-----TU --:-* -- * ------*.- -- RECONSTITUTION OF EPSTEIN-BARR NUCLEAR ANMGEN 1 (EBNA1)- MEDIATED PLASMID SEGREGATION IN BUDDING YEAST REQUIRES EUMAN EBP2
Priya Kapoor
A thesis subrnitted in codomiity with the requirements for the de- of Master of Science Graduate Department of Molecular and Medical Genetics University of Toronto
@ Copyright by Priya Kapoor, 2001 The author has gfaated a non- L'auteur a accordé une licence non exclusive licence dowing the exclusive permettant (i la National Li- of Canada to Biblioth&que nationale du Canada de reproduce, loan, distribute or sell reproduire, prêter, distribuer ou copies of this thesis in microfonn, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/film, de reproduction sur papier ou sur format blectronique.
The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neitha the droit d'auteur qui protège cette thèse. thesis nor substantial extracts &om it Ni la thèse ni des extraits substantiels may be printed or othefwjse de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son Reconstitution of Epstein-Barr Nuclear Antigen 1 (EBNA1)-Mediateâ Plasrnid
-Lu- -- A Segregation In Budding Yeast Rquires Human EBP2
by Priya Kapoor
Masten of Science, Graduate Department of Molecular and Medical Geneticq
University of Toronto, 2001
ABSTRACT
Epstein-Barr nuclear antigm 1 (EBNAl ) govems the stable segregation of Epstein-
Barr virus (EBV)episomes in latently Ulfected B-lymphocytes by mediating their
attachment to the host metaphase chromosomes. The EBNAl residues that mediate
chromosome attachrnent and DNA segregation activity bind to the human EBP2 (hEBP2)
protein, which is a component of the cellular metaphase chromosomes. To investigate
the importance of hEBP2 for EBNAl-mediateà DNA partitioning, 1 attempted to
reconstitute the EBV-based segregation system in cerevisiue. 1 found that stable
segregation of a yeast repücating plasmid containing the EBV segregation element, FR,
required both EBNAl and hEBP2. An EBNAl mutant that cannot bind hEBP2 failed to
support segregation of the FR plasmid, indicating that an EBNAI-hEBP2 interaction is
necessary. My nsults pmvide dinct eviâence that hEBP2 is required for EBNAI-
mediated segregation and dernonstrate that EBV-based segregation can occur in budding yeast . - A -
First and fomnost, 1want to thank my supervisor, Dr. Lori Frappier. 1 thaak her for
her encouragement, support and brilliant ideas. Moreover, 1thank her for being patient
with me and for guidiag me through the tough times. 1also want to thank the wonderful
members of the Frappier laboratory, both past and present, which includes Dr. Vicki
Athanasopoulos, Tina Avolio, Dr. Derek Ceccarelli, Jennifer Cruickshank, Melissa
Holowaty, Kathy Shire and Dr. Hong Wu. Words cannot express what each of them
mean to me. 1 thank them al1 for their support and niendship.
1 also want to thank the members of my supervisory comrnittee, Dr. B. Andrews and
Dr. B. Funnell, for their ideas and encouragement. 1 thank the Andrews laboratory for
providing me with many of the yeast strains and plasmids and for helphil advice.
From the bottom of my heart, 1 want to thank my parents and my brother, the people who mean the world to me. 1thank them for motivating me, loving me and understanding me. Last but ceriainly not the least, 1 thank my best fnend Rahul. Having him by my side has encouraged me to work harder. 1 thank him for pushing me to strive for the best and for always being by my side. TABLE OF CONTENTS
L-
ABSTRACT ii ACKNOWLEDGEMENTS iii TABLE OF CONTENTS IV LIST OF FIGURES vii LIST OF TABLES viii LIST OF ABBREVIATIONS ix
1, INTRODUCTION
1.1. Epstein-Barr Vh:Mode of Infection and Associated Diseases 1.2. Viral Protein Expression During EBV Latent Infection 1.3. EBV Latent Origin of Replication, oriP 1.4. Epstein-Barr Nuclear Antigen 1 (EBNAl ) I.4.a. Protein Domains and Structure of EBNA 1 I.4.b. EBNAl Functions 1.4.b.i. Transcriptional Activities of EBNAl I.4.b.ii. Replication Function of EBNAl 1.4.b.iii. Segregation Function of EBNAl L S. EBNA 1-hEBP2 (human EBNA 1-Binding Protein 2) Interaction 1.5.a. Roperties and Cellular Function of EBPZ 1.5.b. Significance of EBNAl -hEBP2 Interaction in EBV Segregation 1.6. Segregation Mechanism of Extrachromosomal DNA Molecules 1.6.a Papillomavirus 1.6.b. Kaposi Sarcoma-Associateci Herpesvirus 1.6,~.Double Minute Chromosomes I.6.d. Yeast 2-Micron Plasmids I.6.e. Yeast CEN Plasmids 1.7. Thesis Rationale
II. MATERIALS AND METHODS
II.1. Yeast Strains II.2. Construction of the Plasmid Loss Assay Constnicts II.3. Expression of EBNAl and hEBP2 in Buddhg Yeast U.4. Construction of the Yeast Two-Hybrid Plasmids 11.5. Construction of hEBP2-Integrated Yeast Strains 11.6. Plasmid Loss Assays n.7. Yeast Two-Hybnd Assays m.1. Expression of EBNAl and hEBP2 in Budding Yeast III.2. EBNAl Dots Not Mcdiate FR Plasmid Segregation in Yeast iII.3. Human EBP2 is RequKed for EBNAl-Mediateci Segregation in Yeast iII.4. EBNA 1-Mediated Segregation in Yeast Requues EBNA 1 Binding to hEBP2 IJI.5. EBNA1 Does Not Interact with the Yeast EBPZ Protein III.6. FR Plasmid Segregation in hEBP2-Integrated Yeast Sûains by EBNAl
IV. DISCUSSION
IV. 1. Cornparison of EBV-Bssed Segregation in Budding Yeast and Humans IV.2. hEBP2 Plays an Essential Role in EBNAl-Meâiated Segregation N.3.nie Function of hEBP2 in EBV Segregation Cannot be ProMded by yEBP2 N.4. A Role for hEBP2 in the Segregation of BPV and KSHV Episomes?
V. FUTURE DIRECTIONS
V. 1. Cellular Localization Studies of EBP2 and EBNAl in Budding Yeast V.2. Assessing hEBP2 Residues Involved in EBNAl- Mediated Segregation in Yeast V.2.a. Mapping hEBP2 Residues that Interact with EBNAl V.2.b. Role of the EBNAl-Binding Domain of hEBP2 in EBNA 1-Mediatecl Segregation V.2.c. Identifjing the Chromosome-Associathg Region of hEBP2 in Budding Yeast V.3. Identifjing hEBP2 Residues Required for Chromosome Association in Humans V.4. Using Yeast and Human EBP2 Hybnds to Assess the Inability of Yeast EBPZ to Support Segregation by EBNAl V.S. hEBP2-Mediated Segregation of a Plasmid Containhg GAIA Binding Sites in the Presence of the GALA DNA-Binding Domain V.6. Use of a Sectoring Assay to Messure Plasmid Stability in the EBV-Based Segregation System in Yeast VI. 1. Stability of Segregation Plasmids in the Presence and Absence Of EBNAl and hEBP2 - Results from Individual Experiments 85
VIT. REFERENCES 86 LIST OF FIGURES
Figure 1 O*, the EBV latent origin of replication
Figure 2 Schematic diagram of the functional regions of EBNAl
Figure 3 Aiignrnent of EBP2 homologs
Figure 4 Expression plasmids useci in the plasmid loss assay
Figure 5 EBNAl and hEBP2 expression in budding yeast
Figure 6 Segregation test plasmids
Figure 7 EBNAl cannot support stable segregation of YRp7FR in budding yeast
Figure 8 hEBP2 is required for EBNAl-mediated FR-plasmid segregation in budding yeast
Figure 9 EBNAA325-376 expression in budding yeast
Figure 10 Requirement of the hEBP2-binding region of EBNAl in FR plasrnid segregation in yeast
Figure 11 Interaction of EBNAl with EBP2
Figure 12 Expression of hEBP2 in hEBP2-integrated yeast strains
Figure 13 EBNAl-rneâiated segregation in hEBP2-ùltegrated yeast strains
Figure 14 Schematic diagram of the hEBP2 protein LIST OF TABLES
Table 1 Stability of Segregation Plasmids in the Resence 51 and Absence of EBNAI and hEBP2 LIST OF ABBREVIATIONS
.- -
ARS autonomously replicating sequence
AT 3-aminotriazole
NT adenindthymine
ATP adenosine triphosphate bp base pairs BPV1 bovine papillomavirus type 1
BZLF 1 BamHI C teftward fiame 1
CAT chloramphenicol acetylûansferase CEN
CIP calf intestinal alkaline phosphatase
CTL cytotoxic T-lymphocyte m20 de-ionized water DMs doubleminute chromosomes
DNA deoxyribonucleic acid ms deoxynucleotide triphosphate
DS dyad symmetry
DTT dithiothreitol
EBER Epstein-Barr expmsed RNA
EBNA1 Epstein-Barr nuclear antigen 1
EBP2 EBNAl-binding protein 2
EBV Epstein-Barr vinis ECL enbanced cherniluminescence
A..-- 4 EDTA ethylenediaminetetraacetic acid
FISH Fluorescence in-situ hybridization
FR family of repeats
GFP green fluorescent protein
Gly-Ala gl ycine-alanine
Gly -hg glycine-arginine
hEBP2 human EBP2
HHV8 human herpesvirus 8
HMG-1 high-mobility group-1 Kb kilobase
kD kilodaltons
KSHV kaposi sarcoma-associated herpesvirus
lac lactose
LANA latency-associated nuclear antigen
LB Luria Bertani broth
LCLs lymphobiastoid ce11 lines LCV Iymphocrytovirus
Leu leucine
LiAc lithium acetate
LMP latent membrane protein
Mb megabase
MCS multiple cloning sites MME minictuomosome maintenance element -- - - . -.-- - MO minimal o@n of replication NEB New England Biolabs
NLS nuclear localization signal
OD opticai density
0riP origin of plasmid replication
PBS phosphate buffered saline
PCR polymerase chah reactian
PEG polyethylene glycol
PMSF phenylmethylsulfonyl fluoride
MG1 /2 recombination activating genes 1 and 2
RGG arginine glycine glycine
RNPs ribonucleic proteins rRNA ribosomal ribonucleic acid
RPA replication protein A
SC synthetic complete SDS sodium dodecyl sulphate
SDS-PAGE SDS polyacrylamide gel electrophoresis snRNPs small RNPs
SV40 simian vinis 40
TBE tris, bric acid, ethylenediaminetetraacetic acid
Trp tryptophan
Ur a YACs yeast artificial chromosomes
-----A %L --.-&A& - -- - yEBP2 yeast EBP2
YPD yeast extract peptone dextrose
2D two dimensional 1.1 Epstein-Barr Virus: Mode of Infection and Associateà Diseases
A member of the lyrnphocryptovinis (LCV)genus of the gamma herpesvims, Epstein-
Barr virus (EBV)affects humans worldwide. PNnary infection by EBV can take place
during early childhood, when it is usually asymptomatic or during adolescence, when it
cm cause infectious mononucleosis. Transmitted through the saliva, EBV infects the
epithelial cells of the oropharynx, where it undergoes lytic infection, though latent
infection has been observed in rare cases. The vhsalso infects the B-lymphocytes,
where latent infection and long-term persistence occur. in the nucleus of latently infected
cells, multiple copies of the 172 Kb EBV genome are maintained as double-stranded,
circular, DNA episomes at a constant copy nurnber. Viral reactivation occurs in a small
percentage of the infected B cells, where expression of the EBV-encoded BZLFl protein
leads to lytic replication of the virus. In ce11 culture, latent infection of the B-cells causes
proliferation of the cells, leading to irnmortalized lymphoblastoid ce11 lines (LCLs). EBV
infection can lead to several malignant diseases, even years after the primary infection,
such as nasopharyngeal carcinoma, Burkitt's lymphoma, immunosuppression-related
lymphomas, gastric carcinoma, leiomyosarcomas and some forms of Hodgkin's disease
(reviewed in Kieff, 1996; reviewed in Rickinson and Kieff.1996).
1.2. Viral Protein Expression During EBV Latent Infcetion
Only a small hction of the genes encoded by the genome are expressed during latency. These genes, which appear suflicient for cellular proliferation and Wal persistence, code for six nuclear proteins (EBNA-1,2,3A, 3B, 3C, and LP), 3 membrane
#-. - - - proteins (LMP 1, ZA, and 2B) and 2 untranslated RNAs (EBER 1 and 2); (reviewed in
Kieff, 1996). However, each of these genes is not expressed during al1 stages of EBV
latent infkction. The latent genes exptessed are dependent on the fom of latency
established in the infecteci B cell, which is cbaracterizd by the state of the cell. In non-
dividing cells, the "latency program" is established, where none of the latent genes are
expressed, with the possible exception of LMP2a (Babcock and Thorley-Lawson, 2000;
Babcock et al., 2000). In dividing cells, three foms of latency have been identified. The
"growth program" is a latency mechanism where dl latent genes are expressed and is
found in immunoblastic lymphoma and cells grown in culhur. The "EBNAlsnly
program" is a latency mechanism where only EBNAl is expressed and is found in
Burkitt's lymphoma. The "EBNAl/LMP program" is a latency mechanism where only
EBNAl and the LMPs are expressed and is found in Hodgkin's lymphoma (Babcock and
Thorley-Lawson, 2000; Babcock et al., 2000; reviewed in Thorley-Lawson et al., 1996).
Expression of al1 latent genes is not necessary for the stable persistence of the EBV
episome. The stable penistence only requires the Epstein-Barr nuclear antigen 1
(EBNAl) protein (Yates et al., 1984; Yates et al., 1985; Lupton and Levine, 1985).
Interaction of EBNAl with oriP, the EBV latent ongin of DNA synthesis, is responsible
for the replication of the viral episome and the stable partitionhg of the episome during cellular rnitotic division.
1.3. EBV Latent Origin of Replication, orZP
Yates and Sugden first identified the viral ongin of latent infection when they cloned fragments of the EBV gemme into a plasmid containing a selectable marker and
2--- 2 - - - ûansfected these plasmids into both EBV-positive and EBV-negative ce11 lines. They discovered that plasmids containing a 1800 bp sequence, now referred to as oriP, were maintained efficiently in EBV-positive cells through long-term selection (Yates et al.,
1984; Sugden et al., 1985). OnP, which is activated to initiate replication once per ce11 cycle (Yates et al., 1984; Yates and Guan, 1991; Adams, 1987), consists of two elements, namely the family of repeats (FR) and the dyad symmetry (DS) element (Rawlins et al.,
1985; Reisman et al., 1985); (Figure 1). Both elements contain sites recognized by
EBNAl dimers, which is an 18 bp imperfect palindromic sequence (Ambinder et al.,
1990; Rawlins et al., 1985). The FR element is composed of twenty high affinity
EBNAl binding sites, each located within a 30 bp tandem repeat. The DS element consists of four lower attinity EBNAl binding sites, with sites 3 and 4 contained within a
65 bp dyad symmetry. The two elements are separated by a 1 Kb region, which does not appear to be important for the known fbnctions of oriP. Deletion and insertion of sequences within this intervening region are tolerated and do not inhibit the replication, segregation or transactivation nuictions of oriP (Reisman et al., 1985; Yates et al., 1984).
Within oriP, the DS is the site for the initiation of bi-directional DNA synthesis
(Wysokenski and Yates, 1989; Gahn and Schildkraut, 1989; Yates et al., 2000). Binding of EBNAl dimers to two out of the four EBNAl binding sites (sites 1 and 2 or sites 3 and
4) within the DS is sufncient for replication initiation in some ce11 lines but efficient ongin activation in B cells requires al1 four sites (Yates et al., 2000). Though 2D gel electrophoresis analysis demonstrates that replication forks initiate within oriP at or near
DS, the DS is not the only site for DNA replication initiation within the EBV genome ( EBNA 1 binding site
Figure 1. OriP, the EBV latent ongin of replication. The FR (family of repeats) element contains 20 EBNAl binding sites. The DS (dyad symmetry) element contains 4 EBNAl binding sites, with sites 3 and 4 located within a dyad symmetry sequence as inâicated by the arrows. nie two elements are separatd by appmxhately 1 Kb of DNA. itself. Several other replication forks have been noted in the genome, most starting
------L - upstream of the DS (Gahn and Schilâkraut, 1989; Little and Schildkraut, 1995). When
the DS is deleted, replication of the genome hmthese other sites appear sufficient for
the virus to establish a latent infection and penistence in BL30 cells (Nono et al., 2000).
The FR element of oriP has several fuactions. To begin with, FR may enhance
replication hmthe DS element. Association of the EBNAl dimers bound to the FR and
DS elements, which results in the looping out of intervening DNA sequences, may allow
stabilization of the EBNAI-DS complex and subsequent DNA replication (Su et al.,
1991 ; Frappier and O'Donnell, 1991b; Frappier et al., 1994). It was shown that of the
twenty EBNAl binding sites within the FR, only seven were required for this
enhancement of replication (Wysokenski and Yates, 1989). However, ment studies
suggest that this function of FR may not be absolutely essentid, since efficient replication
hmDS in the absence of FR is observeci in 293 cells (Yates et al., 2000).
When bound by EBNA1, FR also acts as a segregation element and a transcription
activator (Chittenden et al., 1989; Gahn and Sugden, 1995; Reisman and Sugden, 1986;
Pugeilli et al., 1996). Both fùnctions are independent of DNA replication and can be
carried out in the absence of the DS (Middleton and Sugden, 1994, Reisman et Sugden
1986; Shire et al., 1999). ~urthemok,removal of up to 13 EBNA 1 binding sites from
the FR still allows efficient segregation of an oriP-containhg plasrnid and transactivation
when placed upstream or downstream of a promoter (Chittenden et al., 1989;
Wysokenski and Yates, 1989).
Lastly, FR has been shown to act as a pause and tedation site for DNA replication forks. In oriP-containhg plastnids and the EBV episomes, replication hmthe DS results in two forks rnoving away hmthe DS, where one fork appears to pause at the FR AL ------= - while the other fork continues around the plasmid and terminates at the FR (Gahn and
Schildkraut, 1989). This occurs due to the binding of EBNAl to its high afnnity sites in
the FR, which blocks the passage of the forks (Ermakova et al., 1996). Studies with an
in-vitro replication system hes shown that six of the twenty EBNAl binding sites are
sufficient for this phenornenon @bar and Schildkraut, 199 1). However, EBNA l does
not completely inhibit the passage of forks thugh the FR as plasmids containing a DS
element flanked by two FR elements replicate efficiently (Kirchrnaier and Sugden, 1995).
1.4. Epstein-Barr Nuclear Antigen 1 (EBNA1)
EBNA1, a 641 amino acid ongin-binding protein, is a multifunctionai protein that is
the only latent viral protein that is expressed in al1 EBV-related diseases (nviewed in
Rickinson and Kieff, 1996). This, dong with the observation that over-expression of
EBNAl in mice cm cause B ce11 lymphomas (Wilson et al., 1996), suggests that EBNAl
plays a role in cellular proliferation leading to cancer. EBNAl is also the only viral
protein that must be expressed in order to maintain the latent genome in proliferating
cells and, as such, is a target for anti-viral drugs.
1.4.a. Protein Domains and Structure of EBNAl
The primary sequence of the EBNAl protein (shown in Figure 2) contains two Gly-
Arg (glycine-arginine) rich regions. The htof such regions is located at the N
terminus, between residues 41-53, while the second mgion is located at the central region,
between residues 325-376 (see Figure 2). Together these Gly-Arg regions contain a total Gl~-&g DNA binding Acidic Gly-Ala npeat domam) NLS & dirnerization tail ...... 4. $ 4 4 ......
segregation, tramactiva tion, and hEBP2-biaditrg
Figure 2. Schematic diagram of the fùnctional regions of EBNAI. Shown are the Gly-Arg regions, Gly-Ala repeat, nuclear localization signal (NLS), DNA bindingldimerization region and acidic tail. The region required for the segregatioaal, transactivational and hEBP2-binding activities is also shown. Arnino acids spanned by each region are indicated. of three 'RGG' (mghine glycine glycine) motifs. Such motifs are implicated in protein-
--.-.------protein interaction, RNA binding and nucleolar targeting of proteins (Snudden et al.,
1994). The 325-376 Gly-Arg region is quind for several EBNAl activities. It mediates
attachment of EBNAl to host chromosomes during rnitosis and is required for the
segregation and transactivation fùnctions of EBNAI. Deletion of this region abrogates al1
three activities of EBNAl (Wu et al., 2000; Shire et al., 1999; Ceccarelli and Frappier,
1998; Ceccarelli and Frappier, 2000). The N-terminal region spanning the 41 -53 Gly-Arg
and flanking sequences was shown to be involved in host chromosome attachment,
segregation and transactivation activities of EBNA1, suggesting that the N-terminal Gly-
Arg sequence itself might contribute to these activities (Marechal
al., 2001; Hong a al., unpublished results; Mackey and Sugdca, 1999). However, a
deletion that removes only the Gly-Arg sequences between amino acids 41 and 53 does
not affect EBNAl 's ability to bind mitotic chromosomes or EBNAl's segregation and
transactivation fhctions (Hong et al., unpublished results). Thus the functional
contribution of the 4 1-53 Gly-Arg is not clear.
