University of IowaIowa Research Online
Theses and Dissertations
Summer 2013
Studies of Proliferating Cell Nuclear AntigenMutant Proteins Defective in Translesion Synthesisand Mismatch RepairLynne Margaret DieckmanUniversity of Iowa
Copyright 2013 Lynne Margaret Dieckman
This dissertation is available at Iowa Research Online: https://ir.uiowa.edu/etd/4839
Follow this and additional works at: https://ir.uiowa.edu/etd
Part of the Cell Biology Commons
Recommended CitationDieckman, Lynne Margaret. "Studies of Proliferating Cell Nuclear Antigen Mutant Proteins Defective in Translesion Synthesis andMismatch Repair." PhD (Doctor of Philosophy) thesis, University of Iowa, 2013.https://doi.org/10.17077/etd.5pihvjjx
STUDIES OF PROLIFERATING CELL NUCLEAR ANTIGEN
MUTANT PROTEINS DEFECTIVE IN TRANSLESION SYNTHESIS
AND MISMATCH REPAIR
by
Lynne Margaret Dieckman
A thesis submitted in partial fulfillment of the requirements for the Doctor of Philosophy degree in Molecular and Cellular Biology in the Graduate College of The University of
Iowa
August 2013
Thesis Supervisor: Associate Professor M. Todd Washington
Graduate College The University of Iowa
Iowa City, Iowa
CERTIFICATE OF APPROVAL
PH.D. THESIS
This is to certify that the Ph.D. thesis of
Lynne Margaret Dieckman
has been approved by the Examining Committee for the thesis requirement for the Doctor of Philosophy degree in Molecular and Cellular Biology at the August 2013 graduation.
Thesis Committee: M. Todd Washington, Thesis Supervisor Marc Wold Kris DeMali John Dagle Jon Houtman
ii
ACKNOWLEDGMENTS
First and foremost, I would like to thank Mr. Joel Hutchison for playing an
integral role in establishing my interest and foundation in science when I was a high
school student. Under his instruction, I gained the knowledge and confidence necessary
to succeed as a college student and persevere as a graduate student. I would not be in
science today if it had not been for his constant direction and encouragement to pursue a
career path in chemistry. Even though his classes only lasted for four years of my
educational life, he has continued to be a cherished mentor since then and will continue to
be for the rest of my life.
I would like to thank all of my family and friends for their continued support and
reassurance throughout my graduate school career. I am extremely lucky to have such an
amazing group of people who have been there for me under any circumstance. Just to
name a couple, my mom has been the most loving and supportive parent that I could ask
for. And Tyson Shepherd has been a fantastic help in and out of the lab with all of our
entertaining and valuable discussions.
Lastly, I would like to thank my thesis advisor, Dr. M. Todd Washington, for
providing an environment adequate for me to learn and grow during both my Master’s
and Ph.D. work. His confidence in my abilities facilitated my development as a scientist,
and I would not be where I am today if it were not for his continued support and guidance
throughout the years. He was a fantastic and respected mentor during graduate school
and will always remain a valued friend.
iii
ABSTRACT
Proliferating cell nuclear antigen (PCNA) is a versatile protein involved in all
pathways of DNA metabolism. It is best known as a processivity factor for classical
polymerases, which synthesize DNA on non-damaged templates during DNA replication
(ex: pol δ). Non-classical polymerases, on the other hand, are those that synthesize DNA
on damaged templates (ex: pol η). PCNA also functions in repair, recombination, and
most other DNA-dependent cellular processes. A number of separation of function
mutant PCNA proteins have been identified, suggesting that PCNA could be a valuable
target to manipulate DNA metabolism. This thesis focuses on the study of PCNA mutant
proteins that affect translesion synthesis (TLS) and mismatch repair (MMR).
During TLS, the process by which DNA polymerases replicate through DNA
lesions, PCNA recruits and stabilizes polymerases at the replication fork. TLS requires
the monoubiquitylation of PCNA, and PCNA and ubiquitin-modified PCNA (Ub-PCNA)
stimulate TLS by classical and non-classical polymerases. Two mutant forms of yeast
PCNA, one with an E113G substitution and one with a G178S substitution, support
normal cell growth but inhibit TLS. To better understand the role of PCNA in TLS, I re-
examined the structures of both mutant PCNA proteins and identified substantial
disruptions of the subunit interface that forms the PCNA trimer. This resulted in reduced
trimer stability in the mutant proteins. The mutant forms of PCNA and Ub-PCNA do not
stimulate TLS of an abasic site by either classical pol or non-classical pol . Normal
replication by pol was also impacted, but normal replication by pol was much less
affected. These findings support a model in which reduced trimer stability causes these
mutant PCNA proteins to occasionally undergo conformational changes that compromise
their ability to stimulate TLS by both classical and non-classical polymerases.
During MMR, PCNA recruits and coordinates proteins involved in the mismatch
recognition, excision, and resynthesis steps. Previously, two mutant forms of PCNA
iv
were identified that cause defects in MMR with little if any other defects. These are the
C22Y and C81R mutant PCNA proteins. In order to understand the structural and
mechanistic basis by which these two substitutions in PCNA proteins block MMR, we
solved the X-ray crystal structures of both mutant proteins and carried out further
biochemical studies. I found that these amino acid substitutions lead to distinct structural
changes in PCNA. The C22Y substitution alters the positions of the -helices lining the
central hole of the PCNA ring, whereas the C81R substitution creates a distortion in the
-sheet at the PCNA subunit interface. I conclude that the structural integrity of the -
helices lining the central hole and the -sheet at the subunit interface are both necessary
to form productive complexes with MutS and mismatch-containing DNA.
As described above, my studies focused on four amino acid substitutions in
PCNA that disrupt TLS and MMR: the E113G and G178S substitutions cause defects in
TLS while the C22Y and C81R substitutions cause defects in MMR. The structures of
these mutant PCNA proteins revealed that three of the four substitutions caused
disruptions near the subunit interface of PCNA. To further examine the importance of
this region, we generated random mutations of the PCNA subunit interface and
performed in vivo genetics assays and in vitro biochemical assays to examine their effects
on TLS and MMR. We determined that the subunit interface of PCNA is very dynamic
and that small changes at this interface can cause drastically different effects on TLS and
MMR. Moreover, we suggest that the integrity of the subunit interface as well as the
nearby β-strands in domain A are crucial for proper PCNA function in vivo and in vitro.
v
TABLE OF CONTENTS
LIST OF TABLES ........................................................................................................... viii LIST OF FIGURES ........................................................................................................... ix LIST OF ABBREVIATIONS .......................................................................................... xiii CHAPTER 1 INTRODUCTION .........................................................................................1
DNA Metabolism and Carcinogenesis .............................................................................1 DNA replication, repair, and recombination ................................................................1 DNA mismatches and mistmatch repair .......................................................................5 Mutagenesis and translesion synthesis .........................................................................7
DNA Polymerases ............................................................................................................8 Overview of DNA polymerases ...................................................................................8 Classical DNA polymerases .........................................................................................9 Classical DNA polymerase δ ......................................................................................10 Non-classical DNA polymerases ................................................................................12 Non-classical DNA polymerase η ..............................................................................13
Eukaryotic MutS Homologues MSH2 and MSH6 .........................................................15 Structure and function of the MSH2 and MSH6 proteins .........................................15 Interactions with PCNA and other repair proteins .....................................................17
Proliferating Cell Nuclear Antigen.................................................................................19 Structure and function of PCNA ................................................................................19 Role of PCNA in regulating DNA metabolism ..........................................................21
Post-Translational Modifications of PCNA ...................................................................24 Overview of PCNA modifications .............................................................................24 Monoubiquitylation of PCNA ....................................................................................24 Polyubiquitylation of PCNA ......................................................................................28 Sumoylation of PCNA ................................................................................................29
Structures of PCNA Complexes .....................................................................................30 Insights from structures of PCNA bound to PIP peptides ..........................................30 Structures of PCNA bound to full-length proteins .....................................................33 Low resolution structures of PCNA complexes .........................................................34 Conclusions from structural studies with PCNA........................................................35
Interactions of Y-family Polymerases with PCNA and Ubiquitylated PCNA ...............36 Interactions with un-modified PCNA .........................................................................37 Interactions with ubiquitin-modified PCNA ..............................................................38
Polymerase Switching and the Tool Belt Model ............................................................40 Polymerase switching during translesion synthesis ...................................................40 The tool belt model of polymerase switching ............................................................41
Mutant PCNA Proteins ...................................................................................................43 Mutant PCNA proteins defective in translesion synthesis .........................................43 Mutant PCNA proteins defective in mismatch repair ................................................45
Thesis Overview .............................................................................................................46 CHAPTER 2 PCNA TRIMER INSTABILITY INHIBITS TRANSLESION
SYNTHESIS BY DNA POLYMERASE η AND BY DNA POLYMERASE δ ...........76
Abstract ..........................................................................................................................76 Introduction ....................................................................................................................77 Materials and Methods ...................................................................................................79
vi
Protein expression and purification ............................................................................79 DNA and nucleotide substrates ..................................................................................80 PCNA trimer stability assays......................................................................................80 Enzyme-linked immunosorbent assays ......................................................................81 Polymerase processivity assays ..................................................................................81 Polymerase activity assays .........................................................................................82
Results ............................................................................................................................83 The E113G and G178S mutant PCNA proteins have altered subunit interfaces .......83 Trimer stability of the E113G and G178S mutant PCNA proteins ............................84 Interactions of the E113G mutant PCNA protein with DNA polymerases ................86 Impact of the E113G mutant PCNA protein on the activity of pol η .........................86 Impact of the E113G mutant PCNA protein on the catalytic efficiency of pol η ......88 Impact of the E113G mutant PCNA protein on the activity of pol δ .........................89 Impact of the E113G mutant PCNA protein on the catalytic efficiency of pol δ .......90 Impact of the G178S mutant PCNA protein on the activity of pol δ .........................91
Discussion ......................................................................................................................91 Inhibition of TLS in yeast by mutant PCNA proteins ................................................91 Mechanism of TLS inhibition by mutant PCNA proteins ..........................................92 Inhibition of TLS is caused by PCNA trimer instability ............................................93 Model of TLS inhibition by PCNA trimer instability ................................................94
CHAPTER 3 DISTINCT STRUCTURAL ALTERATIONS IN PCNA BLOCK DNA MISMATCH REPAIR .......................................................................................113
Abstract ........................................................................................................................113 Introduction ..................................................................................................................114 Materials and Methods .................................................................................................116
Protein expression and purification ..........................................................................116 DNA and nucleotide substrates ................................................................................116 Crystallization of the C22Y and C81R mutant proteins ...........................................117 Data collection and structural determination ............................................................117 PCNA trimer stability assays....................................................................................118 Polymerase δ activity assays ....................................................................................118 Enzyme-linked immunosorbent assays ....................................................................118 Sedimentation assays ................................................................................................119
Results ..........................................................................................................................120 Structure of the C22Y mutant PCNA protein...........................................................120 Structure of the C81R mutant PCNA protein ...........................................................121 Stability of the mutant PCNA proteins .....................................................................121 Impact of the mutant PCNA proteins on DNA polymerase δ activity .....................122 Interactions of the mutant PCNA proteins with MutSα ...........................................123 Interactions of the mutant PCNA proteins with MutSα and DNA ...........................124
Discussion ....................................................................................................................125
CHAPTER 4 IDENTIFICATION AND CHARACTERIZATION OF RANDOM MUTATIONS OF THE PCNA SUBUNIT INTERFACE ..............................................142
Abstract ........................................................................................................................142 Introduction ..................................................................................................................142 Materials and Methods .................................................................................................146
Protein expression and purification ..........................................................................146 DNA and nucleotide substrates ................................................................................146 PCNA trimer stability assays....................................................................................147
vii
Polymerase activity assays .......................................................................................147 Results ..........................................................................................................................148
Genetic analyses of random PCNA interface mutant proteins .................................148 Trimer stability of the PCNA interface mutant proteins ..........................................149 Impact of the PCNA interface mutant proteins on the activity of pol η ...................150 Impact of the PCNA interface mutant proteins on the activity of pol δ ...................151 Structures of the PCNA interface mutant proteins ...................................................153
Discussion ....................................................................................................................154 CHAPTER 5 THE C-TERMINAL REGION OF DNA POLYMERASE η IS INTRINSICALLY DISORDERED AND REQUIRED FOR INTERACTION WITH PCNA AND MONOUBIQUITYLATED PCNA ................................................168
Abstract ........................................................................................................................168 Introduction ..................................................................................................................168 Materials and Methods .................................................................................................170
Protein expression and purification ..........................................................................170 Protein disorder prediction studies ...........................................................................171 Nuclear magnetic resonance spectroscopy ...............................................................171 Enzyme-linked immunosorbent assays ....................................................................172 Isothermal titration calorimetry experiments ...........................................................172 Crystallization of the Ub-PCNA-CTR fusion protein ..............................................173
Results ..........................................................................................................................173 The C-terminal region of pol η is intrinsically disordered .......................................173 Binding studies of the CTR of pol η with PCNA, ubiquitin, and Ub-PCNA ...........174 Crystallography studies of the complex of the CTR of pol η and Ub-PCNA ..........176
Discussion ....................................................................................................................177
CHAPTER 6 DISCUSSION ............................................................................................191
Integrity of the PCNA Interface during Translesion Synthesis ....................................192 Overview of studies with the E113G mutant PCNA proteins ..................................192 Role of Ub-PCNA during TLS .................................................................................193 Possible insights from studies with the E113G mutant PCNA and Ub-PCNA proteins ...................................................................................................194
Defects in Mismatch Repair Caused by Distinct Structural Alterations in PCNA ......195 Overview of studies with the C22Y and C81R mutant PCNA proteins...................195 How do the C22Y and C81R substitutions prevent productive PCNA complex formation during MMR? ..........................................................................................196
Importance of the Subunit Interface of PCNA for Translesion Synthesis and Mismatch Repair ..........................................................................................................198
Overview of studies with random PCNA interface mutant proteins ........................198 Shearing of the PCNA interface ...............................................................................200 Growth phenotypes of the S177G and G178S mutant PCNA proteins ....................201 PCNA-dependent processes and PCNA substitutions that disrupt them..................202
PCNA and Its Interactions with Intrinsically Disordered Proteins ..............................203 Overview of studies with the C-terminal region of pol η .........................................203 Regulation of PCNA interactions .............................................................................204
REFERENCES ................................................................................................................213
viii
LIST OF TABLES
Table 1.1. Classification of DNA polymerases .................................................................55 Table 2.1. Distances between potential hydrogen bond donor and acceptor
atoms at the PCNA subunit interface ................................................................97 Table 2.2. Percentage of PCNA proteins in the monomeric state as determined
by size exclusion chromatography ..................................................................101 Table 2.3. Processivity of pol η on non-damaged and damaged DNA ...........................105 Table 2.4. Steady state kinetics of nucleotide incorporation by pol η .............................107 Table 2.5. Processivity of pol δ on non-damaged and damaged DNA ...........................109 Table 2.6. Steady state kinetics of nucleotide incorporation by pol δ .............................111 Table 3.1. Data collection and refinement statistics ........................................................129 Table 3.2. Relative DNA synthesis by pol δ in the presence of PCNA
mutant proteins ................................................................................................136 Table 4.1. Summary of genetic studies with the PCNA interface mutant proteins .........159 Table 4.2. Distances between potential hydrogen bond donor and acceptor
atoms at the PCNA subunit interface ..............................................................165 Table 4.3. Summary of in vivo and in vitro studies with the PCNA interface mutant
proteins ............................................................................................................167
ix
LIST OF FIGURES
Figure 1.1. Model of DNA replication in eukaryotes ........................................................50 Figure 1.2. Common types of DNA damage .....................................................................51 Figure 1.3. Possible DNA mismatches ..............................................................................52 Figure 1.4. Model of mismatch repair in eukaryotes ........................................................53 Figure 1.5. Model of translesion synthesis .......................................................................54 Figure 1.6. Mechanism of DNA polymerization by polymerases ....................................56 Figure 1.7. Structure of the catalytic subunit (Pol3) from DNA polymerase δ
from yeast bound to DNA ..............................................................................57 Figure 1.8. Structure of DNA polymerase from the bacteriophage RB69........................58 Figure 1.9. Structure of DNA polymerase η from yeast ...................................................59 Figure 1.10. Structural model of the full-length pol η .......................................................60 Figure 1.11. Structure of the ubiquitin-binding zinc-finger (UBZ) of pol η .....................61 Figure 1.12. Structure of the human MutSα dimer bound to a G:T mispair ......................62 Figure 1.13. Structure of yeast PCNA ...............................................................................63 Figure 1.14. Structure of yeast PCNA bound to DNA .....................................................64 Figure 1.15. Structure of ubiquitin-modified PCNA ........................................................65 Figure 1.16. Overlay of the two positions occupied by ubiquitin in the crystal
structure of Ub-PCNA .................................................................................66 Figure 1.17. The potential “ubiquitin-switch” on PCNA .................................................67 Figure 1.18. Overlay of the structures of ubiquitylated and sumoylated PCNA ..............68 Figure 1.19. Structures of PCNA bound to PIP peptides ..................................................69 Figure 1.20. Structure of PCNA bound to FEN1 ...............................................................70 Figure 1.21. The structured and unstructured regions of Y-family polymerases ..............71 Figure 1.22. Structural models of the full-length Y-family polymerases ..........................72 Figure 1.23. Structural model of full length pol bound to
ubiquitin-modified PCNA ............................................................................73
x
Figure 1.24. The tool belt model of translesion DNA synthesis ........................................74 Figure 1.25. Structures of the G178S and E113G mutant PCNA proteins ........................75 Figure 2.1. The subunit interface of the wild-type and mutant PCNA proteins ...............96 Figure 2.2. Analysis of the wild-type and mutant PCNA proteins by native gel
electrophoresis ................................................................................................98 Figure 2.3. Analysis of the wild-type and mutant PCNA proteins by size exclusion
chromatography ..............................................................................................99 Figure 2.4. Stability of the wild-type and mutant Ub-PCNA proteins ............................102 Figure 2.5. Interaction of the wild-type and mutant PCNA and Ub-PCNA
proteins with pol .......................................................................................103 Figure 2.6. Processive DNA synthesis by pol in the presence of the wild-type
and mutant PCNA proteins ...........................................................................104 Figure 2.7. Steady state kinetics of pol in the presence of the wild-type
and mutant PCNA proteins ..........................................................................106
Figure 2.8. Processive DNA synthesis by pol in the presence of the wild-type and mutant PCNA proteins ...........................................................................108
Figure 2.9. Steady state kinetics of pol in the presence of the wild-type and
mutant PCNA proteins .................................................................................110 Figure 2.10. Processive DNA synthesis by pol δ in presence of the E113G
and G178S mutant PCNA proteins ...........................................................112 Figure 3.1. Structure of the C22Y mutant PCNA protein................................................130 Figure 3.2. Structure of the C81R mutant PCNA protein ................................................131 Figure 3.3. Analysis of the PCNA proteins by native gel electrophoresis ......................132 Figure 3.4. Analysis of the PCNA proteins by size exclusion chromatography..............133 Figure 3.5. Analysis of the C81R mutant PCNA protein by size exclusion
chromatography ...........................................................................................134 Figure 3.6. DNA synthesis by pol in the presence of the PCNA proteins ....................135
Figure 3.7. Interactions of the PCNA proteins with MutS ............................................137 Figure 3.8. Sedimentation analysis of the interactions of the PCNA
proteins with MutS and mistmatched DNA ...............................................138 Figure 3.9. Sedimentation analysis of the interactions of the PCNA
proteins with MutS and homoduplex DNA ..............................................140
xi
Figure 4.1. The amino acid residues that comprise the PCNA subunit interface ...........158 Figure 4.2. Analysis of the wild-type and mutant PCNA proteins by native gel
electrophoresis ..............................................................................................160 Figure 4.3. Analysis of the PCNA interface mutant proteins by size exclusion
chromatography ............................................................................................161 Figure 4.4. DNA synthesis by pol η in the presence of the PCNA mutant proteins .......162 Figure 4.5. DNA synthesis by pol on a non-damaged template in the presence
of the PCNA mutant proteins .......................................................................163 Figure 4.6. DNA synthesis by pol on a template containing an abasic site in the
presence of the PCNA mutant proteins ........................................................164
Figure 4.7. Structure of the V180A mutant PCNA protein ............................................166 Figure 5.1. The structured and unstructured regions of yeast pol η ................................179 Figure 5.2. The
1H-
15N heteronuclear single quantum coherence (HSQC)
spectrum of the CTR of pol η ......................................................................180 Figure 5.3. Analysis of the interaction of the PCNA and Ub-PCNA proteins
with pol using ELISA ..............................................................................181 Figure 5.4. Analysis of the interaction of the full-length pol and the
CTR of pol η with Ub-PCNA .......................................................................182 Figure 5.5. Analysis of the interaction of the CTR of pol with PCNA and
ubiquitin using ITC ......................................................................................183 Figure 5.6. Analysis of the interaction of the CTR of pol with
Ub-PCNA using ITC ....................................................................................184 Figure 5.7. The
1H-
15N HSQC spectra of the CTR of pol η bound to PCN ..................185
Figure 5.8. The
1H-
15N HSQC spectra of the CTR of pol η bound to Ub-PCNA ..........187
Figure 5.9. Production and purification of the Ub-PCNA-CTR fusion protein ..............189 Figure 5.10. Crystallization of the Ub-PCNA-CTR fusion protein ................................190 Figure 6.1. Positions of all PCNA substitutions used for studies ..................................207 Figure 6.2. Close-up of positions of all PCNA substitutions used for studies ...............208 Figure 6.3. Potential positions of the ubiquitin moiety on PCNA and
possible roles of these positions during TLS ................................................209 Figure 6.4. SAXS analysis of the E113G mutant Ub-PCNA protein .............................210 Figure 6.5. Non-denaturing gel electrophoresis of wild-type PCNA .............................211
xii
Figure 6.6. B-factors of the PCNA interface .................................................................212
xiii
LIST OF ABBREVIATIONS
8-oxoG – 7,8-dihydro-8-oxoguanine
BER – base excision repair
Can-R – canavanine-resistant
CTR – C-terminal region
dNTP – deoxyribonucleotide
DSB – double-strand break
ELISA – enzyme-linked immunosorbent assay
EM – electron microscopy
EXOI – exonuclease I
FAD – flavin adenine dinucleotide
FEN1 – flap endonuclease I
HSQC – heteronuclear single quantum coherence
HU – hydroxyurea
IDCL – interdomain connector loop
IDL – insertion/deletion loop
ITC – isothermal titration calorimetry
MCM – mini chromosome maintenance
MLH – MutL homologue
MMR – mismatch repair
MMS – methyl methanesulfonate
MSH – MutS homologue
NER – nucleotide excision repair
NHEJ – nonhomologous end joining
NMR – nuclear magnetic resonance
NTR – N-terminal region
PAD – polymerase associated domain
xiv
PAGE – polyacrylamide gel electrophoresis
PCNA – proliferating cell nuclear antigen
PIP – PCNA interacting peptide
Pol – polymerase
PUb-PCNA – polyubiquitylated PCNA
RFC – replication factor C
RPA – replication protein A
SAXS – small angle X-ray scattering
ssDNA – single-stranded DNA
SUMO – small ubiquitin-like modifier
TLS – translesion synthesis
TT dimer – thymine-thymine dimer
UBM – ubiquitin-binding motif
Ub-PCNA – monoubiquitylated PCNA
UBZ – ubiquitin-binding zinc-binding
UV – ultraviolet radiation
XP-V – xeroderma pigmentosum variant
α – alpha
δ – delta
ε – epsilon
ζ – zeta
η - eta
ι – iota
κ – kappa
1
CHAPTER 1
INTRODUCTION
(The section entitled “Structures of PCNA Complexes” has been published in
Dieckman, L.M., Freudenthal, B.D., and Washington, M.T. (2012) Subcell Biochem.
(Springer) 62: 281-299. The section entitled “Interactions of Y-family Polymerases with
PCNA and Ubiquitylated PCNA” is to be published in Pryor, J.M., Dieckman, L.M.,
Boehm, E.M., and Washington, M.T. (2013) Nucl. Acids Mol. Biol. (Springer) (In
press).)
DNA Metabolism and Carcinogenesis
DNA replication, repair, and recombination.
All living organisms inherit their genetic information in the form of DNA. In
order for each daughter cell to acquire a complete genome from generation to generation,
this material must be duplicated during every round of cell division. The process of
duplicating DNA is called DNA replication and involves the coordinated action of
numerous enzymes and proteins. Figure 1.1 shows a model of DNA replication in
eukaryotes. In brief, the process is initiated by the separation of the two complementary
DNA strands in an ATP-dependent fashion by the mini chromosome maintenance
(MCM) complex, the major DNA helicase in eukaryotic organisms. Any single-stranded
DNA generated during replication is protected and prevented from re-annealing by
replication protein A (RPA), a heterotrimeric single-stranded DNA binding protein that
binds DNA in a non-specific manner. From here, DNA synthesis begins bidirectionally,
and replication is divided into leading strand and lagging strand synthesis. All nucleotide
synthesis in cells is carried out in the 5ꞌ to 3ꞌ direction by enzymes called polymerases. In
eukaryotes, the polymerase that accounts for leading strand replication is polymerase ε
2
(pol ε). Lagging strand synthesis is more complex than leading strand synthesis due to
the inability of polymerases to add nucleotides to DNA in a 3ꞌ to 5ꞌ direction. To
overcome this obstacle, lagging strand synthesis is performed in short fragments and,
subsequently, the DNA fragments are joined together. This procedure is initiated by the
introduction of RNA primers by polymerase α (pol α), an RNA primase, and the
remaining DNA replication on this strand is then carried out by polymerase δ (pol δ).
Lagging strand synthesis is completed by the concerted actions of flap endonuclease 1
(FEN1) and a DNA ligase, which remove the RNA primer and seal the gap between the
resulting two fragments of DNA in an ATP-dependent manner. Polymerases are
recruited to the replication fork by a replication accessory factor called proliferating cell
nuclear antigen (PCNA), a homotrimeric sliding clamp that prevents polymerases from
dissociating from the DNA during replication. PCNA also interacts with and coordinates
the actions of all of these proteins involved in lagging strand synthesis at the replication
fork.
It is imperative that DNA be replicated accurately and efficiently to avoid any
type of alteration in the DNA sequence from one cell division to the next. However,
errors occur during each round of replication at a rate of 10-9
[1]. In addition, our
genome is constantly being attacked by byproducts of cellular processes as well as by
environmental agents. Common examples of DNA damage are shown in Figure 1.2.
Exposure to ultraviolet light causes adjacent thymine residues to covalently bond to one
another to create thymine dimers or (6-4) photoproducts [2]. Ionizing radiation also
causes numerous types of DNA damage, including deleterious double-stranded breaks
[3]. Chemical factors, present in both the environment and originating within the cell
frequently damage DNA as well. For example, oxygen free radicals that are released as
by-products of chemical reactions during metabolism in cells cause damage such as 7,8-
dihydro-8-oxoguanine (8-oxoG) [4, 5]. These lesions occur between 1000 and 2000
times per cell per day [6]. In addition to the lesions induced by damaging agents, other
3
types of damage result from spontaneous processes. For instance, uracil (which is
considered damage in the context of DNA) results from the deamination of a cytosine.
Similarly, an abasic site results from the hydrolysis of the N-glycosidic bond that attaches
the base to the sugar-phosphate backbone. This type of damage can occur at an
astounding rate of 10,000 sites per human cell per day [6, 7].
Fortunately, cells have evolved to deal with the presence of DNA replication
errors and lesions. In order to preserve their genetic information, they possess several
biological responses to DNA damage. Two major categories of pathways exist to repair
incorrect or damaged DNA: DNA repair and DNA damage tolerance. There are several
types of DNA repair mechanisms, including direct reversal of the damage, base and
nucleotide excision repair (BER and NER), mismatch repair, and double-strand break
repair, which includes both nonhomologous end joining (NHEJ) and homologous
recombination. The two main types of DNA damage tolerance pathways that exist are
translesion synthesis - the best studied DNA damage tolerance pathway – and the error-
free pathway. The focus of this thesis will be on both mismatch repair and translesion
synthesis and specific proteins involved in these pathways, which will be discussed
below.
Some lesions are repaired by direct reversal of the damaged base. One example
of this is the repair of ultraviolet light-induced damage by a single enzyme (photolyase)
existing in bacteria, plants, and animals. These enzymes utilize energy from light to
activate the flavin adenine dinucleotide (FAD) and break the covalent bonds of the
pyrimidine dimers. Photolyases, however, are not found in placental mammals, and so
these organisms rely on excision repair for removal of the lesion [8].
In general, excision repair requires that the damaged nucleotide(s) are removed,
that the resulting gap is filled using the opposite strand as a template, and finally that the
DNA is sealed by ligation. During BER, the damage is recognized and removed by a
DNA glycosylase [9]. This leaves an abasic site, which is then removed in two steps by
4
an AP endonuclease and an AP lyase [10, 11]. The resulting single nucleotide gap is
filled by a DNA polymerase and subsequently sealed by a DNA ligase [11-13]. Bulky
lesions that cause distortion in the DNA double helix are removed by NER [14]. NER
begins with recognition of the damaged base(s) by an enzyme complex and, unlike BER,
several DNA bases are removed on either side of the lesion, leaving a 27-29 nucleotide
gap. Similar to BER, the gap is filled by a DNA polymerase and sealed by a DNA ligase
[15].
Double-strand breaks (DSB) in DNA occur through several means. The main
cause of DSBs is exposure to ionizing radiation. In addition to this, reactive oxygen
species and chemotherapeutic agents that generate oxidative free radicals also cause
DSBs [16-18]. Moreover, when the replication machinery encounters certain forms of
DNA damage, the replication fork can collapse and form a DSB. These DSBs can be
extremely harmful to the DNA as they are capable of leading to chromosomal
rearrangements and, if left unrepaired, cell death.
Due to the extremely detrimental consequences of DSBs, it is imperative that cells
have a means of fixing them, which can occur through one of two methods – NHEJ and
homologous recombination. In humans, NHEJ is the preferred method, whereas
homologous recombination predominates in yeast. Homologous recombination is
initiated by the formation of a DSB. This is followed by resection, in which sections of
DNA at the 5ꞌ end of the break are processed to give 3ꞌ single-stranded tails. Resection
then stimulates strand invasion, where the single-stranded tails invade and align with
their complementary strands on the homologous chromosome and initiate DNA synthesis
by a DNA polymerase to fill in the degraded DNA [19, 20]. This type of DNA repair is
often accurate as it is able to use an existing complementary strand as a template for
nucleotide incorporation. Conversely, during NHEJ, the DNA ends of DSBs are directly
ligated together without the use of a homologous template. When perfectly compatible
overhangs are present after generation of the DSB, NHEJ accurately repairs the break.
5
When these overhangs are incompatible, however, NHEJ can lead to chromosomal
translocations, telomere fusion [21], and a gain or loss of nucleotides anywhere in the
genome.
DNA mismatches and mismatch repair.
Purine bases only form appropriate hydrogen bonding and structural geometry
when base-paired with their complementary pyrimidine. Inappropriate insertion of a
pyrimidine opposite another pyrimidine forms energetically unfavorable pairing because
the bases are not large enough to form the necessary hydrogen bonds. An example of this
situation is shown in Figure 1.3. Purine-purine mispairing, on the other hand, is
energetically unfavorable because the bases are too bulky and create steric clashes within
the double helix. Even though the incorporation of a mismatched base pair is highly
unfavorable, they are still produced frequently during DNA metabolism. DNA
mismatches arise from the inaccurate insertion, deletion, or mis-incorporation of
nucleotides during DNA replication, repair, and recombination [22, 23]. Like DNA
damage, mismatched base pairs must be removed to maintain genome integrity.
Recognition and repair of these mispairs is accomplished by mismatch repair (MMR)
proteins, thereby reducing mutation rates. Defects in MMR are most known for their
correlation with sporadic and hereditary human cancers, including hereditary non-
polyposis colorectal cancer [24-26].
The MMR pathway has been studied extensively in E. coli. In this system, the
homodimer MutS recognizes base-base mispairs and small nucleotide insertion/deletion
mispairs [25]. MutL, which also acts as a homodimer [27], then interacts with MutS and
enhances mismatch recognition [28]. This complex recruits and activates MutH in an
ATP-dependent manner. The process of MMR is initiated when MutH specifically nicks
the newly synthesized strand of DNA [25, 28]. The bacterial genome is hemimethylated
immediately after DNA replication, and thus the newly synthesized strand is defined by
6
the absence of methylation. This allows for the helicase UvrD to unwind the duplex
DNA from the nick towards the mismatch [29] and for subsequent degradation of the
single-stranded DNA by an exonuclease (EXOI, EXOX, EXOVII, or RecJ depending on
the location of the mismatch). The resulting gap is then resynthesized by DNA
polymerase III and sealed by DNA ligase [25, 30, 31].
The general scheme for which MMR occurs in eukaryotes is shown in Figure 1.4.
While MutS and MutL homologues are present in eukaryotes, no MutH homologue has
been identified. The eukaryotic MutS homologues MSH2, MSH3, and MSH6 exist as
heterodimers. Where the MSH2-MSH6 heterodimer (denoted MutSα) is the most
abundant and involved in the recognition of base-base mismatches and small
insertion/deletion loops (IDLs) [32-34], the MSH2-MSH3 heterodimer (denoted MutSβ)
is involved in the recognition and repair of longer IDLs [32, 34]. MutL homologues
MLH1, MLH2, MLH3, and PMS1 also function as heterodimers. The MLH1-PMS1
complex carries out the majority of MMR in yeast cells, while MLH1-MLH2 and MLH1-
MLH3 have been implicated in playing a minor role in the repair of mismatches [35, 36].
Like the MutL and MutS complex in E. coli, the MLH1-PMS1 heterodimer
interacts with either the MutSα or MutSβ complex at the site of the mismatched DNA
template [37, 38]. Although MMR in yeast and humans is not as well-defined as in
bacteria, the main proteins involved in strand degredation, resynthesis, and ligation have
been identified as exonuclease I (EXOI) [39-42], DNA polymerase δ [43], replication
protein A (RPA) [44, 45], replication factor C (RFC) [46, 47], high-mobility group box 1
[48], and DNA ligase I [49].
Unlike E. coli, eukaryotic DNA does not exhibit the presence of hemi-methylated
DNA to help discriminate strand specificity during MMR. It has been suggested that the
recognition of the newly synthesized daughter strand is determined by the DNA
replication processivity factor proliferating cell nuclear antigen (PCNA) [50, 51]. PCNA
is thought to play important roles in the initiation, excision, and DNA resynthesis steps
7
during MMR, but the specific details are not clear [51, 52]. Importantly, PCNA interacts
with both MutSα and MutSβ [53-55] and helps recruit these complexes to sites of
mismatches [56, 57].
Mutagenesis and translesion synthesis.
Most human cancers result from the accumulation of multiple somatic DNA
mutations [58, 59]. These mutations may result from damaged DNA or misincorporated
nucleotides that are not repaired by the cell. Despite the continuous efforts of several
types of repair mechanisms functioning in our cells, incorrect and damaged DNA bases
persist throughout the cell cycle. Consequently, the DNA replication machinery will
encounter these lesions. This will result in either replication fork collapse or, preferably,
promote DNA damage tolerance. The best understood DNA damage tolerance pathway
is translesion synthesis (TLS), which occurs in both prokaryotes and eukaryotes. During
TLS, the cell’s replication machinery has the unique ability to bypass the DNA lesion(s).
This process has a high risk of being mutagenic, however, as the enzymes involved in
TLS are often error-prone.
Classical polymerases, i.e. those that utilize non-damaged DNA as templates, are
blocked at sites of DNA damage because their stringent active sites cannot accommodate
the bulk and distortion of most lesions. For this reason, specialized polymerases have
evolved to allow bulky structural distortions in their active sites, and are therefore
capable of bypassing the damage [60]. Because of this distinct feature, these
polymerases are called non-classical polymerases. The tolerating nature of these
enzymes, however, also causes them to have reduced fidelity of nucleotide incorporation.
In effect, almost all DNA damage-induced mutations are due to errors associated with
TLS [61].
When a classical polymerase is confronted with a DNA lesion, the replication
fork stalls and triggers a switch to a non-classical polymerase. Progression through
8
damaged DNA by the process of TLS is shown in Figure 1.5. In eukaryotes, the non-
classical polymerases involved in TLS are polymerase η (pol η), polymerase ζ (pol ζ),
polymerase ι (pol ι), polymerase κ (pol κ), and the Rev1 protein. Depending on the
polymerase and the type of damage being replicated, the resulting synthesized DNA may
be either a correct base or a misincorporated base that will create a mutation in the
genome after further rounds of replication. Once the damage is passed, the classical
polymerase once again replaces the non-classical polymerase and continues the
replication process. During both normal DNA replication and translesion synthesis,
polymerases utilize PCNA as the accessory factor that anchors the enzymes to the DNA
substrate. PCNA stabilizes the polymerases on DNA during TLS, and it is thought that
monoubiquitylation of PCNA is the signal to initiate the switch from the classical to the
non-classical polymerase.
DNA Polymerases
Overview of DNA polymerases.
All deoxyribonucleic acid synthesis during DNA replication, repair, and
recombination is performed by a specialized set of enzymes called DNA polymerases.
There are seven different families of DNA polymerases, which are divided according to
sequence homology and function: A, B, C, D, X, Y, and RT [62-64] (see Table 1.1).
Regardless of the low sequence homology and divergent function from one family of
polymerases to another, nearly all polymerase structures determined to date have similar
overall features [65, 66]. Each catalytic core, also known as the polymerase domain, is
comprised of a fingers, palm, and thumb sub-domains that together bind the substrate
DNA and incoming nucleotide and catalyze its incorporation onto the primer strand. The
thumb sub-domain binds the primer-template DNA while the fingers sub-domain binds
and aligns the incoming nucleotide. The catalytic center and active site of the
9
polymerase domain resides in the palm sub-domain. It contains three aspartic acid
residues that are essential in coordinating two divalent ions and catalyzing
phosphodiester bond formation [67]. Excluding the polymerase domain, the architecture
of polymerases is quite diverse between families. This is likely due to the distinctive
functions, interactions, and even post-translational modifications required for proper
DNA synthesis by these enzymes. The unique structures and characteristics of the
polymerases utilized in this thesis will be discussed in more detail below.
Classical DNA polymerases.
Based on their DNA substrates, polymerases can generally be separated into two
categories: classical polymerases and non-classical polymerases. Classical polymerases
are typically found in the B family of polymerases. These enzymes are responsible for
the majority of DNA replication, repair, and recombination and use normal, non-damaged
DNA as their templates. The classical eukaryotic polymerases involved in replication of
the nuclear genome are DNA pol α, pol ε, and pol δ. Pol α is the primase that initiates
lagging strand synthesis by synthesizing short RNA and DNA primers at each origin of
replication. The remaining lagging strand synthesis and Okazaki fragment maturation is
performed by pol δ [68-71], while leading strand synthesis is accomplished by pol ε [72].
It is crucial that leading and lagging strand DNA replication be efficient and accurate.
Therefore, many of these classical polymerases contain a proofreading domain to ensure
increased accuracy during normal DNA synthesis. Both pol δ and pol ε possess 3ꞌ to 5ꞌ
exonuclease activity that significantly increases their fidelity of DNA replication (tens to
thousands fold above polymerases lacking this domain [73]). The proofreading domain
detects the insertion of an incorrect nucleotide and immediately excises it, allowing for
the subsequent incorporation of the correct nucleotide. This has also been shown to
occur in the presence of a correct base pair, in which case the polymerase will likely
resynthesize the same nucleotide which had been excised in the first place. In addition to
10
their exonuclease activity, classical polymerases have highly stringent active sites that
allow for considerably increased fidelity over other their non-classical counterparts. The
X-ray crystal structure of pol δ, for instance, shows that its high fidelity is achieved
because the shape of the active site allows only the conformations of correct Watson-
crick base pairs [74]. Pol δ incorporates nucleotides with an error frequency of 10-4
to 10-
5 [75], which is approximately 100 fold more accurate than the prototypical non-classical
DNA polymerase η (pol η), whose error frequency is 10-2
to 10-4
[76, 77].
Classical polymerases synthesize DNA at a surprisingly slow rate in the absence
of any replication accessory proteins, with a kpol of approximately 1 s-1
in the case of
classical pol δ [75]. In general, all polymerases use the same basic mechanism of DNA
polymerization (Figure 1.6). They first bind the primer-template DNA substrate (step 1)
followed by the incoming nucleotide to form a polymerase-DNA-dNTP ternary complex
(step 2) [78]. A phosphodiester bond is formed and pyrophosphate is released through
nucleophilic attack by the 3ꞌ hydroxyl group of the primer terminus on the α-phosphate of
the dNTP, (step 3). For non-processive DNA synthesis, the rate limiting step of the
polymerase reaction is the dissociation of the polymerase from the DNA substrate (step
4) [79]. (Alternatively, if the next nucleotide to be inserted is present, the polymerase
may translocate forward by a single nucleotide for another round of nucleotide
incorporation.) The mechanism of DNA polymerization is very complicated, and it is
likely that many steps in the mechanism are actually composites of multiple elementary
steps themselves, such as conformational changes in the polymerase.
Classical DNA Polymerase δ.
Classical pol δ is essential for synthesis of the lagging strand of genomic DNA
during normal replication [68-71]. In the absence of pol ε, pol δ can substitute and
perform leading strand synthesis [68, 80]. Conversely, pol δ function is not be rescued
by pol ε. Pol δ not only functions in normal DNA replication and translesion synthesis,
11
but unlike pol ε, it also plays a fundamental role in homologous recombination, BER,
NER, MMR, and DSB repair, thereby demonstrating the extraordinary versatility of the
enzyme [81, 82].
Saccharomyces cerevisiae pol δ is a heterotrimeric protein consisting of a catalytic
subunit denoted Pol3 (125 kDa) and two accessory subunits Pol31 (55 kD) and Pol32 (40
kD). Pol3 possesses both polymerase and 3ꞌ-5ꞌ exonuclease activities [83], and the
protein encoded by the POL3 gene is highly conserved in all eukaryotes [84]. Pol3 and
Pol31 are both essential for viability, whereas Pol32 is nonessential [85]. This is
surprising, however, since Pol32 contains the interacting motif for PCNA, the replication
accessory factor necessary for polymerase recruitment to replication forks. It should also
be noted that deletion of the POL32 gene causes a defect in TLS in yeast [86], although
how this occurs is unclear.
The structure of the catalytic subunit of pol δ, Pol3, bound to DNA has recently
been determined [74] (Figure 1.7). The palm subdomain is made up of six β-strands that
constitute a β-sheet bordered by N-terminal and C-terminal α-helices. This subdomain
houses the Asp-608 and Asp-764 active site residues and interacts with the primer-
terminal end of the DNA. The fingers subdomain contains two antiparallel α-helices that
bind the incoming nucleotide, which is done in a manner consistent with that observed in
other polymerase ternary complexes [66]. This shifts the fingers and palm subdomains
closer together to catalyze the incorporation reaction. The thumb subdomain consists of
two smaller subdomains - the base and the tip – that together interact with the duplex
portion of the DNA substrate. In the crystal structure, the exonuclease domain contains a
single Ca2+
ion that is surrounded by the active site Asp-321, Glu-323, and Asp-407
residues.
Another classical polymerase whose X-ray crystal structure has been determined
is the bacteriophage RB69 polymerase (Figure 1.8). Despite their small degree of
sequence homology, the catalytic subunits of pol δ and the RB69 polymerase are quite
12
similar, with an overall r.m.s. deviation of 2.50 Å and an r.m.s. deviation of 1.71 Å
between the structurally conserved palm subdomains. The catalytic core domains of
these two classical polymerases consist of five separate domains – the palm, fingers,
thumb, exonuclease, and N-terminal domains - to create a circular structure with a central
cavity. The presence of these additional domains in classical polymerases is the basis for
their selective nucleotide discrimination and insertion. The importance of the
exonuclease domain in classical polymerases is apparent based on studies performed in
which mice homozygous for exonuclease-deficient pol δ prematurely developed several
types of cancer [87]. Moreover, in vivo measurements of cellular mutation rates showed
that exonuclease-deficient S. cerevisiae pol δ generated base substitution errors at rates
60 fold higher than those produced by wild type pol δ.
Non-classical DNA polymerases.
The architecture of classical polymerase active sites only tolerates the geometry
of correct Watson-Crick base pairs. Due to the low size constraints of their active sites
and ability to accommodate distortions in DNA, non-classical polymerases have the
unique ability to replicate through damaged DNA templates. They are utilized
exclusively during lesion bypass instead of bulk replication during normal DNA
replication, recombination, and repair like classical polymerases. This is fortunate, as
these non-classical polymerases have significantly lower fidelities than classical
polymerases and are therefore more error-prone when synthesizing from the primer-
terminus. Consequently, non-classical polymerases also replicate DNA with low
processivity, i.e. they only incorporate a few nucleotides per DNA-binding event.
The eukaryotic non-classical polymerases involved in TLS are pol η, pol ζ, pol ι,
pol κ, and the Rev1 protein [60]. Each of these polymerases has evolved to efficiently
replicate through specific lesions, and these are referred to as cognate lesions for that
particular polymerase. Pol η’s cognate lesions are ultraviolet (UV) photoproducts, most
13
notably thymine-thymine dimers [88] and 8-oxoguanines [89]. Pol ι efficiently
incorporates opposite minor groove purine adducts and exocyclic purine adducts [90-92].
