table I. Cells were grown in
rich medium (YPD) as described in Kaiser et al., 1994. The TET collection was obtained
from Timothy Hughes (University of Toronto) (appendix III (Mnaimneh et al., 2004)).
Gene functions were determined using the Saccharomyces Genome Database (SGD)
website.
Ccw]4p analysis of SEC temperature sensitive alleles
Temperature sensitive strains of known SEC genes were grown to exponential
phase in YPD and shifted to 37'C for 2 hours. Cell lysates were made using the Pierce YPER protein extraction reagent. Cell lysates were solubilized in sample buffer (2% SDS,
10% glycerol, 80 mM Tris-HCl pH 6.8, 0.1 mg/ml bromophenol blue, 100 mM DTT) and
analyzed by SDS-PAGE and immunoblotting using anti-Ccwl4p antibody. Rabbit antiCcwl4p was used at 1:1000, HRP conjugated anti-rabbit was used at 1:10,000 (GE
Healthcare). Western blots were analyzed using an Image Station 440CF and 1D Image
Analysis Software (Kodak Digital Services)
Ccwl4p screening
TET strains were grown in YPD plus 10 gg/ml doxycycline for 8 hours to ~ .7
OD 60 0 /
ml. Cells extracts were made using Pierce Y-PER protein extraction reagent. For
each strain, ~1 OD (-1 x 107 cells) was analyzed by SDS-PAGE and immunoblotting
using anti-Ccwl4p antiserum as described above for the SEC temperature sensitive
alleles.
Invertase screening
139
For invertase, TET strains were grown in YPD (2% glucose) + 10 gg/ml
doxycycline for 6 hrs and shifted to low glucose YPD (.1%) + 10 pg/ml doxycycline for
2 hours to induce production of invertase. Cells were harvested, spheroplasted, and spun
down. The supernatant from spheroplasting was kept as the external sample (E) and the
pellet was solubilized as the internal sample (I). Samples were analyzed by SDS-PAGE
and immunoblotting with anti-invertase.
140
Results
CcwJ4p as a model cargo
The inner cell wall protein Ccw l4p (ICWP) was initially discovered in our lab as
a non-invertase band in a polyclonal invertase 137 antibody (M. Elrod-Erickson, PhD
dissertation, 1998). Interestingly this non-invertase band was detectable only in mutants
that affected the secretory pathway, and not in wild type cell lysates. This band was
identified as Ccwl4p. Ccwl4p is modified in the ER by glycosylation and addition of a
GPI anchor (Moukadiri et al., 1997; Mrsa et al., 1999). As it traverses through the Golgi,
it is further glycosylated. At the trans-Golgi GPI-Ccwl4p is packaged into vesicles and
ultimately transported to the inner cell wall, where it becomes covalently attached to the
glucans of the cell wall losing the GPI-anchor (Fig. 1) (Lesage and Bussey, 2006). Once
covalently attached to the cell wall, Ccwl4p is insoluble by most protein extraction
methods, and extractable only by the addition of glucanase to break down the cell wall
(Mrsa et al., 1999). This makes Ccwl4p an ideal marker protein for identification of
secretory pathway defects. In a wild type cell, very little Ccwl4p is detectable by
immunoblotting, as most has reached the cell wall and remains insoluble. If there is even
a slight defect in trafficking of Ccwl4p, it will accumulate in the GPI-anchored form
internally and be in the soluble fraction of cell lysates (M. Elrod-Erickson, PhD
dissertation, 1998).
Characterization of bands of CcwJ4p
The intracellular form of Ccwl4p that accumulates depends on the site of delay or
block in the secretory pathway.
Mutations that affect translocation (Fig. 1, step 1)
accumulate an unmodified form of Ccwl4p that runs at -25 kDa. Mutations that affect
141
ER modifications or ER to Golgi transport (Fig. 1, steps 2 and 3) accumulate bands that
are detectable between 75-100 kDa. Mutations that affect transport from the Golgi or
delivery to the cell wall accumulate higher molecular weight forms of Ccwl4p that are
seen between 120-170 kDa (Fig. 1, step 4). To verify these forms of Ccwl4p, we
analyzed the Ccwl4p phenotype of temperature sensitive alleles of known SEC genes
that act at each step in the pathway (Fig. 2). Wild type cells do not show an internal
accumulation of Ccwl4p. Genes involved in the early stages of the secretory pathway
accumulate two distinct forms of Ccwl4p. Genes involved in COPII vesicle formation
(SEC13, SEC23, SEC24, and SEC31) all accumulate a Ccwl4p band of -100 kDa.
Interestingly, genes that function in later stages of ER to Golgi transport (SEC18) or
genes involved in Golgi to ER retrograde trafficking (SEC20 and SEC21) accumulate a
slightly smaller form of Ccwl4p at ~80 kDa. This suggests that if there is a delay in
Ccw I4p exiting the ER, as would be the case with COPII mutants, Ccw l4p may undergo
additional glycosylation. Temperature sensitive alleles of genes involved in later stages of
the secretory pathway such as the exocyst complex (SEC8, SEC6 and SEC] 0) or other
steps of exocytosis (SEC4, SEC9 and SEC7) accumulate both the ~80 kDa and a higher
molecular weight (~150 kDa) form of Ccwl4p expected after Ccwl4p has been fully
glycosylated in the Golgi. The temperature sensitive allele of the translocation gene
SEC63 is extremely sick, and therefore accumulation of Ccw l4p is undetectable.
Screen of TET collection
In order to identify new genes required for Ccwl4p trafficking, we used the TET
collection of doxycycline regulatable essential genes (Appendix III) (Mnaimneh et al.,
2004). This TET collection replaces the endogenous promoter of each essential gene with
142
a promoter containing tetracycline response elements (TRE) called the TETO 7 promoter
(Fig. 3A. A tet activator (tTA*), encoded by a separate construct inserted at the URA3
locus, binds to the TRE to activate transcription of the downstream gene (Fig. 3B). When
doxycycline is added to the growth media, it enters cells and binds to the tTA* preventing
the activator from binding to the TRE, blocking transcriptional activation (Fig. 3B). This
transcriptional block leads to a reduction in downstream gene product over time based on
the rate of degradation of the already made protein product (Hughes et al., 2000;
Mnaimneh et al., 2004).
As a test to see if Ccwl4p accumulation could accurately identify secretory
proteins, we analyzed known SEC genes in the TET collection (Fig. 4). Each TETO 7SEC strain was grown in doxycycline for 8 hours to shut off protein production and then
analyzed by immunoblotting for Ccwl4p. As expected wild type cells do not show an
internal accumulation of Ccwpl4p. Sec61p is part of the ER translocation machinery and
a loss of Sec61p should prevent Ccwl4p from entering the secretory pathway. TetO7 Sec61 accumulated a low molecular weight band of ~25 kDa, which is the predicted size
of unmodified Ccwl4p. Sec13p and Sec31p are components of the COPII coat required
for vesicle trafficking from the ER and Secl8p is required for vesicle fusion with the
Golgi. As expected, all three of these TetO 7 strains show accumulation of the ER forms
of Ccwl4p (~80-100 kDa). The increase in apparent molecular weight is due to the
addition of the GPI-anchor and partial glycosylation that occur in the ER. Sec4p and
Sec 15p are both involved in post Golgi trafficking. Loss of either shows accumulation of
the higher molecular weight form of Ccwl4p (~150 kDa), likely due to additional
glycosylation in the Golgi.
143
The TET collection was analyzed by growing each strain for 8 hours in
doxycycline and examining Ccwl4p accumulation by Western blots. We use the TETO7 SEC18 strain as a positive control for Ccwl4p accumulation on our Western blots.
Mutations in SEC18 block ER to Golgi transport and accumulate a band of -80 kDa (Fig.
4). Since the wild type strain should transport Ccwl4p rapidly to the cell wall and not
have detectable accumulation, this strain was used as a negative control. A total of 773
strains (-75% of the essential genes in yeast) were screened for Ccwl4p accumulation
and 172 of these showed an accumulation of Ccwl4p (summarized in Tables II and III).
Figure 5 shows one Western blot from the screen.
The genes that when shut off cause an accumulation of Ccwl4p fall into four
categories (Table II). The first category contains 39 known secretory genes expected to
show a defect in Ccwl4p transport (Table III, purple). There are 55 genes in the
collection known to be involved in secretory function. The 16 of these that do not show a
secretory defect may be required for steps that are not required for Ccwl4p trafficking.
The second category contains 48 genes for which there is a known function not related to
the secretory pathway directly but for which it is possible to deduce how their function
may be connected to the secretory pathway (Table III, pink and blue). These include GPI
anchor and glycosylation genes as well as mRNA splicing genes. The third category
contains 79 genes for which there is a known function that does not clearly suggest an
involvement in secretory function (Table III, white). The fourth category contains 6 genes
for which little or nothing is known about their function (Table III, green).
Discussion and Future Directions
Known secretory genes
144
A good test of whether this screen is able to identify genes involved in the
secretory pathway is to examine the results of knockdown of the known secretory genes
in the collection. There are 55 known secretory genes in the starting TET collection. Of
these 55 genes, 39 of them show an accumulation of Ccwl4p (labeled purple in table III).
The 16 genes that did not show accumulation of Ccw l4p in this screen may be genes that
are not required for Ccwl4p trafficking or genes that are still produced at a high enough
level or have a long enough half life to allow function even after 8 hours of growth in
doxycycline. For example the exocyst subunit SEC8 did not show a Ccwl4p
accumulation phenotype even though other exocyst proteins did. Similarly, the COPI
component RET3 showed a Ccwl4p accumulation but SEC26 did not.
Protein modification genes
Many of the genes identified in this screen are required for protein folding and
modification (labeled pink in table III). These include many members of the GPI-anchor
synthesis and glycosylation pathways. As these are both modifications that are known to
be required for Ccwl4p trafficking and function, they are expected to show Ccwl4p
accumulation. Also in this category are ER folding chaperones such as Kar2, actin related
genes that are needed for vesicle transport, and phospholipid biosynthesis genes.
mRNA processing genes
There were 12 genes identified in the screen that are involved in RNA splicing
(labeled in blue in table III). After a eukaryotic gene is transcribed, the RNA transcript
undergoes modifications before being translated into protein. One of these modifications
is RNA intron splicing. Whereas most mammalian transcripts undergo splicing, only 4%
of yeast genes contain introns (Lopez and Seraphin, 1999).
145
One of these intron-
containing genes, SARI, is an essential small GTPase required for COPII vesicle
formation (Barlowe et al., 1994). If the intron is not spliced out of the transcript prior to
translation the protein is not functional and there is a complete block in protein transport.
Previous work identified RSE1 as having a secretory defect (Chen et al., 1998).
Upon further analysis, RSE1 was identified as a component of the U2 snRNP subunit of
the spliceosome. Addition of an intronless copy of the SARI gene in the presence of the
rsel mutant rescues the secretory defect, suggesting that the role of RSE1 is in RNA
processing of SARI, and is only indirectly playing a role in the secretory pathway. The
role of the genes identified in this screen could be tested by adding an intronless SAR1
plasmid and seeing if there is still a defect in Ccwl4p trafficking.
Genes with no clear secretory role
The majority of the genes (79) identified in this screen do not have an obvious
connection to the secretory pathway (labeled white in table III). Some of these may cause
a general disruption in cellular function while others may affect the secretory pathway
through an as yet unknown way. These genes have a wide range of functions in the cell
including RNA processing, ribosome assembly, chromatin remodeling, and transcription.
Genes of unknown function
When this screen was initially done there were 12 genes of unknown function that
showed Ccwl4p accumulation. Since then several of these genes have been assigned
functions. This includes three genes CAB 1, CAB3, and CAB5 that have been shown to
be required for biosynthesis of Coenzyme A (Olzhausen et al., 2009). yLR440c
(SEC39/DSL3) has been shown to be a part of the DSLl complex that is required for
tethering of COPI coated vesicles to the ER membrane (Kraynack et al., 2005; Mnaimneh
146
et al., 2004). There are currently 6 genes for which little is known about their function.
Four of these are listed as dubious ORFs, as they overlap other essential genes. These are
yDLO16c, yJL202c, yLLO37w and yPL142c. RTS2 is a zinc finger protein that may be
involved in DNA replication based on homology to mouse and human proteins (B6hm et
al., 1997). Characterization of yNL149c (PGA2) is shown in appendix II.
Future directions
Candidates identified in this screen as showing defects in trafficking of Ccwl4p
could be tested for trafficking of other known secretory proteins, such as the GPIanchored protein Gasl or the soluble proteins invertase or CPY to see if the effects are
specific to Ccwl4p, to GPI-anchored proteins, or to all secretory cargo. Figure 6 shows
that some of the Tet0 7 candidates do show a defect in invertase trafficking. The Ccwl4p
accumulation assay could be used to examine other collections of gene disruptions, such
as the non-essential deletion collection.
Four of the unknown genes are listed on SGD as dubious ORFs because they
partially overlap another known ORF. yDLO16c overlaps the 3' end of the DNA
replication gene CDC7, yJL202c overlaps the 3' end of the splicing factor PRP21,
yLLO37w overlaps the 3' end of the splicing factor PRP19, and yPL142c almost
completely overlaps the ribosomal subunit RPL33A. Interestingly, all of the genes
overlapped are essential genes. Only PRP21 was part of the TET collection, and it did not
show a Ccwl4p phenotype. Both PRP21 and PRP19 are involved in splicing, so could
have an indirect effect on the secretory pathway. To test if the unknown ORFs are
expressed, they could be genomically tagged in a region of the DNA that does not
overlap another gene, or if this isn't possible cloned into a plasmid for tagging. The
147
tagged constructs could then be used to test for protein expression. If the tagged protein is
detectable, this would suggest that these are true ORFS, which could then be further
characterized to determine if they play a direct role in the secretory pathway.
148
References
Barlowe, C., L. Orci, T. Yeung, M. Hosobuchi, S. Hamamoto, N. Salama, M.F. Rexach,
M. Ravazzola, M. Amherdt, and R. Schekman. 1994. COPII: a membrane coat formed by
Sec proteins that drive vesicle budding from the endoplasmic reticulum. Cell. 77:895-
907.
Bdhm, S., D. Frishman, and H.W. Mewes. 1997. Variations of the C2H2 zinc finger
motif in the yeast genome and classification of yeast zinc finger proteins. Nucleic acids
research. 25:2464-9.
Carlson, M., B.C. Osmond, and D. Botstein. 1981. Mutants of yeast defective in sucrose
utilization. Genetics. 98:25-40.
Chen, E.J., A.R. Frand, E. Chitouras, and C.A. Kaiser. 1998. A link between secretion
and pre-mRNA processing defects in Saccharomyces cerevisiae and the identification of
a novel splicing gene, RSE1. Molecular and cellular biology. 18:7139-46.
Costanzo, M., A. Baryshnikova, J. Bellay, Y. Kim, E.D. Spear, C.S. Sevier, H. Ding,
J.L.Y. Koh, K. Toufighi, S. Mostafavi, J. Prinz, R.P. St Onge, B. VanderSluis, T.
Makhnevych, F.J. Vizeacoumar, S. Alizadeh, S. Bahr, R.L. Brost, Y. Chen, M. Cokol, R.
Deshpande, Z. Li, Z.-Y. Lin, W. Liang, M. Marback, J. Paw, B.-J. San Luis, E. Shuteriqi,
A.H.Y. Tong, N. van Dyk, I.M. Wallace, J.A. Whitney, M.T. Weirauch, G. Zhong, H.
Zhu, W.A. Houry, M. Brudno, S. Ragibizadeh, B. Papp, C. Pal, F.P. Roth, G. Giaever, C.
Nislow, O.G. Troyanskaya, H. Bussey, G.D. Bader, A.-C. Gingras, Q.D. Morris, P.M.
Kim, C.A. Kaiser, C.L. Myers, B.J. Andrews, and C. Boone. 2010. The genetic landscape
of a cell. Science (New York, N.Y.). 327:425-3 1. doi:10.1 126/science. 1180823.
Davierwala, A.P., J. Haynes, Z. Li, R.L. Brost, M.D. Robinson, L. Yu, S. Mnaimneh, H.
Ding, H. Zhu, Y. Chen, X. Cheng, G.W. Brown, C. Boone, B.J. Andrews, and T.R.
Hughes. 2005. The synthetic genetic interaction spectrum of essential genes. Nature
genetics. 37:1147-52. doi:10.1038/ngl640.
Fujita, M., and T. Kinoshita. 2012. GPI-anchor remodeling: potential functions of GPIanchors in intracellular trafficking and membrane dynamics. Biochimica et biophysica
acta. 1821:1050-8. doi:10.1016/j.bbalip.2012.01.004.
Gemmill, T.R., and R.B. Trimble. 1999. Overview of N- and 0-linked oligosaccharide
structures found in various yeast species. Biochimica et biophysica acta. 1426:227-37.
Hashimoto, H., M. Abe, A. Hirata, Y. Noda, H. Adachi, and K. Yoda. 2002. Progression
of the stacked Golgi compartments in the yeast Saccharomyces cerevisiae by
overproduction of GDP-mannose transporter. Yeast (Chichester, England). 19:1413-24.
doi: 10.1002/yea.925.
