Insight Into the Fidelity of Two X-Family Polymerases: Dna Polymerase Mu and Dna Polymerase Beta

INSIGHT INTO THE FIDELITY OF TWO X-FAMILY POLYMERASES: DNA POLYMERASE MU AND DNA POLYMERASE BETA

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

Michelle P. Roettger, B.S.

* * * * *

The Ohio State University 2008

Dissertation Committee: Approved by Professor Ming-Daw Tsai, Advisor

Professor Ross Dalbey, Co-advisor ______Professor Richard Swenson Advisor Ohio State Biochemistry Program Professor Juan Alfonzo

ABSTRACT

DNA polymerase µ (Pol µ) is a recently discovered X-family DNA polymerase,

with yet unknown physiological function. Due to its preferential expression in secondary

lymphoid tissues, this enzyme has been implicated as a potential mutase involved in the

somatic hypermutation (SHM) of immunoglobulin (Ig) genes during antibody affinity

maturation. To evaluate the hypothesis which regards Pol µ as a mutase in Ig maturation,

pre-steady-state kinetic methods were used to accurately measure the fidelity of human

Pol µ based on all 16 possible deoxynucleotide (dNTP) incorporations and four matched

ribonucleotide (rNTP) incorporations into normal DNA primer/template substrates. The

overall fidelity of Pol µ was estimated to be in the range of 10-3-10-5 for both dNTP and

rNTP incorporations and was sequence-independent. Furthermore, to evaluate the

template-independent polymerization of this enzyme, the kinetics of dNTP and rNTP

incorporation into a single-stranded DNA substrate were measured and qualitatively

compared to terminal deoxynucleotidyl transferase (TdT). The potential biological

functions of Pol µ are discussed on the basis of the pre-steady-state kinetic data.

DNA Polymerase β (Pol β), another X-family polymerase has been well

characterized both biochemically and structurally. This enzyme is known to play an important role in short DNA gap-filling during mammalian base excision repair (BER).

ii Based upon its small size (39 kDa) and a lack of intrinsic exonuclease activity, Pol β is an attractive enzyme for model studies on the mechanism by which polymerase fidelity is achieved. Our lab has previously utilized stopped-flow fluorescence to examine the matched dNTP incorporation pathway of Pol β. While monitoring the reaction’s progress utilizing a DNA substrate containing a 2-aminopurine (2-AP) fluorescent probe, a biphasic trace is observed. Extensive studies involving a variety of chemical probes indicate that the fast fluorescence transition corresponds to a dNTP-induced subdomain conformational change occurring prior to the rate-limiting chemistry step, while the slow fluorescence transition corresponds to a post-chemistry conformational change, likely subdomain reopening. In this work, stopped-flow fluorescence assays are further utilized: i) to examine the role of R258 in subdomain reopening by mechanism studies on site-specific Pol β mutant, R258A; ii) to investigate the mechanism of Pol β mismatched dNTP incorporation by wild-type (WT) and I260Q “mutator” mutant; and iii) to evaluate the contribution of the reverse of the conformational closing step to Pol β’s fidelity.

Computational studies have suggested that reorientation of the R258 side chain is rate-limiting during Pol β catalysis, and that the R258A mutant shows facilitated subdomain closing, consistent with a reported increased rate of nucleotide insertion. By varying pH and buffer viscosity, we can decouple the rate of chemistry from the rate of the slow fluorescence transition, thus directly assigning this transition to a conformational event after chemistry, likely subdomain reopening. Analysis of the Pol β R258A mutant suggests that while rotation of the R258 sidechain is not rate-limiting in Pol β’s overall kinetic pathway, it is kinetically significant in subdomain reopening.

iii While matched nucleotide incorporation by DNA polymerase β (Pol β) has been well-studied, a true understanding of polymerase fidelity requires comparison of both matched and mismatched dNTP incorporation pathways. Here we examine the mechanism of misincorporation for wild-type (WT) Pol β and an error-prone I260Q variant using stopped-flow fluorescence assays and steady-state fluorescence spectroscopy. In stopped-flow, a biphasic fluorescence trace is observed for both enzymes during mismatched dNTP incorporation. The fluorescence transitions are in the same direction as that observed for matched dNTP, albeit with lower amplitude.

Assignments of the fast and slow fluorescence phases are designated to the same mechanistic steps previously determined for matched dNTP incorporation. For both WT and I260Q mismatched dNTP incorporation, the rate of the fast phase, reflecting subdomain closing, is comparable to that induced by correct dNTP. Pre-steady-state kinetic evaluation reveals that both enzymes display similar correct dNTP insertion profiles, and the lower fidelity intrinsic to the I260Q mutant results from enhanced efficiency of mismatched incorporation. Notably, in comparison to WT, I260Q demonstrates enhanced intensity of fluorescence emission upon mismatched ternary complex formation. Both kinetic and steady-state fluorescence data suggest that relaxed discrimination against incorrect dNTP by I260Q is a consequence of a loss in ability to destabilize the mismatched ternary complex. Overall, our results provide first direct evidence that mismatched and matched dNTP incorporations proceed via analogous kinetic pathways, and support our standing hypothesis that the fidelity of Pol β originates from destabilization of the mismatched closed ternary complex and chemical transition state.

iv In light of recent emphasis on the significance of the reverse of the

conformational closing step and its relationship to fidelity, stopped-flow fluorescence analyses of Pol β’s putative reverse closing step were conducted. Examination of putative reverse closing under a variety of altered reaction parameters starting from both preformed ternary matched and mismatched complexes, support that the fluorescence decay observed upon addition of EDTA does, in fact, correspond to the reverse of the conformational closing step. Analysis of the relative magnitudes of reverse closing and forward chemistry, still support our standing hypothesis that the fidelity of Pol β is determined by the difference in free energy between matched and mismatched dNTP incorporation pathways at the chemical transition state. However, the results must be interpreted with caution, as additional studies are necessary to fully assess the contribution of the reverse of conformational closing to the fidelity of Pol β.