Pol κ inserts opposite minor groove guanine adducts such as benzo[a]pyrene guanines
[93-96]. The cognate lesions for Rev1 are abasic sites [96] and minor groove and
exocyclic guanine adducts [97-100]. In addition to efficient synthesis opposite specific
lesions, pol ζ and pol κ are also efficient extenders from particular types of damage. In
most instances, though, this extension is performed by pol ζ.
The overall protein structures of non-classical polymerases are generally similar
to one another. The catalytic core of these enzymes contains a polymerase domain as
well as a polymerase-associated domain (PAD) that is unique to non-classical
polymerases. Like classical polymerases, the polymerase domain of non-classical
polymerases is comprised of fingers, thumb, and palm subdomains. In most cases, a
large instrinsically disordered region resides C-terminal to the catalytic core region [101].
These C-terminal regions (CTRs) are crucial for protein-protein interactions during TLS,
namely with PCNA and ubiquitin. Intrinsically disordered regions in proteins have
gained much interest recently, as the high prevalence of these interactions is becoming
apparent, and especially among PCNA-interacting partners. Interactions between
unstructured regions and their intended binding partners are dynamic to allow a rapid
response in vivo, and the flexibility of these interactions are required for the vast and
diverse array of activities necessary for these proteins to function. The specific
interactions between the CTRs of polymerases and PCNA will be discussed in detail
below.
Non-classical DNA polymerase η.
Pol η is the prototypical non-classical polymerase and functions in the replication
of UV-damaged DNA [102] with relatively high efficiency but low fidelity [88, 103]. It
synthesizes DNA with low processivity as well, incorporating only a few nucleotides
14
before dissociating from the DNA template [76]. Mutations in the gene encoding pol η,
RAD30, result in an increase of UV-induced mutations in yeast and humans [102], and
inactivation of pol η in humans causes the variant form of the autosomal recessive
genetic disorder xeroderma pigmentosum [102]. This disease is characterized by
sensitivity to UV light and predisposition to UV-induced cancers [104, 105].
Steady state and pre-steady state kinetic studies have shown that the cognate
lesions for pol η are thymine-thymine dimers and 8-oxoguanines. Pol η incorporates two
adenines opposite the two thymines of the dimer as efficiently as opposite two non-
damaged thymines [77, 106, 107], and it inserts the correct cytosine opposite an 8-
oxoguanine as efficiently as it does opposite a non-damaged guanine [89, 108]. In
contrast, classical polymerases prefer to incorporate adenine bases opposite 8-
oxoguanines due to the fact that the structure of this base pair more closely resembles the
conformation of Watson-Crick base-pairs. Interestingly, pol η is very poor at replicating
through abasic sites because the absence of a templating base poses a large kinetic barrier
to pol η compared to its cognate lesions [109].
Valuable information regarding catalytic activity of pol η was acquired from
crystallography studies of the catalytic core region of pol η bound to a damaged or non-
damaged DNA substrate and an incoming nucleotide [110-112] (Figure 1.9). The DNA
template makes contacts with all three subdomains of the polymerase domain as well as
the PAD. The fingers subdomain of pol η is smaller than in classical polymerases, which
causes the active site of pol η to be larger. This variation in size between classical
polymerases and pol η is what allows the non-classical polymerase to accommodate the
thymine-thymine (TT) dimer in its active site. This additional capacity also permits
normal Watson-Crick base-pairing between the incoming adenines and the two cross-
linked thymine bases. Finally, the structure of the active site also tolerates the distortion
of the TT dimer in a way such that the damaged DNA substrate is in a similar
conformation as when the DNA is non-damaged [110]. Taken together, this
15
demonstrates how pol η is able to synthesize nucleotides opposite its cognate lesions as
accurately and efficiently as it does with non-damaged templates.
Recruitment and regulation of pol η to sites of DNA damage is mediated through
the CTR of the protein. Although this region is not required for catalytic function in
vitro, it is essential for proper localization and catalytic function of pol η in vivo [113].
As with most non-classical polymerases, the CTR is found directly following the
catalytic core and is intrinsically disordered, and was therefore not observed in the crystal
structure of pol η. Figure 1.10 shows a representation of full length pol η in solution and
how the unstructured CTR would compare proportionally to the structured catalytic core
in the context of the full-length protein. Based on disorder predictions (discussed in
below), the CTR is unlikely to contain any structured domains besides the two folded
motifs known to be involved in protein-protein interactions – the ubiquitin-binding zinc-
finger (UBZ) and the PCNA interacting peptide (PIP) motifs. The structures of the UBZ
and PIP motifs have been determined using nuclear magnetic resonance (NMR) and X-
ray crystallography, respectively [114, 115]. The structure of the UBZ of human pol η is
shown in Figure 1.11; it adopts a classical ββα fold seen in other C2H2 zinc-finger
structures. Interactions between the UBZ and PIP motifs and PCNA and how these
interactions regulate pol η at the replication fork will be described in more detail below.
Eukaryotic MutS Homologues MSH2 and MSH6
Structure and function of the MSH2 and MSH6 proteins.
The MMR pathway is crucial in preserving DNA integrity and maintaining a
reduced rate of mutagenesis in an organism’s genome. This pathway is conserved from
bacteria to humans, and it is best characterized in the E. coli system. Two of the major
essential proteins involved in MMR in this system are MutS and MutL. In yeast and
humans, homologues of both MutS and MutL have also been identified to play
16
instrumental roles in MMR. The MutS homologues MSH2, MSH3, and MSH6 exist in
both yeast and humans, whereas yeast contain the MutL homologues MLH1, PMS1,
MLH2, and MLH3, and humans contain the MutL homologues MLH1 and PMS2 [25,
32, 33, 51, 116-120]. Several combinations of these proteins are formed in vivo as
heterodimeric complexes depending on the type of mismatch: MutSα (MSH2 + MSH6),
MutSβ (MSH2 + MSH3), and MutLα (MLH1 + PMS2). The MutSα complex, the most
abundant of these heterodimers, recognizes DNA mispairs and short insertion deletion
loops (IDLs) while MutSβ binds larger IDLs. Human MSH6 is intrinsically unstable and
requires dimerization with MSH2 for proper function. As a result, 80-90% of the MSH2
in cells is found in complexes with MSH6 [121].
Crystallographic studies of the MutSα heterodimer show that the structures of
MSH2 and MSH6 are generally similar to their bacterial MutS homologue. Each protein
is divided into five conserved domains, with MSH6 containing an additional disordered
N-terminal tail region [122, 123]. The X-ray crystal structure of the MSH2-MSH6 dimer
bound to ADP and a G:T mismatch is shown in Figure 1.12. Domain 1 (the mismatch
recognition domain) is the domain that recognizes and binds the DNA mismatch.
Domains 2 and 3 (the connector and core domains, respectively) comprise the connector
and lever that connect domain 1 to domains 4 and 5. Domain 4 is the clamp domain that
binds non-specifically to the rest of the DNA substrate. Domain 5 possesses ATPase
activity in that it binds adenosine and confers subsequent ATP hydrolysis. Lastly, the
helix-turn-helix (HTH) domain is involved in dimer contacts between the two monomers
of MutSα [122-124].
Asymmetric binding of MutSα to DNA is important for detection and
directionality of repair during MMR, where domain 1 of MSH6 binds to mismatched
DNA with higher affinity than the corresponding domain in MSH2. This tighter binding
is achieved via an essential Phe-X-Glu motif that is not found in MSH2 or MSH3 [125].
The aromatic ring of the phenylalanine specifically recognizes the stereo-chemical
17
distortion in the DNA caused by mismatched nucleotides [122, 124, 126]. Furthermore,
interaction with the mispaired nucleotide is enhanced by hydrogen bonds formed between
it and the glutamate residue of the Phe-X-Glu motif [122, 123, 127]. Together, these
interactions render MSH6 extremely specific for binding mistmatched bases and cause it
to do so in an asymmetric manner. Based on this asymmetry, MSH6 drives the
directionality of MMR [128].
Interactions with PCNA and other repair proteins.
The N-terminus of MSH6 (residues 1-389 in human MSH6, residues 1-295 in
yeast MSH6) is predicted to be instrinsically disordered. Similar to the disordered
regions in polymerases, this region is flexible and involved in protein-protein interactions
that are crucial for MutSα function and MMR. One major binding partner of MutSα that
is necessary for appropriate MMR is PCNA, which is required during the initiation,
excision, and DNA resynthesis steps of MMR [51, 52]. Small angle X-ray scattering
(SAXS) studies carried out with the complex of PCNA and MutSα revealed that MutSα
binds tightly to PCNA via a PIP motif in its unstructured N-terminal region (NTR) [57].
The NTR does not gain structure upon binding to PCNA, suggesting that the NTR acts as
an extended tether between these two proteins to help localize MutSα to the vicinity of
PCNA without confining it to one position or in a rigid conformation. Both MutSα and
PCNA would then be capable of dynamic interactions with other proteins involved in
MMR while simultaneously contacting each other. For example, the additional space
created by the flexible tether of MutSα would allow the subsequent association and
transfer of MMR complexes to the site of the DNA mismatch. In support of this, the
MutSα-PCNA complex binds to random homoduplex DNA, but PCNA is released from
the complex upon encountering a mismatched pair [56].
Mutational analysis of the NTR of MSH6 demonstrated that an intact PIP motif is
required for proper MMR, as human MSH6 lacking the first 77 amino acids (containing
18
the PIP sequence) did not localize with PCNA to the replication fork [55]. In agreement,
recruitment of human MutSα to the DNA substrate was inhibited when PIP-binding sites
on the PCNA trimer were occupied by competitive inhibition [129]. Loss of the
interaction between yeast MutSα and PCNA had a profound effect on MMR as well, as
an increase in mutations was observed when the PIP sequence was removed in yeast
MSH6 [53, 130]. Overall, these studies suggest a central role for PCNA in MMR,
possibly by recruiting MutSα to sites of mismatched nucleotides and regulating its
downstream interactions and coordinated functions.
Depending on the type of DNA lesion and its surrounding sequence, MutSα binds
its DNA substrates with differing affinities [125, 131-135]. It has high affinity for
mismatches that also contain a modified base, such as O6-methylguanine base paired with
a thymine or an 8-oxoguanine paired with either a thymine or another guanine [136-139].
Remarkably, the MMR machinery repairs 8-oxoguanine:G and 8-oxoguanine:T
mismatches with similar efficiency as it repairs non-modified G:G and G:T mismatches
[139, 140]. In contrast, processing of G:A and 8-oxoguanine:A is not as efficient [139,
140].
In addition to MMR, MutSα is involved with several other DNA repair processes
such as base excision repair, transcription-coupled repair, and double-strand break repair
[133, 141, 142], and interestingly, it has also been implicated in highly error-prone
MMR-related activities. For instance, MSH2 and MSH6 are both implicated in somatic
hypermutation of C:G and A:T base pairs, which are highly mutated at Ig loci [143-146].
Likewise, humans who are deficient in pol η have normal frequencies of hypermutations
at these loci but fewer A:T mutations, implicating pol η in this process as well [147-149].
This is consistent with the observation that MutSα and pol η interact physically and that
MutSα stimulates the catalytic activity of pol η in vitro [150]. It is the interaction
between these proteins that likely stimulates synthesis of these error-prone mutations at
Ig loci.
19
Proliferating Cell Nuclear Antigen
Structure and function of PCNA.
Sliding clamps exist in eubacteria, archaea, and eukaryotes. Although the
sequence homology of sliding clamps is not conserved throughout the domains of life,
they are all conserved structurally and functionally. They all form ring-shaped
complexes with pseudo-hexametric symmetry, encircle DNA, and are capable of sliding
around on the DNA in both the 3ꞌ and 5ꞌ directions. While the native oligomeric state of
clamps varies from one branch of life to another – eubacterial clamps exist as
homodimers, eukaryotic and T4 bacteriophage clamps as homotrimers, and archaea
clamps as heterotrimers – the overall architecture of each clamp is similar. Each is
composed of two or three domains that together align in a head-to-tail fashion to form the
ring that encircles the DNA substrate. The general structure of all sliding clamps consists
of a DNA-contacting inner cavity made of positively charged α-helices surrounded by an
exterior surface made up of several β-sheets. Proliferating cell nuclear antigen (PCNA) is
the eukaryotic DNA sliding clamp and will be the main focus of this thesis.
PCNA was first described in 1978 as an autoantibody in sera that recognized a
nuclear antigen found in proliferating cells in systemic lupus erythematosis patients
[151]. Two years later, it was found to be differentially expressed during the cell cycle,
peaking during S-phase, [152] and its expression was associated with proliferation [153,
154]. Today, PCNA is appreciated as a vital player in all aspects of DNA metabolism,
including DNA replication, TLS, BER, NER, MMR, chromatin assembly and
remodeling, cell cycle control, sister chromatid cohesion, and prevention of sister-
chromatid recombination [155]. It acts as a scaffold to provide a docking platform for the
recruitment of enzymes and proteins involved in these nuclear processes. This
recruitment is mediated by an array of dynamic protein-protein interactions whose details
still remain to be elucidated.
20
The determination of the crystal structure of the homotrimeric form of yeast
PCNA provided valuable insight into understanding its multifaceted function in nuclear
processes. PCNA was found to be a closed circular ring with pseudo-hexagonal
symmetry with markedly similar shape, size, and architecture as its sliding clamp
homologues in other species (Figure 1.13) [156]. It contains an inner surface made up of
12 α-helices surrounded by an outer layer of 6 β-sheets that together comprise a circular
collar. Each monomer or subunit is approximately 29 kDa and consists of two
independent but topologically identical domains. The N-terminal domain (residues 1-
117) is denoted domain A and the C-terminal domain (residues 135-258) is denoted
domain B, and each is formed with two anti-parallel β-sheets and two α-helices that pack
against a hydrophobic region between the sheets. The two domains are linked together
by a long, flexible loop (residues 118-134) called the interdomain connector loop (IDCL).
The final trimer is assembled by head-to-tail interactions with domain A of one subunit
contacting domain B on the adjacent subunit. The three resulting subunit-subunit
interfaces are then each stabilized by an anti-parallel β-sheet, formed with β-strands from
domain A and β-strands from the adjacent domain B, which is held together by eight
hydrogen bonds.
Overall, the PCNA ring is approximately 80 Å in diameter and 30 Å wide with a
central hole approximately 35 Å in diameter. The α-helices in the inner surface of the
hole contain several lysine and arginine residues that provide a positively charged surface
to allow negatively charged DNA to remain in close proximity while also enabling
PCNA to slide freely back and forth along the DNA. The X-ray crystal structure of the
eukaryotic PCNA bound to DNA, shown in Figure 1.14, revealed that the DNA passes
through the central hole of the ring at an angle tilted 40o away from the axis of symmetry
[157, 158], which may be due to the mode by which PCNA moves along the DNA double
helix. Single-molecule studies have shown that PCNA travels along the DNA by two
possible methods [159]. The first and most common mode of action entails rotation and
21
translation along the helical pitch of duplexed DNA, while the second mode involves fast
translation but not tracking along the helical pitch. Although PCNA is capable of sliding
in both directions, it retains two distinct faces – the front face and the back face. The
front face is characterized by the presence of a hydrophobic pocket and the IDCL that act
as contact sites for most PCNA-protein interactions, several of which perform their
activities on DNA at the front face. The back face of PCNA has recently been
characterized as an active site for post-translational modifications, which serves as a
secondary platform for binding and recruiting enzymes to the replication fork [160, 161].
The protein interactions and post-translational modifications that PCNA undergoes will
be discussed in more detail below.
Since DNA does not exist as short linear fragments in vivo and PCNA exists
primarily as a trimer, loading of PCNA onto its DNA substrate requires the assistance of
an additional factor. Replication factor C (RFC) is present in the nucleus as the clamp
loader, which opens the PCNA ring and loads it onto DNA in an ATP-dependent manner.
RFC is a five-protein complex that recognizes the 3ꞌ end of the primer-template for site-
specific placement of PCNA. ATP binding is necessary to stabilize the PCNA-RFC
complex and for subsequent loading. Hydrolysis of ATP is achieved through complex-
DNA binding followed by RFC dissociation from PCNA [162]. PCNA is loaded on
double-stranded DNA in an orientation-dependent manner, with its front face closest to
the primer-terminus. This ensures that polymerases are positioned correctly on the DNA
for proper synthesis. The orientation of PCNA at the replication fork may also be the
only mode of discrimination between the parental and newly synthesized strands of DNA
in eukaryotes, which, as described later, plays an important role in the initiation of MMR.
Role of PCNA in regulating DNA metabolism.
PCNA is essential during normal DNA replication by classical polymerases as
well as lesion bypass by non-classical polymerases to secure these enzymes at the primer-
22
terminus. The interaction between PCNA and polymerases prevents polymerase
dissociation from the DNA, thereby increasing their efficiency and facilitating several
rounds of nucleotide incorporation. The ability of a polymerase to progress through
multiple nucleotide binding and incorporation events without dissociating from the DNA
is referred to as the processivity of the enzyme. Depending on the type of polymerase,
PCNA has the potential to stimulate processivity substantially [68, 163, 164]. For both
the leading and lagging strand synthesizers pol ε and pol δ, the presence of PCNA
increases their ability to stay associated with the DNA substrate over 100 fold compared
to the polymerases alone [165]. This high level of processivity is due to the increase in
stability of the polymerase-DNA complex. In vitro, calf thymus PCNA stabilizes the pol
δ-template-primer complex by three orders of magnitude [166]. In addition to enhancing
the activity of polymerases during DNA replication, PCNA also serves as a moving
platform for other protein factors involved in this process. It is required for the switch
from the primase, pol α, to a classical polymerase during initiation of leading strand
synthesis and all throughout lagging strand synthesis [68]. Finally, structural and kinetic
studies suggest that PCNA stimulates and coordinates the activities of FEN1 and DNA
Ligase I during the final steps of DNA replication [167, 168].
In addition to replication through both non-damaged and damaged DNA, PCNA
is required for all DNA repair pathways, including MMR, BER, and NER. Repair by
MMR entails recognizing and processing on the newly synthesized strand of DNA only,
which is accomplished by the MMR machinery and likely due to the presence and
orientation of PCNA. PCNA’s participation in later steps in the MMR pathway is also
apparent due to the fact that it directly interacts with MSH6, MSH3, MLH1, EXO1, DNA
Ligase I, and pol δ [47, 54, 55, 169, 170]. Along with human MutSα and RFC, PCNA
stimulates MutLα (MLH1 and PMS2) endonuclease activity to excise DNA that does not
already contain a 5ꞌ nick [46]. Mutational analysis of the PIP motif of MutSα showed
reduced 5ꞌ-directed repair synthesis. This defect was not seen in the synthesis step for 3ꞌ-
23
directed repair, suggesting that PCNA is only required for 5ꞌ-repair synthesis [122, 171].
Overall, PCNA’s role in MMR is similar to its involvement in DNA replication in that it
provides a docking site to regulate the stepwise recruitment of enzymes to the site of
mismatched DNA. Furthermore, it is increasingly apparent that PCNA is utilized
immensely throughout MMR – including the initial strand discrimination and
directionality stages as well as the excision and resynthesis steps. The role of PCNA in
BER and NER is less clear than in MMR, however, it seems to function in a way
analogous to the mistmatch repair pathway by coordinating DNA excision and
resynthesis.
Considering the intimate connection between PCNA and DNA, it is not surprising
that regulation of PCNA can have a dramatic impact on the cell cycle. The cyclin-
dependent kinase inhibitor, p21, is a tumor suppressor protein that functions to regulate
the cell cycle through binding and inhibiting the activity of cyclin-CDK2 or cyclin-CDK1
complexes. This protein also binds and regulates PCNA function during DNA
replication. Binding of p21 to PCNA inhibits interactions between the sliding clamp and
polymerases, thereby not allowing stimulation of DNA synthesis by PCNA [172, 173].
The X-ray crystal structure of the C-terminal region of p21 bound to human PCNA
revealed that this interaction inhibits polymerase binding to PCNA by direct competition
for the hydrophobic binding surface on the front face of PCNA. Likewise, PCNA-p21
complex formation inhibits PCNA interactions with pol δ and FEN1 [174, 175] and
functionally inhibits MMR [51], RFC ATPase activity [176], and many other processes.
This tight control over the association of various protein complexes by p21 illustrates the
importance and magnitude of PCNA-protein interactions throughout the expansive range
of cellular activities.
24
Post-Translational Modifications of PCNA
Overview of PCNA modifications.
Because of the vast array of PCNA-dependent processes, it is not surprising that
the cell has developed several means of coordinating PCNA-associated proteins at the
replication fork. One major mode of regulation comes from post-translational
modifications of PCNA. Studies show that it is subject to acetylation, phosphorylation,
and ubiquitin and ubiquitin-like modification. It is becoming increasingly apparent that
the specificity of PCNA for some of its binding partners is regulated by these
modifications [177-180]. For instance, monoubiquitylation of PCNA promotes TLS by
recruiting non-classical polymerases to sites of DNA damage and modification by the
small ubiquitin-like modifier (SUMO) inhibits unwanted recombination by recruiting
anti-recombinogenic helicases to replication forks. Acetylation and phosphorylation of
PCNA are not as well understood, but it has been suggested that acetylation and
deacetylation of PCNA is associated with promoting and preventing DNA replication by
controlling PCNA’s interactions with the classical polymerases pol δ and pol β [181]. It
appears as though phosphorylation of PCNA is also involved at sites of DNA synthesis,
but that this modification is correlated with PCNA’s interactions with cyclin D1 and
cyclin A [182]. This thesis will focus on the ubiquitin and SUMO modifications of
PCNA.
Monoubiquitylation of PCNA.
PCNA is monoubiquitinylated on lysine 164 in response to DNA damage by
sequential action of the ubiquitin-activiating enzyme E1, the E2 ubiquitin-conjugating
enzyme Rad6, and a RING-finger-conjugating E3 ubiquitin ligase Rad18 [183]. First, the
ubiquitin moiety is attached to the E1 protein in an ATP-dependent manner, followed by
transfer of the ubiquitin from the E1 enzyme to a cysteine residue of Rad6. Finally,
25
Rad18 acts as a bridging protein to bring Rad6 and PCNA together to transfer the
ubiquitin to K164 on PCNA. Genetic studies performed with yeast cells lacking either
Rad6 or Rad18 showed that these cells were extremely sensitive to ultraviolet (UV)
radiation and methyl methanesulfonate (MMS) exposure and defective in UV-induced
mutagenesis as well [184]. It was also determined that yeast deficient in pol η and pol ζ
are epistatic with yeast carrying a K164 mutation in PCNA. Together, these data indicate
that ubiquitylation on K164 of PCNA is crucial for efficient TLS by these non-classical
polymerases.
Ubiquitylation of PCNA is necessary for the recruitment and subsequent
activation of non-classical DNA polymerases in lesion bypass initiation. While these
polymerases specifically interact with ubiquitylated PCNA, it is possible that
monoubiquitylation also functions to prevent binding of other replication factors during
TLS. Moreover, it was recently identified that non-classical polymerases themselves are
prone to post-translational monoubiquitylation [185, 186]. Together, these events may be
a factor in regulating the concerted actions of non-classical polymerases during lesion
bypass. For instance, ubiquitylation of PCNA would recruit the first non-classical
polymerase to incorporate a nucleotide opposite the damaged base, followed by
ubiquitylation of the polymerase to recruit an extender polymerase. Or, conversely,
modification of the first enzyme after nucleotide insertion may inhibit its interaction with
PCNA and any time thereafter. No matter the course of action, PCNA and its
monoubiquitylation are crucial for proper management and stimulation of non-classical
polymerases throughout TLS.
Exactly when and how PCNA ubiquitylation occurs remains unclear. Currently,
it is believed that stalling of the replication fork and the accumulation of single-stranded
DNA (ssDNA) initiates monoubiquitylation [187, 188]. It has been proposed that fork
stalling allows exposure of long stretches of ssDNA because the helicase is still moving
forward to unwind duplexed DNA. The presence of ssDNA then facilitates binding of
26
RPA, the ssDNA binding protein, which is thought to directly interact with and recruit
Rad18 to these sites [188]. Rad18 then recruits Rad6 and together they ubiquitylate
PCNA. It also remains unknown whether initiation of TLS requires one or all three
subunits of PCNA to be ubiquitylated. In vivo experiments demonstrated that, following
DNA damage, pol η is associated with a form of ubiquitylated PCNA (Ub-PCNA) that
has all three monomers modified [189]. It would be interesting, however, to determine if
the presence of only one ubiquitin-modified subunit is sufficient for facilitating TLS.
Monoubiquitinylation of PCNA initiates the TLS pathway by recruiting one or
more non-classical polymerases to the stalled PCNA-DNA-classical polymerase complex
to bypass the DNA damage. The impact of ubiquitin-modified PCNA on the activity of
classical pol δ was determined to be similar to that of un-modified PCNA [190]. Studies
showing the effects of the addition of the ubiquitin to PCNA on non-classical
polymerases, however, have been conflicting. One group demonstrated that Ub-PCNA
stimulates the activity of pol η and Rev1 compared to unmodified PCNA, but had no
effect on the catalytic activity of pol ζ [191]. On the other hand, another group showed
that the monoubiquitylation of PCNA does not stimulate the activity of any of these non-
classical polymerases compared to unmodified PCNA [190]. This same group also
suggested that the presence of the ubiquitin does not enhance the binding affinity of these
polymerases for PCNA. Studies presented in this thesis, however, show that pol η’s
affinity for PCNA is substantially increased by the addition of ubiquitin (see Chapter 5).
This discrepancy may be due to a difference in the assays used in these investigations
with differing stringencies.
Our lab recently established a novel approach to producing large quantities of Ub-
PCNA protein without the need for purified ubiquitinating enzymes. This approach
involves splitting the PCNA monomer at the site of ubiquitylation and inserting ubiquitin
in-frame at this position fused by two glycines to mimic the canonical isopeptide linkage
to lysine 164. The two polypeptides used to overexpress Ub-PCNA are shown in Figure
27
1.15A. The first polypeptide, called the N fragment, consisted of residues 1-163 of
PCNA with an N-terminal FLAG-tag. The second polypeptide, called the C fragment,
consisted of residues 1-76 of ubiquitin followed by a two glycine linker to residues 165-
258 of PCNA with an N-terminal 6xHis-tag. These two polypeptides were then co-
expressed and self-assembled in vivo. From this, milligram quantities of
monoubiquitylated PCNA were produced and purified, which allowed biochemical
analysis and X-ray crystal structure determination of this protein.
The Ub-PCNA fusion protein produced in our lab was shown to behave as
expected for the native, modified protein in that it stimulated the activity of pol η in vitro
and it supported cell viability and TLS in vivo [160]. The crystal structure of Ub-PCNA,
determined to 2.8 Å, revealed that the addition of the ubiquitin moiety did not
significantly alter the conformation of PCNA (Figure 1.15B). Interestingly, the structure
also revealed ubiquitin in two distinct positions (Figure 1.16) [160]. The positions of the
ubiquitin moieties were similar, however, in that they both occupied the back face of the
PCNA ring and were both oriented the same way, separated by only 2.5 Å. These
moieties contacted PCNA within domain B of PCNA and at the canonical hydrophobic
binding surface of ubiquitin used to interact with most binding partners, centered on Leu-
8, Ile-44, and Val-70. Together, these studies suggest that Ub-PCNA facilitates the
recruitment of non-classical polymerases by forming a novel interacting surface, namely
for their ubiquitin binding domains to interact.
To determine if there are possible alternative positions of the ubiquitin moiety on
PCNA besides those observed in the crystal structure, SAXS and computational modeling
experiments were performed with the split-fusion Ub-PCNA protein [192]. These
experiments did, in fact, identify two additional positions. Computational molecular
modeling showed the ubiquitin at the side of the PCNA ring in the groove directly above
the subunit interface, while SAXS results indicated ubiquitin to be in a flexible
conformation distinct from the other two observed positions in the crystallography and
28
molecular modeling experiments. Together, these studies suggest that the ubiquitin on
PCNA is dynamic and able to adopt at least three positions - 1) the back face, 2) the side
of the ring, and 3) in a flexible, extended position - and that all of these conformations are
important for the recruitment and positioning of non-classical polymerases during TLS.
Polyubiquitylation of PCNA.
Similar to monoubiquitylation of PCNA, polyubiquitylation of PCNA (PUb-
PCNA) also occurs on K164 and requires both Rad6 and Rad18. Unlike
monoubiquitylation, however, polyubiquitylation also requires the RING-finger ubiquitin
ligase Rad5 and the heterodimeric E2 enzyme (Ubc13 and Mms2) that specifically
catalyzes K63-linked polyubiquitin chains [193]. It seems reasonable that the
monoubiquitylation of PCNA is a prerequisite for polyubiquitylation, but whether this is
correct remains speculative. This potential “ubiquitin-switch” from mono- to
polyubiquitylation is shown in Figure 1.17. PUb-PCNA promotes the error-free pathway
of DNA damage tolerance [183, 184]. This pathway is not well understood, but it is
believed to bypass DNA damage in an error-free fashion by using the non-damaged sister
strand as the template.
It is currently unclear as to how PUb-PCNA activates the error-free pathway.
Here, I will discuss four possible scenarios. First, PUb-PCNA may promote template
switching by facilitating reversion of the replication fork, possibly producing a chicken
foot type structure, and exposing the newly replicated non-damaged sister duplex strand
for recombination. Second, it could induce dissociation of PCNA from the polymerase
and the replication fork. In doing so, ssDNA gaps would be exposed and possibly filled
in by TLS or recominbation and a new PCNA molecule may be recruited to the newly
synthesized DNA without affecting normal DNA replication. Third, the presence of
K63-linked polyubiquitin chains might recruit specific factors to initiate error-free bypass
of damage. Lastly, PUb-PCNA may inhibit TLS by specific non-classical polymerases.
29
Binding studies with mammalian pol η and pol ι have demonstrated that they are able to
bind polyubiquitin chains [194], which might suggest that these polymerases are removed
from the PCNA molecule to interact with the polyubiquitin modification itself, thereby
preventing TLS by these polymerases. Novel approaches to producing PUb-PCNA
would greatly enhance our understanding of PCNA’s role in the initiation and
progression of the error-free pathway.
Sumoylation of PCNA.
Sumoylation of PCNA mostly occurs at the same lysine residue as ubiquitylation,
Lys-164 (Figure 1.17). SUMO modification at this site on PCNA in yeast recruits the
antirecombinogenic helicase Srs2 [195]. Srs2 disrupts Rad51 filament formation on
DNA, which prevents unwanted recombination during DNA replication that can lead to
detrimental chromosomal rearrangements. This, in turn, may promote the RAD6-
dependent processes that utilize ubiquitylated PCNA to bypass DNA damage. Hence,
SUMO and ubiquitin modifications on PCNA may act as switches between the DNA
replication and the DNA damage tolerance pathways. Indeed, some evidence suggests
that PCNA is hyper-sumoylated during S phase, but upon exposure to DNA damage, the
SUMO is switched to ubiquitin to initiate the DNA damage response [183, 184, 196].
Interestingly, Srs2 is not present in higher eukaryotes. Since sumoylation of PCNA still
occurs in these cells, however, it is likely that other antirecombinogenic enzymes exist in
these organisms to prevent unwanted recombination. PCNA is also sumoylated to a
lesser extent at residue Lys-127 [183, 197]. Lys-127 resides within the IDCL of PCNA,
which is the same location that binds the PIP motifs of target proteins. Therefore,
sumoylation of PCNA at Lys-127 is thought to prevent the binding of proteins containing
a PIP sequence.
Recently, the structure of SUMO-modifed PCNA at Lys-164 was determined in
our lab. As in the ubiquitylated PCNA structure, the SUMO modification resided on the
30
back face of the PCNA ring, but in a position distinct from that occupied by ubiquitin in
the crystal structure of Ub-PCNA. Whereas the ubiquitin was positioned almost straight
backward from the PCNA ring in monoubiquitylated PCNA, the SUMO was positioned
in an angle more radial from the PCNA axis in sumoylated PCNA (Figure 1.18). Similar
to ubiquitin-modified PCNA, though, addition of the SUMO moiety does not alter the
overall structure of PCNA. Based on these results and the results of the SAXS analysis
of Ub-PCNA, one would conclude that these modifications likely adopt multiple
conformations relative to the PCNA ring, and that they recruit their binding partners by
providing an additional binding site for these factors. This structural characteristic would
then allow ubiquitylated PCNA to recruit non-classical polymerases and sumoylated
PCNA to recruit Srs2 to the back or side of the PCNA ring without interfering with
ongoing processes at the replication fork near the front face of PCNA.
Structures of PCNA Complexes
Insights from structures of PCNA bound to PIP peptides.
PCNA interacts with many of the enzymes involved in DNA replication and
repair. Structural studies of PCNA bound to several of its binding partners have been
carried out and these have provided valuable insights into how PCNA interacts with these
proteins. Most proteins that bind PCNA do so through a conserved PIP motif [198-200].
The sequences of PIP motifs from several proteins are shown in Figure 1.19A. These
motifs usually interact with PCNA on a single subunit in a region between the two
domains near the IDCL. The canonical PIP motif contains eight amino acid residues. The
conserved glutamine of the PIP motif normally inserts into a small pocket in PCNA (Fig.
1.19B). The last five residues of the PIP motif, which include the conserved hydrophobic
residue (methionine, leucine, or isoleucine) and the two conserved phenylalanine or
tyrosine residues, form a 310 helix that binds in a large hydrophobic pocket between the
31
two domains and also contacts the IDCL. Generally, the structure of PCNA is not
changed upon the binding of PIP peptides; only small alterations in the structure of the
IDCL are observed.
PIP motifs are often thought to be a flexible tether that anchors the PCNA-binding
protein to PCNA. PIP motifs are often found at the C-termini of PCNA binding proteins,
such as classical DNA polymerase δ (the Pol32 subunit in yeast and the p66 subunit in
humans), non-classical DNA polymerase η, and p21. PIP motifs, however, can occur
elsewhere in the primary structure of the PCNA-binding proteins, including the N-termini
(such as DNA ligase I) and the interiors of the proteins (such as non-classical DNA
polymerase ι). Deletion of the PIP motif or mutations in its conserved residues can
significantly weaken or abolish PCNA interactions in vivo and in vitro. Thus, even
though the PCNA-PIP interactions involve rather small regions of these proteins, these
interactions are often necessary to recruit many enzymes to replication forks.
Classical DNA polymerases are responsible for synthesizing DNA during DNA
replication and DNA repair. They achieve high processivity by interacting with PCNA,
and this interaction is dependent on their PIP motifs. In humans, DNA polymerase δ is
composed of four subunits (p125, p66, p50, and p12). The catalytic activity resides in the
p125 subunit. DNA polymerase δ interacts with PCNA via the PIP motif on the p66
subunit. The X-ray crystal structure of PCNA bound to the PIP peptide of p66 shows that
the PIP motif forms the normal 310 helix that fits into the large hydrophobic pocket of
PCNA [201].
Upon encountering DNA damage in the template strand, the replication fork
stalls. This is because classical DNA polymerases are unable to incorporate nucleotides
across from damaged DNA templates. Non-classical DNA polymerases, such as DNA
polymerases η, κ, and ι, are recruited to stalled replication forks to carry out translesion
synthesis [60, 202, 203]. The recruitment of these non-classical DNA polymerases is
governed in part by the monoubiquitylation of PCNA; this aspect of non-classical
32
polymerase recruitment will be described later. Nevertheless, the PIP motifs of these non-
classical polymerases are necessary for their recruitment to stalled replication forks. The
X-ray crystal structures of PCNA bound to the PIP motifs of DNA polymerases η, κ, and
ι have been determined [115]. The structures of the PIP motifs of DNA polymerases η
and κ are similar to that of the classical DNA polymerase δ in that they form the normal
310 helix (Fig. 1.19C). There are, however, some minor differences in the specific
contacts made by these PIP motifs, because the sequences of the PIP motifs of these non-
classical polymerases differ slightly from the PIP consensus sequence. For example,
neither of these PIP motifs have the conserved glutamine residue. Pol η, for instance, has
a methionine residue that inserts into the small pocket where the glutamine normally fits.
The structure of the PIP motif of DNA pol ι, however, differs significantly from that of
any other PIP motif structure. It does not form the normal 310 helix, but instead forms a β-
bend-like structure (Fig. 1.19D). Taken together, it is likely that the divergence of the
non-classical polymerase PIP motifs from the consensus PIP sequence reduces their
affinities for PCNA relative to other PIP motifs [115]. This could be important for
preventing the recruitment of non-classical polymerases to replication forks until the
PCNA is monoubiquitylated and their activities are needed.
In the X-ray crystal structures of PCNA bound to some PIP peptides, secondary
contacts (i.e., those that occur outside of the PIP motif) are observed between PCNA and
the portions of the peptide flanking the PIP motif. For example, DNA ligases catalyze
the linkage of 5ꞌ phosphates and a 3ꞌ OH groups during DNA repair and Okazaki
fragment processing. The yeast Cdc9 DNA ligase has a PIP motif that forms the
conventional 310 helix. However, the residues flanking the N-terminal sides of the PIP
motif form an anti-parallel β-sheet with the C-terminus of PCNA [204]. The presence of
DNA damage triggers an increase in expression of the tumor suppressor protein p21
leading to DNA replication arrest. The inhibition of DNA replication by p21 requires that
it bind directly to PCNA [173, 205, 206]. The X-ray crystal structure of the p21 PIP motif
33
bound to PCNA reveals that this PIP motif binds in the normal manner. However,
secondary contacts between PCNA and the peptide in the regions immediately flanking
the PIP motif are observed. The N-terminal and the C-terminal flanking regions form
anti-parallel β-sheets with the C-terminus and the IDCL of PCNA, respectively [207]. It
has been suggested that these extensive interactions are responsible for the higher affinity
PIP motif-PCNA interaction observed with the p21 PIP motif relative to other PIP motifs.
As discussed above, this tighter binding may allow the p21 PIP to inhibit DNA
replication by effectively competing with DNA polymerases for binding PCNA.
Structures of PCNA bound to full-length proteins.
While most structures of PCNA have been of complexes of PCNA with PIP motif
peptides, a few structures have been determined of complexes of PCNA with full-length
proteins. These have provided insights into the secondary contacts between PCNA and
PCNA-binding proteins that occur in addition to and alongside the contacts mediated by
PIP motifs. For example, the X-ray crystal structure of PCNA bound to full-length FEN1,
which catalyzes the removal of 5ꞌ single-stranded DNA overhangs that occur during
DNA repair and during the processing of the ends of Okazaki fragments, has been
determined (Fig. 1.20A) [208]. FEN1 consists of a nuclease core domain (residues 1–
332) and a C-terminal tail region (333–380). The main PCNA-interacting interface of
FEN1 is the N-terminal half of the C-terminal tail region, which contains a PIP motif.
Although the primary contact made between FEN1 and PCNA is mediated by the PIP
motif, there are secondary contacts between PCNA and the regions flanking the PIP motif
and between PCNA and the core domain of FEN1. Residues of the core domain make
several intramolecular contacts with the PIP motif as well as several intermolecular
interactions with both the IDCL and C-terminus of PCNA. Moreover, the core domain of
FEN1 is connected to its C-terminal tail through a 4-residue linker. It has been suggested
34
that this linker acts as a hinge to allow the core domain of FEN1 to be positioned near its
DNA substrate.
The structure of the FEN1-PCNA complex had three FEN1 molecules bound to
PCNA in each asymmetric unit, and each FEN1 molecule was in a different position
relative to the PCNA subunit to which it was bound (Fig. 1.20B). One of the observed
FEN1 positions had the active site of the core domain swung away from the front face of
PCNA, and this may represent an inactive conformation of FEN1. In the other two
positions, the core domain is located closer to the PCNA central cavity near the expected
position of the DNA. These latter positions may reflect active conformations in which
FEN1 can bind the DNA flap and bring itself into a position to cleave it.
Another interacting partner whose full-length structure has been determined in
complex with PCNA is RFC, the ATP-dependent clamp loading protein that binds to the
sliding clamp, opens the ring, and deposits it on the DNA. RFC sits on the front face of
the closed PCNA ring [209]. The five subunits of RFC form a right-handed spiral that is
tilted by approximately 9° relative to the threefold axis of PCNA. Only three of the five
subunits of RFC (RFC-A, RFC-B, and RFC-C) make contacts with the PCNA. In the
case of RFC-A and RFC-C, these are contacts mediated by PIP motifs. RFC-B, by
contrast, makes several secondary contacts with PCNA at the intersubunit regions.
Low resolution structures of PCNA complexes.
Several PCNA complexes have been examined using lower resolution structural
techniques such as small angle X-ray scattering (SAXS) and single particle electron
microscopy (EM). Using SAXS, the conformation of the archaeal PCNA protein from
Sulfolobus solfataricus was determined in complex with DNA ligase [210, 211]. In the
case of S. solfataricus, PCNA is a heterotrimer composed of the PCNA1, PCNA2, and
PCNA3 subunits. As in all sliding clamps, these three subunits have the same overall
structure [211]. In the SAXS structure of the ligase-PCNA complex, only one DNA
35
ligase was bound to the PCNA trimer, and it was found on the PCNA3 subunit. The
DNA ligase was observed to be in an open conformation extending out from the side of
the PCNA ring. DNA ligase, however, would need to also exist in a closed state in vivo,
which suggests that the interaction between these two proteins must be dynamic to allow
for the conformational change induced in the ligase during DNA replication and repair.
The structure of PCNA in complex with DNA polymerase B and DNA was
examined using EM with the archaeal Pyrococcus furiosus proteins [212]. Unlike in S.
solfataricus, PCNA from P. furiosus is similar to eukaryotic PCNA in that it forms a
homotrimer. Interestingly, the polymerase made contacts with PCNA on more than one
subunit. It formed the canonical PIP-PCNA interaction at the front face of PCNA, but
contacted an adjacent subunit as well. One purpose for making this secondary interaction
may be that it inhibits PCNA binding to other proteins during DNA replication.
However, it is also possible that two contact sites are necessary between PCNA and the
polymerase to help position the polymerase in an active conformation and promote its
activity.
Conclusions from structural studies with PCNA.
The organization and coordination of the vast number of proteins that interact
with PCNA is very complex, and the mechanism by which this regulation occurs remains
to be elucidated. It is likely that several factors affect the sequential loading and
unloading of particular PCNA-binding proteins. It is likely that local protein
concentrations as well as competition between target proteins influence who binds and
when. For instance, the affinity of the PIP motifs of non-classical polymerases for PCNA
is lower than that of the PIP motifs of classical polymerases for PCNA [115]. This
difference may prevent error-prone non-classical polymerase access to the replication
fork until Ub-PCNA is present. Furthermore, one of the highest affinity PIP-PCNA
interactions determined is with p21 [201]. As a cell cycle regulator protein, it seems
36
sensible that p21 would compete for PCNA binding with other proteins, including
classical polymerases, to inhibit further rounds of replication in the event of aberrant
DNA replication or in the presence of DNA damage.
Another factor that probably has a large impact on PCNA-binding specificity is
the existence of secondary contacts outside the PIP motif, including both those that
immediately flank the PIP motif or those that are completely separate from it. Studies
show that the affinity of PIP-PCNA interactions can vary by as much as 1000-fold
depending on the sequences surrounding the PIP motifs112 Bret
. DNA in the context of
cellular processes is also likely a contributing factor to specificity. For instance, the
presence of damaged DNA initiates recruitment of a non-classical polymerase during
TLS, and the presence of a 5ꞌ flap-containing DNA substrate probably instigates
recruitment of FEN1 during DNA replication. Lastly, post-translational modifications on
both the target protein and PCNA itself play a role in regulating PCNA interactions.
PCNA is subject to ubiquitylation and sumoylation, both of which recruit specific
interacting partners. In contrast, modification of PCNA-binding proteins usually inhibits
complex formation. For example, ubiquitylation of non-classical polymerases may
preclude binding to Ub-PCNA, whereas phosphorylation of p21 and FEN1 has been
shown to inhibit their interactions with PCNA [213, 214].
Interactions of Y-family Polymerases with PCNA and Ubiquitylated PCNA
The Y-family polymerases are recruited to stalled replication forks and regulated
in part by their interactions with the key replication accessory factor PCNA. When cells
are exposed to DNA damaging agents, PCNA is ubiquitylated on lysine-164 by the Rad6-
Rad18 ubiquitin-conjugating complex [183, 184, 189], and ubiquitin-modified PCNA
recruits Y-family polymerases to replication forks. In the structure of ubiquitin-modified
PCNA, the ubiquitin moiety sits on the back face of the PCNA ring [215]. In this section,
37
we will discuss the interactions of Y-family polymerases with unmodified and ubiquitin-
modified PCNA.
Interactions with un-modified PCNA.
Pol , pol , and pol all possess one or more PIP motifs in their disordered C-
terminal regions (Figure 1.21 and Figure 1.22). These motifs all bind on the front face of
the PCNA ring near the inter-domain connector loop. The conserved hydrophobic and
aromatic residues bind in a pocket at the interface of the two domains of PCNA. While
the pol PIP motif binds to PCNA by forming the same 310 helix that other PIP motifs
form, the pol PIP binds to PCNA by forming a novel -bend structure. The significance
of this unusual PIP conformation is unclear. It should be noted that a structure of PCNA
bound to the pol PIP has also been determined [115], but in this case, additional amino
acid residues not found in pol were added to the PIP construct to allow PCNA binding.