149
Heider, M.R., and M. Munson. 2012. Exorcising the exocyst complex. Traffic
(Copenhagen, Denmark). 13:898-907. doi:10.1111 /j.1600-0854.2012.01353.x.
Hughes, T.R., M.J. Marton, A.R. Jones, C.J. Roberts, R. Stoughton, C.D. Armour, H.A.
Bennett, E. Coffey, H. Dai, Y.D. He, M.J. Kidd, A.M. King, M.R. Meyer, D. Slade, P.Y.
Lum, S.B. Stepaniants, D.D. Shoemaker, D. Gachotte, K. Chakraburtty, J. Simon, M.
Bard, and S.H. Friend. 2000. Functional discovery via a compendium of expression
profiles. Cell. 102:109-26.
Kaiser, C., S. Michaelis, A. Mitchell, and C.S.H. Laboratory. 1994. Methods in yeast
genetics: a Cold Spring Harbor Laboratory course manual. Cold Spring Harbor
Laboratory Press.
Kraynack, B. a, A. Chan, E. Rosenthal, M. Essid, B. Umansky, M.G. Waters, and H.D.
Schmitt. 2005. Dsllp, Tip20p, and the novel Dsl3(Sec39) protein are required for the
stability of the Q/t-SNARE complex at the endoplasmic reticulum in yeast. Molecular
biology of the cell. 16:3963-77. doi: 10.1091 /mbc.E05-01-0056.
Lesage, G., and H. Bussey. 2006. Cell wall assembly in Saccharomyces cerevisiae.
Microbiology
and
molecular
biology
reviews:
MMBR.
70:317-43.
doi: 10.11 28/MMBR.00038-05.
Lopez, P.J., and B. Seraphin. 1999. Genomic-scale quantitative analysis of yeast premRNA splicing: implications for splice-site recognition. RNA (New York, N.Y.).
5:1135-7.
Mnaimneh, S., A.P. Davierwala, J. Haynes, J. Moffat, W.-T. Peng, W. Zhang, X. Yang, J.
Pootoolal, G. Chua, A. Lopez, M. Trochesset, D. Morse, N.J. Krogan, S.L. Hiley, Z. Li,
Q. Morris, J. Grigull, N. Mitsakakis, C.J. Roberts, J.F. Greenblatt, C. Boone, C. a Kaiser,
B.J. Andrews, and T.R. Hughes. 2004. Exploration of essential gene functions via
titratable promoter alleles. Cell. 118:31-44. doi: 10.101 6/j.cell.2004.06.013.
Moukadiri, I., J. Armero, A. Abad, R. Sentandreu, and J. Zueco. 1997. Identification of a
mannoprotein present in the inner layer of the cell wall of Saccharomyces cerevisiae.
Journal of bacteriology. 179:2154-62.
Mrsa, V., M. Ecker, S. Strahl-Bolsinger, M. Nimtz, L. Lehle, and W. Tanner. 1999.
Deletion of New Covalently Linked Cell Wall Glycoproteins Alters the Electrophoretic
Mobility of Phosphorylated Wall Components of Saccharomyces cerevisiae. J. Bacteriol.
181:3076-3086.
Nothwehr, S.F., and T.H. Stevens. 1994. Sorting of membrane proteins in the yeast
secretory pathway. The Journal of biological chemistry. 269:10185-8.
150
Novick, P., C. Field, and R. Schekman. 1980. Identification of 23 complementation
groups required for post-translational events in the yeast secretory pathway. Cell.
21:205-15.
Olzhausen, J., S. Schibbe, and H.-J. Schi6ler. 2009. Genetic analysis of coenzyme A
biosynthesis in the yeast Saccharomyces cerevisiae: identification of a conditional
mutation in the pantothenate kinase gene CAB1. Current genetics. 55:163-73.
doi: 10.1007/s00294-009-0234- 1.
Palade, G. 1975. Intracellular aspects of the process of protein synthesis. Science (New
York, N.Y.). 189:347-58.
Schuldiner, M., S.R. Collins, N.J. Thompson, V. Denic, A. Bhamidipati, T. Punna, J.
Ihmels, B. Andrews, C. Boone, J.F. Greenblatt, J.S. Weissman, and N.J. Krogan. 2005.
Exploration of the function and organization of the yeast early secretory pathway through
an epistatic miniarray profile. Cell. 123:507-19. doi: 10.101 6/j.cell.2005.08.03 1.
Weerapana, E., and B. Imperiali. 2006. Asparagine-linked protein glycosylation: from
16:91R-101R.
Glycobiology.
systems.
to
prokaryotic
eukaryotic
doi: 10.1093/glycob/cwj099.
Wuestehube, L.J., R. Duden, A. Eun, S. Hamamoto, P. Kom, R. Ram, and R. Schekman.
1996. New mutants of Saccharomyces cerevisiae affected in the transport of proteins
from the endoplasmic reticulum to the Golgi complex. Genetics. 142:393-406.
151
Figures
Figure 1 Ccwl4p trafficking pathway. Following translation, Ccwl4p is translocated into
the ER (1) where it undergoes glycosylation and GPI-anchor addition (2). GPI-anchored
Ccwl4p is transported to the Golgi via COPII coated vesicles (3).
In the Golgi, GPI-
Ccwl4p is further glycosylated, before being transported via the plasma membrane to the
inner cell wall (4) where it becomes covalently attached to cell wall glucans, resulting in
loss of the GPI anchor.
152
= GPI anchor
= Ccw14p
= ER translocon
=covalent bond formation
153
Figure 2 Ccwl4p phenotypes of SEC temperature sensitive alleles. Wild type and
temperature sensitive SEC alleles were grown at 24'C to mid log phase and shifted to
37'C for 2 hours. Cells were harvested and cell lysates were made using Pierce Y-PER
protein
extraction
reagent.
Cell
lysates
were
analyzed
by
SDS-PAGE
and
immunoblotting with anti-Ccwl4p antibody. Arrows indicate the ER and post Golgi
forms of Ccwl4p. ER 1 is ~80 kDa, ER 2 is ~100 kDa and post Golgi is -150 kDa
154
C/N
N#///q
f
155
4-
Post-Golgi
4-
ER 2
ER 1
Figure 3 TET collection strains. (A) TetO 7-promotor alleles were constructed in strain
RI 158, which has the Tet "off' activator (tTA*) integrated at the URA3 locus, by
replacing the 100 bases upstream of the start codon of each gene with a cassette (kanRTetO 7-TATACYC 1) from plasmid RP188 by one-step integration and selection on G418
plates (Mnaimneh et al., 2004). (B) Schematic of TET collection regulation by
doxycycline. In the absence of doxycycline, the Tet activator (tTA*) binds to the TetO 7
promoter, and YFG (your favorite gene) is transcribed. When doxycycline is added, it
binds to tTA*, preventing binding to TetO 7 and therefore prevents transcription of YFG.
156
A.
=
in+
TAT
-
2
AF*
-
bp f55W tp
strain R1158 (WThaploid th
tTA* interated at URA3)
B.
tTA
In absence of doxycycline
41K
In presence of doxycycline
157
Figure 4 Sec genes in TET collection. TetO 7-SEC gene strains were grown for 8 hours in
YPD plus 10 gg/ ml doxycycline. Cells were harvested and cell lysates were made using
Pierce Y-PER protein extraction reagent. Cell lysates were analyzed by SDS-PAGE and
immunoblotting with anti-Ccwl4p antibody. Each sample was loaded into two adjacent
lanes. Arrows indicate the unmodified (-25 kDa), ER (80 and 100 kDa) and post Golgi
(~150 kDa) forms of Ccwl4p.
158
0.
4
4,
t.
,
4)
N
6
A$;
44
4
-
*
159
44-
Post-Golgi
4-
ER 1
4-
unmodified
ER2
Figure 5 Example of Western blot from TET screen. TetO 7 strains were grown for 8
hours in YPD plus 10 ptg/ ml doxycycline. Cells were harvested and cell lysates were
made using Pierce Y-PER protein extraction reagent. Cell lysates were analyzed by SDSPAGE and immunoblotting with anti-Ccw l4p antibody. Arrows indicate the ER (80-100
kDa) and post Golgi (~150 kDa) forms of Ccwl4p. Candidates from this gel are TRS 120,
yNL149c (PGA2), IQG1, and CDC54, all of which were retested to confirm the Ccwl4p
accumulation phenotypes.
160
I
N
Ne
ONO4"
a
161
4- Post Golgi
.sof
*
-
.M -
4-4--
ER 2
ER 1
Figure 6 Invertase analysis of TetO 7 strains. TetO 7 strains were grown 6 hours in YPD
(2% glucose) plus 10 pg/ ml doxycycline and shifted to YPD (.1% glucose) plus 10pg/
ml doxycycline for 2 hours to induce invertase expression. Cells were harvested and
spheroplasted. The supernatant from the spheroplasts was collected as the external
sample (E). Spheroplasts were solubilized for the internal sample (I). Lysates were
analyzed by SDS-PAGE and immunoblotting with anti-invertase antibody. Arrows
indicate the ER (-90 kDa) and post Golgi (~120 kDa) forms of invertase.
162
IE
E
I .E
I
E
I
E
I
E
44-
163
Post-Golgi
ER
Tables
Table I Strain Used
Strain
Genotype
Source
R1158
MATa URA3 :CMV-tTA his3A1 leu2AO
(Hughes et al., 2000)
metl5AO
TET Collection
MATa URA3::CMV-tTA his3Al leu2AO
met15AO kanR-TetO 7-TA TAcycj
CKY45
Matasec13-1 ura3-52 his4-619
CKY64
Matasec20-1 ura3-52 his4-619
CKY69
Mata sec21-1 ura3-52 his4-619
CKY78
Matasec23-1 ura3-52 his4-619
CKY149
Matasec4-8 ura3-52 leu2-3,112
CKY151
Mata sec6-4 ura3-52 leu2-3,112
CKY152
Matasec8-9ura3-52 leu2-3,112
CKY153
Mata sec9-4 ura3-52 leu2-3,112
CKY154
Mata seclO-2 ura3-52 leu2-3,112
CKY158
Mata sec63-1 ura3-52 leu2-3,112
CKY159
Mata sec7-1 ura3-52 leu2-3,112
CKY189
Matasec18-1 ura3-52 leu2-3,112
CKY263
Mata ura3-52 leu2-3,112
CKY545
Matasec24-2 ura3-52 leu2-3,112 Gal+
CKY555
Mata sec3l-2 ura3-52 leu2-3,112
164
(Mnaimneh et al., 2004)
Table II Summary of Screen
Category
Number of strains
Total strains in TET collection
773
Strains with Ccwl4p accumulation
172
Known Secretory genes in collection
55
Known SEC genes with Ccwl4p accumulation
39
Secretory pathway related genes
36
RNA splicing genes
12
Genes of unknown function
6
Other genes
79
165
Table III Summary of candidate genes
Gene Na
Ccw14p Phenotype
Protein Function
ALR1I
3
M02+ transporter
5
ARH1
ARP7
2
2
Mitochondrial oxidoreduts
Chromatin remoei
5
5
BRR6
1Nuclear
Function Categor
Eneoe5
CAR2
CDC1 1
CDC42
CDC45
CDC54
CFT2
4
2
3
2
2
4
Arginine catabolism
Cytokinesis
Cell polarity
DNA replication
DNA replication
mRNA processing
5
5
5
5
5
5
DAD2
DFR1
DRS1
4
2
4
Mitotic spindle organization
Tetra hy rofolate biosynthesis
Ribosome assembly
5
5
5
EPL1
1
H4/H2A acetyltransferase complex
5
tESA1
FAD1
FOL2
-GAR1
-GCR1
2
2
2
3
1
GLE1
GRC3
com lex
Histone acetyltransferase
FAD synthesis
Folic Acid synthesis
rRNA processing
Glycolysis Regulation
5
5
5
5
4
mRNA nuclear export
5.
2
Ribosome synthesis5
166
Table III continued
5
HEM12
2
Heme biosynthesis
HSF1
I
Heat shock transcription factor
5
HSP60
HTS1
HYP2
2
3
3
Mitochondria folding chaperone
Histidine tRNA synthetase
Translation elongation factor
5
5
5
Chromosome condensation
5
Ribosome assembly
Ribosome assembly
5
5
LOC7
LTO1
MAK11
3
2
MED8
3
5
proceing
RN
Methionine biosynthesis
MET30
2
MOB1
MRT4
2
2
Mioi euain5
Rbsmase
NDD1
NH P2
2
3
Clcylreuain5
rNprcsig5
NOP8
2
Riooeasml5
PG11
1
POB33
POL30
POP4
POP5
POP7
3
3
3
4
2
5
ly5
5olsi
Ncesmoraiton5
DArpiaon5
rcsig5
RN
RN
rcsig51
rcsig51
RN
RNA processing
167
MI
Table III continued
PZF1
3
rRNA snthesis
5
QNS1
4
NAD biosynthesis
5
RAD3
REB1
2
3
DNA helicase
RNA transcription
5
5
RFC5
2
DNA reolication
5
-RHC18
RNA15
ROK1
RPB1 1
-RPC40
RPP1
RPT2
-RRN1 1
RRP46
3
2
2
2
1
3
1
3
1
4
RSC4
1RNA
DNA repair
mRNA processing
rRNA processing
RNA polymerase 11
RNA polymerase 1l
RNA processing
Proteosome5
rRNA transcription
processing
5
5
5
5
5
5
Chromatin remodeling
5
168
5
51
Table III continued
SSU72
STU11
2
2
RNA synthesis
Mitotic spindle organization
b
5
TAF1_1
2
TEID transcription factor complex
5
TAF13
2
TFIID transcription factor complex
5
TAF3
TAF7
TEN1
TFC7
TFG1
TIM22
TIM50
1
2
2
4
2
3
3
TFIID transcription factor complex
TEID transcription factor complex
Telomere maintenance
RNA transcristion
RNA transcription
Mitochondrial translocation
Mitochondrial translocation
5
5
5
5
5
51
5
Chromatin remodeling
5
TTl1
1
UTP20
VHT1
1
2
rRNA synthesis
Vitamin H trans orter
5
5
YCG1
1
Chromosome condensation
5
Ccwl4p phenotypes: 1. Strong ER form 2. Strong ER and Golgi forms 3. Weak ER,
strong Golgi 4. Strong ER, weak Golgi 5. Strong unmodified 6. Streak
Function Category: 1. Known Secretory pathway (Purple) 2. Known function linkable to
secretory pathway (Pink) 3. RNA splicing (blue) 4. Unknown function (green) 5. Known
function not related to secretory pathway (white)
169
Appendix II: Analysis of PGA2
170
Abstract
The secretory pathway in S. cerevisiae is essential for the proper processing and
delivery of newly synthesized proteins to internal organelles and the plasma membrane.
Following translocation into the ER, properly folded proteins are packaged into COPII
coated vesicles for transport to the Golgi network. How cargo proteins are distinguished
from ER resident proteins and loaded into COPII vesicles is still unclear for many cargo
proteins.
I identified the gene PGA2 in a screen of essential S. cerevisiae genes that
affect intracellular protein trafficking of the GPI-anchored protein Ccw l4p (Appendix I).
PGA2 shares physical characteristics with EXPJ, which we show is required as a cargo
adaptor protein for transport of the plasma membrane ATPase Pmalp (Chapter 2). This
raises the possibility that Pga2p may function as a cargo adaptor for ER export. In this
chapter I present a preliminary functional analysis of Pga2p. When the amount of Pga2p
in the cell is decreased by using the TetO7 -PGA2 construct grown in doxycycline, the
cells show a defect in trafficking of not only Ccwl4p, but also the soluble cargo proteins
CPY and invertase. Pga2p is a type Ib integral membrane protein with a single Nterminal transmembrane domain and a cytosolic C terminus and localizes to the ER by
microscopy and sucrose gradients. Mutagenesis of the conserved regions of the C
terminus and analysis of fusions with Explp are currently underway with preliminary
results shown here.
171
Introduction
The secretory pathway is required for proper folding, modification, and targeting
of many newly synthesized proteins. This pathway consists of an ordered set of
organelles: the ER, Golgi, and plasma membrane, with a side branch to the endosome and
vacuole (Palade,
1975). Upon translocation
into the ER, a protein undergoes
modifications needed for proper folding and function including signal sequence cleavage,
GPI anchor addition, glycosylation, and disulfide bond formation (Weerapana and
Imperiali, 2006; Gemmill and Trimble, 1999; Fujita and Kinoshita, 2012). From the ER,
properly folded proteins are transported to the Golgi, a series of compartments that
further modify proteins through the addition of specific sugar residues (Hashimoto et al.,
2002). Fully processed proteins are packaged into secretory vesicles that go on to fuse
with either the plasma membrane or endosome.
Proteins necessary for formation of
vesicles at each stage are returned to the donor membranes for reuse.
In the ER, properly folded cargo proteins are separated from misfolded proteins
and ER resident proteins and packaged into COPII coated vesicles (Kuehn et al., 1998).