Dedicated to my loving husband, Jeffrey Roettger, and to my parents, Geoffrey and Marcia Pomeroy

ACKNOWLEDGMENTS

I wish to express my sincere gratitude to my advisor, Dr. Ming-Daw Tsai, for the honor of joining his lab. I am truly grateful to have found a positive and stimulating environment in which to pursue my graduate studies. Dr. Tsai’s patience and provision have not only promoted the development of my intellect, but have more importantly allowed me to renew my love for science. I also extend a special thanks to my dissertation committee members, Dr. Ross Dalbey, Dr. Richard Swenson, and Dr. Juan

Alfonzo, for their time and vested interest in my education.

I would like to acknowledge past and present colleagues, including Kevin Fiala,

Yuxia Dong, Marina Bakhtina, Sandeep Kumar, Shengjiang Tu, Yu Wang, Haiyan Song,

Anjali Mahajan, Hyun Lee, and Brandon Lamarche. Everyone in this group of incredibly talented scientists and has taught me something new not only about science, but also about being human. It has been a privilege to have worked with each of these individuals, and in multiple respects I am excited to see how their future successes will continue to contribute to a better world. I would specifically like to thank Marina for her generosity in sharing her time and expertise with me from my very first week in the Tsai lab. Her mentorship has shown me first-hand that it is possible to possess a successful

vii career in science, while also maintaining a balanced family-life. I am also forever

grateful for the personal and professional encouragement and advice of Lena Furci.

Thanks to several others who have contributed to my work: Ruth Luketic for her

instrumental role in maintaining organization in the Tsai lab; Dr. Ross Dalbey for use of

his FPLC and fluorimeter; Dr. Zucai Suo for initial training in enzyme kinetics and

protein purification – the third chapter of this dissertation was completed in his

laboratory; Dr. Dale Ramsden and Dr. Beverly Mitchell for providing TdT. This work

was supported in-part by fellowships and grants from the Ohio State Biochemistry

Program and the National Institutes of Health Chemistry-Biology Interface Training

Program at The Ohio State University.

Most importantly, I thank God for blessing me with the opportunity to pursue

such studies on the intricate design of His Creation, and for the people with which He has

surrounded me during the process. I am extremely grateful for the unconditional love

and support of my parents, Geoffrey and Marcia, and siblings, Melody and Sean. Since

childhood, my parents have sacrificed to provide me with superior educational

opportunities that have ultimately afforded my current endeavors. While my father has

promoted my interest in the math and sciences, my mother has promoted the importance of music and the arts. It is the balance of these disciplines that makes me the individual I am today. I would also like to thank the Roettgers, Fashians, Abts, McKenzies, and

Duce, for all their support over the past several years. Lastly, my husband, Jeff, has been my most solid emotional support throughout this entire process. It has been such a privilege to have him by my side during this time in my life, as his unwavering love, optimism, and encouragement have been vital to my success.

viii

VITA

August 9, 1979...... Born – Columbus, Ohio.

June 2001...... Bachelor of Science in Biochemistry, Otterbein College.

2002 – 2003...... Ohio State Biochemistry Program Fellow, The Ohio State University.

2003 – 2006...... NIH Chemistry-Biology Interface Fellow, The Ohio State University.

2006 – present...... Graduate Teaching and Research Associate, The Ohio State University.

PUBLICATIONS

Research Publications

1. Roettger, M. P., Bakhtina, M., and Tsai, M. D. (2008) Mismatched and matched dNTP incorporation by DNA polymerase β proceed via analogous kinetic pathways. Biochemistry, submitted.

2. Bakhtina, M., Roettger, M. P., Kumar, S., and Tsai, M. D. (2007) A unified kinetic mechanism applicable to multiple DNA polymerases. Biochemistry 46, 5463-72.

3. Roettger, M. P., Fiala, K. A., Sompalli, S., Dong, Y., and Suo, Z. (2004) Pre- steady-state kinetic studies of the fidelity of human DNA polymerase mu. Biochemistry 43, 13827-38.

FIELDS OF STUDY

Major Field: Biochemistry

TABLE OF CONTENTS

Page

Abstract ...... ii

Dedication ...... vi

Acknowledgments ...... vii

Vita ...... ix

List of Tables...... xvi

List of Figures ...... xvii

List of Abbreviations ...... xx

Chapters:

1. Introduction ...... 1

1.1 DNA Polymerase Fidelity...... 1

1.2 The Mammalian DNA Polymerase X Family ...... 2

1.3 DNA Polymerase µ Background ...... 5

1.4 DNA Polymerase β Background ...... 6

1.5 The Kinetic Mechanism of Nucleotide Incorporation by Pol β ...... 9

xi 2. Materials and Methods ...... 17

2.1 Materials ...... 17

2.1.1. Materials for Pol µ studies ...... 17

2.1.2. Materials for Pol β studies ...... 17

2.2 Cloning and Enzyme Purification ...... 18

2.2.1. Cloning and purification of full-length Pol µ ...... 18

2.2.2. Cloning and purification of full-length Pol β ...... 21

2.2.3. Cloning and purification of Pol β R258A mutant ...... 22

2.2.4. Cloning and purification of Pol β I260Q mutant ...... 23

2.3 Pol µ Experimental Details ...... 24

2.3.1. DNA substrates ...... 24

2.3.2. Reaction buffer optimization ...... 25

2.3.3. Enzyme stability assays ...... 26

2.3.4. Substrate specificity measurements ...... 26

2.3.5. DNA trap experiment ...... 27

2.3.6. Product and data analysis ...... 27

2.4 Pol β Experimental Details ...... 28

2.4.1. DNA substrates ...... 28

2.4.2. Reaction buffer composition ...... 29

2.4.3. Stopped-flow fluorescence assays ...... 30

2.4.4. Rapid chemical quench ...... 31

2.4.5. Steady-state fluorescence ...... 31

2.4.6. Product and data analysis ...... 31

xii 3. Pre-steady-state Kinetic Analysis of the Fidelity of Human DNA Polymerase µ ...... 33