Both Gly-Arg regions are involved in protein-protein interactions, as both regions
interact with p32/TAP, a cellular protein of UnkRown huiction residing mostly in the
mitochondria and plasma membrane (Wang et al., 1997; Van Scoy et al., 2000). In
addition, the central 325-376 Gly-Arg sequence interacts with the human EBNAI-
binding protein 2 (hEBPZ), which is important for EBNA l -mediated segregation (Shire
et al., 1999). The central Gly-Arg region is also termed the DNA looping domain, as it
allows association between EBNAl dimers bound to the FR and DS, resulting in the looping out of the region between the FR and DS. Whm association is between EBNA
-- ---A dimers bound to diffmnt DNA molecules, this results in the linking of the two DNA
molecules (Laine and Frappier, 1995; Frappier and O'Donnell, 1991 b; Su et al., 1991;
Middleton and Sugden, 1992).
Located between the two Gly-Arg regions is a large Gly-Ala (glycine-alanine) repeat,
encompassing amino acids 90-324. Deletion of this region does not affect the
transactivation, replication or segregation function of EBNA 1 (Yates and Camiolo, 1988;
Yates et al., 1985). However, the presence of this region allows EBNA 1 to escape the
host immune system and prevents EBNAl hmeliciting a cytotoxic T-lymphocyte
(CTL)response (Blake et al., 1997; Levitskaya et al., 1995). The nuclear localization
signal (NLS) of EBNAl maps to amino acids 379-386. The presence of this region is
important in order for EBNAl to gain entry into the nucleus, where it resides during
intephase (Ambinder et al., 1991). Entry of EBNAl into the nucleus may also be aided
by the karyopherin a protein, a nuclear transport factor that interacts with EBNAl (Kim
et al., 1997; Ito et al., 2000). At the C-terminus of EBNAl resides the acidic tail, which
spans amino acids 620-641 .This region is not required for the transactivation, replication
or segregation huictions of EBNAl (Yates and Camiolo, 1988; Ceccarelli and Frappier,
2000; Polvino-Bodnar and Schaffer, 1992).
The DNA binding and dimerkation region of EBNAl is well characterized and maps
to amino acids 459-607 (Ambinder et al., 1991). This region is responsible for the
formation of EBNAl dimers both in solution and when bound to DNA, as welI as for the
binding of the dirners to the 18 bp EBNAl recognition sequence (Frappier and
O'Domeil, 1991 a; Arnbinder et al., 1991; Shah et al., 1992; Chen et al., 1993; Summers et al., 1996). Crystal structures of the EBNAl DNA binding and dimerization region
------have been solved both in the presence and absence of the EBNAl DNA recognition sites
(Bochkarev et al., 1995; Bochkarev et al., 1996;Bochkarev et al., 1998). These
structures revealed that the region consists of two domains, the con domain (residues
504604) and the flanking domain (residues 461-503). The core domain, which does not
interact with the DNA in the crystal structure, is nsponsible for the dimerization of the
EBNAl protein. It is an eight-stranded antiparallel B-barre1 that consists of four B-
strands and two a helix hmeach monomer (Bochkarev et al., 1995; Bochkarev et al.,
1996). The core domain is homologous in structure to the DNA binding domain of the E2
protein of bovine papillomavirus (Hedge et al., 1992). The flanking d~mainis involved
in sequence-specific contacts with the bases within the EBNAl recognition sites. This
domain, which is co~e~tedto the core domain by a loop, consists of an a helix that
contacts the outer portion of the recognition site and an extended chah that is inserted in
the minor gmove of the DNA (Bochkarev et al., 1995; Bochkarev et al., 1996; Bochkarev
et al., 1998). Recmt mutational studies have demonstrated that the flanking domain may
not be the only structure that interacts with DNA.The core domain also makes sequence-
specific contacts with the DNA, contacts that are otherwise not observed in the crystal
structure. Interaction of the core domain with the DNA is through one a helix fbm each
rnonomer, tenned the recognition helix. Point mutations in potential hyârogen bond donors residhg in this helix impair the association of EBNAl with its recognition site
(Cniickshank et al., 2000). The EBNAl recognition helix is a structural homolog of the
E2 recognition helix, which mediates al1 of the sequence-specific DNA contacts in the E2
DNA bhding domain (H6dge et al., 1992). I.4.b. EBNAl Functions ------EBNAl makes several fiinctional contributions to EBV latency. It is responsible for
the transactivation of several latent viral genes and the replication and segregation of the
EBV episomes. EBNAl most likely canies out these fiinctions thugh interactions with
host cellular proteins.
L 4. hi Transcriptional Activities of EBNA l
EBNAl has the ability to both activate and repress transcriptional activity. EBNAl is
able to enhance transcnption of reporter genes by binding to the FR enhancer element
(Reisman and Sugden, 1986; Sugden and Warren, 1989). EBNAl bound to FR can also
activate transcription hmthe EBV Cp and Wp promoters. These promoters are
responsible for the transcription of al1 EBNAs, which includes EBNAl, during the
"growth program fom of latency. EBNAl also activates transcription from promoters
responsible for the expression of the LMPs (Gahn and Sugden, 1995; Puglielli et al.,
1996; Woisetschlaeger et al., 1990; Sugden and Warren, 1989). Furthemiore, EBNA 1
has been implicated in transcriptional activation of some cellular genes. Specifically,
EBNAl can activate the expression of MG1 and RAG2, which are recombination
activating proteins involved in the immune system (Srinivas and Sixbey, 1995).
In addition to dancing its own expression, EBNAl can also repress its expression
hmthe Qp promoter. The Qp promoter is responsible for the expression of EBNAl in
dividing cells that exhibit a fom of latency other than the "growth program (Schaefer et
al., 1995). Repression occurs through bindiiig of EBNAl to two recognition sites located
upstream of the promoter (Sample et al., 1992; Schaefer et al., 1991). The transactivation hction of EBNAl is separable from its replication fiinction. An :------.-- - .------EBNAl mutant lacking residues 325-376 is able to replicate at wild-type levels but is
unable to tnuisactivate a chloramphenicol acetylüansferase (CAT)reporter gene placed
downstream hmthe FR element (Ceccarelli and Frappier, 1998; Ceccarelli and
Frappier, 2000). This separation becomes important when using the oriPEBNA1 system
to replicate plasmids in mammalian cells by ensuring that the introduction of EBNAl
does not affect the expression of cellular genes. The transactivation function of EBNAl
is less species-specific than its replication fbction, as EBNAl and FR can enhance
transcription in various organisms, such as dents and yeast (Wysokenski and Yates,
1989; Shire et al., 1999).
L 4.6.ii. Replication Function of EBNAl
Interaction of EBNAl with the DS results in the initiation of replication of EBV
episomes and oriP-containhg plasmids once per S phase. The binding of EBNAl to the
DS is CO-operativeand likely involves unwinding of the DNA (Smerset al., 1996).
However, association of EBNAl with the DS is not sufficient for replication to occur.
The fact that EBNAl is bound to these elements throughout the ce11 cycle (Hsieh et al.,
1993) and that EBNAl possesses no enzymatic activities (Frappier and O'Domeii,
1991a) suggests that replication most likely occurs through interaction with cellular
factors (Middleton and Sugden, 1992). EBNAl likely recruits the cellular replication
machinery to the DS during S phase. In fact, interaction of DNA-bound EBNAl with
RPA has been connmied. RPA is the human single-strand DNA binding protein that
helps generate and stabilize single-stranded regions that serve as templates for DNA -.
synthesis (Zhang et al., 1998). The existence of host factors for EBV replication is also
-----L--A. . -7- SAX---- 2 - -2 - - -- - = - suggested by the species-specificity of EBNAl -mediated replication. Replication by
EBNAl only occurs in primates and fails to occur in rodent cells (Yates et al., 1985).
This species specificity may be due to the ability of EBNAl to interact with replication
proteins hmprimates but not rodent cells.
L 4. b.iii. Segregation Fùnction of EBNAI
Replication of the EBV episomes is not sufficient to maintain the life-long persistence
of the virus. The virus must also ensure that the replicated episomes are efficiently
partitioned between dividing cells during cellular mitosis, such that each daughter ce11
harbors the Wus. The EBV episome does not use a segretion mechanism similar to
that of the host chromosomes, as no long centmmeric regions have been found dong the
virus genome. Segregation of the episomes requires the FR element as well as EBNAl in
tram (Krysm et al., 1989). Certain mutations in EBNAl have been shown to inhibit
stable segregation of EBV plasrnids without afliecting the replication of the piasmids
(Yates and Camiolo, 1988; Shire et al., 1999), indicating that EBNAl is required for
segregation. The requirement for FR was first show when addition of the FR element to
unstable plasmids containing the DS element and a selectable drug marker conferred
stability upon the plasmids in the presence of EBNAl (Reisman and Sugden, 1986;
Reisman et al., 1985). EBNAl- expressing cells containing plasmids with the DS, the FR
and the selectable marker were able to survive under cimg selection for several weeks,
while cells containing plasmids with the DS and the selectable marker did not survive the
hgselection. Interaction of EBNAl with the FR is required for EBV segregation (Chittenden et al., 1989; Reisman and Sugden, 1986), though this interaction cannot be
-- -A- e L - -- sufficient for segregation to occur. The existing model for EBV segregation states that
EBV episomes segregate by attachment to host mitotic chromosomes. Specifically,
interaction of EBNAl with both the condensed chromosomes during mitosis and with FR
tethers the EBV episomes to the host chromosomes, thus enabling the episomes to
segregate along with the host genome.
Several lines of evidence exist to validate this model, which suggests that segregation
and mitotic chromosome association of EBV is coupled. First, during mitosis, when
EBV episomes segregate to daughter cells, both EBNAl and EBV episomes have been
shown to associate with the chromosomes through irnmunofluorescence microscopy and
fluorescence in-situ hybridization (FISH) studies (Petti et al., 1990; Delecluse et al.,
1993). OnP-containhg artificial chromosomes and plasmids have also been shown to
associate with mitotic DNA in the presence of EBNAl (Simpson et al., 1996; Kanda et
al., 2001 ;Wu et al., 2000). In one study, the addition of oriP sequences to YACs (yeast
artificial chromosomes) confmed stability upon the YAC and enabled it to bind
metaphase chromosomes, when introduced into human 293 cells expressing EBNA 1
(Simpson et al., 1996). FISH analysis on these YACs showed brightly staining dots
located throughout the metaphase chromosomes. Similar results were obtained using plasmids containhg oriP, the EBNAl gene and lac operator repeats (Kanda et al., 2001).
When such plasmids were transfated into HeLa cells expressing the lac repressor fûsed to GFP, the plasmids were observed as fluorescent dots (due to the binding of the GFP- tagged lac repressor to the lac operator sites) scattered over the mitotic chromosomes, which co-locaiized with EBNAI . The observations that EBNAl ,EBV episomes and onP-containing constructs localize to mitotic chromosomes without any distinct patterns
-..------L.-* := A -- - - A - suggest that binding to the chromosomes is not sequence-specific.
Second, like EBV segregation, the association of EBV plasmids with metaphase
chromosomes aiso requins both FR and EBNAI. A recent study by Kanda et al (2001)
(mentioned above) demonstrateci that oriP- plasmids tagged with the lac operator repeats
and expressing EBNAl associated with the mitotic chromosomes when both EBNAl and
FR were present. Deletion of the EBNAl gene or the oriP sequence hmthe plasmid
abrogated the ability of the EBV plasmids to tether ta the mitotic chromosomes.
However, deletion of the DS alone did not affect this ability, suggesting that the presence
of EBNAI and FR were suflicient for the rnitotic chromosome association of the
plasmids.
Third, EBNAl regions required for the mitotic chromosome association of EBNA l
(and therefore, of EBV plasmids) are the same as those required for the segregation
function of EBNAl, which suggests that these two huictions are coupled. An EBNAl
mutant lacking residues 325-376 (previously mentioned as the central Gly-Arg region) is
nuclear and can efficiently replicate an oriP plasmid. However, such a mutant cannot
segregate these plasrnids nor cm it bhd to the mitotic chromosomes, suggesting the
requirement of the Gly-Arg region in both fhctions (Shire et al., 1999; Wu et al., 2000).
Furthexmore, an N-terminal region of EBNAI,which maps to amino acids 8-67, was also
shown to be involved in both fiinctions. Fusion of EBNAl amino acids 8-67 to GFP
resulted in the association of the fision protein with the condensed chromosomes
(Marechal et al., 1999). An EBNAl mutant lacking the same sequences exhibited a weak
segregation activity as compared to wild-type (Wu et al., unpublished results). Lastly, a GFP-taggeâ C-temiinal EBNAl fragment (containhg residues 379-641). which lacks
-LL--x- - - - A L------PL both regions 325-376 and 8-67, was unable to bind mitotic chromosomes and did not
allow the stable segregation or the mitotic chromosome association of onP plasmids
(Hung et al., 2001).
Recently, work by Hung et al. (2001) provided fùrther proof that mitotic chromosome
association is required for EBV segregation. They first detemined that the N-terminal
region of EBNAl (residues 1-386), which contains both the N-terminal sequences and
the 325-376 Gly-Arg region implicated in mitotic chromosome association, was
necessary for onP-containingplasmid segregation and for binding of both EBNAl and
oriP plasmids to mitotic chromosomes. Hung et al. then fused the C-terminal portion of
EBNAl, lacking amino acids 1-386, to cellular mitotic chromosome binding proteins.
These cellular proteins, which are amino acids 1-90 of HMG-1 (high-mobility group-1)
and histone Hl, were able to mediate the localization of the EBNAl C-terminal portion
to metaphase chromosomes. These fusion proteins were aîso show to recruit oriP
plasmi& to the chromosomes and to maintain these plasmids under long-term selection.
1.5. EBNAI-hEBP2 (human EBNAl-Binding Protein 2) Interaction
hEBP2 was first discovered in a two hybrid scmn in budding yeast to isolate
EBNAl -interacting cellular proteins hma human B-lymphoma cDNA library. nie
interaction between EBNAl and hEBP2 was also observed through co- immunoprecipitation experiments of baculovinis-produced proteins in insect cells.
Furthemore, yeast one-hybnd experiments have shown that hEBP2 can also bind
EBNA 1 that is bound to the FR (Shire et al., 1999). Though it is not conhed whether the interaction betwecn the proteins is indirect or direct, the observation that they are *.- - - - - capable of an interaction in more than one organism suggests that the interaction is likely
direct.
Two other cellular proteins, p32/TAP and karyopherin a,have been isolated as
EBNAl-interacting proteins fiam yeast two-hybrid screens (Wang et al., 1997; Kim et
al., 1997). However, the significance of the interaction between these cellular factors and
EBNAl is not known. p32RAP has been hypothesized to be involved in either EBNAI-
mediated replication or transactivation, but this has not been connmied (Wang et al.,
1997; Van Scoy et al., 2000). Based on its known cellular hction, karyopherin a may
bind EBNAl to mediate its entry into the nucleus (Kim et al., 1997). The interaction
between bEBP2 and EBNAl, however, is the only interaction thus far implicated in EBV
segregation.
I.5.a. Properties and CelIular Function of EBPZ
hEBP2 is a 306 amino acid protein that is not welltharacterized. It is nucleolar during
interphase and appears to be expressed in al1 proliferating human cells (Chatterjee et of.,
1987). Its protein sequence consists of a central coiled-coi1 region (approximately amino acids 100-208) that may be involved in protein-protein interactions. This central region, along with the C-terminal region of hEBP2, is highly conserved among the various hEBP2 homologs (Figure 3). Homologs of hEBP2 exist in a variety of organisms, such as Sacchmyces cerevisiae (budding yeast), Schizosaccharornyces ponibe (fission yeast) and Caenorhabditis elegans. Both budding yeast and Celegans contain an EBPZ that is larger in size than hEBP2 due to an extended N-terminus (Shire et aL, 1999; Huber
et al., 2000).
& - - Information acquired thus far on the cellular hction of EBP2 has corne hmstudies
of yEBP2, the yeast homolog of hEBP2. yEBP2 is a 427 amino acid protein that is also
nucleolar. yEBPr is an essential gene; cells containhg a yEBP2 gene disruption undergo
a maximum of two divisions and are unable to form coionies (Huber et al., 2000; Tsujii et
al., 2000). The essential fiuiction provided by yEBP2 requires the central and C-terminal
sequences of the protein, while the N-terminal extension, which is not present in hEBP2,
is dispensable for the function (Huber et al., 2000).
Two groupe have investigated the function of yEBP2 (Huber et al., 2000; Tsujii et
aL,2000), and discovered a role for yEBP2 in ribosome biogenesis. A temperature
sensitive mutant of yEBP2 was shown to be defective in the processing of rRNA for the
large ribosomal subunit (60s subunit). One processing step in particular (27SA to 27SB),
which is involved in the pmcessing of the 35s pre-rRNA into 25s and 5.8s rRNA, was
found to be affécted at the non-permissive temperature. The deof yEBP2 in these
processes is fiuther connmied by the observation that yEBP2 interacts with the Rrsl
protein. Rrsl protein is a nuclear protein that also plays a role in nbosome production,
mainly in 25s rRNA processing and 60s subunit assembly. Given that hEBP2 is also
nucleolar and is highly homologous to yEBP2, it is ükely that hEBP2 plays a sllnilar role
in nbosome synthesis in human cells.
1.S.b. Sidficance of EBNAl-hEBPZ Interaction in EBV Segrejgation
nie htclue to the involvement of hEBP2 in EBNAl-mediatecl segregation arose when hEBP2 was found to interact with an EBNAl region that is essential for its segngation activity (Shin et al., 1999). Deletion-J- of this region, which corresponds to
------a------L residues 325-376, not only lead to a rapid loss of oriP plamnids and the inability of
EBNAl to bind chmatin, but also abrogated the interaction of hEBP2 with EBNAl
(Shire et al., 1999; Wu et al., 2000). Smdler interna1 deletions within this region that did
not affect segregation of oriP plasmids by EBNAl or the ability of EBNAl to bind
I chromosomes also did not affect the ability of EBNAl to associate with hEBP2.
Secondly, hEBP2 was found to associate with the mitotic chromosomes, with a
distribution that was much like EBNAl (Wu et al., 2000). These findings lead to the
hypothesis that hEBP2 connects EBNAI, and hence, EBV episomes, to the Mtotic
chromosomes. In other words, binding of EBNAl to hEBP2, which is associated with
the mitotic chromosomes in the absence and presence of EBNAI, allows EBNAl to
tether the EBV episomes to the chromosomes for proper episomal segregation.
Association of hEBP2 with the mitotic chromosomes is through the chromosome periphery, which is also called the perichromosomal layer. This layer is located at the
surface of chromosomes during mitosis. The layer is characterized by closely packed
fibrils and granules, which consists of ribonucleic proteins (RNPs) as well as proteins hmthe nuclear matrix, nuclear membrane and nucleolus. Nucleolar proteins that associate with the chromosome penphery are those involved in rRNA processing. During different stages of mitosis, these proteins associate with the surface of the chromosomes to msure that they are equally disûibuted to the daughter cells and not lost during the separation process (reviewed in Hemandez-Verdun and Gautier, 1994). Proteins such as snRNPs (small RNPs) associate with the layer dudg metaphase and anaphase, while nucleolar pmteins, such as fibriliarin and B23, associate with the layer throughout mitosis, hmearly prophase to telophase (Gautier et al., 1994; Fomproix et al., 1998).
-A-- - -L - Localization of hEBP2 to the chromosome periphery throughout mitosis places hEBP2 at
the nght place at the nght thne to aid in EBNAl-mediateci attachment of the EBV
episomes to the mitotic chromosomes for proper EBV segregation.
1.6. Segregation Mecbanism of Extrachromosomal DNA molecules
Chromosomes are stably segregated to daughter cells during mitosis by the attachent
of their centmmeric (CEN) sequences to mitotic spindles. Extrachmmosomal DNA
molecules also require a partitionhg mechanism that will ensure their retention within
dividing cells. Segregation of the EBV episomes by attachent to host mitotic
chromosomes is a segregation mechanism that is not unique to this viras. Low copy
number episomal viwes and double minute chromosomes also appear to ensure their
persistence within cells by a mechanisrn that involves hitchhiking of the DNA molecules
on mitotic chromosomes. Yeast 2-micron plasmids and yeast CEN plasmids ensure their
maintenance in cells by using the host segregation machinery.