The COPII coat consists of Sarlp, Sec23p/Sec24p and Secl3p/Sec3lp (Barlowe et al.,
1994). Sarlp, a small GTPase, is activated and recruited to the ER membrane by Sec12p,
a guanine nucleotide exchange factor (Lee et al., 2005). Activated Sarlp-GTP recruits
the Sec23p/Sec24p subcomplex, which in turn recruits the Seci3p/Sec31p subcomplex
leading to vesicle formation (Belden and Barlowe, 1996; Stagg et al., 2006). Sec24p is
responsible for recognition and loading of many cargo proteins into vesicles and has
multiple distinct cargo binding sites along its surface (Mossessova et al., 2003; Miller et
al., 2003). However, not all cargo proteins contain known recognition sites for Sec24p
172
binding, suggesting that there are unidentified recognition sites or that some cargo
proteins do not interact directly with Sec24p.
While some cargo proteins interact directly with the COPII coat, soluble cargo
proteins in the ER lumen and some transmembrane cargo proteins require the assistance
of membrane spanning accessory proteins, or cargo adaptors, for efficient uptake into
COPII vesicles. We show in chapter 2 that Explp is required as a cargo adaptor for the
trafficking of the plasma membrane ATPase Pmalp. Pga2p, like Explp is a small single
membrane spanning protein with a highly charged C terminus. Due to this similarity and
the fact that loss of Pga2p interferes with trafficking of Ccwl4p (appendix I) we sought
to determine if Pga2p also acts as a cargo adaptor and to identify the cargo specificity of
Pga2p. Here we show that Pga2p, like Explp, is an ER localized single transmembrane
spanning protein with a cytosolic C terminus. We also show that in addition to Ccwl4p, a
loss of Pga2p causes a trafficking delay in both CPY and invertase. We are currently
analyzing Pga2p protein function by alanine mutagenesis and Pga2p/Exp 1p chimeras and
present some preliminary results here.
173
Materials and Methods
Media, Strains,and Plasmids
S. cerevisiae strains used in this study are CKY8 (MATa ura3-52 leu2-3,112),
TetO7-PGA2 (see appendix I) and Pga2p-3HA. Pga2p-3HA was made by homologous
recombination using the pFA6a-3HA-kanMX6 construct (Longtine et al., 1998).
Cells
were grown in either rich medium (YPD) or supplemented minimal medium (SMM), as
described in (Kaiser et al., 1994). For shut off of Pga2p in TetO 7-PGA2, the cells were
grown for 8 hours in YPD + 10 gg/ ml doxycycline at 24'C. pDMO1 was made by PCR
amplifying PGA2 from CKY8 genomic DNA and cloned into pRS315 (CEN LEU2).
Immunoblotting
Proteins were solubilized in sample buffer (2% SDS, 10 % glycerol, 80 mM TrisHCl pH 6.8, 0.1 mg/ml bromophenol blue, 100 mM DTT) and resolved by SDS-PAGE
according to standard protocols. Pga2p-3HA was detected using mouse monoclonal
12CA5 anti-HA at 1:1000. Rabbit anti-Explp, anti-Pga2p, and anti-Pdilp were used at a
dilution of 1:1000. HRP conjugated anti-rabbit and anti-mouse were used at 1:10,000
(GE Healthcare).
Microscopy
Pga2p-3HA strains were grown in YPD to -. 5 OD/ml. Cells were harvested and
fixed in formaldehyde for 1 hour, spheroplasted, attached to lysine coated slides and
blocked for 1 hour in PBS + 1% BSA.
After extensive washes in PBS, cells were
incubated with oHA 12CA5 (1:500) for 90 minutes, followed by 1 hour in secondary
(Alexa488 a-mouse). Digital photomicrographs were taken with a Hamamatsu Digital
camera attached to a Nikon Eclipse E800 microscope and visualized with OpenLab
174
software (Improvision). Alexa488-conjugated anti-mouse IgG (Molecular Probes) was
used at a dilution of 1:200. Mounting medium was supplemented with 4', 6-diamidino-2phenylindole (DAPI).
Protease accessibilityofExpip
Wild type (CKY8) cells were converted to spheroplasts with lyticase in 1.2 M
sorbitol, 10 mM Tris-HCl pH 7.4 at 30'C after treatment for 15 minutes with 0.5% Pmercaptoethanol. Washed spheroplasts were gently lysed in lysis buffer (250 mM
sorbitol, 150 mM potassium acetate, 20 mM HEPES pH 6.8, 1 mM magnesium acetate)
by Dounce homogenization on ice to create microsomes, and cell debris was removed by
centrifugation at 500 x g for 5 minutes.
Microsomal membranes were collected by
centrifugation of cleared cell extracts for 15 minutes at 13, 000 x g and treated with 0.75
mg/ml Trypsin in the presence or absence of 1% Triton X-100.
Proteolysis was
terminated with 25 mM Pefabloc (Roche) at the indicated times. Proteins were separated
by SDS-PAGE and analyzed by immunoblotting with antiserum to Pga2p, Explp or
Pdilp.
Cellfractionation
Subcellular distribution of Pga2p was examined as follows. Pga2p-3HA cells
were grown to exponential phase in YPD, converted to spheroplasts, and gently lysed by
agitation with glass beads in lysis buffer (50 mM Tris-HCl pH7.5, ImM EDTA, 200 mM
sorbitol) containing protease inhibitors. The cell extract was sequentially centrifuged at
500 x g for 10 minutes, 13,000 x g for 10 minutes, and 100,000 x g for 30 minutes.
Proteins solubilized in sample buffer were separated with SDS-PAGE and analyzed by
immunoblotting with antiserum to Exp Ip, HA (12CA5), or Dpmlp.
175
Release of Explp from the particulate/membrane bound fraction was performed
as described in Kaiser et al., 2002 with the following modifications. Pga2p-3HA cells
growing logarithmically in YPD were harvested and converted to spheroplasts.
Spheroplasts were lysed on ice by douncing in lysis buffer plus protease inhibitors. After
a clearing spin (5 minutes at 500 x g), cell extracts were treated with either 100 mM
Na 2 CO 3 pH 11.5; 1 M NaCl; 2.5 M urea; 1%Triton; or buffer alone and incubated on ice
for 1 hour. Treated extracts were centrifuged at 100,000 x g for 75 minutes to separate
membrane bound and soluble proteins. Samples were solubilized in sample buffer and
analyzed by immunoblotting with antiserum to Explp or HA (12CA5).
Sucrose density gradient fractionation was performed as described in Kaiser et al.,
2002. The presence of Pga2p, Explp, and Erolp in each fraction was detected by SDSPAGE followed by immunoblotting.
The relative amounts of each protein in cell
fractions were determined using an Image Station 440CF (Kodak Digital Sciences) and
1D Image Analysis Software (Kodak Digital Sciences). The presence of Golgi GDPase
activity was detected enzymatically as described in (Kaiser et al., 2002)
CPY Pulse Chase
TetO-PGA2 was grown to exponential phase in selective minimal medium
(SMM) with doxycycline (10 jg/ml) and resuspended at 5 OD600 units/ml in SMM minus
methionine. Cells were labeled with [ 35S] methionine and [ 35S] cysteine (EXPRESS;
PerkinElmer Life and Analytical Sciences) for 7 minutes and chased 30 minutes with
unlabelled methionine. Time points were collected at 10 minute intervals. At each time
point 2 OD6 0 0 equivalents were collected in 20 mM NaN3 and lysed by agitation with
glass beads in 30 pl of sample buffer containing 1 mM PMSF. Extracts were diluted to 1
176
ml in IP buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1%Triton X-100, 0.1% SDS).
Proteins were immunoprecipitated with polyclonal anti-CPY. Immunoprecipitates were
collected with protein A-sepharose (GE Healthcare), washed twice in IP buffer, once in
detergent-free IP buffer, and solubilized in sample buffer with 100 mM DTT. Samples
were resolved by SDS-PAGE (8% gel) and analyzed with a 445si Phosphorlmager.
Detection of Ccwl 4p and Invertase
For Ccwl4p detection, TetO 7-PGA2 was grown for 8 hours in YPD + 10 pg/ ml
doxycycline. Cell lysates were made using the Pierce Y-PER protein extraction reagent,
and lysates were analyzed by immunoblotting with anti-Ccwl4p antibody.
For invertase, TetO 7-PGA2 was grown in YPD (2% glucose) + 10 gg/ml
doxycycline for 6 hrs and shifted to low glucose YPD (.1%) + 10 pg/ml doxycycline for
2 hours to induce production of invertase. Cells were harvested, spheroplasted, and spun
down. The supernatant from spheroplasting was kept as the external sample (E), and the
pellet was solubilized as the internal sample (I). Samples were analyzed by SDS-PAGE
and immunoblotting with anti-invertase.
PGA2 Mutagenesis and Chimera Construction
Alanine mutants were created by site directed alanine mutagenesis of pDM1
(PGA2 LEU2) using the Stratagene Quik-Change PCR mutagenesis kit. Each construct
was confirmed by sequencing and transformed into pga2A + pDM09 (PGA2 URA3).
Strains were grown to exponential phase in SMM-Ura, Leu and spotted to SMM-Ura,
Leu or SMM + 5-FOA at 30'C for 3 days.
Chimeric proteins were constructed using a two step overlapping PCR reaction. A
pair of reverse complement primers were designed at the fusion site such that each primer
177
contained DNA sequence from both genes. In the first round of PCR each of these
primers was used with a primer to the 5' upstream or 3'downstream flanking region of
the genes creating two PCR fragments which each contained the sequences flanking the
fusion site. The two PCR products were mixed and PCR amplified in the absence of
primers for 5 cycles, followed by 30 cycles with the 5' or 3' flanking primers to amplify
the full length chimeric DNA sequence. The fusions were confirmed by sequencing,
cloned into pRS315 (LEU2) and transformed into pga2A + pDM09 (PGA2 URA3).
Transformants were grown on SMM-Ura, Leu or SMM + 5-FOA media for 3 days at
24 0 C.
178
Results
Trafficking defects in TetO 7-PGA2
In appendix I we describe a Western blot based assay to identify genes involved
in the trafficking of the GPI-anchored protein Ccwl4p. We used this assay to screen the
TET collection of essential genes under the control of a TetO 7 doxycycline regulatable
promoter (Mnaimneh et al., 2004). One of the genes identified in this screen was the then
uncharacterized ORF yNL149c (Fig. IA). During the course of this work, Yu and
colleagues identified yNL 149c in a screen of the TET collection using flow cytometry to
analyze the cellular DNA content after promoter shut off by addition of doxycycline. In
their study TetO 7-PGA2 grown in doxycycline showed a 3C/4C DNA content profile,
suggesting a defect in cell separation and named this gene PGA2 (Yu et al., 2006). We
used the doxycycline regulatable construct to determine if Pga2p was required for
trafficking of other cargo proteins. TetO 7 -PGA2 was grown in YPD + doxycycline for 8
hours and analyzed for processing of the soluble cargo proteins CPY and invertase.
CPY is a vacuolar protease that undergoes glycosylation in the Golgi and signal
peptide cleavage upon arrival in the vacuole. This allows for monitoring of trafficking, as
the size of CPY changes as the protein progresses through the secretory pathway with the
ER form (P 1) running smaller than the mature vacuolar form (M) due to a lack of
complete glycosylation, and the Golgi form (P2) running larger due to the presence of
both complete glycosylation and the uncleaved vacuolar signal sequence. A radioactive
pulse chase analysis was used to determine the ability of the Pga2p depleted cells to
transport CPY. Following a 7 minute labeling with
35
S methionine, cells were chased
with unlabelled methionine for 30 minutes with time points collected at 10 minute
179
intervals followed by immunoprecipitation of CPY. In wild type cells, CPY is seen only
in the mature form (M) by 10 minutes (Fig. 1B). In the TetO7-PGA2 strain, even at 30
minutes a large percentage of CPY remains in the ER unglycosylated form (P1) (Figure
1B). Yu et al. also examined the trafficking of CPY. However, they looked at total
protein by Western blot analysis after 15 hours of doxycycline treatment and did not see a
defect in CPY processing. They also demonstrated that TetO 7-PGA2 grown in
doxycycline for 15 hours has a defect in processing of alkaline phosphatase and possibly
the GPI-anchored protein Gasip (Yu et al., 2006).
Invertase is a glycosylated soluble enzyme encoded by the SUC2 gene, which is
required for hydrolysis of sucrose into glucose and fructose. In a wild type cell, invertase
is secreted into the extracellular space between the plasma membrane and cell wall. If
there is a delay in trafficking, invertase will accumulate internally and the location of
accumulation can be determined based on the amount of glycosylation. To determine if
Pga2p is needed for secretion of invertase, we examined the TetO7-PGA2 strain after
growth for 8 hours in doxycycline. Invertase is only produced under conditions of
glucose starvation, so cells were shifted to YPD with .1% glucose for the last 2 hours of
the doxycycline treatment. Cells were spheroplasted to release the invertase in the
extracellular space, and the supernatant was collected (E). Spheroplasted cells were
solubilized to examine intracellular invertase (I). Western blot analysis showed that in
wild type cells, invertase is found exclusively in the extracellular fraction and at a high
molecular weight (Fig. 1C). In Pga2p depleted cells, most of the invertase remains in the
intracellular fraction and runs at a much lower molecular weight suggesting a
glycosylation defect (Fig. 1C). As a control, TetO 7-SEC18 was also examined. Sec 18p is
180
required for COPII vesicle fusion at the Golgi. As expected, the loss of Sec 18p resulted
in accumulation of an intracellular lower molecular weight form of invertase (Fig. 1C).
Pga2p is a type Ib membrane protein
PGA2 encodes a 129 amino acid protein of unknown function with one predicted
transmembrane domain near the N terminus. To examine the cellular localization and
membrane association of Pga2p, I made a C-terminal genomically tagged protein using a
27 amino acid triple HA epitope (Pga2p-3HA). Since Pga2p is essential and Pga2p-3HA
is the only copy in the cell, the tagged protein must be functional. As an experimental test
of the intracellular distribution of Expip, lysates from Pga2p-3HA cells were converted
to spheroplasts and subjected to a series of centrifugation steps designed to separate
soluble proteins from those that are membrane bound. At both the 13,000 x g and
100,000 x g centrifugation steps, Pga2p-3HA was located in the pellet along with the
membrane bound markers Dpmlp and Explp, indicating that Pga2p is tightly associated
with cellular membranes (Fig. 2A).
To confirm that Pga2p is an integral membrane protein, spheroplasted cell lysates
were subjected to chemical treatments prior to centrifugation at 100,000 x g. Incubation
of cell lysates with high salt (500 mM NaCl), high pH (200 mM Na 2CO 3 pH 11.5), or 2.5
M urea, treatments known to perturb the interaction of peripheral membrane proteins
with cell membranes, had no effect on the association of Pga2p-3HA with the 100,000 x
g insoluble fraction. In contrast, treatment with 1% Triton was sufficient to release
Pga2p-3HA into the soluble fraction (Fig. 2B). This suggests that like Expl, Pga2p is an
integral membrane protein.
181
The highly charged C-terminal tail of Pga2p is predicted to be cytosolic based on
hydrophobicity and the positive-cytosolic rule using TopPred. The orientation was
experimentally determined by testing the accessibility of Pga2p-3HA in microsomes to
digestion with Trypsin. The C terminus of Pga2p-3HA in intact microsomal membranes
was efficiently degraded by Trypsin as shown by loss of HA signal, suggesting that the C
terminus is accessible to protease and therefore cytosolic (Fig. 2C). As a control for the
integrity of the microsomes under these conditions, we also examined the protease
accessibility of the lumenal ER chaperone PDI and found that PDI remained stable and
not accessible to protease unless the membranes had first been solubilized by detergent
(1% Triton) (Figure 2C). Together, these results show that Pga2p is a type lb integral
membrane protein, anchored to the membrane through its N-terminal transmembrane
domain with the C terminus facing the cytosol.
Pga2p localizes to the ER
To determine the cellular location of Pga2p, we used Pga2p-3HA for
immunofluorescence. Pga2p-3HA staining is seen in rings around the nucleus, suggesting
that Pga2p-3HA is in the ER (Fig. 3A). DAPI staining shows the location of the nucleus
in each cell. The ER localization of Pga2p was also examined using sucrose density
gradients. Different cellular membranes have distinct densities based on their lipid and
protein content, which can be used to separate organelle membranes using equilibrium
sedimentation on a sucrose gradient. ER membranes are unique in the fact that their
density is altered based on extraction conditions.
In the absence of magnesium (by
addition of EDTA), ribosomes dissociate from the ER resulting in a relatively low density
membrane similar to that of Golgi membranes. When magnesium is present, ribosomes
182
remain associated with the ER membrane, giving it a higher density such that the ER now
fractionates with the plasma membrane. Pga2p is detected in low density fractions in
EDTA buffer (which chelates magnesium) and high density fractions in Mg2 buffer (Fig.
3B). This shift in fractional location suggests that Pga2p is an ER protein.
Pga2p mutagenesis
Regions of genes vital for function are usually conserved in orthologous proteins.
Pga2p has orthologs in other Saccharomyces species and in other yeast species (Fig. 4).
There are no identifiable orthologs in higher eukaryotes.