3.1 Reaction Buffer Optimization ...... 33

3.2 Enzyme Stability Assays ...... 37

3.3 Rapid Equilibrium of Nucleotide Binding to Binary E•DNA Complex . . 39

3.4 Rapid DNA Dissociation and Slow Polymerization ...... 41

3.5 Substrate Specificity of Incoming Correct dNTP ...... 42

3.6 Substrate Specificity of Incoming Incorrect dNTP ...... 47

3.7 Substrate Specificity of Incoming Matched rNTP ...... 49

3.8 Weak Strand-displacement Activity ...... 52

3.9 Distributive Elongation of a Primer/Template Substrate ...... 53

3.10 Pol µ Displays Template-independent Terminal Transferase Activity in the Presence of Mg2+ ...... 55

3.11 Substrate Specificity of dNTP and rNTP Incorporation into Single-strand DNA ...... 58

3.12 Discussion and Conclusion ...... 59

4. Reopening Studies of DNA Polymerase β and R258A Mutant – Contribution of R258 to Subdomain Motions ...... 69

4.1 Arg258 Background and Importance ...... 69

4.2 Dissection of the Chemistry and Conformational Subdomain-reopening Steps of WT Pol β using Differential pH and Viscosity ...... 71

4.3 The Slow Phase of Fluorescence Transition does not Originate from Enzyme Translocation ...... 76

4.4 R258A Characterization ...... 77

4.5 R258A Mutant of Pol β has a Facilitated Subdomain-reopening Step . . . 79

xiii 4.6 Discussion and Conclusion ...... 81

5. Mechanism of Mismatch Incorporation by DNA Polymerase β and I260Q “Mutator” Mutant ...... 83

5.1 Mismatched dNTP Incorporation ...... 83

5.2 Ile260 Background ...... 85

5.3 Mechanism of Mismatched dNTP Incorporation by WT Pol β ...... 85

5.4 dNTP Concentration Dependence of the Fast and Slow Fluorescence Transitions during Mismatch Incorporation ...... 89

5.5 A Biphasic Trace is Observed for Multiple Mismatches ...... 91

5.6 Kinetic Properties of I260Q and WT are Similar for Correct dNTP Incorporation ...... 92

5.7 Kinetic Properties of I260Q and WT are Different for Mismatched dNTP Incorporation ...... 96

5.8 The Fast Fluorescence Transition is also Present at Physiological pH . . 100

5.9 Steady-state Fluorescence Studies ...... 103

5.10 Discussion and Conclusion ...... 104

6. DNA Polymerase β Reverse Closing ...... 111

6.1 Significance of the Reverse of the Conformational Closing Step ...... 111

6.2 Monitoring Reverse Closing in Stopped-flow ...... 112

6.3 Dependence of Reverse Closing on Altered Reaction Parameters . . . . . 114

6.4 Dependence of Reverse Closing on Differential Viscosity ...... 119

6.5 Reverse Closing in the Presence of a Mismatched dNTP ...... 121

6.6 R258A Demonstrates Facilitated Reverse Closing ...... 126

6.7 Discussion and Conclusion ...... 127

xiv

Bibliography ...... 130

LIST OF TABLES

Table Page

2.1 Pol µ DNA substrates ...... 25

2.2 Pol β DNA substrates ...... 29

3.1 Pre-steady-state kinetic parameters of human DNA polymerase µ . . . . 46

3.2 Pre-steady-state kinetic parameters of Pol µ incorporation into single- strand DNA ...... 59

3.3 Comparison of kinetic parameters of human Pol µ and rat Pol β for dNTP incorporation ...... 64

4.1 Effect of viscosity on kfast, kslow, and kquench in stopped-flow during correct dNTP incorporation by WT Pol β at pH 8.3 ...... 75

5.1 Kinetic rate and binding constants for WT versus I260Q nucleotide incorporation into 19/36AP(T)/15 and 19/36AP(C)/15 DNA substrates at pH 8.5 ...... 88

5.2 Kinetic rate and binding constants for WT versus I260Q nucleotide incorporation into 19/36AP(T)/15 and 19/36AP(C)/15 DNA substrates at pH 7.6 ...... 102

6.1 Dependence of the rate of reverse closing on EDTA, free Mg2+, and matched dATP ...... 119

6.2 Viscosity dependence of the rate of reverse closing ...... 120

6.3 Comparison of WT versus R258A reverse closing and forward chemistry rates ...... 127

xvi

LIST OF FIGURES

Figure Page

1.1 DNA Polymerase X Family common architecture ...... 4

1.2 Subdomain organization of Pol β ...... 8

1.3 Key active site residues in the Pol β “open” state and “closed” state ...... 10

1.4 Nucleotidyl transfer mechanism of Pol β ...... 11

1.5 Typical matched dNTP incorporation trace by Pol β as monitored by 2- AP fluorescence in stopped-flow ...... 12

1.6 Kinetic scheme of single-nucleotide incorporation ...... 16

2.1 Expression and purification of Pol µ ...... 20

2.2 Expression and purification of Pol β ...... 22

3.1 Pol µ reaction buffer optimization ...... 36

3.2 Pol µ stability assay ...... 38

3.3 Incorporation of [α-32P]dGTP into D-8 ...... 42

3.4 Time course of product formation for correct dTTP incorporation into D-1 ...... 43

3.5 Substrate specificity of incoming correct nucleotide ...... 45

3.6 Comparison of matched and mismatched dNTP incorporation by Pol µ ...... 48

xvii 3.7 Pol µ template-dependent ribonucleotide incorporation ...... 49

3.8 Substrate specificity of incoming correct ribonucleotide ...... 51

3.9 Pol µ deoxynucleotide incorporation into a single-nucleotide gapped DNA substrate ...... 53

3.10 Pol µ DNA trap assay ...... 54

3.11 Processivity evaluation of Pol µ template-independent incorporation ...... 56

3.12 Template-independent dNTP incorporation by Pol µ ...... 57

3.13 Template-independent rNTP incorporation by Pol µ ...... 58

4.1 Motion of Arg258 upon subdomain closure...... 70

4.2 pH dependence of kslow and kquench in stopped-flow and chemical quench by WT Pol β ...... 72