The native pol PIP does not seem to bind PCNA, so this particular structure is of
limited value.
Purified pol , pol , and pol physically interact with unmodified PCNA [113,
216-218]. Unlike the interactions between PCNA and classical polymerases, the
interactions between PCNA and the Y-family polymerases do not substantially increase
the processivity of DNA synthesis. Nevertheless, steady state kinetics shows that
interacting with PCNA significantly increases the catalytic efficiency of nucleotide
incorporation by all three of these enzymes on both non-damaged and damaged
templates. For example, in the case of pol , the increase in efficiency of incorporation
opposite non-damaged DNA ranges from 3-fold to 10-fold and the increase in efficiency
on a template abasic site, a non-cognate lesion, ranges from 3-fold to as much as 300-fold
depending on experimental conditions [113, 216, 219]. These physical and functional
interactions with PCNA are dependent on intact PIP motifs. Moreover, in human cells,
intact PIP motifs are required for both pol and pol to localize to nuclear foci
38
containing PCNA following DNA damage [185, 220, 221]. In the case of human pol ,
there are two PIP motifs, named PIP1 and PIP2. Disruptions of the individual PIP motifs
have only a moderate effect on localization to nuclear foci and pol -dependent TLS
suggesting, that the two PIP motifs are able to functionally substitute for one another.
Simultaneous disruption of both PIP motifs, however, completely eliminates localization
and pol -dependent TLS in vivo [220].
Like the other Y-family polymerases, Rev1 physically interacts with PCNA [222,
223], and this interaction stimulates the catalytic activity of Rev1 [223]. Unlike these
other polymerases, however, Rev1 does not contain a canonical PIP motif, and there has
been some debate about the regions of Rev1 that are required to interact with PCNA. It
has been reported that the localization of Rev1 to nuclear foci containing PCNA requires
either the N-terminal half of Rev1 (residues 1 to 730) or the C-terminal half (residues 730
to 1251) [224]. Another report, however, showed that localization requires the C-terminal
region of Rev1 (residues 826 to 1251), but not the N-terminal region [225]. It has also
been reported that the N-terminal BRCT domain of Rev1 is required for localization to
foci in non-damaged cells, but is not required in UV-treated cells [222]. This too is
controversial as another study failed to detect a direct interaction between PCNA and the
Rev1 BRCT domain [226]. Moreover, the stimulation of Rev1’s catalytic activity by
PCNA does not require an intact BRCT domain [223]. Thus questions remain regarding
the structural basis of the PCNA-Rev1 interaction.
Interactions with ubiquitin-modified PCNA.
All four Y-family polymerases possess one or more small ubiquitin-binding
domains in their disordered C-terminal regions (Figure 1.21 and Figure 1.22). In the case
of pol and pol , these small domains are UBZs, which contain about 20 amino acid
residues and form a short, two-stranded anti-parallel -sheet followed by an -helix
(Figure 1.23) [114]. Two conserved cysteine residues and two conserved histidine
39
residues coordinate a zinc ion, which likely provides structural stability to this small
domain. In the case of pol and Rev1, these small domains are UBMs, which contain
about 30 amino acid residues and form a helix-turn-helix motif [227, 228]. NMR
titrations have shown that the UBZs and the UBMs interact in slightly different ways
with the canonical protein-protein interaction surface of ubiquitin, which is made up of a
conserved hydrophobic patch containing leucine-8, isoleucine-44, and valine-70. Neither
the conformation of the ubiquitin nor the conformation of the ubiquitin-binding domains
seems to change upon complex formation.
In vitro pull-downs using purified Rev1 have shown that this polymerase interacts
with ubiquitin-modified PCNA (Ub-PCNA) with qualitatively higher affinity than it
interacts with unmodified PCNA [223]. The difference in affinities between the pol η-
PCNA interaction and pol η-Ub-PCNA interaction has not been determined. However, in
human cells, pol specifically interacts with Ub-PCNA, but not unmodified PCNA.
Immunoprecipitation of PCNA from normal cells pulled down only unmodified PCNA,
whereas immunoprecipitation of PCNA from UV-irradiated cells pulled down Ub-PCNA
and pol . Moreover, localization of pol to nuclear foci and pol -dependent TLS
require that the UBZ be intact [185]. Localization of pol to foci requires that both
UBMs be intact [185, 227]. Thus ubiquitin-binding domains are important for
localization to nuclear foci.
The complex of Y-family polymerases and Ub- PCNA is likely flexible. First, the
PIP and ubiquitin-binding domains are located within large regions of the Y-family
polymerases that are intrinsically disordered. Second, experimental evidence obtained
using SAXS and computational studies using Brownian dynamics simulations show that
the ubiquitin moieties of Ub- PCNA are dynamic [229]. Nevertheless, while the ubiquitin
moieties (still attached to lysine-164 of PCNA) are capable of moving around, they have
preferred positions on the back face and the side of the PCNA ring. This is important as
nearly all PCNA binding proteins interact with the front face of PCNA. This suggests that
40
Y-family polymerases can bind to the back or side of the PCNA ring without affecting
ongoing activity of other proteins bound to the front face of PCNA. Thus the Y-family
polymerases can be held in reserve on the back or side of PCNA until their activities are
required. Then because of the flexible nature of this complex, they can move to the front
face of PCNA and engage the primer-terminus of the DNA substrate. A model of pol
bound to the DNA substrate on the front face of ubiquitin-modified PCNA is shown in
Figure 1.23A. In addition, a more detailed speculation of the function of each ubiquitin
position on PCNA is provided in Chapter 6.
Polymerase Switching and the Tool Belt Model
Polymerase switching during translesion synthesis.
Most PCNA-dependent DNA replication and repair pathways require that
multiple enzymes access the replication fork in a sequential and intricate manner. The
switching of one enzyme to another on the DNA substrate is coordinated by their
interactions with PCNA. How this switching occurs is not clear and is currently an active
area of research, and especially in the field of TLS. This is because, like all protein
exchanges at the replication fork, the regulation of TLS polymerase switching must be
tightly controlled as to prevent aberrant use of error-prone polymerases. During TLS,
classical polymerases are blocked at sites of DNA damage. This triggers the
monoubituitylation of PCNA, which recruits one or more non-classical polymerases to
the replication fork to synthesize through the damaged DNA. After this process is
complete, a second switch occurs back to the classical polymerase to allow normal
replication to proceed.
The TLS polymerase switching event has been studied more thoroughly in E. coli
than in eukaryotes. In an in vitro reconstituted system, the E. coli non-classical
polymerases Pol II and Pol IV can freely exchange with the classical Pol III through
41
interactions with the β-clamp [230]. Establishment of the Pol II- or Pol IV-β-clamp
complex slows down the replisome [230], possibly to allow time for these low fidelity
and low efficiency polymerases to perform nucleotide incorporation opposite damaged
DNA. Complementary in vivo studies indicate that the non-classical polymerases can
access a DNA substrate that is already engaged in DNA replication [230, 231]. Together,
these results support the idea that classical and non-classical polymerases can act in a
coordinated fashion at the primer-terminus.
Compared to prokaryotes, eukaryotic polymerase switching is more complex and
more stringently regulated. Efficient exchange from pol δ to pol η at sites of lesions
requires both stalling of the PCNA-classical polymerase holoenzyme as well as the
monoubiquitylation of PCNA. Pol η requires both its PIP and UBZ motifs as well as the
presence of Ub-PCNA to undergo polymerase switching with pol δ [232]. In contrast,
pol η was unable to exchange with pol δ on the DNA when PCNA was unmodified, even
when the replication fork was stalled. In a human reconstituted system, the formation of
a stable PCNA-pol δ-DNA complex stimulated Rad6/Rad18-mediated
monoubiquitylation of PCNA, suggesting that Rad6/Rad18 prefers a stalled PCNA-pol δ
complex as its substrate for modification [233]. Using Xenopus laevis egg extracts, it
was demonstrated that the modification of PCNA likely occurs after replication fork
stalling, but before pol δ dissociation from the primer-terminus [234]. Along with the
finding that monoubiquitylation of PCNA does not destabilize the PCNA-pol δ
interaction, this suggests that addition of the ubiquitin moiety to PCNA does not facilitate
pol δ dissociation from PCNA [232].
The tool belt model of polymerase switching.
How Y-family polymerases are recruited to stalled replication forks and how their
activities are coordinated with other polymerases and enzymes involved in DNA
replication and repair is unknown. However, it is clear that the switch from the classical
42
to the non-classical polymerase requires the monoubiquitylation of PCNA. In light of
this, two models of polymerase switching have been proposed: 1) ubiquitylation of
PCNA induces a conformational change in PCNA that promotes switching and 2)
ubiquitylation of PCNA strictly provides an additional binding surface for non-classical
polymerases. In the first model, the presence of the ubiquitin moiety must cause a
conformational change in PCNA that would either reduce its affinity for classical
polymerases, increase its affinity for non-classical polymerases, or both. This model is
unlikely, though, as recent structural studies of Ub-PCNA indicated that the addition of
ubiquitin does not alter the structure of PCNA [215]. Instead, several key pieces of data
have emerged lately that suggest that the second model of polymerase switching, also
referred to as the tool belt model, is more probable.
In the tool belt model, a polymerase is recruited to the back face or side of the
PCNA ring while another polymerase is simultaneously engaged at the DNA primer-
terminus at the front face of PCNA, as shown in Figure 1.24. The recruitment of the
second polymerase occurs on a different subunit than the one previously occupied on
PCNA and does not interfere with the conformation of the existing PCNA-polymerase
complex. Consequently, the second polymerase is able to rapidly replace the original
polymerase on the DNA substrate. This would require considerable flexibility of the
second polymerase, which is likely mediated via their intrinsically disordered regions.
Evidence to support the tool belt model comes from structural and biochemical studies
that revealed unique characteristics of both PCNA and PCNA-interacting proteins. First,
the presence of the ubiquitin-binding motif in non-classical polymerases and the lack of
one in the classical proteins suggest a competitive advantage for non-classical
polymerase access to the replication fork following PCNA ubiquitylation. This is also
likely the case with SUMO-PCNA binding proteins while handing off at the replication
fork during DNA recombination. Second, the classical pol δ forms a stable complex with
Ub-PCNA [232], supporting the notion that pol δ is capable of remaining bound to Ub-
43
PCNA while a non-classical polymerase displaces it at the primer-terminus to carry out
TLS.
Finally, the best evidence in support of the tool belt model comes from studies of
sliding clamps in archaea and prokaryotes. In S. solfataricus, simultaneous binding of the
DNA polymerase, FEN1, and DNA ligase to a single PCNA trimer has been observed
with GST pull-down assays [235]. Similarly, simultaneous binding of pol III and pol IV
to the β-clamp from E. coli was seen using fluorescence-based binding assays [236]. As
in the eukaryotic system, recruitment of the non-classical pol IV to PCNA is dependent
on the stalling of pol III, and a subsequent switch back to pol IV occurs immediately after
the stall is relieved. Crystallography with PCNA bound to pol IV revealed how pol IV
can be recruited to the sliding clamp and held in reserve after fork progression is blocked
during TLS [237]. A C-terminal peptide of pol IV contacts PCNA, thereby forming
similar interactions to those seen between the PIP motif of eukaryotic polymerases and
PCNA. Interestingly, an additional binding surface between these two proteins is also
observed. The C-terminal domain of pol IV and the subunit interface of PCNA make
secondary contacts, which maintains pol IV in an inactive orientation. This secondary
interaction in prokaryotes may be analogous to contacts between the ubiquitin of Ub-
PCNA and the ubiquitin binding domains of eukaryotic non-classical polymerases in that
they are necessary for non-classical polymerase recruitment to the replication fork.
Currently, no evidence for a tool belt model in eukaryotes exists, but considering the
parallels between the two domains, this scenario appears very feasible.
Mutant PCNA Proteins
Mutant PCNA proteins defective in translesion synthesis.
Isolation of mutant PCNA proteins from yeast has been a valuable tool for
improving our knowledge of PCNA function. In 2006, pol30-61 (formerly known as the
44
rev6-1 allele) was identified as an allele of POL30 (the gene which encodes PCNA) in
yeast that caused an increased sensitivity to UV-induced DNA damaging agents and a
deficiency in UV-induced mutagenesis [238]. This allele encodes a G178S substitution
in PCNA, and yeast cells containing this allele are defective in lesion bypass by pol η, pol
ζ, and Rev1 as well as the error-free damage tolerance pathway, but seem to support
normal cell growth [238]. Therefore it was not surprising when steady state kinetic
studies demonstrated that, unlike the stimulation seen in the presence of wild-type PCNA,
the G178S mutant PCNA protein inhibited pol η activity [219].
Another PCNA mutation, encoded by the pol30-113 allele, was identified in 1996
in a screen for mutant PCNA proteins sensitive to methylmethane sulfonate (MMS)
[239]. This allele results in an E113G substitution in PCNA and renders the cells
deficient in UV-induced mutagenesis [240]. As with the G178S mutant PCNA protein,
the E113G mutation causes increased sensitivity to DNA damage and loss of TLS, while
cell growth is normal [240]. In vivo studies carried out by this group also showed that
PCNA containing the E113G mutation is capable of being monoubiquitylated in response
to DNA damage. This suggests that the loss in TLS caused by this substitution is not due
to its inability to be modified, but is likely due to some defect downstream of
ubiquitylation.
In order to gain a better understanding of how the G178S and E113G substitutions
in PCNA cause defects in TLS, our lab determined the X-ray crystal structures of these
mutant proteins [219, 241]. Both of the Gly-178 and Glu-113 residues are located at the
subunit-subunit interface of PCNA. They reside directly across from each other within
the β-strands on adjacent subunits. Analysis of the crystal structures of the G178S and
E113G mutant PCNA proteins revealed similar overall structure to wild-type PCNA as
well as similar local structural alterations. Both substitutions resulted in a shift in an
extended loop near the subunit interface called loop J (in the domain where the Glu-113
is located) compared to the wild-type protein (Figure 1.25). However, it should be noted
45
that the shift seen in the E113G mutant PCNA protein was less pronounced than the shift
seen in the G178S mutant PCNA protein. From these results, it was suggested that the
position of loop J in PCNA is essential for stimulating proper TLS.
Although these structures do provide insight into how the mutations in PCNA
may be disrupting TLS, more experiments utilizing these proteins are necessary to
determine the underlying mechanism of this defect. For instance, the impact of these
amino acid substitutions on the subunit interface was not described in these previous
studies. We reanalyzed these structures and identified substantial disruptions to the
interface that provide a more plausible cause for TLS inhibition, which is discussed in
more detail in Chapter 2. In addition, it is expected that non-classical polymerases only
associate with the ubiquitin-modified PCNA during TLS in vivo. However, no work has
been done with the monoubiquitylated form of PCNA that contains either of these
mutations. My studies with the G178S and E113G mutant forms of Ub-PCNA described
in Chapter 2 examine the mechanisms by which these substitutions interfere with TLS in
the context of the cell.
Mutant PCNA proteins defective in mistmatch repair.
Over time, numerous mutations in POL30 have been identified that cause
increased mutation rates due to defects in MMR. However, most of these mutations
generate defects in other replication and repair pathways as well. In 2002, two PCNA
mutations were isolated that appear to disrupt MMR with little or no other defects in vivo
[242]. These are the pol30-201 and pol30-204 mutations, which represent the amino acid
substitutions C22Y and C81R, respectively. The Cys-22 residue is located within the
inner surface of the central hole on PCNA and its substitution to Tyr-22 results in a
robust defect in MutSα-dependent MMR [242]. The Cys-81 residue is located near the
monomer-monomer interface of the PCNA trimer, and its substitution to Arg-81 results
in a partial impairment in both MutSα- and MutSβ-dependent MMR [242]. In order to
46
determine if these two mutations had synergistic or epistatic effects in vivo, one group
generated a yeast strain containing both the C22Y and C81R mutations in PCNA and
measured mutation rates using a canavanine-resistant (Can-R) assay [243]. They
determined that the double mutant strain exhibited a mutation rate 10-fold greater than
that of either of the single mutants. In fact, the phenotype of the double mutant was
similar to that seen in the absence of both MSH3 and MSH6, suggesting that the C22Y
and C81R mutations in PCNA are synergistic [243]. This is likely due to a complete loss
of MMR through disruption of both MutSα- and MutSβ-dependent repair.
Prior to my work, very little was known about these mutant proteins and how they
fail to support MMR. Sedimentation analysis with MutSα and the wild-type and mutant
PCNA proteins indicated that the C81R mutant PCNA protein, but not the C22Y mutant
PCNA protein, is incapable of interacting with MutSα [242]. From these data, the
authors predicted that the C81R mutant PCNA protein inhibits MMR because it does not
allow the interaction between PCNA and MutSα, whereas the C22Y mutant PCNA
protein may interact with MutSα in an inappropriate manner to reduce MMR. My studies
with these proteins, however, suggest that these conclusions are incorrect. In Chapter 3, I
discuss the X-ray crystal structures of these two mutant PCNA proteins and show results
of biochemical studies that together more precisely address how these mutant proteins
cause defects in MMR.
Thesis Overview
In Chapter 2, I investigate the mechanism of TLS inhibition by two PCNA mutant
proteins – the E113G and G178S mutant proteins. The X-ray crystal structures of these
two proteins have been determined, and from these structures, it was suggested that the
position of loop J was important for stimulation of TLS by PCNA [219]. However after
analyzing these structures more thoroughly, I concluded that the subunit interface of
47
PCNA is significantly altered in both the E113G and G178S mutant PCNA proteins and
that this is likely the cause of TLS disruption in cells. Due to these observations, I
examined if these structural alterations affect PCNA trimer stability. Results showed that
the G178S mutant PCNA protein had a significantly reduced stability compared to wild-
type PCNA, and that the E113G mutant PCNA protein had a moderately reduced
stability. The reduced trimer stability of the E113G mutant PCNA protein did not inhibit
binding of PCNA to polymerases, but it did inhibit PCNA from stimulating both pol η
and pol δ activity opposite abasic sites. Interestingly, the E113G mutant PCNA protein
still allowed for efficient synthesis opposite normal, non-damaged DNA by pol δ. In
vivo, non-classical polymerases only associate with the ubiquitin-modified form of
PCNA, suggesting that the ubiquitin modification on PCNA is necessary and sufficient
for promoting TLS by these polymerases. The presence of ubiquitin on the E113G
mutant PCNA protein, however, did not rescue the inhibition caused by the mutation.
In Chapter 3, I discuss two mutant PCNA proteins (with C22Y and C81R
substitutions) that are defective in MMR but appear to have no other replication or repair
defects. To understand the structural and mechanistic basis by which these two amino
acid substitutions in PCNA proteins block MMR, I solved the X-ray crystal structures of
both mutant proteins and carried out further biochemical studies in collaboration with
Elizabeth Boehm. We found that these two amino acid substitutions lead to distinct
structural changes in PCNA. The C22Y substitution in PCNA creates a distortion of the
α-helices that comprise the central hole of PCNA, which causes it to form an aberrant
PCNA-MutSα complex on DNA containing a mismatch. The C81R substitution, in
contrast, causes local changes in the β-sheet at the PCNA subunit-subunit interface and
has reduced affinity for MutSα compared to the wild-type PCNA and the C81R mutant
PCNA protein. Similar to the C22Y mutant PCNA protein, however, the C81R mutant
PCNA protein also forms an aberrant PCNA-MutSα complex on a mismatched DNA
substrate. From these results, we conclude that the structural integrity of the -helices
48
lining the central hole and the -sheet at the subunit interface are both necessary to form
productive complexes with MutS and mismatch-containing DNA.
In Chapter 4, I characterized a set of mutant PCNA proteins that were made in our
laboratory by Dr. Christine Kondratick. She has generated 12 strains of yeast that
express mutant forms of PCNA with mutations at the subunit interface. Her work using
UV survival and UV-induced mutagenesis assays with these strains shows that they all
have varying effects on cell growth and mutagenesis. Five of these mutant proteins, each
with an amino acid substitution at a distinct residue at the interface, have been purified
for X-ray crystallography and biochemical analysis. These mutant PCNA proteins
contain the S177G, G178S, S179T, V180A, and I181R substitutions. I examined the
trimer stability of these mutant PCNA proteins and their abilities to stimulate TLS by pol
η and pol δ using DNA polymerase activity assays. I determined that, similar to their in
vivo effects, these five mutant PCNA proteins show varying effects on the activities of
pol η and pol δ opposite both non-damaged DNA and abasic sites. My data shows a
strong correlation between PCNA trimer stability and the ability to stimulate TLS by both
classical and non-classical polymerases.
In Chapter 5, I investigate the interaction between un-modified PCNA or
ubiquitin-modified PCNA and the C-terminal region of pol η. The CTR of pol η contains
both the PCNA-binding (PIP) and ubiquitin-binding (UBZ) motifs. Using disorder
prediction software and NMR, I determined that the CTR of pol η is intrinsically
unstructured. These disorder predictions were also used for all of the other eukaryotic
non-classical polymerases as well, which showed that each of these proteins have large
regions of disorder. As expected, pol η binding to PCNA or Ub-PCNA does not induce
folding of this region. Using binding experiments, I show that the CTR of pol eta is
sufficient for binding to PCNA, ubiquitin, and Ub-PCNA, and that, in contrast to
previous beliefs, this region binds to Ub-PCNA with much higher affinity than it does to
either PCNA or ubiquitin alone. Lastly, I discuss attempts at obtaining the crystal
49
structure of the CTR of pol η bound to Ub-PCNA as well as a novel method of producing
this complex in a 1:1 ratio.
In Chapter 6, I provide a summary of the results described within this thesis. I
also discuss implications of the conclusions from each Chapter and how they further the
field of DNA replication and repair. Finally, I propose prospective future directions and
experimentations in this Chapter.
50
Figure 1.1. Model of DNA replication in eukaryotes. Leading and lagging strand
synthesis are shown and the key proteins involved in each are indicated.
MCM
RPA
Polymerase ε
PCNA
Leading-strand
synthesis
Lagging-strand
synthesis
Polymerase δ
Polymerase α
FEN1
DNA ligase
51
Figure 1.2. Common types of DNA damage. (A) Thymine dimers and (B) (6-4)
photoproducts are created by exposure to ultraviolet light. Both lesions result from two
adjacent thymine residues being covalently cross-linked together. (C) DNA strand breaks
may arise from exposure to ionizing radiation. (D) 8-oxoG lesions are caused by oxygen
free radicals, which result in the addition of an oxygen atom to carbon 8 of the guanine
base. (E) Abasic sites are spontaneously produced by hydrolysis of the glycosidic bond
that attaches the base to the sugar-phosphate backbone. (F) Uracil residues are produced
in DNA by spontaneous deamination of a cytosine residue.
OO
O
OO
O
R
P
N
NH
O
O
CH3
OH
O
R
N
NH
O
O
CH3
OO
O
OO
O
R
P
N
NH
O
O
CH3
OH
O
R
N
NH
O
CH3
OO
O
N
NH
N
NH
NH2
O
O
R
RO
O
O
N
NH
O
OR
R
OHO
O
O
R
R
OO BaseR
P OH
O
O
OH
OOH
O
Base
R
A B C
D E F
52
Figure 1.3. Possible DNA mismatches. (A) A cytosine residue mispaired with a thymine
residue. (B) A guanine residue mispaired with an adenine residue. Hydrogen bonds
formed are indicated with a dotted line and glycosidic bonds are indicated with thick blue
lines.
53
Figure 1.4. Model of mismatch repair in eukaryotes. The mismatched base pair
(shown as a yellow star) is recognized by the MSH2-MSH6 complex, which is recruited
by PCNA. The key factors involved in subsequent DNA excision, resynthesis, and
ligation of the gap produced during mismatch repair are indicated.
54
Figure 1.5. Model of translesion synthesis. Normal DNA synthesis by classical
polymerases is blocked at sites of DNA damage (indicated by the red X), causing the
polymerase to stall. This stall in replication stimulates the monoubiquitination of PCNA,
which recruits a non-classical polymerase to the replication fork. The non-classical
polymerase then replaces the classical polymerase at the site of nucleotide incorporation
and proceeds to incorporate opposite the DNA lesion. After translesion synthesis is
complete, the classical polymerase may once again associate with the DNA substrate to
continue normal DNA polymerization.
55
Table 1.1. Classification of DNA polymerases.
56
Figure 1.6. Mechanism of DNA polymerization by polymerases. Nucleotide
incorporation by pol δ is initiated upon binding of the polymerase (E) to the DNA
substrate (DNA25) (step 1). This step is defined by the binding affinity of the polymerase
for the DNA (KdDNA
). Pol δ then binds the incoming nucleotide to form a pol δ-DNA-
dNTP ternary complex (step 2), which is limited by the binding affinity of pol δ for the
incoming nucleotide (KddNTP
). Through nucleophilic attack by the 3’ hydroxyl group of
the primer terminus on the α-phosphate of the dNTP, a phosphodiester bond is formed
and pyrophosphate is released (step 3). This step is described by the rate constant of
polymerization (kpol). The result is a DNA substrate that is one nucleotide longer than the
original substrate (DNA26). After nucleotide incorporation, the rate-limiting step occurs
in which the polymerase dissociates from the DNA (step 4), which is described by pol δ’s
rate constant of dissociation for the DNA (koff). If the next dNTP is available, pol δ with
translocate a single nucleotide downstream to prepare for another round of nucleotide
incorporation. DNA synthesis continues in this manner until the polymerase dissociates
from the polymerase-DNA complex.
57
Figure 1.7. Structure of the catalytic subunit (Pol3) from DNA polymerase δ from
yeast bound to DNA. (A) and (B) Two different views of the structure of the Pol3
subunit of pol δ bound to DNA. The overall shape of the catalytic domain resembles a
right hand, with subdomains referred to as the palm (green), fingers (blue), and thumb
(purple), which is similar to the structure observed for nearly all polymerase domains
studied thus far The N-terminal (red) and exonuclease (yellow) domains are typically
found in classical polymereases. DNA is shown in orange. (C) Linear representation of
the domains of Pol3. Residues defining the domain boundaries are indicated above the
illustration.
58
Figure 1.8. Structure of DNA polymerase from the bacteriophage RB69. (A)
Structure of the RB69 polymerase. The overall shape of the polymerase resembles a
right hand, with subdomains referred to as the palm (green), fingers (blue), and thumb
(purple). The N-terminal (red) and exonuclease (yellow) domains are indicated. (B)
Structure of the RB69 polymerase in complex with DNA (orange/green/pink) with dATP
opposite dTMP. (C) Linear representation of the domains of the RB69 polymerase.
Residues defining the domain boundaries are indicated above the illustration.
59
Figure 1.9. Structure of DNA polymerase η from yeast. (A) Structure of pol η with
the palm (green), fingers (blue), and thumb (purple) subdomains indicated. The
polymerase associated domain (PAD), unique to non-classical polymerases is shown in
yellow. (B) Linear representation of the domains of pol η. The C-terminal region (CTR)
is shown in white, with the ubiquitin-binding zinc-finger (UBZ) and PCNA interacting
peptie (PIP) motifs contained within the CTR shown in grey. Residues defining the
domain boundaries are indicated above the illustration.
60
Figure 1.10. Structural model of the full-length pol η. The catalytic core region of
yeast pol η (residues 1-510) is shown at the far left and the disordered CTR (residues
510-632) is shown as a random coil in grey. The UBZ and PIP motifs are indicated.
61
Figure 1.11. Structure of the ubiquitin-binding zinc-finger (UBZ) of pol η. The
structure of the UBZ of human pol eta determined by NMR is shown from the N-
terminus to the C-terminus (blue to red). The zinc ion is modeled in as a grey sphere.
62
Figure 1.12. Structure of the human MutSα dimer bound to a G:T mispair. (A) Side
view of the structure of the MutSα dimer with MSH6 shown in the foreground and MSH2
shown in lighter colors in the background. DNA is shown in orange. (B) View of the
structure of the MutSα dimer rotated by 90o compared to (A), with MSH6 shown on the
left and MSH2 shown in lighter colors on the right. (C) Linear representation of the
domains of each monomer of MutSα. Residues defining the domain boundaries are
indicated above the illustration.
63
Figure 1.13. Structure of yeast PCNA. (A) Front view and (B) side view of the PCNA
trimer with individual subunits colored in pink, green, and blue. The two domains of
each subunit and the inter-domain connector loop (IDCL) are indicated.
64
Figure 1.14. Structure of yeast PCNA bound to DNA. (A) Front view and (B) side
view of the PCNA trimer (purple) with a double-stranded DNA helix (orange, green,
blue) contacting the central hole of the PCNA ring. The two domains of each subunit and
the IDCL are indicated.
65
Figure 1.15. Structure of ubiquitin-modified PCNA. (A) Linear diagram of the two
polpeptides used to create the split Ub-PCNA with residue numbers of PCNA (shown in
green) and ubiquitin (shown in blue) indicated. (B) Front and side view of the structure
of Ub-PCNA with the ubiquitin moieties shown in blue and the PCNA trimer shown in
green.
66
Figure 1.16. Overlay of the two positions occupied by ubiquitin in the crystal
structure of Ub-PCNA. PCNA is shown in green, the ubiquitin moiety in position 1 is
shown in blue, and the ubiquitin moiety in position 2 is shown in pink.
67
Figure 1.17. The potential “ubiquitin-switch” on PCNA. Schematic for the possible
switch between mono- and polyubiquitylation and sumoylation of PCNA and the
downstream pathways that are affected by these modifications. The PCNA trimer is
shown in purple, ubiquitin is shown as yellow circles, and SUMO is shown as pink
circles. Potential modification sites on PCNA are indicated.
68
Figure 1.18. Overlay of the structures of ubiquitylated and sumoylated PCNA. The
ubiquitin (blue) and SUMO (pink) moieties occupy different positions on the back face of
yeast PCNA (green).
69
Figure 1.19. Structures of PCNA bound to PIP peptides. (A) Sequence alignment of
PIP peptides from several human PCNA-binding proteins. In the PIP consensus
sequence, the ‘h’ can be isoleucine, leucine or methionine, and the ‘a’ can be
phenylalanine or tyrosine. (B) The structure of the canonical PIP motif from FEN1
binding to PCNA is shown in yellow. (C) The structure of the PIP motif from DNA
polymerase η bound to PCNA shown in red overlaid with the structure of the PIP motif
from FEN1 shown in yellow. (D) The structure of the PIP motif from DNA polymerase ι
bound to PCNA shown in green overlaid with the structure of the PIP motif from FEN1
shown in yellow. Panels B-D are courtesy of Bret Freudenthal.
70
Figure 1.20. Structure of PCNA bound to FEN1. (A) Ribbon diagram of the PCNA
trimer shown in blue bound to three molecules of FEN1 shown in red, yellow, and green.
(B) Overlay showing the three positions of FEN1 relative to the PCNA subunit to which
they are bound. The PCNA is shown in blue, the inactive conformation is shown in red,
and the active conformations are shown in yellow and green.
71
Figure 1.21. The structured and unstructured regions of Y-family polymerases. The
graphs of disorder probability for (A) pol , (B) pol , (C) pol , and (D) Rev1 were
obtained using the meta approach for predicting disordered regions of proteins. In the
diagrams of each polymerase, the structured regions are shown as thick rectangles, and
the disordered regions are shown as thin rectangles. The polymerase (Pol) domain and
PAD of each protein are indicated. The N-clasp (NC) of pol as well as the N-digit
(ND), the BRCT domain, and the CTD of Rev1 are indicated. PCNA-binding, ubiquitin-
binding, and Rev1-binding motifs are indicated by P, U, and R, respectively. Courtesy of
Todd Washington.
72
Figure 1.22. Structural models of the full-length Y-family polymerases. The models
of full-length (A) pol , (B) pol , (C) pol , and (D) Rev1 were built using Coot starting
with the X-ray crystal structures of the catalytic core regions of these polymerases; and
the NMR structures of the UBZ of pol , the UBM of pol , the Rev1 CTD, and the Rev1
BRCT domain. The UBZ of pol was modeled based on the UBZ of pol , and the UBM
of Rev1 was modeled based on the UBM of pol . The disordered regions were then built
as random coils. The various PIP motifs, UBZs, UBMs, and RIR motifs are indicated.
Courtesy of Todd Washington and Elizabeth Boehm.
73
Figure 1.23. Structural model of full length pol bound to ubiquitin-modified
PCNA. (A) The model of pol bound to ubiquitin-modified PCNA was built using Coot
starting with the X-ray crystal structures of the catalytic core region of pol , ubiquitin-
modified PCNA, and PCNA bound to the pol -PIP motif and with the NMR structure of
the pol UBZ. (B) A close up of the pol PIP motif bound to the PCNA portion of
ubiquitin-modified PCNA is shown. (C) A close up of the pol UBZ bound to the
ubiquitin portion of ubiquitin-modified PCNA is shown. Courtesy of Todd Washington
and Elizabeth Boehm.
74
Figure 1.24. The tool belt model of translesion DNA synthesis. In this model, the
classical polymerase remains bound to the ubiquitylated PCNA while a non-classical
polymerase is engaged at the primer-terminus. After damage bypass by the non-classical
polymerase, the classical polymerase gains access back to the primer-terminus to
continue DNA replication.
75
Figure 1.25. Structures of the G178S and E113G mutant PCNA proteins. (A)
Overlay of the G178S mutant PCNA protein and wild-type PCNA. Positions of the
G178S residue, the E113G residue, and the J-loop are indicated. (B) Close-up view of
the positions of the J-loops of the G178S and E113G mutant PCNA proteins and wild-
type PCNA. The E113 residue, located on the same subunit as the J-loop, is indicated.
76
CHAPTER 2
PCNA TRIMER INSTABILITY INHIBITS TRANSLESION
SYNTHESIS BY DNA POLYMERASE η AND BY DNA
POLYMERASE δ
Abstract
Translesion synthesis (TLS), the process by which DNA polymerases replicate
through DNA lesions, is the source of most DNA damage-induced mutations. Sometimes
TLS is carried out by classical DNA polymerases that have evolved to synthesize DNA
on non-damaged templates. Most of the time, however, TLS is carried out by specialized
non-classical DNA polymerases that have evolved to synthesize DNA on damaged
templates. TLS requires the mono-ubiquitylation of the replication accessory factor
proliferating cell nuclear antigen (PCNA). PCNA and ubiquitin-modified PCNA (Ub-
PCNA) stimulate TLS by classical and non-classical polymerases. Two mutant forms of
PCNA, one with an E113G substitution and one with a G178S substitution, support
normal cell growth but inhibit TLS thereby reducing mutagenesis in yeast. A re-
examination of the structures of both mutant PCNA proteins revealed substantial
disruptions of the subunit interface that forms the PCNA trimer. Both mutant proteins
have reduced trimer stability with the G178S substitution causing a more severe defect.
The mutant forms of PCNA and Ub-PCNA do not stimulate TLS of an abasic site by
either classical pol or non-classical pol . Normal replication by pol was also
impacted, but normal replication by pol was much less affected. These findings support
a model in which reduced trimer stability causes these mutant PCNA proteins to
occasionally undergo conformational changes that compromise their ability to stimulate
TLS by both classical and non-classical polymerases. (The work described in this
77
Chapter will be published in Dieckman, L.M. and Washington, M.T. (2013) DNA
Repair.)
Introduction
Classical DNA polymerases have evolved to synthesize DNA on non-damaged
templates during normal DNA replication. In general, these enzymes catalyze the
template-directed incorporation of nucleotides with high fidelity. For example, eukaryotic
DNA polymerase (pol ), a member of the B-family of polymerases, is responsible for
the majority of lagging strand synthesis during normal DNA replication [244-246]. It also
plays important roles in base excision repair, nucleotide excision repair, mismatch repair,
and double strand break repair [247]. Yeast pol is a heterotrimer comprised of a
catalytic subunit (Pol3) and two accessory subunits (Pol31 and Pol32) [248]. A low
resolution structure of pol shows that the protein is elongated with Pol31 acting as a
bridge to connect the Pol3 and Pol32 subunits [249]. The X-ray crystal structure of the
catalytic Pol3 subunit shows that it possesses three domains: an N-terminal domain, a
polymerase domain, and an exonuclease domain for proofreading [74]. The polymerase
domain contains fingers, thumb, and palm sub-domains similar to those found in other B-
family DNA polymerases. In the absence of the proofreading function, pol incorporates
nucleotides with error frequencies of 10-4
to 10-5
[75, 87, 250]. Kinetic analysis shows
that this high fidelity arises at multiple steps along the reaction pathway, including both
the initial nucleotide-binding step and the subsequent nucleotide-incorporation step [75].
Despite the remarkable catalytic activities of classical DNA polymerases, most of
these enzymes are unable to efficiently incorporate nucleotides opposite template DNA
damage. Consequently, cells possess several specialized non-classical DNA polymerases,
which have evolved to replicate through DNA lesions [60, 61, 202, 203, 251-254]. For
example, eukaryotic DNA polymerase (pol ), a member of the Y-family of DNA
78
polymerases, is responsible for bypassing template thymine dimers and 8-oxoguanine (8-
oxoG) lesions [77, 88, 89]. This enzyme is a monomer, and X-ray crystal structures show
that it possesses two domains: a polymerase domain and a polymerase-associated domain
[110-112, 255]. The polymerase domain contains fingers, thumb, and palm sub-domains
similar to those found in other Y-family polymerases. It also contains an active site that is
larger than those of classical polymerases, and this larger active site allows pol to
readily accommodate thymine dimers [110, 112]. Kinetic analyses show that the
mechanisms of nucleotide incorporation opposite non-damaged templates, opposite
thymine dimers, and opposite 8-oxoG lesions are identical [107, 108, 256]. Thus these
forms of DNA damage present no barrier to nucleotide incorporation by pol .
Translesion synthesis (TLS) is the process by which DNA polymerases replicate
through DNA lesions by directly using the damaged DNA as a template. Although
classical DNA polymerases can carry out TLS in a few contexts, most TLS is
accomplished by non-classical polymerases. In these cases, the classical polymerase is
replaced at the replication fork by a non-classical polymerase. This polymerase-switching
event is facilitated by a key replication accessory factor, proliferating cell nuclear antigen
(PCNA). In addition to its role in TLS, PCNA participates in a wide range of functions,
including DNA replication, base excision repair, nucleotide excision repair, mismatch
repair, recombination, chromatin remodeling, and cell cycle regulation [155, 199, 257-
259]. PCNA directly regulates the activities of both classical and non-classical
polymerases. For instance, it increases both the processivity and catalytic efficiency of
DNA synthesis by classical pol [75, 165]. Similarly, it increases the catalytic efficiency
of DNA synthesis by non-classical pol [113]. During TLS, PCNA is monoubiquitylated
on Lys-164 by the Rad6-Rad18 complex [183]. Ubiquitin-modified PCNA (Ub-PCNA)
plays an important role in TLS, as non-classical polymerases preferentially interact with
Ub-PCNA through their ubiquitin-binding elements [185].
79
Two mutant forms of yeast PCNA, one with a G178S substitution and the other
with an E113G substitution, have been identified that support normal cell growth but
inhibit TLS thereby reducing mutagenesis in yeast [238-240]. It remains unclear how
these mutant forms of PCNA inhibit TLS. The X-ray crystal structures of both of these
mutant PCNA proteins have been determined, and both of these amino acid substitutions
are located at the subunit interface that forms the PCNA trimer [219, 260]. I have re-
examined these structures and noticed substantial disruptions to the subunit interface.
This suggested that these mutant PCNA proteins may have defects in trimer stability
compared to the wild-type PCNA protein. I found that both mutant PCNA proteins had
reduced trimer stability with the G178S substitution causing a more severe defect. I also
found that mutant forms of PCNA and Ub-PCNA did not stimulate TLS of an abasic site
by either classical pol or non-classical pol . Normal replication by pol was also
impacted, but normal replication by pol was much less affected. These findings support
a model in which reduced trimer stability causes these mutant PCNA proteins to
occasionally undergo conformational changes that compromise their ability to stimulate
TLS by both classical and non-classical DNA polymerases.
Materials and Methods
Protein expression and purification.
The wild-type and E113G mutant PCNA proteins from S. cerevisiae were over-
expressed as N-terminal His6-tagged proteins and purified from E. coli as previously
described [219]. The wild-type and E113G mutant Ub-PCNA proteins from S. cerevisiae
were over-expressed and purified from E. coli using the split-fusion strategy as
previously described [215]. S. cerevisiae replication factor C (RFC) was over-expressed
and purified from E. coli as previously described [261]. S. cerevisiae pol and pol
were over-expressed and purified from S. cerevisiae as previously described [75, 256].
80
DNA and nucleotide substrates.
To measure DNA polymerase activity, a 68-mer oligodeoxynucleotide with the
sequence 5'- GAC GGC ATT GGA TCG ACC TCX AGT TGG TTG GAC GGG TGC
GAG GCT GGC TAC CTG CGA TGA GGA CTA GC with biotins on both ends was
used as the template strand. The X represents the position of a non-damaged G or an
abasic site. For steady state kinetic studies, a 31-mer oligodeoxynucleotide with the
sequence 5'-GGT AGC CAG CCT CGC ACC CGT CCA ACC AAC T was used as a
primer. For the processivity assays, a 31-mer oligodeoxynucleotide with the sequence 5'-
TCG CAG GTA GCC AGC CTC GCA CCC GTC CAA C was used as a primer. The
primer strands were 5'-32
P-end-labeled with T4 polynucleotide kinase and (γ-32
P)ATP.
The primer and template strands were annealed at 200 nM in 25 mM TrisCl, pH 7.5, and
100 mM NaCl at 90°C for 2 min and slowly cooled to 30°C. Solutions of each of the four
dNTPs (10 mM) were obtained from New England Biolabs and stored in 5 l-aliquots at
-80°C.
PCNA trimer stability assays.
To assay for PCNA trimer stability, I used non-denaturing polyacrylamide gel
electrophoresis (PAGE) and size exclusion chromatography. For non-denaturing PAGE,
the wild-type and mutant PCNA and Ub-PCNA proteins (0.05 to 5 mg/ml) were
incubated in 60 mM TrisCl, pH 6.8, 0.01% bromophenol blue, and 10% glycerol for 5
min. The protein samples were then run on a TrisCl pre-cast 4-20% gradient non-
denaturing polyacrylamide gel (Bio-Rad) at a constant 25 mA using 25 mM Tris, pH 8.3,
and 0.2 M glycine as a running buffer at 4°C. Protein bands were visualized by
coomassie staining. For size exclusion chromatography, PCNA proteins were diluted to
various concentrations (0.005 to 0.1 mg/ml) and loaded onto a 120 ml HiLoad 16/60
Superdex 200 PG column (GE Healthcare). The column was calibrated with the
81
following molecular weight standards (Bio-Rad): thyroglobulin (670 kDa), γ-globulin
(158 kDa), ovalbumin (44 kDa), myoglobin (17 kDa), and vitamin B12 (1.35 kDa).
Enzyme-linked immunosorbent assays.
The wells of a 96 well EIA/RIA plate (Corning) were coated with 1 µg of pol in
PBS (4.3 mM Na2HPO4, 1.4 mM KH2PO4, pH 7.4, 137 mM NaCl, 2.7 mM KCl) for one
hour. The wells were then washed three times with PBS, 0.2% Tween-20, blocked for 30
min. with PBS with 5% milk, and washed again. Various amounts of wild-type or mutant
PCNA or Ub-PCNA proteins or bovine serum albumin (BSA) (0.5 µg to 20 µg) in PBS
with 5% milk were added to the wells and incubated for one hour, followed by washing.
For the wild-type PCNA and E113G mutant PCNA proteins, a 1:500 dilution of rabbit
polyclonal anti-PCNA antibody in PBS with 5% milk was added to the wells and
incubated for 30 min. For the wild-type and E113G mutant UbPCNA proteins, a 1:200
dilution of rabbit polyclonal anti-His tag antibody (Santa Cruz Biotechnology) was used.
Wells were washed before adding a 1:10,000 dilution of goat anti-rabbit antibody
conjugated with horseradish peroxidase (Jackson ImmunoResearch) in PBS with 5% milk
for 30 min. The plate was washed, and 0.8 mg/ml of O-phenylenediamine (Aldrich) in
0.05 M phosphate-citrate buffer with 0.03% sodium perborate (Sigma) was added.
Absorbance at 450 nm was measured after 5 to 35 min. with an iMark microplate reader
(Bio-Rad). The BSA control absorbance values were subtracted from the absorbance of
each sample at the corresponding protein concentration. All steps were performed at
25°C.
Polymerase processivity assays.
Reactions were performed in 40 mM Tri-Cl, pH 7.5, 8 mM MgCl2, 150 mM
NaCl, 1 mM DTT, and 100 µg/ml BSA at 30°C. Annealed DNA substrates were
incubated with a 10-fold molar excess of streptavidin for 5 min to block the ends of the
82
DNA to prevent PCNA dissociation. Wild-type or mutant PCNA or Ub-PCNA protein
(90 nM) was loaded onto 20 nM of the DNA substrate with 25 nM RFC and 500 µM
ATP. Either pol or pol (40 nM) was added to each reaction and pre-incubated for 20
min. Reactions were initiated by the addition of 200 µM of each dNTP and 1 mg/ml of
salmon sperm DNA and quenched after 1 min for Pol or 2 min for pol by the addition
of 10 volumes of 80% deionized formamide, 10 mM EDTA, pH 8.0, 1 mg/ml xylene
cyanol, and 1 mg/ml bromophenol blue. The products were then analyzed on a 15%
polyacrylamide sequencing gel containing 8 M urea and the intensities of the labeled gel
bands were determined using the Storm 860 (GE Healthcare). Percent of polymerases
that incorporated at least N nucleotides was calculated by dividing the sum of the
intensities of all gel bands resulting from N nucleotide incorporations or greater by the
sum of the intensities of all gel bands resulting from one nucleotide incorporation or
greater.