Sequence alignments of the
orthologs show several conserved regions in the C terminus as well as conservation of
many residues of the transmembrane region (Fig. 4). These residues are being mutated by
alanine substitution and tested for their ability to rescue pga2A lethality by plasmid
shuffle. A starting pga2A strain containing a wild type copy of PGA2 on a URA3 marked
plasmid is used. Alanine mutants in a LEU2 marked plasmid are co-transformed into this
strain. Mutations that disrupt Pga2p function can be selected based on their sensitivity to
the toxic pyrimidine analog 5-fluoroorotic acid (5-FOA), which indicates that they need
to maintain the wild type PGA2 URA3 plasmid for viability. Table I shows preliminary
results of the alanine mutagenesis. The transmembrane region has two conserved arginine
residues (R26 and R38) and a conserved tyrosine (Y34). Mutagenesis of any of these
three residues to alanine does not disrupt Pga2p function.
So far 4 mutants in the C
terminus have been identified that disrupt Pga2p function. Mutation of residues 85-87
(GWG) to alanine results in no growth on 5-FOA plates. Mutations of residues 106-109
(EEAK), 125-137 (ELL) or 128-129 (EE) results in weak growth on 5-FOA suggesting a
partial defect in function. As was the case for Explp, disruption of the putative di-acidic
183
(DxD/E) motifs (116-124) did not affect Pga2p function. Additional alanine mutants are
being generated and tested.
184
Discussion and future directions
Soluble cargo proteins and some transmembrane proteins require cargo adaptor
proteins for ER exit via COPII coated vesicles. These cargo adaptors generally act as a
link between the cargo protein and the COPII coat and aid in recruitment of the cargo
protein into the vesicle. Cargo adaptor proteins share a set of common characteristics: (i)
transmembrane protein (ii) ER localization (iii) cycle between the ER and Golgi (iv) loss
of cargo adaptor causes a decrease the ability of one or more cargo proteins to exit the ER
(v) interact with Sec24p or its paralogs and/or load into COPII coated vesicles. In chapter
2 we show that Explp fits these criteria and is required as a cargo adaptor for Sec24p
mediated ER export of the plasma membrane ATPase Pmalp. Our screen for essential
genes that are involved in trafficking of the inner cell wall protein Ccwl4p identified
Pga2p (Appendix I). Like Explp, Pga2p is a small single transmembrane protein with a
highly charged C terminus raising the possibility that Pga2p may be acting as a cargo
adaptor. We have shown that Pga2p is a small transmembrane protein with a cytosolic C
terminus and that Pga2p localizes to the ER at steady state. A loss of Pga2p results in ER
export defects in the GPI-anchored protein Ccw l4p as well as the soluble cargo proteins
CPY and invertase. Thus, Pga2p meets criteria i, ii and iv above. We are currently
addressing the other criteria as described below. We are also continuing to analyze the
alanine mutants and Pga2p/Explp fusion proteins.
Does Pga2p cycle between the ER and Golgi
In wild type cells at steady state, Pga2p is localized to the ER membrane. If Pga2p
is acting as a cargo adaptor, it would likely be loaded into COPII coated vesicles and
therefore need to be retrieved from the Golgi. As an experimental test of the cycling of
185
Pga2p, we will compare the subcellular localization of Pga2p in wild type strains and in
strains expressing a mutant version of the COPI component Sec2lp. In cells expressing
sec21-1 at the semi-permissive temperature (30'C), retrograde transport from the Golgi to
the ER is blocked while anterograde transport from the ER to the Golgi remains
unaffected (Gaynor and Emr, 1997). Under these conditions, proteins that are normally
recycled from the Golgi back to the ER will accumulate in the Golgi, in later secretory
compartments such as the vacuole, or be secreted from the cell. Cell extracts from wild
type and sec21-1 strains grown at a semi-permissive temperature (30'C) will be
fractionated on a linear sucrose gradient in buffer containing magnesium, to separate
Golgi and vacuolar membranes (low density) from the ER and plasma membrane (high
density). If Pga2p is cycling between the ER and Golgi there should be a shift of Pga2p to
co-fractionate with the lower density fractions along with the Golgi marker GDPase in
sec21-1 compared to wild type. If Pga2p is not cycling, it will co-fractionate with the
higher density ER markers in both sec2l-1 and wild type cells.
Pga2p/Explp fusion analysis
The structural similarities between Pga2p and Explp raises the possibility that
there may be functional overlap of their domains. To test this we made fusion proteins
that combine different domains of each gene. A schematic of these fusions is shown in
figure 5. These fusions will be tested for functionality by their ability to rescue lethality
of a pga2A by the same method used to analyze the alanine mutants. Preliminary results
suggest that fusion PEP, which replaces the transmembrane region of Pga2p with the
transmembrane region of Exp lp, is able to rescue pga2A lethality but replacing the C
terminus of Pga2p with the C terminus of Exp lp (PPE) is not (data not shown).
186
Does Pga2p physically interact with the COPH coat proteins?
Since Pga2p is an essential protein, and a decrease in Pga2p shows intracellular
transport defects for at least three different cargo proteins, it is possible that Pga2p has a
general role in COPII vesicle trafficking. To determine the potential role of Pga2p in
COPII coat formation, I will look at interactions of Pga2p with known secretory genes, in
particular components of the COPII coat. The extreme C terminus of Pga2p contains two
previously identified Sec24p binding motifs (DxE and di-leucine) suggesting that it may
interact directly with Sec24p.
To test for direct binding of Pga2p with components of the COPII coat
biochemically, I have constructed a GST-Pga2p fusion protein replacing the N terminus
and transmembrane domain of Pga2p with GST (GST-Pga2p 42 -129).
Purified GST-
Pga2p 4 2 - 12 9 will be immobilized on glutathione-Sepharose beads to form an affinity
matrix. Wild type yeast cell extracts will be applied to this matrix as well as a matrix of
GST alone.
Elutions will be analyzed for binding of Sec24p to GST-Pga2 42-129 by
immunoblotting with Sec24p antibody. Binding of other COPII components or SEC
proteins can also be tested using antibodies for each. Comparison of quantities of protein
bound to GST-Pga2 42-129 and GST alone will identify proteins that bind non-specifically.
To determine the sites of Pga2p required for binding to Sec24p or other genes
identified above, I can use GST-Pga2p fusions containing mutants created based on
sequence homology. These mutant proteins will be tested for their ability to interact with
Sec24p or other proteins that bound GST-Pga2 42-129 as described above. If a mutation is
in a region of Pga2p required for a specific protein interaction, that protein will not bind
to the GST-Pga2p mutant protein matrix.
187
Is there a genetic interaction between Pga2p and known SEC genes?
Mutations in COPII component genes are synthetically lethal with other COPII
mutant alleles, but not with mutant alleles of genes involved in vesicle tethering, fusion to
the Golgi, or COPI mediated retrieval. Explp and the cargo adaptor Ervl4p both show
genetic interaction specifically with COPII coat proteins, suggesting this may be a
common characteristic of cargo adaptors (Chapter 2 and (Powers and Barlowe, 2002). To
investigate the relationship between Pga2p and the COPII coat proteins, we could
combine TetO-PGA2 and thermosensitive alleles of known secretory genes. These
strains could be grown in medium with or without doxycycline and at both the permissive
and restrictive temperatures to determine if the decreased production of Pga2p
exacerbates the temperature sensitivity of the COPII alleles or the growth rate of TetO7 PGA2. I can also test whether overexpression of Pga2p can rescue the temperature
sensitivity of the SEC mutants.
188
References
Barlowe, C., L. Orci, T. Yeung, M. Hosobuchi, S. Hamamoto, N. Salama, M.F. Rexach,
M. Ravazzola, M. Amherdt, and R. Schekman. 1994. COPII: a membrane coat formed by
Sec proteins that drive vesicle budding from the endoplasmic reticulum. Cell. 77:895907.
Belden, W.J., and C. Barlowe. 1996. Erv25p, a component of COPII-coated vesicles,
forms a complex with Emp24p that is required for efficient endoplasmic reticulum to
Golgi transport. J Biol Chem. 271:26939-26946.
Fujita, M., and T. Kinoshita. 2012. GPI-anchor remodeling: potential functions of GPIanchors in intracellular trafficking and membrane dynamics. Biochimica et biophysica
acta. 1821:1050-8. doi:10.1016/j.bbalip.2012.01.004.
Gaynor, E.C., and S.D. Emr. 1997. COPI-independent anterograde transport: cargoselective ER to Golgi protein transport in yeast COPI mutants. J Cell Biol. 136:789-802.
Gemmill, T.R., and R.B. Trimble. 1999. Overview of N- and O-linked oligosaccharide
structures found in various yeast species. Biochimica et biophysica acta. 1426:227-37.
Hashimoto, H., M. Abe, A. Hirata, Y. Noda, H. Adachi, and K. Yoda. 2002. Progression
of the stacked Golgi compartments in the yeast Saccharomyces cerevisiae by
overproduction of GDP-mannose transporter. Yeast (Chichester, England). 19:1413-24.
doi: 10.1002/yea.925.
Kaiser, C., E.J. Chen, and S. Losko. 2002. Subcellular fractionation of secretory
organelles. Methods in enzymology. 351:325-38.
Kaiser, C., S. Michaelis, A. Mitchell, and C.S.H. Laboratory. 1994. Methods in yeast
genetics: a Cold Spring Harbor Laboratory course manual. Cold Spring Harbor
Laboratory Press.
Kuehn, M.J., J.M. Herrmann, and R. Schekman. 1998. COPII-cargo interactions direct
protein sorting into ER-derived transport vesicles. Nature. 391:187-190.
Lee, M.C.S., L. Orci, S. Hamamoto, E. Futai, M. Ravazzola, and R. Schekman. 2005.
Sarlp N-terminal helix initiates membrane curvature and completes the fission of a
COPII vesicle. Cell. 122:605-17. doi: 10.1016/j.cell.2005.07.025.
Longtine, M.S., a McKenzie, D.J. Demarini, N.G. Shah, a Wach, a Brachat, P.
Philippsen, and J.R. Pringle. 1998. Additional modules for versatile and economical
PCR-based gene deletion and modification in Saccharomyces cerevisiae. Yeast
doi: 10.1 002/(SICI) 109714:953-61.
England).
(Chichester,
0061(199807)14:10<953::AID-YEA293>3.0.CO;2-U.
189
Miller, E.A., T.H. Beilharz, P.N. Malkus, M.C. Lee, S. Hamamoto, L. Orci, and R.
Schekman. 2003. Multiple cargo binding sites on the COPII subunit Sec24p ensure
capture of diverse membrane proteins into transport vesicles. Cell. 114:497-509.
Mnaimneh, S., A.P. Davierwala, J. Haynes, J. Moffat, W.-T. Peng, W. Zhang, X. Yang, J.
Pootoolal, G. Chua, A. Lopez, M. Trochesset, D. Morse, N.J. Krogan, S.L. Hiley, Z. Li,
Q. Morris, J. Grigull, N. Mitsakakis, C.J. Roberts, J.F. Greenblatt, C. Boone, C. a Kaiser,
B.J. Andrews, and T.R. Hughes. 2004. Exploration of essential gene functions via
titratable promoter alleles. Cell. 118:31-44. doi: 10.101 6/j.cell.2004.06.013.
Mossessova, E., L.C. Bickford, and J. Goldberg. 2003. SNARE selectivity of the COPII
coat. Cell. 114:483-495.
Palade, G. 1975. Intracellular aspects of the process of protein synthesis. Science (New
York, N.Y.). 189:347-58.
Powers, J., and C. Barlowe. 2002. Ervl4p directs a transmembrane secretory protein into
COPI-coated transport vesicles. Mol Biol Cell. 13:880-891.
Stagg, S.M., C. Giirkan, D.M. Fowler, P. LaPointe, T.R. Foss, C.S. Potter, B. Carragher,
and W.E. Balch. 2006. Structure of the Secl3/31 COPII coat cage. Nature. 439:234-8.
doi: 10.1038/nature04339.
Weerapana, E., and B. Imperiali. 2006. Asparagine-linked protein glycosylation: from
eukaryotic
to
prokaryotic
systems.
Glycobiology.
16:91R-101R.
doi: 10.1093/glycob/cwjO99.
Yu, L., L. Peia Castillo, S. Mnaimneh, T.R. Hughes, and G.W. Brown. 2006. A survey
of essential gene function in the yeast cell division cycle. Molecular biology of the cell.
17:4736-47. doi:10.1091/mbc.E06-04-0368.
190
Figures
Figure 1 Pga2p is required for trafficking of Ccwl4p, CPY and invertase. (A) TetO 7PGA2 was grown in YPD + 10 pg/ml doxycycline for 8 hrs. Cells were harvested and
cell extracts were made using Pierce Y-PER protein extraction reagent. Lysates were
analyzed by immunoblotting with anti-CCwl4p. (B) TetO 7-PGA2 was grown in SMM +
10 pg/ml doxycycline for 8 hrs, resuspended in SMM-methionine and radio-labeled with
35S
methionine for 7 minutes. Samples were taken at indicated time points after initiation
of chase with unlabeled methionine. Each sample was lysed with glass beads,
immunoprecipitated with x-CPY, and analyzed by SDS-PAGE and phosphoimaging. (C)
TetO 7-PGA2 was grown in YPD + 10 pg/ml doxycycline for 6 hrs and shifted to low
glucose YPD (.1%) + 10 gg/ml doxycycline for 2 hours to induce production of
invertase. Cells were harvested, spheroplasted, and spun down. The supernatant was kept
as the external sample (E) and the pellet was solubilized as the internal sample (I).
Samples were analyzed by SDS-PAGE and immunoblotting with anti-invertase.
191
A.
-
120 kDa
4.
80 kDa
Is
*No
cp%
z~
:
B.
P2
Wild Type
+- P1
+ m
TetO7-PGA2
($b\
Je
C.
0'~
I E
I
E
p
I E
.-
120 kDa
90 kDa
192
Figure 2 Pga2p is a type lb integral membrane protein. (A) Pga2p-3HA cells were grown
to exponential phase in YPD. Cells were harvested, spheroplasted and lysed with glass
beads. Cleared cell extracts were sequentially centrifuged at 500 x g for 10 minutes,
13,000 x g for 10 minutes, and 100,000 x g for 30 minutes. An aliquot of the cleared cell
extract was removed prior to centrifugation (Total lysate). Proteins in the soluble and
particulate fractions after each spin were analyzed by SDS-PAGE and immunoblotting
with antisera against HA (12CA5), PGK, or Dpmlp. (B) Pga2p-3HA cell extracts were
treated for 1 hour at 4'C with 1%Triton X-100, 100 mM Na2 CO 3 pH 11.5, 2.5 M urea,
500 mM NaCl, or buffer alone. Treated samples were separated into soluble (S) or
particulate (P) fractions by centrifugation at 100,000 x g for 75 minutes. Samples were
solubilized in sample buffer and analyzed by SDS-PAGE and immunoblotting with
antiserum to HA (12CA5) or Dpmlp. (C) Microsomes generated from Pga2p-3HA cell
extracts were digested with 0.75 mg/ml Trypsin in the presence or absence of 1%Triton
X-100.
Proteins were analyzed by immunoblotting with antiserum to HA (12CA5),
ExpIp or Pdilp, a lumenal ER protein.
193
'9
B.
~.1
A.
.0
AS
'.1
4-,
'-I
a.2-
Q
~,
1.0.
~#)
fl~
Pga2p
~
~,
4
n,
-
Q
#.A
s PS P S PSP S P
QO'
o0
Pga2p
Expip
El
C.
e
*
(1Z
c
Time (minutes)
PGA2
(C-terminal HA tag)
05 15 3060 05 15 30 60 05 15 3060 05 15 30 60
-ALMLAU*
Aww",
wMwW-
PDI
(lumenal ER marker)
194
*IMiW
Figure 3 Pga2p localizes to the ER. (A) Pga2p-3HA strains were grown in YPD to ~.5
OD 600 /ml.
Cells were harvested and fixed in formaldehyde for 1 hour, spheroplasted,
attached to lysine coated slides and blocked for 1 hour in PBS + 1% BSA.
After
extensive washes cells were incubated with axHA (12CA5) for 90 minutes, followed by 1
hour in secondary (Alexa488 a-mouse).
DAPI was added to the mounting medium.
Images were taken with a Nikon E800 microscope and image analysis was performed
using Improvision OpenLab software. (B) Wild type (CKY8) cell membrane extracts
were fractionated on 20-60% sucrose density gradients containing either 10 mM EDTA
or 2 mM MgCl2. Fractions were collected from the top of the gradient. Relative levels of
Pmalp, Pga2p, and Dpmlp in each fraction were quantitated by immunoblotting and
densitometry. GDPase activity in each fraction was assayed enzymatically. Lines under
the Western blots show the fraction locations of Dpml (ER) and GDPase (Golgi).
195
A.
DAPI
PGA2-HA
B.
Fraction Number
12 3 4 5 6 7 8 9101112 13141516
EDTA
mi ER
-
Golg
Mg 2.
Golgi
ER
A;
196
Figure 4 Pga2p has orthologs in other yeast species. Yeast orthologs were identified
using BLASTP. Alignments were generated using ClustalW with highlighting set at 50%
identity. The predicted transmembrane region is indicated.
197
Transmembrane
20
10
30
S.
K.
A.
D.
C.
V.