4.3 Qualitative free energy profile of correct dNTP incorporation by Pol β ...... 73

4.4 Viscosity dependence of kfast and kslow in stopped-flow during correct dNTP incorporation by WT Pol β at pH 7.0 ...... 74

4.5 Viscosity dependence of kfast and kslow in stopped-flow during correct dNTP incorporation by WT Pol β at pH 8.3 ...... 76

4.6 Pol β stopped-flow fluorescence assay of correct dNTP incorporation into single-nucleotide gapped DNA substrate ...... 77

4.7 R258A mutant of Pol β stopped-flow and chemical quench overlay of dATP incorporation into 19/36AP(T) ...... 79

4.8 Viscosity dependence of kslow in stopped-flow during correct dNTP incorporation by WT Pol β and R258A at pH 7.5 versus pH 8.3 ...... 80

4.9 Qualitative free energy profile of correct dNTP incorporation by Pol β versus R258A ...... 81

5.1 Pol β mismatched dNTP incorporation in stopped-flow ...... 87

xviii 5.2 Viscosity dependence of kfast in stopped-flow during mismatched dNTP incorporation by WT Pol β and I260Q ...... 89

5.3 Biphasic fluorescence trace observed for multiple mismatched dNTP incorporation by WT Pol β ...... 92

5.4 I260Q correct dNTP incorporation in stopped-flow ...... 94

5.5 Viscosity dependence of kfast and kslow in stopped-flow during correct dNTP incorporation by WT Pol β versus I260Q ...... 95

5.6 Comparison of WT Pol β and I260Q T:G mismatched incorporation at pH 8.5 ...... 97

5.7 I260Q T:G mismatch titration at pH 8.5 ...... 99

5.8 Comparison of WT Pol β and I260Q T:G mismatched incorporation at pH 7.6 ...... 101

5.9 Steady-state fluorescence spectra of matched and mismatched ternary complexes for WT Pol β and I260Q ...... 104

5.10 Qualitative free energy profile of matched and mismatched dNTP incorporation by Pol β versus I260Q ...... 109

6.1 Kinetic scheme of single-nucleotide incorporation with emphasis on reverse closing ...... 113

6.2 Comparison of forward and reverse closing as monitored in stopped-flow ...... 114

6.3 Effect of altered reaction conditions on reverse closing ...... 118

6.4 Viscosity dependence of reverse closing...... 120

6.5 Mismatched reverse closing ...... 123

6.6 Effect of alternate matches/mismatches on reverse closing ...... 125

xix

LIST OF ABBREVIATIONS

AP apurinic/apyridinic

BER base excision repair

BRCA1 breast cancer 1, early onset

BRCT BRCA1 C-terminal

BSA bovine serum albumin

°C degrees Celsius

DEAE diethylaminoethyl dNTP 2’-deoxynucleoside 5’-triphosphate dRP 5’-deoxyribosephosphate

DSB double strand break

DTT dithiothreitol

EDTA ethylenediaminetetraacetic acid g gram(s) h hour(s)

HhH helix-hairpin-helix

Ig immunoglobulin

κ kappa

xx µ micro m milli

M moles per liter

MD molecular dynamics min minute(s) mol mole(s)

NHEJ nonhomologous end-joining

NLS nuclear localization signal

PAGE polyacrylamide gel electrophoresis

PMSF phenylmethanesulphonylfluoride

Pol β DNA polymerase beta

Pol λ DNA polymerase lambda

Pol µ DNA polymerase mu rNTP ribonucleoside 5’-triphosphate s second

SHM somatic hypermutation

TBE tris/borate/EDTA electrophoresis buffer

TdT terminal deoxynucleotidyl transferase

WT wild-type

xxi

CHAPTER 1

INTRODUCTION

1.1 DNA Polymerase Fidelity.

DNA replication fidelity is crucial to the survival of any organism. Without accurate DNA replication during cell propagation and DNA repair processes, cell death will certainly ensue. Fidelity, or base substitution error, can be qualitatively regarded as a measure of the frequency by which a polymerase incorporates a correct nucleotide versus an incorrect nucleotide. Quantitatively, fidelity is defined as a function of the ratios of the apparent nucleotide dissociation constant (Kd,app) and the maximum rate of polymerization (kp or kpol) measured in vitro for both correct and incorrect nucleotide incorporation. A replicative polymerase functionally requires high fidelity in order to accurately preserve genetic information, as well as to prevent mutations that may promote human disease such as cancer (1). Low fidelity polymerases are also of importance to human survival, as they possess the capacity to bypass replication stalling lesions, and play a large role in developing the diversity required for proper immune response. To date, there are over fifteen known eukaryotic polymerases, with base substitution error rates ranging from 10-1 to 10-6, plus an additional fidelity enhancement of 10-1 to 10-2 when intrinsic exonuclease proofreading function is considered (2, 3).