1.6.a. Papillomavirus
Studies conducted on bovine papillomavhs type 1 (BPV-1)show that this vins
exists as low-copy number, episomal plasmids within infected epithelial cells (reviewed
in Howley, 1996). During latent infection, the BPV-1 genome is stably maintaineci at a
constant copy number within host cells thugh interaction of the El and E2 proteins with
cis-acting sequences (Chiang et al., 1992). El, which is a Wal ongin recognition factor
and a helicase, is nsponsible for the replication of the viral genome (Seo et al., 1993; Hughes and Romanos, 1993). E2 is a üanscriptional factor that also plays a dein BPV
------=--A--..-LL------replication and is required for the segngation of the virus (McBride et al., 1991). It
consists of a N-terminal transactivation domain linked to a C-terminal DNA-binding
domain by a hinge region (Bastien and McBride, 2000). Binding of El and E2 to the cis-
acting minimal origin of replication (MO) is required for the replication of the virus.
Presence of El, E2 and the cis minichromosome maintenance elexnent (MME),which
contains multiple E2 binding sites, is requited for stable partitioning of the BPV episomes
(Piirsw et al., 1996).
Through antibody staining, E2 was found to attach to the mitotic chromosomes
through al1 stages of mitosis, both in the absence and presence of the BPV DNA (Bastien
and McBnde, 2000; Skiadopoulos and McBride, 1998). nirough the use of viral DNA
probes, BPV genomes were alsa shown to associate with the mitotic DNA (Skiadopoulos
and McBride, 1998). Both protein and DNA appear as random dots on the chromosomes,
suggesting that binding is not sequence-specific. Furthemore, plasmids containing E2
binding sites are also able to bind the condensed chromosomes in the presence of E2
(Ilves et al., 1999). These data suggest that E2 may mediate BPV genome segregation
through attachent of the genome to host mitotic chromosomes. Specifically, attachent
of the DNA to the chromosomes requires an intact MME and a version of E2 that can
mediate plasmid segregation. With El and E2 expression, BPV-based constnicts that are
capable of replicating (due to the presence of the replication ongin MO) can also
segregate and attach to chmmatin in the presence of an intact MME. However, the sarne
plasmids are lost rapidly hmcells and fail to attach to chromatin without an intact
MME. The El binding site in the BPV MO (or El itself) cm be removed without affecting the association of BPV plddswith chromath. However, such plasmids
--& ---- -2 A -.. - - - cannot persist efficiently due to the lack of replication (Iïves et al., 1999).
E2 mutants that fail to maintain the viral DNA alsa fail to attach the DNA to mitotic
chromosomes. Lehrnan and Botchan (1998) dernonstrated that a mutation in E2 (which
they called an A4 mutation) that abolished the ability of the mutant to segregate the BPV
DNA also did not allow the attachent of itself and the viral DNA to the mitotic
chromosomcs. Such a mutation did not hinder the replication function of EZ but instead,
lead to an E2 mutant that replicated the viral DNA more efficiently than wild-type E2.
Recently, Bastien and McBride (2000) mapped the mitotic chromosome association
region of E2 to the E2 transactivation domain. This domain was both necessary and
sufficient for the association of E2 with host chromosomes. Point mutations in this
domain that abrogated the transactivation fùnction of E2 also destroyed the ability of E2
to bind mitotic chromosomes. Since the DNA-binding domain of E2 is dispensable for its
mitotic chromosome association, it is unlikely this association occurs through direct
binding of E2 to DNA. Association of E2 with the host chromosome likely involves
protein-protein interactions.
1.6. b. Kaposi Sarcoma-Associated Herpesvhs
The Kaposi sarcoma-associateci hcrpesvi~~~(KSHV) is an episornal virus that may
also facilitate segregation through association with host chromosomes. KSHV,which is
also known as human herpesvirus 8 (HHVB), is a gamma herpesvinis like EBV and is a
major cause of cancer in AIDS patients (reviewed in Beral and Newton, 1998; Chang et
al., 1994). Primary latent infection by KSHV likely occw in CD19' and C~22+B- lymphocytes, with the infection thni moviag to the endothelial cells (Li et al., 1996).
&--A - -.LA------Though little is hown about how the multiple copies of KSHV episomes (Decker et al.,
1996) are maintaineci witbthe Uifected celis, the KSHV genome is seen to associate
with the mitotic chromosomes (Cotter and Robertson, 1999b; Ballestas et al., 1999).
immunofluorescence studies to determine the locaiization of LANA, a latency associated
nuclear antigen that is always detected in KSHV-infécted cells (Kedes et al., 1997),
revealed that LANA CO-localizeswith the KSHV genome. In the absence of the viral
DNA, LANA is dimisely associated with the chmmosomes, while in the presence of
DNA, both LANA and the DNA stain the chromosomes in dots, suggesting that the presence of viral DNA might recruit LANA to specific chromosomal sequences
(Ballestas et al., 1999). However, the spacing of the "dots" does not appear to form a regular pattern, which contradicts the idea of sequence-specific binding. Apart hmCO- localizing with KSHV genome, LANA is also respoasible for the segregation of KSHV
DNA. Transfection of B cells expressing LANA with KSHV DNA comprised of a region nferred to as 26 (located within the first 34 Kb of the KSHV genome), which contains LANA binding sites, leads to the stable maintenance of the DNA. Furthemore, the 26-containing KSHV DNA was also obsaved to attach to chromosomes in the presence of LANA. In the absence of LANA, the 26 DNA was lost hmcells and did not associate with chromosomes (Cotter and Robertson, 1999b).
A ment study ushg GFP-tagged LANA hgments amplified hmBBGI cells
(which are KSHV positive and express a fom of LANA that consists of 1036 arnino acids) mapped the mitotic chromosome association region of LANA to amino acids 5-22
(Piolot et al., 2001). Transfection of HeLa cells with LANA amino acids 5-22 fùsed to GFP lad to the association of the hision protein with mitotic chromosomes, as observed
-&--.A------by fluorescence micmopy. Fusion of LANA lacking this region (LANA sequences 33-
1036) to GFP did not show any association of the fusion protein with mitotic
chromosomes. In both cases, the SV40 NLS was also added to the fusion proteins to
ensun their entry into the nucleus, since neither fusion proteins contain the LANA NLS,
which was mapped to amino acids 24-30 in the same study.
I.6.c. Double Minute Chromosomes
Viral genomes are not the only entities that are implicated to segregate via host
chromosomes. DNA molecules like the double minute chromosomes (DMs) an also
predicted to ride on the host chromosomes during chromosomal segregation. DMs are
circular acentric molecules that are usually 1-2 Mb in size and are autonomously
replicating and highly stable in dividing hurnan cells. They reside in cancer cells and
usually cany oncogenes or hg-resistance genes (reviewed in Hahn, 1993). Light and
electron rnicroscopy revealed that DMs are ofien found associated with the mitotic
chromosomes. Integration of GFP-tagged histone H2B into DMs also showed clusten of
DM attached to the chromosomes as the chromosomes move away hmeach other to
daughter cells during cellular mitosis. These DNA molecules are found to tether to the
periphery of the chromosomes, near the telomeric regions (Kanda et al., 2000). This
tethering may be the reason for its stable persisteme within cancer cells.
I.6.d. Yeast 2-Micron Plasmids
2-micron plasrnids are autonomously rcplicathg molecules that natiually exist in yeast at high copy nurnbcrs (Murray and Szostak, 1983). Their continued existence in cells
--- 2 2 2 -- requKes both a mechanian for amplification and for partitionhg (Vokert et al., 1986;
Som et al., 1988; Velmurugan et al., 2000). Recently, analysis of the 2-micron plasmid of
budding yeast lead to the hypothesis that these plasmids likely use the host segregation
machinery to ensure their peïsistence in cells (Velmunigan et al., 1998; Velmunigan et
al., 2000). Partitioning requires the Rep 1 and Rep2 proteins, which are encoded by the
plasmid, and the cis-acting locus STB (also REP3); (Kikuchi, 1983; Wu et al., 1987).
Microscopy studies indicate that both the Repl and Rep2 protein CO-localizeon the 2-
micron plasmid through interaction with STB and that the entire complex CO-localizesin
the nucleus (Velmurugan et al., 1998; Velmurugan et al., 2000; Scott-Drew and Murray,
1998). Specifically, during nuclear division, the complex is mainly located near the poles
of the yeast mitotic spindles (Velmunigan et al., 2000). These data are equally consistent
with the possibility that the 2-micron plasmids attach directly to mitotic spindles or
associate with the chromosomes at the centromeric regions.
Fluorescence microscopy of synchronized cells at diffmnt stages of mitosis revealed
that both the 2-micron plasmids and chromosomes segregate in a similar manner
(Velmunigan et al., 2000). Specifically, they exhibited similar movements and positions
in the cells at the sarne time. This similarity was dependent on the presence of Repl and
Rep2 (Velrnurugan et al., 2000). Absence of these proteins led to missegregation of the
plasmids and gave a spatial distribution for the plasmi& that was different hmthe
spatial distribution of the chromosomes at various times during the ce11 cycle. Mutations
in genes that abrogated the ability of host chromosomes to segregate also lead to
missegregation of the plasmid (Chan and Botsteh, 1993; Velmunigan et al., 2000). The data suggest that the RepllRep2lSTB complex allows association of the plasmid with the
.s- = - - - - mitotic spindles or the host chromosomes, where the plasmids can utilize the host
segregation machinery to ensure its proper partitioning.
I.6.e. Yeast CEN Plasmids
Apart kom the 2-micron plasmids, yeast can also harbor DNA plasmids that contain
the ARS (autonomously replicating sequence) and CEN (centromeric) chromosomal
elements, which are replication and segregation elements, respectively (reviewed in
Newlon and Theis, 1993; reviewed in Fitzgerald-Hayes, 1987). Plasmids containing the
ARS element alone cm replicate but are mitotically unstable while plasmids containing
the ARS and CEN elements are mitotically stable and are maintainecl at a low copy
nurnber (Hieter et al., 1985). The mechanism by which CEN elements on plasmids
mediate segregation is assumed to be the same as the mechanism used for chromosome
segregation, which involves attachment of the plasmids to mitotic spindles (reviewed in
houe, 198 1). but this has never been demonstrateci. CEN elements do not work as efficiently on plasmids as they do on chromosomes, as CEN-containhg plasmids are subjected to a loss rate of at lest 1% per generation and the known loss rate for some yeast centromeres is only 0.002% per generation (Hieter et al., 1985).
The ktcentromeric region in budding yeast was identifid hmchromosome III and was subsequently named CEN3 (Clarke and Carbon, 1980). CEN3 was found in a 1.6 Kb fiagrnent of chromosome III DNA which is located near the CDClO locus. Addition of this hgment to an ARSplasmid led to the stabiüzation of the plasmid, both mitotically and meiotically. Shce this discovery, several other centromeric regions have been identified and cloned, such as those hmchromosomes 11,6,4 and 5 (C.11, CEN6,
-- - L - 2 A------CEN4 and CENS) (Hieter et al., 1985).
Unlike centromeres in fission yeast and higher eukaryotes, which are relatively large,
centrometes in buddhg yeast are short and compact (Bloom, 1993). The budding yeast
centrornere is a stretch of 125 bp of DNA that consists of three conserved elements:
CDEI, CDEII and CDEIII (reviewed in Hegemann and Fleig, 1993). CDEI, which is
located at one end of the centromeric region, is a 8 bp conserved element that plays a
minor dein chromosome segregation. Deletion of this region of the centromere
increases the rate of chromosome loss by 10-30 fold (Baker and Masison, 1990;
Niedenthal et al., 1991). Adjacent to CDEI is CDEII, which is a 78-86 bp region with a
sequence that can Vary among the various chromosomes but that has a high (90%) AR
content. CDEII is essential for proper chromosome partitioning, since deletion of this
region leads to an inactive centromm. However, smaller deletions are tolerated within
this sequence without causing a high delpee of chromosome missegregation, as long as
the percentage of AIT is conserved and the lmgth of 78-86 bp is maintaincd (reviewed in
Gaudet and Fitzgerald-Hayes, 1990). Beside CDEII, lies CDEIII, which is the most
highly conserved element of the centmmere. CDEXiI is a 25 bp element in which single
point mutations lead to increased rates of chromosome loss (McGrew et al., 1986).
Orientation of CDEIII with respect to CDEII is dso important for the proper fùnctioning
of the centromere (Murphy et al., 1991).
The centromere acts as a binding site for many proteins. Interaction of these proteins
with the centmmeric DNA foms a DNA-protein complex called the kinetochore.
Formation of a kinetochore is cellcycle regulated (Kbgsbury and Koshland, 1991). At mitosis, each kinetochore attaches to a single microtubde (Peterson and Ris, 1976), ------where attachment of a paired kinetochore to microtubules extending hmopposite poles
of the mitotic spindle allows the separation of the condenseci sister chromatids and the
movement of each chromatid into a newly fonning cell. Inability to attach to the
microtubules leads to chromosome missegregation.
Thus far, several protein components of the kinetochore have been identified, but not
al1 are required for segregation. The CBF3 complex, which is a multi-subunit protein
complex that binds CDEIII upon phosphorylation, is an essential component of the
kinetochore (Lechner and Carbon, 1991). Deletions in CBF3, like CDEIII, are lethal and
abrogate the stability of the chromosomes (Kingsbury and Koshland, 1991). Interaction
of CDEIII with CBF3. which was shown through DNA Wtychromatography (Lechner
and Carbon, 1991), is absolutely required for the microtubule binding activity of the
kinetochore. However, the interaction is not sufficient for this activity. The presence of
additional proteins that associate with CBF3 and aid in the binding of the centromeres to
the microtubules is also required (Kingsbwy and Koshland, 1993). In fact, the current
mode1 suggests that the main microtubule attachent site of the kinetochore resides in
CDEUI. The CDEIII-CBF3 complex may fonn a basis upon which additional
microtubule-binding proteins can assemble, with CDEII and. to a lesser extent, CDEI
acting as stabilizers of the complex (Hyman and Sorger, 1995).
1.7. Thesis Rationale
Although the data to date suggest the importance of hEBP2 in EBNAl-mediated partitioning, the role of hEBP2 in this process has not been directly tested. The fiuictional contribution of hEBP2 in EBV partitionhg could be tested through the
---- S-L -- - A- A- a* - - cnation of an EBV-based segregation system in a background lacking endogenous
hEBP2. An ideal system would depend on EBNAl for segregation but not for
replication. Furthemore, the system should allow the addition, removal or mutation of
hEBP2, without afkcthg the viability of the host in which the system resides. Lastly,
such a system should be easy to manipulate. Al1 these criteria cm potentially be met by
developing the segregation system in budding yeast. Budding yeast is an attractive
choice of organism because it expresses its own EBP2, and therefore would not be
dependent on hEBP2 for survival, and because it is easy to manipulate and handle. More
importantly, it is an organism where plasrnids can be propagated through replication and
segregation elements that are distinct and easily separable.
The purpose of my project is to develop an EBV-based plasmid segregation system in
budding yeast and to use this system to assess the role of hEBP2 in EBV segregation. I
have done this by inserting the EBV segregation element, FR, into an unstable yeast
replicating plasmid. 1 then determincd whetha this FR-containing plasmid can be stably
maintained in yeast by EBNAl and whether hEBP2 is required for this process. In
addition to understanding the function of hEBP2 in EBV segregation, development of
such a system will allow assessrnent of EBNAl segregation hction, independent from
its role in EBV replication. Moreover, the system will help elucidate the mechanism of
EBV segregation, including the identification of any additional cellular proteins required
for this process to occur. Understanding the mechankm of EBV segregation may allow
development of strategies to disrupt the persisteme of EBV genomes in latently-infected cells, thus b~gingus closer to preventing malignancies caused by EBV. n.1. Yeast Strains
The yeast strains KY320 (WTa leu2-PETS6 ura3-52 trpl-Al lys2-80lam ade2-
IOloc his3-A200 GA+)and W303-1A (MATa leu2-3,112 ura3-1 hpl-I ade2-l hisi-1 l
can1-100) were uscd for the plasmid los assays (Chen and Struhi, 1988; Bailis and
Rothstein, 1 990). Yeast strains KY32O.hEBP2 (MATa ura3-52 trpl-Al lys2-80 1am ade2-
1Oloc Lis34200 GAL+)and W303.hEBP2 (MA Ta um3-1 ml-l ade2-I hk3-11 can 1-
100), which contain an integrated copy of the hEBP2 gme, were also used. The
construction of the hEBP2-intepted strains is discussed later (Section iI.5).
The yeast strain Y 190 (MT&leut-3,112 uru3-52 ml-901 his3-A200 adel-101
GAL4Agal80A URA3 GA-lad LYS GAL-HIS3Cyh?, which contains integrated HIS3
and LAC2 reporter genes under the control of GALA binding sites, was used for the yeast
two-hybnd assays (Harper et al., 1993).
II.2. Construction of the Plasmid LOSSAssay Constrocb
The stability of three plasmids, pRS3 14, YRp7 and YRp7FR, was determined through
plasmid loss assays. pRS3 14, which sen& as a positive control, is a yeast centromenc
plasmid that contains the ARSH4 and C'elements and the TRW selection marker
(Sikorski end Hieter, 1989). YRp7, which was the negative contml, is a yeast replicating
plasmid that contains the ARSI element and as well as the TRPl selection marker
(Stinchcornb et al., 1979). YRp7FR, which is essentially YRp7 containing the EBV segregation element, FR, sdas the segregation test plasmid. To constnict YRpWR, YRp7 was lineacized with BamH 1and Msc 1. pGEMoriP (Frappier and
--a.------O'Domell, 1Wa) was also digested with the same enzymes to yield the FR element of
ofl The linearized YRp7 plamid and the FR element were rnixed at a ratio of 1:5 ami
ligated ovemight at 16OC with T4 DNA ligase. The ligation mixture was used to
transfomi the E. coli strain DHSa and plasmids isolated hmthe transfonnants were
screened for positive clones through restriction enzyme digestions. The appropriate
YRp7FR clone was sequenced by the sequencing facility at the Hospital for Sick
Chi ldren .
The p4 16MET. EBNA 1 plasmid, which contains EBNA 1 inserted downstrearn to the
MET25 promoter nsiding in the p416MET25 vector, was used for expression of EBNAl
in the plasmid loss assays. p4 16MET25 is a low copy number expression plasmid that
contains the CEN6/ARSHQelernent and the UR13 selection marker (Mumberg et al.,
1994). To construct the EBNAlsxpressing vector, p416MET25 was digested with fia 1
and BamH 1. Similady, pEBNAl (Shire et al., 1999) was also digested with Xba 1 and
BamH 1 to give the EBNAl gene lacking most of the Gly-Ala repeat. In the context of this work, this EBNAl will be referred to as full-length. The linearized plasmid and the
EBNAl gene were ligated and the resulting constructs were screened as described for
YRp7FR.
The p416METZS.EBNAA325-376plasmid, which expresses the EBNAl mutant lacking amino acids 325 to 376 as well as most of the Gly-Ala repeat, was also constnicted for the plasmid los assays. This involved digestion of p416MET25 with
Sma 1, followed by the dephosphorylatioa of the linearized plddwith CIP (calf intestinal alkaline phosphatase). The EBNAA325-376gene was excised hmthe -.. -
pASZ.EA325-376 (Shire et al., 1999) plasmid by digestion with Nde 1 and BamH 1. The
.~ .<-- u J7L - - - - 5' overhangs of the EBNAl fiagrnent wmfilled in with DNA Polymerase 1 Klenow
hgment. The linear plasmid and the EBNAl insert were ligated and the ligation
products were screened for the appropriate construct as described above.
Expression of hEBP2 in the assays was achieved hughthe plasmid
p425PGK.hEBP2, which is the pR425RGK plasrnid containing the IiEBP2 gene
incorporated downstnam of the PGK promoter. pR42S/PGK itself is a hi&-copy
number, 2-micron expression vector that contains the LEU2 marker for selection (Marcus
et al., 1995). The hEBP2cxpressing plasmid was made by digesting the pR425lPGK
plasrnid with Bgl II, which cuts the plasmid between the PGK promoter and the PGK
teminator sequmces. The sticky ends of the digested plasmid were then filled-in with
Klenow and treated with CIP. The hEBP2 gene was obtaineci from pVLEBP2 (Shin et
al., 1999) by PCR (polymerase chah reaction) amplification with VENT DNA
Polymerase using the following prhers:
ebp2 N 5' ACC TCT CGA GAT GGA CAC TCC CCC 3'
ebp2 C 5' AAT GCT CGA GCT ATT TOT GTT CTG TTC T 3'
The amplification reaction was initiated by heating the PCR mixture at 94OC for 3
minutes, followed by 25 rounds of heating at 94OC for 1 minute (to denature the DNA
template), cooling to 41°C for 30 seconds (to meal primers to the template) and heating
back up to 72OC for I minute (for elongation of the prirners). The PCR product was
subjected to phosphorylation by T4 polynucleotide hase. Ligation of linear pR42YPGK with the hEBP2 PCR product was performed and the appropriate clone was identified as
--PL-& = -Lr L--- -& --- - for the other constmcts.