S.
corevisiae
lactis
gossypii
hansenii
albicans
polyspora
pombe
Y. lipolytica
1
1I
1
1I
1
1
1
1
00
K. lactis
A. gossypii
D. hansenii
C. albicans
V. polyspora
c
S. pombe
Y. lipolytica
55
53
53
48
50
48
52
47
1mylas
C. albicans
V. polyspora
S. pombe
y. lipolytica
...........
.......
....
...
......
. ...
---
SSAmm--
ElW
80
90
100
110
120
105
DEEIG SK------DLID-KPD--
AE&3
RSVAA------GTSDGSTADQGZS
S
-------LVEDNQAPZAESTP---
1 106
102
-------
100
S 102
------
-D
GEVDGTIDPT II'mI111
KRit L=TAAR- -A
100
TDa
S
150
140
104
GSAI
160
. . I....I.... I... .l....I
S. cerevisiae
54
52
52
47
49
47
51
46
----
ayISBTAKQ
130
K. lactis
A. gossypii
D. bansenii
60
50
llT~aaM~i---
70
S. carevisia*
40
1
.1.
106
107
103
D--D ~
101
103 D -----KSL
D
XIMLK
-101
112
105
D
-
W
X--
D
L
-----
DTE------------LLE
----LL
-1E-, -------D
------
129
130
124
125
127
129
132
137
Figure 5 Schematic of Pga2p/Explp fusion proteins. Chimeric proteins combining the N
terminus (N), transmembrane (TM) and C terminus (C) of Explp and Pga2p were made
using overlapping primers in a 2-step PCR reaction. The functionality of each chimera
will be tested by rescue of pga2A lethality using plasmid shuffle.
199
N
TM
C
EEEPEE
PPE
EPE
EPP
EEP1
PEP
200
Tables
Table I Alanine mutagenesis of Pga2p
Mutated Residues
Rescue pga2A lethality
R26
YES
G32
YES
G33
YES
Y34
YES
R38
YES
P39
YES
GWG85-87
NO
RRR91-93
YES
VKR94-96
YES
LFE100-102
YES
EEAK106-109
PARTIAL
PDS117-119
YES
DIE122-124
YES
ELL125-127
PARTIAL
EE128-129
PARTIAL
Alanine mutants (LEU2) were transformed into pga2A + PGA2 (URA3) and spotted on
5-FOA containing medium to test ability to lose the PGA2 (URA3) plasmid.
201
Appendix III: Exploration of Essential Gene Functions via
Titratable Promoter Alleles
The contents of this appendix have been published as:
Mnaimneh S, Davierwala AP, Haynes J, Moffat J, Peng WT, Zhang W, Yang X,
Pootoolal J, Chua G, Lopez A, Trochesset M, Morse D, Krogan NJ, Hiley SL, Li Z,
Morris
Q, Grigull J, Mitsakakis N, Roberts
CJ, Greenblatt JF, Boone C, Kaiser CA,
Andrews BJ, Hughes TR. Exploration of essential gene functions via titratable
promoter alleles. Cell. 2004 Jul 9;1 18(1):31-44.
Copyright C 2004 Cell Press All rights reserved.
Cell, Volume 118, Issue 1, 31-44, 9 July 2004
I contributed the analysis of yLR440c (SEC39) shown in figure 6 and used this gene
collection in the Ccwl4p screen described in appendix I.
202
Cell, Vol. 118, 31-44, July 9, 2004, Copyright @2004 by Cell Press
Exploration of Essential Gene Functions
via Titratable Promoter Alleles
lar biology, due in part to the ease of making targeted
integrations in the genome. Recently, a consortium of
laboratories created a collection of isogenic knockouts
for virtually all yeast genes, enabling an unparalleled
degree of unbiased functional analysis (Giaever et al.,
2002). We previously utilized this collection to generate
a "compendium" of microarray expression profiles
(Hughes et al., 2000). Application of statistical classification algorithms to this compendium showed that patterns of transcriptional changes could be used to distinguish perturbation of diverse cellular functions and
thereby assign gene function or link bioactive compounds to their intracellular target pathways (Hughes et
al., 2000).
A major outcome of the deletions consortium effort
was the demonstration that 1,105 (18.7%) of the -5,800
protein-coding yeast genes (Cliften et al., 2003; Kellis
et al., 2003) are essential for haploid viability under standard laboratory conditions (growth at 300C in rich medium with glucose as the carbon source). Transcription,
splicing, ribosome biosynthesis, translation, cell wall
and membrane biogenesis, DNA replication, nuclear
transport, and basic cytoskeletal functions are all required for cell proliferation, and genes involved in these
processes tend to be essential. Specific aspects of other
cellular functions are also required for viability. For example, although the vast majority of mitochondrial proteins are nonessential, some components involved in
mitochondrial import and iron-sulfur complex assembly
are essential, presumably because of the role of mitochondria in cellular iron metabolism (Baker and Schatz,
1991; Lill and Kispal, 2000). Essential genes tend to
be more highly conserved in humans; 38% of essential
yeast proteins have counterparts in humans, versus
20% for nonessential genes (with a Blastp E value of
E-50; Hughes, 2002).
The precise molecular and genetic functions of many
essential yeast proteins have not been studied in detail,
at least in part because it is difficult to study essential
genes using deletion mutants. Heterozygous diploids
have proven to be useful in screens for haploinsufficiency in the presence of specific insults (Giaever et al.,
1999, 2004; Lum et al., 2004), but in the absence of an
insult most of them have little or no growth defect in
rich medium and do not manifest the full range of phenotypes associated with alleles that cause growth inhibition (C.R. and T.R.H., unpublished data). Genetic analysis of essential proteins has traditionally relied on
conditional mutants, typically temperature-sensitive (ts)
alleles (e.g., Hartwell et al., 1970). A systematic method
for constructing ts alleles involves the construction of
a gene fusion of the coding sequence to create a heatinducible-degron domain, which modulates the stability
of the protein (Dohmen et al., 1994; Kanemaki et al.,
2003). Another systematic approach involves replacement of the native promoter with one that can be rapidly
repressed. We and other researchers (Hughes et al.,
2000; Peng et al., 2003; Zhang et al., 2003) have employed the tetracycline (tet)-regulatable promoter (Gari
et al., 1997) to create mutant alleles of essential genes.
Sanie Mnaimneh,'' Armaity P. Davierwala,'-5
2
Jennifer Haynes,2 Jason Moffat, Wen-Tao Peng,'
12
Wen Zhang, , Xueqi Yang,' Jeff Pootoolal,'
Gordon Chua,' Andres Lopez,' Miles Trochesset,'
Darcy Morse,3 Nevan J. Krogan,1 ,2
Shawna L. Hiley,' Zhijian Li, 1,2 Quaid Morris,'
J6rg Grigull,' Nicholas Mitsakakis,'
Christopher J. Roberts,' Jack F. Greenblatt,1' 2
Charles Boone,1 , 2 Chris A. Kaiser,
Brenda J. Andrews,2 and Timothy R. Hughes,2,*
'Banting and Best Department of Medical Research
University of Toronto
112 College Street
Toronto, ON M5G 1L6
Canada
2 Department of Medical Genetics and Microbiology
University of Toronto
1 Kings College Circle
Toronto, ON
Canada
'Department of Biology
Massachusetts Institute of Technology
Cambridge, Massachusetts 02139
4 Rosetta Inpharmatics LLC, a wholly owned
subsidiary of Merck & Co., Inc.
401 Terry Avenue North
Seattle, Washington 98109
Summary
Nearly 20% of yeast genes are required for viability,
hindering genetic analysis with knockouts. We created
promoter-shutoff strains for over two-thirds of all essential yeast genes and subjected them to morphological analysis, size profiling, drug sensitivity screening,
and microarray expression profiling. We then used this
compendium of data to ask which phenotypic features
characterized different functional classes and used
these to infer potential functions for uncharacterized
genes. We identified genes involved in nbosome biogenesis (HASI, URBI, and URB2), protein secretion
(SEC39), mitochondrial import (MIMI), and tRNA
charging (GSN1). In addition, apparent negative feedback transcriptional regulation of both ribosome biogenesis and the proteasome was observed. We furthermore show that these strains are compatible with
automated genetic analysis. This study underscores
the importance of analyzing mutant phenotypes and
provides a resource to complement the yeast knockout collection.
Introduction
The budding yeast Saccharomyces cerevisiae is one of
the most thoroughly characterized organisms in molecu*Correspondence: t.hughes@utoronto.ca
6
These authors contributed equally to this work.
203
Cell
32
A major advantage of promoter-replacement systems
is that the native open reading frame of the gene is
maintained. Moreover, with the tet system, repression
is controlled by addition of doxycycline to the growth
medium, which has little effect on yeast physiology and
virtually no effect on global gene expression at concentrations used for promoter shutoff, which is critical for
analysis using microarray expression profiling (Hughes
et al. 2000; Peng et al., 2003).
To facilitate uniform and systematic analysis of yeast
essential genes, we have now created TetO7-promoter
alleles for over two-thirds of all yeast essential genes.
We describe an initial phenotypic characterization of
602 of these strains that encompasses morphological
characterization, size profiling, drug sensitivity analysis,
and expansion of our previous "compendium" of expression profiles to include 215 of the TetO7-promoter mutants. We show that precise statistical inferences regarding gene functions can be drawn from patterns
within all of these data types, and we present in-depth
examination of several mutants in previously uncharacterized essential genes. These strains and the data presented here are a critical counterpart to sequence features, localization, and protein-protein interactions that
have been compiled on a subset of yeast essential genes
(Rout et al., 2000; Uetz et al., 2000; Ito et al., 2001; Gavin
et al., 2002; Ho et al., 2002; Kumar et al., 2002; Huh et
al., 2003; Hazbun et al., 2003; Sickmann et al., 2003;
Krogan et al., 2004). In some cases, however, we obtained dramatic phenotypes from mutants in genes for
which there is little or no functional information from
sequence or other large-scale studies. These strains are
publicly available, and since they are derived from the
deletions consortium wild-type MATa strain, they are
compatible with most screens designed to use haploid
yeast cells. To illustrate this, we show that the array of
TetOr-promoter mutants is compatible with synthetic
genetic array analysis (Tong et al., 2001, 2004).
previously described (Chun and Goebl, 1997) (Figure
2A), and several mutants in the secretion pathway displayed a "wide bud neck" (TetO7-SEC8 is shown in Figure 2A). Second, we measured the size distribution of
individual cells in cultures of each strain using a Coulter
Channelizer (Figure 2B). Third, we measured colony size
under 15 diverse growth conditions, using a colony
arraying robot (Tong et al., 2001) (Figure 2C). Fourth, we
examined 215 of the TetO7-promoter strains by microarray expression profiling (Figure 2D). These 215 strains
were hand selected to represent a wide variety of functional classes and to encompass all mutants in uncharacterized genes that displayed a growth defect.
Results
Construction of 602 TetOT-Promoter Strains
We constructed 602 yeast strains, each with a kanR
tetO7-TATA cassette integrated into the promoter of a
different essential gene in strain R1 158 (Hughes et al.,
2000) (Figure 1A). In the presence of doxycycline (10
pg/ml), 74% (447) yielded colonies smaller than wildtype upon visual inspection (Figures 1B and 1C). Among
these, 20% (123) yielded smaller colonies even in the
absence of doxycycline (Figure 1B, boxed in red; Figure
1C), suggesting that these genes are highly sensitive to
perturbation of promoter activity. Colony size reflected
growth rate in liquid (Figure 1D).
Systematic Phenotypic Analysis
We applied four general purpose phenotypic assays to
initially characterize the 602 strains. First, we examined
the morphology of each strain in the presence and absence of 10 pg/mI doxycycline by manually scoring each
strain with one or more of 16 morphological descriptors.
This confirmed expected phenotypes for many previously characterized genes: for example, the TetOrCDC53 strain displayed an "elongated" morphology, as
204
Phenotypic Features that Characterize Mutants
in Specific Functional Categories
We next sought to objectively determine whether specific features or patterns in these data correlated with
known functions of the mutated genes, and whether
these patterns could be used as diagnostics to identify
new genes with related functions. While clustering analysis (Eisen et al., 1998) (e.g., Figure 2D) is useful for
gaining an initial overview of the data, it is not ideal
for surveying all relationships among gene classes and
features in the data set (Wu et al., 2002; Gasch and
Eisen, 2002) and it does not automatically determine
how well pattems in the data correspond to distinct
functional classes. Therefore, we applied Support Vector Machines (SVM) (Brown et al., 2000), a machine learning algorithm (i.e., a computer program) that seeks features or patterns of features in the data that separate two
classes, e.g., genes that are known to be in a functional
category and those that are known to not be in the
category. SVM has previously been established to work
well for predicting gene functions in yeast on the basis
of microarray expression data, and it can also be applied
to other types of data that can be represented in a
spreadsheet, such as the size, drug sensitivity, and morphology data that we collected. SVM considers each
functional category separately and outputs a single "discriminant value" for each gene that represents relative
confidence that the gene is in the given category. These
values can be processed to obtain an estimate of the
probability that the prediction for each gene in each
category is correct (i.e., "precision"), on the basis of
how well previously annotated genes in the given category can be distinguished from previously annotated
genes that are not in the category. This is accomplished
by performing the analysis several times, on each repetition "masking" the known functions of a portion of the
genes to ask how well they are classified. In this way,
SVM can identify which gene categories are distinguishable in the data and can also provide a relative confidence that an uncharacterized gene is in any given category, on the basis of the phenotype displayed by the
mutant in that gene.
For each of 643 gene ontology (Ashburner et al., 2000)
"biological process" categories represented among the
602 mutants, we trained separate SVMs for each data
set. We also trained SVMs for each data set in which
the mutants considered were restricted to the 215 mutants that were included in the microarray analysis. After
excluding eleven broad categories in which predictions
Exploring Essential Gene Functions
33
B
A
YPD
YPD+Dox
WT
TetO7-SEC4
TetO 7-PSA1
WT
TetO7-YGR198W
TetO 7-NUP192
TetO7-YDR327
WT
TetO 7-RPL18A
TetQ7-YNL149C
48 hours
WI
WT
TetO 7-RPL18A
TetO 7-YNL149C
WTM,
94$
|0
72 hours
D
C
Constitutive Reverse
Slow Growth
10%
3%
Normal
26%
Very
Severe
24%
= 0
Slight
19%
Severe
35
30
251
20
15
10
5
0
Growth on
agar plates:
WT
Const. Normal Slight Severe Very
Gsow
Severe
18%
Figure 1. Construction and Growth Characteristics of tot-Promoter Strains
(A) Strain construction by one-step homologous replacement of native promoters with a TetO-containing cassette. YFG1, your favorite gene;
NGC1, next gene on chromosome.
(B) Examples of different growth phenotypes on YPD-agar plates. TetO-SEC4 and TetO-PSA1, very severe; TetO-YGR198W, severe; TetOr
NUP192, normal; TetO-YDR327W, reverse (i.e., growth slightly better in presence of doxycycline); TetO-RPL18A, TetO-YNL149C, constitutive
slow growth.
(C) Proportion of the 602 strains with different growth phenotypes. A subset of strains in the slight, severe, and very severe categories also
exhibit constitutive slow growth that is accentuated by growth on doxycycline.
(D) Correspondence between growth on YPD-agar plates and doubling time measured in liquid SC medium (2% dextrose) between 18 and
24 hr after addition of doxycycline (10 sig/ml).
shows phenotypic data for mutants in selected categories that were identified in this analysis as being "predictable" by one or more data types. These mutants represent genes that were previously known to be in the
category as well as those that are not known to be in
the category. The drug/media sensitivity profiles were
less informative for these categories and are not shown.
Since it is difficult to determine exactly which parts
of the data the SVM uses to make each prediction, we
used Fisher's score to ask which of the individual features in the morphology, drugs, and expression data
distinguish mutants in each of the functional categories
(values for all measurements in all categories are listed
in Supplemental Data online). For example, among the
mutants in Figure 3, the morphology category "amorphous" was characteristic of mutants in ubiquitindependent protein catabolism, whereas the categories
"large" and "round" were characteristic of mutants in
do not lead to testable hypotheses (listed in Supplemental Data at http://www.cell.com/cgi/content/fuV118/1/
31/DC1), we asked how many categories and how many
genes achieved precision values of over 50% (i.e., in a
blind test, among previously annotated genes identified
by the algorithm as being in the class, at least 50%
are indeed in that class). This was accomplished for 61
categories in the morphology data, 43 categories in the
size data, 64 categories in the drugs data, and 79 categories in the microarray expression data. Using the same
SVMs, a total of 75 unannotated genes were classified
in at least one category with at least 50% precision
by at least one assay. Frequently, these classifications
were supported by multiple data types, particularly if
lower confidence thresholds were allowed. All of the
predictions with precision values above 5%, for both
characterized and uncharacterized genes, are found in
Supplemental Table S1 on the Cell website. Figure 3
205
Cell
34
A
B
256 size bins on
Coulter counter
small
250
big
-
0
toc
%
0
600 7
500
Median over 602 strains
TetO,-MAK 16 (smal, unbudded)
-TelO-ARC40
400
4-TeOrPREO
("ysis debir)
(hyperpoarized)
300
-5
A
2200
100
0
0z
ins- ou
Size bins on coulter counter
C
- GTetOy-T/M50
Q TetO-TOM22
TetO7-RIM2
YPD + doxycycline
YPglycerol + doxycycline
-0.5
D
0
Log(ratio)
Functional categories of mutated gene
0.5
scrpt
6307tra
Figure 2. Four Phenotypic Assays Applied to the TetO-Promoter Strains
(A) Morphological characteristics, assayed by microscopy. The strains shown are round, TetOrY-P1; pointed, TetOr-NUP192; elongated,
TetO-CDC53; amorphous, TetOrRPT6; multiple buds, TetO-COG4; wide bud neck, TetO7-SEC8; hyperpolarized buds, TetO-PRE6; lysis,
TetO7-ARC40; clumped, TetO-HRR25.