Polymerase activity assays.
Reactions were performed using the same buffer conditions as described in the
polymerase processivity assays, and the wild-type and mutant PCNA and Ub-PCNA
proteins were loaded on the DNA substrates as described in the processivity assays. For
nucleotide incorporation by pol opposite a non-damaged template residue, different
concentrations of dCTP (2 to 200 µM) were mixed with the PCNA-loaded DNA substrate
(20 nM) and pol (2 nM). The reactions were quenched at various times up to 15 min.
For nucleotide incorporation by pol opposite a template abasic site, different
concentrations of dGTP (2 to 200 µM) were used, and the reactions were quenched at
various times up to 20 min. For nucleotide incorporation by pol opposite a non-
damaged template residue, different concentrations of dCTP (2 to 200 µM in the absence
of any PCNA; and 0.5 to 50 µM in the presence of the wild-type or mutant PCNA or Ub-
PCNA proteins) were mixed with the PCNA-loaded DNA substrate (20 nM) and pol (2
83
nM). The reactions were quenched at various times up to 2 min. For nucleotide
incorporation by pol opposite a template abasic site, different concentrations of dATP
(2 to 200 µM in the absence of any PCNA, and 0.5 to 50 µM in the presence of the wild-
type or mutant PCNA or UbPCNA proteins) were used, and the reactions were quenched
at various times up to 15 min. The products were then analyzed on a 15% polyacrylamide
sequencing gel containing 8 M urea and the intensities of the labeled gel bands were
determined using the Storm 860 (GE Healthcare). The rate of product formation was
plotted as a function of incoming dNTP concentration, and the Vmax and Km parameters
were obtained from the best fit of the data to the Michaelis-Menten equation. All
experiments were carried out at least three times to ensure reproducibility.
Results
The E113G and G178S mutant PCNA proteins have
altered subunit interfaces.
Two variant forms of PCNA, the E113G and G178S mutant PCNA proteins,
support normal cell growth but inhibit TLS thereby reducing mutagenesis in yeast [238-
240]. These amino acid substitutions are located on -strand I1 and -strand D2,
respectively, which constitute the PCNA subunit interface. The X-ray crystal structures
of both of these mutant PCNA proteins have been determined [219, 260]. However, the
precise impact of these amino acid substitutions on the structure of the subunit interface
was not described in these previous studies. Here we have re-examined these structures
and noticed substantial disruptions to this interface (Fig. 2.1 and Table 2.1). In the case of
the wild-type PCNA protein, there are seven backbone hydrogen bonds between -
strands I1 and D2 with the distances between the amide nitrogen atoms and the
corresponding carbonyl oxygen atoms ranging from 2.8 to 3.1 Å. In the E113G mutant
PCNA protein, there are only five backbone hydrogen bonds between these -strands.
84
This is because the backbone of -strand I1 has shifted so that the distances between two
of the amide nitrogen atoms and the corresponding carbonyl oxygen atoms are now 4.5
and 5.9 Å, which is too large to form hydrogen bonds. In the G178S mutant PCNA
protein, there are only three backbone hydrogen bonds between these -strands. This is
because the position of the backbone of -strand I1 has shifted dramatically so that the
distances between four of the amide nitrogen atoms and the corresponding carbonyl
oxygen atoms now range from 4.1 to 9.3 Å, which again is too large to form hydrogen
bonds.
Trimer stability of the E113G and G178S mutant
PCNA proteins.
Because the E113G and G178S mutant PCNA proteins had fewer hydrogen bonds
at the subunit interface, I examined the stability of the mutant PCNA trimers. Non-
denaturing polyacrylamide gel electrophoresis (PAGE) showed that the wild-type PCNA
protein formed a stable trimer under all concentrations tested (0.05 to 5 mg/ml) (Fig. 2.2).
The E113G mutant PCNA protein formed a stable trimer at the higher concentrations (0.2
to 5 mg/ml). At lower concentrations (0.05 to 0.1 mg/ml), however, the gel bands were
consistently less intense than those of the wild-type protein and were streaking. This
suggested that although the mutant protein was primarily a trimer, it was less stable than
the wild-type PCNA timer. By contrast, the G178S mutant PCNA protein was a
monomer under all concentrations tested.
To test further the trimer stability of these mutant PCNA proteins, we used size
exclusion chromatography (Fig. 2.3 and Table 2.2). At all concentrations tested (0.005 to
0.1 mg/ml), the wild-type PCNA protein elutes from the size exclusion column at a
volume of approximately 74 ml, which corresponds to the 90-kDa trimer. At all
concentrations tested, the E113G mutant PCNA protein was primarily a trimer. However,
at the lower concentrations (0.005 and 0.01 mg/ml), a significant proportion (20 % and 7
85
%, respectively) eluted from the column at a volume of approximately 86 ml, which
corresponds to the 30-kDa monomer. This result is consistent with a previous study
showing decreased trimer stability of the E113G mutant PCNA protein [260]. By
contrast, at all concentrations tested, the G178S mutant PCNA protein was primarily a
monomer. This result seems to be at odds with the X-ray crystal structure of the G178S
mutant protein showing the mutant protein to be a trimer [219]. It should be noted that
the formation of this was due to the extremely high concentration of protein in the
crystalline state. Taken together, the native PAGE and size exclusion chromatography
results show that both mutant proteins form less stable trimers than does the wild-type
protein. Furthermore, the trimers formed by the G178S mutant PCNA protein were far
less stable than those formed by the E113G mutant PCNA protein.
In cells, PCNA is monoubiquitylated on Lys-164 during TLS. In order to study
these mutant proteins in the more biologically relevant context of ubiquitin-modified
PCNA (Ub-PCNA), I produced both wild-type and mutant Ub-PCNA proteins using the
split-fusion method [215]. This method, which involves co-expressing the Ub-PCNA as
two polypeptide fragments that self-assemble in vivo to form ring-shaped trimers, was
used previously to determine the X-ray crystal structure of Ub-PCNA [215]. Moreover,
the split-fused Ub-PCNA functions the same as genuine Ub-PCNA in vitro [215]. I used
native PAGE to examine the stability of the wild-type and mutant Ub-PCNA trimers (See
Fig. 2.4). The wild-type Ub-PCNA protein formed a stable trimer under all
concentrations tested (0.01 to 5 mg/ml). At lower concentrations (0.01 to 0.1 mg/ml), the
E113G mutant Ub-PCNA protein was not a trimer. However, at higher concentrations
(0.5 to 5 mg/ml), the E113G mutant Ub-PCNA protein was a trimer. At all concentrations
tested (0.5 to 5 mg/ml), the G178S mutant Ub-PCNA proteins was completely aggregated
indicating that it is highly unstable. Because of the extreme instability of the G178S
mutant PCNA and Ub-PCNA proteins, all subsequent biochemical studies were
performed only with the E113G mutant Ub-PCNA protein.
86
Interactions of the E113G mutant PCNA protein with
DNA polymerases.
One possible way that the E113G substitution in PCNA prevents TLS is by
interfering with the ability of PCNA to interact with non-classical polymerases. To
examine this possibility, I have used enzyme-linked immunosorbent assays (ELISAs) to
determine if the E113G mutant PCNA protein can bind pol , the prototypical non-
classical polymerase. Pol was immobilized in the wells of a microtiter plate, and
various concentrations of the wild-type or E113G mutant PCNA protein was added (Fig.
2.5A). The absorbance measured is proportional to the amount of PCNA bound to the
immobilized pol . There was no difference in binding observed between the wild-type
and mutant PCNA proteins. I carried out similar experiments with the wild-type or
E113G mutant Ub-PCNA proteins (Fig. 2.5B). There was also no observed difference in
binding in this case. Therefore, the E113G substitution does not affect the ability of either
PCNA or Ub-PCNA to bind the non-classical polymerase pol .
Impact of the E113G mutant PCNA protein on the
activity of pol .
Although the E113G mutant PCNA protein interacts with the non-classical
polymerase pol , the mutant PCNA protein may be unable to enhance the catalytic
activity of pol . To test this possibility, I measured the effect of the E113G substitution
on the ability of PCNA and Ub-PCNA to increase activity of pol by examining DNA
synthesis on non-damaged DNA templates in the presence of an excess of unlabeled
competitor DNA (Fig. 2.6A). Under these conditions, little DNA synthesis by pol was
observed. However, in the presence of the wild-type PCNA and Ub-PCNA proteins,
DNA synthesis by pol was clearly visible with some of the products extending to the
end of the DNA template. Significantly less pol activity, however, was observed with
87
the E113G mutant PCNA and Ub-PCNA proteins. Thus, the E113G mutant PCNA and
Ub-PCNA proteins are unable to fully enhance the catalytic activity of pol .
Because these experiments were carried out in the presence of an excess of
unlabeled competitor DNA, it was possible to examine the effect of the E113G
substitution on the ability PCNA and Ub-PCNA to increase the processivity of pol .
Processivity is a measure of the number of nucleotides a DNA polymerase incorporates
per DNA-binding event. For example, in the presence of the wild-type and the E113G
mutant PCNA proteins, 10 to 20 % of the total extension products resulted from at least
four nucleotide incorporations (Table 2.3). In the presence of the wild-type and the
E113G mutant Ub-PCNA proteins, 20 to 30 % of the extension products resulted from at
least four nucleotide incorporations. Overall, these results show Ub-PCNA increases the
processivity of pol to a somewhat greater extent than does PCNA. Moreover, these
results show that despite a clear effect on the ability of these mutant proteins to enhance
the catalytic activity of pol , they do not significantly affect the processivity of pol .
I next examined the activity of pol on a DNA substrate containing an abasic site
at the sixth position in template following the primer terminus (Fig. 2.6B and Table 2.3).
Again, in the absence of PCNA or Ub-PCNA, little DNA synthesis by pol was
observed. With the wild-type and E113G mutant PCNA proteins, 10 to 30 % of the
extension products resulted from DNA synthesis up to the abasic site, and 1 to 5 % of the
products resulted from incorporation opposite the abasic site. With the wild-type and
E113G mutant Ub-PCNA proteins, 30 to 50 % of the products and 10 to 20 % of the
products resulted from incorporation up to and opposite the abasic site, respectively.
These results further show that despite a clear effect on the ability of these mutant
proteins to enhance the catalytic activity of pol , they do not significantly affect the
processivity of pol .
88
Impact of the E113G mutant PCNA protein on the
catalytic efficiency of pol .
To quantitatively examine the effects of the E113G substitution in PCNA and in
Ub-PCNA to increase the activity of non-classical pol , I used steady state kinetics to
measure the catalytic efficiency (Vmax/Km) of nucleotide incorporation by pol (Fig. 2.7
and Table 2.4). For the non-damaged template, the wild-type PCNA protein increases the
catalytic efficiency of nucleotide incorporation by 2.7-fold on average and the wild-type
Ub-PCNA protein increases it by 9.5-fold on average. These values are consistent with
previous reports [215]. Interestingly, neither the E113G mutant PCNA protein nor the
E113G mutant Ub-PCNA protein increases the catalytic efficiency of nucleotide
incorporation by pol on non-damaged templates. These results show that the E113G
substitution in both PCNA and Ub-PCNA has a significant effect on the efficiency of pol
.
I next examined the impact of the E113G substitutions on the ability of PCNA
and Ub-PCNA to increase the catalytic efficiency of nucleotide incorporation by non-
classical polymerases when they are confronted with a kinetic barrier, such as a non-
cognate lesion. Abasic sites are non-cognate lesions for pol and reduce its efficiency of
nucleotide incorporation substantially [109]. For the template abasic site, the wild-type
PCNA protein increases the catalytic efficiency of nucleotide incorporation by 2.5-fold
on average and the wild-type Ub-PCNA protein increases it by 4.2-fold on average.
Similar to what was observed with the non-damaged DNA, neither the E113G mutant
PCNA protein nor the E113G mutant Ub-PCNA protein notably increase the catalytic
efficiency of nucleotide incorporation by pol on non-damaged templates. Thus in the
context of both non-damaged DNA and a non-cognate lesion, the E113G substitution
interfered with the ability of PCNA and Ub-PCNA to increase the efficiency of pol .
Moreover, this substitution also interferes with the ability of Ub-PCNA to increase the
efficiency of incorporation to a greater extent than PCNA does.
89
Impact of the E113G mutant PCNA protein on the
activity of pol .
Because the E113G substitution interferes with the ability of both PCNA and Ub-
PCNA to increase the efficiency of nucleotide incorporation by the non-classical
polymerase pol , I next examined whether this substitution affects the activity of pol ,
the prototypical classical DNA polymerase. I first examined the ability of pol to
synthesize DNA on non-damaged DNA templates (Fig. 8A). In all cases (pol alone, pol
with the wild-type PCNA protein, pol with the wild-type Ub-PCNA protein, pol
with the E113G mutant PCNA protein, and pol with the E113G mutant Ub-PCNA
protein), robust DNA synthesis by pol was clearly visible with many of the products
extending to the end of the DNA template. Again, because these experiments were
carried out in the presence of an excess of unlabeled competitor DNA, it was possible to
measure the processivity of pol (Table 2.5). In all cases, 70 to 80 % of the extension
products resulted from at least four incorporations, and 50 to 60 % of the products were
full length. Because pol synthesizes with such high processivity under these conditions,
I cannot conclude that the E113G mutant PCNA and Ub-PCNA proteins increase the
processivity of pol , but I can conclude that they do not reduce it.
I next examined the activity and the processivity of pol on the abasic site-
containing DNA substrate (Fig. 2.8B and Table 2.5). In the absence of PCNA, 50 % of
the extension products resulted from DNA synthesis up and opposite the abasic site. In all
other cases, 70 to 80 % of the extension products resulted from DNA synthesis up to the
abasic site, and 60 to 70 % of the products resulted from incorporation opposite the
abasic site. In addition, in all cases, 30 to 40 % of the products were full length. Again,
because pol synthesizes with such high processivity, I can only conclude that the
E113G mutant PCNA and Ub-PCNA proteins do not reduce the processivity of pol .
90
Impact of the E113G mutant PCNA protein on the
catalytic efficiency of pol .
I next examined whether this substitution interferes with the ability of PCNA and
Ub-PCNA to increase the efficiency of pol (Fig. 2.9 and Table 2.6). For the non-
damaged template, the wild-type PCNA protein increases the catalytic efficiency of
nucleotide incorporation by 10-fold on average and the wild-type Ub-PCNA protein
increases it by 20-fold on average. The E113G mutant PCNA protein increases the
catalytic efficiency of nucleotide incorporation by 6.7-fold on average and the E113G
mutant Ub-PCNA protein increases it by 4.8-fold on average. Although the E113G
substitution in PCNA and Ub-PCNA does cause a reduction in the efficiency of
nucleotide incorporation by pol , the mutant PCNA and Ub-PCNA proteins still
significantly increase the efficiency of nucleotide incorporation. Thus, the impact of this
substitution on the activity of classical pol on non-damaged DNA templates is quite
different from the impact on the activity of non-classical pol .
I next examined the impact of the E113G substitutions on the ability of PCNA
and Ub-PCNA to increase the catalytic efficiency of nucleotide incorporation by classical
polymerases when they are confronted with a kinetic barrier, such as a DNA lesion.
Abasic sites significantly reduce the efficiency of nucleotide incorporation by pol , but
the presence of PCNA increases this efficiency [262]. Thus, I examined whether the
E113G substitution interferes with the ability of PCNA and Ub-PCNA to increase the
efficiency nucleotide incorporation opposite abasic sites by pol . I found that the wild-
type PCNA protein increases the catalytic efficiency of nucleotide incorporation by 14-
fold on average and the wild-type Ub-PCNA protein increases it by 20-fold on average.
By contrast, the E113G mutant PCNA protein increases the catalytic efficiency of
nucleotide incorporation by only 3.2-fold on average and the E113G mutant Ub-PCNA
protein increases it by only 3.6-fold on average. Therefore, the E113G substitution does
interfere with the ability of both PCNA and Ub-PCNA to increase the efficiency of
91
incorporation opposite abasic sites by pol . Moreover, this effect is more pronounced
than the effect on nucleotide incorporation opposite non-damaged templates by pol .
Impact of the G178S mutant PCNA protein on the
activity of pol .
Finally, because the G178S mutant PCNA protein exists primarily in the
monomeric form, I examined whether this substitution interferes with the ability of
PCNA to increase the activity of classical pol on non-damaged DNA templates (Fig.
2.10). Interestingly, the G178S mutant PCNA protein does not stimulate the activity of
pol to the same degree as does the wild-type or the E113G mutant PCNA proteins. In
fact, the amount of DNA synthesis observed in the presence of the G178S mutant PCNA
protein is approximately the same as that observed in the absence of PCNA. This is
surprising because yeast producing the wild-type PCNA protein and yeast producing the
G178S mutant PCNA protein grow at the same rate [238]. This suggests either that the
presence of other replication fork components help stabilize the trimeric form of the
G178S mutant PCNA protein in vivo or that the slower rate of DNA synthesis by pol in
the presence of the G178S mutant PCNA protein does not limit the cell cycle in yeast.
Discussion
Inhibition of TLS in yeast by mutant PCNA proteins.
Sporadic mutations play an important role in causing a wide range of diseases,
such as cancer, autism, diabetes, and schizophrenia [59, 263-266]. Because many
mutations results from TLS, it is important to understand how this process can be
inhibited. Here I focused on two variant forms of PCNA that block TLS and thereby
mutagenesis in yeast without affecting normal replication and cell growth [238-240]. The
E113G and the G178S amino acid substitutions are adjacent to one another at the subunit
92
interface of PCNA. Although both substitutions inhibit TLS by non-classical polymerases
Rev1 and pol , there is some disagreement regarding their impact on pol . It has been
suggested that the G178S mutant PCNA protein blocks TLS by pol [238], whereas the
E113G mutant protein does not [267]. I believe that these differences are overstated. For
example, the lack of inhibition of pol -dependent TLS by the E113G substitution is
based on epistasis analysis showing that the UV sensitivity of strains producing the
E113G mutant protein is made greater by the lack of pol [267]. However, the
difference in UV sensitivities in the absence and presence of pol is very slight and is
only apparent at the highest dose of UV radiation used in the assay. Moreover, it should
be noted that this result only shows that pol is functional to some extent in these cells;
this result is consistent with a partial defect in pol -dependent TLS. In any case, my
findings reported here clearly show that the E113G mutant PCNA protein does not
support TLS by pol in vitro. Therefore, E113G and G178S substitutions likely inhibit
TLS by the same underlying mechanism.
Mechanism of TLS inhibition by mutant PCNA proteins.
The mechanism by which these two mutant PCNA proteins block TLS in cells
remains unclear. It was previously shown that the E113G mutant PCNA protein is
ubiquitylated in response to DNA damage [240]. This shows that the inhibition of TLS is
not due to an inability of these PCNA mutant proteins to become modified with ubiquitin.
It was also reported that the non-classical DNA polymerases Rev1 and pol do not
physically interact with the E113G mutant PCNA protein in vitro [267]. This lead to the
view that the inhibition of TLS is due to the inability of these polymerases to interact
with the mutant PCNA protein [267]. It is not known, however, whether these
polymerases fail to interact with the mutant Ub-PCNA protein, which is the biologically
relevant state. This is particularly important given that non-classical polymerases possess
ubiquitin-binding motifs for interacting with the ubiquitin moiety on Ub-PCNA [185]. In
93
any event, I have shown here that non-classical pol binds the wild-type and the E113G
mutant PCNA proteins with the same affinity. Others have shown that the E113G mutant
PCNA protein binds classical pol [267]. Thus, although a lack of binding may explain
the inability of this mutant protein to support TLS by non-classical polymerases Rev1
and pol , it does not explain the inability to support TLS by pol and pol that I
observed here.
PCNA has been shown to stimulate nucleotide incorporation by pol on both
non-damaged and damaged DNA templates [113, 219], and Ub-PCNA stimulates this
activity even further [191, 215]. One of the objectives of this study was to assess the
impact of the G178S and E113G amino acid substitutions on the ability of PCNA and
Ub-PCNA to stimulate the activity of pol . Because of the high instability of the G178S
mutant Ub-PCNA protein, I have focused on the E113G substitution. I showed that the
E113G mutant PCNA protein does not increase the efficiency of nucleotide incorporation
by pol in either the absence or presence of ubiquitylation. This finding has two direct
implications. First, the effect of this substitution on the activity of pol occurs even with
the non-ubiquitylated form of PCNA. Second, the additional ability to stimulate the
activity of pol afforded to PCNA by its monoubiquitylation is entirely eliminated by
this substitution. Overall, these findings show that these mutant PCNA proteins blocks
TLS by failing to form productive complexes with non-classical polymerases that
stimulate their activity.
Inhibition of TLS is caused by PCNA trimer instability.
The structures of the G178S and E113G mutant PCNA proteins have been
determined [219, 260]. It was suggested based on these structures that the aberrant
position of loop J (residues 105 to 110), which is immediately adjacent to the subunit
interface, might be responsible for the inability of these mutant proteins to support the
activity of non-classical polymerases. It was suggested that this loop may be a key
94
binding site for non-classical polymerases. However, I have re-examined these structures
and noted that the separation of the -strands constituting the subunit interface is a more
substantial change than the movement of loop J. In fact, the separation of these -strands
is directly responsible for the movement of loop J in these mutant proteins. I also found
that this structural perturbation causes a decrease in trimer stability, which I believe to be
the cause for the inability of the mutant forms of PCNA to facilitate TLS. It should be
noted that although these decreases in trimer stability are sufficient to inhibit TLS, they
are not severe enough to cause a defect in normal DNA replication and cell growth. In
further support of this notion, I found that TLS by classical pol is also impacted by the
E113G substitution as well. It is unlikely that classical pol and all non-classical
polymerases share the same key binding site at the subunit interface. It is more likely that
the effect of these mutant proteins on TLS is an indirect effect mediated by their trimer
instability. Finally, we have carried out random mutagenesis of the PCNA subunit
interface, and we found that these mutant PCNA proteins lead to a range of phenotypes.
While we found that some of these mutant PCNA proteins cause defects in cell growth,
we identified over ten additional mutant proteins that have no impact on cell growth, but
block TLS in vivo (see Chapter 4). This further confirms the importance of the integrity
of the PCNA subunit interface and trimer stability for TLS.
Model of TLS inhibition by PCNA trimer instability.
To understand how PCNA trimer instability inhibits TLS by both classical and
non-classical polymerases alike, we must first consider how the wild-type PCNA protein
increases the catalytic efficiency of classical an non-classical polymerases. The catalytic
efficiency is determined by the individual steps of the overall polymerase reaction,
especially the nucleotide-binding step and the nucleotide-incorporation step. Thus PCNA
must either increase the binding affinity for the incoming nucleotide, increase the rate of
the nucleotide-incorporation step, or both. The only way that PCNA can have such an
95
effect is by maintaining the polymerases in a more catalytically competent
conformational state – i.e., one that either binds the incoming nucleotide tighter,
incorporates it faster, or both. Interestingly, PCNA must be able to maintain polymerases
in this more catalytically competent state for different periods of time depending on the
DNA polymerase and the template. In the case of fast nucleotide incorporation, such as
with classical polymerases on non-damaged templates, PCNA would only need to
maintain the polymerase in the more competent state for a very brief time. In the case of
slow nucleotide incorporation, such as with both classical and non-classical polymerases
on damaged templates, PCNA would need to maintain the polymerase in this state
significantly longer.
Based on these considerations, I propose that a moderate degree of trimer
instability is causing the mutant PCNA proteins to occasionally undergo a conformational
change that is compromising its ability to maintain the more catalytically competent state
of the bound polymerase. This might involve a transient opening of the PCNA ring at the
subunit interface, but other possibilities exist. In any event, in cases where PCNA only
needs to maintain the competent state for a short period of time (such as a classical
polymerase with non-damaged templates), the likelihood of the mutant PCNA protein
undergoing this conformational change and disrupting the competent state of the
polymerase is small. Consequently, normal replication is largely unaffected. However, in
cases where PCNA needs to maintain the competent state for a longer period of time
(such as a classical polymerase with damaged templates or a non-classical polymerase
with either non-damaged or damaged templates), the likelihood of the mutant PCNA
protein undergoing this conformational change and disrupting the competent state of the
polymerase is much greater. While this model nicely explains all of my observations,
more work on the structure and dynamics of the wild-type and mutant PCNA proteins
and their complexes with DNA polymerases will be necessary to fully understand the
mechanism of TLS inhibition.
96
Figure 2.1. The subunit interface of the wild-type and mutant PCNA proteins.
Structures of the subunit interface of the wild-type PCNA protein (1PLQ.pdb), the
E113G mutant PCNA protein (3GPM.pdb), and the G178S mutant PCNA protein
(3F1W.pdb). -strand I1 (residues 110 to 117) is shown in blue and -strand D2 (residues
175 to 182) is shown in red. The dashed lines show the positions of the backbone
hydrogen bonds that constitute the PCNA subunit interface. The distances between the
backbone hydrogen bond donors and acceptors are provides in Table 2.1.
97
Table 2.1. Distances between potential hydrogen bond donor and acceptor atoms at
the PCNA subunit interface.
Donor atom a
Acceptor
atom a
Wild-type
PCNA
protein b
E113G mutant
PCNA
protein b
G178S mutant
PCNA
protein b
K117 (N)
I175 (O) 3.0 Å 3.5 Å 2.9 Å
S177 (N)
S115 (O) 3.1 Å 3.0 Å 2.9 Å
S115 (N)
S177 (O) 2.8 Å 3.0 Å 2.8 Å
S179 (N)
E113 (O) 2.9 Å 3.2 Å (4.1 Å)
E113 (N)
S179 (O) 2.9 Å 3.2 Å (5.6 Å)
I181 (N)
I111 (O) 3.1 Å (4.5 Å) (7.5 Å)
I111 (N)
I181 (O) 3.1 Å (5.9 Å) (9.3 Å)
a (N) refers to the amide nitrogen atom of the residue, and (O) refers to the carbonyl
oxygen atom of the residue.
b
Values in parentheses are distances that are too large to allow for hydrogen bonding.
98
Figure 2.2. Analysis of the wild-type and mutant PCNA proteins by native gel
electrophoresis. (A) Coomassie stained non-denaturing polyacrylamide gradient gel (4 to
20 %) in which solutions of the wild-type and mutant PCNA proteins (0.05 to 0.2 mg/ml)
were run. The positions of the PCNA monomer and PCNA trimer are shown. (B)
Coomassie stained non-denaturing polyacrylamide gel in which solutions of the wild-type
and mutant PCNA proteins (0.50 to 5.0 mg/ml) were run. The positions of the PCNA
monomer and PCNA trimer are shown.
0.0
5
0.1
0
0.2
0
0.0
5
0.1
0
0.2
0
0.0
5
0.1
0
0.2
0
WT
PCNA
E113G
PCNA
G178S
PCNA
WT
PCNA
E113G
PCNA
G178S
PCNA
Trimer
Monomer
Trimer
Monomer
(mg/ml)
A
B
0.5
0
1.0
2.0
5.0
0.5
0
1.0
2.0
5.0
0.5
0
1.0
2.0
5.0
(mg/ml)
99
Figure 2.3. Analysis of the wild-type and mutant PCNA proteins by size exclusion
chromatography. (A) Elution profiles of a size exclusion chromatography column in
which 0.005 mg/ml (), 0.01 mg/ml (), or 0.1 mg/ml () solutions of the wild-type
PCNA protein were run. The amounts of the monomeric and trimeric species calculated
from the area under these curves are provided in Table 2.2. (B) Elution profiles of 0.005
mg/ml (), 0.01 mg/ml (), or 0.1 mg/ml () solutions of the G178S mutant PCNA
protein. The amounts of the monomeric and trimeric species are provided in Table 2.2.
(C) Elution profiles of 0.005 mg/ml (), 0.01 mg/ml (), or 0.1 mg/ml () solutions of
the wild-type PCNA protein. The amounts of the monomeric and trimeric species are
provided in Table 2.2.
100
70 75 80 85 90
0.0000
0.0005
0.0010
0.0015
0.0020
70 75 80 85 90 95
0.0000
0.0002
0.0004
0.0006
0.0008
0.0010
0.0012
70 75 80 85 90
0.0000
0.0002
0.0004
0.0006
0.0008
0.0010
0.0012
Elution volume (ml)
Elution volume (ml)
Elution volume (ml)
Ab
sorb
ance
A
bso
rban
ce
Ab
sorb
ance
WT
PCNA
E113G
PCNA
G178S
PCNA
A
B
C
101
Table 2.2. Percentage of PCNA proteins in the monomeric state as determined by
size exclusion chromatographya.
0.005 mg/ml 0.01 mg/ml
0.1 mg/ml
WT PCNA
7 % 1 % < 1 %
E113G PCNA
20 % 7 % 2 %
G178S PCNA
100 % 100 % 100 %
a These percentages were obtained by calculating the area under the curve of the size
exclusion chromatograms.
102
Figure 2.4. Stability of the wild-type and mutant Ub-PCNA proteins. The oligomeric
state of the Ub-PCNA proteins (0.01 to 5 mg/ml) was analyzed by non-denaturing
polyacrylamide gradient (4-20%) gel electrophoresis. The gels were coomassie stained,
and the positions of the PCNA trimer, the Ub-PCNA trimer, and the aggregated protein
are indicated.
WT
UbPCNA
E113G
UbPCNA
G178S
UbPCNA
0.5
0
1.0
2.0
5.0
0.5
0
1.0
2.0
5.0
0.5
0
1.0
2.0
5.0
5.0
WT
PCNA
UbPCNA Trimer
Aggregated protein
PCNA Trimer
(mg/ml)
UbPCNA Trimer
PCNA Trimer
WT
PCNA
0.0
1
0.0
5
0.1
E113G
PCNA 0
.01
0
.05
0
.1
WT
UbPCNA
0.0
1
0.0
5
0.1
E113G
UbPCNA
0.0
1
0.0
5
0.1
103
Figure 2.5. Interaction of the wild-type and mutant PCNA and Ub-PCNA proteins
with pol . (A) Results of an ELISA assay showing the interaction of the wild-type
PCNA protein () and the E113G mutant PCNA protein () with pol . (B) Results of
an ELISA assay showing the interaction of the wild-type Ub-PCNA protein () and the
E113G mutant Ub-PCNA protein () with pol .
0 2 4 6 8 10
0.0
0.2
0.4
0.6
0.8
0 2 4 6 8 10
0.0
0.2
0.4
0.6
0.8
1.0
PCNA (g)
Ab
sorb
anc
e
UbPCNA (g)
Ab
sorb
ance
A
B
104
Figure 2.6. Processive DNA synthesis by pol in the presence of the wild-type and
mutant PCNA proteins. (A) Autoradiogram of the extension products of pol on a non-
damaged DNA template in the absence of PCNA (N) or in the presence of the wild-type
PCNA protein (P), the wild-type Ub-PCNA protein (U), the E113G mutant PCNA
protein (P*), or the E113G mutant Ub-PCNA protein (U*). The percentages of extension
products at least 4 nt. in length, at least 9 nt. in length, or full length are provided in Table
2.3. (B) Autoradiogram of the extension products of pol on a DNA template containing
an abasic site in the absence of and presence of the wild-type and E113G mutant PCNA
and Ub-PCNA proteins. The gel band representing extension products 6 nt. in length,
which corresponds to incorporation opposite the abasic site, is indicated by the arrow.
The percentages of extension products at least 5 nt. in length, at least 6 nt. in length, or
full length are provided in Table 2.3.
P U P* U* N
B
A
P U P* U* N
WT PCNA P
N
U
P*
U*
No PCNA
WT UbPCNA
E113G PCNA
E113G UbPCNA
105
Table 2.3. Processivity of pol η on non-damaged and damaged DNA.
≥ 4 nt. a ≥ 9 nt.
a
Full length
PCNA
Non-damaged 20 % < 1 % < 1 %
UbPCNA
Non-damaged 30 % 9 % < 1 %
E113G PCNA
Non-damaged 10 % < 1 % < 1 %
E113G
UbPCNA
Non-damaged 20 % 2 % < 1 %
≥ 5 nt. c ≥ 6 nt.
c
Full length
b
PCNA
Abasic site 30 % 5 % nd
UbPCNA
Abasic site 50 % 20 % nd
E113G PCNA
Abasic site 10 % 1 % nd
E113G
UbPCNA
Abasic site 30 % 10 % nd
nd, not detectible
a These percentages reflect the amount of extended products at least 4 nt. in length or at
least 9 nt. in length.
b These percentages reflect the amount of products that were extended all the way to the
end of the template.
c These percentages reflect the amount of extended products at least 5 nt. in length (which
corresponds to incorporation opposite the template residue on the 3' side of the abasic
site) or at least 6 nt. in length (which corresponds to incorporation opposite the abasic
site).
106
Figure 2.7. Steady state kinetics of pol in the presence of the wild-type and mutant
PCNA proteins. (A) The catalytic efficiency (Vmax/Km) of nucleotide incorporation by
pol on a non-damaged template G in the absence of PCNA (N) or in the presence of the
wild-type PCNA protein (P), the wild-type Ub-PCNA protein (U), the E113G mutant
PCNA protein (P*), or the E113G mutant Ub-PCNA protein (U*). The individual Vmax
and Km parameters are provided in Table 2.4. (B) The catalytic efficiency of
incorporation by pol on a template abasic site in the absence of and presence of the
wild-type and E113G mutant PCNA and Ub-PCNA proteins. The Vmax and Km
parameters are provided in Table 2.4.
Vm
ax/K
m
0.000
0.002
0.004
0.006
0.008
0.010
0.012
0.014
Vm
ax/K
m0.00
0.05
0.10
0.15
0.20
0.25
0.30
A
B
P U P*
U*
N
P U P*
U*
N
107
Table 2.4. Steady state kinetics of nucleotide incorporation by pol η.
Template PCNA Vmax (nM/min) a Km (M)
a
Vmax/Km
Non-damaged G
No PCNA 0.44 0.14 22 1 0.021 0.007
Non-damaged G
PCNA 0.40 0.10 7.1 2.4 0.056 0.024
Non-damaged G
UbPCNA 0.43 0.15 2.2 0.1 0.20 0.07
Non-damaged G
E113G PCNA 0.29 0.08 11 4 0.026 0.014
Non-damaged G
E113G UbPCNA 0.28 0.10 12 2 0.023 0.009
Abasic site
No PCNA 0.090 0.040 42 11 0.0021
0.0011
Abasic site
PCNA 0.14 0.04 27 8 0.0052
0.0022
Abasic site
UbPCNA 0.14 0.04 16 3 0.0088
0.0032
Abasic site
E113G PCNA 0.047 0.009 16 2 0.0029
0.0007
Abasic site
E113G UbPCNA 0.047 0.005 16 1 0.0029
0.0003
a Mean and standard errors for the Vmax and Km values were calculated from five
independent experiments.
108
Figure 2.8. Processive DNA synthesis by pol in the presence of the wild-type and
mutant PCNA proteins. (A) Autoradiogram of the extension products of pol on a non-
damaged DNA template in the absence of PCNA (N) or in the presence of the wild-type
PCNA protein (P), the wild-type Ub-PCNA protein (U), the E113G mutant PCNA
protein (P*), or the E113G mutant Ub-PCNA protein (U*). The percentages of extension
products at least 4 nt. in length, at least 9 nt. in length, or full length are provided in Table
2.5. (B) Autoradiogram of the extension products of pol on a DNA template containing
an abasic site in the absence of and presence of the wild-type and E113G mutant PCNA
and Ub-PCNA proteins. The gel band representing extension products 6 nt. in length,
which corresponds to incorporation opposite the abasic site, is indicated by the arrow.
The percentages of extension products at least 5 nt. in length, at least 6 nt. in length, or
full length are provided in Table 2.5.
A B
P U P* U* N P U P* U* N
109
Table 2.5. Processivity of pol δ on non-damaged and damaged DNA.
≥ 4 nt. a ≥ 9 nt.
a
Full length b
No PCNA
Non-damaged 70 % 60 % 50 %
PCNA
Non-damaged 80 % 70 % 50 %
UbPCNA
Non-damaged 80 % 70 % 60 %
E113G PCNA
Non-damaged 70 % 60 % 50 %
E113G UbPCNA
Non-damaged 70 % 70 % 50 %
≥ 5 nt. c ≥ 6 nt.
c
Full length b
No PCNA
Abasic site 50 % 50 % 30 %
PCNA
Abasic site 70 % 70 % 30 %
UbPCNA
Abasic site 80 % 70 % 40 %
E113G PCNA
Abasic site 70 % 60 % 30 %
E113G UbPCNA
Abasic site 70 % 70 % 30 %
a These percentages reflect the amount of extended products at least 4 nt. in length or at
least 9 nt. in length.
b These percentages reflect the amount of products that were extended all the way to the
end of the template.
c These percentages reflect the amount of extended products at least 5 nt. in length (which
corresponds to incorporation opposite the template residue on the 3' side of the abasic
site) or at least 6 nt. in length (which corresponds to incorporation opposite the abasic
site).
110
Figure 2.9. Steady state kinetics of pol in the presence of the wild-type and mutant
PCNA proteins. (A) The catalytic efficiency (Vmax/Km) of nucleotide incorporation by
pol on a non-damaged template G in the absence of PCNA (N) or in the presence of the
wild-type PCNA protein (P), the wild-type Ub-PCNA protein (U), the E113G mutant
PCNA protein (P*), or the E113G mutant Ub-PCNA protein (U*). The individual Vmax
and Km parameters are provided in Table 2.6. (B) The catalytic efficiency of
incorporation by pol on a template abasic site in the absence of and presence of the
wild-type and E113G mutant PCNA and Ub-PCNA proteins. The Vmax and Km
parameters are provided in Table 2.6.
Vm
ax/K
m
0.0
0.1
0.2
0.3
0.4
0.5
0.6
Vm
ax/K
m
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
A
B
P U P*
U*
N
P U P*
U*
N
111
Table 2.6. Steady state kinetics of nucleotide incorporation by pol δ.
Template PCNA Vmax (nM/min) a Km (M)
a
Vmax/Km
Non-damaged G
No PCNA 0.59 0.10 2.2 0.6 0.27 0.09
Non-damaged G
PCNA 1.7 0.2 0.60 0.20 2.8 1.1
Non-damaged G
UbPCNA 1.6 0.1 0.30 0.05 5.3 0.9
Non-damaged G
E113G PCNA 0.77 0.08 0.44 0.11 1.8 0.5
Non-damaged G
E113G UbPCNA 0.78 0.06 0.59 0.21 1.3 0.5
Abasic site
No PCNA 0.078 0.011 3.7 0.8 0.021 0.005
Abasic site
PCNA 0.23 0.02 0.77 0.19 0.30 0.08
Abasic site
UbPCNA 0.25 0.03 0.58 0.14 0.43 0.12
Abasic site
E113G PCNA 0.14 0.02 2.1 0.6 0.067 0.021
Abasic site
E113G UbPCNA 0.13 0.02 1.7 0.5 0.076 0.025
a Mean and standard errors for the Vmax and Km values were calculated from five
independent experiments.
112
Figure 2.10. Processive DNA synthesis by pol δ in presence of the E113G and G178S
mutant PCNA proteins. (A) Autoradiogram of the extension products of pol δ on a
non-damaged template in the absence of PCNA (N) or in the presence of the wild-type
PCNA protein (P), the E113G mutant PCNA protein (E), or the G178S mutant PCNA
proteins (G) after a 1 minute reaction time. (B) Autoradiogram of these same extension
products after a 2 minute reaction time.
A
P E G N P E G N
B
113
CHAPTER 3
DISTINCT STRUCTURAL ALTERATIONS IN PCNA BLOCK DNA
MISMATCH REPAIR
Abstract
During DNA replication, mismatches and small loops in the DNA resulting from
insertions or deletions are repaired by the mismatch repair (MMR) machinery.
Proliferating cell nuclear antigen (PCNA) plays an important role in both the mismatch-
recognition stage and the resynthesis stage of MMR. Previously, two mutant forms of
PCNA were identified that cause defects in MMR with little if any other defects. The
C22Y mutant PCNA protein completely blocks MutS-dependent MMR, and the C81R
mutant PCNA protein partially blocks both MutS-dependent and MutS-dependent
MMR. In order to understand the structural and mechanistic basis by which these two
amino acid substitutions in PCNA proteins block MMR, we solved the X-ray crystal
structures of both mutant proteins and carried out further biochemical studies. We found
that these amino acid substitutions lead to distinct structural changes in PCNA. The
C22Y substitution alters the positions of the -helices lining the central hole of the
PCNA ring, whereas the C81R substitution creates a distortion in the -sheet at the
PCNA subunit interface. We conclude that the structural integrity of the -helices lining
the central hole and the -sheet at the subunit interface are both necessary to form
productive complexes with MutS and mismatch-containing DNA. (The work described
in this Chapter has been submitted for publication in Biochemistry. Dieckman, L.M.*,
Boehm, E.M.*, Hingorani, M.M., and Washington, M.T. (2013). *These authors
contributed equally to this work)
114
Introduction
Inaccurate DNA replication can result in base-base mismatches and small loops
arising from insertions or deletions. These mismatches and loops are recognized and
repaired by the mismatch repair (MMR) machinery. The mechanisms of MMR in E. coli
have been studied extensively and are relatively well understood [22, 25, 125, 141, 268,
269]. The MutS protein is a homodimer that recognizes base-base mismatches and small
nucleotide insertions and deletions. The MutL protein is a homodimer that interacts with
MutS in an ATP-dependent manner to initiate MMR [270-274]. Next, the MutH
endonuclease is activated and generates a nick in the newly synthesized, unmethylated
DNA strand of a hemimethylated duplex [275]. Subsequent steps include unwinding and
degradation of the newly synthesized DNA strand and filling in of the resulting gap by
DNA polymerase III [22, 25, 125, 141, 268, 269].
The mechanisms of MMR in eukaryotes are more complicated and are not as well
resolved. In yeast, there are six MutS homologs designated MSH1 to MSH6; in
mammals, there are five, MSH2 to MSH6. These proteins function as heterodimers with
specialized functions. For example, MSH2 and MSH6 form a heterodimer called MutS,
which recognizes base-base mismatches and small loops [33, 34, 276-278]. By contrast,
MSH2 and MSH3 form a heterodimer called MutS, which recognizes longer loops [32,
279, 280]. In addition to the MutS homologs, there are several MutL homologs, including
MLH1, MLH2, MLH3, and PMS1, which also function as heterodimers. The best
characterized of these is MutL (MLH1/PMS1 in yeast and MLH1/PMS2 in humans),
which functions with both MutS and MutS [37, 116, 281-283]. Mutations in both
MutS and MutL homologs that disrupt mismatch repair cause sporadic and hereditary
human cancers, including hereditary nonpolyposis colorectal cancer (HNPCC) [24-26,
284-286]. Other key proteins involved in the subsequent excision and resynthesis steps
include exonuclease I (EXOI) [39-42, 287], DNA polymerase delta (pol ) [43],
115
replication protein A (RPA) [44, 45], replication factor C (RFC) [47], and proliferating
cell nuclear antigen (PCNA) [51, 54, 288].
The PCNA clamp is an essential replication accessory protein that forms a ring-
shaped homotrimer and encircles duplex DNA [156]. In addition to serving as a
processivity factor for DNA polymerases, it interacts with a wide variety of proteins and
plays important roles in DNA replication, repair, recombination, translesion synthesis,
chromatin remodeling, sister chromatid cohesion, and cell cycle regulation [155, 199,
200, 257, 258, 289]. During MMR, PCNA functions in the initiation and mismatch
recognition stage as well as the excision and resynthesis stage. The role of PCNA in the
initial stage of MMR is not well understood. PCNA interacts with both MutS and
MutS and is thought to facilitate their recruitment to mismatches [53-57]. Moreover, it
has been suggested that PCNA plays a role in strand discrimination, i.e., the recognition
of the newly synthesized daughter strand [50-52].
Various PCNA mutant alleles have been identified that lead to elevated mutation
rates [242, 288, 290, 291]. Genetic studies have shown that two of these mutant alleles,
pol30-201 and pol30-204, specifically disrupt the MMR pathway with little if any effect
on other DNA metabolic processes [242]. The pol30-201 allele, which encodes the C22Y
mutant PCNA protein, causes a strong defect in MutS-dependent MMR, and the pol30-
204 allele, which encodes the C81R mutant PCNA protein, causes a partial defect in both
MutS-dependent and MutS-dependent MMR [242]. In order to understand the
structural and mechanistic basis by which these mutant PCNA proteins disrupt MMR, we
solved the X-ray crystal structures of both mutant proteins and carried out related
biochemical studies. We found that these two amino acid substitutions lead to distinct
structural changes in PCNA. The C22Y substitution alters the positions of the -helices
lining the inside of the PCNA ring, whereas the C81R substitution disrupts the -sheet at
the PCNA subunit interface. We conclude that the structural integrity of the -helices
lining the central hole and the structural integrity of the -sheet at the subunit interface
116
are both necessary to form productive complexes with MutS and mismatch-containing
DNA.
Materials and Methods
Protein expression and purification.
The wild-type and the C22Y and C81R mutant PCNA proteins from S. cerevisiae
were over-expressed as N-terminally His6-tagged proteins in E. coli and were purified as
described previously [219]. Replication factor C (RFC) from S. cerevisiae was over-
expressed in E. coli and purified as previously described [261]. DNA polymerase delta
(pol ) from S. cerevisiae was over-expressed in S. cerevisiae and purified as previously
described [75]. MutS from S. cerevisiae was over-expressed in E. coli and purified as
described previously [292].