1 Studies examining the energetics of base pairing in solution have estimated the

free energy difference between correct and incorrect base pairing to be 1-3 kcal/mol (4,

5). This translates into a fidelity of 10-1 to 10-2 if a polymerase were to offer no selective

preference during DNA elongation synthesis (6). Clearly, the observation of higher

fidelity owned by the majority of polymerases indicates that these enzymes possess a

unique ability to provide significant selectivity enhancement during catalysis. After

decades of research, the question still remains: what is the nature of the mechanism by

which such selectivity is amplified in the active site of a polymerase? The initial portion

of this work (Chapter 3) delineates the fidelity characterization of the recently discovered

DNA polymerase µ, while the remainder (Chapters 4, 5, and 6) focuses on the

advancement of our understanding of the underlying mechanism of fidelity through

model studies on DNA polymerase β.

1.2 The Mammalian DNA Polymerase X Family.

To date, the X-family of mammalian DNA polymerases is comprised of DNA polymerase β (Pol β), terminal deoxynucleotidyl transferase (TdT), DNA polymerase λ

(Pol λ), and DNA polymerase µ (Pol µ). All members possess DNA polymerization

capability, albeit with differing DNA substrate specificities, while Pol β and Pol λ

uniquely possess an additional deoxyribose-phosphate (dRP) lyase activity. Among the

X-family polymerases, alignment of the human protein sequences indicates that Pol µ and

TdT are most homologous, with 41% sequence identity and 61% homology. Closely

following are Pol β and Pol λ, which share 34% identity and 54% homology (7, 8).

2 The X-family members possess several architectural features in common: a

nuclear localization signal (NLS), an N-terminal BRCA1 C-terminal domain (BRCT)

domain (with the exclusion of Pol β), and a C-terminal Pol-β-like catalytic core

containing two internal helix-hairpin-helix (HhH) motifs (Figure 1.1) (7-11). The BRCT domain was first identified by sequence analysis of the C-terminal portion of the BRCA1

(breast cancer 1, early onset) susceptibility gene product. These domains are now known to mediate a vast assortment of DNA-protein and protein-protein interactions, and are present in many proteins which play intricate and diverse roles in cell cycle control, DNA

damage response, and DNA repair (12, 13). Recently, BRCT domains have been

reported to possess phospho-protein binding specificity (14-16), potentially implicating phosphorylation as a modulation mechanism in cellular polymerase function (17). The

HhH motif is also common to many DNA repair proteins, and is thought to mediate non- sequence-specific DNA recognition (18).

Figure 1.1: DNA Polymerase X Family common architecture. Hs, Homo Sapiens; BRCT, BRCA1 (Breast Cancer Gene 1) C-terminal domain; HhH, helix-hairpin-helix motif; NLS, nuclear localization signal. Figure modified from Ref. (11).

It is well established that Pol β is a DNA repair polymerase involved in both short-patch and long-patch mammalian base excision repair (BER) (19, 20). TdT is known to function in antigen receptor diversification during V(D)J recombination through addition of nucleotides (N additions) to gene segment junctions in a template- independent manner (21-23). The more recently discovered Pol λ and Pol µ do not yet

have clearly defined biological roles. Overall, both have been implicated in general

4 double strand break repair (DSB), but the specific context of their roles remains to be

established [for review see Ref. (17)].

1.3 DNA Polymerase µ Background.

Pol µ is a recently identified member of the X-family of DNA polymerases, with

yet unknown physiological function. While expression of this enzyme is noted

predominantly in secondary lymphoid tissues (ie: thymus, lymph nodes, spleen), basal

levels are also present in a variety of other tissues (7, 8). Based upon its lymphoid-

specific expression and marked error-prone tendency in vitro, this enzyme was initially

proposed to be a mutase involved in somatic hypermutation (SHM) (9, 10). SHM is a

secondary diversification process after V(D)J recombination in lymphocytes by which

high affinity antigen binding sites are created in the variable (V) region of

immunoglobulin (Ig) genes [for review see Refs. (24-26)]. Additional support for this theory was derived from the fact that overexpression of Pol µ in constitutively SHM active cells caused increased base substitution alterations specifically at IgV gene regions

(27).

However, a large body of more recent evidence points to involvement of Pol µ in repair of double strand breaks (DSBs) in both V(D)J recombination and non-homologous end joining (NHEJ) events. V(D)J recombination is the lymphocyte-specific process by which Ig V, D, and J gene segments are rearranged to generate the primary antibody repertoire [for review see Ref. (28)]. NHEJ is the major pathway for the repair of DSBs introduced by exogenous ionizing radiation and endogenous oxidative metabolism in all cell types, and is an essential mechanism that repairs DSBs generated during V(D)J 5 recombination [for review see Refs. (29-31)]. Several factors implicate the involvement

of Pol µ in V(D)J recombination and NHEJ, rather than SHM: (i) Pol µ shows ubiquitous

expression in a variety of tissues, in addition to lymphoid-specific tissues (7, 8). (ii) Pol

µ-deficient mice exhibit normal SHM patterns, yet show altered light chain Ig gene rearrangement (32, 33); (iii) Pol µ associates with known NHEJ participants, such as Ku and the XRCC4-ligase IV complex (34); (iv) In vitro, Pol µ possesses the unique ability to frequently generate frameshift mutations, a potentially useful characteristic for promotion of microhomology searching and pairing during NHEJ (35).

Similar to TdT, Pol µ can incorporate both rNTPs and dNTPs into either DNA or

RNA primers (36, 37). In addition, Pol µ can bypass several common DNA lesions, including abasic sites, 8-oxoguanines, cis-syn thymidine-thymidine dimers, as well as bulkier lesions such as 1, N6-ethenoadenines, N-2-acetylaminoflourines, (+)- and (-)-

trans-anti-benzo[a]pyrene-N2-dG adducts, and platinum DNA adducts, through a deletion

mechanism (38-41). These abilities are also in keeping with a proposed role for this

enzyme in V(D)J recombination and NHEJ.