(B) Top, cell size distributions, from smallest to largest overall distribution. Bottom, examples of cell size distributions from individual strains,
in comparison to the median cell size distribution over all experiments (blue line).
(C) Growth phenotypes assayed by robotically depositing colonies on plates.
(D)2D hierarchical clustering analysis (Elsen et al., 1998) of microarray expression data for 6,307 yeast genes in 215 different TetO-promoter
strains. The full data matrix is given in the Supplemental Data. Representation of nineteen GO categories among the mutants, selected on
the basis of being "learnable" by SVM, are shown at right.
mutants in genes involved in amino acid activation featured transcriptional induction of genes involved in
amino acid biogenesis; these are most likely regulated
by Gcn4p, a transcription factor that is known to regulate
amino acid biogenesis and whose translation is coupled
to tRNA charging (Vazquez de Aldanaet al., 1994) (Figure
3). Mutants In genes involved in protein-mitochondrial
the secretory pathway. The microarray measurements,
however, were able to correctly distinguish the largest
number of distinct categories (Figure 3 and Supplemental Data). In several cases, the identities of the genes
whose transcript levels increased are consistent with
what is known about the functional categorization of
the mutants that caused their induction. For example,
206
I
Exploring Essential Gene Functions
35
Microarrav exoression
Current annotations
Morphology
ORF
Gene
GO-Biological Process
incatabolism YOR261C RPN8
ubqu in-dependent
RPT4
YR259C
catabolism
ubiquitidependent
ublqit'n-dependent protein cataboIsm YGLO48C RPT6
ubU tindependent protein catabolism YOL.038W PRE6
ubiqulfin-dependent protein catabolism YDL.132W CDC53
ubin-dependent protein catabolism YDRO54C CDC34
ubiquitilnependent protein catabolism YBLO84C CDC27
mitotic chromosome segregation YLRO86W SMC4 U E
ubiquifin-deperdent prn cataiolism YDL132W CDC53
G2IM transition ofmitotic cell cycle YLL003W SFIlNE%
EU
establishment of cell polarity YLR229C CDC42
membrane transportYMR079W SEC14
Golgi to
IologicaLprocess unknown YDR373W FRO
cyoinesis YFLO05W SEC4U
9
blological-procs unknown YLR4 C
protein-ERteting YLR378C SEC61I
RET3U
retrograde transport, GlItoER YPLO1OW
YPR105C COG4
intra-olgi
0
YJL097W
biological .proes unkn
0
biologicalarcess unknown YGLO47W
0
APK Kcascade YGR198W
NAinked gycosylation YBR243C ALG7
Size
Predicted GO
Biological Process
Ubiqubin-dependent
protein catabolism
Chromosome
segregation
G2IM transition of
mitotic cellcycle
pathway
Protein amino acid
glycosylation
protein amino acid gyosylain YBR02C RER2I
biologicalprocess
YDL055C
acid g1 un otion YNL128
ammno
protein
is YJL091C
GPI anchor bios
attachment of GPI anchor to protein YLR459W
attachment of GPI anchor to protein YHR188C
ergosterol biosynthesis YNRO43W
ergosterol biosynthesis YGR 0W
PSAl
S
GWT1
CDC91
GPI16
MVD1
ERG25
Protein
hpidaUon
Sterol
metabolism
ation YOR168W GLN4
NAm
glutam
lutevNA aminoacylton YGRO94W VAS1
YGL246W GSNI
amh!noAcyffn
9RK4
guitt"M
hitRNA aminoacyltlon YPR033C HTS1
NAa--""-'yation YLR06QW FRS1
P~heryw~
ndna matrix protein 'nport YLR008C PAM18
m
WI
bb kof mcm unkown VOL026C
ocess unknown YNL310C FMP28
mitochondrial processing YLR163C MASI
btical roces unknown YKL033W FMP47
mitochond ma protein import YNL131W TOM22
mitochondrial matrix oteinknportVOR232W MGE1
Amino acid
activation
Protein-mitochondrial
targeting
i YOL077C BRX1
Rib. ge subunit assembly
0
biooglcalprocess unknown YNR054C
DNA dependent DNA replication YELOSC PO5
rRNA roc n YDR412W
0
A YGR103W NOP7
processing of 208
0
tr lion YHR122W
NA YGR09OW UTP22
processing of 20S
processing YDL031W DBP10 E
35S primary tran
rRprocessing YOR145C PNO1
rRNA processing YHR085W IPl1
060W TSR
riRNA
p
nibosome-nutuept YHRI97W RIXi
0I
unknown Y0L022C
biologicaL-process
processing of 20S pre A YPL126W NANi
PAYGR251W
pre
of
20S
processm
n40
processing of 20S pre-rRNA YOR1O4W
35S primarytusrtjoean YGR0O5C RRP46
presig f_ 05perNYMR128WEGM16
processing of20S pr-RA YBLO04W UTP20
rRNA moiicto YJR002W MPPIO
pre-rRNA YPRI44C
of 208 unknow
processing
UR82
YJR04fC NOC4
bifkfrocess
nuclear mRNA splicn, via spiceosome YER 12W LSM4
prcsIn YDR365C ESFI
ARNA
processing of 20pe"MRNA YORI19C RIO1
Rib. Ige subunit assembly and maint, YNL182C IP13
rlbosomne-nucleus export Y0R206W NOC2
establishment of cell polarity YKLO82C RRP4
protein monoubiquitnation YPR169W JIP5
ribosome-nucleus export YPL093W NOG1
Ribosome biogenesis
and assembly
Protein-nucleus
export
>1000 Cells/size bin
0
<-0.3
0
<-100
1
>0.3
Log(ratio) (individual genes)
>100 Log(P) (gene categories) (Positive for increased, negative for reduced)
Figure 3. Relationships among Gene Functions and Phenotypes Allows Systematic Prediction of the Functions of Uncharacterized Genes
Phenotypic data (morphology, size profiles, and microarray expression) are shown for mutants in functional classes that can be distinguished
by SVM in at least one data type. Mutants in bold at left are further characterized in Figures 5 and 6. The microarray measurements shown
were selected on the basis of Fisher's score (see Experimental Procedures). The labels at bottom indicate gene properties enriched among
the separate groups and representative genes. The rightmost panel of microarray data represents the behavior of functional classes of genes
where the values indicated are p values output by the Kolmogorov-Smimov test (Sachs, 1982), which estimates the likelihood (i.e., assigns a
p value) that the distribution of a subset of ranks is nonrandom. Not all mutants are shown for all categories.
207
Cell
36
transport specifically featured transcriptional induction
of genes involved in iron transport, which is consistent
with the idea that iron regulation is the essential mitochondrial function (Baker and Schatz, 1991; Lill and Kispal, 2000) (Figure 3). Genes specifically induced upon
perturbation of sterol metabolism included many members of the sterol metabolic pathway (Figure 3), as previously described (Daum et al., 1998; Hughes et al., 2000;
Vik and Rine, 2001), reconfirming what appears to be
negative feedback regulation of this essential pathway.
Negative Feedback Regulation of Trans-Acting
Ribosome Biogenesis Factors
To more systematically identify examples of feedback
transcriptional regulation, we examined whether the
overall induction or repression of genes in entire functional categories were characteristic of mutants in any
given category. The Supplemental Data contain the full
analysis. Selected results are shown at the right in Figure
3 and expose characteristic trends that are not evident
from looking at transcriptional responses of individual
genes. For example, the proteasome appears to be subject to negative feedback regulation at the level of transcription, consistent with results obtained previously
using chemical inhibition (Fleming et al., 2002). Another
prominent example is that the majority of mutants affecting ribosome biogenesis and assembly displayed a specific induction of genes encoding trans-acting factors
involved in ribosome biogenesis and assembly. This
suggests that, like many other cellular activities, this
process is controlled by a negative feedback loop that
regulates transcript levels. Although autoregulation of
the ribosome itself has been previously observed (Zhao
et al., 2003), to our knowledge feedback control of transcripts encoding the trans-acting factors has not been
previously noted. Figure 4 shows in detail the identities
of relevant mutants and individual induced genes in
these two categories.
In some cases, our data confirmed gene functions that
have been suspected from sequence features, physical
associations, and/or localization. For example, Hasi p
("helicase associated with Set1 ") has been found in
physical association with a variety of nucleolar rRNA
processing complexes (Gavin et al. 2002; Ho et al., 2002;
Fatica et al., 2002) and is localized to the nucleus, nuclear membrane, and/or nucleolus (Rout et al., 2000;
Kumar et al., 2002). In the TetO7-HAS1 mutant, we observed a striking loss of 18S rRNA and 20S pre-rRNA,
as well as accumulation of the 35S pre-rRNA and the
aberrant 23S species (A3-B1), which is indicative of a
defect in cleavage of the pre-rRNA at sites AO, A1 , or A2
(Fatica and Tollervey, 2002) (Figures 5A and 5B). Virtually
identical results were published recently in an independent study (Emery et al., 2004).
YKLO14C (URBI) and YJRO41C (URB2) Are
Required for Normal Ribosome Biogenesis
We also observed a reduction in levels of both 18S and
25S rRNAs in TetO7-promoter mutants of YKLO14C and
YJRO41C (Figure 5C). We confirmed by pulse-chase
analysis that there is a delay in rRNA synthesis in both
of these mutants (Figure 5D; TetO7-YJRO41C has a more
severe growth defect and displays a more prominent
delay). Both of the encoded proteins have been localized
to the nucleolus (Huh et al., 2003), and although they
have been detected in physical association with rRNA
processing proteins (Gavin et al., 2002; Ho et al., 2002),
they have also been reported to be associated with
proteins in other processes (Ho et al., 2002). Using a
high-stringency affinity purification procedure (Krogan
et al., 2004), we found that Ykl01 4cp copurified with
Yjr041cp almost exclusively (Figure 5E). Although their
precise role in ribosome biogenesis remains to be determined, these observations together indicate that these
two proteins share a close functional relationship and
are directly required for ribosome biogenesis and/or
rRNA processing. We have named the genes URB1
(YKLO14C) and URB2 (YJR041C) (URB = unhealthy ribosome biogenesis).
Verification of Inferred Gene Functions
As with any screening approach, directed experimentation (preferably coupled with independently obtained
supporting observations) is required to characterize
gene and protein functions with confidence. To show
explicitly that patterns and features in the data are predictors of new gene functions, we further pursued de
novo predictions made regarding the functions of uncharacterized genes in several different functional categories.
YGL245W (GSNI) Displays a tRNA Charging-like
Defect and Associates with tRNA-Glu In Vivo
Among the mutants that displayed transcriptional induction of genes involved in amino acid biosynthesis was
TetO7-YGL254W (Figure 3), suggesting that the product
of this gene is required for amino acid activation. In fact,
the Yg1254w protein sequence indicates that it is atRNAGlu synthetase, and the encoded protein has previously
been shown to physically associate with both methionyltRNA synthetase (Mesi p) and Arci p (Simos et al., 1996).
However, to our knowledge it has not been previously
shown that YGL245W functions as a tRNA synthetase
in vivo, and the gene has not been assigned a standard
name. To provide further support that Ygl254wp is a
bona fide tRNA-Glu synthetase in vivo, we epitopetagged and purified YgI254wp, phenol extracted any
associated RNAs, labeled them with fluors, and hybridized them to a microarray containing oligonucleotides
complementary to all known yeast noncoding RNAs
(Xing et al., 2004). This revealed that the most highly
HASI Is Required for 18S rRNA Biogenesis
In addition to the genes encoding the 78 ribosomal proteins, over 100 essential genes are already known to be
involved in ribosome biogenesis (Fatica and Tollervey,
2002). In our analyses, depletion of these factors was
not only characterized by the aforementioned transcriptional response, but also by generally smaller cell size
and absence of other morphological defects (Figure 3),
as has been previously described for nonessential genes
(Jorgensen et al., 2002). These phenotypes were shared
by TetO-promoter mutants in several poorly characterized genes, which we then examined by other assays.
208
Exploring Essential Gene Functions
37
-0.3
0
Log(ratio)
Function of
mutated gene
II
0.3
QM
YH11
y
Y
BC
16
45W
192
Y '7
642
YYYY
Y
Yoftj
Y
OY171YY
YORI
Y31
1
jYJ
7W
VYL114
YY
YQR1
I7A
0
YAQmll
YOR095
AWW
Y
11
YYY
2
R
1
P9
Y
YYiYI
Y
1
Y"YY
IYJYH
YM I
Y
Y
qnR=
0
s9c
C
1111111M
11111
1..
"'F-Tlt
%WMI-
in c-ta.ohm
-7
nP.
Figure 4. Negative Feedback Regulation of Ribosome and Proteasome Biogenesis
Induction or repression of individual genes (horizontal axis) in Individual mutants (vertical axis) are shown. The mutants are sorted such that
those that induce transcript levels of ribosome biogenesis factors (not ribosomal proteins) are shown at the top, those that induce transcript
levels of proteasome components most highly are shown at the bottom. The remainder are shown in the middle, sorted by the overall
proteasome induction level. The transcripts are sorted on their level of induction; i.e., the ribosome biogenesis transcripts that are induced
most highly are farthest to the left.
209
Cell
38
B
A
Primary transcipt
A
8
33S
323
20S
E
ITS1
5-ETS
2
3 TS
ITS2
I,
35S
18$
-
-
4,4
-
j
188
7Sf|-
27SA2
25
258
27SA
258
27SB,
- + ,
(18S)
25S
U2 1 -
4-
2S
Probes:
D-A2,
A2-A3,
U2
27SA2
(25S
205
25S WIe
CC)
'
18S PO
D
27SA2
(25S)
23S
-1
do
4
E
TAP tage
Tet07-YKLO14C Teto-YJR041C
35SI
Probes:
18S,
25S
Wr
(URBI)
(URB2)
0 2 515 0 2 515 0 2 5 15 min
4
27S
Yklp
Yjr41cp
-97kD
(Urb2p)-TAP
2..
20S1
66 kD
(185)
45 kD
U2
25S
18S
Nopip -+
F
tRNA-GluiE
tRNA-Gu2EE
tRNA-Gu3EUE
tRNA-Metl E
tRNA-Ser2U
tRNA-le1
tRNA-Lys3E
tRNA-Tyr2
tRNA-Leu4U tRNA-Asp1
e
Green = RNA from
Gsn1p-TAP
Red = total RNA
18SEEE
25SME
snR64WUE
Figure 5. Essential Genes Involved in RNA Metabolism
(A) Schematic representation of rRNA processing pathway.
(B) Northern analysis of rRNA and pre-rRNA in TetO-HAS1. The blot in the top panel was probed simultaneously with oligonucleotides
complementary to D-A2, A2-A3, and U2 snRNA (which serves as a loading control). The blot was then stripped and probed with oligonucleotides
complementary to the 1S and 25S rRNAs.
(C) Northern analysis of rRNA and pre-rRNA in TetO-promoter mutants of YKLO14C (URB1) and YJRO41C (URB2), as in (B).
(D) Pulse-chase analysis verifying delayed rRNA processing in TetO-URB1 and TetO7-URB2.
(E) TAP-tagged Urb2p primarily recovers Urbip.
(F) Ygr245w (Gsnl)-TAP associates with tRNA-Glu in vivo. A preparation of purified Gsn1 -TAP was phenol extracted, and the resulting RNA
was DNase I-treated, labeled with fluorescent dyes, and hybridized to a microarray containing probes complementary to tRNA, rRNA, and
other yeast noncoding RNAs. In the image shown, the Gsnl-TAP-purified RNA was labeled with Cy3 (green), yeast total RNA was labeled
with Cy5 (red), and the two were hybridized simultaneously.
YLR440C (SEC39) Is Required
for Protein Secretion
In our analysis, mutants in the secretory pathway were
characterized by a large, round morphology, often with
a wide bud neck and/or large size distribution, and a
enriched sequences (relative to total yeast RNA that was
hybridized in the other channel on the array) were tRNAGlu, followed by tRNA-Met (Figure 5F). We have named
the YGL245W gene GSN1 (GSN = tRNA glutamine synthetase).