11.3. Expression of EBNAl and hEBP2 in Budding Yeast
Once the EBNAI- and hEBP2- expressing vectors were created, expression of the
proteins hmthese vectors was confirmecl in both KY320 and W303-1A yeast strains.
This was accomplished by transformhg the strains with the appropriate plasmid using a
modification of the lithium acetate (LiAc) method developed by Gietz and Schiestl
(1995) and perfonning Western analysis on lysates prepared hmthe transfomants.
For LiAc h.ansfonnation, yeast strains KY320 or W303-lk were grown in YPD
complete medium (Sherman et al., 1979) until a density of 1-2 x 1O' celldml was
reacheâ. Cells were pelleted, washeà with âH20,then with 0.1 M LiAc and resuspended
in 500 pl LiAc. 50 pl of the Li& culture was used for each transformation. Cells in the
50 pl culture were pelleted and the followlng components were added in order to the
pellet: 240 pl of 50% PEG 3500.36 pl 1.0 M LiAc, 25 pl of 2 mglml denatured hemng
sperm DNA and 50 pl of dH20containhg 1 pg of plasmid (EBNAl-or hEBP2-
expressing plasmid). The mixture was vortexed to resuspend the pellet and then
incubated at 30°C for 30 minutes, followed by heat shock at 42OC for 23 minutes. Mer
heat shock, the cells were pelleta resuspended in dH20and plated on synthetic
complete (SC) plates (Sherman et al., 1979) lacking the appropriate amino acid to allow
selection for the transfomants. In the case of p416MET.EBNAl transfomants, uracil
was omitted hmthe medium (SC-Ura), while in the case of p42SPGKhEBP2
transformants, leucine was omitted fitom the medium (SC-Leu). For Western blotting, transformants were grown in the appropriate selective medium ------to a ce11 density of 5 x 1o6 - 1 x 10' cellshnl. Appmximately 2 x 10' cells were harvested . and resuspended in dH20containing 1 mM of phenylmethylsulfonyl fluonde (PMSF), 1 mM benzamidine and SDS loading buffet. The ce11 lysate was boiled for 5 minutes,
ciarified by centrifugation and analyzed by SDS-PAGE(SDS polyacrylamide gel
electrophoresis). The separated proteins were transferred to a nitrocellulose filter ushg
the BioRad wet transfer apparatus (according to manufacturer's instructions). The filter
was blocked for one hour at room temperature in 5% mik powder in PBS (phosphate
buffereâ saline). Blocking was followed by a one hour incubation with the pnmary
antibody (1 5,000 dilution of rabbit polyclonal anti-EBNAl #67 or anti-hEBP2 in
blocking buffer) and three 5-minutes washes with PBS. The secondary antibody (1 5,000
dilution in blocking buffer of goat anti-rabbit conjugated to horseradish peroxidase) was
then placed on the filter for one hour, followeâ by another thm 5-minutes washes in
PBS. Finally, 4 ml of the enhanced cherniluminescence solutions (ECL by NEN Inc.) was
added to the filter for 1 minute and the filter was exposed on Kodak X-Omat LS for
various lengths of tirne to obtain a signal.
The expression of EBNAA325-376 was aiso determined in the manner detailed above,
but only in KY320. Furthemore, to ensure that both EBNAl and hEBP2 were able io
express at the sarne tirne in seains also containing the segregation test plasmid. KY320
and W3O3-1 A were transfomecl with the thme comsponding piasmids, trans formants
wen selected for on SC-Ura,Leu,Trp and Western blotting was performed. IL4. Construction of the Yemt Two-Hybrid Plasmiâs
-- -.--- 2 -- - = L -- In the. yeast two-hybnd assays, the expression of full length EBNAl and EBNAl amino acids 452 to 641, as fluions to the GALA DNA binding domain, was achieved thmugh the plasmids pAS2.EBNA 1 and pASZ.E452-64 1, respectively. Both pAS2.EBNA 1 and pAS2.E452-64 1 are 2-micron plasmids that contain a TRPI selection marker and express theu respective proteins fimm the ADH promoter. These plasmids were kindly provided by K. Shire and their construction has been previously described
(Shire et al., 1999).
The plasmid pACT63 wu used to express hEBP2 amino acids 2 1 to 306 fûsed to the
GUactivation domain hmthe ADH promoter. pACT63 is a cDNA library construct that was isolated through interaction with EBNAl in a two-hybnd screen (Shire et al.,
1999) and contains the 2-micron ongin and the LEU2 marker.
The pACT.yEBP2 pladd expresses yEBP2 as a fusion to the GAL4 activation domain hmthe ADH promoter located in the pACTIl vector. pACTïI is a 2-micron plasmid that contains the LEU2 marker (Li et al., 1994). To construct pACT.yEBP2, the yEBPZ gene was PCR amplified hmpETlSbyEBP2 (ptovided by K. Shire) with the following primers containing the Sma 1 and fia 1 restriction enzyme sites:
N yebp2 (SmaI) 5' GGG ATC CCG GGG ATG GCT AAA GGT TTC AAG TTG 3'
C yebp2 (MOI)5' ACG CGG ATC CTC GAG TTA GAA TCT TCT GGC ACG TC 3'
The PCR protocol used was as previously described for hEBP2, with the exception of the annealing temperature and the elongation tirne, which wm53OC and 1.5 minutes for yEBP2, respectively. The PCR product was digested with Sma 1 and X?zo I and the
---- 2 - - - - A------A --- --.. -- - resdting fragments wmligated to pACTII, which had also been digested by Sma 1 and
X71o 1. The ligation mixture was screened for the appropriate construct as described for
the plasmid loss assay constnicts. Expression of the yEBP2 fusion protein hmthe
appropriate construct was conhed by Western analysis with an antibody raised against
a peptide contained within the y- gene.
II.5. Construction of hEBPt1ntegrated Yeast Smins
KY320.hEBP2 and W303.hEBP2 were consûucted using the plasmid pRS305.hEBP2,
which is a pRS305 vector containing a copy of the hEBP2 gene that is expressed fiom the
PGK promoter. pRS305 is a yeast integrahg plasmid containing the LEU2 marker
(Sikorski and Hieter, 1989). To make pRS305 .hEBP2, p425PGK.hEBP2 was digested
with Hind III, which yielded a fhgment containing the PGK promoter, hEBP2 gene and
PGK terminator sequmce. The hgment was ligated to pRS305, which has been
digested with Hind III and dephosphorylated. The nsulting plasmid contained two
EcoR V sites, one located within the multiple cloning site (MCS)and the other in the
LEU2 gene. To eliminate the &OR V site within the MCS (so that LEU2 contains a
unique restriction enzyme site), the plasmid was partially digested with EcoR V. Plasmid
molecules that had been digested only once (at either of the EcoRV sites) was isolated by
agame gel electrophoresis and digesteci to completion with BamH 1 (also located in the
MCS). The DNA Aagrnents were separated by agarose gel electrophoresis and the
appropriate band was excised. The BamH YEcoR V ends of the plasmid were filled-in
with Klenow and re-ligated, giving pRS305.hEBPZ. This plasmid was linearized with EcoR V withh the LEU2 gene, then used to transfomi KY320 and W303-1A by the LiAc
--- - =- - --A------method for site-dincted integration at the LEUL locus (Rothstein, 1991). eu'
transfonnants that grew on SC-Leu plates were pickeâ, grown ovemight in YPD medium
and subjected to Western blot analysis to conhhEBP2 expression.
IL6. Plasmid Loss Assays
Plasmid loss assays were conducted by transformation of KY320, W3O3- 1 A,
KY320.hEBPZ or W303.hEBP2 to Trp protoûophy with a segregation test plasmid
(pRS314,YRp7 or YRp7FR) and to Ura prototrophy with a p426MET25 vector
expressing or not expressing EBNAI . When required, KY320 and W 303-1 A were also transformed to Leu prototrophy with a pR42WGK vector expressing or not expressing
hEBP2. The desireci transfomiants were selected by plating on SC-Trp, üra or SC-Trp,
Ura, Leu (when a pR425/PGK vector was present) plates. Single colonies were picked
hmthe plates and grown in 3 ml of the same selective media at 30°C9until mid-log
phase (ODm of 0.4-0.8; OD = optical density). The cultures were then diluted to 1 x 1O'
celldml (ODm = 0.01) in 6 ml of non-selective medium (SC-Ura or SC-Ura,Leu when
pR425RGK is present) with respect to the segregation test plasmid. AAer pwth in this
medium for 1 1 generations (except where indicated), the OD of the cultures was
determined using a spectrometer and ten-fold seriai dilutions of the cultures were
generated9where the least diluted culture consisteà of 1 x 106 celldml (ODm = 0.1). 5 pi
of each dilution (a total of 4 dilutions) were then spotteci ont0 selective (SC-Trp,Ura or
SC-Trp,Ura, Leu) and non-selective (SC-Uraor SC-Ura,Leu) plates with respect to the
segregation plamid. Qualitative analysis of the relative numba of colonies on selective 7x
and non-selective plates was carrieâ out to determine the stability of the various
--.Lx& - - - - segregation plasmiâs.
To quanti@ the percentage of cells that retained the segregation plasmids, equal
volumes of the diluted cultures were plated on both selective and non-selective plates,
such that about 350-450 colonies were obtained on the non-selective plates. Colonies that
appeared on the plates were counted and the ratio of the number of cells on selective
versus non-selective plates was used to rneasure plasmid stability.
11.7. Yeast Two-Hybrid Assays
For the yeast two-hybrid assays (Fields and Song, 1989; Chien et al., 1991), yeast
strain Y 190 was traasformed by the LiAc method to Trp and Leu prototrophy with
PAS 1.EBNAl and pACT63 (positive control), PAS1 .E452-64 1 and pACT63 (negative
control) or pASLEBNA1 and pACT.yEBP2. Positive transformants were identi fied by
plating the transfomation mixtures on selective plates (SC-Trp,Leu). Single colonies
hmeach plate were picked and used to inoculate 5 ml of the same selective media.
Cultures were grown in a 30°C shaker until saturation and then diluted into 5 ml of SC-
Trp, Leu, His medium at various concentrations, such that after overnight growth, the
cultures contained about 5.0 x 106to 1.O x 10' cells per ml (ODm of 0.5 to 1.O). The ce11
density of the ovemight cultures was deteaed and the concentration of each culture
was adjusted to 5 x 106cells paml (ODm = 0.5). The cultures at ODm of 0.5 were
Merdiluted three times by 10-fold serial dilutions and 5 pl of each dilution (four in
total) were spotted on SC-Trp, Leu and SC-Trp,Leu, His plates that contained 50 mM 3-
aminotriazol (AT). Interaction between GAL4 DNA binding domain and GALA -- -
activation domain fusion proteins was inâicateâ by the activation of the HIS3 gene, which
--A ------A - -- - resulted in the growth of the cultures on both types of plates. Lack of interaction between
the fbions was indicated by the pwthof the cultures on SC-Trp, Leu plates only. ïII.1. Expression of EBNAl and hEBP2 in Budding Yeast
Before using the EBNAl- and hEBP2- containhg plasmids (shown in Figure 4B and
4A, respectively) in the plasmid loss assays, 1 asked whether the proteins were properly
expressed hmthe plasmids in budàing yeast. To this end, KY320 and W303-1A yeast
süains were transformed with EBNAI- or hEBP2- containhg expression plasmids and
the lysates of the resulting transfomiants were analyzed by Westem blot, using the
appropriate antibody for each protein. Western analysis indicated that both proteins were
expressed in both strains (Figure 5). Though both EBNAl and hEBP2 were stable and
full-length (running at appmximately 70 kD and 42 kD, respectively) in KY320,
significant protein degradation was observed in W3O3-1 A. The ratio of hill-length protein
versus degraded protein was obsmed to be higher for EBNAl than for hEBP2 in this
strain. However, detectable levels of the full-length proteins were still observed.
In some instances, expression of foreign proteins in yeast can compromise the
viability of the organism. However, expression of neither EBNAl nor hEBP2 had any
obvious affects on the viability of the yeast. When observed under a light microscope, cells expressing either protein were UIdistinguishable hmcells transformed with vector alone. Cells at dl stages of the ce11 cycle @oth budded and unbudded of various sizes) were visible, indicating that they were growing uniformly and not blocked at any particular stage.
The plasmid loss assays require expression of both EBNAl and hEBP2 as well as the presence of the FR-containkg plasmid. Therefore, it tasnecessary to ensure that hEBP2 PGK terrninator
Figure 4. Expression plasmids used in the plasmid loss assay. A. p425PGK.hEBP2, which contains the 2 micron ongin and the LEU2 rnarker, expressed hEBP2 hma PGK prornoter. In cases when hEBP2 expression was not reqquired, the same vector lacking the EBP2 gene was used. B and C. p416MET.EBNAI and p416MET.EA325-376, both of which contain the CEN6IARSH4 eiement and the (IRA3 marker. was used to express EBNAl and EBNAA325-376 fiorn the MET25 prornoter, respectively. When EBNAl or EBNAA325-376 expression was not needed, the empty vecton lacking the gene was used Figure 5. EBNAl and hEBP2 expression in budding yeast. A Western blots of KY320 and W303-1A lysates expressing no EBNAl (-), EBNA1 alone (EBNAl ) or EBNAl and hEBP2 in the presence of the FR plasmid (EBNAI *). Blots were probed with anti-EBNAl antibody. The positions of hill-length EBNAl (arrow) and EBNAl degrdation products (*a) are indicated. BeWestern blots of KY320 and W303-1A lysatts atprtssing no hEBP2 (-), hEBP2 alone (hEBP2) or hEBP2 and EBNAl in the presence of the FR plasmid (hEBP2+). Blots were pmbed with anti-hEBP2 antibody. The positions of Ml- length hEBP2 (arrow) and hEBP2 degradation products (**) are indicated. EBNAl and hEBP2 were expresseci in yeast containhg the EBNAl and hEBP2
A=- -- - -A - -- expression plasmids and the FR plaJmid. To do this, both KY320 and W303-1 A yeast
strains were transfonned with al1 three plasmids and Westerns were perfomed on the ce11
lysates. Both EBNAl and hEBP2 were expressed in this condition, in both strains
(results for W3O3-1A are shown in Figure 5, lanes labeied EBNA 1* and hEBP2*). The
level of expression for both proteins in this case was slightly different from those
observed when EBNAl and hEBP2 were expressed individually (W303-1 A column,
cornparhg EBNAl with EBNAI * and hEBP2 with hEBPZ*; each W303-1A lane in a cornparison contains an equal amount of lysate).
1ïI.2. EBNAl Does Not Mediate FR Plasmid Segrqation in Yeast
In human cells, EBNAl is involved in both the replication and segregation of EBV
episomes and oriP-containhg plasmids. As a result, studies aimed at understanding the
segregation function of EBNAl are often complicated by its role in replication. For
instance, loss of an oriP plasmid in long-term plasmid maintenance assays which are
done to assess the segregation function of EBNAl is always accompanied by transient
plasmid replication assays to ensure that the loss is not due to a failure in EBNAl 's
ability to replicate the plasmid. In an attempt to separate the segregation activity of
EBNAl hmits replication activity, I asked whether EBNAl allowed long-term
maintenance of plasmids containhg the EBV segregation element (FR) in budding yeast.
To test this the segregation test plasmid YRp7FR was created, which contains the yeast
replication origin, ARS, and the EBV segregation element, FR (see figure 6). This plasmid replicates using the yeast replication machinery but lacks the yeast segregation ARS
Figure 6. Segregation test plasrnids. Stability of three plasmids, each containhg an ARS element and a TRPl selectable marker, was tested in the yeast plasmid loss assay. A The pRS3 14 plasmid, which contains the yeast CEN element, was used as a positive control. B. The YRp7 plasmid, which lacks a segregation element, serveâ as a negative contiol. C. The experimental plasmid, YRp7FR, contains the EBV segregation element, FR element (CEN). To test whether the plamid could segrcgate ushg the FR element,
A-.------YRp7FR and the EBNAl-expressing pldd(p416MET.EBNAl) were used to transform the yeast strain KY320 to Trp and Ura prototrophy, respectively, and the stability of YRp7FR was determined using the plasmid loss assay (as described in
Materials and Methods). KY320 was used because EBNAl was observed to be more stable in this strain as compared to the strain W303-IA. KY320 transfomed with
YRp7FR, YRp7 (lacks CEN and FR; negative control) or pRS314 (contains ARS and
CEN; positive control); (Figure 6) and p416MET25 were also included in the assays as controls. Note that in these cases where no EBNAl expression was required, the p416MET25 plasmid, which is the backbone plasrnid used for EBNAl expression and does not express EBNAl itself, was used. This protocol allowed growth of yeast cultures in similar conditions (in SC-Ura,Trp when selecting and SC-Urawhen not selecting for the segregation plasmid).
In the plasmid loss assay, the various KY320 transformants were grown in non- selective media for 48 hours (appmximately 8 divisions) before dilutions were spotted ont0 selective and non-selective plates. A typical result hmsuch an assay is shown in
Figure 7. As expecteâ, the CEN-containing plasmid was efficiently maintained during growth in non-selective media, as indicated hmthe similar number of colonies on both selective and non-selective plates. On the other hand, YRp7, which contains no segregation element, was lost hmthe culture at a high rate. This is indicated by the low ratio of colonies on the selective plate versus the non-selective plate. Note that the difference in the stability of pRS3 14 and YRp7 is primarily due to the presence and absence of the CEN element and not due to the difference in the ARS elements (ARSH4 Selection for test dasrnid CEN -
FR - FR + CEN -
Figure 7. EBNAl cannot support stable segrqation of YRp7FR in budding yeast. Stability of YRp7FR was tested in the yesst KY320, both in the pnsence and absence of EBNAI. Yeast were transformed with a segregation plasmid (shown in Figure
5) a.a üW3 expression plasmicl arprcJsmg(+) or not expressing (0) EBNAI . A&r growth without selection for the segngation plasmids for 8 genaations, IO-fold dilutions of the cultures were plated on SC-Ura (no selection for test plasmid) or SC-Ura,Trp (selection for test plasmid) plates. --
and ARS 1, respectively). Both ARSH4 and ARS 1 are similar in their ability to stabiüze a
----L ------LA-- - * ------2 ------yeast plasmid (Hieter et ai., 1985). The FR-containhg plasmid, YRp7FR, was also
unstable and exhibiteci a plasmid loss profile that was identical to YRp7 in the absence
and presence of EBNAl expression. These resuits indicate that EBNAl was unable to
mediate the stable partitionhg of the FR-containing plasmid in yeast.
m.3. Human EBP2 is Required for EBNAl-Mediated Segregation in Yeast
Since EBNA1 was unable to support the stable segregation of the FR-containing
plasmid, 1 next asked whether hEBP2 was requireâ to rescue the segregation function of
EBNAl in yeast. A requirement for hEBP2 was tested because previoüs studies
conducted in our lab have shown that hEBP2 associates with EBNAl and that thete is a
comlation between the ability of EBNAl to bhd hEBP2 and the ability to mediate
segregation of FR-containhg plasmids in human cells (Shire et al., 1999). To test the
possible contribution of hEBP2 to EBNAl-mediated segregation in yeast, the plasmid
loss assay was conducted in KY320, where the stability of YRp7FR was detemiined in
the presence of both EBNAl and hEBP2. The stability of both pRS3 14 and YRp7 was
also monitored, as positive and negative controls. In thesecontrols, expression of
EBNAl and hEBP2 hmthe UR43 and LEU2 expression vectors was not required.
Therefore, yeast containhg these segregation plasmids were transforrned with UR43 and
LEU2 vectors lacking the EBNAl and hEBP2 gene, respectively, in order to enable
growth of al1 transformants under identical conditions (ie. grow in selective SC-Ura, Trp,
Leu media). The various transfonnants were grom in SC-Ura, Leu media, which is
selective for the expression vectors but non-selective for the segregation test plasmids, for 11 generations (approximately 66-70 hours) and then plateâ to assess segregation
--- --LE --' ------L - - plamid stability.