1.4 DNA Polymerase β Background.

Pol β is ubiquitously expressed in mice, with increased presence in the testes, thymus, and brain (42). Pol β-deficient mice demonstrate inviability, as they complete embryogenesis but die shortly after birth (43, 44). Studies both in vivo and in vitro have unequivocally demonstrated Pol β’s important role in single-nucleotide gap filling during the short-patch BER process (45-48). This pathway is modeled as follows [for review see Refs. (19, 49, 50)]. A DNA glycosylase creates an apurinic/apyridinic (AP) site via 6 removal of a damaged base in duplex DNA, through breakage of the N-glycosidic bond.

An AP endonuclease then cleaves the phophodiester backbone 5’ to the sugar, thus generating a 3’-hydroxyl group and a 5’-deoxyribosephosphate (dRP) flap. Pol β then fills the gap with an undamaged nucleotide, and afterward Pol β’s 8-kD lyase domain is responsible for the removal of the dRP via a β-elimination mechanism (51). Finally, the nicked duplex DNA is sealed by either DNA ligase I or DNA ligase III (52, 53).

Multiple crystal structures of human Pol β solved in unliganded and various liganded states are of significant importance to Pol β mechanism studies. These include: free enzyme (54, 55), binary complexes of enzyme with gapped, nicked, or blunt-ended

DNA (56, 57), and ternary complexes of enzyme, DNA, and correct incoming dNTP (55,

57-59). More recently, insights into mismatch incorporation have been attained via binary structures of Pol β and DNA containing varied mismatches at the primer terminus

(60-62).

Pol β topology consists of two domains: an N-terminal 8-kD domain and a C- terminal 31-kD catalytic domain. The N-terminal domain binds the 5’-phosphate of the downstream primer in short gapped DNA substrates (63), and possesses dRP lyase activity that proves essential for successful completion of BER (64, 65). Upon functional alignment, the C-terminal domain of Pol β, possessing DNA polymerization capability, resembles the overall structure of other polymerases, such as HIV-1 reverse transcriptase

(66), bacteriophage T7 DNA polymerase (67), and Klenow Fragment of Escherichia coli

DNA polymerase I (68), in maintaining the canonical polymerase architecture which resembles a right hand. Accordingly, the C-terminal domain is further subdivided into fingers, palm, and thumb subdomains. According to the nomenclature suggested by

7 Steitz (69), the fingers, palm, and thumb subdomains functionally correspond to actions of nascent base pair binding, catalysis, and double-strand DNA binding, respectively

(Figure 1.2).

Figure 1.2: Subdomain organization of Pol β. The C-terminal catalytic domain of Pol β, is subdivided into three subdomains: the thumb subdomain (purple), which participates in duplex DNA binding; the palm subdomain (gold), which harbors the three catalytic aspartate residues; and the fingers subdomain (green), which is involved in nascent base pair binding. The N-terminal lyase domain is highlighted in white. Figure modified from Ref. (70).

8 1.5 The Kinetic Mechanism of Nucleotide Incorporation by Pol β.

Based upon the differences in free energy for correct and incorrect base pairing in

solution, Pol β’s measured fidelity of 10-4 - 10-5 (71-73) translates into a selectivity enhancement of up to 200-fold in Pol β’s active site (74). Crystal structures of many polymerases to date, including Pol β, implicate the existence of a large conformational change that occurs upon correct nucleotide binding, from an “open” E•DNA binary complex conformation to a “closed” E’•DNA•dNTP ternary complex conformation [for review see Refs. (75, 76)]. Formation of the “closed” ternary complex is a prerequisite to chemistry. Existence of this conformational change has led to proposal of a general

“induced-fit” mechanism, in which this conformational change is deemed the rate- limiting step in polymerization, and is, in turn, postulated to be the major contributor to polymerase fidelity (77, 78). Comparison of Pol β’s binary gapped DNA complex with the ternary gapped DNA complex containing correct incoming nucleotide ddCTP, does indeed reveal that there is a dNTP-induced subdomain closure originating from the fingers subdomain in which there is a 30° rotation of α-helix N toward the nascent base pair along the hinge axis of α-helix M (Figure 1.3) (57). In light of recent studies from

our lab, it is now commonly accepted that this conformational change, at least in the case

of Pol β, is not rate-limiting (79, 80).

Figure 1.3: Key active site residues in the Pol β “open” state (left) and “closed” state (right). Figure modified from Ref. (81).

Pol β utilizes a “two metal-ion” mechanism for nucleotide incorporation chemistry (82). This mechanism is likely conserved for all DNA and RNA polymerases

(69, 83). Crystal structures of Pol β ternary complexes were the first to validate this mechanism on a structural basis (55, 57). The active site of Pol β is located in the palm subdomain, and consists of three catalytic residues, Asp 190, Asp 192, and Asp 256, which coordinate two metal ions, usually Mg2+ ions (Figure 1.4). Metal ion A

coordinates dNTP in a tridentate fashion, and presumably enters the active site as a

Mg•dNTP complex. Metal ion B coordinates the 3’-oxygen of the primer, as well as the

2+ pro-Rp α-phosphate of the incoming nucleotide, and is referred to as the catalytic Mg

ion in mechanism studies. Both hexacoordinated metal ions serve to stabilize the

structure and charge of the pentacovalent transition state formed upon in-line

10 nucleophilic attack of the α-phosphate of an incoming nucleotide by the primer’s 3’- oxygen. The nucleotidyl transfer reaction continues through the transition state as the primer increases one nucleotide in length and a pyrophosphate leaving group is formed.

Figure 1.4: Nucleotidyl transfer mechanism of Pol β. Figure from Ref. (57).