DNA and nucleotide substrates.
The template strand used to measure pol activity was a 68-mer
oligodeoxynucleotide with the sequence 5'- GAC GGC ATT GGA TCG ACC TCX AGT
TGG TTG GAC GGG TGC GAG GCT GGC TAC CTG CGA TGA GGA CTA GC with
biotins on both ends. The X represents the position of a non-damaged G or an abasic site.
The primer strand was a 31-mer oligodeoxynucleotide with the sequence 5'-TCG CAG
GTA GCC AGC CTC GCA CCC GTC CAA C. The primer strand was 5'-32
P-end-
labeled with T4 polynucleotide kinase and (γ-32
P)ATP and annealed to the template
strand at 1 µM in 25 mM TrisCl, pH 7.5, and 100 mM NaCl at 90oC for 2 min and slowly
cooled to 30oC. A mixture of all four dNTPs (10 mM each) was purchased from New
England Biolabs.
Two 37-mer duplex DNA substrates were used in the sedimentation analysis, one
with a G/C pair at position 19 (the homoduplex) and one with a G/T mispair at position
117
19 (the heteroduplex). The DNA substrates were formed by annealing a top strand to a
bottom strand. The top strand, which had a Cy3 fluorescent tag on the 3' end, had the
sequence: 5'-ATT TCC TTC AGC AGA TAG GAA CCA TAC TGA TTC ACA T. The
bottom strand had the sequence: 5'-ATG TGA ATC AGT ATG GTT XCT ATC TGC
TGA AGG AAA T, where the X represents the position of either a C in the case of the
homoduplex or a T in the case of the heteroduplex. The annealing reactions were carried
out as described above.
Crystallization of the C22Y and C81R mutant proteins.
The C22Y mutant PCNA protein and the C81R mutant PCNA protein were
crystallized using the hanging drop method with 400 nl drops prepared using a Mosquito
Crystallization Robot (TTP Labtech). The best diffractiing C22Y mutant PCNA protein
crystals were obtained by combining an equal volume of protein (30 mg/ml) with a
reservoir containing 1.6 M ammonium sulfate and 0.1 M citric acid. Crystals formed after
3 days at 18C. The best diffracting C81R mutant PCNA protein crystals were obtained
by combining an equal volume of protein (20 mg/ml) with a reservoir containing 20%
PEG1000, 0.2 M MgCl2 hexahydrate, and 0.1 M sodium cacodylate trihydrate, pH 6.5.
Crystals formed after 3 days at 18C.
Data collection and structural determination.
The C22Y and C81R mutant PCNA protein crystals were soaked in a mother-
liquor solution containing 10% (v/v) glycerol prior to flash-cooling in liquid nitrogen.
Data were collected at the 4.2.2 synchrotron beamline at the Advanced Light Source in
the Ernest Orlando Lawrence Berkeley National Laboratory. The data were collected at
100 K with a crystal to detector distance of 200 mm and were processed and scaled using
d*TREK [293]. For the C22Y mutant protein crystals, the space group was determined to
be P212121. For the C81R mutant protein crystals, the space group was determined to be
118
P213. Molecular replacement was performed using the known structure of wild-type
PCNA [PDB 1PLQ] with PHASER [294]. Refinement was done using REFMAC5 from
CCP4 [295] and PHENIX. All model building was carried out using Coot [296].
PCNA trimer stability assays.
For non-denaturing polyacrylamide gel electrophoresis (PAGE), the wild-type
and C22Y mutant and C81R mutant PCNA proteins (0.1 to 1.0 mg/ml) were incubated in
25 mM TrisCl, pH 7.4, 150 mM NaCl, 0.01% bromophenol blue, and 10% glycerol for 5
min and then run on a TrisCl pre-cast 4-20% gradient non-denaturing polyacrylamide gel
(Bio-Rad) at 4oC at a current of 25 mA using 25 mM Tris, pH 8.3, and 0.2 M glycine
running buffer. Protein bands were visualized using Coomassie blue staining. For size
exclusion chromatography, wild-type and mutant PCNA proteins were diluted to various
concentrations (0.1 to 10 mg/mL) in 25 mM Tris, pH 7.4, 150 mM NaCl, 5% glycerol
and run on a 120 ml HiLoad 16/60 Superdex 200 PG column (GE Healthcare).
Polymerase δ activity assays.
Running start assays were performed as described previously [161]. Reactions
were carried out in the absence of PCNA and in the presence of 90 nM wild-type or
mutant PCNA proteins (trimer concentration), and contained 20 nM pol , 25 nM DNA,
and 100 µM of each of the four dNTPs. Reactions were stopped after 30 min, and the
extension products were analyzed on a 15% polyacrylamide sequencing gel containing
8 M urea.
Enzyme-linked immunosorbent assays.
The wells of a 96 well EIA/RIA plate (Corning) were coated with 0.75 µg of
MutS in PBS (4.3 mM Na2HPO4, 1.4 mM KH2PO4, pH 7.4, 137 mM NaCl, 2.7 mM
KCl) for two hours. The wells were then washed four times with PBS, 0.2% Tween-20,
119
blocked for one hour with PBS with 5% milk, and washed again. Various amounts of the
wild-type, the C22Y mutant, or the C81R mutant PCNA proteins (1 to 20 µg) in 100 µl
of PBS with 5% milk were then added to the wells and incubated for one hour. After
washing, a 1:1000 dilution of rabbit polyclonal anti-PCNA antibody in PBS with 5%
milk was added to the wells and incubated for 30 minutes. The wells were washed again,
and a 1:10,000 dilution of goat anti-rabbit antibody conjugated with horseradish
peroxidase (Jackson ImmunoResearch) in PBS with 5% milk was added and incubated
for 30 minutes. The plate was then washed, and 0.8 mg/mL of O-phenylenediamine
(Aldrich) in 0.05M phosphate-citrate buffer with 0.03% sodium perborate (Sigma) was
added. Absorbance was measured at 450 nm after various amounts of time (5 to 35
minutes) using an iMark microplate reader (BioRad). Parallel control reactions using
bovine serum albumin instead of MutS were carried out and these background
absorbance values were subtracted from the absorbance of each sample at the
corresponding PCNA protein concentration. All steps were performed at 25°C.
Sedimentation assays.
Samples (100 µl) were prepared with 300 nM MutS, 300 nM of either the wild-
type or mutant PCNA proteins (trimer concentration), and 300 nM of either the
homoduplex or heteroduplex DNA in 1xTBS buffer. The samples were incubated on ice
for 30 minutes prior to loading on a 5 ml glycerol gradients (15-30%) and were then spun
for 20 h at 45,000 rpm at 4⁰C in a Thermo Sorvall WX ultracentrifuge using an AH-651
swing bucket rotor. Sixteen 300 µl aliquots were collected from the bottom of each
gradient, were concentrated using Millipore Amicon® Ultra 10K centrifugal filters, and
were analyzed by SDS PAGE using 4-15% pre-cast gradient gels (BioRad). The Cy3-
labeled DNA in each fraction was visualized using a BioRad ChemiDoc-MP Imaging
System after which the gels were silver stained according to the BioRad Polyacrylamide
Gel Staining Procedure.
120
Results
We have examined two mutant forms of PCNA that are known to cause defects in
MMR with little if any defects in other DNA metabolic processes [242]. The C22Y
mutant PCNA protein completely blocks MutS-dependent MMR, and the C81R mutant
PCNA protein partially blocks both MutS-dependent and MutS-dependent MMR. In
order to understand the structural and mechanistic basis by which they disrupt MMR, we
solved the X-ray crystal structures of both mutant proteins and analyzed their
biochemical properties.
Structure of the C22Y mutant PCNA protein.
We first determined the X-ray crystal structure of the C22Y mutant PCNA protein
to a resolution of 2.7 Å (Table 3.1). Overall, its structure resembles that of the wild-type
PCNA protein with each subunit comprised of two domains, an N-terminal domain
(residues 1-117) and a C-terminal domain (residues 135-258) linked by a long, inter-
domain connector loop (residues 118-134) (Figure 3.1A). The subunits are arranged in a
head-to-tail fashion to form a ring-shaped trimer. Residue 22 is located in the N-terminal
domain on a loop following α-helix A1 (residues 9-20), which along with α-helices B1,
A2, and B2, form the inside ring of the PCNA trimer (Figure 3.1B). Because of its larger
size, the substituted tyrosine side chain is unable to occupy the same position as the wild-
type cysteine side chain. Consequently, the tyrosine side chain points toward the front of
the PCNA ring rather than toward α-helix B2 (Figure 3.1C). This rearrangement induces a
set of other structural changes that ultimately alter the positions of both α-helix A2 and α-
helix B2. First, the α-carbon of residue 22 is shifted 1.5 Å from its position in the wild-
type protein structure. This in turn causes the α-carbon of Asp-21 to move 0.6 Å and the
δ-oxygen of Asp-21 to move 0.9 Å from their positions in the wild-type protein structure.
This causes the -nitrogen of Lys-217, which is located in α-helix B2 and forms a
121
hydrogen bond with the δ-oxygen of Asp-21, to move 3.0 Å, and this causes the α-
carbons in α-helix B2 (residues 209-221) to move up to 1.8 Å from their positions in the
structure of the wild-type protein and the α-carbons in α-helix A2 (residues 141-153) to
move up to 1.7 Å from their positions in the wild-type protein structure. The changes in
the positions of the two α-helices in the C-terminal domain are the most notable structural
alterations in the C22Y mutant PCNA protein.
Structure of the C81R mutant PCNA protein.
We next determined the X-ray crystal structure of the C81R mutant PCNA protein
to a resolution of 3.0 Å (Table 3.1 and Figure 3.2A). Residue 81 is located in the N-
terminal domain on a loop following α-helix B1 (residues 72-79) near the subunit
interface (Figure 3.2B). The substituted arginine side chain points toward β-strand I1 and
forms two new hydrogen bonds with the η-oxygen of Tyr-114 (Figure 3.2C). This
interaction causes the α-carbon of Tyr-114, which is located in β-strand I1, to move 1.6 Å
from its position in the wild-type protein structure. This change in turn disrupts the
hydrogen bond between the carbonyl oxygen of Tyr-114 and the amide nitrogen of Leu-
101 located in β-strand H1. As a result, there is a localized distortion within the β-sheet
formed by β-strands H1 (residues 98-104) and I1 (residues 109-117). This distortion of the
β-sheet in the N-terminal domain at the PCNA subunit interface is the only notable
structural alteration in the C81R mutant PCNA protein.
Stability of the mutant PCNA proteins.
Because the C81R substitution causes a distortion in β-strand H1 at the PCNA
subunit interface, we examined the stability of the mutant PCNA trimers. First, Elizabeth
Boehm used native polyacrylamide gel electrophoresis (PAGE) to determine the
oligomeric form of the wild-type and the two mutant PCNA proteins at various
concentrations (Figure 3.3). Both the wild-type and the C22Y mutant PCNA proteins
122
were trimers at all concentrations tested (0.1 to 1.0 mg/ml). By contrast, the C81R mutant
PCNA protein did not form stable trimers as indicated by the higher mobility species at
low concentrations (0.1 and 0.2 mg/ml) and the smeared gel bands at higher
concentrations (0.5 and 1.0 mg/ml).
In order to further assess PCNA trimer stability, I analyzed the proteins by size
exclusion chromatography. When loaded onto the size exclusion column at high
concentration (10 mg/ml), the wild-type PCNA protein and the C22Y mutant protein
eluted in a narrow peak as expected for the 90-kDa trimer (Figure 3.4A and Figure 3.4B).
When the C81R mutant protein was loaded at high concentration (10 mg/ml), it eluted in
a broad peak corresponding to a mixture of trimer and dimer species (Figure 3.4C). At
lower protein concentration (0.01 mg/ml), the C81R mutant protein eluted as the 30-kDa
monomer (Figure 3.5). Taken together, both the native PAGE and size exclusion
chromatography results show that the C81R mutant PCNA protein is far less stable than
either the wild-type PCNA protein or the C22Y mutant PCNA protein.
Impact of the mutant PCNA proteins on DNA
polymerase activity.
DNA polymerase (pol ) is responsible for the majority of lagging strand
synthesis during normal DNA replication and is also involved in base excision repair,
nucleotide excision repair, mismatch repair, and double strand break repair [244-247].
To determine whether the two mutant PCNA proteins can stimulate DNA synthesis by
pol , I used running start experiments to measure pol activity in the absence and
presence of the wild-type and mutant PCNA proteins on non-damaged DNA (Figure
3.6A). The wild-type PCNA protein stimulates DNA synthesis by pol with 5-fold
more full-length runoff products formed in the presence of the wild-type PCNA protein
than in its absence. The C22Y mutant PCNA protein and the C81R mutant PCNA protein
stimulated DNA synthesis by pol to differing extents. In the presence of the C22Y
123
mutant protein, 3-fold more full-length runoff products were observed than in its
absence. In the presence of the C81R mutant protein, 2-fold more full-length runoff
products were observed.
PCNA also facilitates the bypass of abasic sites by pol , and I examined if this
activity was affected by the mutant PCNA proteins (Figure 3.6B). In the presence of the
wild-type PCNA protein, 5-fold more extension products resulting from incorporation
opposite the abasic site were observed than in the absence of PCNA. Moreover, 4-fold
more full-length runoff products were observed in the presence of the wild-type PCNA
protein. Similarly, in the presence of the C22Y mutant PCNA protein, there were 6-fold
more extension products resulting from incorporation opposite the abasic site and 3-fold
more full-length runoff products than in the absence of PCNA. In the presence of the
C81R mutant PCNA protein, there were 2-fold more extension products resulting from
incorporation opposite the abasic site and no increase in the amount of full-length
products than in the absence of PCNA.
Taken together, these experiments show that, like the wild-type PCNA protein,
the C22Y mutant PCNA protein fully supports pol function in both normal and
translesion DNA synthesis, whereas the C81R mutant PCNA protein only partially
supports pol function. It should be noted that under these assay conditions, the
concentration of PCNA was low (0.01 mg/ml), and most of the C81R mutant PCNA
protein was not expected to be trimeric. Thus, the reduced ability of the C81R mutant
PCNA protein to stimulate the activity of pol is very likely due to this mutant protein
not forming stable trimers in these assays.
Interactions of the mutant PCNA proteins with MutS.
Since PCNA is known to interact with MutS during MMR and since the C22Y
and C81R mutant PCNA proteins are defective in MMR, I examined the ability of these
mutant proteins to bind MutS. I used an enzyme-linked immunosorbent assay (ELISA)
124
to monitor this interaction (Figure 3.7A). A fixed concentration of the MutS was
immobilized in a microtiter plate and titrated with various concentrations of the wild-type
PCNA protein, the C22Y mutant PCNA protein, or the C81R mutant PCNA proteins. The
absorbance signal is proportional to the amount of PCNA bound to MutS. No
significant difference was observed between the binding of the wild-type PCNA protein
and the C22Y mutant PCNA protein to MutS. The C81R mutant PCNA protein, by
contrast, exhibited weaker binding than the wild-type and the C22Y mutant PCNA
proteins, though still significantly greater than the background. These results are
generally consistent with previously published co-sedimentation studies that showed that
the C22Y mutant PCNA protein binds MutS in vitro, but the C81R mutant PCNA
proteins does not [242]. It should be pointed out again that the concentrations of PCNA
used in both the ELISA assays reported here and the co-sedimentation assay reported
previously are less than 0.2 mg/ml. Under these conditions, most of the C81R mutant
protein is not expected to be trimeric. Thus, the apparent weaker binding of the C81R
mutant PCNA protein to MutS is very likely due to this mutant protein not forming
stable trimers in these assays.
Interactions of the mutant PCNA proteins with
MutS and DNA.
It has previously been shown that PCNA, MutS, and mismatch-containing DNA
form a stable ternary complex [54]. Elizabeth Boehm carried out sedimentation analysis
to determine if these amino acid substitutions in PCNA affect its ability to form ternary
complexes with MutS and DNA. The DNA contained a G/T mismatched base-pair
flanked on both sides by 18 matched base-pairs. In the absence of PCNA, MutS was
found mainly in fractions 5 to 9 along with DNA (Figure 3.8A), suggesting that a single
MutS protein binds to the centrally positioned G/T mismatch. In the presence of the
wild-type PCNA protein, MutS was found mainly in fractions 5 to 9 along with PCNA
125
and DNA (Figure 3.7B), suggesting that a single MutS protein and a single PCNA
protein binds to the mismatch. Interestingly, in the presence of either the C22Y mutant
PCNA protein or the C81R mutant PCNA protein, MutS was found mainly in fractions
1 to 4 along with the mutant PCNA proteins and DNA (Figure 3.8C and Figure 3.8D).
This means that the MutS-containing complex is larger in the presence of the mutant
PCNA proteins than in the presence of the wild-type PCNA protein. This may be due to
the fact that in the case of the mutant PCNA proteins, MutS no longer specifically binds
the centrally positioned G/T mismatch, and more than one MutS protein can bind the
same DNA. The precise nature of these larger complexes with the mutant PCNA proteins
remains unclear, and further structural analysis will be required to understand how and
why they form. Nevertheless, these results convincingly demonstrate that the complexes
of the mutant PCNA proteins, MutS, and mismatch-containing DNA are aberrant.
Moreover, these aberrant complexes are strictly dependent both on the presence of the
mutant PCNA proteins and on the presence of a mismatch in the DNA, as they do not
form in the presence of fully matched DNA (Figure 3.9).
Discussion
PCNA plays a critical role in many aspects of DNA metabolism and the
maintenance of genome stability. It interacts with a wide variety of proteins, recruits
them, and coordinates their activities at sites of DNA synthesis. It functions in DNA
replication, translesion synthesis, base excision repair, nucleotide excision repair,
mismatch repair, homologous recombination, chromatin remodeling, sister chromatid
cohesion, and cell cycle regulation [155, 199, 200, 257, 258, 289]. Genetic studies,
especially in yeast, have identified a number of mutations in PCNA that disrupt one or
more of these processes. For example, simple amino acid substitutions in PCNA have
been identified that interfere with translesion synthesis [238, 240], error-free
126
postreplication repair [297], MMR [242, 288, 290, 291], and chromatin remodeling [298,
299]. In this study, we focused on two important substitutions that block MMR.
A systematic study of amino acid substitutions in PCNA that interfere with MMR
has shown that many of the changes impact cell growth at non-permissive temperatures
and are sensitive to DNA damaging agents such as methyl methanesulfonate (MMS),
ultraviolet radiation (UV), and hydroxyurea (HU) [242]. Most of these substitutions in
PCNA cause an increase in spontaneous mutations in assays measuring the frequency of
reversion mutations of a one-nucleotide insertion in the hom3-10 allele, the frequency of
reversion mutations of a four-nucleotide insertion in the lys2-bgl allele, and the frequency
of forward mutations in the CAN1 gene. Two of these amino acid substitutions in PCNA,
the C22Y substitution (encoded by the pol30-201 allele) and the C81R substitution
(encoded by the pol30-204 allele), specifically impact MMR and do not cause notable
defects in other DNA replication and repair processes [242]. Neither substitution causes
temperature-dependent growth defects or increased sensitivity to DNA damaging agents.
Both substitutions lead to an increase in the frequency of reversion mutations in the
hom3-10 and lys2-bgl alleles and in the frequency of forward mutations in the CAN1
gene. Genetic analysis of these mutant forms of PCNA in combination with the msh2,
msh3, and msh6 mutations, which disrupt MutS or MutS, imply that the C22Y
mutant PCNA protein causes a strong defect in MutS-dependent MMR and the C81R
mutant PCNA protein causes a moderate defect in both MutS-dependent and MutS-
dependent MMR [242].
In the present study, we examined the structural changes in PCNA induced by
these two amino acid substitutions to understand the basis for their specific defects in
MMR. Interestingly, these two substitutions caused two distinct structural alterations in
PCNA. The C22Y mutant PCNA protein has shifts in the α-helices that line the central
hole of the PCNA ring that encircles DNA. This mutant protein forms stable trimers and
stimulates DNA synthesis by pol . The C81R mutant PCNA protein has a localized
127
distortion in the β-sheet at the PCNA subunit interface. This mutant protein does not form
as stable trimers as do the wild-type and the C22Y mutant PCNA proteins. Consequently,
this mutant protein exhibits only a moderate stimulation of DNA synthesis by pol ,
which is still higher than in the absence of PCNA.
Since both PCNA mutant proteins are known to block MutS-dependent MMR,
we examined their ability to interact with MutS. The C22Y mutant PCNA protein
interacts with MutS with the same affinity as does the wild-type PCNA. The C81R
mutant PCNA protein also interacts with MutS, but does so with lower affinity
compared to the wild-type and C22Y mutant proteins. Again, this apparent weaker
binding is likely due to the instability of the C81R mutant PCNA protein trimers, and I
believe that the few C81R mutant PCNA protein trimers that do form under these
experimental conditions still bind MutS. It should be noted that the conformational
changes induced by these amino acid substitutions do not perturb the hydrophobic
binding pocket on PCNA which binds the canonical PCNA-interacting protein (PIP)
motif in the N-terminal region of the Msh6 subunit of MutS. Overall, these findings
suggest that the disruption of MMR by these PCNA mutant proteins does not arise from
substantial defects in the interaction with MutS, especially in the case of the C22Y
mutant PCNA protein.
We therefore examined the ability of the PCNA mutant proteins to form ternary
complexes with MutS and mismatch-containing DNA. Surprisingly, we found that
despite relatively subtle changes in the structures of the C22Y and C81R mutant PCNA
proteins, they both formed aberrant complexes with MutS and DNA. In the presence of
the mutant PCNA proteins and a mismatch, the MutS-containing complexes were larger
than in the presence of the wild-type PCNA protein. A possible explanation for this is
that improper mismatch recognition leads to multiple MutS proteins on the DNA
substrate. Understanding why these higher-ordered complexes form awaits further
structural studies. However, we conclude that these complexes are aberrant and that the
128
formation of proper, productive complexes between MutS and mismatch-containing
DNA depends on both the structural integrity of the -helices lining the central hole of
the PCNA ring and the structural integrity of the -sheet at the PCNA subunit interface.
Previously, our lab determined the structures of two mutant forms of PCNA that
block translesion synthesis (TLS) [219, 260]. These two mutant proteins have amino acid
substitutions (E113G and G178S) that are in the -strands constituting the subunit
interface. Both substitutions create distortions at the subunit interface and decrease the
stability of the mutant PCNA trimers [Dieckman and Washington, DNA Repair, in press].
The E113G substitution is believed to block TLS with little other defects, and there is no
evidence that it interferes with MMR. This is interesting, because Glu-113 is directly
adjacent to Tyr-114, which is affected by the C81R substitution. The G178S substitution,
by contrast, does appear to have a defect in both TLS and in MMR. This substitution
leads to an increase in the frequencies of both spontaneous forward and reversion
mutations [242]. Further structure-function analysis is necessary to understand how
disruptions at the PCNA subunit interface can affect MMR in some cases (such as the
C81R mutant PCNA protein), TLS in some cases (such as the E113G mutant PCNA
protein), and both MMR and TLS in other cases (such as the G178S mutant PCNA
protein).
129
Table 3.1. Data collection and refinement statistics.
C22Y mutant protein C81R mutant protein
(A) Data collection statistics
Resolution (Å)a 20.9-2.7 (2.8-2.7)
21.5-3.0 (3.1-3.0)
Wavelength (Å) 1.54 1.00
Space Group P212121 P213
Cell (Å) a = 85.9, b = 90.6, c = 140.6 a=b=c=121.8
Completeness (%) a 100 (100) 100 (100)
Redundancy a 7.1 (7.2) 19.9 (20.2)
<I/(I)> a 7.0 (2.0) 9.1 (2.0)
Rmerge (%) a 13.3 (65.9) 14.2 (77.3)
(B) Refinement statistics
Resolution range (Å) 19.9-2.7 20.5-3.0
R (%) 21.5 23.9
Rfree (%) 27.4 28.5
rms bonds (Å) 0.013 0.011
rms angles () 1.15 1.50
Number of protein atoms 6026 (761 residues) 1975 (254 residues)
Number of water molecules 0 0
Ramachandran analysis (%)
Most favored 91. 5 92.5
Allowed 7.8 6.8
PDB ID code 4L6P 4L60
a Values in parentheses are for the highest resolution shell
130
Figure 3.1. Structure of the C22Y mutant PCNA protein. (A) Front view of the C22Y
mutant PCNA protein trimer with one of the subunits shown in red. (B) Side view of one
subunit of the C22Y mutant PCNA protein with the -helices A2, B2, and A1 shown in
red. The wild-type PCNA structure is overlaid with the same -helices shown in purple.
(C) Close up of the region near the C22Y substitution. The mutant PCNA protein is
shown in red and the wild-type PCNA protein is shown in purple.
131
Figure 3.2. Structure of the C81R mutant PCNA protein. (A) Front view of the C81R
mutant PCNA protein trimer with one of the subunits shown in orange. (B) Side view of
one subunit of the C81R mutant PCNA protein with the -helix B1 and the -strands H1
and I1 shown in orange. The wild-type PCNA structure is overlaid with the same -helix
and -strands shown in purple. (C) Close up of the region near the C81R substitution.
The mutant PCNA protein is shown in orange and the wild-type PCNA protein is shown
in purple.
132
Figure 3.3. Analysis of the PCNA proteins by native gel electrophoresis. Solutions
containing the wild-type or mutant PCNA proteins (0.1 to 1.0 mg/ml) were run on a non-
denaturing polyacrylamide gradient gel (4 to 20%) and Coomassie stained. The positions
of the PCNA monomer and trimer are indicated.
133
Figure 3.4. Analysis of the PCNA proteins by size exclusion chromatography. (A)
The elution profile of a size exclusion chromatography column in which a solution of the
wild-type PCNA protein (10 mg/ml) was run. (B) The elution profile of a size exclusion
chromatography column in which a solution of the C22Y mutant PCNA protein (10
mg/ml) was run. (C) The elution profile of a size exclusion chromatography column in
which a solution of the C81R mutant PCNA protein (10 mg/ml) was run.
134
Figure 3.5. Analysis of the C81R mutant PCNA protein by size exclusion
chromatography. The elution profile of a size exclusion chromatography column in
which a solution of the C81R mutant PCNA protein (0.01 mg/ml) was run.
135
Figure 3.6. DNA synthesis by pol in the presence of the PCNA proteins. (A) An
autoradiogram of the products of pol -catalyzed DNA synthesis on a non-damaged DNA
substrate in the presence of no PCNA, the wild-type PCNA protein, the C22Y mutant
PCNA protein, and the C81R mutant PCNA protein. The position of the fully extended,
runoff product is indicated with an arrow. (B) An autoradiogram of the products of pol -
catalyzed DNA synthesis on an abasic site-containing DNA substrate in the presence of
no PCNA, the wild-type PCNA protein, the C22Y mutant PCNA protein, and the C81R
mutant PCNA protein. The position of the abasic site is indicated by an X, and the
position of the fully extended, runoff product is indicated with an arrow.
136
Table 3.2. Relative DNA synthesis by pol δ in the presence of PCNA mutant
proteinsa.
Synthesis of full
length product on
non-damaged DNA
Synthesis
opposite abasic
site
Synthesis of full
length product on
damaged DNA
No PCNA
1 1 1
WT PCNA
4.9 4.8 4.4
C22Y PCNA 2.8 6.2 2.9
C81R PCNA
1.5 1.7 1.3
a These figures were obtained by dividing the nucleotide incorporation of pol δ in the
presence of each PCNA protein by the nucleotide incorporation of pol δ in the
absence of PCNA.
137
Figure 3.7. Interactions of the PCNA proteins with MutS. The results of an ELISA
assay showing the interactions of the wild-type PCNA (), the C22Y mutant PCNA
protein (), and the C81R mutant PCNA protein () with MutS. Control experiments
using BSA instead of MutS have been subtracted from each of the values.
138
Figure 3.8. Sedimentation analysis of the interactions of the PCNA proteins with
MutS and mistmatched DNA. (A) Fractions of a glycerol gradient (15-30%)
containing MutS and heteroduplex DNA containing a G:T base pair in the absence of
PCNA were analyzed by denaturing polyacrylamide gradient gel electrophoresis (4-
15%). The fractions ranged from 1 (the bottom of the gradient) to 14 (the top of the
gradient). The proteins were visualized by silver staining, and the DNA substrate was
visualized by Cy3 fluorescence. (B) Fractions of a glycerol gradient containing MutS
and heteroduplex DNA containing a G:T base pair in the presence of the wild-type PCNA
protein were analyzed by denaturing polyacrylamide gradient gel electrophoresis. (C)
Fractions of a glycerol gradient containing MutS and heteroduplex DNA containing a
G:T base pair in the presence of the C22Y mutant PCNA protein were analyzed by
denaturing polyacrylamide gradient gel electrophoresis. (D) Fractions of a glycerol
gradient containing MutS and heteroduplex DNA containing a G:T base pair in the
presence of the C81R mutant PCNA protein were analyzed by denaturing polyacrylamide
gradient gel electrophoresis. (Data from Elizabeth Boehm.)
139
140
Figure 3.9. Sedimentation analysis of the interactions of the PCNA proteins with
MutS and homoduplex DNA. (A) Fractions of a glycerol gradient (15-30%)
containing MutS and homoduplex DNA in the absence of PCNA were analyzed by
denaturing polyacrylamide gradient gel electrophoresis (4-15%). The fractions ranged
from 1 (the bottom of the gradient) to 14 (the top of the gradient). The proteins were
visualized by silver staining, and the DNA substrate was visualized by Cy3 fluorescence.
(B) Fractions of a glycerol gradient containing MutS and homoduplex DNA in the
presence of the wild-type PCNA protein were analyzed by denaturing polyacrylamide
gradient gel electrophoresis. (C) Fractions of a glycerol gradient containing MutS and
homoduplex DNA in the presence of the C22Y mutant PCNA protein were analyzed by
denaturing polyacrylamide gradient gel electrophoresis. (D) Fractions of a glycerol
gradient containing MutS and homoduplex DNA in the presence of the C81R mutant
PCNA protein were analyzed by denaturing polyacrylamide gradient gel electrophoresis.
(Data from Elizabeth Boehm).
141
142
CHAPTER 4
IDENTIFICATION AND CHARACTERIZATION OF RANDOM
MUTATIONS OF THE PCNA SUBUNIT INTERFACE
Abstract
Proliferating cell nuclear antigen (PCNA) function is essential for proper DNA
replication, repair, and recombination. During translesion synthesis (TLS), PCNA
recruits and stabilizes non-classical polymerases at the replication fork to bypass DNA
lesions. During mismatch repair (MMR), PCNA recruits and coordinates proteins
involved in the initiation, excision, and resynthesis steps. Four amino acid substitutions
have been identified in PCNA that disrupt TLS and MMR: the E113G and G178S
substitutions cause defects in TLS while the C22Y and C81R substitutions cause defects
in MMR. The structures of these mutant PCNA proteins revealed that three of the four
substitutions caused disruptions near the subunit interface of PCNA. Here, we generated
random mutations of the PCNA subunit interface and performed in vivo genetics assays
and in vitro biochemical assays to examine their effects on TLS and MMR. We
determined that the subunit interface of PCNA is very dynamic and that small changes at
this interface can cause drastically different effects on TLS and MMR. Moreover, we
suggest that the integrity of the subunit interface as well as the nearby β-strands in
domain A are crucial for proper PCNA function in vivo and in vitro.
Introduction
Classical polymerases are those that function in normal DNA replication and
repair on non-damaged DNA templates. In general, these enzymes incorporate
nucleotides with high fidelity and processivity. DNA polymerase δ (pol δ) is the classical
143
polymerase that is responsible for lagging strand synthesis during DNA replication [244-
246], and it also plays a major role in base excision repair, nucleotide excision repair, and
double strand break repair [247]. In yeast, pol δ is a heterotrimer, composed of a
catalytic subunit (Pol3) and two accessory subunits (Pol31 and Pol32). Mutations in pol
delta accelerates tumorigenesis in mice [300-302] and have also been identified in several
human cancer cell lines [303, 304].
Even though classical polymerases are extremely efficient, normal DNA
replication by these polymerases is blocked at sites of DNA damage due to their highly
stringent active sites that cannot accommodate the structure of most DNA lesions. As a
result, cells employ several non-classical polymerases to replicate through the damage.
These polymerases have much larger active sites than classical polymerases that allow
the inclusion of bulky and distorted bases lesions [60, 61, 202, 203, 251-254]. One such
non-classical polymerase is DNA polymerase η (pol η), a monomeric enzyme that is
responsible for replication through template thymine dimers and 8-oxo-guanine lesions
[77, 88, 89]. Human cells carrying mutations in their pol η genes were found to have
xeroderma pigmentosum variant (XP-V), which is an inherited disorder associated with
increased incidence of sunlight-induced skin cancers [105].
The process of replicating through DNA damage by polymerases is called
translesion synthesis (TLS). Most TLS is performed by non-classical polymerases,
however classical polymerases are also able to carry out TLS in a few contexts.
Typically, TLS occurs when a classical polymerase stalls at a replication fork containing
a DNA lesion and is replaced by a non-classical polymerase to facilitate damage bypass.
The protein accessory factor that recruits and stabilizes polymerases to the replication
fork is proliferating cell nuclear antigen (PCNA), a ring-shaped homotrimeric protein that
encircles the DNA template and regulates polymerase activity. PCNA increases the
processivity and catalytic efficiency of DNA synthesis by classical pol δ, [75, 165, 305]
as well as the catalytic efficiency of non-classical pol η [113, 305]. During TLS, PCNA
144
is monoubiquitylated on lysine 164 by the Rad6-Rad18 complex. This post-translational
modification event is thought to initiate the switch between these classical and non-
classical polymerases at the primer-template [183], as non-classical polymerases
preferentially interact with the ubiquitin-modified form of PCNA [185].
Inaccurate DNA replication can result in base-base mismatches and small loops
arising from insertions or deletions (insertion/deletion loops or IDLs). In order to reduce
mutation rates, these mismatches and loops are recognized and repaired by mismatch
repair (MMR) proteins. In humans, defects in MMR are known for their correlation with
sporadic and hereditary human cancers, including hereditary non-polyposis colorectal
cancer [24-26, 284-286].
The mechanisms of MMR in eukaryotes are more complicated than in
prokaryotes and are not well defined. Six MutS homologs exist in yeast (MSH1 to
MSH6) and five exist in mammals (MSH2 to MSH6). In all eukaryotes, MSH2 and
MSH6 form a heterodimer called MutSα that recognizes base-base mismatches and small
IDLs [33, 34, 276-278]. MSH2 and MSH3 form a deterodimer called MutSβ that
recognizes longer loops [32, 279, 280]. Several MutL homologs also exist, including
MLH1, MLH2, MLH3, and PMS1, which function as heterodimers in complex with
MutSα and MutSβ to enhance binding of the mismatch [37, 116, 281-283]. Other key
proteins involved in the subsequent excision and resynthesis steps include exonuclease I
(EXOI) [39-42, 287], DNA polymerase delta (pol ) [43], replication protein A (RPA)
[44, 45], replication factor C (RFC) [47], and proliferating cell nuclear antigen (PCNA)
[51, 54, 288].
In addition to functioning as a processivity factor for DNA polymerases, PCNA
plays important roles in DNA replication, repair, recombination, translesion synthesis,
chromatin remodeling, sister chromatid cohesion, and cell cycle regulation [155, 199,
200, 257, 258, 289]. PCNA’s role in MMR is not well understood, but it is thought to be
important for the initiation and mismatch recognition stage as well as the excision and
145
resynthesis stages. During the initial mismatch recognition stage of MMR, PCNA
interacts with and recruits both MutSα and MutSβ to sites of mismatched bases [53-57].
It has also been suggested that PCNA is involved in strand discrimination of the daughter
strand of DNA during MMR in eukaryotes since hemi-methylation does not exist in these
systems [50-52].
Several mutant PCNA alleles have been identified in genetic screens that lead to
defects in various nuclear processes. Two of these mutations are defective in TLS by
non-classical polymerases but support normal cell growth [238, 240, 267]. The pol30-
178 allele encodes a mutation in PCNA where the glycine 178 residue is substituted with
a serine (G178S), while the pol30-113 allele encodes a PCNA protein in which glutamic
acid 113 is substituted with a glycine residue (E113G). In addition, genetic studies have
shown that two other mutant alleles, pol30-201 and pol30-204, specifically disrupt the
MMR pathway with little if any effect on other DNA metabolic processes [242]. The
pol30-201 allele, which encodes the C22Y mutant PCNA protein, causes a strong defect
in MutS-dependent MMR, and the pol30-204 allele, which encodes the C81R mutant
PCNA protein, causes a partial defect in both MutS-dependent and MutS-dependent
MMR [242].
The structures of each of these four mutant PCNA proteins that disrupt TLS and
MMR have been determined in our lab (Chapter 3 and [219, 241]). Interestingly, three of
the four amino acid substitutions (E113G, G178S, and C81R) are located near the PCNA
subunit-subunit interface. The E113G and G178S substitutions in PCNA disrupt the
hydrogen bonding between the two adjacent subunits, while the C81R mutation disrupts
the β-sheet in the N-terminal domain of PCNA at the subunit interface. Moreover, all
three of these mutant proteins have reduced trimer stability (Chapter 3 and [305]) and
cause defects in pol δ activity (Chapter 3 and [305]). Furthurmore, the E113G and
G178S mutant PCNA proteins have been shown to disrupt pol η activity [305].
146
Altogether, the studies presented in this thesis suggest that the integrity of the subunit
interface of PCNA is crucial for proper function in both TLS and MMR.
Because of these striking results, our lab chose to focus on the subunit interface
and generated a set of random mutations in PCNA in this region of the protein. Christine
Kondratick and Viana Nguyen determined that these amino acid substitutions cause
differing effects on cell growth, TLS, and MMR in vivo. I determined that these amino
acid substitutions cause differing effects in vitro on PCNA trimer formation and DNA
synthesis by pol δ and pol η. My results show that there is a direct correlation between
PCNA trimer stability and TLS function in vitro. Together with the in vivo data, we
show that the interface of PCNA is very dynamic and that small changes in the interface
can cause drastically different effects on TLS and MMR. Moreover, we suggest that the
integrity of the subunit interface as well as the nearby β-strands in domain A are crucial
for proper PCNA function in vivo and in vitro.
Materials and Methods
Protein expression and purification.
The wild-type and the S177G, G178S, S179T, V180A, and I181R mutant PCNA
proteins from S. cerevisiae were over-expressed as N-terminally His6-tagged proteins in
E. coli and were purified as described previously [219]. Replication factor C (RFC) from
S. cerevisiae was over-expressed in E. coli and purified as previously described [261].
DNA polymerase δ (pol ) and DNA polymerase η (pol η) from S. cerevisiae were over-
expressed in S. cerevisiae and purified as previously described [75, 256].
DNA and nucleotide substrates.
The template strand used to measure pol and pol η activity was a 68-mer
oligodeoxynucleotide with the sequence 5'- GAC GGC ATT GGA TCG ACC TCX AGT
147
TGG TTG GAC GGG TGC GAG GCT GGC TAC CTG CGA TGA GGA CTA GC with
biotins on both ends. The X represents the position of a non-damaged G or an abasic site.
The primer strand was a 31-mer oligodeoxynucleotide with the sequence 5'-TCG CAG
GTA GCC AGC CTC GCA CCC GTC CAA C. The primer strand was 5'-32
P-end-
labeled with T4 polynucleotide kinase and (γ-32
P)ATP and annealed to the template
strand at 1 µM in 25 mM TrisCl, pH 7.5, and 100 mM NaCl at 90oC for 2 min and slowly
cooled to 30oC. A mixture of all four dNTPs (10 mM each) was purchased from New
England Biolabs.
PCNA trimer stability assays.
For non-denaturing polyacrylamide gel electrophoresis (PAGE), the wild-type
and S177G, G178S, S179T, V180A, and I181R mutant PCNA proteins (0.1 to 1.0
mg/ml) were incubated in 25 mM TrisCl, pH 7.4, 150 mM NaCl, 0.01% bromophenol
blue, and 10% glycerol for 5 min and then run on a TrisCl pre-cast 4-20% gradient non-
denaturing polyacrylamide gel (Bio-Rad) at 4oC at a current of 25 mA using 25 mM Tris,
pH 8.3, and 0.2 M glycine running buffer. Protein bands were visualized using
Coomassie blue staining. For size exclusion chromatography, wild-type and mutant
PCNA proteins were diluted to 10 mg/ml in 25 mM Tris, pH 7.4, 150 mM NaCl, 5%
glycerol and run on a 120 ml HiLoad 16/60 Superdex 200 PG column (GE Healthcare).
Polymerase activity assays.
Running start assays were performed as described previously [161, 305].
Reactions were carried out in the absence of PCNA and in the presence of 90 nM wild-
type or mutant PCNA proteins (trimer concentration), and contained 20 nM pol or pol
η, 25 nM DNA, and 100 µM of each of the four dNTPs. Reactions were stopped after 1
and 2 minutes in the case of pol δ and after 30 min in the case of pol η. The extension
products were analyzed on a 15% polyacrylamide sequencing gel containing 8 M urea.
148
Results
Genetic analyses of random PCNA interface
mutant proteins.
Previous work in our lab has suggested that amino acid substitutions at the
subunit-subunit interface of PCNA that cause defects in TLS and MMR cause structural
perterbations at the interface (Chapter 3 and [219, 241, 305]). To determine the
importance of the PCNA interface in these DNA replication and repair processes, Todd
Washington generated several random mutations at the subunit interface using site-
directed mutagenesis with primers containing randomized nucleotides at each of the
individual codons. The mutations used for the studies presented herein are shown in
Figure 4.1. To examine how these mutant proteins affect the function of PCNA in TLS
and MMR in vivo, Christine and Viana carried out cell survival and mutagenesis studies
with yeast cells expressing PCNA containing each of these individual interface
mutations. PCNA containing the K164R mutation was used as a negative control in these
studies because it cannot be ubiquitylated at residue 164 and is therefore defective in
promoting TLS. As shown in Table 4.1, column I, several of these interface mutants
displayed normal cell growth, while four (the S177G, G178M, G178L, and V180A
substitutions) grew slower than wild-type PCNA. Interestingly, all PCNA mutants
generated were more sensitive to UV exposure than wild-type PCNA (Table 4.1, column
II).
They next assayed the PCNA interface mutants for induced mutations after UV
exposure (Table 4.1, column III). In this assay, cells expressing the mutant PCNA
proteins were plated on synthetic complete medium lacking arginine and containing
canavanine to determine the frequency of CAN1S to can1
r forward mutations. Only three
substitutions in PCNA (S177G, G178M, and V180A) had normal levels of induced
mutagenesis, suggesting that all other substitutions are defective in TLS in vivo. Lastly,
149
to examine the effect of these PCNA interface mutants on MMR, they performed a
spontaneous mutagenesis assay in which mutagenesis in cells was measured in the
absence of any DNA-damaging agents (Table 4.1, column IV). Results showed that the
E113G, Y114A, G178S, G178M, and V180A mutant PCNA proteins have increased
rates of spontaneous mutations and are therefore likely defective in MMR. Overall, the
genetic studies suggest that these PCNA interface mutant proteins have differing effects
on TLS and MMR.
From these in vivo results, we identified five PCNA interface mutants that
encompass an array of phenotypes and span the β-strand βD2 to isolate and use as a
representative sample for in vitro assays. All subsequent biochemical studies were
performed with these five PCNA interface mutant proteins – S177G, G178S, S179T,
V180A, and I181R.
Trimer stability of the PCNA interface mutant proteins.
Since I have observed reduced trimer stability with several other mutant PCNA
proteins that disrupt the subunit interface, I examined the ability of the five PCNA
interface mutant proteins to form trimers. Using non-denaturing polyacrylamide gel
electrophoresis (PAGE), I showed that these mutant proteins have differing trimer
stabilities (Figure 4.2). Wild-type PCNA and the S179T and I181R mutant PCNA
proteins formed stable trimers under all concentrations tested (0.1 to 1.0 mg/ml). By
contrast, the G178S mutant PCNA protein was a monomer under all concentrations
tested. Interestingly, the S177G and V180A mutant proteins were less stable than wild-
type PCNA, but formed trimers more readily than the G178S mutant PCNA protein. At
all concentrations, the S177G mutant protein bands were mostly trimeric and streaking.
This suggests that, although this protein is generally a trimer, it is slightly less stable than
wild-type PCNA. In comparison, the oligomeric form of the V180A mutant PCNA
protein ranged from monomeric to trimeric as protein concentration increased. This
150
shows that the V180A mutant protein is capable of forming trimers at higher
concentrations, but is even less stable than the S177G mutant PCNA protein. From these
results, I suggest that the trimer stabilities of these PCNA interface mutant proteins span
a wide range as such: G178S < V180A < S177G < S179T, I181R, and wild-type PCNA,
where the G178S mutant PCNA protein is the least stable, and wild-type PCNA and the
S179T and I181R mutant PCNA proteins are the most stable.