As seen in Figure 8 (ieft panel), the FR-plasmid was stably maintained when both
EBNAl and hEBP2 were expressed. Cultures containing the FR-plasmid exhibited
growth on selective and non-selective plates that was sirnilar to cultures transfomed with
the CEN-containhg plasmid. Once again, YRp7 was found to be unstable. 1 quantified
the percentage of cells that maintained the various segregation plasmids (Table 1, KY320
column); YRp7FR was slightly more stable than pRS3 14 in the presence of EBNAl and
hEBP2 and approximately 50-fold more stable than YRp7 over the period of 11
generations. Neither YRp7FR nor pRS3 14 demonstratecl 100% retention in KY320,
which is expected. The stable pRS3 14 plasmid itself has a loss rate of 2.6% per
generation in wild-type cells (Merker and Klein, 1998), which can vary slightly
depending on yeast strain and conditions used. This loss of a CEN-containing plasmid,
which does not ofien occur for CEN-containing chromosomes, is likely due to the
decreased efficiency of CEN fiinction on mal1 pldds (Hieter et al., 1985).
In order to individually assess the importance of FR, EBNAl and hEBP2 in the
observed stability of YRp7FR, plasmid loss assays were perfomed where each of these
components was separately omitted. When plasmids that express EBNAl or hEBP2 were
replaced with empty vectors, YRp7FR was not stably maintained and was lost at the
same rate as negative control plasmid (Figure 8, KY320 panel; Table 1, KY320 column).
When both EBNAl and hEBP2 were expmsed but YRp7, which lacks FR, was used as
the segregation plasmid, this plasmid was also unstable (Figure 8). 1 conclude that
efficient EBV-based segregation in KY320 requires EBNAI, hEBP2 and the FR element. Selection test plam
-
+
.Figure 8. hEBP2 is requircd for EBNAl-mediated FR-plasmid segregation in budding yeast. Segregation of YRp7FR was evaluated in the pnsence of bth EBNAl and hEBP2. Plasmid Ioss assays wae conducted in both KY320 and W303-1A yeast strains. Yeast were transfonnd with a segregation plasmid, a URAJ plisrnid that contains (+) or lacks (9) the EBNAl genc, and a MU2 plasmid that contains (+) or lacks (-) the hEBP2 gene. Mer growth for 1 1 gencrations without selection for the test plasmid, dilutions of the cultures were plated on SC-Un,Leu (no selection for test plasrnid) or SC-Ura, Leu, Trp (selection for the test plasmid) plates Table 1. Stability of Segregation Plasmids in the Presence and Absence of EBNAl and
- :---4-- & -- =- - - -- MBP2 -- - - A - -- -
Segregation EBNA 1 hEBP2 Plasmid Expression ~x~ression~ pu314 ., 9 pu314 - + YRp7FR + + YRp7FR + - YRp7FR - + YRp7 + + YRp7FR LUZS-376 + YRp7 9 - YRD~ - +
'Values shown for plasmid stability were determined &er 1 1 generations of growth in non-selective media as described in Materials and Methods and refer to plasmid retention. Values shown are an average of 1-1 1 experimmts (see Appendix). b~nKY320 and W303-l A, hEBP2 was expresseci hmthe p425PGK.hEBP2 plasmid, while in KY320.hEBP2 and W303.hEBP2, hEBP2 was expnssed fiom the integrated gene. 'non applicable. At this point, al1 plasmid los~assays bad been performed in the yeast ~trainKY320.
= .-L------To ensure that the obsmed stability of YRp7FR in the presence of EBNAl and hEBP2
was not strain-specific, 1repeated the plasmid loss experiments in the yeast strain W303-
1A, which has a growth rate that is almost tbree times as fast as KY320. The results of
the plasmid loss assays performed over 11 generations (approximately 24 hours) in
W303-1A are shown in Figure 8 (right panel) and Table 1 (W303-1A colurnn). Plasmid
loss results obtained with W303-1A were very similar to those obtained fmm KY320. As
in KY320, YRpFR was rendered stable when both EBNAl and hEBP2 were expressed
but was unstable when EBNAl, hEBP2 or FR was removed. Stability of both the
CEN/ARS plasmid @RS3 14), and YRp7FR (with EBNA 1hEBP2 expression) was
slightly higher in W303-1A than KY320. The moderate level of EBNAl and EBP2 degradation observed in W3O3-1 A did not inhibit the ability of these proteins to mediate
FR plasmid segregation.
111.4. EBNAl-Mediated Segregadon in Yeast Requires EBNAl Binding to hEBP2
To gain merinsight into the mechanism by which hEBP2 fiinctions in EBNAl- mediated segregation, the ability of the EBNAA325-376 mutant to mediate FR plasmid segregation in yeast was exarnined. EBNAA325-376supports replication of an oriP plasmid in human cells but is unable to support both hEBP2 binding and plasmid segregation in human cells. The EBNAA325-376 gene was cloned into p425MET25 and expression of the mutant from the resulting plasmid (Figun 4C) was confïmed through western blot. A representative blot is shown in Figure 9, where the expression of
EBNAA325-376 was compared to that of EBNAl in the KY320 strain. As shown in ihis Figure 9. EBNAA325-3 76 expression in budding yeast. Western analysis of lysates of 2 x IO7 KY320 cells, transformed with the p42SMET.EA325-376 plasmid (EBNAA325-376). was conducted to check for EBNAA325-376 expression using an antibody against EBNA1. Equal amounts of KY320 lysates expressing no EBNAl (-) or wild-type EBNAl(EBNA1) were also included for cornparison. figure, EBNAA325-376 is expressed hmthe MET25 promoter at a level similar to
- - *L - .. - -- - - EBNAI. Similar results were obtained in the W303-1A yeast strain (data not show),
though as previously noted, degradation products were evident in this strain.
The segregational ability of EBNAA325-376 in yeast was tested in the plasmid loss
assay in both KY320 and W3O3-1 A and the results were compared to those obtained with
wild-type EBNA1. Unlike wild-type EBNAl ,EBNAA325-376 was unable to mediate
stable segregation of the FR plasmid in either strain (Figure 10 and Table 1-KY320 and
W3O3- 1A columns). Afler 11 divisions of pwthin non-selective media, cultures
containing the FR plasmid, hEBP2 and EBNAA325-376 show a considerable decrease in
growth on selective plates as compared to non-selactive plates and this plasmid loss
profile is similar to negative control cultures containing FR -nd hEBP2 but no EBNAI.
These results indicate that the hEBP2-binding region of EBNAl is required for stable FR
plasmid segregation in yeast. The behavior of EBNAA325-376 in EBV segregation in
yeast is thus sirnilar to that in human cells, suggesting that a similar mechanism is used
by EBNAl to govern plasmid segregation in both organisms.
111.5. ElBNAl Does Not Interact with the Yeast EBP2 Protein
The inability of EBNAl to support FR plasmid segregation in yeast in the absence of hEBP2 indicated that the yeast homolog of hEBP2, which is tenned yEBP2, could not substitute for hEBP2 in EBNAl-mediated segregation. Since my results suggested that an interaction between EBNAl and hEBP2 is reqwred for segregation, 1wanted to detemine if the failun of yEBP2 to support EBNAl-mediated segregation was due to its hability to bind EBNAl. To investigate whekinteraction between EBNAl and Selection for test plrissnid EBNAl
Figure 10. Requirement of the hEBP2 binding region of EBNAl in FR plasmid segngation in yeast. The stability of YRp7FR was determined in the presence of either wild-type EBNAl (+) or EBNAA325-376 (A325-376). a mutant which lacks the WBP2 binding region, or in the absence of EBNAI(-). in each case, hEBP2 was also exprcssed. Plasmid loss assays were conducted as described in Fig. 7, in both KY320 (left panel) and W3O3- 1A (nght panel) strains. -*
yEBP2 occurs, the yeast two-hybrid assay was used. Activation of the HIS3 and LAC2
- -7 A-&,. ------reporter genes (uada the control of GAIA sites) was assayod when EBNA1, fiised to the
GAIA DNA binding domain, and yEBP2, fiised to the GALA activation domain, were CO-
expressed in yeast strain Y190. EBNAl or EBNA452-641 (which contains EBNAl
residues 452 to 641 and lacks the hEBP2-binding domain) fusad to the GALA DNA
binding domain and hEBP2 fused to the GAIA activation domain were also CO-expressed
in Y 190 as positive and negative controls, respectively. The results for the two-hybnd
assays using the HIS3 reporter gene are shown in Figure 11. As expected, CO-expression
of EBNAl and hEBP2 fusion protek, which do interact, always lead to the activation of
the HIS3 gene, as indicated by the growth of cells expressing these proteins on plates
lacking histidine (SC-Trp,Leu, His containhg AT). Co-expression of EBNA45 1-641
and hEBP2 fusion proteins, which do not interact, did not activate the HIS3 gene, as seen
hmthe lack of growth of cells containing these proteins on plates lacking histidine. Co-
expression of EBNAl and yEBP2 fusion pmteins also did not activate the HïS3 gene and
exhibited a pwthpattern that was similar to the negative control. Similar results were
observed for each pair of fusion proteins when the LRCZ reporter gene was used in the
two-hybrid assays (results not shown). Since expression of the yEBP2 fision protein was
confirmed by Western blot (data not shown), the results indicate that yEBP2 either does
not interact or interacts very weakly (below the level of detection of the assay) with
EBNA l in budding yeast.
IE.6. FR Plasmid Segregation in hEBP2-Integrated Yeast Strains by EBNAl
So far, my plasmid loss assays wmconducted in strains carrying three plasmids, a Figure 11. Interaction of EBNAl with EBP2. EBP2 hmhumans (hEBP2) and yeast (yEBP2) were tested for binding to EBNAl in a ycast two-hybrid assay. A positive interaction was indicated by the activation of the HIS3 reporter gene in yeast strain Y 190. As a negative control, hEBP2 was tested for binding to EBNA452-641, which lacks the hEBP2 binding domain. Y 190 was transformed to Trp and Leu prototrophy with a plamid cxpmsing cither EBNAl or EBNA452-641 hised to the GAIA DNA binding domain md a plasmid expressing EBP2 hsed to the GAIA activation domain. Tm-fold serial dilutions of the cultures were grown on SC-Trp,Leu (lefi panel) and SC-Trp, Leu, His plus 50 mM AT (right panel). less than ideai situation. In an attempt to ducethis selective pressure on the
------transformants, I integrated the hEBP2 gene into KY320 and W3O3- 1A by homologous
tecombination at the LEU2 locus located on chromosome II, resulting in strains that
constitutively expressed hEBP2 hmthe PGK promoter. Expression of hEBP2 in these
strains was confimed by western analysis (Figure 12). The hEBP2 protein was less
stable when expressed hmthe integrated gene as compared to that expressed From a
plasmid in both yeast strains (Figure 12 and Figure SB). Despite this decrease in stability,
the ratio of full-length hEBP2 versus degraded hEBP2 was still high in the hEBP2-
integrated KY320 strain. However, in the hEBP2-integrated W303-lA strain, htll-length
hEBP2 was barely detectable as compareâ to degradeû IiEBP2.
To ensure that the observed stability of the FR plasmid in the ptesence of EBNAl and
hEBP2 was not compromiseci in these integrated strains, the fate of the plasmid in the
presence of the two proteins was deterrnined after 11 generations of growth in these
strains. The tirne taken to undergo 11 generations in the plasmid Ioss assays by
KY320.hEBP2 and W303.hEBP2 (transformeci with TRPl and UR43 plasmids) was not
significantly lower fkom that of KY320 and W303-1A (transfomed with TWI,
UR43 and LEU2 plasmids), respectively. The results hmthese assays are outlined in
Figure 13 and Table 1 (KY320.hEBP2 and W303.hEBP2 colurnns). As expected, the FR plasmid was found to be stable in these integrated strains when EBNAl was expressed, but not when either EBNAl or the FR element was omitted. Furthermore, expression of
EBNAA325-376 did not confer stability upon the FR plamid, as in the non-integrated strains. In KY32O. hEBP2, the stability of the FR plasmid with EBNAl expression was Figure 12. Expression of hEBP2 in hEBP2-integrated yeast strains. A KY320 (-) and KY320.hEBP2 (hEBP2) lysates were subjected to gel electmphoresis and analyzed by Western blot with anti-hEBP2 antibody to confimi expression hmthe integrated gene. B. W3O3- 1 A (-) and W303.hEBP2 (hEBP2) lysates were analyzed by western blot using the sarne antibody as in A. In A and B, 2 x I07cellswere loaded in each lane. The mw and ** represent full-lcngth hEBP2 and degraâation products, nspectively. Selec 'on for test Df asmd
Figure 13. EBNAl-mediated scgregation in hEBP2-integrated yeast strains. The ability of EBNAl to stabilize the FR plasmid was determined in the yeast strains KY320.hEBP2 and W303.hEBP2, which contain an integrated hEBP2 gene under the controt of a constitutively active promoter. PIasmid loss assays were camed out for 11 generatiom as describcd in Figure 6, in the presence of either wild-type EBNAl (+) or the EBNAA325-376mutant (A325-376). or in the absence of EBNAl (-). slightly higher than the CEN/ARS plasmid, as seen in KY320 and W3O3-1 A. In
.Cz------A. - - - - - &- L- -- --A- - W303.hEBP2, the stability of the FR plddwas lowa than the CENIARS plasmid (See
Appendix; 70% and 67% retention for the CENIARS plasmid and 59% and 21%
retention for the FR plasmid with EBNAl expression). However, the stability of the FR
plasrnid with EBNAl expression was still higher than the stability without EBNAl
expression andlar the FR elemmt. The low stability of the FR plasmid in the presence of
EBNAl is likely due to the high level of hEBP2 degradation in this strain, which leaves a
limiting amount of fùnctional hEBP2 protein for EBNAl-mediated segregation. Though
W303.hEBP2 may not be useful for EBV-based segregation in yeast plasmid loss assays,
KY32O.hEBP2 will be quite usefûl in assays where hEBP2 is an unchanging variable,
such as those that test the ability of various EBNAl mutants to support FR plasmid
segregation in yeast. IV. DISCUSSION - - - .- --
IV.1. Cornparison of EBV-Based Segregaoii in Budding Yeast and Humans
My studies have demonstrateci that an unstable yeast plasmid can be stabilized in
budding yeast by an EBV-based plamid segregation system. This segregation system
requires the EBV segregation element, FR, and the presence of EBNAl and IiEBP2.
Interaction between EBNAl and hEBP2 is also required. An ARS-containing plasmid
segregating under such a system is maintained in yeast with a loss rate of approximately
2-4% per generation. In such plasmids, the FR huictionally substitutes for the CEN
element in the presence of EBNAt and hEBP2. The howu requirements for the EBV-
based segregation in yeast are the same as that in human cells. Stable segregation of
oriP-containing plamids in humans also requires the FR element and EBNAl. The
EBNAl region involved in plddsegregation and hEBP2 binding is the same,
suggesting a role for hEBP2 in EBNAl-mediated plasmid segregation. The EBNAl
mutant lacking this region cannot bind hEBP2 and cannot mediate plasmid segregation.
Studies in a number of hman ce11 lines have indicated that the rate of loss of an oriP- containing plasmid is 26% per generation (Kirchmaier and Sugden, 1995; Yates et al.,
1984), which is simil81 to the rate of loss of the FR-containhg yeast pIasmid.
Similarities between EBV-base-segregation in yeast and humans indicate that insights gaimd on the mechanism of EBV segregation through studies in yeast should truly npnsent the segregation mechaaisrn used by EBV to persist in latently infected human cells. IV.2. bEBP2 Plays an Essentfil ROI~in EBNAl-Meàiated Segregatîon
- - L 4------Before the EBV-based segngation system was developed in yeast, a role for hEBP2 was only implicated in EBNAl-mediated segregation. hm2was found to interact with
EBNAl residues 325-376, which is the same region required by EBNAl for its plasmid segregation and chmmosome attachent activities. Deletion of these residues abrogated al1 three activities, namely hEBP2 binding, plasmid segregation and chromosome attachment. hEBP2 was also found to coat the mitotic chromosomes, much in the same way as EBNAl. These results suggest that EBNAl bound to the EBV episomes might associate with the mitotic chromosomes by interaction with hEBP2 and in this mannn, mediate partitionhg of the episomes during mitosis. However, evidence that suggested a role for hEBP2 in EBNAl-mediated segregation was only correlative and direct evidence was required to verify that hEBP2 was involved in EBV segregation. Such proof, which should demonstrate that EBNAl-mediated segregation occurs in the presence and not in the absence of hEBP2, was not feasible in human cells because hEBP2 is likely required for ce11 viability, like its yeast homologue. My studies in budding yeast have provided direct evidence of the importance of hEBP2 in EBNAl-mediated segregation. 1 have shown that the EBV segregation element FR and EBNAl are not sufficient to confer stability upon an unstable yeast plasmid. However, in the presence of hEBP2, the segregation function of EBNAl is rescued and EBNAl-mediated segregation does occur.
Though the role of hEBP2 in EBNA 1-mediated segrcgation has been confinneci in budding yeast, we do not know whether the EBV segregation mechanism in yeast is the same as in human cells. In humans, EBV episomes are believed to segregate during mitosis by association with the host chromosomes. This ükely occurs by the binding of EBNAI, which is bound to the FR elomcnt of oriP through its DNA binding and ------dimetization region, to hEBP2 through its hEBP2-binding doW.Binding of hEBP2 to
EBNAI, through its EBNAl-binding domain, and to the mitotic chromosomes, through
its chromosome-associating domain, allows locaiization of the EBV episomes to the
chromosomes. The tethered episomes are then partitioned dong with the chromosomes
as the chromosomes sepatate and move to the daughter cells during rnitosis. This mode1
impües that the interaction between EBNAl and hEBP2 is direct, though this has not
been confimed. Lack of purifiecl hEBP2 protein has impeded the testing of whether the
interaction between EBNAl and hEBP2 is direct or requires interaction with other
cellular proteins. However, it is likely that the interaction between the two proteins is
direct, as the interaction has bem shown to occur in both yeast and insect cells (Shire et al., 1999).
IV.3. The Funetion of hEBP2 in EBV Segregatioii Cannot be Provided by yEBP2
EBNAl was unable to stabilize the FR-containing yeast plasmid in the absence of hEBP2, suggesting that the yeast EBP2 (yEBP2) cannot substitute for hEBP2 in EBNAI- mediated segregation. yEBP2 may not support EBNAl-mediated segregation for several reasons. First, yEBP2 fails to bind EBNAI.The specific sequences to which EBNA 1 bimls in hEBP2 may not be preseat in yEBP2. Altematively, the EBNAl -binding sequences may be present, but the structure of yEBP2 may inhibit EBNAl fkom binding to these sequences. The yEBP2 protein contains aa N-terminal extension that is not found in hEBP2. This extension in the teïtiary structure may mask the region of yEBP2 that can bind EBNA1. Second, interaction between EBNAl and yEBP2 rnay occur but the interaction may be weak and wt detected by the yeast two-hybrid assays performed in
-A - -2 the present study. The low affhity between the two proteins may cause them to easily dissociate. Thid, the inability of yEBP2 to mediate EBV segregation may be due to the localkation of yEBP2 in yeast cells during mitosis. Unlike the nucleolus in human cells, the nucleolus in buddllig yeast does not dissociate during mitosis (hotand Snyder,
1991). As a result, nucleolar proteins, like yEBP2, main associated with the rDNA regions and do not coat the chromosomes via the chromosome periphery, as done by hEBP2 in human cells. During chromosome division, the intact nucleolus in budding yeast aiso appears to associate with parts of the nucleus that allows it to segregate along with the buk of the DNA,rather than trailing behiml with the telomeric DNA (Granot
# and Snyder, 1991). Since yEBP2 is localized to only specific parts of the chromosome and may be associated with other nuclear entities during mitosis, it rnay not be easily accessible to EBNAl. hEBP2, on the other hand, may be localized throughout the yeast chromosomes, either in mitosis or duMg the entire ce11 cycle, and thus tether EBNAl to the chromosomes for proper FR plasmid segregation as hypothesized in human cells.
It cm be argued that the stable segregation of a FR-containhg plasmid in the presence of EBNAl and hEBP2 can involve yEBP2. Perhaps the mechanism behind EBNA1- mediated segregation in budding yeast involves interaction between yEBP2 and hEBP2.