Stopped-flow studies of Pol β, using both 2-aminopurine (2-AP), a fluorescent analog of adenine and guanine, and Pol β’s sole Trp325 residue, have been very important in advancing Pol β mechanism studies (59, 79, 80, 84). Both probes serve as independent reporters of global enzyme-DNA conformational changes which occur during the enzyme’s catalysis of nucleotide incorporation. Whether monitoring 2-AP or

11 Trp325, the stopped-flow trace of fluorescence emission upon correct nucleotide

incorporation into DNA substrate possesses two phases: a first (fast) phase followed by a second (slow) phase. In these experiments, the reaction is initiated by rapid mixing of preincubated E•DNA binary complex with correct incoming dNTP in the presence of saturating amounts of Mg2+ (Figure 1.5).

Figure 1.5: Typical matched dNTP incorporation trace by Pol β as monitored by 2- AP fluorescence in stopped-flow. The black trace shows correct dNTP incorporation -1 into 2-AP containing primer/template; a double exponential fit yields kfast = 126 ± 1 s -1 and kslow = 13.0 ± 0.1 s . The reaction was initiated by rapid mixing of solution A (containing 1 µM Pol β and 400 nM 19/36AP(T)) with solution B (containing 1 mM dATP). Reactions were performed at pH 7.6 and 37 °C in Pol β standard assay buffer containing 5 mM Mg2+.

Further studies involving a variety of chemical probes have been conducted in

order to determine the nature of these two phases. These studies support a model in 12 which the first (fast) phase of fluorescence is attributed to dNTP-induced subdomain

closure to form a ternary E’•DNA•dNTP complex and the second (slow) phase of

fluorescence is attributed to a conformational step rate-limited by chemistry. Stopped-

flow results constituting evidence for such a model include, in short: (i) The rate of the

slow phase of fluorescence emission matches the rate of chemistry as determined by

rapid chemical quench under identical reaction conditions (79). This suggests that

chemistry is the rate-limiting step in Pol β’s nucleotide incorporation mechanism. Since it is unlikely that the actual chemistry step would demonstrate a fluorescent transition in stopped-flow, it is more likely that the slow phase is representative of a fast step, such as enzyme reopening, occurring after the rate-limiting chemistry step. This is addressed in

Chapter 4 of this work. (ii) Use of dideoxy-terminated primer, in which the 3’-OH of the primer’s terminal nucleotide is absent, eliminates the slow phase of fluorescence transition in stopped-flow experiments. Since chemistry cannot physically occur with this primer substrate, the existence of fast phase alone is consistent with the expected formation of a closed ternary complex, and the elimination of the slow phase is consistent with the impossibility of chemistry under such conditions (79). (iii) Use of thio- substituted analogs (dNTPαS), is commonplace in kinetic analyses of polymerases for selective perturbation of the chemical step. Stopped-flow analysis of dNTPαS demonstrates an unperturbed fast phase transition followed by the typical slower fluorescence transition, albeit with a slowed rate compared to that of natural dNTP incorporation. This supports correlation of the slow phase to a step rate-limited by chemistry (80). (iv) Viscosity studies indicate that upon increasing sucrose or glycerol concentrations in the reaction mixture at neutral pH, the fast phase gradually slows, while

13 the slow phase proves unperturbed. This is consistent with the idea that a large conformational closing would be subject to molecular restraints imposed by increased solution viscosity, whereas chemistry would remain unaffected (80).

In addition to characterization of the two phases observed in stopped-flow during

Pol β’s nucleotide incorporation, studies using exchange-inert chromium(III) and rhodium(III) dNTP complexes have been further able to examine the order of magnesium binding in the active site. Enzyme binding of two magnesium ions is required for catalysis, namely consisting of a nucleotide-binding Mg2+ and a catalytic Mg2+. Binding

of substitution-inert Cr(III)•dNTP to Pol β•DNA binary complex in the absence of Mg2+

induces the fast phase of fluorescence alone. Upon subsequent addition of Mg2+ the slow phase of fluorescence is restored (85). Similarly, mixing of preformed Pol

β•DNA•Rh(III)dNTP ternary complex with Mg2+ demonstrates only the second (slow)

fluorescence phase. Upon mixing of free Mg2+ with preformed Pol β•DNA•Rh(III)dNTP

ternary complex in which the DNA primer is dideoxy-terminated, no stopped-flow signal

is observed at all (80). All of these results suggest that a fast conformational change

occurs upon Mg•dNTP binding to the E•DNA binary complex, and support that binding

of the catalytic Mg2+ occurs after the formation of the E’•DNA•dNTP ternary complex.

In response to long-standing criticisms regarding the possibility of obtaining unnatural results through use of the exchange-inert metal•dNTP complexes mentioned above, the respective difference in binding affinity for each of the two magnesium ions

Mg2+ Mg•dATP (Kd, app = 1.0 mM and Kd, app = 46 µM (79)) was employed to further test the

Mg2+ binding order (80). Stopped-flow results showed that upon rapid mixing of Pol

β•DNA with dNTP under limiting magnesium concentrations which allow Mg•dNTP

14 binding site saturation, yet do not provide sufficient Mg2+ to support catalysis, the fast phase of fluorescence only was observed. Paralleling the rhodium(III) experiments

mentioned above, fluorescence monitoring of the Pol β•DNA•MgdNTP ternary complex,

preformed under limited magnesium concentration so as to prevent chemistry, mixed

with excess Mg2+ demonstrates only the slow phase. Such results with the use of non-

analogs further corroborate mechanism studies using exchange-inert metal•dNTP

complexes.

Further validation regarding the metal binding order proposed for the Pol β

mechanism is found in the crystal structure of a pathway intermediate, Pol

β•DNA•Cr(III)dTMPCPP (59). This complex structural intermediate lacking bound

catalytic Mg2+ is found in the closed ternary conformation. It implicates that the fingers

subdomain closing occurs prior to catalytic Mg2+ binding.