In order to further assess PCNA trimer stability, Christine and Viana analyzed the
proteins by size exclusion chromatography. When loaded onto the size exclusion column
at a high concentration (10 mg/ml), the wild-type PCNA protein and the S179T and
I181R mutant proteins eluted in a narrow peak as expected for the 90-kDa trimer (Figure
4.3). When the S177G and V180A mutant proteins were loaded at a high concentration
(10 mg/ml), they eluted in a broad peak, likely corresponding to a mixture of trimeric and
non-trimeric species (Figure 4.3). The G178S mutant PCNA protein has yet to be
evaluated using size exclusion chromatography, but its elution profile is expected to be
similar to or even more broad than those of the S177G and V180A mutant PCNA
proteins. Taken together, both the native PAGE and size exclusion chromatography
results show that the S177G and V180A mutant PCNA proteins (and expectedly the
G178S mutant PCNA protein) are less stable than the S179T and I181R mutant proteins,
which form stable trimers like the wild-type PCNA protein.
Impact of the PCNA interface mutant proteins
on the activity of pol η.
In order to test the ability of the mutant PCNA proteins to stimulate the catalytic
activity of the non-classical polymerase pol η, I used running start experiments to
measure DNA synthesis by pol η on a non-damaged template in the presence of each
PCNA interface mutant protein (Figure 4.4A). Under these conditions, the presence of
wild-type PCNA stimulated DNA synthesis by pol η more than in its absence. Similar
151
stimulation was also observed in the presence of both the S179T and I181R mutant
PCNA proteins. In contrast, the G178S and V180A mutant PCNA proteins did not
stimulate the activity of pol η beyond that of the polymerase itself, whereas the S177G
mutant PCNA protein may show an intermediate stimulation of pol η that is above that of
pol η alone and less than that seen with wild-type PCNA. Therefore, these five PCNA
interface mutant proteins have varying degrees of stimulation of pol η on non-damaged
DNA.
To determine the ability of these mutant PCNA proteins to stimulate TLS by pol
η, I measured DNA synthesis by pol η on a DNA substrate containing an abasic site
located at the sixth position in the template following the primer terminus (Figure 4.4B).
In the absence of PCNA, pol η had little activity opposite the templating abasic site.
Similar to synthesis on non-damaged DNA, pol η’s activity to incorporate nucleotides
opposite the abasic site was stimulated in the presence of wild-type PCNA and the S179T
and V180A mutant PCNA proteins, with 1.3 fold, 1.7 fold, and 1.6 fold stimulation above
that seen with no PCNA, respectively. In the presence of the S177G, G178S, and V180A
mutant PCNA proteins, however, little to no stimulation was observed (0.9 to 1.1 fold for
each compared to pol η alone). Thus, the PCNA interface mutant proteins have differing
effects on the process of TLS by pol η. Altogether, these studies with pol η suggest that
the five mutations at the subunit interface of PCNA exhibit a range of abilities to
stimulate the activity of pol η, where the S177G, G178S, and V180A mutant proteins
show little to no stimulation and the S179T and I181R mutant proteins behave like wild-
type PCNA.
Impact of the PCNA interface mutant proteins
on the activity of pol δ.
Classical pol is responsible for lagging strand synthesis during normal DNA
replication and is also involved in base excision repair, nucleotide excision repair,
152
mismatch repair, and double strand break repair [244-247]. To examine the effects of the
five PCNA interface mutant proteins on the activity of pol δ, I carried out running start
experiments with pol δ on non-damaged DNA in the absence and presence of wild-type
PCNA and each of the mutant proteins (Figure 4.5). The PCNA interface mutant proteins
stimulated DNA synthesis by pol δ to differing extents. In the presence of either the
wild-type PCNA protein, the S179T mutant protein, or the I181R mutant protein,
approximately 4.7, 4.9, or 5.6 fold more full-length runoff products formed, respectively,
compared to in the absence of any PCNA. As observed with DNA synthesis by pol η,
neither of the G178S and V180A mutant PCNA proteins were able to stimulate pol δ
activity on non-damaged DNA past that of the polymerase alone. The amount of full
length product formed by pol δ in the presence of either of these mutant proteins was
~1.1 fold of the amount formed in their absence. Interestingly, the S177G did stimulate
pol δ’s activity (2 fold over the polymerase alone), but to a lesser extent than in the
presence of wild-type PCNA.
In addition to non-damaged DNA, pol δ is capable of bypassing abasic sites. To
determine if the PCNA interface mutant proteins affect TLS by pol δ, I examined DNA
synthesis by the classical polymerase on a DNA template containing an abasic site
(Figure 4.6). Approximately 2.4, 2.1, or 2.7 fold more extension products were observed
opposite the abasic site in the presence of the wild-type PCNA protein, the S179T mutant
PCNA protein, and the I181R mutant PCNA protein than in their absence, respectively.
Furthermore, ~4.8, 4.0, and 4.8 fold more full-length runoff products were formed in the
presence of the three PCNA proteins, respectively. In contrast, both of the G178S and
V180A mutant PCNA proteins failed to stimulate pol δ’s activity opposite the abasic site,
with ~0.9 fold pol δ activity seen in their presence compared to pol δ alone. The amount
of full-length runoff products formed in the presence of these two mutant PCNA proteins
was similar to this, as only ~0.8 and 0.9 fold extension was observed for the G178S and
V180A mutant PCNA proteins, respectively. In contrast to these results, in the presence
153
of the S177G mutant PCNA protein, there were ~1.3 fold more extension products
resulting from incorporation opposite the abasic site and ~1.4 fold more full-length runoff
products than in the absence of PCNA. Thus, the PCNA interface mutant proteins have
differing effects on DNA synthesis on an abasic-containing DNA template by pol δ.
Altogether, the running start experiments with pol δ suggest that five mutations at the
subunit interface of PCNA exhibit a range of abilities to stimulate the activity of pol δ,
where the G178S and V180A mutant proteins show no stimulation, the S179T and I181R
mutant proteins behave like wild-type PCNA, and the S177G mutant protein shows an
intermediate stimulation.
Structures of the PCNA interface mutant proteins.
Experiments to determine the X-ray crystal structures of each of the five PCNA
interface mutants have been underway by Christine Kondratick, Viana Nguyen, and Kyle
Powers. So far, the structures of the S177G, G178S, S179T, and V180A mutant PCNA
proteins have been solved and fully refined. The major changes observed between these
mutant proteins and the wild-type PCNA protein are generally located at the subunit
interface. As described in Chapter 2, the G178S substitution causes significant shearing
of the interface, in which only three of the original seven hydrogen bonds are maintained
between the β-strands that constitute the PCNA interface (Table 4.2). This same shearing
of the interface is also seen in the case of the S177G, S179T, and V180A mutant PCNA
proteins, however to lesser extents (Table 4.2). In the case of the S177G mutant protein,
three hydrogen bonds out of the original seven are broken. Similar to that observed with
the protein stability assays, the S179T mutation causes less distortion of the subunit
interface than either the G178S and S177G substitutions, i.e. only two of the original
seven hydrogen bonds are broken in this mutant PCNA protein. Therefore, the S177G
and G178S amino acid substitutions perturb the subunit interface of PCNA to varying
154
degrees and are likely the source of the differences in trimer stabilities and in vitro
functions of each mutant PCNA protein.
Contrary to what was expected, the V180A substitution only caused two hydrogen
bonds to be disrupted at the interface. However, upon further inspection, I noticed that
there were large perturbations of the V180A mutant PCNA protein outside of the subunit
interface. While the V180A substitution is positioned in domain B of PCNA, the largest
changes in structure are observed in domain A (Figure 4.7A). Figure 4.7B shows that the
β-strands in domain A (βA1, βG1, βH1, and βI1) are shifted significantly from the wild-
type PCNA structure. For instance, backbone residue shifts of 0.5 Å, 0.9 Å, and 0.8 Å
are observed between the mutant and wild-type PCNA for strands βI1, βH1, and βG1,
respectively. In the case of βA1, these shifts are as large as 3.0 Å. Together, the
propagated disruptions in these β-strands along with the small changes at the subunit
interface likely account for the reduced trimer stability seen with the V180A mutant
PCNA protein. Determination of the structures of the I181R mutant PCNA protein is still
underway, but because it behaves like wild-type PCNA in every assay performed, I
predict that the I181R mutant PCNA protein will have the least significant changes of all
of these mutant PCNA proteins.
Discussion
Studies with the five PCNA interface mutant proteins presented here suggest that
the subunit interface of PCNA is very dynamic and that small changes at the interface can
cause drastically different effects on TLS and MMR. Table 4.3 summarizes all of the
genetic and biochemical results characterizing the five mutant PCNA proteins. My
biochemical studies with the five mutant PCNA proteins suggest that there is a strong
correlation between PCNA trimer stability and TLS in vitro. Based on the structures of
these five mutant proteins, this trimer instability usually arises due to instability at the
155
interface, i.e. number of hydrogen bonds disrupted between the two β-strands that
constitute the interface. For instance, the S179T mutant PCNA protein behaves like
wild-type PCNA in stimulating both pol δ and pol η activity opposite damaged DNA and
has at least five of the original seven hydrogen bonds observed in wild-type PCNA. In
contrast, the G178S mutant PCNA protein is completely defective in stimulating TLS by
pol δ or pol η and only preserves three hydrogen bonds, resulting in a severely sheared
subunit interface. The S177G mutant PCNA protein, however, shows an intermediate
phenotype for stimulating DNA synthesis by pol δ and pol η on damaged DNA and
maintains four hydrogen bonds at the interface. The structure of the I181R mutant PCNA
protein has not yet been determined, but I expect that it will contain at least five hydrogen
bonds at the subunit interface due to its high trimer stability and its ability to stimulate
both pol δ and pol η on damaged and non-damaged DNA templates. These results are
consistent with the studies described in Chapters 2 and 3 in which the reduced trimer
stability of the E113G and C81R mutant PCNA proteins at the PCNA subunit interface
resulted in a decreased ability to stimulate DNA synthesis by polymerases.
The integrity of the PCNA subunit interface is clearly important for PCNA
function in vitro. However, the structure of the V180A mutant protein shows that amino
acid substitutions at the interface can also cause perturbations outside of the interface
itself. This substitution causes structural conformational shifts of up to 3.0 Å, which are
propagated up to the N-terminus of the mutant PCNA protein. These residues are
actually positioned in trans to the mutation site and together make up four β-strands in
domain A. Therefore, the defect in stimulating TLS by pol δ and pol η observed in the
presence of the V180A mutant PCNA protein is likely due to these large structural
changes. Together, my results suggest that, in addition to the PCNA subunit interface,
the first ten residues of PCNA are crucial for proper PCNA function in vitro.
The correlation between the crystal structures of the five PCNA interface mutant
proteins and their functions in vitro is quite evident. This correlation, however, is less
156
well-defined with their functions in vivo, as many more factors are likely involved in TLS
and MMR in the cell. Probably the clearest case is of the G178S mutant PCNA protein.
The subunit interface is significantly altered compared to the interface of the wild-type
protein, which is likely the cause of the defects seen in both the TLS and MMR in the
genetic assays. Since the structure of the I181R mutant PCNA protein has not been
determined, it is unclear how this mutation causes reduced UV-induced mutagenesis in
vivo despite the high trimer stability and ability to stimulate pol δ and pol η observed with
this mutant protein in vitro. It is plausible, though, that this structure is similar to the
structure of the S179T mutant PCNA protein, as their phenotypes are almost identical.
Based on conclusions drawn from Chapter 2, it is likely that the small disruption at the
interface seen in the S179T mutant PCNA protein causes a defect in TLS in vivo, but that
the disruption is not large enough to observe significant defects in vitro. The correlation
between the structure and in vivo phenotype of the S177G mutant PCNA protein is
unclear, as the structure of this mutant protein suggests that it would cause defect in TLS
or MMR. More in vivo studies will be necessary with this mutant protein to elucidate the
relationship between its structure and function.
The case of the V180A mutation PCNA protein is unique in that the largest
structural perturbations caused by this mutation is actually within domain A of the
protein instead of at the subunit interface. This causes defects in both growth rate and
MMR but not TLS in vivo. One possible explanation for the elevated mutation rate and
reduced growth rate is that the structural changes produced by the V180A mutation
interfere with the physical interaction between PCNA and pol δ. Although the integrity
of the interface of PCNA is not required for PCNA-polymerase interactions (see Chapter
2), the integrity of the β-strands of domain A near the inter-domain cleft may be required
for proper interactions with pol δ. Another possible explanation for this is that these
structural changes disrupt the functional interaction between PCNA and pol δ. The large
decrease in stability in PCNA may interfere with the activity of the exonuclease domain
157
or the active site in pol δ. In support of this, exonuclease-deficient S. cerevisiae pol δ
generated base substitution errors at rates 60 fold higher than those produced by wild type
pol δ in in vivo [87]. Therefore, disruption of the functional or physical interaction
between PCNA and pol δ caused by the introduction of the V180A substitution may
result in slow cell growth and defective MMR but not TLS in vivo. This may also be the
case with the S177G mutant PCNA protein since it also shows a slow growth phenotype.
More genetic and binding studies will be necessary to distinguish between these
possibilities.
158
Figure 4.1. The amino acid residues that comprise the PCNA subunit interface.
Residues located on β-strand βI1 are shown surrounded by green and residues located on
β-strand βD2 are shown surrounded by purple. Amino acid substitutions shown in both
grey and red were used in genetic studies, while the amino acid substitutions shown in
red were used for biochemical assays.
159
Table 4.1. Summary of genetic studies with the PCNA interface mutant proteins.
I II III IV
Cell growth UV survival UV-induced mutagenesis
Spontaneous mutagenesis
WT Normal Normal Normal Normal
K164R Normal Sensitive Reduced Normal
E113G Normal Sensitive Reduced Elevated
Y114F Normal Intermediate Reduced Normal
Y114A Normal Sensitive Intermediate Elevated
S115N Normal Intermediate Reduced Normal
S177G Slow Intermediate Normal Normal
G178S Normal Sensitive Reduced Elevated
G178M Slow Intermediate Normal Elevated
G178L Slow Sensitive Reduced Normal
S179R Normal Intermediate Reduced Normal
S179T Normal Intermediate Reduced Normal
V180A Slow Sensitive Normal Elevated
I181R Normal Sensitive Reduced Normal
Reduced UV-induced mutagenesis indicates a defect in TLS. Elevated spontaneous mutagenesis indicates a defect in MMR. (Data from Christine Kondratick and Viana Ngyuen.)
160
Figure 4.2. Analysis of the wild-type and mutant PCNA proteins by native gel
electrophoresis. Coomassie stained non-denaturing polyacrylamide gradient gel (4 to
20%) in which solutions of the wild-type and mutant PCNA proteins (1.0 to 1.0 mg/ml)
were run. The position of the PCNA trimers and monomers are shown.
161
Figure 4.3. Analysis of the PCNA interface mutant proteins by size exclusion
chromatography. The elution profiles of a size exclusion chromatography column in
which a solution of the wild-type PCNA protein or the S177G, S179T, V180A, or I181R
mutant PCNA protein (10 mg/ml) was run. (Data from Christine Kondratick and Viana
Ngyuen.)
162
Figure 4.4. DNA synthesis by pol η in the presence of the PCNA mutant proteins.
(A) An autoradiogram of the products of pol η-catalyzed DNA synthesis on a non-
damaged DNA substrate in the presence of no PCNA, the wild-type PCNA protein, and
the S177G, G178S, S179T, V180A, and I181R mutant PCNA proteins. (B) An
autoradiogram of the products of pol η-catalyzed DNA synthesis on an abasic site-
containing DNA substrate in the presence of no PCNA, the wild-type PCNA protein, and
the S177G, G178S, S179T, V180A, and I181R mutant PCNA proteins. The gel band
representing extension products 6 nt. in length, which corresponds to incorporation
opposite the abasic site, is indicated by an X.
163
Figure 4.5. DNA synthesis by pol on a non-damaged template in the presence of
the PCNA mutant proteins. (A) An autoradiogram of the products of pol -catalyzed
DNA synthesis on a non-damaged DNA substrate in the presence of no PCNA, the wild-
type PCNA protein, and the S177G, G178S, S179T, V180A, and I181R mutant PCNA
proteins after a 1 minute reaction time. The position of the fully extended, runoff product
is indicated with an arrow. (B) An autoradiogram of the products of pol -catalyzed DNA
synthesis on a non-damaged DNA substrate in the presence of no PCNA, the wild-type
PCNA protein, and the S177G, G178S, S179T, V180A, and I181R mutant PCNA
proteins after a 2 minute reaction time. The position of the fully extended, runoff product
is indicated with an arrow
164
Figure 4.6. DNA synthesis by pol on a template containing an abasic site in the
presence of the PCNA mutant proteins. (A) An autoradiogram of the products of pol -
catalyzed DNA synthesis on an abasic site-containing DNA substrate in the presence of
no PCNA, the wild-type PCNA protein, and the S177G, G178S, S179T, V180A, and
I181R mutant PCNA proteins after a 1 minute reaction time. The position of the abasic
site is indicated by an X, and the position of the fully extended, runoff product is
indicated with an arrow. (B) An autoradiogram of the products of pol -catalyzed DNA
synthesis on an abasic site-containing DNA substrate in the presence of no PCNA, the
wild-type PCNA protein, and the S177G, G178S, S179T, V180A, and I181R mutant
PCNA proteins after a 2 minute reaction time. The position of the abasic site is indicated
by an X, and the position of the fully extended, runoff product is indicated with an arrow.
165
Table 4.2. Distances between potential hydrogen bond donor and acceptor atoms at
the PCNA subunit interface.
Donor Acceptor Wild-type S177G G178S S179T V180A
K117 (N) I175 (O) 3.0 Å 5.8 Å* 2.9 Å 3.0 Å 2.9 Å
S177 (N) S115 (O) 3.1 Å 2.7 Å 2.9 Å 2.9 Å 3.0 Å
S115 (N) S177 (O) 2.8 Å 2.8 Å 2.8 Å 2.9 Å 2.8 Å
S179 (N) E113 (O) 2.9 Å 3.0 Å 4.1 Å* 2.8 Å 2.8 Å
E113 (N) S179 (O) 2.9 Å 3.4 Å 5.6 Å* 3.2 Å 3.0 Å
I181 (N) I111 (O) 3.1 Å 4.3 Å* 7.5 Å* 4.1 Å* 3.6 Å*
I111 (N) I181 (O) 3.1 Å 4.9 Å* 9.3 Å* 5.0 Å* 4.0 Å*
*Indicates no hydrogen bond.
166
Figure 4.7. Structure of the V180A mutant PCNA protein. (A) One subunit of the
V180A mutant PCNA protein (green) overlayed with wild-type PCNA (cyan). The
location of the V180A substitution and domains A and B are indicated. (B) Close up of
the region affected by the V180A substitution in domain A. β-strands disrupted by the
V180A substitution are indicated. The mutant PCNA protein is shown in green and the
wild-type PCNA protein is shown in cyan.
167
Table 4.3. Summary of in vivo and in vitro studies with the PCNA interface mutant
proteins.
Reduced UV-induced mutagenesis indicates a defect in TLS. Elevated spontaneous mutagenesis indicates a defect in MMR. *V180A contains structural changes in addition to at the subunit interface. ND – not determined.
S177G G178S S179T V180A I181R
Growth rate Slow Normal Normal Slow Normal
UV sensitivity Intermediate Sensitive Intermediate Sensitive Sensitive
UV-induced
mutagenesis Normal Reduced Reduced Normal Reduced
Spontaneous
mutagenesis Normal Elevated Normal Elevated Normal
Trimer stability Intermediate Unstable Normal Intermediate Normal
Pol δ activity Intermediate Reduced Normal Reduced Normal
Pol η activity Intermediate Reduced Normal Reduced Normal
Interface H-
bonds broken 3 4 2 2* ND
168
CHAPTER 5
THE C-TERMINAL REGION OF DNA POLYMERASE η IS
INTRINSICALLY DISORDERED AND REQUIRED FOR
INTERACTION WITH PCNA AND MONOUBIQUITYLATED PCNA
Abstract
Non-classical polymerases are those that have evolved to efficiently synthesize
DNA opposite DNA lesions, and they utilize the replication accessory factor proliferating
cell nuclear antigen (PCNA) for their recruitment to the replication fork. PCNA is
monoubiquitylated during the DNA damage response, and non-classical polymerases are
thought to preferentially bind this form of PCNA. DNA polymerase η (pol η), the
prototypical non-classical polymerase, contains a PCNA interacting peptide (PIP) motif
as well as a ubiquitin-binding zinc-binding (UBZ) motif in its C-terminal 120 residues
(510-632), called the C-terminal region (CTR). Here, I show that the CTR of pol η is
intrinsically disordered, and that binding to either PCNA or Ub-PCNA does not induce
folding of this region. In addition, the CTR binds to Ub-PCNA with 19 fold higher
affinity than either PCNA or free ubiquitin alone, suggesting that the UBZ and PIP motifs
of pol η work independently to specifically bind ubiquitin and PCNA on Ub-PCNA.
Introduction
Classical polymerases, i.e. those that utilize non-damaged DNA templates,
synthesize DNA with high efficiency and accuracy. Unfortunately, these classical
polymerases are blocked at sites of DNA damage because they have highly stringent
active sites that cannot accommodate the structure of most DNA lesions. Therefore, non-
classical polymerases, i.e. those that are able to efficiently synthesize DNA on damage-
169
containing templates, have evolved to bypass DNA lesions when classical polymerases
are stalled at the replication fork. The process of replication through DNA damage is
called translesion synthesis (TLS). DNA mutations result from subsequent rounds of
replication following the insertion of an incorrect nucleotide by a non-classical
polymerase during TLS. One of the best characterized non-classical polymerases is DNA
polymerase η (pol η). Pol η, a member of the Y-family of DNA polymerases, is encoded
in yeast by the RAD30 gene. It is a monomeric protein that is capable of nucleotide
incorporation opposite several forms of damage in DNA templates, most notably thymine
dimers and 8-oxo-guanine (8-oxo-G) lesions [77, 88, 89]. Pol η consists of three
domains: the catalytic domain, the polymerase-associated domain (PAD), and the C-
terminal region (CTR). The X-ray crystal structure of the catalytic core of pol η shows
that it possesses two domains: a polymerase domain and a polymerase-associated domain
[110-112, 255]. The polymerase domain contains fingers, thumb, and palm sub-domains
similar to those found in other Y-family polymerases. It also contains an active site that is
larger than those of classical polymerases, and this larger active site allows pol to
readily accommodate thymine dimers [110, 112]. Mutations in the pol η gene in humans
causes the variant form of xeroderma pigmentosum (XP-V), an inherited disorder in
which patients have an increased sensitivity to sunlight and are at high risk of developing
skin cancers [105, 306].
The protein accessory factor that recruits and stabilizes polymerases to the
replication fork during DNA replication and TLS is proliferating cell nuclear antigen
(PCNA), a ring-shaped homotrimer encoded by the POL30 gene in yeast. PCNA is
monoubiquitylated on Lys-164 by the Rad6-Rad18 complex when cells are subjected to
DNA damaging agents. This monoubiquitylation of PCNA promotes TLS, as non-
classical polymerases preferentially associate with the ubiquitin-modified form of PCNA
(Ub-PCNA) in vivo. Where non-modified PCNA stimulates the catalytic efficiency of
DNA synthesis by pol η [113], the presence of the ubiquitin moiety on PCNA increases
170
pol η activity even further [160, 305]. Polymerases are recruited to the DNA template via
their PCNA interacting peptide (PIP) motifs that bind PCNA along its inter-domain
connector loop (IDCL). This motif is necessary for the function of pol η in vivo. Most
non-classical polymerases have an additional interaction motif that binds Ub-PCNA; in
the case of pol eta, this motif is called the ubiquitin-binding zinc-binding (UBZ) motif.
Pol -dependent TLS requires that the UBZ be intact [185].
Both of the PIP and UBZ motifs in yeast pol η reside within the CTR - the last
122 residues of the polymerase following the catalytic core (510-632). Using X-ray
crystallography and nuclear magnetic resonance (NMR), the structures of the individual
PIP and UBZ motifs of pol η were determined [114, 115]; however, the structure of the
CTR of pol η has not been examined. The CTR of pol η is not necessary for catalytic
activity of the enzyme in vitro. In contrast, the CTR of pol η is required for its
localization to nuclear foci and function in vivo [185]. However, the specific role of the
CTR in the recruitment of pol η to sites of DNA damage is unknown. Here I show that
the CTR of pol η is intrinsically disordered, and that binding to PCNA or Ub-PCNA does
not induce folding of this region. Moreover, my results suggest that the CTR of pol η is
necessary and sufficient for binding to PCNA, ubiquitin, and Ub-PCNA, and that this
region binds to the ubiquitin-modified form of PCNA with much higher affinity than it
does to either un-modified PCNA or ubiquitin alone.
Materials and Methods
Protein expression and purification.
PCNA from S. cerevisiae was over-expressed as an N-terminal His6-tagged
protein and purified from E. coli as previously described [219]. The Ub-PCNA protein
from S. cerevisiae was over-expressed and purified from E. coli using the split-fusion
strategy as previously described [215]. Full-length S. cerevisiae pol was over-
171
expressed and purified from S. cerevisiae as previously described [256]. The CTR of pol
η was over-expressed as an N-terminal His6-tagged protein and purified from E. coli
using an NTA-agarose affinity chromatography column (Qiagen), followed by a HiTrap
CM Sepharose Fast Flow column (GE Healthcare), and finally a 120 ml HiLoad 16/60
Superdex 200 PG column (GE Healthcare) in buffer containing 50 mM Arg and 50 mM
Glu. The split Ub-PCNA plasmid was produced previously in our lab, in which the gene
encoding the N-terminally Flag-tagged N-fragment of PCNA was cloned into
multicloning site 1 of the pET-Duet1 plasmid, and the gene encoding either the C
fragment or the N-terminally His6-tagged Ub-C-fragment was cloned into multicloning
site 2 of the same plasmid. To produce the split Ub-PCNA-pol η CTR fusion protein, the
gene encoding the CTR of pol η was cloned in-frame onto the C-terminus of the N-
terminally His6-tagged Ub-C-fragment of PCNA in the pET-Duet1 plasmid. The two
fragments of the Ub-PCNA-CTR protein were simultaneously overexpressed in E. coli
Rosetta-2 (DE3) cells and purified as described previously for the Ub-PCNA protein
[215].
Protein disorder prediction studies.
Disorder probability for pol η was determined using the meta method for
predicting disordered regions of proteins (metaPrDOS protein disorder meta-prediction
server). This approach predicts the tendency of each residue to be disordered by
combining results of seven different prediction methods [307].
Nuclear magnetic resonance spectroscopy.
TROSY-HSQC NMR experiments were performed with 300 µM (15
N, 1H)-
labeled sample of pol η CTR in 50 mM Tris-Cl, pH 8.0, 10 mM DTT, 100 mM NaCl, 20
mM Arg, 20 mM Glu. NMR spectra were recorded at 25oC using either a Varian 600
MHz or Bruker 500 MHz spectrometer. Titrations of 1 mM PCNA or Ub-PCNA diluted
172
in the buffer described above were added until the CTR of pol η was saturated in at least
a ratio of 1:3 of CTR:PCNA or CTD:Ub-PCNA.
Enzyme-linked immunosorbent assays.
The wells of a 96 well EIA/RIA plate (Corning) were coated with 1 µg of full-
length pol or pol η truncations in PBS (4.3 mM Na2HPO4, 1.4 mM KH2PO4, pH 7.4,
137 mM NaCl, 2.7 mM KCl) for one hour. The wells were then washed three times with
PBS, 0.2% Tween-20, blocked for 30 min. with PBS with 5% milk, and washed again.
Various amounts of PCNA or Ub-PCNA proteins or bovine serum albumin (BSA) (0.5
µg to 20 µg) in PBS with 5% milk were added to the wells and incubated for one hour,
followed by washing. A 1:200 dilution of rabbit polyclonal anti-FLAG antibody (Santa
Cruz Biotechnology) in PBS with 5% milk was added to the wells and incubated for 30
min. Wells were washed before adding a 1:10,000 dilution of mouse anti-rabbit antibody
conjugated with horseradish peroxidase (Jackson ImmunoResearch) in PBS with 5% milk
for 30 min. The plate was washed, and 0.8 mg/ml of O-phenylenediamine (Aldrich) in
0.05 M phosphate-citrate buffer with 0.03% sodium perborate (Sigma) was added.
Absorbance at 450 nm was measured after 5 to 35 min. with an iMark microplate reader
(Bio-Rad). The BSA control absorbance values were subtracted from the absorbance of
each sample at the corresponding protein concentration. All steps were performed at
25°C.
Isothermal titration calorimetry experiments.
ITC experiments were performed using a VP-ITC calorimeter (MicroCal) in 1x
PBS (pH 8.0), 50 mM Arg, 50 mM Glu. The concentration of each protein was measured
at A280 using a NanoDrop ND-1000 spectrophotometer (Thermo Scientific). All proteins
and buffers were degassed before each experiment. Heats of dilution were substracted
173
from the raw data and then the data was analyzed using ORIGIN (MicoCal) with the
single-site model and floating n-value.
Crystallization of the Ub-PCNA-CTR fusion protein.
The Ub-PCNA-CTR fusion protein was crystallized using the hanging drop
method with 400 nl drops prepared using a Mosquito Crystallization Robot (TTP
Labtech). Microcrystals were obtained by combining an equal volume of protein (16.6
mg/ml) with a reservoir containing 1% PEG 2.0, 0.1 M HEPES, pH 7.0, 1.0 M succinic
acid. Crystals formed after 45 days at 18oC.
Results
The C-terminal region of pol η is intrinsically disordered.
The structure of the CTR of pol η has not been determined. I expressed and
purified the CTR of pol η using size exclusion chromatography, which showed that the
protein was soluble and behaved as a monomer in solution. Therefore, I used this protein
in crystal screens in an effort to crystallize and solve its structure. Despite several
attempts, however, I was unable to obtain any crystals of the CTR of pol η. To examine
the possibility that the CTR is too disordered for structural determination by
crystallography, I performed disorder probability predictions on the primary sequence of
full-length pol η using the meta protein disorder prediction system [307]. Figure 5.1
shows the disorder probability of each residue in pol η. From this, I determined that the
CTR of pol η (residues 510-632) is likely to be completely disordered.
To experimentally confirm the disorder predictions of the CTR of pol η, I
performed nuclear magnetic resonance (NMR) with the CTR of pol η. The CTR of pol η
was labeled with 14
N and 1H and analyzed by heteronuclear single quantum coherence
(HSQC) spectroscopy. The resulting spectrum suggested that the CTR of pol η is
174
intrinsically disordered due to the fact that there are very few peaks in the spectrum and
many of them are broad and not widely dispersed as observed with ordered proteins
(Figure 5.2). Together, the disorder prediction analysis and NMR studies show that the
CTR of pol η is unstructured.
Binding studies of the CTR of pol η with PCNA,
ubiquitin, and Ub-PCNA.
The mechanism of polymerase recruitment to the replication fork following DNA
damage is unclear. The CTR of pol η is required for localization and function of the
polymerase in vivo, and both of the Ub-PCNA binding elements of pol η - the UBZ and
PIP motifs - are located within the CTR. To test if the CTR of pol η is necessary and
sufficient for binding to the replication fork, I examined the interactions between pol η
and PCNA using enzyme-linked immunosorbent assays (ELISAs). The full-length or
CTR of pol η was immobilized in the wells of a microtiter plate, and various
concentrations of PCNA or Ub-PCNA were added. The absorbance measured is
proportional to the amount of PCNA protein that bound to pol η. The full-length pol η
protein bound with higher affinity to Ub-PCNA than the un-modified form of PCNA
(Figure 5.3A). Similar results were seen when only the CTR of pol η was used in this
assay. The CTR of pol η bound to un-modified PCNA with lower affinity than it bound
the ubiquitin-modified form of the protein (Figure 5.3B). Furthermore, the CTR of pol η
bound to Ub-PCNA with at least the same affinity as full-length pol η bound to Ub-
PCNA (Figure 5.4), suggesting that this region of the polymerase is sufficient for the pol
η-Ub-PCNA interaction.
In order to quantitatively measure the interactions between the CTR of pol η and
the PCNA and Ub-PCNA proteins, I used isothermal titration calorimetry (ITC). A
solution containing the CTR of pol η was placed in the sample cell, and concentrated
solutions of either PCNA, ubiquitin, or Ub-PCNA were titrated into the cell. Heats of
175
injection were measured for each protein and these data were fit to a single-site binding
isotherm to obtain binding affinities for each interaction. The CTR of pol η bound to the
un-modified form of PCNA with a Kd of 3.8 µM (Figure 5.5A,B). This value is probably
a direct measure of the interaction between the PIP motif of pol η and PCNA. The CTR
of pol η bound to free ubiquitin with a similar affinity of Kd of 2.2 µM (Figure 5.5C,D).
This value is probably a direct measure of the interaction between the UBZ of pol η and
ubiquitin. In contrast, the CTR of pol η bound to Ub-PCNA with a Kd of 0.2 µM, which
was 19 fold higher than either PCNA or ubiquitin alone (Figure 5.6). This value is
probably a measure of the interactions between both the PIP and UBZ motifs of pol η and
Ub-PCNA. Similar results were observed in an analogous system, where an unstructured
region of the anti-recombinogenic helicase Srs2 containing both SUMO-interacting and
PCNA-interacting motifs bound sumoylated PCNA with higher affinity than either
SUMO or PCNA alone [308]. The results from my binding studies show that the PIP and
UBZ motifs of pol η bind ubiquitin and PCNA independently.
To examine the importance of the unstructured region between the PAD and UBZ
of pol η in its interaction with Ub-PCNA, I generated a truncation of pol η containing
only residues 546 through 632 (the UBZ motif is defined as residues 548 to 577). ITC
analysis showed that, while it bound to PCNA with the expected affinity, this truncation
protein did not interact with free ubiquitin or Ub-PCNA. Together, the ITC data shows
that both the UBZ and PIP motifs are required for pol η’s high affinity interaction with
Ub-PCNA and that additional contacts between the PAD and UBZ of pol η are necessary
for this interaction.
Because the CTR of pol η is unstructured but binds tightly to both un-modified
and ubiquitin-modified PCNA, it is possible that this interaction induces folding of the
CTR. To investigate this possibility, I performed NMR experiments with the CTR of pol
η in complex with either PCNA or Ub-PCNA. The CTR was 14
N- and 1H-labeled and a
solution of 300 µM was placed in a 600 MHz (in the case of PCNA) or an 800 MHz (in
176
the case of Ub-PCNA) NMR instrument to obtain HSQC spectra at 25oC. Then
concentrated solutions of either PCNA or Ub-PCNA (1 mM) were titrated in until the
CTR was saturated. Resulting spectra show that, even at saturating conditions, the
addition of PCNA (Figure 5.7) and Ub-PCNA (Figure 5.8) do not cause significant shifts
in the peaks. This suggests that the CTR of pol η does not gain structure upon binding to
either PCNA or Ub-PCNA.
Crystallography studies of the complex of the CTR
of pol η and Ub-PCNA.
Determination of the structure of the CTR of pol η in complex with Ub-PCNA
would provide valuable insight into how pol η is recruited to sites of DNA damage by
Ub-PCNA. However, based on my studies of the CTR of pol η, I predict that the
interaction between the CTR and Ub-PCNA is likely transient. Therefore, in order to
ensure that each subunit of PCNA is bound to a CTR peptide, I engineered a strategy to
fuse the CTR of pol η to the end of the second fragment of the Ub-PCNA polypeptide
(Figure 5.9A). This is possible because the N-terminus of the CTR of pol η likely binds
to Ub-PCNA near the C-terminus of the second fragment. Using this method of protein
expression, I have been able to purify substantial quantities of the fusion protein. Size
exclusion chromatography experiments show that it forms a stable trimer and, as
expected, that these trimers are slightly larger than the trimers formed by Ub-PCNA
(Figure 5.9B). Because of this, I prepared several 96-well crystallization trays with the
Ub-PCNA-CTR fusion protein. Under certain conditions, I was able to obtain some
small crystals (Figure 5.10), but more experimentation is necessary in order to produce
crystals large enough to be used to gather diffraction data for X-ray crystallography
analysis.
177
Discussion
The CTR of pol η is required for its function in vivo [185]. In this Chapter, I
show that the CTR of pol η binds with at least the same affinity to Ub-PCNA as the full-
length polymerase. Because Ub-PCNA is required for pol η localization in cells, these
studies suggest that the CTR of pol η is necessary and sufficient for pol η recruitment to
the replication fork during the DNA damage response. In addition, binding studies
discussed here show that the UBZ-ubiquitin and PIP-PCNA interactions involved in pol
η-Ub-PCNA complex formation bind independently of each other. Since classical
polymerases do not possess ubiquitin-binding motifs, this could suggest a model of TLS
where, after monoubiquitylation, PCNA preferentially interacts with a non-classical
polymerase such as pol η during TLS. This would ensure that a non-classical polymerase
is recruited to the replication fork following DNA damage and facilitate the switch from
the classical polymerase to a non-classical polymerase.
The CTR of pol η is intrinsically disordered, and protein disorder prediction
studies of other Y-family DNA polymerases suggest that they also contain intrinsically
disordered C-terminal regions (see Chapter 1, Figure 1.21). This is common, as nearly
one-third of eukaryotic proteins and one-half of mammalian proteins are partially or fully
disordered [309, 310]. As with pol η and the other eukaryotic Y-family polymerases,
these unstructured regions are often involved in a multitude of protein-protein
interactions in the cell [309-311]. Despite the high affinity interaction between the CTR
of pol η and Ub-PCNA, the binding of these proteins do not induce folding of the CTR,
which is likely the case with many intrinsically disordered proteins. This could be due to
the transient nature of the interactions formed between pol η and Ub-PCNA because the
activity of pol η is only required at the replication fork for non-processive DNA
synthesis. Lack of rigid protein conformations in the CTR of pol η would also ensure
enough flexibility in the polymerase to allow interaction with the front face of PCNA
178
(where the PIP binding region is located) and interaction with the back face of PCNA
(where the ubiquitin moiety is located) simultaneously (see Chapter 1, Figure 1.23).
PCNA is involved in a very dynamic network of interactions during DNA
replication and repair. For instance, PCNA coordinates the switch from the classical
polymerase to the non-classical polymerase and back during TLS. It has been suggested
that PCNA accomplishes this by acting as a tool belt, in which the non-classical
polymerase is recruited to the back face or side of the PCNA ring while a classical
polymerase is engaged at the DNA primer terminus at the front face. Flexible
interactions between PCNA and the polymerases would allow efficient and rapid
polymerase switching at the replication fork. Because PCNA is crucial for an array of
cellular processes, it is likely that many PCNA-interacting proteins possess intrinsically
disordered regions in order to allow PCNA to perform sequential and efficient protein
switching on DNA substrates.
179
Figure 5.1. The structured and unstructured regions of yeast pol η. The graph of
disorder probability for pol was obtained using the meta approach for predicting
disordered regions of proteins at a prediction false positive rate of 5.0% [307]. In this
diagram, residues with disorder probabilities below 0.5 are likely to be structured, and
residues with probabilities above 0.5 are likely to be disordered. The structured regions
are shown as thick rectangles, and the disordered regions are shown as thin rectangles.
The polymerase domain, PAD, and PIP and UBZ motifs are indicated.
180
Figure 5.2. The 1H-
15N heteronuclear single quantum coherence (HSQC) spectrum
of the CTR of pol η. Peaks resulting from backbone and sidechain amide bonds are
shown in blue.
181
Figure 5.3. Analysis of the interaction of the PCNA and Ub-PCNA proteins with pol
using ELISA. (A) Results of an ELISA assay showing the interaction of the un-
modified PCNA protein (red) and the Ub-PCNA protein (green) with full-length pol .
(B) Results of an ELISA assay showing the interaction of the un-modified PCNA protein
(red) and the Ub-PCNA protein (green) with the CTR of pol .
182
Figure 5.4. Analysis of the interaction of the full-length pol and the CTR of pol η
with Ub-PCNA. Results of an ELISA assay showing the interaction of the full-length pol
η protein () and the CTR of pol η () with Ub-PCNA.
183
Figure 5.5. Analysis of the interaction of the CTR of pol with PCNA and
ubiquitin using ITC. (A) Raw measured heat changes as a function of time as PCNA
(200 µM) was injected into a cell containing the CTR of pol η (10 µM) at 25oC. (B)
Normalized measured heats of injection and the best-fit values for these heats.
Thermodynamic parameters were estimated to be N = 1.0, K = 2.6 x 105/M, and ΔH = -
8.7 kcal/mol for this interaction. (C) Raw measured heat changes as a function of time as
ubiquitin (100 µM) was injected into a cell containing the CTR of pol η (10 µM) at 25oC.
(B) Normalized measured heats of injection and the best-fit values for these heats.
Thermodynamic parameters were estimated to be N = 0.5, K = 4.6 x 105/M, and ΔH = -
9.3 kcal/mol for this interaction.
184
Figure 5.6. Analysis of the interaction of the CTR of pol with Ub-PCNA using
ITC. (A) Raw measured heat changes as a function of time as Ub-PCNA (100 µM) was
injected into a cell containing the CTR of pol η (10 µM) at 25oC. (B) Normalized
measured heats of injection and the best-fit values for these heats. Thermodynamic
parameters were estimated to be N = 0.13, K = 2.3 x 106/M, and ΔH = -434 kcal/mol for
this interaction.
185
Figure 5.7. The 1H-
15N HSQC spectra of the CTR of pol η bound to PCNA. Data
were collected on a 600MHz NMR instrument. (A) Spectrum of the CTR of pol η alone.
Peaks resulting from backbone and sidechain amide bonds are shown in red. (B)
Spectrum of the CTR of pol η after addition of PCNA. Peaks resulting from backbone
and sidechain amide bonds are shown in black. (C) Spectrum of the CTR of pol η (red)
overlayed with the spectrum of the CTR of pol η after addition of PCNA (black).
186
187
Figure 5.8. The 1H-
15N HSQC spectra of the CTR of pol η bound to Ub-PCNA.
Data were collected on an 800 MHz NMR instrument. (A) Spectrum of the CTR of pol η
alone. Peaks resulting from backbone and sidechain amide bonds are shown in blue. (B)
Spectrum of the CTR of pol η after addition of Ub-PCNA. Peaks resulting from
backbone and sidechain amide bonds are shown in red. (C) Spectrum of the CTR of pol
η (blue) overlayed with the spectrum of the CTR of pol η after addition of Ub-PCNA
(red).
188
189
Figure 5.9. Production and purification of the Ub-PCNA-CTR fusion protein. (A)
Polypeptides used to produce the Ub-PCNA-CTR protein. (B) Size exclusion
chromatography of the Ub-PCNA-CTR fusion protein. The protein formed a trimer with
the expected molecular weight for one CTR bound per PCNA monomer.
190
Figure 5.10. Crystallization of the Ub-PCNA-CTR fusion protein. Microcrystals that
formed in 1% PEG 2.0, 0.1 M HEPES, pH 7.0, 1.0 M succinic acid after 45 days at 18oC.
191
CHAPTER 6
DISCUSSION
When pol δ encounters damaged DNA, normal replication is stalled, which results
in either replication fork collapse, error-free replication, or mutagenic translesion
synthesis. By understanding PCNA’s role in TLS and other replication and repair
pathways, we could learn to manipulate PCNA’s functions and promote less mutagenic
processes such as error-free replication. In order to gain a better understanding of the
role of PCNA during replication and repair, I carried out structural and functional
characterization of several mutant forms of PCNA. The positions of all eight PCNA
substitutions used in my studies are shown in Figures 6.1 and 6.2. The focus of this
thesis is on the role of PCNA during TLS and MMR, and my overall objective was to
understand the mechanism by which specific mutations in PCNA cause defects in these
processes. This lead to the finding that the subunit-subunit interface of PCNA is
extremely dynamic. Therefore, I also focused on how random mutations in this region of
PCNA affect TLS and MMR. I identified that mutations that reduce PCNA trimer
stability caused the most significant defects in TLS in vitro, whereas differing structural
alterations in PCNA can impact MMR. Lastly, I examined the mechanism for
recruitment of the non-classical DNA pol η to the replication fork by Ub-PCNA during
TLS. My studies suggest that the unstructured C-terminal region of pol η is necessary
and sufficient for its recruitment to Ub-PCNA, and provide evidence that other non-
classical polymerases likely utilize this same mechanism of recruitment to the replication
fork.
192
Integrity of the PCNA Interface during Translesion Synthesis
Overview of studies with the E113G mutant
PCNA proteins.
Two mutant yeast PCNA proteins, encoded by the pol30-178 allele and the pol30-
113 allele, support normal cell growth in yeast but exhibit reduced mutagenesis and are
defective in promoting TLS [238, 262]. These alleles encode the G178S mutant PCNA
protein and the E113G mutant PCNA protein, respectively. The X-ray crystal structures
of these mutant proteins have been determined [219, 260]. Based on these structures,
previous work has suggested that the position of a loop near the subunit interface of
PCNA, called loop J, is important for TLS and it was proposed that loop J is involved in
interactions with non-classical polymerases during TLS [219]. After careful re-
evaluation of the structures of the G178S and E113G mutant PCNA proteins, I observed
significant alterations of the subunit interface, as described in Chapter 2. These
alterations caused drastic trimer instability in the G178S mutant PCNA protein and a
slight destabilization of the E113G mutant PCNA trimer, but did not inhibit binding to
pol η by either mutant protein. While the presence of wild-type PCNA or Ub-PCNA
stimulated the activity of both classical pol δ and non-classical pol η, the G178S mutant
PCNA protein failed to stimulate DNA synthesis on both damaged and non-damaged
templates by either polymerase. In comparison, the presence of the E113G mutant
PCNA protein did not stimulate TLS opposite an abasic site by either polymerase or
normal replication by pol η, however normal DNA replication by pol δ was not
significantly affected. These findings indicate that reduced trimer stability of the G178S
and E113G mutant PCNA proteins causes them to undergo conformational changes that
compromise their ability to stimulate TLS by both classical and non-classical
polymerases.