However, this mechanism is unlikely, as yeast two-hybrid studies have shown that hEBP2 is able to interact with itself, but does not interact with yEBP2 (Shire, mpublished results). IV.4. A Role for hEBP2 in the Segregation of BPV and KSRV Episornes? ------Low copy numbcr episomal vinises like BPV and KSHV are also believed to stably
partition between daughter cells during rnitosis by attaching to host chromosomes. As
stated earlier, sbidies suggest that association of E2 and LANA segregation proteins with
the chromosomes and with the BPV and KSHV DNA, respectively, allows tethering of
the viral DNA to the chromosomes. Lüce EBV, it is likely that association of E2 and
LANA with the chromosomes is through interaction with chromosomal proteins. The
idea that hEBP2 is a common chromosomal protein that is also used by E2 and LANA to
attach to the condensed chromosomes is tempting but appears unlikely. First, both E2
and LANA share no sequence homology with EBNAI.Thus, the sequences in EBNAl
that bind hEBP2 do not appear to exist in E2 or LANA. Second, neither E2 nor LANA
exhibits a chromosome localization pattern that is similar to hEBP2. Antibodies raised
against hEBP2 completely stain the rnitotic chromosomes and give a staining pattem that
is similar to that generated by DAPI staining of the chromosomes (Wu et al., 2000). A
similar staining pattern is also obsmed when chromosomes are stainecl with antibodies
against EBNAl (Petti et al., 1990; Marechal et al., 1999; Wu et al., 2000; Hung et al.,
2001). However, E2 gives a random speckled pattern on chromosomes, which is unlike
the pattern generated by hEBP2 (SIUadopouIos and McBride, 1998). Though LANA
alone associates with the chromosomes diffisely, it becomes concentrated in dots on the
chromosomes in the presence of KSHV DNA (Ballestas et al., 1999). However, the
difference in the stalliing pattern of LANA and E2 as cornparcd to hEBP2 and EBNAl could simply reflect the lower abundance of LANA and E2 as compared to hEBP2 and
EBNA1. Third, pretiminary work by K. Shire has shown that interaction of E2 with hEBP2 is not detected in ycast two-hybrid assays. Simüarly, ment unpublished work by
L A- -A-.- + - -- - - A. McBride has revealed that expression of hEBP2 does not rescue the ability of E2 to segregate ARS plasmids containing the BPV segregation element in yeast.
nie E2 protein has been show to associate with the nuclear matrix (Hubbert et al.,
1988). This association could be through nuclear matrix proteins that localize to the chromosome periphery during mitosis and thenfore, could be important for E2 attachrnent to mitotic chromosomes. The LANA protein has been shown to interact with histone Hl through CO-immunoprecipitationexperiments (Cotter and Robertson, 1999b).
Both nuclear matrix proteins and histone Hl could be potential candidates for chromosomal proteins that tether E2 and LANA, respectively, to mitotic chromosomes.
The EBV-based segregation system serves as the htdemonstration that viral segregation can be reconstituted in yeast. Though hEBP2 may not mediate the segregation of BPV and KSHV episomes, yeast systems can be used to identify cellular proteins that do mediate segregation of these viral episomes. For instance, a human cDNA library cm be used to screen for cellular proteins that allow the stable segregation in yeast of ARS plasmids containing the BPV or KSHV segregation element in the presence of El or LANA, respectively. Such screens may identify human proteins that tether BPV or KSHV episomes to human chromosomes. V. FUTURE DIRECTIONS
----c------% L - - The work presented in this thesis is only the beginning of the characterization of the
EBV-bas& segregation system in S. cerevisiae. Some of the basic requimnents of the
system have been determined. However, we still neeâ to elucidate the precise mechanism
by which EBV-based segregation occurs. We also need to mercharacterize hEBP2
and EBNAl and determine theu specific roles and sequence requirernents in the system.
1 have already started to pursue these areas of study and plan to continue with these
studies in my Ph.D. research.
V.1. CeUular Locallzation Studies of EBP2 and EBNAl in Budding Yeast
If the mechanism by which EBNAl and hEBP2 mediate FR plasmid segregation in
yeast is sllnilar ta that in humans, then one would expect both proteins to exhibit cellular
localization in yeast that is sirnila.to that in humans, at least during mitosis. EBNAl is
nuclear thughout the ce11 cycle and hEBP2 is nucleolar during interphase and nuclear
during mitosis in human cells. To determine whekthis also holds hue in yeast, 1 will
perform immunofluorescence micmscopy shrdies on yeast expressing these proteins.
Logphase or nocodazole-amsted mitotic yeast cells expressing hEBP2 and/or EBNAl
will be spheroplasted, fixed on microscope slides, stauied with antibodies against hEBP2
andfor EBNAl and observed under the light microscope to determine the localization of
both proteins. Since the yeast nucleolus is difficult to locate, yeast will also be stained
with an antibody against yEBP2, which wiîl serve as a nucleolar marker. In these microscopy studies, the localization of hEBP2 and yEBP2 will also be
L------compared. Cornparison of the cellular loc~tionof hEBP2 and yEBP2 in interphase
and mitosis will help us understand whether diffetences in locolization contribute to the
failure of yEBP2 to mediate EBNAl segregation. If hEBP2 is always nucleolar like
yEBP2, then locaiization of yEBP2 in yeast is unlikely to be the reason behind its
inability to support EBNAl segregation. However, if hEBP2 is ahvays nuclear or is
nucleolar in interphase and nuclear in mitosis in yeast, then the nucleolar localization of
yEBP2 in mitosis may conûibute to its failure to aid EBNAl-mediated segregation.
V.2. Assessing hEBP2 Residues IiivoIved in EBNAl-Medhted Segrqation in Yeast
To undentand the contribution of hEBP2 to EBNAl-mediated segregation in budding
yeast, 1 will attempt to identiQ hEBP2 sequmces that are requirsd for this process.
Based on the hypothesis that hEBP2 tethers EBNAI to the mitotic chromosome, 1 expect
two domains of hEBP2 to be quired for segregation, an EBNAl-binding domain and a chrornosome-associating domain.
V.2.a. Mapping hEBP2 Residues that Interact with EBNAl
1 have begun to assess the functional contribution of hEBP2 sequences by rnapping the
EBNAl-binding domain of hEBP2. 1 performed a yeast two-hybnd assay where 1 have tested the interaction of a fusion protein containhg the GALA DNA-binding domain and
EBNAl with various hEBP2 fragments, fiised to the GALA activation domain. The preâicted structure of hEBP2 indicates that hEBP2 consists of a coiled-coi1 domain at its central region (approximately amino aci& 100-208) that is the most conserved betwem the EBP2 homologs (Shin et al., 1999; Huber et al., 2000). This region was used to
- - - divide hEBP2 into three regions, the N-terminal, the central coiled-coi1 and the C-
terminal region (Figure 14). and the various heBP2 fiagments tested in the two-hybnd
consistai of one or more of these regions.
For the assay, the EBNAl hision protein, along with a hEBP2 fûsion protein, was CO-
expressed in yeast stralli Y 190, which contains integrated HI'S3 and LAC2 reporter genes
under the control of the GAL4 sites, and the activation of the reporter genes were
detemiineà. Assay results with the various hEBP2 fiagments indicated that the C-
terminal hEBP2 region (amino acids 220-306) was sufficient to bind EBNAl and,
conversely, that a C-terminal bEBP2 tnuication mutant (hEBP2 mutant containing arnino acids 1-220) lacking this ngion failed to bind EBNAl . Thus, the C-terminal sequences of hEBP2 contain the EBNAl-binding domain.
V.2.b. Role of the EBNA 1Binding Domain of hEBP2 in EBNAl -Mediateâ Segregation
1 showed that the hEBP2-binding region of EBNAl was required for EBV segregation
in budding yeast, suggesting that sepgation reqwred the interaction of EBNAl with
EBP2. To mertest this hypothesis, 1 asked whether the EBNAl-binding domain of
hEBP2 was required for EBV segregation. 1 assessed the ability of the hEBP2 mutant
(containing amino acids 1-220) that lacks this domain to mediate FR plasmid segregation
in the presence of EBNA l in the yeast plasmid loss assay. 1found that the EBNA 1-
binâing mutant of hEBP2 was unable to segregate the FR plasmid, strengthening the
conclusion that the hEBP2-EBNAI interaction is essential for EBV segregation. Figure 14. Schematic diagram of the hEBP2 protein. Based on its predicted structure, which identified a consmed, coiled coi1 domain at the central region, the hEBP2 protein was divided into three regions as shown. Amino acids spanning each region are indicated. --
To determine whethex the EBNAl-binding region of hEBP2 was sufficient for
-----.L--. ------A- -- EBNAl-mediated segregation, I also tested the ability of the hEBP2 220-306 hgment to
stably segregate the FR plamid in the presence of EBNAl in the plasmid loss assay.
The nsults demonstrated that the FR plamid was rapidly lost hmthe yeast cells in the
presence of the EBNAl-binding domain of hEBP2, indicating that this domain was not
sufficient to mediate EBNAl segregation. 1 conclude that hEBP2 sequences in addition
to the EBNAl-binding sequaices are required to mediate EBV segregation in yeast.
Although we do not biow the functional contribution of these additional hEBP2
sequmces in the yeast segregation system, 1 will refer to these sequences as the
bbchrom~some-associatingregion". This name reflects the fact that, in human cells,
hEBP2 likely binds mitotic chromosomes through sequences that are distinct hmthe
EBNAl binding domain. The binding of -2 to mitotic chromosomes through these
sequences will tether EBNAl to the chromosomes, when EBNAl is bound to hEBP2.
V.2 .c. Identifjing the Chromosome-Associating Region of hEBP2 in Budding Yeast
1 will undertake two approaches to map the chromosome-associating region of hEBP?.
First, in the yeast plasmid loss assay, hEBP2 fragments containhg the C-tenninal
EBNAl-binding region and another hEBP2 region (figure 14) will be tested to determine
which region serves as the chromosome-associatiag region that is required for EBNA1-
mediated segregation of the FR plasmid. 1will start by using hEBP2 kagrnent 95-306,
which will test the centra1 hEBP2 region for chromosome associating activity. If such a
£@mentdoes not mediate segregation, 1 will detmnine if the N-terminal hEBP2 region is the chromosome ansociathg rcgion by testhg an hEBP2 fiagrnent containhg amho
-. - - L- -- - L acids 1- 100 and 220-306 (EBNAl -binding domain) in the plasmid loss assay.
Second, 1will construct fuJion proteins consisting of an EBNAl fiagrnent containing
the DNA-binding domain (amino acids 452-641) hsed to the C-terminus of bEBP2 and
each of the thehEBP2 regions (figure 14). The nuclear localization signal (NLS) of
EBNAl will be fised to the N-terminus of the fusion proteins to msure the enûy of the
proteins into the nucleus. The ability of the resulting fusion proteins to maintain the FR
plasmid will be tested in the yeast plasmid loss assay. In such an approach, the
chromosome-apsociating region of hEBP2 can be rnapped without requiring the presence
of the EBNAl-binding domain in each hEBP2 hgment tested. This approach is based on
the assumption that the segregation fiinction of EBNAl has only two requirements,
which are binding to the FR and hEBP2. In the plddloss assay, the interaction of
EBNAl with the FR plasmid will be mediated through the DNA-binding domain and the
interaction with hEBP2 will be mediated hughfusion with hEBP2. The fusion protein
containing the EBNAl DNA-binding domain and fùll-length hEBP2 will serve as a
positive control and is expected to stably maintain the FR plasrnid in yeast. If the
positive control doa not maintain the FR plamid, I will try using a larger EBNAl
fiagrnent (amino acids 377-641) that contahs both the NLS (amino acids 376-386)and
the DNA-binding domain in the hision proteins. The EBNAl DNA-binding domain alone
will serve as the negative control and is not expected to support stable segregation of the
FR plasmid. Through this approach, I cm ask whether the hEBP2 C-terniinal ngion is
only required for association with EBNAl or whether it is also hvolved in chromosome
attachment by testing a fusion protein containhg the hEBP2 hgment lackhg the C- tdalregion. The mswer to this question cannot be obtaineâ through the fht ------approach descnbed.
If the fusion protein containing the EBNAl DNA-binding domaln and hEBP2
(positive control) does not stably segregate the FR plasmid, then EBNAl sequences
outside of the DNA-binding domain may be contributhg to more than just the EBP2
interaction. In this case, 1 will fuse EBNAl or EBNAd325-376(lacking the hEBP2-
binding region) to hEBP2 and determine whetha the resulthg fusion proteins can
mediate FR plasmid segregation in the plasmid loss essay. If fusion proteins containing
both version of EBNAl are functional for segregation, then this would imply that the
EBNAl sequmces, in addition to the DNA-binding domain and the hEBP2-bhding
domain, are likely required for segregation. Alternatively, if the hision protein
containing EBNAl is hctional for segregation and that containing EBNM325-376 is
not, then additional EBNAl sequences quired Iikely map to the hEBP2-bUiding region.
Such a result would suggest that EBNAl amino acids 325-376 are not only involved in
binding hEBP2 but have an additional unidentified role in EBV segregation.
If a version of EBNAl (EBNA1 or EBNAA325-376) that mediates segregation when
tiised to hill-length hEBP2 can be identified, I will use this version to make the various -- EBNAI-hEBP2 &ion proteins that will be used to map the chromosomeassociating
region of hEBP2 by the plasmid loss assay. Ethc -ion proteins are not well expressed
in yeast due to theu high molecular weight, 1can attempt to delete EBNAl sequences
that are known to be dispensable for its segregation hction, such as the N-tednd
sequences spdgthe fht Gly-Arg region and the C-temiinal acidic tail (Wu et al.,
unpublished resuits; Ceccarelli ami Frappier, 2000), hmthe hion proteins. 1wilî ensure that deletion of these EBNAl sequences does not Scct the ability of the hision
-- A-*-& = - - A ------proteins to mediate FR plasmid segregation through the plamiid loss assay.
V.3. Idenwng hEBP2 Residues Required for Chromosome Association in Humans
Thus fa., 1 have talked about mapping the hEBP2 residues that are required for
EBNA 1-mediated segregation, in addition to the EBNAl -binding residues, in budding
yeast and have referreâ to these residues as the chromosome-associating region. I also
plan to identify residues that are required for hEBP2 chromosome association in human
cells. Association of hEBP2 with the mitotic chromosomes has been shown in humans
(Wu et al., 2000) and it will be important to detemiine whether the hEBP2 residues that
associate with the chromosomes in humans are the sarne as those mapped in yeast.
For mapping studies in humans, I will generate constructs that express hEBP2
hgrnents fùsed at the N-terminus to the rnyc epitope. hEBP2 fhgments used will
consist of either one of the thehEBP2 regions (cefer to figure 14) and as well, pair wise
combinations of these regions. Before these fusion proteins are made, positive and
negative control constructs that express full-length hEBP2 fused at the N-teminus to rnyc
or rnyc alone, respectively, will be constructeci and tested for mitotic chromosome
association (as described below). If the positive control does not associate with the
chromosomes, then I will try tagging hEBP2 at the C-terminus with rnyc. The rnyc
epitope is an ideal choice for tagging hEBP2 because it is only 13 arnino acids long,
making it unlikely that the small size of the epitope will interfére with the chromosome-
associating function of hEBP2.
Once the myc-containhg expression consüucts are made, they will be used to transf~thuman ce11 lines and the percentage of cells expressing the rnyc fusion proteins
----A------will be detennined by immunofluonscence microscopy using an antibody against the
rnyc epitope. Once the expression of the rnyc hision proteins is connmied, transfected
human cells will be blocked in mitosis by colcemid treatment. Mitotic chromosomes will
be spread, stained with DAPI to visualize them and stained with anti-myc antibody to
determine whether the rnyc fusion proteins localize on the chromosomes. For full-length
hEBP2 fused to rnyc, 1expect to see overlap of the DAPI and rnyc signals, while for the
rnyc tag done, 1 expect to see no rnyc signal on the chromosomes.
V.4. Using Yeast and Human EBP2 Hybrids to Assas the Inability of Yeast EBP2 to
Support Segregation by EBNAl
1 have shown that yeast EBP2 (yEBP2) does not mediate segregation by EBNAl and
does not appear to interact with EBNAl in a yeast two-hybrid assay. If the failure of
yEBP2 to support EBNAl 's segregation huiction is only due to its failure to bind
EBNAl, then addition of the EBNAl-bindipg domaln of hEBP2 to yEBP2 should allow
yEBP2 to bind EBNAl and support EBNAl-mediateâ segregation. To test this
possibility, I will fuse the C-terminal EBNAl-binding domain of hEBP2 (arnlno acids
220-306) to the C-tamuiai end of yEBP2 that has been üuncated after the central coiIed-
coi1 domain. The tnmcated yEBP2, which will contain amino acids 1-347 (yEBP2
consists of 427 amino acids), will lack those amino acids at the C-teminus that
corresponds to hEBP2 amino acids 220-306 (refer to figue 3). Since removing the C-
terminal region of yEBP2 might affect the folding of the other yEBP2 domains, I will - also construct a fusion protein that contains the bEBP2 EBNAl-binding domain fused to
-a------2%- --=- - -- - %- - the C-terminus of full-length yEBP2.
The ability of the two EBP2 fusion proteins to interact with EBNAl will be tested by
the yeast two-hybnd assay. Since the C-temiinal amino acids 220-306 of hEBP2 is
sufficient to bind EBNAI, it is expected that the fusions containhg these amino acids
should interact with EBNAl. Once the interaction is confinneci, I will detennine whether
the hybnd proteins mediate segregatian of the FR plasmid in the presence of EBNAl in
the yeast plasmid loss assay.
The EBP2 fusion proteins rnay interact with EBNAl but rnay not mediate FR plasmid
segregation. Two possible re85ons for this result are as follows: first, yEBP2 rnay not
have a chromosome-associating domain that is hctionally equivalent to th3t of hEBP2.
Second, yEBP2 rnay contain the chromosome-associating domain but this domain rnay
not be accessible in the tertiary structure of the fusion proteins. As mentioncd previously,
yEBP2 contains a N-terxninal extension that does not exist in hEBP2 and the presence of
this extension rnay inhibit the ability of yEBP2 to function in EBNAl -mediated
segregation.
Once 1 have mapped the chromosome-associating domain of hEBP2,I can test the
fht possibility. Upon the identification of this domain in hEBP2, hEBP2 amino acids
responsible for interaction with EBNAl and chromosome association can be hised to
yEBP2 or cari be used to replace yEBP2 amino acids that occupy the same position as the
hEBP2 amino acids (as detennined hmFigure 3). The segregation ability of the
resulting hybrid proteins cmbe tested in the plannid loss assay. The second possibility,
that the N-terminal sequaices of yEBP2 masks the chromosome-associating region, can be tested by removing the N-taminel sequaces of yEBP2 hmthe fisions containhg
--.L-=- -- -.A------AL ------full-length or C-tenninal truncated yEBP2 and the EBNAl-binding domain. The ability
of the N-terminal truncated fusion proteins to mediate FR plasmid segregation in the
presence of EBNAl can then be tested.
V.5. bEBP2-Mediated Segregation of a Phsrnid Containing GAIA Binding Sites in
the Presence of the GAL4 DNA-Binding Domain
If the hypothesis that hEBP2 tethers the segregation plasrnid containing the FR
element to the mitotic chromosomes in the yeast segregation system is true, then the
apparent role of the FR element and EBNAl in this system is to allow association
between the plasmid and hEBP2.I will examine whether FR and EBNAl are specifically
requirrd for hEBP2-modiated segregation in yeast or whether these viral components can
be replaced by non-virai components that allow association between hEBP2 and the
segregation plasmid. To this end, 1 will replace the FR element hmthe yeast
segregation plasmid, YRp7FR (refer to Materials and Methods, 11.2). wi th multiple
GALA-binding sites and test the stability of the resulting plasmid under two conditions.
First, 1 will test whether such a plasmid can be maintaineci in yeast in the pnsence of a
hion protein containhg the GAIA DNA-binding domain hedto the N-terminus of
hEBP2 or to the chromo~~mes-associatingdomain of hEBP2. Second, 1 will test the
stability of the plasmid in the presence of a fusion protein containhg the GAL4 DNA-
binding domain hised to the N-terminus of EBNAl(1-386) (EBNAl amino acids 1-386,
one of the smallest EBNAl hgments tested that binds hEBP2) and hEBP2. In this
second condition, plasmid segregation will be dependent on interaction between the GALA-EBNA1 fusion protein and hEBP2. If the GALA-containing plasmid is maintaineci
% ------under both of these conditions, it will hdicate that hEBP2-mediated segregation in yeast
cm be gencralized and does not specifically requin EBV components. This will imply
that the main requimnent for segregation by hEBP2 is that hEBP2 is in some way
associateci with the segregation plasmid.
If the GAZA plasmid segregation system stabilizes an unstable yeast plasmid, 1 will
attempt to determine whether the segregation system also works in mammalian cells.
Mammalian plasmids expressing the GAL4-hEBP2 or the GALA-EBNAl(1-386) fusion
proteins and containing a drug selection marker, the DS element and the GAL4 binding
sites will be used to transfect a marnmalian ce11 line that expresses EBNAl (for
replication). The GAL4-hEBP2 fusion protein may attach to chromosomes inefficiently
due to cornpetition with the endogenous hEBP2 for chromosome-binding sites. However,
this possibility would be eliminated when the GALA-EBNAl(1-386) fusion protein is
us& since this fusion protein should hteract with the endogenous hEBP2, which is requid for plasmid segregation. The stability of the mammalian plasmid under drug
selection for two weeks will be determincd by a long-tem plasmid maintenance assay as descnbed in Shire et al. (1999) and Ceecarelli and Frappier (2000).