210
Exploring Essential Gene Functions
39
microarray expression phenotype that is similar to (but
distinguishable from) mutants in glycosylation. Previously characterized genes that shared one or more of
these distinctive phenotypes included SEC4, SEC14,
SEC61, COG4 (SEC38), and RET3 (a vesicle coat component; Cosson et al., 1996). This group also included
the TetO 7-promoter mutant of YLR440C, a gene whose
product has been localized to the endoplasmic reticulum (Huh et al., 2003) and for which a two-hybrid interaction has been observed with Dsl p (Ito et al., 2001), a
protein required for Golgi-to-ER retrograde traffic (Andag et al., 2001; Reilly et al., 2001). We further characterized the phenotype of TetO7-YLR440C by pulse-chase
analysis, which confirmed a striking defect in processing
of Carboxypeptidase Y (CPY), a marker of protein secretion (Conibear and Stevens, 1998). On the basis of these
observations, we named the YLR440C gene SEC39.
Discussion
The analysis of mutant phenotypes is a singularly powerful means of defining gene functions. The "compendium" approach we previously described employed
microarray expression profiles as a one-size-fits-all phenotypic assay (Hughes et al., 2000). This approach was
the most fruitful for mutants that had a growth defect
since these consistently exhibited gene expression
changes relative to wild-type. This suggests that the
systematic analysis of conditional alleles of essential
genes should yield highly informative data. Here, we
show that the majority of a set of ~700 different TetO7 promoter strains display dramatic alterations in expression profiles. Furthermore, we show that similar "compendia" of other phenotypes provide complementary
perspectives on the mutant phenotype and can also be
used independently to generate functional inferences.
Although all four data types were informative (Figure
2 and Supplemental Data online), microarray expression
profiles appear to be able to discern the largest number
of distinct functional classes (Figure 3 and Supplemental
Data), perhaps due to the large number of independent
measurements made by microarrays; the yeast genome
encodes at least 100 different transcription factors, implying that microarrays can measure at least 100 different variables. The drug/media sensitivity profiles did not
correlate as tightly with distinct functional classes as
might have been expected on the basis of how extensively such tests have previously been used as screening tools (Supplemental Data). This may be due to the
fact that the robotic arraying system we used deposited
a large number of cells on the test plates, possibly leading to false negatives (data not shown). We did, however,
detect a number of expected sensitivities; for example,
our TetO7-ALG7 mutant was sensitive to tunicamycin,
consistent with previous observations (Barnes et al.,
1984; Giaever et al., 1999), and our TetO7-POL30 mutant
was sensitive to hydroxyurea, consistent with its role in
DNA replication (Pol30p is the yeast homolog of PCNA)
(Bauer and Burgers, 1990) (Supplemental Data). Figures
2C and 6B show that several TetO7-promoter mutants
in genes involved in mitochondrial function (TIM50,
TOM22, RIM2, MIM1) are compromised for growth on
glycerol.
YOLO26C (MIMI) Is Required
for Mitochondrial Import
Mutants in four genes known to be involved in mitochondrial import (MAS1, TOM22, MGE1, and PAM18) displayed similar phenotypic features to one another, most
notably induction of genes involved in iron homeostasis
(Figure 3). Mutants in three additional genes also displayed these features: FMP28, FMP47, and YOLO26C.
Fmp28p and Fmp47p have recently been localized to
mitochondria (Sickmann et al., 2003; FMP = found in
mitochondrial proteome). Aside from cytoplasmic localization (Kumar et al., 2002), there is currently no supporting data regarding properties of Yol026cp. A growth
defect of the TetO7-YOL026C strain on medium with
glycerol as the sole carbon source was dependent on
addition of doxycycline to the medium, supporting a
role in mitochondrial function (Figure 6B). Moreover,
Western blotting showed that the TetO7-YOL026C mutant accumulates uncleaved Atp2p precursor, a classical mitochondrial import defect shared by mutants in
other mitochondrial import proteins (Figure 6C). We have
named the YOLO26C gene MIM1 (for mitochondrial
import).
Synthetic Genetic Array Analysis
using TetO7-Promoter Alleles
The microarray expression data were particularly effective for discerning perturbation of biosynthetic pathways. This appears to be due at least in part to negative
feedback loops in which perturbation of a pathway results in transcriptional induction of the same pathway
or a related pathway. Previous analyses in which much
smaller numbers of experiments were examined (Birrell
et al., 2002; Giaever et al., 2002) suggested that there
is a poor correlation between the genes that are required
to survive a condition and the genes that are transcriptionally induced under the same condition. Our results
(Figures 3 and 4) suggest that there is often a meaningful
correspondence between the perturbed pathway and
the genes that are specifically induced, although a large
number of experiments may be required to determine
this specificity. In particular, we observed an apparent
negative feedback loop controlling trans-acting factors
involved in ribosome biogenesis. To our knowledge, this
As a final illustration of the utility of the TetO 7-promoter
strain collection for exploring essential gene functions,
we conducted synthetic genetic array (SGA) analysis
(Tong et al., 2001) by crossing a cdc40-1 strain to the
arrayed TetO7-promoter strains (Figure 2C). Consistent
with the known functions of CDC40 in both DNA replication and pre-mRNA splicing (Ben-Yehuda et al., 2000),
we identified synthetic genetic interactions with DNA
replication factors (PSF2, RFA2) and splicing factors
(PRP16, PRP24, PRP38) (Figure 7); in fact, the genetic
interaction we observed between TetO7-PRP16 and
cdc40-1 8 recapitulates a previous result (Ben-Yehuda
et al., 2000). Genetic interactions were obtained with
genes in other pathways, including protein targeting
(HSP60, KAP95) and membrane lipid biosynthesis
(GPl10, GWT1, TSC11), indicating that CDC40 function
is also linked to these processes.
211
Cell
40
A
P1
Chase time (min)
0 10
Wild Type
20 30
10
20 30 0
TetO7-YLR440c
(SEC39)
B
TetO7-YOLO26c (MIMI)
WT
TetO,-YOLO26c (MIMI)
YP Glycerol -dox
Qe
Q
s
YP Glycerol +dox
Figure 6. Mutants with Defects in Protein
Processing
(A) SDS-PAGE and autoradiography of antiCPY immunoprecipitates showing that TetOr
YLR440C (SEC39) accumulates CPY precursor. The ER precursor (P1) and mature
vacuolar (m)forms of CPY are indicated.
(B) Dilution spot assays, on agar plates containing glycerol as the sole carbon source,
showing that the TetO-YOLO26C (MIMI)
growth defect on glycerol is dependent upon
addition of doxycycline to the medium.
(C) Western blot showing that the TetOYOL026C (MIMI) strain accumulates Atp2p
precursor. TetO7-PSA1 is shown as a negative control (PSA1 is involved in cell wall function); TetO7-MASI and TetO7-PAM18 are positive controls.
wr
C
C)
A0)
'I)
0
Io
83 kD -
0
0
404-
V
ii-404
iii
P
m
62 kD a recent study described systematic analysis of 100
uncharacterized essential yeast genes using four such
assays (affinity purification and mass spectrometry,
two-hybrid, GFP localization, and PSI-BLAST) (Hazbun
et al., 2003). Among the 602 mutants we analyzed, 49
overlapped with the 77 for which Hazbun et al. (2003)
reported results. Among these, our inferences are in
agreement on twelve genes, which is a higher level of
concordance than would be expected by chance. However, we are in disagreement on fifteen genes, which
will require further study. There are also 19 genes in
which our SVM analysis produced predictions with expected precision of 30% or higher, but for which Hazbun
et al. made no functional assignments. These include
feedback loop has not been previously described; this
shows that microarray expression profiling can reveal
new regulatory pathways governing even the most fundamental cellular processes.
In many cases, our data are complementary to data
from other large-scale studies. The development and
implementation of techniques for systematic characterization of protein localizations and complexes have led
to the generation of some type of functional information
for a majority of yeast genes, including many that are
essential for viability (Rout et al., 2000; Uetz et al., 2000;
Ito et al., 2001; Gavin et al., 2002; Ho et al., 2002; Kumar
et al., 2002; Huh et al., 2003; Hazbun et al., 2003; Sickmann et al., 2003; Krogan et al., 2004). In particular,
YGLO47W
PRP38
Figure 7. Network
PRP24
PRP16
FRQ1
TSC 11
ISC1
CxS1
*mRNA
splicing
Cell cycle and DNA replication
Protein targeti ng
Membrane lipid biosynthesis
*Regulation
of signal transduction
gUnclassified
KAP95 HSP60
212
Diagram
Summarizing
Synthetic Genetic Interactions between
cdc4O-1' and Strains in the TetOr-Prornoter
Array
Exploring Essential Gene Functions
41
Size Analysis
Cultures were grown in YPD t 10 sg/ml doxycycline at 30*C to 0.6
to 3 x 10 cells/mI, a density range in which wild-type cell size
distributions do not vary. Cells were sonicated as above and size
distributions were obtained with a Coulter Channelizer Z2 (Beckman-Coulter) as described previously (Jorgensen et al., 2002).
Batches of up to 84 mutants at a time were analyzed, with multiple
wild-type cultures interspersed among the mutants. Batches were
normalized by translating the data across the 256 size bins and
adjusting the spread by rescaling so that each batch had an identical
overall distribution. The counts for each profile were then renormalized to an identical sum. Empty leading and trailing bins were filled
in by linear extrapolation. We confirmed that this transform brought
the wild-types into register.
correct predictions for YHRO40W (BCD1), YHRO85W
(IP/1), YHR197W (RIX1), YNL182C (1P13), and YOR004W;
although these proteins were still unannotated at the
time of our analysis, we have recently shown in Independent work that all of them are involved in ribosome
biogenesis (Wu et al., 2002; Peng et al., 2003; Krogan
et al., 2004). Similarly, among the six genes whose functions we analyzed here using specific assays, only four
(HAS1, URB1, GSN1, and SEC39) could have been predicted with confidence (i.e., on the basis of agreement
among more than one data set) from any of the functional
genomics or proteomics data currently in the literature.
For URB2 and MIM1, previous evidence was either conflicting or insufficiently informative.
The strains described here are derived from the deletions consortium wild-type MATa strain (BY4741) (Giaever et al., 2002) and consequently are compatible with
any assay developed for the deletion strains that does
not require the strains to be ura3 (since URA3 marks
the tet activator). For example, the TetO7 strains are
compatible with synthetic genetic array analysis (Tong
et al., 2001, 2004) (Figure 7), and this may provide a
means for mapping large-scale interaction networks
among essential genes. We plan to ultimately generate
TetO7 strains encompassing all essential yeast genes,
and we have already created an additional 100 strains,
bringing the current collection to over 700 strains. In
addition, an effort is also underway to integrate molecular barcodes into these strains in order to facilitate genetic analysis in pools using a microarray readout (Giaever et al., 2002), as is already possible with the
deletions consortium strains (C. Nislow and G. Giaever,
personal communication).
Drug Sensitivity Analysis and Synthetic
Genetic Array Analysis
Strains were transferred onto agar plates at a density of 768 colonies
per plate (384 strains in duplicate, as in Figure 2C) using a VersArray
Colony Arrayer (BioRad) (Tong et al., 2001). Drug concentrations
were hydroxyurea, 100 mM; benomyl, 15 sg/ml; tunicamycin,
1 Ig/ml; rapamycin, 50 ng/ml; 6-azauracil, 100 sg/ml; cycloheximide, 0.1 pg/ml; caffeine, 8 mM; 5-fluorouracil, 20 mM, 40 mM, 80
mM; glucosamine, 3%; sorbitol, 1.5 M; all with and without doxycycline, 10 sg/ml. The strains were also arrayed on YPglycerol (3%
glycerol) and YPgalactose media. Drug sensitivity screens were performed in duplicate; SGA analysis was performed in triplicate. Sensitivities were scored using an automated system that measures the
colony area from digital images of the plates and reports the probability (p value) that the colony area is different between reference
and treatment plates (C.B. and H. Ding, unpublished data). For SGA
analysis, the CDC40 gene was replaced with a natR-marked cdc4O1" allele in the haplold MATa strain Y3656 (Tong et al., 2004) to
create strain Y5065. Y5056 was crossed to the TetO7-promoter array
following Tong et al. (2001) with modifications. The Supplemental
Data contain a full protocol. Interactions were confirmed by random
spore analysis following Tong et al. (2004). Supplemental Figure S1
shows random spore analyses for each of the interactions shown
in Figure 7.
Microarray Analysis
Experimental Procedures
Strain Construction
TetO7-promoter alleles were constructed in strain R1 158 (Hughes
et al., 2000), which expresses the tot "off" activator (tTAJ, by replacing the 100 bases upstream of the start codon with a cassette
(kanR-TetO7-TATAcyc1) from plasmid RP188 via one-step integration.
Colony-purified G418-resistant transformants were frozen as glycerol stocks and confirmed by PCR. The Supplemental Data on the
Cell website includes a table with all of the primer sequences, as
well as a table of strains. The full collection is available from Open
Biosystems (Huntsville, Alabama).
Growth Assays
Growth rates on solid medium (YPD) were scored by manual inspection of spotted dilution series in the presence and absence of 10
pLg/ml doxycyline. Growth rates in liquid (Figure 1D) were calculated
from the cell densities (at 18 and 24 hr after the addition of doxycycline) of cultures used for microarray expression profiling.
Morphological Characterization
Strains were grown 12-16 hr to mid-log phase at 30*C in YPD with
and without 10 ±g/ml doxycycline, fixed by addition of formaldehyde
(3.7%), and sonicated gently for 8 s to disperse aggregated cells.
At least 100 cells of each strain were examined using phase-contrast
microscopy. Mutants were scored for 16 categories: an increase of
unbudded, small-budded, or large-budded cells; small, large, or
variable size; round, pointed, elongated, or amorphous shape; hyperpolarized buds; multiple buds; wide bud neck; lysis and clumped.
Elongated and amorphous categories were subclassified as slight,
moderate, or severe. Mutants that did not have a phenotype were
classified as wild-type.
213
Mutant cultures were grown in SC + 2% glucose + 10 sg/ml doxycycline for 24 hr, harvested, and processed in parallel with corresponding wild-type control cultures (Ri 158) as described (Hughes et al.,
2000). Samples were hybridized to spotted arrays containing 70mer oligonucleotides complimentary to 6,307 yeast genes (Operon),
with spotting and hybridizations following Hegde et al., 2000. Arrays
were scanned, images were quantified, and physical artifacts (dust
and salt residue) edited as described previously (Hughes et al., 2000)
and normalized by loess smoothing (Yang et al., 2002). Spatial trends
were removed by applying a high-pass filter with a Gaussian window
in the frequency domain (FFT transformed slides with missing values
set to the slide mean). Remaining spatial trends were manually
flagged. Replicate slides (dye-swaps) were then compared, and
spots with poor correlation were manually flagged on both arrays.
Dye-swaps were combined by averaging the ratio of the two measurements. The data were assembled into a 215 x 6,307 data matrix
and missing values (9.8%) were imputed using BPCAfill (Oba et
al., 2003). The final data are in the Supplemental Data online and
individual microarray experiments have been submitted to the GEO.
Northern Blotting and Pulse-Chase Analysis
RNA extraction and Northem blotting were performed as described
(Peng et al., 2003). For pulse-chase analysis, strains were transformed with pRS411, grown to 0.8 x 101 cells/mI in SD-Met+10
tg/ml doxycycline, and labeled following Kressler et al. (1999). Aliquots (20,000 cpm) were run on a 1% agarose-glyoxal gel, transferred to Hybond-N+ (Amersham) by overnight downward capillary
transfer, dried, and visualized by autoradiography.
TAP Tagging and Mass Spectrometry
Four liters of TAP strains for URB1, URB2, and GSN1 (Ghaemmaghami et al., 2003) were grown in YPD to 1.5 x 101 cells/mi. Proteins
Cell
42
were extracted with high-speed clarification, TAP-purified, separated by 10% SDS-PAGE gels, visualized by silver staining, and
identified by MALDI-TOF mass spectrometry as previously described (Krogan et al., 2004).
Secretion Assays
Strains were grown to the exponential phase in selective minimal
medium (with doxycycline) at 24*C, suspended at OD600 = 5 in
selective minimal medium without methionine, pulse-labeled with
35S-methionine for 7 min, and chased for 0-30 min as indicated.
The pulse and chase were done at room temperature. Extracts prepared by lysis in SDS were diluted into Triton X-100 and immunoprecipitated with anti-CPY.
Mitochondrial Import Assay
ATP2 was TAP-tagged in each of the strains shown in Figure 6C by
transformation with C-terminal ATP2 PCR products from an ATP2TAP strain (Ghaemmaghami et al., 2003). Strains were grown in SCHis+10 sg/ml doxycycline for 24 hr, extracted, and precipitated by
vortexing in 20% TCA in the presence of glass beads and subjected
to Western blotting with rabbit IgG to recognize the protein A component of the TAP tag.
Statistical Analysis
For SVM, we used Gist (http://svm.sdsc.edu) version 2.0.5 for Unux
with default parameters. Precision and recall were established by
3-fold cross validation. GO annotations were downloaded from Saccharomyces Genome Database (Issel-Tarver et al., 2002) and all
categories were "propagated up" the GO graph. Morphology and
drug sensitivity data were input into SVM verbatim. To avoid overfitting, the size profile for each mutant used by the SVM was a sixdimensional vector consisting of the projections of the original 256dimensional profile on to the first six principal components of the
covariance matrix of the normalized size profiles. The expression
profile for each mutant used by the SVM was the projection of the
initial profile onto the first 20 principal components of the expression
covariance matrix. A separate SVM for each mutant utilized a 65dimensional vector composed of 65 groups of coregulated genes
identified manually from clustering diagrams.