193
Role of Ub-PCNA during TLS.
Monoubiquitylation of PCNA plays an important role during TLS. This event
triggers the switch between the classical and non-classical polymerases, and non-classical
polymerases preferentially bind this form of PCNA. Small angle X-ray scattering
(SAXS) analysis, in combination with X-ray crystallography and multiscale
computational modeling, showed that the ubiquitin moiety is capable of adopting
multiple positions on PCNA [192]. The three most likely positions are shown in Figure
6.1, which demonstrates that the ubiquitin can sit on the side of the PCNA ring, the back
face of the PCNA ring, or in a flexible conformation extended away from the PCNA ring.
Based on these results, it was proposed that each of these positions has a specific function
during TLS [192]. The “flexible” and “back” positions may allow for non-classical
polymerase binding and recruitment to the back face of PCNA without disrupting the
classical polymerase that is already engaged on the DNA substrate at the front face. The
“side” position would then allow the non-classical polymerase to gain access to the
primer-terminus to progress through TLS.
Monoybiquitylated PCNA functions during TLS in several ways. First, it acts as
both a recruiter and a scaffold by interacting with DNA polymerases and providing a
docking site for polymerases to carry out DNA synthesis. Second, it is a processivity
factor and enhances the activity of these polymerases. In Chapter 2, I showed that the
E113G mutation in Ub-PCNA causes a defect in TLS by inhibiting stimulation of DNA
synthesis on an abasic site by classical and non-classical polymerases, but not by
interfering with binding or enhancing the processivity of these enzymes. However, I did
not address the possibility that the E113G substitution inhibits TLS by producing a
change in the potential positions of the ubiquitin moiety on Ub-PCNA. Interestingly, the
Glu-113 residue is located at the interface of PCNA, and the ubiquitin positioned on the
side of PCNA could potentially contact this residue. Therefore, it is possible that the
E113G substitution disrupts the “side” position of ubiquitin, thereby not allowing the
194
non-classical polymerase access to the replication fork. To address this possibility, I
collaborated with Dr. Susan Tsutakawa of Lawrence Berkeley National Labs to carry out
SAXS analysis with the wild-type Ub-PCNA and the E113G mutant Ub-PCNA proteins
(Figure 6.2). Results show that the E113G substitution does not significantly change the
conformation of Ub-PCNA, as the scattering profiles for both proteins display the same
overall shape. Therefore, the defect in TLS observed with this mutant protein is not
likely due to restricting the ubiquitin moiety on Ub-PCNA to adopt any particular
positions.
Possible insights from studies with the E113G
mutant PCNA and Ub-PCNA proteins.
The kinetic studies with classical pol δ described in Chapter 2 show that the
presence of the ubiquitin on PCNA appears to stimulate pol δ’s activity opposite an
abasic site more so than non-modified PCNA. Along with the fact that Ub-PCNA
stimulates DNA damage bypass by non-classical pol η, these data could indicate that the
ubiquitin modification on PCNA stimulates the process of TLS by both classical and non-
classical polymerases. When the E113G substitution and the ubiquitin moiety are both
present on the PCNA protein simultaneously, however, it is clear that the inhibition by
the mutation prevails over the stimulation by the ubiquitin. The fact that the presence of
the ubiquitin moiety on PCNA stimulates synthesis on damaged DNA templates by both
classical and non-classical polymerases is interesting, as pol δ does not possess an
ubiquitin-binding motif. How this occurs is unclear, and future studies will be needed to
determine the mechanistic basis for these findings.
The intrinsic processivity of pol δ is very efficient on short stretches of DNA, but
pol δ requires PCNA to be fully processive over thousands of base pairs [163]. Previous
studies on the processivity of pol η, on the other hand, have been conflicting. In Chapter
2, I show that pol η requires PCNA for processive DNA synthesis, whereas Ub-PCNA
195
increases the processivity even further. I also show that the presence of the E113G
mutant PCNA protein inhibits the processivity of pol η to very low levels. However, the
attachment of the ubiquitin to PCNA rescues the processivity to levels higher than that
with wild-type PCNA. Therefore, in the case of pol η, stimulation of processivity by the
presence of ubiquitin on the PCNA trimer predominates over the inhibition by the
presence of the E113G mutation, which is in complete contrast to their effects on the
efficiency of the polymerase activity.
Overall, the E113G substitution in PCNA slightly decreases stability of the PCNA
trimer. This compromises both classical pol δ and non-classical pol η activity during
TLS, but does not significantly decrease DNA replication by non-classical pol δ. Results
of the processivity assays with pol δ and pol η suggest that the addition of the ubiquitin
on the E113G mutant PCNA protein may rescue the processivity of these polymerases
(see Figures 2.6 and 2.8). A possible explanation for this is that the presence of the
ubiquitin on PCNA may stabilize PCNA-polymerase complexes at primer-termini,
especially in the case of pol η and other non-classical polymerases due to the additional
ubiquitin-UBZ contacts. Future studies measuring the stability of PCNA-polymerase-
DNA complexes will be necessary to explore this possibility.
Defects in Mismatch Repair Caused by Distinct Structural Alterations in PCNA
Overview of studies with the C22Y and C81R
mutant PCNA proteins.
The role of PCNA during MMR is not well understood. Studies suggest that it is
required for several steps during the repair of mismatched bases; including the initiation,
mismatch recognition, excision, and resynthesis steps. To better understand the function
of PCNA during MMR, I performed structural and biochemical experiments on two
mutant PCNA proteins shown to be defective in the MMR pathway. As described in
196
Chapter 3, two mutant PCNA alleles (pol30-201 and pol30-204) were shown to inhibit
MMR without causing defects in other cellular processes in vivo [242]. The pol30-201
allele encodes the C22Y mutant PCNA protein and causes a strong defect in MutSα-
dependent MMR, while the pol30-204 allele encodes the C81R mutant PCNA protein
and causes a partial defect in both MutSα-dependent and MutSβ-dependent MMR [242].
We determined the X-ray crystal structure of each of these mutant PCNA proteins and
identified that they produce structural alterations at different locations within PCNA.
The C22Y substitution creates a distortion in the α-helices lining the central hole of the
PCNA ring, whereas the C81R substitution distorts the β-sheet at the PCNA subunit
interface. The work described in Chapter 3 demonstrates that the C81R mutation
significantly reduces PCNA binding to MutSα, while the C22Y mutation in PCNA does
not affect MutSα binding. Similarly, the C81R mutant PCNA protein stimulates DNA
synthesis by pol δ considerably less than the C22Y mutant PCNA protein, which
stimulated pol δ activity almost to the same level as wild-type PCNA. However,
sedimentation analysis demonstrated that both the C22Y and C81R mutant proteins
formed aberrant complexes with MutSα and DNA in the presence of a mismatched base
pair. Therefore, we proposed that the α-helices within the central hole and the β-sheet at
the subunit interface are both important for proper PCNA function during MMR.
How do the C22Y and C81R substitutions prevent
productive PCNA complex formation during MMR?
Despite the fact that the C22Y and C81R mutations cause distinct structural
changes in PCNA, they both inhibit MMR by forming aberrant complexes with MutSα
and mismatch-containing DNA. One possible explanation for this result is that the
presence of these mutant PCNA proteins induces aggregation of MutSα or causes
multiple MutSα complexes to bind one PCNA trimer simultaneously. This idea is
supported by the fact that MutSα is seen only in the high molecular weight fractions
197
during sedimentation analysis, suggesting that more than one MSH2 and one MSH6
protein is present in these complexes. A second possibility is that the PCNA mutant
proteins may not be in the preferred oligomeric state for proper MMR. It is possible that
the C22Y and C81R mutant proteins are forming monomers, dimers, trimers, or even
hexamers under these experimental conditions. This is not unlikely, as the C81R mutant
PCNA protein was shown to exist in different oligomeric states (monomers and dimers)
at various protein concentrations (Figures 3.3 to 3.5). Different oligomeric forms are also
seen with PCNA from Pyrococcus furiosus, as it mostly forms ring-shaped hexamers but
can also exist as C-shaped tetramers or pentamers [212, 312]. Furthermore, even though
PCNA is usually observed as a trimer in vitro, it is tempting to speculate that eukaryotic
PCNA may actually exist as a hexamer in vivo, as leading and lagging strand DNA
replication are coupled and would require one trimer per DNA templating strand. Thus it
is feasible to consider that PCNA functions as a hexamer during MMR as well. In
support of this, I consistently observe a small proportion of PCNA in higher-order
oligomers when performing non-denaturing gel electrophoresis with wild-type protein
(Figure 6.3). Therefore, it is possible that the C22Y and C81R mutant proteins are unable
to support normal MMR due to their inability to form correct trimeric or higher-order
oligomeric species.
Another possible explanation for the results described in Chapter 3 is that the
C22Y and C81R substitutions in PCNA are disrupting MMR by causing PCNA to form
non-productive complexes with MutSα or with mismatch-containing DNA. The C22Y
mutant protein could potentially be forming a non-productive complex with DNA. The
structure of this mutant PCNA protein revealed perturbations of the α-helices that line the
inner hole, and these residues contact DNA during normal PCNA-dependent processes.
Thus these contacts between PCNA and DNA may be disrupted in the presence of a
mismatched base pair. In contrast, the C81R mutant PCNA protein is more likely to form
a non-productive complex with MutSα. The crystal structure of this mutant protein
198
exhibited a local distortion in a β-strand near the subunit interface. This led to greatly
reduced trimer stability and prevented the C81R mutant PCNA protein from properly
interacting with MutSα. From these results, we draw the possible conclusion that the
subunit interface is a potential interaction site for MutSα. The N-terminal region of
MSH6 contains a PCNA-interacting peptide (PIP) motif that binds PCNA in its canonical
PIP-interacting hydrophobic binding pocket. Several PCNA-interacting proteins,
however, are known to make secondary contacts with PCNA in addition to the PIP
interaction site [208, 209]. Thus MutSα binding may also require secondary contacts
with the subunit interface to form a productive PCNA-MutSα complex.
The last plausible explanation for aberrant PCNA-MutSα-DNA complex
formation by the mutant PCNA proteins is that they prevent release of MutSα from
PCNA after the mismatched DNA has been identified. Mispair binding studies have
demonstrated that PCNA and MutSα form a ternary complex with homoduplex DNA, but
that the presence of a mispair causes PCNA to dissociate from the complex [56]. These
results suggest that, during mismatch recognition, MutSα interacts with PCNA that is
pre-bound to the DNA and is then transferred to the mismatched base pair [56]. The
structural changes created by the C22Y and C81R substitutions may therefore lock
MutSα in complex with the PCNA and DNA substrate, preventing it from release onto
the mistmatch.
Importance of the Subunit Interface of PCNA for Translesion Synthesis and
Mismatch Repair
Overview of studies with random PCNA
interface mutant proteins.
Chapters 2 and 3 describe my work with four mutant PCNA proteins that block
TLS and MMR. The E113G mutant PCNA protein inhibits TLS, while the C22Y and
199
C81R mutant proteins inhibit MMR. Interestingly, three of these four mutant proteins
cause structural disruptions near the PCNA subunit interface. We therefore concentrated
our efforts on this region of the protein, and Christine Kondratick and Viana Nguyen
generated a large set of random PCNA interface mutant proteins to characterize
genetically, structurally, and biochemically. My studies focused on in vitro biochemical
characterization of five of these mutant PCNA proteins - consisting of the S177G,
G178S, S179T, V180A, and I181R substitutions. I demonstrated that these substitutions
cause varying degrees of trimer instability, which correlated directly with these mutant
proteins’ abilities to stimulate DNA synthesis by pol δ and by pol η, and particularly on
damaged DNA templates. Structural analysis of these mutant PCNA proteins also
correlated with the biochemical results, in that larger structural perturbations at the
interface resulted in less stable trimers and a greater reduction in TLS. In the case of the
V180A mutant PCNA protein, however, the most substantial structural changes were
seen in the β-strands nearby the subunit interface, which still caused significant trimer
instability and defective TLS in vitro. Genetic analysis of these five mutant PCNA
proteins by Christine suggested that, in vivo, these five mutant PCNA proteins exhibit
slightly different effects on TLS than what I observed in vitro. Genetic analysis was also
performed with these five mutant PCNA proteins to determine their effects on MMR in
vivo. In order to examine the effect of each PCNA interface mutant protein on MMR in
vitro, we plan to perform sedimentation analysis with MutSα, mismatch-containing DNA,
and each of the PCNA interface mutant proteins. This will show if these amino acid
substitutions in PCNA affect its ability to form ternary complexes with MutSα and DNA,
which was observed to be the case with both of the MMR-defective C81R and C22Y
mutant PCNA proteins (see Chapter 3). Together, our studies show that the subunit
interface of PCNA is very dynamic and that small changes at the interface can cause
drastically different effects on TLS and MMR, however more experimentation will be
necessary to elucidate the mechanism by which these mutant proteins function in vivo.
200
Shearing of the PCNA interface.
The interface of PCNA is held together by seven hydrogen bonds between
monomer backbone residues and is comprised of β-strand βD2 from one monomer and βI1
from another monomer (see Figure 4.1). The crystal structures of numerous mutant
PCNA proteins that display changes at the subunit interface have been determined in our
lab. These include the C81R, E113G, S177G, G178S, S179T, and V180A mutant
proteins. These structures all show shearing of the interface resulting from loss of one or
more hydrogen bond, which always occurs on the same side of the PCNA ring (the
interface shears such that the opening faces toward the back face of PCNA) and even the
same β-strand (βI1). This suggests that the residues at the N-terminus of βI1 (R110, I111,
etc.) are more dynamic and easily destabilized than those in β-strand βD2 or at the C-
terminus of βI1. Interestingly, these residues are located directly C-terminal to loop J,
whose position was shown to be shifted in the E113G and G178S mutant PCNA protein
structures. The flexibility of these residues in βI1 and loop J are apparent in Figure 6.4,
which shows the B-factors of the interface of wild-type PCNA according to size. The
extremity of βD2 at the back face of PCNA is also adjacent to a dynamic extended loop,
but this does not seem to affect the dynamics of the atoms involved in hydrogen bonding
at the interface. The residues that make up the interface near the front face of PCNA
extend into small, tight turns that appear to be rigid. Together, these data suggest that
loop J and the residues immediately C-terminal to it (R110, I111, etc.) are dynamic,
which is likely the cause of interface shearing and trimer instability of the PCNA
interface mutant proteins. We plan to solve the structure of at least one more interface
mutant (I181R), and I predict that the structure will exhibit perturbations at the N-
terminus of β-strand βI1 as well.
How does disruption of the interface of PCNA cause defects in multiple DNA
metabolic pathways? A likely scenario is described in Chapter 2, in which reduced
trimer stability of the PCNA interface causes the protein to undergo conformational
201
changes that compromises its ability to stimulate the PCNA-binding proteins involved in
TLS and MMR. One explanation for this is that instability of the trimer promotes PCNA
dissociation from the DNA substrate. In this situation, the presence of kinetic barriers as
well as the presence of certain TLS and MMR protein factors will likely play an
important role as to when this dissociation will occur. For instance, the presence of a
polymerase or MutSα or MutSβ may help stabilize PCNA at the replication fork.
Experiments investigating the dissociation of PCNA mutant proteins from DNA in the
absence and presence of PCNA-interacting proteins (similar to those described in [313])
are necessary to elucidate if this is in fact the case.
It is also appealing to speculate that many defects caused by disruptions of the
PCNA interface are due to lack of physical or functional interactions that normally occur
near this region. Factors involved in TLS or MMR may require the interface to be
dynamic to allow their proper function. Several PCNA-interacting proteins make
secondary contacts with PCNA outside of the PIP-binding pocket, and it is likely that
many sites on PCNA that are required for enzyme activity during DNA replication and
repair have not yet been discovered. For instance, it is plausible that MutSα contacts
PCNA near the interface, as perturbations in this region render the C81R mutant PCNA
protein defective in MutSα binding. It is also possible that the integrity of the interface is
required for functional interactions with many PCNA binding partners, however future
work will be necessary to determine if this is the case.
Growth phenotypes of the S177G and G178S
mutant PCNA proteins.
The biochemical assays described in this thesis have shown that the G178S
mutant PCNA protein is defective in promoting TLS by both classical and non-classical
polymerases as well as normal DNA synthesis by classical pol δ. Genetic studies, on the
other hand, have shown that PCNA containing the G178S substitution display normal cell
202
growth. How does this mutant protein support normal cell growth in vivo, but not normal
DNA replication in vitro? My work described in Chapter 2 suggests that the defect in
stimulating polymerase activity in vitro is due to the substantial trimer instability
produced by the presence of the mutation, as the G178S mutant PCNA protein does not
form stable trimers under any concentration tested (Figures 2.2 and 2.3). In the cell,
however, this mutant protein may actually be stable as a trimer due to high local
concentrations of the protein and the presence of many other protein factors involved in
DNA replication.
Genetic studies described in Chapter 4 show an opposite phenotype with the
S177G mutant PCNA protein compared to the G178S mutant PCNA protein. The S177G
mutant PCNA protein behaves like wild-type PCNA in experiments that examined its
ability to function in TLS and MMR. However, cells expressing this mutant protein
display slow growth rates. Therefore, it is likely that other DNA metabolic processes or
proteins are being affected by this mutation besides TLS or MMR. For example, the
nucleotide or base excision repair pathways could be defective in these cells, or the
substitution may be impacting interactions with protein factors such as DNA ligase or the
polymerases involved in leading and lagging strand replication (pol δ, pol ε, pol α – see
Figure 1.1) to impair DNA replication in general. Future experiments with this the
G178S mutant PCNA protein will be necessary to distinguish between these possibilities.
PCNA-dependent processes and PCNA substitutions
that disrupt them.
PCNA plays a multifaceted role in maintaining proper DNA metabolism and
genome integrity. This requires PCNA for the recruitment and coordination of numerous
proteins involved in DNA replication, recombination, and repair to the replication fork.
PCNA functions not only during DNA replication, translesion DNA synthesis, base
excision repair, nucleotide excision repair, MMR, and recombination, but also during
203
chromatin assembly and remodeling, sister chromatid cohesion, and cell cycle control
[155, 199, 200, 257, 258]. Over time, an increasing number of PCNA mutations have
been identified that disrupt one or more of these nuclear processes. For example,
mutations in PCNA that cause defects in MMR [242], translesion DNA synthesis [238,
240], error-free post-replication repair [297], and chromatin remodeling [298, 299] have
been identified and are beginning to be characterized.
In some cases, mutant PCNA proteins seem to specifically disrupt one pathway in
vivo. However, in many cases, a single amino acid substitution can lead to extensive and
general defects in multiple PCNA-dependent processes. For instance, numerous
substitutions have been identified that disrupt MMR, but most of these mutations are also
sensitive to temperature, MMS treatment (suggesting a deficiency in base excision repair
and nucleotide excision repair), or UV light (suggesting a deficiency in TLS) [242]. In
addition, the G178S mutation in PCNA causes global disruptions in several pathways
including TLS, MMR, and DNA synthesis by classical and non-classical polymerases.
Many mutations in PCNA that appear analogous are actually capable of causing
drastically differing phenotypes, and most of these mutations disrupt more than one
cellular process. These data show that we need to be cautious about which mutant PCNA
proteins we utilize to study individual pathways. Thus future work will require detailed
experimentation to determine exactly how mutant proteins behave in vivo and identify
PCNA substitutions appropriate for studying individual pathways.
PCNA and Its Interactions with Intrinsically Disordered Proteins
Overview of studies with the C-terminal region of pol η.
DNA is constantly bombarded by internal and external DNA damaging agents,
causing lesions at an astounding rate. Consequently, classical DNA polymerases will
encounter lesions and normal DNA replication will be blocked at the damaged base(s).
204
During TLS, the classical polymerase is replaced at the primer terminus by a non-
classical polymerase that is capable of synthesizing DNA opposite the damage. The best
studied non-classical polymerase is pol η. The C-terminal region (CTR) of pol η
(residues 510-632) contains a PIP motif as well as a ubiquitin-binding zinc-finger (UBZ)
motif that binds ubiquitin with high affinity. PCNA is monoubiquitylated in response to
DNA damage exposure, and pol η preferentially binds Ub-PCNA over un-modified
PCNA in vivo. My work described in Chapter 5 shows that the CTR of pol η is
intrinsically disordered, which does not become structured upon binding either PCNA or
Ub-PCNA. Using several binding assays with pol η and PCNA or Ub-PCNA, I showed
that the PIP and UBZ together bind 19-fold tighter to Ub-PCNA than either PCNA or
ubiquitin alone. These binding studies also suggest that the CTR of pol η is both
necessary and sufficient for interaction with Ub-PCNA and is therefore likely to be solely
responsible for recruitment of pol η to the replication fork during TLS.
Regulation of PCNA interactions.
Not many proteins exist in cells that function in so many processes and associate
with so many other protein factors as PCNA. It acts as a protein recruiter, scaffold,
processivity factor, and enhancer of a multitude of proteins during all processes involved
in DNA metabolism. PCNA is involved in DNA replication and recombination, TLS,
MMR, base excision repair, nucleotide excision repair, chromatin reassembly, cell cycle
control, sister chromatid cohesion [155, 199, 200, 257, 258]. Therefore, it is crucial that
PCNA functions properly and efficiently to regulate the position and function of proteins
involved in all of these pathways.
The means by which PCNA orchestrates control of so many proteins is currently
an active area of research. So far, several modes of regulation have been proposed that
are likely to help facilitate sequential action of PCNA-interacting proteins. The first and
perhaps the most obvious mode of regulation is competition by these proteins. For
205
instance, p21 has the highest affinity for PCNA than any other PIP motif-containing
protein [201]. Since p21 is a cell cycle regulator, it is essential that this protein have the
ability to displace any other protein from PCNA at any moment. Along these same lines
is the idea that post-translational modifications of PCNA regulate the binding and/or
removal of specific interacting proteins. Several studies have shown that the
monoubiquitylation of PCNA recruits non-classical polymerases that contain ubiquitin-
binding motifs, whereas sumoylation of PCNA recruits the anti-recombinogenic helicase
Srs2 to the DNA. Similarly, several PCNA-binding factors are known to be post-
translationally modified themselves. For example, FEN1 is phosphorylated by the
CDK2-PCNA complex, which leads its dissociation from PCNA [314]. In addition, the
observation that many non-classical polymerases are capable of ubiquitylation has led to
the notion that this event acts as a switch to prevent re-association of these enzymes after
TLS is completed.
PCNA is a homotrimer and therefore has the capability of binding up to three PIP
motif-containing proteins simultaneously. This would allow the ordered binding of
factors involved in multi-protein processes. Thus one protein could be recruited to one
subunit of PCNA, which would stimulate binding of a second factor to another subunit,
and so on. Indeed, studies show that this is likely the case with Sulfolobus solfataricus
PCNA, in which FEN1, DNA ligase, and a DNA polymerase can bind one PCNA
heterotrimer simultaneously [235].
After all of these factors bind, how does PCNA regulate the dissociation of these
proteins? As with loading on to DNA substrates, PCNA is also unloaded by RFC. This
would act to remove all PCNA-interacting partners from the DNA. It has also been
suggested that sumoylation of PCNA at K127 causes dissociation of proteins from
PCNA, although this pathway is not well understood [315]. Lastly, PCNA-protein
complexes can be permanently eliminated by PCNA degradation through ubiquitylation.
This ubiquitylation is separate from the RAD6 pathway and causes reduced DNA
206
synthesis and defects in mismatch repair, but specific details of this pathway remain to be
elucidated.
My work in Chapter 5 shows that the interaction between PCNA and pol η is
dependent on the unstructured CTR of pol η. With all of these complicated mechanisms
to carry out in a sequential order, it is likely that PCNA often interacts with its partners in
a transient and dynamic manner. Since unstructured regions of proteins provide a great
deal of flexibility for protein-protein interactions, it is conceivable that many of the
proteins that interact with PCNA contain one or more of these intrinsically disordered
regions. It has been estimated that nearly a third of eukaryotic proteins and a half of
mammalian proteins are partially or fully disordered [309, 310]. These disordered regions
are often involved in multiple interactions with several protein partners [309, 310, 316].
Another example of this is the interaction between sumoylated PCNA and the Srs2
helicase, which contains a large unstructured region that possesses ordered motifs that
bind SUMO and PCNA independently [308]. There are many ways in which PCNA
regulates cellular processes, and it is probable that many more exist. It is likely, though,
that intrinsically disordered proteins play a major role, and future studies will be focused
on the novel modes of interaction that exist through these unstructured gold mines.
207
Figure 6.1. Positions of all PCNA substitutions used for studies. (A) Front view and
(B) side view of the PCNA trimer with individual subunits colored in pink, green, and
blue. The location of each of the eight residues that were mutated in my studies are
indicated in relation to the front and back faces, the IDCL, and the domains of each
subunit of PCNA.
208
Figure 6.2. Close-up of positions of all PCNA substitutions used for studies. Close-
up view of a PCNA monomer with the eight individual residues that were mutated in my
studies indicated in relation to the front and back faces, the IDCL, and the domains of
PCNA.
209
Figure 6.3. Potential positions of the ubiquitin moiety on PCNA and possible roles
of these positions during TLS. PCNA is shown in purple, ubiquitin is shown in yellow,
and the polymerase is shown in green. The possible roles of each position during TLS
and the location of the E113G substitution are indicated.
210
Figure 6.4. SAXS analysis of the E113G mutant Ub-PCNA protein. (A) Comparison
of SAXS curves of the wild-type Ub-PCNA and the E113G mutant Ub-PCNA proteins.
(B) Close-up of the comparison of SAXS curves of the wild-type Ub-PCNA and the
E113G mutant Ub-PCNA proteins at low angles. (Data from Susan Tsutakawa.)
211
Figure 6.5. Non-denaturing gel electrophoresis of wild-type PCNA. Trimeric,
hexameric, and higher-order oligomeric states of wild-type PCNA are shown at
increasing protein concentrations.
212
Figure 6.6. B-factors of the PCNA interface. The dynamics of each backbone atom in
wild-type PCNA is represented according to size. The image was rendered using B-
factor putty as implemented in pymol. Domain A is shown in blue and domain B is
shown in red. The β-strands that constitute the interface and the position of loop J are
indicated.
213
REFERENCES
1. Devlin, T.M., ed. Textbook of Biochemistry: with Clinical Correlations. 6 ed. 2006, John Wiley & Sons, Inc.: Hoboken, NJ.
2. Matsunaga, T., K. Hieda, and O. Nikaido, WAVELENGTH DEPENDENT FORMATION OF THYMINE DIMERS AND (6-4)PHOTOPRODUCTS IN DNA BY MONOCHROMATIC ULTRAVIOLET-LIGHT RANGING FROM 150 TO 365 NM. Photochemistry and Photobiology, 1991. 54(3): p. 403-410.
3. Karran, P., DNA double strand break repair in mammalian cells. Current Opinion in Genetics & Development, 2000. 10(2): p. 144-150.
4. Beckman, K.B. and B.N. Ames, Oxidative decay of DNA. Journal of Biological Chemistry, 1997. 272(32): p. 19633-19636.
5. Helbock, H.J., et al., DNA oxidation matters: The HPLC-electrochemical detection assay of 8-oxo-deoxyguanosine and 8-oxo-guanine. Proceedings of the National Academy of Sciences of the United States of America, 1998. 95(1): p. 288-293.
6. Lindahl, T. and D.E. Barnes, Repair of endogenous DNA damage. Cold Spring Harbor Symposia on Quantitative Biology, 2000. 65: p. 127-133.
7. Lindahl, T., INSTABILITY AND DECAY OF THE PRIMARY STRUCTURE OF DNA. Nature, 1993. 362(6422): p. 709-715.
8. Lucas-Lledo, J.I. and M. Lynch, Evolution of Mutation Rates: Phylogenomic Analysis of the Photolyase/Cryptochrome Family. Molecular Biology and Evolution, 2009. 26(5): p. 1143-1153.
9. Parikh, S.S., et al., Base excision repair initiation revealed by crystal structures and binding kinetics of human uracil-DNA glycosylase with DNA. Embo Journal, 1998. 17(17): p. 5214-5226.
10. Guillet, M. and S. Boiteux, Origin of endogenous DNA abasic sites in Saccharomyces cerevisiae. Molecular and Cellular Biology, 2003. 23(22): p. 8386-8394.
11. Wilson, D.M. and D. Barsky, The major human abasic endonuclease: formation, consequences and repair of abasic lesions in DNA. Mutation Research-DNA Repair, 2001. 485(4): p. 283-307.
12. Wilson, S.H., Mammalian base excision repair and DNA polymerase beta. Mutation Research-DNA Repair, 1998. 407(3): p. 203-215.
13. Caldecott, K.W., XRCC1 and DNA strand break repair. DNA Repair, 2003. 2(9): p. 955-969.
14. Sancar, A., DNA excision repair. Annual Review of Biochemistry, 1996. 65: p. 43-81.
214
15. Fousteri, M. and L.H. Mullenders, Transcription-coupled nucleotide excision repair in mammalian cells: molecular mechanisms and biological effects. Cell Research, 2008. 18(1): p. 73-84.
16. Downs, J.A., Chromatin structure and DNA double-strand break responses in cancer progression and therapy. Oncogene, 2007. 26(56): p. 7765-7772.
17. Ward, J.F., THE YIELD OF DNA DOUBLE-STRAND BREAKS PRODUCED INTRACELLULARLY BY IONIZING-RADIATION - A REVIEW. International Journal of Radiation Biology, 1990. 57(6): p. 1141-1150.
18. Song, X.Y., et al., Phosphorylation of the SQ H2A.X motif is required for proper meiosis and mitosis in Tetrahymena thermophila. Molecular and Cellular Biology, 2007. 27(7): p. 2648-2660.
19. van Attikum, H. and S.M. Gasser, The histone code at DNA breaks: A guide to repair? Nature Reviews Molecular Cell Biology, 2005. 6(10): p. 757-765.
20. Symington, L.S., Role of RAD52 epistasis group genes in homologous recombination and double-strand break repair. Microbiology and Molecular Biology Reviews, 2002. 66(4): p. 630-+.
21. Espejel, S., et al., Mammalian Ku86 mediates chromosomal fusions and apoptosis caused by critically short telomeres. Embo Journal, 2002. 21(9): p. 2207-2219.
22. Iyer, R.R., et al., DNA mismatch repair: Functions and mechanisms. Chemical Reviews, 2006. 106(2): p. 302-323.
23. Hsieh, P. and K. Yamane, DNA mismatch repair: Molecular mechanism, cancer, and ageing. Mechanisms of Ageing and Development, 2008. 129(7-8): p. 391-407.
24. Kolodner, R.D. and G.T. Marsischky, Eukaryotic DNA mismatch repair. Current Opinion in Genetics & Development, 1999. 9(1): p. 89-96.
25. Modrich, P. and R. Lahue, Mismatch repair in replication fidelity, genetic recombination, and cancer biology. Annual Review of Biochemistry, 1996. 65: p. 101-133.
26. Lynch, H.T. and A. de la Chapelle, Genetic susceptibility to non-polyposis colorectal cancer. Journal of Medical Genetics, 1999. 36(11): p. 801-818.
27. Ban, C. and W. Yang, Crystal structure and ATPase activity of MutL: Implications for DNA repair and mutagenesis. Cell, 1998. 95(4): p. 541-552.
28. Junop, M.S., et al., Composite active site of an ABC ATPase: MutS uses ATP to verify mismatch recognition and authorize DNA repair. Molecular Cell, 2001. 7(1): p. 1-12.
29. Dao, V. and P. Modrich, Mismatch-, MutS-, MutL-, and helicase II-dependent unwinding from the single-strand break of an incised heteroduplex. Journal of Biological Chemistry, 1998. 273(15): p. 9202-9207.
215
30. Li, F., et al., Escherichia coli mismatch repair protein MutL interacts with the clamp loader subunits of DNA polymerase III. Mutation Research-Fundamental and Molecular Mechanisms of Mutagenesis, 2008. 637(1-2): p. 101-110.
31. Lopez de Saro, F.J., et al., The beta sliding clamp binds to multiple sites within MutL and MutS. The Journal of biological chemistry, 2006. 281(20): p. 14340-9.
32. Palombo, F., et al., hMutS beta, a heterodimer of hMSH2 and hMSH3, binds to insertion/deletion leaps in DNA. Current Biology, 1996. 6(9): p. 1181-1184.
33. Drummond, J.T., et al., ISOLATION OF AN HMSH2-P160 HETERODIMER THAT RESTORES DNA MISMATCH REPAIR TO TUMOR-CELLS. Science, 1995. 268(5219): p. 1909-1912.
34. Acharya, S., et al., hMSH2 forms specific mispair-binding complexes with hMSH3 and hMSH6. Proceedings of the National Academy of Sciences of the United States of America, 1996. 93(24): p. 13629-13634.
35. Flores-Rozas, H. and R.D. Kolodner, The Saccharomyces cerevisiae MLH3 gene functions in MSH3-dependent suppression of frameshift mutations. Proceedings of the National Academy of Sciences of the United States of America, 1998. 95(21): p. 12404-12409.
36. Harfe, B.D., B.K. Minesinger, and S. Jinks-Robertson, Discrete in vivo roles for the MutL homologs Mlh2p and Mlh3p in the removal of frameshift intermediates in budding yeast. Current Biology, 2000. 10(3): p. 145-148.
37. Habraken, Y., et al., Enhancement of MSH2-MSH3-mediated mismatch recognition by the yeast MLH1-PMS1 complex. Current Biology, 1997. 7(10): p. 790-793.
38. Prolla, T.A., D.M. Christie, and R.M. Liskay, DUAL REQUIREMENT IN YEAST DNA MISMATCH REPAIR FOR MLH1 AND PMS1, 2 HOMOLOGS OF THE BACTERIAL MUTL GENE. Molecular and Cellular Biology, 1994. 14(1): p. 407-415.
39. Genschel, J., L.R. Bazemore, and P. Modrich, Human exonuclease I is required for 5 ' and 3 ' mismatch repair. Journal of Biological Chemistry, 2002. 277(15): p. 13302-13311.
40. Schmutte, C., et al., Human exonuclease I interacts with the mismatch repair protein hMSH21. Cancer Research, 1998. 58(20): p. 4537-4542.
41. Tishkoff, D.X., et al., Identification of a human gene encoding a homologue of Saccharomyces cerevisiae EXO1, an exonuclease implicated in mismatch repair and recombination. Cancer Research, 1998. 58(22): p. 5027-5031.
42. Tishkoff, D.X., et al., Identification and characterization of Saccharomyces cerevisiae EXO1, a gene encoding an exonuclease that interacts with MSH2. Proceedings of the National Academy of Sciences of the United States of America, 1997. 94(14): p. 7487-7492.
216
43. Longley, M.J., A.J. Pierce, and P. Modrich, DNA polymerase delta is required for human mismatch repair in vitro. Journal of Biological Chemistry, 1997. 272(16): p. 10917-10921.
44. Lin, Y.L., et al., The evolutionarily conserved zinc finger motif in the largest subunit of human replication protein a is required for DNA replication and mismatch repair but not for nucleotide excision repair. Journal of Biological Chemistry, 1998. 273(3): p. 1453-1461.
45. Ramilo, C., et al., Partial reconstitution of human DNA mismatch repair in vitro: Characterization of the role of human replication protein A. Molecular and Cellular Biology, 2002. 22(7): p. 2037-2046.
46. Kadyrov, F.A., et al., Endonucleolytic function of MutL alpha in human mismatch repair. Cell, 2006. 126(2): p. 297-308.
47. Dzantiev, L., et al., Defined human system that supports bidirectional mismatch-provoked excision. Molecular Cell, 2004. 15(1): p. 31-41.
48. Yuan, F.H., et al., Evidence for involvement of HMGB1 protein in human DNA mismatch repair. Journal of Biological Chemistry, 2004. 279(20): p. 20935-20940.
49. Zhang, Y.B., et al., Reconstitution of 5 '-directed human mismatch repair in a purified system. Cell, 2005. 122(5): p. 693-705.
50. Pluciennik, A., et al., PCNA function in the activation and strand direction of MutL alpha endonuclease in mismatch repair. Proceedings of the National Academy of Sciences of the United States of America, 2010. 107(37): p. 16066-16071.
51. Umar, A., et al., Requirement for PCNA in DNA mismatch repair at a step preceding DNA resynthesis. Cell, 1996. 87(1): p. 65-73.
52. Gu, L.Y., et al., ATP-dependent interaction of human mismatch repair proteins and dual role of PCNA in mismatch repair. Nucleic Acids Research, 1998. 26(5): p. 1173-1178.
53. Clark, A.B., et al., Functional interaction of proliferating cell nuclear antigen with MSH2-MSH6 and MSH2-MSH3 complexes. Journal of Biological Chemistry, 2000. 275(47): p. 36498-36501.
54. Flores-Rozas, H., D. Clark, and R.D. Kolodner, Proliferating cell nuclear antigen and Msh2p-Msh6p interact to form an active mispair recognition complex. Nature Genetics, 2000. 26(3): p. 375-378.
55. Kleczkowska, H.E., et al., hMSH3 and hMSH6 interact with PCNA and colocalize with it to replication foci. Genes & Development, 2001. 15(6): p. 724-736.
56. Lau, P.J. and R.D. Kolodner, Transfer of the MSH2 center dot MSH6 complex from proliferating cell nuclear antigen to mispaired bases in DNA. Journal of Biological Chemistry, 2003. 278(1): p. 14-17.
217
57. Shell, S.S., C.D. Putnam, and R.D. Kolodner, The N terminus of Saccharomyces cerevisiale Msh6 is an unstructured tether to PCNA. Molecular Cell, 2007. 26(4): p. 565-578.
58. Loeb, K.R. and L.A. Loeb, Significance of multiple mutations in cancer. Carcinogenesis, 2000. 21(3): p. 379-385.
59. Loeb, L.A., K.R. Loeb, and J.P. Anderson, Multiple mutations and cancer. Proceedings of the National Academy of Sciences of the United States of America, 2003. 100(3): p. 776-781.
60. Prakash, S., R.E. Johnson, and L. Prakash, Eukaryotic translesion synthesis DNA polymerases: Specificity of structure and function. Annual Review of Biochemistry, 2005. 74: p. 317-353.
61. Guo, C., et al., Y-family DNA polymerases in mammalian cells. Cellular and Molecular Life Sciences, 2009. 66(14): p. 2363-2381.
62. Braithwaite, D.K. and J. Ito, COMPILATION, ALIGNMENT, AND PHYLOGENETIC-RELATIONSHIPS OF DNA-POLYMERASES. Nucleic Acids Research, 1993. 21(4): p. 787-802.
63. Ohmori, H., et al., The Y-family of DNA polymerases. Molecular Cell, 2001. 8(1): p. 7-8.
64. Steitz, T.A., DNA polymerases: Structural diversity and common mechanisms. Journal of Biological Chemistry, 1999. 274(25): p. 17395-17398.
65. Doublie, S., M.R. Sawaya, and T. Ellenberger, An open and closed case for all polymerases. Structure with Folding & Design, 1999. 7(2): p. R31-R35.
66. Rothwell, P.J. and G. Waksman, Structure and mechanism of DNA polymerases. Fibrous Proteins: Muscle and Molecular Motors, 2005. 71: p. 401-+.
67. Brautigam, C.A. and T.A. Steitz, Structural and functional insights provided by crystal structures of DNA polymerases and their substrate complexes. Current Opinion in Structural Biology, 1998. 8(1): p. 54-63.
68. Garg, P. and P.M.J. Burgers, DNA polymerases that propagate the eukaryotic DNA replication fork. Critical Reviews in Biochemistry and Molecular Biology, 2005. 40(2): p. 115-128.
69. Echols, H. and M.F. Goodman, FIDELITY MECHANISMS IN DNA-REPLICATION. Annual Review of Biochemistry, 1991. 60: p. 477-511.
70. Goodman, M.F. and D.K. Fygenson, DNA polymerase fidelity: From genetics toward a biochemical understanding. Genetics, 1998. 148(4): p. 1475-1482.
71. Kunkel, T.A. and R. Bebenek, DNA replication fidelity. Annual Review of Biochemistry, 2000. 69: p. 497-529.
72. Pursell, Z.F., et al., Yeast DNA polymerase epsilon participates in leading-strand DNA replication. Science, 2007. 317(5834): p. 127-130.
218
73. McCulloch, S.D. and T.A. Kunkel, The fidelity of DNA synthesis by eukaryotic replicative and translesion synthesis polymerases. Cell Research, 2008. 18(1): p. 148-161.
74. Swan, M.K., et al., Structural basis of high-fidelity DNA synthesis by yeast DNA polymerase delta. Nature Structural & Molecular Biology, 2009. 16(9): p. 979-U107.
75. Dieckman, L.M., et al., Pre-Steady State Kinetic Studies of the Fidelity of Nucleotide Incorporation by Yeast DNA Polymerase delta. Biochemistry, 2010. 49(34): p. 7344-7350.
76. Washington, M.T., et al., Fidelity and processivity of Saccharomyces cerevisiae DNA polymerase eta. Journal of Biological Chemistry, 1999. 274(52): p. 36835-36838.
77. Washington, M.T., et al., Accuracy of thymine-thymine dimer bypass by Saccharomyces cerevisiae DNA polymerase eta. Proceedings of the National Academy of Sciences of the United States of America, 2000. 97(7): p. 3094-3099.
78. Ng, L., et al., ENZYMOLOGIC MECHANISM OF CALF THYMUS DNA POLYMERASE-DELTA. Journal of Biological Chemistry, 1991. 266(18): p. 11699-11704.
79. Johnson, K.A., Rapid Quench Kinetic-Analysis of Polymerases, Adenosine-Triphosphatases, and Enzyme Intermediates. Enzyme Kinetics and Mechanism, Pt D, 1995. 249: p. 38-61.
80. Johnson, A. and M. O'Donnell, Cellular DNA replicases: Components and dynamics at the replication fork, in Annual Review of Biochemistry. 2005, Annual Reviews: Palo Alto. p. 283-315.
81. Podust, V.N., et al., Reconstitution of human DNA polymerase delta using recombinant baculoviruses - The p12 subunit potentiates DNA polymerizing activity of the four-subunit enzyme. Journal of Biological Chemistry, 2002. 277(6): p. 3894-3901.
82. Fortune, J.M., et al., RPA and PCNA suppress formation of large deletion errors by yeast DNA polymerase delta. Nucleic Acids Research, 2006. 34(16): p. 4335-4341.
83. Zhou, J.Q., et al., Purification and characterization of the catalytic subunit of human DNA polymerase delta expressed in baculovirus-infected insect cells. Journal of Biological Chemistry, 1996. 271(47): p. 29740-29745.
84. Hindges, R. and U. Hubscher, DNA polymerase delta, an essential enzyme for DNA transactions. Biological Chemistry, 1997. 378(5): p. 345-362.
85. Gerik, K.J., et al., Characterization of the two small subunits of Saccharomyces cerevisiae DNA polymerase delta. Journal of Biological Chemistry, 1998. 273(31): p. 19747-19755.
219
86. Hanna, M., et al., Pol32 is required for Pol zeta-dependent translesion synthesis and prevents double-strand breaks at the replication fork. Mutation Research-Fundamental and Molecular Mechanisms of Mutagenesis, 2007. 625(1-2): p. 164-176.
87. Fortune, J.M., et al., Saccharomyces cerevisiae DNA polymerase delta - High fidelity for base substitutions but lower fidelity for single- and multi-base deletions. Journal of Biological Chemistry, 2005. 280(33): p. 29980-29987.
88. Johnson, R.E., S. Prakash, and L. Prakash, Efficient bypass of a thymine-thymine dimer by yeast DNA polymerase, Pol eta. Science, 1999. 283(5404): p. 1001-1004.
89. Haracska, L., et al., Efficient and accurate replication in the presence of 7,8-dihydro-8-oxoguanine by DNA polymerase eta. Nature Genetics, 2000. 25(4): p. 458-461.
90. Washington, M.T., et al., Efficient and error-free replication past a minor-groove N-2-guanine adduct by the sequential action of yeast Rev1 and DNA polymerase zeta. Molecular and Cellular Biology, 2004. 24(16): p. 6900-6906.
91. Wolfle, W.T., et al., Human DNA polymerase iota promotes replication through a ring-closed minor-groove adduct that adopts a syn conformation in DNA. Molecular and Cellular Biology, 2005. 25(19): p. 8748-8754.
92. Pence, M.G., et al., Lesion Bypass of N-2-Ethylguanine by Human DNA Polymerase i. Journal of Biological Chemistry, 2009. 284(3): p. 1732-1740.
93. Ogi, T., et al., Pol kappa protects mammalian cells against the lethal and mutagenic effects of benzo a pyrene. Proceedings of the National Academy of Sciences of the United States of America, 2002. 99(24): p. 15548-15553.
94. Takenaka, K., et al., Involvement of vertebrate Pol kappa in translesion DNA synthesis across DNA monoalkylation damage. Journal of Biological Chemistry, 2006. 281(4): p. 2000-2004.
95. Avkin, S., et al., Quantitative analysis of translesion DNA synthesis across a benzo[a] pyrene-guanine adduct in mammalian cells - The role of DNA polymerase kappa. Journal of Biological Chemistry, 2004. 279(51): p. 53298-53305.