If the GAUplasmid sepgation system does maintain the mammalian plasmid, then it will prove to be very usefbi for the development of constructs for gene therapy.
Cumntly, several viral systems exist for replicating and maintainhg extrachromosomal
DNA in human cells (reviewed in Cotter and Robertson, 1999a). However, these systems requin expression of viral proteins that may dktcellular gene expression. Ideaily, these viral proteins should be eliminated to duce effccts on the target cells. Human sequences that cm hinction as on@ on plasmids have besa reporteci (Krysan et al., 1989; Frappier
--A-- * r .--. ------& and 2huis-Hadjopouios, 1987; Leffhk and James, 1989) but to date, the only sequence
shom to enable segregation of plasmids in human cells is FR. If the FRlEBNAl
segregation system cm be replaceû by GAIA sites end a GAIA fusion protein, it would
be an important step towards the development of gene therapy consûucts that do not alter
cellular gene expression. Furthmore, the GAL4 system may prove to be a better
segregation system than the FRlEBNAl system. EBNAl can maintain oriP-containing
plasmids in cells for only a few months, since these plasmids are gradually lost fiom the
cells. The GAL4 fusion protein may maintain the plasmids containing the GALA sites for
a longer period of tirne and may serve as a more efficient segregation system.
V.6. Use of a Sectoring Assay to Measare Plasmid StabUity in the EBV-Based
Segregation System in Yeast
hEBP2 rnay not be unique in its ability to mediate EBV segregation. It is possible that
other human proteins might tether EBNAl to mitotic chromosomes and support EBV
segregation in humans and yeast. To test this possibility, 1will conduct a screen to search
for human proteins that mediate EBNAl-mediated segregation in yeast. For this screen,
the use of the plasmid loss assay descn'bed in this study irefer to MatenaIs and Methods,
LI.6) may not be feasible. The plasmid loss assay is a direct method to measure plasmid
stability that is nproducible and has worked quite welî for the EBV-based segregation
system in yeast. However, it will prove to be very laborious and tirne-consuming when
the stability of a plasmid is to be measured under many ciiffernt conditions and quick
results are tequirad, as in the scmn. To make the detdation of plasmid stability casier and faster in these cases, I will develop a sectoring assay (also called adenine
- - s- --.3 - L - - - genetic assay and colony color assay) for the FR segregation system.
The satoring assay (Hieter et al., 1985; Koshlmd et al., 1985) is an indirect method to measure plamid stability. It pmvides a visual distinction between colonies containing a stable plasmid and an unstable plasmid and also allows a visual measwment of the rate of plasmid loss. The basic principle behind the assay is as follows. The ocb suppressing tRNA gene, SUPI 1, is cloned into the plasmid whose stability is to be determineci. The SVPI I -expressing plasmid is then used to transformed a yeast haploid strain that contains the ade2-IO1 ochre mutation. Due to a defect in purine (adenine) synthesis, ade2- 101 cells accumulate a nd pigment, leading to the formation of red colonies when adenine in the growth medium is lhiting. The presence of a RNA
SWPI I gene on the plasmid suppresses the ade2-IO1 ochre mutation in the cells allowing the synthesis of adenine and causing the cells to fom white colonies. When the SUPI 1- containing plasmid is lost duMg the growth of a colony, red regions or sectors appear, where the arnount of red sectoring provides a visual assay for the rate of plasmid loss.
Cells tnuisformed with a stable plasmid will result in colonies that are al1 or mostly white, indicating a low rate of plasmid loss, while cells containing an utable plasmid will form colonies that are mostly red, indicathg a high rate of plasmid loss.
For the use of the sectorhg assay in the EBV-based segregation system in yeast, 1 will insert a SUPl l expression cassette into the FR-containhg plasmid, YRp7FR. 1will confirm that the assay works for the FR plasmid by transfonning yeast strain KY320
(which is ude2-101) with the SUPIIantaining FR plasmid, the EBNAl expression vector and the hEBP2 expression vector. The resuiting traiisfomants will be pwnin non-selective medium (SC-Ura,Leu) with respe!ct to the segregation plasmid as done in ,------this study and plated on SC-Ura, Leu plates contaihg an experimentally detennined
amount of adenine. The mount of admine that will be used in the plates will be such
that it will allow pwth of cells lacking the SüPZl gene (cells fhxn which the
segregation plasmid is lost) but will also be limiting and therefore, will allow the
formation of the red color. If the colour assay works in the system, 1 will expect colonies
the are mostly white in colour to be visible on these plates. When FR, EBNAl or hEBP2
is omiîted, 1 expect to set colonies that are mostly red in colow on these plates.
Once I have confirmed that the sectoring assay can be used in the EBV-based segregation system, I will use this assay to search for human proteins that mediate EBV segregation in yeast, in addition to hEBP2, by screening a human cDNA library. Ideally,
1 would require a cDNA library in a yeast expression plasmid containing the LEU2 marker, since the library plasmids will be replacing the hEBP2 expression plasmid, which also contains the LEU2 marker, in the EBV-based segregation system. If the required library is not comrnercially available, 1will attempt to constnict the library myself or through the aid of a biotechnology Company. Altematively, 1 cm use a human cDNA library that was constmcted by Clontech Laboratones, calleâ the Lambda
Maxl/pYEUra3 library, which will be provided by Dr. Alison McBride. Such a Iibrary is expnssed bma yeast expression plasmid containing the LIRA3 marker. If the screen is conducted using this library, 1will express EBNAl in the assay hma LEUZ-containing expression plasmid mther than the IBPA3-containing expression plasmid as done in this study. As well, 1will express hEBP2 (which was expressed hma LEU'-containing expression plasmid in this study) firom an LIRA3antaining expression plasmid, as
--..- - - L ------L - - - - expression of this protein wiil be required for the positive control.
To conduct the scnen, KY320 will be transfod with the SUPII-containhg FR
plasmid and positive üansfonnants will be selected for on SC-Trp plates. The
ûansfonnants will be grown in SC-Trp liquid medium and re-transfod with the
EBNAl and hEBP2 expression plasmids (positive control), the EBNAl and a LEU2 or
LIRA3 expression vectors (negative control) or the EBNAl expression plasmid and a
human cDNA library (on LELJ2- or U'i-containing plasmids). Whether the EBNAl
and hEBP2 expression plasmids used contain the LEU2 or the CIRA3 marker will be
dependent on the plasmid library useci. If these transformations prove to inefficient due to
the introduction of three plasmids into the cells the EBNAl gene may be intqpted into
the KY320 genome. The transformants will be plated on SC-Ura, Leu plates with
limiting adenine (as describeci above). Colonies (containhg the library) that are mostly
white in colour will be picked hmthe plates and streaked ont0 the sarne type of
selective plates to connmi that they represent positive clones (as compared to the positive
control). Positive clones will be süeaked ont0 SC-Leuor SC-Ura (depending on the
marker on the library plasmids) to obtain clones that only contain the library plasmid.
Colonies on the SGLeu or SC-Un plates wifl tCmi be repIica-plated ont0 SC-Um or SC-
Leu and SC-Trp to ensure that the EBNAl expression plasmid and the segregation plasmid, respectively, are lost hmthe cells. The library plasmids wiU be recovered in bacteria and the abiiity of the library plamnids to maintain the segregation plasmid will be retested through the sectoring assay and the plasmid loss assay described in this study.
The library plasmids that are positive in these assays will be tested for EBNAL- and FR- dependence by assessing their aôility to maintain the segregation plasmid in the absence
Pd---=------of FR andior EBNAl. Genes hmthe library plasmids that enable FR plasmid
segregation in an EBNA- and FR- dependent mmerwill be subsequently identified by
sequence analysis. Localization of the proteins encoded by these genes in human cells in
mitosis will then be exarnined using antibodies that will be obtained or generated, to
determine whether they CO-localizewith EBNAl on mitotic chromosomes. Appendlr VI.1. Stability of Segrcgation Plasmi& in the Rescncc and Absence of EBNA 1 and hEBP2 - Results hmindividual Experiments
- ilitv " Scgrcgation EBNAl Plasrnid Expression pRS3 14 pRS3 14
YRP~FR +
YRp7FR +
YRp7FR yR~7 4-
YRpm A325-376 np7 np7
%lues shown for plasxnid stabüity wae determincd &er II geacntionr of growth in non-selective media as dtscnid in Materials and Mcthods and rcfcr to plasmid retention. '1n KY320 and W303-IA,hEBP2 was expresscd hmthe p425PGK.bEBP2 plasmid, whik in KY320.hEBP2 and W303.hEBP2, hEBP2 was expresscd ftom the integratcd gene. 'refen to the number of tîmes the givcn expcriment was donc bon applicable. Adams, A. (1987). Replication of latent Epstein-Barr virus gmomes in Raji cells. J.
Virol. 61, 1743- 1746.
Ambinder, RF., Shah, WA., Rawlins, DR., Hayward, GS., and Hayward, SD. (1990).
Definition of the sequence requirements for binding of the EBNAl protein to its
palindrornic target sites in Epstein-Barr virus DNA. JeVirol. 64,2369-2379.
Ambinder, RF., Mullen, M., Chang, Y-N.,Hayward, GS., and Hayward, SD. (1991).
Functional domains of Epstein-Barr virus nuclear antigen EBNA-1. J. Virol. 6.5, 1466-
1478.
Babcock, GJ., and Thorley-Lawson, DA. (2000). Tonsillar memory B cells, latently
infectecl with Epstein-Barr Wus, express the restricted pattern of latent genes previously
found only in Epstein-Barr virus-associated tuxnors. Roc. Netl. Acad. Sci. USA, 97,
12250-12255,
Babcock, GJ., Hochberg, D., and Thorley-lawson, DA. (2000). The expression pattern of
EpstebBarr virus latent genes in vivo is dependent upon the differentiation stage of the infected B cell. Immunity 13,497-506.
Bailis, AM., and Rothstein, R. (1990). A defect in mismatch repair in Saccharomyces
86 Baker, RE., and Masison, DC. (1990). Isolation of the gene encoding the Saccharomyces
cerevisiae centromere-binding protein CP 1. Mol. Cell. Biol. 10, 1863- 1872.
Ballestas, ME., Chatis, PA., and Kaye, KM. (1999). Efficient persistent of
extrachromosomal KSHV DNA mediated by latency-associated nuclear antigen. Science
284,64 1-644.
Bastien, N., and McBride, AA. (2000). Interaction of the Papillomavirus E2 protein with
mitotic chromosomes. Virology 270, 124-134.
Beral, V., and Newton, R. (1998). Oveniew of the epidemiology of irnrnunodeficiency-
associated cancers. J. Natl. Cancer Inst. Monogr. 23, 1-6.
Blake, N., Lee, S., Redchenko, 1.. Thomas, W., Steven, N., Leese, A., Steigerwald-
Mullen, P., Kurilla, MG., Frappier, L., and RiclSnson, A. (1997). Human CD~+Tce11
responses to EBV EBNAI : HLA clas 1presenfation of the (Gfy-Ma)-containhg protein
requires exogenous processing. Immunity 7.79 1-802.
Bloom, C. (1993). The centromere hntier: hetochore components, microtubule-based motility, and the CEN-value paradox. Cell73,62 1624. - - Bochkam, A., BarweU, IA., Phetma, RA., Furey, W., Edwards, AM., and Frappier, L.
-----AL - -- & L -- - - (1995). Crystal structure of the DNA-binding domain of the Epstein-Barr virus ongin- binding protein EBNAI . Ce11 83,3946.
Bochkarev, A., Barwell, JA., Pfbetzner, RA., Bochkareva, E., Frappier, L., and Edwards,
AM. (1996). Crystal stnicture of the DNA-binding domain of the Epstein-Barr virus
origin-binding protein, EBNAI ,bound to DNA. Ce11 84,79 1-800.
Bochkarev, A., Bochkareva, E., Frappier, L., and Edwards, AM. (1998). The 2.2 A
structure of a permanganate-sensitive DNA site bound by the Epstein-Barr virus origin binding protein, EBNAI .J. Mol. Biol. 284, 1273-1278.
Ceccarelli, DFJ., and Frappier, L. (1998). Separation of the DNA replication and
ûansactivation activities of EBNAl, the origin binding protein of Epstein-Barr virus.
Gene Ther. Mol. Biol. 3,1910.
Ceccarelli, DFJ., and Frappier, L. (2000). Functional analyses of the EBNAl origin DNA
binding protein of Epstein-Barr virus. J. hl.74,44394948.
Chan, CS., and Botstein, D. (1993). Isolation and characterization of chromosome-gain
and increase-in-ploidy mutants in yeast. Genetics 135,677-691.
Chang, Y., Cesman, E., Pessin, M. S., Lee, F., Culpepper, J., Knowles, D. M. and P. S.
88 Moore. (1 994). Identification of HerpesvVus-üke DNA sequences in AIDS-associated
---A- ---A%- - -. - - - Kaposi's sarcoma Science 266,1865- 1869.
Chatterjee, A., Fmman, JW.,and Busch, H. (1987). Identification and partial
characterization of a M, 40,000 nucleolar antigen associated with cell proliferation.
Cancer Res. 47,1123-1 129.
Chen, M-R.,Middeldorp, JM., and Hayward, SD. (1993). Separation of the complex
DNA binding domain of EBNA-1 into DNA recognition and dimerization subdomains of
novel structure. S. Virol. 67,4875485.
Chen, W., and Struhl, K. (1988). Saturation mutagenesis of a yeast hid "TATA
element": Genetic evidence for a specific TATA-binding protein. Proc. Natl. Acad. Sci.
USA 85,2691-2695.
Chiang, C., Ustav, M., Stenlund, A., Ho, T., Broker, Te,and Chow, Le(1 992). Viral El
and E2 proteins support replication of homologous and heterologous papillomaviral
ongins. Proc. Natl. Acad. Sci. USA 89,5799-5803.
Chien, CT., Bartel, PL., Stemglanz, Re,and Fields, S. (1991). The two-hybrid system: A
method to identify and clone genes for proteins that interact with a protein of interest.
Proc. Natl. Acad. Sci. USA 88,9578-9582. Chittenden, T., Lupton, S., and Levine, AJ. (1989). Functional limits of oriP, the Epstein-
----.xc.------Barr virus plasmid ongin of replication. J. Virol. 63,301 6-3025.
Clarke, L,,and Carbon, J. (1980). Isolation of a yeast centromere and construction of
functional srnaIl circular chromosomes. Nature 287,5040509.
Cotter, MA., and Robertson, ES. (1999a). Molecular genetic analysis of herpesvirus and
their potentiai use as vecton for gene therapy applications. Curr. Opin. Mol. Ther. 1,633-
644.
Cotîer, MA., and Robertson, ES. (1999b). The latency-associated nuclear antigen tethers
the Kaposi's Sanoma-associated herpesvirus genome to host chromosomes in body
cavity-based lymphoma cells. Virology 264,254-264.
Cruickshank, J., Shire, K., Davidson, AR., Edwards, A., and Frappier, L. (2000). Two
domains of the Epstein-Barr virus origin DNA-binding protein, EBNAI, orchestrate
sequence-specific DNA binding. J. Biol. Chem. 275,22273-22277.
Decker, LL., Shankar, P., Khan, G., Freeman, RB., Dezube, BJ., Lieberman, J.,
Thorley-Lawson, DA. (1996). The Kaposi sarcoma-associated herpesvhs (KSHV) is
present as an intact latent gmome in KS tissue but replicates in the peripheral blood
mononuclear cells of KS patients. J. Exp. Md.184,283-288. * -- Delecluse, HJ., Bartnizke, S., Hamtnerschmidt, W., Bulleràiek, J., and Bornkamm, GW.
-----a---lA- - - - L------(1993). Episomal and integristeci copies of Epstein-Barr virus coexist in Burkitt
Lymphoma ce11 lines. S. Virol. 67,1292-1299.
Dhar, V., and Schildkraut, CL. (1991). Role of EBNAl in arresting replication forks at
the Epstein-Barr Wus oriP family of tandem repeats. Mol. Cell. Biol. 11,6268-6278.
Ermakova, OV., Frappier, L., and Schildhut, CL. (1996). Role of the EBNA-1 protein
in pausing of replication forks in the Epstein-Ban Wus genome. J. Biol. Chem. 271,
3309-33017.
Fields, S., and Song, 0. (1989). A novel genetic system to detect protein-protein
interactions. Nature 340,245-246.
Fitzgerald-Hayes, M. (1987). Yeast centromeres. Yeast 3, 187-200.
Fomproix, N., Gebrane-Younes, J., and Hemandez-Verdun, D. (1998). Effects of anti-
fibriiarh antibodies on buiigof fhctiona1 nucleoü at the end of m*tosis. J. Ce11 Sci.
11 1,359-372.
Frappier, L., and O'Do~elî,M. (1991a). Overproduction, purification and
characterization of EBNA 1, the origin binding protein of Epstein-Barr virus. J. Biol.
Chem. 266,78 19-7826. Frappier, L. and O'Donnell, M. (1991 b). Epstein-Barr nuclear antigen 1 mediates a DNA
-- A.*- .------A-A--- &- - -- -+ loop within the latent replication origin of Epstein-Barr virus. Proc. Natl. Acad. Sci. USA
88,10875- 10879.
Frappier, L., Goldsmith, K.. and Bendell, L. (1994). Stabiiization of the EBNAl protein
on the Epstein-Barr virus latent origin of DNA replication by a DNA looping mechanism.
J. Biol. Chem. 269,1057-1062.
Frappier, L., and Zannis-Hadjopoulos, M. (1987). Autonomous replication of plasmids
bearing monkey DNA ongin-enriched sequences. Proc. Natl. Acad. Sci. USA 84,6668-
6672.
Gahn, TA., and Schildkraut, CL. (1989). The Epstein-Barr virus ongin of plasmid
replication, oriP, contains both the initiation and temination sites of DNA replication.
Ce11 58,527-535.
Gahn, TA., and Sugden?B. (1995). An EBNA-1 dependent enhancer acts hma distance
of 10 kilobase pairs to increase expression of the Epstein-Barr virus LMP gene. J. Virol.
69,263302636.
Gaudet, AM., and Fitzgerald-Hayes, M. (1990). The function of centrorneres in
chromosome segregation. In The eukaryotic nucleus, PR. Strauss and SH.Wilson ,ed.
(West Caldwell: The Telford Press), pp. 845-881. Gautier, T., Fomproix, N., Masson, C., Azum-Gelade, MC., Gas, N., and Hemandez-
--Y= -_--Y .- i __* --Li - . - - .- - Verdun, D. (1994). Fate of specific nucleolar perichromosomal proteins during rnitosis:
cellular distribution and association with U3 snRNA. Bi01 Cell82,S 1-93.
Gietz, RD. and Schiestl, RH. (1995). Transforming yeast with DNA. Meth. Mol. Cell.
Biol. 5,255-269.
Granot, D., and Snyder, M. (1 991). Segregation of the nucleolus during rnitosis in
budding and fission yeast. Ce11 Motil. Cytoskeleton 20,4744.
Hahn, PJ. (1993). Molecular biology of double-minute chromosomes. Bioessays 15,477-
484.
Harper, W., Adami, GR., Wei, N., Keyomarsi, K. and Elledge, SJ. (1993). The p2 1 Cdk-
interaction protein Cipl is a patent inhibitor of G1 cyclin-dependent kinases. Ce11 75,
805-8 16.
Hedge, RS., Grossman, SR., Laimins, LA., and Sigler, PB. (1 992). Crystal structure at
1.7A of the bovine papillomavirus-1 E2 DNA-binding domain bound to its DNA target.
Nature 359,505-512.
Hegemann, JH., and Fleig, UN. (1993). The centromere of budding yeast. Bioessays 15,
45 1-460. Hernandez-Verdun, D., and Gautier, T. (1994). nie chromosome periphery during
------% - -L -2 - mitosis. Bioessays I6, 179-1 85.
Hieter, P., Mann, C., Snyder, M., and Davis, RW. (1985). Mitotic stability of yeast
chromosomes: A colony color assay that measures nondisjunction and chromosome loss.
Cell40,38 1-392.
Howley, PM. (1996). Papilomaviriitae: the viruses and their replication. In Virology,
2nd ed., BC. Fields, DM. Knipe, and PM. Howley, eds. (Philadelphia: Lippincott-Raven
Publishers), pp. 2045-2076.
Hsieh, D-J.,Camiolo, SM., and Yates, JL. (1993). Constitutive binding of EBNAl
protein to the Epstein-Barr virus replication origin, oriP, with distortion of DNA structure
during latent replication. EMBO Je12,4933-4944.
Hubbert, NL., Schiller, JT., Lowy, DR., and Anhphy, EJ. (1988). Bovine papilloma
virus-transformed cells contain multiple E2 proteins. Pmc. Netl. Acad. Sci. USA 85,
5864-5868.