Fisher's score (Bishop, 1995) was computed for all features in
each data type for all GO-BP categories. Matrices of all Fisher's
scores are given in the Supplemental Data. Microarray data shown
in Figure 3 scored above the 99* percentile for at least one of the
indicated categories (corresponding to a Fisher's score of 1.8).
The Kolmogorov-Smimov test and hypergeometric p value calculations used tools at http://kstest.med.utoronto.ca/ and http://
funspec.med.utoronto.ca/ (Robinson et al., 2002).
Acknowledgments
We thank HuiMing Ding, Amy Tong, Renee Brost, Diana Ho, Nina
Enriques, Darlene Ellenor, Mariana Kekis, Natalie Gabovic, Melissa
Ballantine, and Jialin Wu for technical contributions; Dawn Richards
and Victoria Canadien for assistance with mass spectrometry; Michael Yaffe, Howard Bussey, Roberto J. Rodriguez-Suarez, and
Amis Kuksis for advice and assistance in analyzing specific strains;
and Grant Brown for evaluation of the manuscript. This work was
supported by grants from the CIHR and Genome Canada to T.R.H.
and B.J.A. A.D. was supported by a Charles H. Best fellowship and
J.H. was supported by an Estate of Betty Irene West/CIHR doctoral
research award.
Received: April 3, 2004
Revised: May 13, 2004
Accepted: May 17, 2004
Published: July 8, 2004
References
Andag, U., Neumann, T., and Schmitt, H.D. (2001). The coatomerinteracting protein Dsllp is required for Golgi-to-endoplasmic reticulum retrieval in yeast. J. Biol. Chem. 276, 39150-39160.
Ashbumer, M., Ball, C.A., Blake, JA., Botstein, D., Butler, H., Cherry,
214
J.M., Davis, A.P., Dolinski, K., Dwight, S.S., Eppig, J.T., et al. (2000).
Gene ontology: tool forthe unification of biology. The Gene Ontology
Consortium. Nat. Genet. 25, 25-29.
Baker, K.P., and Schatz, G. (1991). Mitochondrial proteins essential
for viability mediate protein import into yeast mitochondria. Nature
349, 205-208.
Barnes, G., Hansen, W.J., Holcomb, C.L., and Rine, J. (1984). Asparagine-linked glycosylation in Saccharomyces cerevisiae: genetic
analysis of an early step. Mol. Cell. Biol. 4, 2381-2388.
Bauer, G.A., and Burgers, P.M. (1990). Molecular cloning, structure
and expression of the yeast proliferating cell nuclear antigen gene.
Nucleic Acids Res. 18, 261-265.
Ben-Yehuda, S., Dix, I., Russell, C.S., McGarvey, M., Beggs, J.D.,
and Kupiec, M. (2000). Genetic and physical interactions between
factors involved in both cell cycle progression and pre-mRNA splicing in Saccharomyces cerevisiae. Genetics 156, 1503-1517.
Birrell, G.W., Brown, J.A., Wu, H.I., Giaever, G., Chu, A.M., Davis,
R.W., and Brown, J.M. (2002). Transcriptional response of Saccharomyces cerevisiae to DNA-damaging agents does not identify the
genes that protect against these agents. Proc. NatI. Acad. Sci. USA
99, 8778-8783.
Bishop, C.M. (1995). Neural Networks for Pattern Recognition (New
York: Oxford University Press).
Brown, M.P., Grundy, W.N., Un, D., Cristianini, N., Sugnet, C.W.,
Furey, T.S., Ares, M., Jr., and Haussler, D. (2000). Knowledge-based
analysis of microarray gene expression data by using support vector
machines. Proc. Nati. Acad. Sci. USA 97, 262-267.
Chun, K.T., and Goebl, M.G. (1997). Mutational analysis of Cakip,
an essential protein kinase that regulates cell cycle progression.
Mol. Gen. Genet. 256, 365-375.
Cliften, P., Sudarsanam, P., Desikan, A., Fulton, L., Fulton, B., Majors, J., Waterston, R., Cohen, B.A., and Johnston, M. (2003). Finding
functional features in Saccharomyces genomes by phylogenetic
footprinting. Science 301, 71-76.
Conibear, E., and Stevens, T.H. (1998). Multiple sorting pathways
between the late Golgi and the vacuole in yeast. Biochim. Biophys.
Acta 1404, 211-230.
Cosson, P., Demolliere, C., Hennecke, S., Duden, R., and Letoumeur,
F.(1996). Delta- and zeta-COP, two coatomer subunits homologous
to clathrin-associated proteins, are involved in ER retrieval. EMBO
J. 15, 1792-1798.
Daum, G., Lees, N.D., Bard, M., and Dickson, R. (1998). Biochemistry, cell biology and molecular biology of lipids of Saccharomyces
cerevisiae. Yeast 14, 1471-1510.
Dohmen, R.J., Wu, P., and Varshavsky, A. (1994). Heat-inducible
degron: a method for constructing temperature-sensitive mutants.
Science 263,1273-1276.
Eisen, M.B., Spellman, P.T., Brown, P.O., and Botstein, D. (1998).
Cluster analysis and display of genome-wide expression patterns.
Proc. NatI. Acad. Sci. USA 95, 14863-14868.
Emery, B., de la Cruz, J., Rocak, S., Deloche, 0., and Under, P.
(2004). Has1p, a member of the DEAD-box family, is required for
40S ribosomal subunit biogenesis in Saccharomyces cerevisiae.
Mol. Microbiol. 52, 141-158.
Fatica, A., and Tollervey, D. (2002). Making ribosomes. Curr. Opin.
Cell Biol. 14, 313-318.
Fatica, A., Cronshaw, A.D., Dlakic, M., and Tollervey, D.(2002). Ssf1p
prevents premature processing of an early pre-60S ribosomal particle. Mol. Cell 9, 341-351.
Fleming, J.A., Ughtcap, E.S., Sadis, S., Thoroddsen, V., Bulawa,
C.E., and Blackman, R.K. (2002). Complementary whole-genome
technologies reveal the cellular response to proteasome inhibition
by PS-341. Proc. NatI. Acad. Sci. USA 99,1461-1466.
Gari, E., Piedrafita, L., Aldea, M., and Herrero, E. (1997). A set of
vectors with a tetracycline-regulatable promoter system for modulated gene expression in Saccharomyces cerevisiae. Yeast 13,
837-848.
Gasch, A.P., and Eisen, M.B. (2002). Exploring the conditional co-
Exploring Essential Gene Functions
43
regulation of yeast gene expression through fuzzy k-means clustering. Genome Biol. 3, RESEARCH0059.
Gavin, A.C., Bosche, M., Krause, R., Grandi, P., Marzioch, M., Bauer,
A., Schultz, J., Rick, J.M., Michon, A.M., Cruciat, C.M., et al. (2002).
Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 415,141-147.
Ghaemmaghami, S., Huh, W.K., Bower, K., Howson, R.W., Belle,
A., Dephoure, N., O'Shea, E.K., and Weissman, J.S. (2003). Global
analysis of protein expression in yeast. Nature 425, 737-741.
Giaever, G., Shoemaker, D.D., Jones, T.W., Liang, H., Winzeler, E.A.,
Astromoff, A., and Davis, R.W. (1999). Genomic profiling of drug
sensitivities via induced haploinsufficiency. Nat. Genet. 21,278-283.
Giaever, G., Chu, A.M., Ni, L., Connelly, C., Riles, L., Veronneau, S.,
Dow, S., Lucau-Danila, A., Anderson, K., Andre, B., et al. (2002).
Functional profiling of the Saccharomyces cerevisiae genome. Nature 418, 387-391.
Giaever, G., Flaherty, P., Kumm, J., Proctor, M., Nislow, C., Jaramillo,
D.F., Chu, A.M., Jordan, M.I., Arkin, A.P., and Davis, R.W. (2004).
Chemogenomic profiling: identifying the functional interactions of
small molecules in yeast. Proc. Nati. Acad. Sci. USA 101, 793-798.
Hartwell, L.H., Culotti, J., and Reid, B. (1970). Genetic control of the
cell-division cycle in yeast. I.Detection of mutants. Proc. Natl. Acad.
Sci. USA 66, 352-359.
Hazbun, T.R., Malmstrom, L., Anderson, S., Graczyk, B.J., Fox, B.,
Riffle, M., Sundin, B.A., Aranda, J.D., McDonald, W.H., Chiu, C.H.,
et al. (2003). Assigning function to yeast proteins by integration of
technologies. Mol. Cell 12,1353-1365.
Hegde, P., Qi, R., Abemathy, K., Gay, C., Dharap, S., Gaspard, R.,
Hughes, J.E., Snesrud, E., Lee, N., and Quackenbush, J. (2000). A
concise guide to cDNA microarray analysis. Biotechniques 29,
548-556.
Ho, Y., Gruhier, A., Heilbut, A., Bader, G.D., Moore, L., Adams, S.L.,
Millar, A., Taylor, P., Bennett, K., Boutilier, K., et al. (2002). Systematic identification of protein complexes in Saccharomyces cerevisiae by mass spectrometry. Nature 415, 180-183.
Hughes, T.R. (2002). Yeast and drug discovery. Funct. Integr. Genomics 2, 199-211.
Hughes, T.R., Marton, M.J., Jones, A.R., Roberts, C.J., Stoughton,
R., Armour, C.D., Bennett, H.A., Coffey, E., Dai, H., He, Y.D., et
al. (2000). Functional discovery via a compendium of expression
profiles. Cell 102, 109-126.
Huh, W.K., Falvo, J.V., Gerke, L.C., Carroll, A.S., Howson, R.W.,
Weissman, J.S., and O'Shea, E.K. (2003). Global analysis of protein
localization in budding yeast. Nature 425, 686-691.
Issel-Tarver, L., Christie, K.R., Dolinski, K., Andrada, R., Balakrishnan, R., Ball, C.A., Binkley, G., Dong, S., Dwight, S.S., Fisk, D.G., et
al. (2002). Saccharomyces Genome Database. Methods Enzymol.
350, 329-346.
Ito, T., Chiba, T., Ozawa, R., Yoshida, M., Hattori, M., and Sakaki,
Y. (2001). A comprehensive two-hybrid analysis to explore the yeast
protein interactome. Proc. NatI. Acad. Sci. USA 98, 4569-4574.
Jorgensen, P., Nishikawa, J.L., Breitkreutz, B.J., and Tyers, M.
(2002). Systematic identification of pathways that couple cell growth
and division in yeast. Science 297, 395-400.
Kanemaki, M., Sanchez-Diaz, A., Gambus, A., and Labib, K. (2003).
Functional proteomic identification of DNA replication proteins by
induced proteolysis in vivo. Nature 423, 720-724.
Kellis, M., Patterson, N., Endrizzi, M., Birren, B., and Lander, E.S.
(2003). Sequencing and comparison of yeast species to identify
genes and regulatory elements. Nature 423, 241-254.
Kressler, D., Doere, M., Rojo, M., and Under, P. (1999). Synthetic
lethality with conditional dbp6 alleles identifies rsal p, a nucleoplasmic protein involved in the assembly of 60S ribosomal subunits.
Mol. Cell. Biol. 19, 8633-8645.
Krogan, N.J., Peng, W.T., Cagney, G., Robinson, M.D., Haw, R.,
Zhong, G., Guo, X., Zhang, X., Canadien, V., Richards, D.P., et al.
(2004). High-definition macromolecular composition of yeast RNAprocessing complexes. Mol. Cell 13, 225-239.
Kumar, A., Agarwal, S., Heyman, J.A., Matson, S., Heidtman, M.,
215
Piccirillo, S., Umansky, L, Drawid, A., Jansen, R., Uu, Y., et al.
(2002). Subcellular localization of the yeast proteome. Genes Dev.
16, 707-719.
Lill, R., and Kispal, G. (2000). Maturation of cellular Fe-S proteins:
an essential function of mitochondria. Trends Biochem. Sci. 25,
352-356.
Lum, P.Y., Armour, C.D., Stepaniants, S.B., Cavet, G., Wolf, M.K.,
Butler, J.S., Hinshaw, J.C., Gamier, P., Prestwich, G.D., Leonardson,
A., et al. (2004). Discovering modes of action for therapeutic compounds using a genome-wide screen of yeast heterozygotes. Cell
116, 121-137.
Oba, S., Sato, M., Takemasa, I., Monden, M., Matsubara, K., and
Ishii, S. (2003). A Bayesian missing value estimation method. Bioinformatics 19, 2088-2096.
Peng, W.T., Robinson, M.D., Mnaimneh, S., Krogan, N.J., Cagney,
G., Morris, Q., Davierwala, A.P., Grigull, J., Yang, X., Zhang, W., et
al. (2003). A panoramic view of yeast noncoding RNA processing.
Cell 113, 919-933.
Reilly, B.A., Kraynack, B.A., VanRheenen, S.M., and Waters, M.G.
(2001). Golgi-to-endoplasmic reticulum (ER) retrograde traffic in
yeast requires Ds1 p, a component of the ER target site that interacts
with a COPI coat subunit. Mol. Biol. Cell 12, 3783-3796.
Robinson, M.D., Grigull, J., Mohammad, N., and Hughes, T.R. (2002).
FunSpec: a web-based cluster interpreter for yeast. BMC Bioinformatics 3, 35.
Rout, M.P., Aitchison, J.D., Suprapto, A., Hjertaas, K., Zhao, Y., and
Chait, B.T. (2000). The yeast nuclear pore complex: composition,
architecture, and transport mechanism. J. Cell Biol. 148, 635-651.
Sachs, L. (1982). Applied Statistics (Berlin: Springer-Veriag).
Sickmann, A., Reinders, J., Wagner, Y., Joppich, C., Zahedi, R.,
Meyer, H.E., Schonfisch, B., Perschil, I., Chacinska, A., Guiard, B.,
et al. (2003). The proteome of Saccharomyces cerevisiae mitochondria. Proc. NatI. Acad. Sci. USA 100, 13207-13212.
Simos, G., Segref, A., Fasiolo, F., Hellmuth, K., Shevchenko, A.,
Mann, M., and Hurt, E.C. (1996). The yeast protein Arcip binds to
tRNA and functions as a cofactor for the methionyl- and glutamyltRNA synthetases. EMBO J. 15, 5437-5448.
Tong, A.H., Evangelista, M., Parsons, A.B., Xu, H., Bader, G.D., Page,
N., Robinson, M., Raghibizadeh, S., Hogue, C.W., Bussey, H., et al.
(2001). Systematic genetic analysis with ordered arrays of yeast
deletion mutants. Science 294, 2364-2368.
Tong, A.H., Lesage, G., Bader, G.D., Ding, H., Xu, H., Xin, X., Young,
J., Berriz, G.F., Brost, R.L., Chang, M., et al. (2004). Global mapping
of the yeast genetic interaction network. Science 303, 808-813.
Uetz, P., Giot, L., Cagney, G., Mansfield, T.A., Judson, R.S., Knight,
J.R., Lockshon, D., Narayan, V., Srinivasan, M., Pochart, P., et al.
(2000). A comprehensive analysis of protein-protein interactions in
Saccharomyces cerevisiae. Nature 403, 623-627.
Vazquez de Aldana, C.R., Wek, R.C., Segundo, P.S., Truesdell, A.G.,
and Hinnebusch, A.G. (1994). Multicopy tRNA genes functionally
suppress mutations in yeast elF-2 alpha kinase GCN2: evidence for
separate pathways coupling GCN4 expression to unchanged tRNA.
Mol. Cell. Biol. 14, 7920-7932.
Vik, A., and Rine, J. (2001). Upc2p and Ecm22p, dual regulators of
sterol biosynthesis in Saccharomyces cerevisiae. Mol. Cell. Biol.
21, 6395-6405.
Wu, L.F., Hughes, T.R., Davierwala, A.P., Robinson, M.D., Stoughton,
R., and Altschuler, S.J. (2002). Large-scale prediction of Saccharomyces cerevisiae gene function using ovelapping transcriptional
clusters. Nat. Genet. 31, 255-265.
Xing, F., Hiley, S.L., Hughes, T.R., and Phizicky, E.M. (2004). The
specificities of four yeast dihydrouridine synthases for cytoplasmic
tRNAs. J. Biol. Chem. 279,17850-17860. Published online February
16, 2004. 10.1074fjbc.M401221200.
Cell
44
Yang, Y.H., Dudoit, S., Luu, P., Lin, D.M., Peng, V., Ngai, J., and
Speed, T.P. (2002). Normalization for cDNA microarray data: a robust
composite method addressing single and multiple slide systematic
variation. Nucleic Acids Res. 30, el 5.
Zhang, N., Osborn, M., Gitsham, P., Yen, K., Miller, J.R., and Oliver,
S.G. (2003). Using yeast to place human genes in functional categories. Gene 303, 121-129.
Zhao, Y., Sohn, J.H., and Warner, J.R. (2003). Autoregulation in the
biosynthesis of ribosomes. Mol. Cell. Biol. 23, 699-707.
Accession Numbers
Individual microarray experiments have been submitted to the Gene
Expression Omnibus with accession number GSE1 404.
216