96. Pryor, J.M. and M.T. Washington, Pre-steady state kinetic studies show that an abasic site is a cognate lesion for the yeast Rev1 protein. DNA Repair, 2011. 10(11): p. 1138-1144.
97. Washington, M.T., et al., Efficient and error-free replication past a minor-groove DNA adduct by the sequential action of human DNA polymerases iota and kappa. Molecular and Cellular Biology, 2004. 24(13): p. 5687-5693.
98. Nair, D.T., et al., Protein-template-directed synthesis across an acrolein-derived DNA adduct by yeast rev1 DNA polymerase. Structure, 2008. 16(2): p. 239-245.
220
99. Choi, J.Y. and F.P. Guengerich, Kinetic analysis of translesion synthesis opposite bulky N-2- and O-6-alkylguanine DNA adducts by human DNA polymerase REV1. Journal of Biological Chemistry, 2008. 283(35): p. 23645-23655.
100. Zhang, Y.B., et al., Response of human REV1 to different DNA damage: preferential dCMP insertion opposite the lesion. Nucleic Acids Research, 2002. 30(7): p. 1630-1638.
101. Ohmori, H., et al., SEPARATE ROLES OF STRUCTURED AND UNSTRUCTURED REGIONS OF Y-FAMILY DNA POLYMERASES, in Advances in Protein Chemistry and Structural Biology, Vol 78, A. McPherson, Editor. 2009. p. 99-146.
102. Washington, M.T., et al., Mismatch extension ability of yeast and human DNA polymerase eta. Journal of Biological Chemistry, 2001. 276(3): p. 2263-2266.
103. Haracska, L., S. Prakash, and L. Prakash, Replication past O-6-methylguanine by yeast and human DNA polymerase eta. Molecular and Cellular Biology, 2000. 20(21): p. 8001-8007.
104. Johnson, R.E., et al., hRAD30 mutations in the variant form of xeroderma pigmentosum. Science, 1999. 285(5425): p. 263-265.
105. Masutani, C., et al., The XPV (xeroderma pigmentosum variant) gene encodes human DNA polymerase eta. Nature, 1999. 399(6737): p. 700-704.
106. Johnson, R.E., et al., Fidelity of human DNA polymerase eta. Journal of Biological Chemistry, 2000. 275(11): p. 7447-7450.
107. Washington, M.T., L. Prakash, and S. Prakash, Mechanism of nucleotide incorporation opposite a thymine-thymine dimer by yeast DNA polymerase eta. Proceedings of the National Academy of Sciences of the United States of America, 2003. 100(21): p. 12093-12098.
108. Carlson, K.D. and A.T. Washington, Mechanism of efficient and accurate nucleotide incorporation opposite 7,8-dihydro-8-oxoguanine by Saccharomyces cerevisiae DNA polymerase eta. Molecular and Cellular Biology, 2005. 25(6): p. 2169-2176.
109. Haracska, L., et al., Inefficient bypass of an abasic site by DNA polymerase eta. Journal of Biological Chemistry, 2001. 276(9): p. 6861-6866.
110. Silverstein, T.D., et al., Structural basis for the suppression of skin cancers by DNA polymerase eta. Nature, 2010. 465(7301): p. 1039-U96.
111. Ummat, A., et al., Human DNA Polymerase eta Is Pre-Aligned for dNTP Binding and Catalysis. Journal of Molecular Biology, 2012. 415(4): p. 627-634.
112. Biertuempfel, C., et al., Structure and mechanism of human DNA polymerase eta. Nature, 2010. 465(7301): p. 1044-U102.
113. Haracska, L., et al., Interaction with PCNA is essential for yeast DNA polymerase eta function. Molecular Cell, 2001. 8(2): p. 407-415.
221
114. Bomar, M.G., et al., Structure of the ubiquitin-binding zinc finger domain of human DNA Y-polymerase eta. Embo Reports, 2007. 8(3): p. 247-251.
115. Hishiki, A., et al., Structural Basis for Novel Interactions between Human Translesion Synthesis Polymerases and Proliferating Cell Nuclear Antigen. Journal of Biological Chemistry, 2009. 284(16): p. 10552-10560.
116. Li, G.M. and P. Modrich, RESTORATION OF MISMATCH REPAIR TO NUCLEAR EXTRACTS OF H6 COLORECTAL TUMOR-CELLS BY A HETERODIMER OF HUMAN MUTL HOMOLOGS. Proceedings of the National Academy of Sciences of the United States of America, 1995. 92(6): p. 1950-1954.
117. Modrich, P., Strand-specific mismatch repair in mammalian cells. Journal of Biological Chemistry, 1997. 272(40): p. 24727-24730.
118. Papadopoulos, N., et al., MUTATIONS OF GTBP IN GENETICALLY UNSTABLE CELLS. Science, 1995. 268(5219): p. 1915-1917.
119. Fishel, R. and T. Wilson, MutS homologs in mammalian cells. Current Opinion in Genetics & Development, 1997. 7(1): p. 105-113.
120. Alani, E., The Saccharomyces cerevisiae Msh2 and Msh6 proteins form a complex that specifically binds to duplex oligonucleotides containing mismatched DNA base pairs. Molecular and Cellular Biology, 1996. 16(10): p. 5604-5615.
121. Modrich, P., Mechanisms in eukaryotic mismatch repair. Journal of Biological Chemistry, 2006. 281(41): p. 30305-30309.
122. Warren, J.J., et al., Structure of the human MutS alpha DNA lesion recognition complex. Molecular Cell, 2007. 26(4): p. 579-592.
123. Obmolova, G., et al., Crystal structures of mismatch repair protein MutS and its complex with a substrate DNA. Nature, 2000. 407(6805): p. 703-710.
124. Lamers, M.H., et al., The crystal structure of DNA mismatch repair protein MutS binding to a G center dot T mismatch. Nature, 2000. 407(6805): p. 711-717.
125. Kunkel, T.A. and D.A. Erie, DNA mismatch repair, in Annual Review of Biochemistry. 2005. p. 681-710.
126. Lee, S.D., J.A. Surtees, and E. Alani, Saccharomyces cerevisiae MSH2-MSH3 and MSH2-MSH6 complexes display distinct requirements for DNA binding domain I in mismatch recognition. Journal of Molecular Biology, 2007. 366(1): p. 53-66.
127. Drotschmann, K., et al., Asymmetric recognition of DNA local distortion - Structure-based functional studies of eukaryotic Msh2-Msh6. Journal of Biological Chemistry, 2001. 276(49): p. 46225-46229.
128. Cristovao, M., et al., Single-molecule multiparameter fluorescence spectroscopy reveals directional MutS binding to mismatched bases in DNA. Nucleic Acids Research, 2012. 40(12): p. 5448-5464.
222
129. Masih, P.J., D. Kunnev, and T. Melendy, Mismatch repair proteins are recruited to replicating DNA through interaction with proliferating cell nuclear antigen (PCNA). Nucleic Acids Research, 2008. 36(1): p. 67-75.
130. Clark, A.B., et al., Multiple functions for the N-terminal region of Msh6. Nucleic Acids Research, 2007. 35(12): p. 4114-4123.
131. Matton, N., J. Simonetti, and K. Williams, Identification of mismatch repair protein complexes in HeLa nuclear extracts and their interaction with heteroduplex DNA. Journal of Biological Chemistry, 2000. 275(23): p. 17808-17813.
132. Edelbrock, M., et al., Recognition and binding of mismatch repair proteins at an oncogenic hot spot. Bmc Molecular Biology, 2005. 6.
133. Bellacosa, A., Functional interactions and signaling properties of mammalian DNA mismatch repair proteins. Cell Death and Differentiation, 2001. 8(11): p. 1076-1092.
134. Duckett, D.R., et al., Human MutS alpha recognizes damaged DNA base pairs containing O-6-methylguanine, O-4-methylthymine, or the cisplatin-d(GpG) adduct. Proceedings of the National Academy of Sciences of the United States of America, 1996. 93(13): p. 6443-6447.
135. Mu, D., et al., Recognition and repair of compound DNA lesions (base damage and mismatch) by human mismatch repair and excision repair systems. Molecular and Cellular Biology, 1997. 17(2): p. 760-769.
136. Colussi, C., et al., The mammalian mismatch repair pathway removes DNA 8-oxodGMP incorporated from the oxidized dNTP pool. Current Biology, 2002. 12(11): p. 912-918.
137. Russo, M.T., et al., Role of MUTYH and MSH2 in the Control of Oxidative DNA Damage, Genetic Instability, and Tumorigenesis. Cancer Research, 2009. 69(10): p. 4372-4379.
138. Mazurek, A., M. Berardini, and R. Fishel, Activation of human MutS homologs by 8-oxo-guanine DNA damage. Journal of Biological Chemistry, 2002. 277(10): p. 8260-8266.
139. Larson, E.D., K. Iams, and J.T. Drummond, Strand-specific processing of 8-oxoguanine by the human mismatch repair pathway: inefficient removal of 8-oxoguanine paired with adenine or cytosine. DNA Repair, 2003. 2(11): p. 1199-1210.
140. Arcangeli, L., et al., Site- and strand-specific mismatch repair of human H-ras genomic DNA in a mammalian cell line. Carcinogenesis, 1997. 18(7): p. 1311-1318.
141. Jiricny, J., The multifaceted mismatch-repair system. Nature Reviews Molecular Cell Biology, 2006. 7(5): p. 335-346.
223
142. Edelbrock, M., Kaliyaperumal, S., Williams, K., Structural, molecular, and cellular functions of MSH2 and MSH6 during DNA mismatch repair, damage signaling and other noncanonical activities. Mutation Research: Fundamental and molecular mechanisms of mutagenesis, 2013.
143. Chahwan, R., et al., AIDing antibody diversity by error-prone mismatch repair. Seminars in Immunology, 2012. 24(4): p. 293-300.
144. Rada, C., et al., Immunoglobulin isotype switching is inhibited and somatic hypermutation perturbed in UNG-deficient mice. Current Biology, 2002. 12(20): p. 1748-1755.
145. Wiesendanger, M., et al., Somatic hypermutation in MutS homologue (MSH)3-, MSH6-, and MSH3/MSH6-deficient mice reveals a role for the MSH2-MSH6 heterodimer in modulating the base substitution pattern. Journal of Experimental Medicine, 2000. 191(3): p. 579-584.
146. Martin, A., et al., Msh2 ATPase activity is essential for somatic hypermutation at A-T basepairs and for efficient class switch recombination. Journal of Experimental Medicine, 2003. 198(8): p. 1171-1178.
147. Zeng, X.M., et al., DNA polymerase eta is an A-T mutator in somatic hypermutation of immunoglobulin variable genes. Nature Immunology, 2001. 2(6): p. 537-541.
148. Faili, A., et al., DNA polymerase eta is involved in hypermutation occurring during immunoglobulin class switch recombination. Journal of Experimental Medicine, 2004. 199(2): p. 265-270.
149. Zeng, X.M., et al., Absence of DNA polymerase eta reveals targeting of C mutations on the nontranscribed strand in immunoglobulin switch regions. Journal of Experimental Medicine, 2004. 199(7): p. 917-924.
150. Wilson, T.M., et al., MSH2-MSH6 stimulates DNA polymerase eta, suggesting a role for A : T mutations in antibody genes. Journal of Experimental Medicine, 2005. 201(4): p. 637-645.
151. Miyachi, K., M.J. Fritzler, and E.M. Tan, AUTOANTIBODY TO A NUCLEAR ANTIGEN IN PROLIFERATING CELLS. Journal of Immunology, 1978. 121(6): p. 2228-2234.
152. Bravo, R. and J.E. Celis, SEARCH FOR DIFFERENTIAL POLYPEPTIDE-SYNTHESIS THROUGHOUT THE CELL-CYCLE OF HELA-CELLS. Journal of Cell Biology, 1980. 84(3): p. 795-802.
153. Bravo, R., et al., Identification of a nuclear polypeptide ("cyclin") whose relative proportion is sensitive to changes in the rate of cell proliferation and to transformation. Progress in clinical and biological research, 1982. 85 Pt A: p. 235-48.
154. Celis, J.E., et al., CYCLIN - A NUCLEAR-PROTEIN WHOSE LEVEL CORRELATES DIRECTLY WITH THE PROLIFERATIVE STATE OF NORMAL AS WELL AS TRANSFORMED-CELLS. Leukemia Research, 1984. 8(2): p. 143-157.
224
155. Moldovan, G.L., B. Pfander, and S. Jentsch, PCNA, the maestro of the replication fork. Cell, 2007. 129: p. 665-679.
156. Krishna, T.S.R., et al., Crystal structure of the eukaryotic DNA polymerase processivity factor PCNA. Cell, 1994. 79(7): p. 1233-1243.
157. Georgescu, R.E., et al., Structure of a sliding clamp on DNA. Cell, 2008. 132(1): p. 43-54.
158. McNally, R., et al., Analysis of the role of PCNA-DNA contacts during clamp loading. Bmc Structural Biology, 2010. 10.
159. Kochaniak, A.B., et al., Proliferating Cell Nuclear Antigen Uses Two Distinct Modes to Move along DNA. Journal of Biological Chemistry, 2009. 284(26): p. 17700-17710.
160. Freudenthal, B.D., et al., Structure of monoubiquitinated PCNA and implications for translesion synthesis and DNA polymerase exchange. Nature Structural & Molecular Biology, 2010. 17(4): p. 479-U123.
161. Freudenthal, B.D., et al., Crystal structure of SUMO-modified proliferating cell nuclear antigen. J Mol Biol, 2011. 406(1): p. 9-17.
162. Gomes, X.V. and P.M.J. Burgers, ATP utilization by yeast replication factor C I. ATP-mediated interaction with DNA and with proliferating cell nuclear antigen. Journal of Biological Chemistry, 2001. 276(37): p. 34768-34775.
163. Langston, L.D. and M. O'Donnell, DNA Polymerase delta Is Highly Processive with Proliferating Cell Nuclear Antigen and Undergoes Collision Release upon Completing DNA. Journal of Biological Chemistry, 2008. 283(43): p. 29522-29531.
164. de Saro, F.J.L., Regulation of Interactions with Sliding Clamps During DNA Replication and Repair. Current Genomics, 2009. 10(3): p. 206-215.
165. Chilkova, O., et al., The eukaryotic leading and lagging strand DNA polymerases are loaded onto primer-ends via separate mechanisms but have comparable processivity in the presence of PCNA. Nucleic Acids Research, 2007. 35(19): p. 6588-6597.
166. Einolf, H.J. and F.P. Guengerich, Kinetic analysis of nucleotide incorporation by mammalian DNA polymerase delta. Journal of Biological Chemistry, 2000. 275(21): p. 16316-16322.
167. Chapados, B.R., et al., Structural basis for FEN-1 substrate specificity and PCNA-mediated activation in DNA replication and repair. Cell, 2004. 116(1): p. 39-50.
168. Pascal, J.M., et al., Human DNA ligase I completely encircles and partially unwinds nicked DNA. Nature, 2004. 432(7016): p. 473-478.
169. Lee, S.D. and E. Alani, Analysis of interactions between mismatch repair initiation factors and the replication processivity factor PCNA. Journal of Molecular Biology, 2006. 355(2): p. 175-184.
225
170. Montecucco, A., et al., DNA ligase I is recruited to sites of DNA replication by an interaction with proliferating cell nuclear antigen: identification of a common targeting mechanism for the assembly of replication factories. Embo Journal, 1998. 17(13): p. 3786-3795.
171. Iyer, R.R., et al., The MutS alpha-proliferating cell nuclear antigen interaction in human DNA mismatch repair. Journal of Biological Chemistry, 2008. 283(19): p. 13310-13319.
172. Rousseau, D., et al., Growth inhibition by CDK-cyclin and PCNA binding domains of p21 occurs by distinct mechanisms and is regulated by ubiquitin-proteasome pathway. Oncogene, 1999. 18(30): p. 4313-4325.
173. Waga, S., et al., THE P21 INHIBITOR OF CYCLIN-DEPENDENT KINASES CONTROLS DNA-REPLICATION BY INTERACTION WITH PCNA. Nature, 1994. 369(6481): p. 574-578.
174. Ducoux, M., et al., Mediation of proliferating cell nuclear antigen (PCNA)-dependent DNA replication through a conserved p21(Cip1)-like PCNA-binding motif present in the third subunit of human DNA polymerase delta. Journal of Biological Chemistry, 2001. 276(52): p. 49258-49266.
175. Warbrick, E., et al., Homologous regions of Fen1 and p21(Cip1) compete for binding to the same site on PCNA: A potential mechanism to co-ordinate DNA replication and repair. Oncogene, 1997. 14(19): p. 2313-2321.
176. Oku, T., et al., Functional sites of human PCNA which interact with p21 (Cip1/Waf1), DNA polymerase delta and replication factor C. Genes to Cells, 1998. 3(6): p. 357-369.
177. Bergink, S. and S. Jentsch, Principles of ubiquitin and SUMO modifications in DNA repair. Nature, 2009. 458(7237): p. 461-467.
178. Shaheen, M., I. Shanmugam, and R. Hromas, The Role of PCNA Posttranslational Modifications in Translesion Synthesis. J Nucleic Acids, 2010. 2010.
179. Ulrich, H.D. and H. Walden, Ubiquitin signalling in DNA replication and repair. Nature Reviews Molecular Cell Biology, 2010. 11(7): p. 479-489.
180. Watts, F.Z., Sumoylation of PCNA: Wrestling with recombination at stalled replication forks. DNA Repair, 2006. 5(3): p. 399-403.
181. Naryzhny, S.N. and H. Lee, The post-translational modifications of proliferating cell nuclear antigen - Acetylation, not phosphorylation, plays an important role in the regulation of its function. Journal of Biological Chemistry, 2004. 279(19): p. 20194-20199.
182. Prosperi, E., et al., PROLIFERATING CELL NUCLEAR ANTIGEN BOUND TO DNA-SYNTHESIS SITES - PHOSPHORYLATION AND ASSOCIATION WITH CYCLIN-D1 AND CYCLIN-A. Experimental Cell Research, 1994. 215(2): p. 257-262.
183. Hoege, C., et al., RAD6-dependent DNA repair is linked to modification of PCNA by ubiquitin and SUMO. Nature, 2002. 419(6903): p. 135-141.
226
184. Stelter, P. and H.D. Ulrich, Control of spontaneous and damage-induced mutagenesis by SUMO and ubiquitin conjugation. Nature, 2003. 425(6954): p. 188-191.
185. Bienko, M., et al., Ubiquitin-binding domains in Y-family polymerases regulate translesion synthesis. Science, 2005. 310(5755): p. 1821-1824.
186. Guo, C.X., et al., Ubiquitin-binding motifs in REV1 protein are required for its role in the tolerance of DNA damage. Molecular and Cellular Biology, 2006. 26(23): p. 8892-8900.
187. Huttner, D. and H.D. Ulrich, Cooperation of replication protein A with the ubiquitin ligase Rad18 in DNA damage bypass. Cell Cycle, 2008. 7(23): p. 3629-3633.
188. Davies, A.A., et al., Activation of ubiquitin-dependent DNA damage bypass is mediated by replication protein A. Molecular Cell, 2008. 29(5): p. 625-636.
189. Kannouche, P.L., J. Wing, and A.R. Lehmann, Interaction of human DNA polymerase eta with monoubiquitinated PCNA: A possible mechanism for the polymerase switch in response to DNA damage. Molecular Cell, 2004. 14(4): p. 491-500.
190. Haracska, L., et al., Ubiquitylation of yeast proliferating cell nuclear antigen and its implications for translesion DNA synthesis. Proceedings of the National Academy of Sciences of the United States of America, 2006. 103(17): p. 6477-6482.
191. Garg, P. and P.M. Burgers, Ubiquitinated proliferating cell nuclear antigen activates translesion DNA polymerases eta and REV1. Proceedings of the National Academy of Sciences of the United States of America, 2005. 102(51): p. 18361-18366.
192. Tsutakawa, S.E., et al., Solution X-ray scattering combined with computational modeling reveals multiple conformations of covalently bound ubiquitin on PCNA. Proc Natl Acad Sci U S A, 2011. 108(43): p. 17672-7.
193. Ulrich, H.D. and S. Jentsch, Two RING finger proteins mediate cooperation between ubiquitin-conjugating enzymes in DNA repair. Embo Journal, 2000. 19(13): p. 3388-3397.
194. Plosky, B.S., et al., Controlling the subcellular localization of DNA polymerases iota and eta via interactions with ubiquitin. Embo Journal, 2006. 25(12): p. 2847-2855.
195. Papouli, E., et al., Crosstalk between SUMO and ubiquitin on PCNA is mediated by recruitment of the helicase Srs2p. Molecular Cell, 2005. 19(1): p. 123-133.
196. Parker, J.L., et al., SUMO modification of PCNA is controlled by DNA. Embo Journal, 2008. 27(18): p. 2422-2431.
197. Haracska, L., et al., Opposing effects of ubiquitin conjugation and SUMO modification of PCNA on replicational bypass of DNA lesions in Saccharomyces cerevisiae. Molecular and Cellular Biology, 2004. 24(10): p. 4267-4274.
227
198. Hingorani, M.M. and M. O'Donnell, Sliding clamps: A (tail)ored fit. Current Biology, 2000. 10(1): p. R25-R29.
199. Maga, G. and U. Hubscher, Proliferating cell nuclear antigen (PCNA): a dancer with many partners. Journal of Cell Science, 2003. 116(15): p. 3051-3060.
200. Tsurimoto, T., PCNA binding proteins. Frontiers in Bioscience, 1999. 4(CITED APRIL 7, 2000): p. d849-858.
201. Bruning, J.B. and Y. Shamoo, Structural and thermodynamic analysis of human PCNA with peptides derived from DNA polymerase-delta p66 subunit and flap endonuclease-1. Structure, 2004. 12(12): p. 2209-2219.
202. Prakash, S. and L. Prakash, Translesion DNA synthesis in eukaryotes: A one- or two-polymerase affair. Genes & Development, 2002. 16(15): p. 1872-1883.
203. Washington, M.T., et al., Variations on a theme: Eukaryotic Y-family DNA polymerases. Biochim Biophys Acta, 2009.
204. Vijayakumar, S., et al., The C-terminal domain of yeast PCNA is required for physical and functional interactions with Cdc9 DNA ligase. Nucleic Acids Research, 2007. 35(5): p. 1624-1637.
205. Floresrozas, H., et al., CDK-INTERACTING PROTEIN-1 DIRECTLY BINDS WITH PROLIFERATING CELL NUCLEAR ANTIGEN AND INHIBITS DNA-REPLICATION CATALYZED BY THE DNA-POLYMERASE-DELTA HOLOENZYME. Proceedings of the National Academy of Sciences of the United States of America, 1994. 91(18): p. 8655-8659.
206. Gibbs, E., et al., The influence of the proliferating cell nuclear antigen-interacting domain of p21(CIP1) on DNA synthesis catalyzed by the human and Saccharomyces cerevisiae polymerase delta holoenzymes. Journal of Biological Chemistry, 1997. 272(4): p. 2373-2381.
207. Gulbis, J.M., et al., Structure of the C-terminal region of p21(WAF1/CIP1) complexed with human PCNA. Cell, 1996. 87(2): p. 297-306.
208. Sakurai, S., et al., Structural basis for recruitment of human flap endonuclease 1 to PCNA. Embo Journal, 2005. 24(4): p. 683-693.
209. Bowman, G.D., M. O'Donnell, and J. Kuriyan, Structural analysis of a eukaryotic sliding DNA clamp-clamp loader complex. Nature, 2004. 429(6993): p. 724-730.
210. Pascal, J.M., et al., A flexible interface between DNA ligase and PCNA supports conformational switching and efficient ligation of DNA. Molecular Cell, 2006. 24(2): p. 279-291.
211. Williams, G.J., et al., Structure of the heterotrimeric PCNA from Sulfolobus solfataricus. Acta Crystallographica Section F-Structural Biology and Crystallization Communications, 2006. 62: p. 944-948.
228
212. Mayanagi, K., et al., Architecture of the DNA polymerase B-proliferating cell nuclear antigen (PCNA)-DNA ternary complex. Proceedings of the National Academy of Sciences of the United States of America, 2011. 108(5): p. 1845-1849.
213. Henneke, G., S. Koundrioukoff, and U. Hubscher, Phosphorylation of human Fen1 by cyclin-dependent kinase modulates its role in replication fork regulation. Oncogene, 2003. 22(28): p. 4301-4313.
214. Scott, M.T., N. Morrice, and K.L. Ball, Reversible phosphorylation at the C-terminal regulatory domain of p21(Waf1/Cip1) modulates proliferating cell nuclear antigen binding. Journal of Biological Chemistry, 2000. 275(15): p. 11529-11537.
215. Freudenthal, B.D., et al., Structure of monoubiquitinated PCNA and implications for translesion synthesis and DNA polymerase exchange. Nat Struct Mol Biol, 2010. 17(4): p. 479-84.
216. Haracska, L., et al., Physical and functional interactions of human DNA polymerase eta with PCNA. Molecular and Cellular Biology, 2001. 21(21): p. 7199-7206.
217. Haracska, L., et al., Targeting of human DNA polymerase iota to the replication machinery via interaction with PCNA. Proceedings of the National Academy of Sciences of the United States of America, 2001. 98(25): p. 14256-14261.
218. Haracska, L., et al., Stimulation of DNA synthesis activity of human DNA polymerase kappa by PCNA. Molecular and Cellular Biology, 2002. 22(3): p. 784-791.
219. Freudenthal, B.D., et al., Structure of a Mutant Form of Proliferating Cell Nuclear Antigen That Blocks Translesion DNA Synthesis. Biochemistry, 2008. 47(50): p. 13354-13361.
220. Acharya, N., et al., Roles of PCNA-binding and ubiquitin-binding domains in human DNA polymerase eta in translesion DNA synthesis. Proceedings of the National Academy of Sciences of the United States of America, 2008. 105(46): p. 17724-17729.
221. Vidal, A.E., et al., Proliferating cell nuclear antigen-dependent coordination of the biological functions of human DNA polymerase iota. Journal of Biological Chemistry, 2004. 279(46): p. 48360-48368.
222. Guo, C.X., et al., REV1 protein interacts with PCNA: Significance of the REV1 BRCT domain in vitro and in vivo. Molecular Cell, 2006. 23(2): p. 265-271.
223. Wood, A., P. Garg, and P.M.J. Burgers, A ubiquitin-binding motif in the translesion DNA polymerase Rev1 mediates its essential functional interaction with ubiquitinated proliferating cell nuclear antigen in response to DNA damage. Journal of Biological Chemistry, 2007. 282(28): p. 20256-20263.
224. Tissier, A., et al., Co-localization in replication foci and interaction of human Y-family members, DNA polymerase pol eta and REV1 protein. DNA Repair, 2004. 3(11): p. 1503-1514.
229
225. Murakumo, Y., et al., Analyses of ultraviolet-induced focus formation of hREV1 protein. Genes to Cells, 2006. 11(3): p. 193-205.
226. de Groote, F.H., et al., The Rev1 translesion synthesis polymerase has multiple distinct DNA binding modes. DNA Repair, 2011. 10(9): p. 915-925.
227. Bomar, M.G., et al., Unconventional Ubiquitin Recognition by the Ubiquitin-Binding Motif within the Y Family DNA Polymerases iota and Rev1. Molecular Cell, 2010. 37(3): p. 408-417.
228. Burschowsky, D., et al., Structural Analysis of the Conserved Ubiquitin-binding Motifs (UBMs) of the Translesion Polymerase iota in Complex with Ubiquitin. Journal of Biological Chemistry, 2011. 286(2): p. 1364-1373.
229. Tsutakawa, S.E., et al., Solution X-ray scattering combined with computational modeling reveals multiple conformations of covalently bound ubiquitin on PCNA. Proceedings of the National Academy of Sciences of the United States of America, 2011. 108(43): p. 17672-17677.
230. Indiani, C., et al., Translesion DNA polymerases remodel the replisome and alter the speed of the replicative helicase. Proceedings of the National Academy of Sciences of the United States of America, 2009. 106(15): p. 6031-6038.
231. Kuban, W., et al., Mutator phenotype resulting from DNA polymerase IV overproduction in Escherichia coli: Preferential mutagenesis on the lagging strand. Journal of Bacteriology, 2005. 187(19): p. 6862-6866.
232. Zhuang, Z.H., et al., Regulation of polymerase exchange between Pol eta and Pol delta by monoubiquitination of PCNA and the movement of DNA polymerase holoenzyme. Proceedings of the National Academy of Sciences of the United States of America, 2008. 105(14): p. 5361-5366.
233. Masuda, Y., J. Piao, and K. Kamiya, DNA Replication-Coupled PCNA Mono-Ubiquitination and Polymerase Switching in a Human In Vitro System. Journal of Molecular Biology, 2010. 396(3): p. 487-500.
234. Leach, C.A. and W.M. Michael, Ubiquitin/SUMO modification of PCNA promotes replication fork progression in Xenopus laevis egg extracts. Journal of Cell Biology, 2005. 171(6): p. 947-954.
235. Dionne, I., et al., A heterotrimeric PCNA in the hyperthermophilic archaeon Sulfolobus solfataricus. Molecular Cell, 2003. 11(1): p. 275-282.
236. van der Kemp, P.A., et al., PCNA monoubiquitylation and DNA polymerase ubiquitin-binding domain are required to prevent 8-oxoguanine-induced mutagenesis in Saccharomyces cerevisiae. Nucleic Acids Research, 2009. 37(8): p. 2549-2559.
237. Bunting, K.A., S.M. Roe, and L.H. Pearl, Structural basis for recruitment of translesion DNA polymerase Pol IV/DinB to the beta-clamp. Embo Journal, 2003. 22(21): p. 5883-5892.
230
238. Zhang, H.S., P.E.M. Gibbs, and C.W. Lawrence, The Saccharomyces cerevisiae rev6-1 mutation, which inhibits both the lesion bypass and the recombination mode of DNA damage tolerance, is an allele of POL30, encoding proliferating cell nuclear antigen. Genetics, 2006. 173(4): p. 1983-1989.
239. Amin, N.S. and C. Holm, In vivo analysis reveals that the interdomain region of the yeast proliferating cell nuclear antigen is important for DNA replication and DNA repair. Genetics, 1996. 144(2): p. 479-493.
240. Northam, M.R., et al., A novel function of DNA polymerase zeta regulated by PCNA. Embo Journal, 2006. 25(18): p. 4316-4325.
241. Freudenthal, B.D., et al., A charged residue at the subunit interface of PCNA promotes trimer formation by destabilizing alternate subunit interactions. Acta Crystallogr D Biol Crystallogr, 2009. 65(Pt 6): p. 560-6.
242. Lau, P.J., H. Flores-Rozas, and R.D. Kolodner, Isolation and characterization of new proliferating cell nuclear antigen (POL30) mutator mutants that are defective in DNA mismatch repair. Molecular and Cellular Biology, 2002. 22(19): p. 6669-6680.
243. Stone, J.E., et al., Role of proliferating cell nuclear antigen interactions in the mismatch repair-dependent processing of mitotic and meiotic recombination intermediates in yeast. Genetics, 2008. 178(3): p. 1221-1236.
244. Burgers, P.M.J., Polymerase Dynamics at the Eukaryotic DNA Replication Fork. Journal of Biological Chemistry, 2009. 284(7): p. 4041-4045.
245. Kunkel, T.A. and P.M. Burgers, Dividing the workload at a eukaryotic replication fork. Trends in Cell Biology, 2008. 18(11): p. 521-527.
246. McElhinny, S.A.N., et al., Division of labor at the eukaryotic replication fork. Molecular Cell, 2008. 30(2): p. 137-144.
247. Shcherbakova, P.V., K. Bebenek, and T.A. Kunkel, Functions of eukaryotic DNA polymerases. Sci Aging Knowledge Environ, 2003. 2003(8): p. RE3.
248. Johansson, E., J. Majka, and P.M.J. Burgers, Structure of DNA polymerase delta from Saccharomyces cerevisiae. Journal of Biological Chemistry, 2001. 276(47): p. 43824-43828.
249. Jain, R., et al., Structural Insights into Yeast DNA Polymerase delta by Small Angle X-ray Scattering. Journal of Molecular Biology, 2009. 394(3): p. 377-382.
250. Hashimoto, K., et al., Fidelity of DNA polymerase delta holoenzyme from Saccharomyces cerevisiae: The sliding clamp proliferating cell nuclear antigen decreases its fidelity. Biochemistry, 2003. 42(48): p. 14207-14213.
251. Lehmann, A.R., Replication of damaged DNA by translesion synthesis in human cells. Febs Letters, 2005. 579(4): p. 873-876.
252. Lehmann, A.R., Replication of damaged DNA. Cell Cycle, 2003. 2(4): p. 300-2.
231
253. Waters, L.S., et al., Eukaryotic Translesion Polymerases and Their Roles and Regulation in DNA Damage Tolerance. Microbiology and Molecular Biology Reviews, 2009. 73(1): p. 134-+.
254. Sale, J.E., A.R. Lehmann, and R. Woodgate, Y-family DNA polymerases and their role in tolerance of cellular DNA damage. Nature Reviews Molecular Cell Biology, 2012. 13(3): p. 141-152.
255. Trincao, J., et al., Structure of the catalytic core of S-cerevisiae DNA polymerase eta: Implications for translesion DNA synthesis. Molecular Cell, 2001. 8(2): p. 417-426.
256. Washington, M.T., L. Prakash, and S. Prakash, Yeast DNA polymerase eta utilizes an induced-fit mechanism of nucleotide incorporation. Cell, 2001. 107(7): p. 917-927.
257. Naryzhny, S.N., Proliferating cell nuclear antigen: a proteomics view. Cellular and Molecular Life Sciences, 2008. 65(23): p. 3789-3808.
258. Zhuang, Z.H. and Y.X. Ai, Processivity factor of DNA polymerase and its expanding role in normal and translesion DNA synthesis. Biochimica Et Biophysica Acta-Proteins and Proteomics, 2010. 1804(5): p. 1081-1093.
259. Dieckman, L.M., B.D. Freudenthal, and M.T. Washington, PCNA Structure and Function: Insights from Structures of PCNA Complexes and Post-translationally Modified PCNA. Subcell Biochem, 2012. 62: p. 281-99.
260. Freudenthal, B.D., et al., A charged residue at the subunit interface of PCNA promotes trimer formation by destabilizing alternate subunit interactions. Acta Crystallographica Section D-Biological Crystallography, 2009. 65: p. 560-566.
261. Finkelstein, J., et al., Overproduction and analysis of eukaryotic multiprotein complexes in Escherichia coli using a dual-vector strategy. Analytical Biochemistry, 2003. 319(1): p. 78-87.
262. Garg, P., et al., Proliferating cell nuclear antigen promotes translesion synthesis by DNA polymerase zeta. Journal of Biological Chemistry, 2005. 280(25): p. 23446-23450.
263. Hanahan, D. and R.A. Weinberg, Hallmarks of Cancer: The Next Generation. Cell, 2011. 144(5): p. 646-674.
264. O'Roak, B.J., et al., Exome sequencing in sporadic autism spectrum disorders identifies severe de novo mutations. Nature Genetics, 2011. 43(6): p. 585-U125.
265. Xu, B., et al., Strong association of de novo copy number mutations with sporadic schizophrenia. Nature Genetics, 2008. 40(7): p. 880-885.
266. Veltman, J.A. and H.G. Brunner, De novo mutations in human genetic disease. Nature reviews. Genetics, 2012. 13(8): p. 565-75.
232
267. Sharma, N.M., O.V. Kochenova, and P.V. Shcherbakova, The Non-canonical Protein Binding Site at the Monomer-Monomer Interface of Yeast Proliferating Cell Nuclear Antigen (PCNA) Regulates the Rev1-PCNA Interaction and Pol zeta/Rev1-dependent Translesion DNA Synthesis. Journal of Biological Chemistry, 2011. 286(38): p. 33557-33566.
268. Li, G.-M., Mechanisms and functions of DNA mismatch repair. Cell Research, 2008. 18(1): p. 85-98.
269. Kunz, C., Y. Saito, and P. Schaer, DNA Repair in Mammalian Cells. Cellular and Molecular Life Sciences, 2009. 66(6): p. 1021-1038.
270. Grilley, M., et al., ISOLATION AND CHARACTERIZATION OF THE ESCHERICHIA-COLI MUTL GENE-PRODUCT. Journal of Biological Chemistry, 1989. 264(2): p. 1000-1004.
271. Galio, L., C. Bouquet, and P. Brooks, ATP hydrolysis-dependent formation of a dynamic ternary nucleoprotein complex with MutS and MutL. Nucleic Acids Research, 1999. 27(11): p. 2325-2331.
272. Spampinato, C. and P. Modrich, The MutL ATPase is required for mismatch repair. Journal of Biological Chemistry, 2000. 275(13): p. 9863-9869.
273. Schofield, M.J., et al., Interaction of Escherichia coli MutS and MutL at a DNA mismatch. Journal of Biological Chemistry, 2001. 276(30): p. 28291-28299.
274. Acharya, S., et al., The coordinated functions of the E-coli MutS and MutL proteins in mismatch repair. Molecular Cell, 2003. 12(1): p. 233-246.
275. Au, K.G., K. Welsh, and P. Modrich, INITIATION OF METHYL-DIRECTED MISMATCH REPAIR. Journal of Biological Chemistry, 1992. 267(17): p. 12142-12148.
276. Palombo, F., et al., GTBP, A 160-KILODALTON PROTEIN ESSENTIAL FOR MISMATCH-BINDING ACTIVITY IN HUMAN-CELLS. Science, 1995. 268(5219): p. 1912-1914.
277. Iaccarino, I., et al., MSH6, a Saccharomyces cerevisiae protein that binds to mismatches as a heterodimer with MSH2. Current Biology, 1996. 6(4): p. 484-486.
278. Mazurek, A., et al., Sequence context effect for hMSH2-hMSH6 mismatch-dependent activation. Proceedings of the National Academy of Sciences of the United States of America, 2009. 106(11): p. 4177-4182.
279. Habraken, Y., et al., Binding of insertion/deletion DNA mismatches by the heterodimer of yeast mismatch repair proteins MSH2 and MSH3. Current Biology, 1996. 6(9): p. 1185-1187.
280. Genschel, J., et al., Isolation of MutS beta from human cells and comparison of the mismatch repair specificities of MutS beta and MutS alpha. Journal of Biological Chemistry, 1998. 273(31): p. 19895-19901.
233
281. Habraken, Y., et al., ATP-dependent assembly of a ternary complex consisting of a DNA mismatch and the yeast MSH2-MSH6 and MLH1-PMS1 protein complexes. Journal of Biological Chemistry, 1998. 273(16): p. 9837-9841.
282. Bowers, J., et al., MSH-MLH complexes formed at a DNA mismatch are disrupted by the PCNA sliding clamp. Journal of Molecular Biology, 2001. 306(5): p. 957-968.
283. Raschle, M., et al., Mutations within the hMLH1 and hPMS2 subunits of the human MutL alpha mismatch repair factor affect its ATPase activity, but not its ability to interact with hMutS alpha. Journal of Biological Chemistry, 2002. 277(24): p. 21810-21820.
284. Kolodner, R.D., MISMATCH REPAIR - MECHANISMS AND RELATIONSHIP TO CANCER SUSCEPTIBILITY. Trends in Biochemical Sciences, 1995. 20(10): p. 397-401.
285. de la Chapelle, A., Genetic predisposition to colorectal cancer. Nature Reviews Cancer, 2004. 4(10): p. 769-780.
286. Rowley, P.T., Inherited susceptibility to colorectal cancer, in Annual Review of Medicine. 2005. p. 539-554.
287. Bertrand, P., et al., Physical interaction between components of DNA mismatch repair and nucleotide excision repair. Proceedings of the National Academy of Sciences of the United States of America, 1998. 95(24): p. 14278-14283.
288. Johnson, R.E., et al., Evidence for involvement of yeast proliferating cell nuclear antigen in DNA mismatch repair. Journal of Biological Chemistry, 1996. 271(45): p. 27987-27990.
289. Dieckman, L.M., B.D. Freudenthal, and M.T. Washington, PCNA Structure and Function: Insights from Structures of PCNA Complexes and Post-translationally Modified PCNA. Sub-cellular biochemistry, 2012. 62: p. 281-99.
290. Chen, C., et al., Saccharomyces carevisiae pol30 (proliferating cell nuclear antigen) mutations impair replication fidelity and mismatch repair. Molecular and Cellular Biology, 1999. 19(11): p. 7801-7815.
291. Kokoska, R.J., et al., A mutation of the yeast gene encoding PCNA destabilizes both microsatellite and minisatellite DNA sequences. Genetics, 1999. 151(2): p. 511-519.
292. Antony, E. and M.M. Hingorani, Mismatch recognitioh-coupled stabilization of Msh2-Msh6 in an ATP-bound state at the initiation of DNA repair. Biochemistry, 2003. 42(25): p. 7682-7693.
293. Pflugrath, J.W., The finer things in X-ray diffraction data collection. Acta Crystallographica Section D-Biological Crystallography, 1999. 55: p. 1718-1725.
294. Read, R.J., Pushing the boundaries of molecular replacement with maximum likelihood. Acta Crystallographica Section D-Biological Crystallography, 2001. 57: p. 1373-1382.
234
295. Bailey, S., The Ccp4 Suite - Programs for Protein Crystallography. Acta Crystallographica Section D-Biological Crystallography, 1994. 50: p. 760-763.
296. Emsley, P. and K. Cowtan, Coot: model-building tools for molecular graphics. Acta Crystallographica Section D-Biological Crystallography, 2004. 60: p. 2126-2132.
297. TorresRamos, C.A., et al., Requirement of proliferating cell nuclear antigen in RAD6-dependent postreplicational DNA repair. Proceedings of the National Academy of Sciences of the United States of America, 1996. 93(18): p. 9676-9681.
298. Zhang, Z.G., K. Shibahara, and B. Stillman, PCNA connects DNA replication to epigenetic inheritance in yeast. Nature, 2000. 408(6809): p. 221-225.
299. Sharp, J.A., et al., Yeast histone deposition protein Asf1p requires Hir proteins and PCNA for heterochromatic silencing. Current Biology, 2001. 11(7): p. 463-473.
300. Goldsby, R.E., et al., High incidence of epithelial cancers in mice deficient for DNA polymerase delta proofreading. Proceedings of the National Academy of Sciences of the United States of America, 2002. 99(24): p. 15560-15565.
301. Goldsby, R.E., et al., Defective DNA polymerase-delta proofreading causes cancer susceptibility in mice. Nature Medicine, 2001. 7(6): p. 638-639.
302. Venkatesan, R.N., et al., Mutation at the polymerase active site of mouse DNA polymerase delta increases genomic instability and accelerates tumorigenesis. Molecular and Cellular Biology, 2007. 27: p. 7669-7682.
303. Flohr, T., et al., Detection of mutations in the DNA polymerase delta gene of human sporadic colorectal cancers and colon cancer cell lines. International Journal of Cancer, 1999. 80(6): p. 919-929.
304. Dacosta, L.T., et al., Polymerase-Delta Variants in Rer Colorectal Tumors. Nature Genetics, 1995. 9(1): p. 10-11.
305. Dieckman, L.M. and M.T. Washington, PCNA trimer instability inhibits translesion synthesis by DNA polymerase eta and by DNA polymerase delta. DNA Repair, 2013.
306. Masutani, C., et al., Xeroderma pigmentosum variant (XP-V) correcting protein from HeLa cells has a thymine dimer bypass DNA polymerase activity. Embo Journal, 1999. 18(12): p. 3491-3501.
307. Ishida, T. and K. Kinoshita, Prediction of disordered regions in proteins based on the meta approach. Bioinformatics, 2008. 24(11): p. 1344-1348.
308. Armstrong, A.A., F. Mohideen, and C.D. Lima, Recognition of SUMO-modified PCNA requires tandem receptor motifs in Srs2. Nature, 2012. 483(7387): p. 59-U99.
309. Fink, A.L., Natively unfolded proteins. Current Opinion in Structural Biology, 2005. 15(1): p. 35-41.
235
310. Dunker, A.K., et al., Function and structure of inherently disordered proteins. Current Opinion in Structural Biology, 2008. 18(6): p. 756-764.
311. Cortese, M.S., V.N. Uversky, and A.K. Dunker, Intrinsic disorder in scaffold proteins: Getting more from less. Progress in Biophysics & Molecular Biology, 2008. 98(1): p. 85-106.
312. Mayanagi, K., et al., Three-dimensional electron microscopy of the clamp loader small subunit from Pyrococcus furiosus. Journal of Structural Biology, 2001. 134(1): p. 35-45.
313. Kumar, R., et al., Stepwise loading of yeast clamp revealed by ensemble and single-molecule studies. Proceedings of the National Academy of Sciences of the United States of America, 2010. 107(46): p. 19736-19741.
314. Koundrioukoff, S., et al., A direct interaction between proliferating cell nuclear antigen (PCNA) and Cdk2 targets PCNA-interacting proteins for phosphorylation. Journal of Biological Chemistry, 2000. 275(30): p. 22882-22887.
315. Moldovan, G.L., B. Pfander, and S. Jentsch, PCNA controls establishment of sister chromatid cohesion during S phase. Molecular Cell, 2006. 23(5): p. 723-732.
316. Cortese, M.S., V.N. Uversky, and A. Keith Dunker, Intrinsic disorder in scaffold proteins: Getting more from less. Progress in Biophysics and Molecular Biology, 2008. 98(1): p. 85-106.