Huber, MD., Dworet, W., Shire, K., Frappier, L., and McAlear, MA. (2000). The
budding yeast homolog of the human EBNAl-binding protein 2 (Ebp2p) is an essential
nucleolar protein required for pre-rRNA pmcessing. J. Biol. Chem. 275,28764-28773. Hughes, F., and Romanos, M. (1993). El protein of human papillomavirus is a DNA
----A ----A - .- . ------helicase/ATPase. Nucleic Acids Res. 21,58 17-5823.
Hung, S., Kang, MS., and Kieff, E. (2001). Maintenance of Epstein-Barr virus (EBV)
oriP-based episomes requires EBV-encoded nuciear antigen-l chromosome-binding
domains, which can be replaced by high-mobility group-1 or histone Hl. Proc. Natl.
Acad. Sci. 98, t 865- 1870.
Hyman, AA., and Sorger, PK. (1995). Stnictwe and bction of kinetochores in budding
yeast. Annu. Rev. Cell. Dev. Biol. 11,471495.
Ilves, I., Kivi, S., and Ustav, M. (1999). Long-tenn episomal maintenance of bovine papillomavirus type 1 plasmids is determined by attachent to host chromosomes, which
is mediated by the viral E2 protein and its binding sites. J. Virol. 73,4404-4412.
houe, S. (198 1). Ce11 division and the mitotic spindle. J. Cell. Biol. 9l,l3 1-147.
Ito, S., Ma,M., Kato, N., Matsumoto, A., Ishikawa, Y., Kumakubo, S., and Yanagi, K.
(2000). Epstein-Barr virus nuclear antigen- 1 binds to nuclear transporter Karyopherin
al/NPI-1 in addition to Karyopherin a2/Rchl. Virology 266,110-1 19.
Kanda, T., Otter, M., and Wahl, GM. (2000). Mitotic segregation of Wal and cellular
acentric extrachromosomal molecules by chromosome tethering. J. Ce11 Sci. lM,49-58. ---- - L - --A> - - Kanda, T., Otter, M.,and Wahl, GM.(2001). Couplhg of mitotic chromosome tethering
and replication cornpetence in Epstein-Ban virus-based plasrnids. Mol. Cell. Biol. 21,
357603588.
Kedes, DH., Lagunoff, M., Re~e,R., and Ganem, D. (1997). Identification of the gene
encoding the major latency-associated nuclear antigen of the Kaposi's sarcoma-
associated herpesvinis. J. Clin. Invest. 100,2606-2610.
Kieff, E. (1996). Epstein-Barr virus and its replication. In Fields Vuology, BN. Fields, DM. Knipe and PM. How ley, eds. (Philadelphia: Lippincott-Raven Publishers), pp.2343 - 2396.
Kim, AL., Maher, M., Hayman, JB.,Ozer, J., Zerby, D., Yates, K.,and Lieberman, PM.
(1997). An imperfect correlation between DNA replication activity of Epstein-Barr virus
nuclear antigen 1 (EBNAI) and binding to the nuclear import receptor, Rch 1Amportin
alpha. Virology 239,340-3 5 1.
Kikuchi, Y. (1 983). Yeast plasmid requires a ch-acting locus and two plasmid proteins
for its stable maintenance. Ce11 35,4879493.
Kingsbury, J., and Koshland, D. (1991). Centromere-dependent binding of yeast
minichromosomes to microtubules in vitro. Ce11 66,483-495. Kingsbury, J., and Koshland, D. (1993). Centromere hction on minichromosomes
-.w- -P.- & -- - isolated hmbudding yeast. Mol. Biol. Cell. 4,859-70.
Kirchmaier, AL., and Sugden, B. (1995). Plasmid maintenance of derivatives of oriP of
Epstein-Barr virus. J. Virol. 69, 1280-1283.
Komaki, S., and Vos, JM. (2000). Epstein-barr virus vectors for gene therapy. Adv. Virus
Res. 55,453-462.
Koshland, D., Kent, JC., and Hartwell, LH. (1985). Genetic analysis of the mitotic
transmission of minichromosomes. Ce11 40,393-403.
Krysan, PJ., Haase, SB., and Calos, MP. (1989). Isolation of human sequences that
replicate autonomously in human cells. Mol. Cell. Biol. 9, 1026-1033.
Laine, A., and Frappier, L. (1995). Identification of Epstein-Barr virus nuclear antigen I
protein domains that direct interactions at a distauce between DNA-bound proteins. J.
Biol. Chem. 270,30914-30918.
Lechner, J., and Carbon, J. (1991). A 240 kD multisubunit protein complex (CBF3) is a
major component of the budding yeast centromere. Ce11 64,7 17-725.
Leffak, M., and James, CD. (1989). Opposite replication polarity of the gem line c-rnyc
97 gene in HeLa cells compareci with that of two Burkitt lyrnphoma ce11 lines. Mol. Cell.
Sv---- - <. 2 - - Biol. 9,586-593.
Lehman, CW., and Botchan, MR. (1998). Segregation of wal plasrnids depends on
tethering to chromosomes and is regulated by phosphorylation. Proc. Natl. Acad. Sci. 95,
433894343.
Levitskaya, J., Coram, M., Levitsky, V., Imreh, S., Steigerwald-Mullen, PM., Klein, G.,
Kurilla, MG., and Masucci, MG. (1995). Inhibition of antigen processing by the interna1
repeat region of the Epstein-Barr Wus nuclear antigen-l . Nature 375,685-688.
Li, JJ., Friedman-Ken, AE.,and Huang, YQ. (1996). Detection of HHV-8in subsets of
blood cells fkom patients with AIDS-related Kaposi's sarcoma. In Program and Abstracts
of The XI International Conference on AIDS, Vancouver, Canada. Abstract WeAl63.
Li, L., Elledge, SJ., Peterson, CA., Bales, ES., and Legerski, RI. (1994). Specific
association between the human DNA repair proteins XPA and ERCC1. Proc. Natl. Acad.
Sci. USA 91,5012-5016.
Liang, C., and Stillman, B. (1997). Persistent initiation of DNA replication and
chromath-bound MCM proteins during the ce11 cycle in cdc6 mutants. Genes Dev. 2 I,
337503386. Little, RD., and Schildkraut, CL. (1995). initiation of latent DNA replication in the
----% ----% -2 d - - Epstein-Barr virus genome cm occur at sites other than the gmetically defined origin.
Mol. Cell. Biol. f 5,2893-2903.
Lupton, S., and Levine, AJ. (1985). Mapping of genetic elements of Epstein-Barr virus
that facilitate extrachromosornai persistence of Epstein-Barr virus-derived p lasrnids in
human cells. Mol. Cell. Biol. 5,2533-2542.
Mackey, D., and Sugden, B. (1 999). The linking regions of EBNA 1 are essential for its
support of replication and transcription. Mol. Cell. Biol. 19,3349-3358.
Marcus, SL., Miyata, KS., Rachubinski, RA. And Capone, JP. (1995). Transactivation by
PPAWRXR heterodimers in yeast is potmtiated by exogenous fatty acid via a pathway
requiring intact peroxisomes. Gene Expression 4,227-239.
Marechal, V., Dehee, A*, Chikhi-Brachet, R., Piolot, T., Coppey-Moisan, M., and
Nicolas, JC. (1999). Mapping EBNA-1 domains involveci in binding to metaphase
chromosomes. J. Virol. 73,4385-4392.
McBride, A., Romanczuk, H., and Howley, P. (1991). The papillomavirus E2 regulatory proteins. J. Biol. Chem. 266, 1841 1 - 18414.
McGrew, J., Diehl, B., and Fitzgerald-Hayes, M. (1986). Single base-pair mutations in centromere elment III cause aberrant chromosome segregation in Saccharomyces ------.-- A- . - - cerevisiae. Mol. Cell. Biol. 6,530-538.
Merker, R., and Klein, Hannah. (1998). The effect of hrpl deh on piasmid stability in
Saccharomyces cerevisiae. In Yeast Genetics and Molecular Biology Conference
Abstract, College Park, Maryland.
Middleton, T., and Sugden, B. (1992). EBNAl cm link the damerelement to the
initiatot element of the Epstein-Barr virus plasmid ongin of DNA replication. Je Virol.
66,489-495,
Middleton, T., and Sugden, B. (1994). Retention of plasmid DNA in mammalian cells is
enhanced by binding of the Epstein-Barr virus replication protein EBNAl. J. Virol. 68,
4067-407 1.
Mumberg, D., Muller, R., and Funk, M. (1994). Regulatable promoters of Saccliaromyces
cerevisiae: Cornparison of transcriptional activity and their use for heterologous
expression. Nucl. Acids Res. 22,5767-5768.
Murphy, MR., Fowlkes, DM., Fitzgerald-Hayes, M. (1991). Analysis of centromere
huiction in Saccharomyces ccrwisiae using synthetic centmmere mutants. Chromosorna
101,189-197. Murray, AW., and Smstak, JW.(1983). Pedigree analysis of plamnid segregation in
. . .- -A - .. . - - --- yeast. Ce11 34,961-970.
Newlon, CS., and Theis, JF. (1 993). The stnicture and fwiction of yeast ARS elements.
Cm. Opin. Genet. Dev. 3,752-758.
Niedenthal, R., Stoll, R., and Hegemann, JH. (1991). In vivo characterization of the
Saccharomyces cerevisiae centmmere DNA element 1, a binding site for the helix-loop-
helix protein, CPFI. Mol. Cell. Biol. 11,3545-3553.
Norio, P., Schilàkraut, CL., and Yates, JL. (2000). Initiation of DNA replication within
oriP is dispensable for stable replication of the latent Epstein-Ban virus chromosome
derinfection of established ce11 lines. J. Virol. 74,8563-8574.
Peterson, JB., and Ris, H. (1976). Electron-microscopie study of the spindle and
chromosome movernent in the yeast Saccharomyces cerevisiae. Je Ce11 Sci. 22,2 19-242.
Petti, L., Sample, C., and Kieff, E. (1990). Subnuclear locaiization and phosphorylation
of Epstein-Barr virus latent infection nuclear proteins. Virology 1 76,563-574.
Piirsoo, M., Ustav, E., Mandel, T., Stenlund, A., and Ustav, M. (1996). Cis and tram
requirements for stable episomal maintenance of the BPV-1 replicator. EMBO J. 15, 1-
11. Piolot, T., Tramier, M., Coppey, M., Nicolas, JC., Marechal, V. (2001). Close but distinct
-,-----.- =------2 ------.. - regions of human herpesWus 8 latency-associated nuclear antigen 1 are responsible for
nuclear targeting and binding to huaian mitotic chromosomes. J. Virol. 75,3948-3959.
Polvino-Bodnar, M., and Schser, PA. (1992). DNA binding activity is required for EBNAl -dependent transcnptional activation and DNA replication. Virology l87,59 1- 603.
Pugeilli, MT., Woisetschlaeger, M., and Speck, SH. (1996). oriP is essential for EBNA
gene promoter activity in Epstein-Barr virus-immortalized lymphoblastoid ce11 iines. I.
Virol. 70,5758-5768.
Rawlins, DR., Milman, G., Hayward, SD., and Hayward, GS. (1 985). Sequence specific
DNA binding of the Epstein-Barr virus nuclear antigen (EBNA-1)to clustered sites in the
plamid maintenance region. Ce11 42,859-868.
Reisrnan, D.,Yates, I., and Sugden, B. (1985). A putative origin of replication of
plasmids derived hmEpstein-Barr virus is composed of two cis-acting components.
Mol. Cell. Biol. 5, 18224832.
Reisman, D., and Sugden, B. (1986). Tramactivation of an Epstein-Barr viral
transcnptional enhancer by the Epstein-Barr viral nuclear antigen 1. Mol. Cell. Biol. 6,
3838-3846. Rickinson, AB., and Kie@ E. (1996). Epstein-Barr Wua In Fields Virology, BN. Fields,
---3 --A- -. - . ------DM. Knipe and PM. Howley, eds. (Philadelphia: Lippincott-Ravm Publishers), pp. 2397-
Rothstein, R. (1991). Targeting, disruption, replacement and allele rescue: Integrative
DNA transformation in yeast. Meth. Enzomol. 194,281-301.
Sample, J., Henson, EBD.,and Sample, C. (1992). The Epstein-Ban nuclear antigen 1
promoter active in type 1 latency is autoregulated. J. Virol. 66,465404661.
Schaefer, BC., Woisetschlaeger, M., Strorninger, K., and Speck, SH. (1 991). Exclusive
expression of Epstein-Barr virus nuclear antigen 1 in Burkitt lymphoma arises hma
third promoter, distinct hmthe promoters used in latently infected lymphocytes. Proc. Natl. Acad. Sci. USA 88,6550-6554.
Schaefer, BC., Strominger, K., and Speck, SH. (1995). Rodefining the Epstein-Barr
virusncoded nuclear antigen EBNAl geae promoter and transcription initiation site in
group 1 Burkitt lymphoma ce11 lines. Roc. Natl. Acad. Sci. USA 92, 10565-10569.
Scott-Drew, S., and Mmy, JA. (1 998). Localisation and interaction of the protein
components of the yeast 2 mu circle plasmid partitionhg system suggest a rnechanism for
plasrnid inheritance. J. Ce11 Sci. 111,1779- 1789. Seo, Y., Muller, F., Lusiry, M., and Hurwitz, J. (1993). Bovine papillomaviw (BPV)-
- . - ---%A=.A .: ------encoded El protein contains multiple activities requucd for BPV DNA repücation. Proc.
Natl. Acad. Sci. USA 90,702-706.
Shah, WA., Ambinder, RF., Hayward, CS., and Hayward, SD. (1992). Binding of
EBNA-1 to DNA creates a protease-resistant domain that encompasses the DNA
recognition and dimerization fiinctions. J. Virol. 66,33553362,
Sherman, F., Fink, GR., and Lawrence, CW. (1979). Methoâs in Yeast Genetics. Cold
Spring Harbor Laboratory Press, New York.
Shire, IL, Ceccmlli, DFJ., Avolio-Huuter, TM., and Frappier, L. (1999). EBPZ, a human
protein that interacts with sequences of the EpstehBarr virus nuclear antigen 1 important
for plasmid maintenance. J. Virol. 73,2587-2595.
Sikorski, RS., and Hieter, P. (1 989). A systern of shuttle vectors and yeast host strains
designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics 122,
19-27.
Simpson, K., McGuigan, A., and Huxley, C. (1996). Stable episomal maintenance of
yeast artificial chromosomes in human cells. Mol. Cell. Biol. 16.5 1 17-5126.
Skiadopoulos, MH., and McBride, AA. (1998). Bovine papillomavirus type 1 genomes and the E2 transactivator protein are closely associateci with mitotic chromosomes. J.
---AL -- -- % - L- - 2 - Vhl. 72,2079-2088.
Snudden, DK., Hearing, JO, Smith, PR., Grasser, FA., and Griffin, BE. (1994). EBNA-1,
the major nuclear antigen of Epstein-Barr vims, resembles 'RGG' RNA binding proteins.
EMBO J. f 3,4840-4847.
Som, T., Armstrong, KA., Volkert, FC., and Broach, IR. (1988). Autoregulation of 2 pm
circle gene expression provides a mode1 for maintenance of stable plasmid copy levels.
Ce11 52,27937.
Snnivas, SK.,and Sixbey, W. (1 995). Epstein-Barr virus induction of recombinase-
activating genes RAGl and RAG2. J. Virol. 69,8 155.
Stinchcomb, DT., Shuhl, K., and Davis, RW. (1979). Isolation and charactenzation of a
yeast chromosomal replicator. Nature 283,3943.
Su, W., Middleton, T., Sugden, B., and Echols, H. (1991). DNA looping between the
ongin of replication of Epstein-Barr vims and its enhancer site: stabilization of an origin
complex with Epstein-Barr nuclear antigen 1. Roc. Natl. Acad. Sci. USA 88,10870-
10874.
Sugden, B., Marsh, K., and Yates, J. (1985). A vector that replicates as a plasmid and can be efficiently selected in B-lyrnphoblasts transfod by Epstein-Barr virus. Mol. Cell.
---A 2 L-- - . -.. - - -. - Biol. 5,41043.
Sugden, B., and Wanen, N. (1989). A prombter of Epstein-Barr virus that can function
during latent inféction can be tnuisactivated by EBNA-1, a viral protein required for viral
DNA replication during latent infection. J. Vhl. 63,26442649.
Summers, H., Barwell, JA., Pfiietzner, RA., Edwards, AM., and Frappier, L. (1 996).
Cooperative assembly of EBNAl on the Epstein-Barr virus latent origin of replication. J.
Virol. 70,1228-1231.
Thorley-Lawson, DA., Miyashita, EM., and Khan, G. (1996). Epstein-Barr virus and the
B cell: îhat's dlit takes. Trends Micmbiol. 5,204-208.
Tsujii, R., Miyoshi, K., Tsuno, A., Matsui, Y., Toh-e, A., Miyakawa, T., and Mizuta, K.
(2000). Ebp2p. yeast homologue of a human protein that interacts with Epstein-Barr virus
nuclear antigen 1, is required for pre-rRNA processing and ribosomal subunit assembly.
Genes Cells 5,543-553.
Van Scoy, S., Watakabe, I., Krainer, AR., and Hearing, J. (2000). Human p32: A
coactivator for Epstein-Barr vhsnuclear aatigea- 1-mediatecl transcriptional activation
and possible role in viral latent cycle DNA replication. Virology 275, 145- 157. Veimurugan, S., Ahn, YT., Yang, XM.,Wu, XL., Jayaram, M. (1 998). The 2 pplasmid
-- 3 A u .------r . - stability system: dysesof the interactions among plasmid- and host-encoded
components. Mol. Cell. Biol.l8,7466-7477.
Velmurugan, S., Yang, XM., Chan, CSM.,Dobson, M., and Jayaram, M. (2000).
Partitionhg of the 2-pn circle plasmid in Saccharomyces cerevisiae: functional
coordination with chromosome segregation and plasmid-encoded Rep protein
distribution. J. Ce11 Biol. 149,553-566.
Vokert, FC., Wu, LC., Fisher, PA., and Broach, JR. (1986). SuMval strategies of the
yeast plasmid two-micron circle. Basic Life Sci. 40,375-396.
Wang, Y., Finan, JE., Middeldorp, IM.,and Hayward, SD. (1997). P32/TAP, a cellular
protein that interacts with EBNA-1 of Epstein-Ban virus. Virology 236, 18-29.
Wilson, JB., Bell, JL., and Levine, AL (1 996). Expression of Epstein-Ban virus nuclear
antigen-1 induces B ce11 neoplasia in transgenic mice. EMBO J. I5,3 117-3126.
Woisetschaeger, M., Yandava, CN., Funnanski, J., Straminger, JL., and Speck, SH.
(1990). Promoter switching in Epstein-Barr virus durllrg the initial stages of infection of
B lymphocytes. Roc. Natl. Acad. Sci. USA 87,1725- 1729.
Wu, H., Ceccarelli, DFJ., and Frappier, L. (2000). The DNA segregation mechanism of Epstein-Barr Wus nuclear antigen 1. EMBO Rep.l,140-4.
--?------* -
Wu, LC., Fisher, PA., and Broach, JR. (1987). A yeast plasmid partitionhg protein is a
karyoskeletal component. J. Biol. Chem. 262,883-891.
Wysokenski, DA., and Yates, JL. (1989). Multiple EBNAl -binding sites are required to
form an EBNAl -dependent enhancer and to activate a minimal replicative origin within
or* of EpsteibBarr virus. J. Vhl. 63,265792666.
Yates, JL., Warren, N., Reisman, D., and Sugden, B. (1984). A cis-acting e1emer.t fiom
the Epstein-Barr viral genome that pennits stable replication of recombinant plasmids in
latently infected cells. Proc. Natl. Acad. Sci. USA 81,3806938 10.
Yates, IL., Warren, N., and Sugden, B. (1985). Stable replication of plasrnids derived
nom Epstein-Barr virus in various mammalian cells. Nature 313,8 12-815.
Yates, JL ., and Camiolo, SM. (1988). Dissection of DNA replication and enhancer
activation îùnctions of Epstein-Barr nuclear antigen 1. Cancer Cells 6, l97-2O!i.
Yates, JL., and Guan, N. (1991). Epstein-Barr virus derived plasmids replicate only once
per ce11 cycle and are not amplifid derentry into cells. J. Virol. 65,483-488.
Yates, JL., Camiolo, SM., Bashaw, M.(2000). The minimal replicator of Epstein-Barr
108 Zhang, D.,Frappier, L., Gibbs, E., Hunvitz, J., and O'Donnell, M. (1998). Human RPA
(hSSB) interacts with EBNAI, the latent ongin binding protein of Epstein-Barr virus.
Nucleic Acids Res. 26,63 1-637.