In summary, through the use of viscogens, dideoxy-terminated primers, thio-

substituted analogs, and exchange-inert chromium(III) and rhodium(III) dNTP complexes, our lab has been able to probe the nature of two phases observed in stopped- flow, as well as dissect the functional roles of the two observed magnesium ions in the active site of Pol β which are required for catalysis. The combination of the results from

the aforementioned studies support our working hypothesis that the fast phase of

fluorescence corresponds to a fast subdomain-closing conformational change (Step 2,

Figure 1.6) to form an E’•DNA•dNTP ternary complex before binding of the catalytic

Mg2+, whereas the slow phase of fluorescence corresponds to a step rate-limited by

chemistry (Step 4, Figure 1.6), likely enzyme re-opening after the chemistry step (Step 6,

Figure 1.6), occurring after the binding of the catalytic Mg2+ (Step 3, Figure 1.6).

Figure 1.6: Kinetic scheme of single-nucleotide incorporation. E = Pol β in open finger conformation; E’ = Pol β in closed finger conformation; Dn = DNA; Dn+1 = DNA elongated by addition of one nucleotide; N = M•dNTP; M = catalytic metal ion; P = M•PPi. Figure from Ref. (80).

CHAPTER 2

MATERIALS AND METHODS

2.1 Materials.

2.1.1 Materials for Pol µ studies.

dNTPs were purchased from Gibco-BRL (Rockville, MD); rNTPs from TriLink

32 Biotechnologies (San Diego, CA); [γ- P]ATP from Perkin Elmer Life Sciences (Boston,

MA); calf intestine alkaline phosphatase from Fermentas (Hanover, MD); T4 polynucleotide kinase from USB (Cleveland, OH); Biospin columns from Bio-Rad

Laboratories (Hercules, CA); and activated calf thymus DNA from Sigma-Aldrich (St.

Louis, MO). All synthetic oligonucleotides used in DNA substrate preparation, were provided by TriLink Biotechnologies (San Diego, CA). Materials and reagents not listed were of standard molecular biology grade.

2.1.2 Materials for Pol β studies.

Ultra-pure dNTP’s and G-25 microspin columns were from GE Healthcare

32 (Piscataway, NJ); [γ- P]ATP from MP Biomedicals (Irvine, CA); T4 polynucleotide kinase from USB (Cleveland, OH); and reverse phase C18 cartridges from Waters

(Milford, MA). All synthetic oligonucleotides used in DNA substrate preparation, were

17 provided by Integrated DNA Technologies (Coralville, IA). Materials and reagents not listed were of standard molecular biology grade.

2.2 Cloning and Enzyme Purification.

2.1.1 Cloning and purification of full-length Pol µ.

Human Pol µ was subcloned into the Nde I/Xho I sites of the vector pET28b from

a previously published plasmid pEGUh6-POLM provided by Dr. Z. Wang (35). Similar

to other published reports (7, 8, 35), a hexahistidine tag was added to the N- and C-

termini of the Pol µ gene for the convenience of protein purification. Although human

DNA Pol µ (494 amino acid residues, 55.9 kDa) has been reported previously to be

overexpressed in E. coli strains BL21(DE3) (7) and BL21(DE3) pLysS (8), we failed to

achieve clear induction of this protein using IPTG in these strains under various

conditions including different induction temperatures, induction times, and IPTG

concentrations. The codon usage in human Pol µ could cause its undetectable expression

in E. coli BL21(DE3): 13 of 44 arginine codons of human Pol µ are rare (AGA/AGG)

(86). Similar results have been found in the expression of murine TdT in E. coli (87).

Therefore, the constructed plasmid pET28b-Pol µ was transformed into E. coli strain

BL21-CodonPlus(DE3)-RIL (Stratagene) to express human Pol µ fused to both N- and C- terminal hexahistidine tags.

Transformed E. coli cells were grown at 37 °C in the presence of 40 µg/ml kanamycin and 25 µg/ml chloramphenicol until reaching an OD600 of 0.55. At this point,

the cultures were induced with 0.1 mM IPTG and incubated at 18 °C for 15 hours. Cells

18 were harvested (4,000 rpm, 15 min) at 4 °C and resuspended in buffer A (10 mM KHPO4,

0.5 M NaCl, 10 mM MgCl2, 10% glycerol, 0.1% beta-mercaptoethanol, 5 mM imidazole,

pH 7.0 at 4 °C). After the addition of 1 mM PMSF and 1 tablet of protease inhibitor

cocktail (Roche), resuspended cells were lysed by French Press. The resulting lysate was

cleared by ultracentrifugation (35,000 rpm, 40 min). The supernatant was pooled and

incubated overnight at 4 °C with nickel-NTA superflow resin (Qiagen). The supernatant

was removed by centrifugation in a swing-bucket centrifuge (2,500 rpm, 10 min), and the

Pol µ-bound nickel resin was subsequently packed into a column. Bound proteins were

eluted through a linear gradient of 20-500 mM imidazole in buffer B (10 mM KHPO4,

0.35 M NaCl, 2.5 mM MgAc2, 10% glycerol, 0.1% 2-mercaptoethanol, pH 7.0 at 4 °C).

Pol µ-containing fractions were pooled and dialyzed against 2 liter buffer C (10 mM

KHPO4, 0.25 M NaCl, 10% glycerol, 1 mM EDTA, 0.1% beta-mercaptoethanol, pH 7.0 at

4 °C) at 4 °C. The dialyzed protein solution was loaded into a pre-packed ssDNA-

cellulose column (Sigma). After washing, Pol µ was then eluted with 250-1000 mM NaCl

gradient in buffer C. The fractions containing Pol µ were pooled and dialyzed against

buffer D (25 mM HEPES, 200 mM NaCl, 10% glycerol, 1 mM EDTA, 0.1% beta-

mercaptoethanol, pH 7.5 at 4 °C). The dialyzed Pol µ was passed through 10 ml DEAE-

Sepharose column (Amersham Pharmacia Biotech). The flow-through was subsequently

applied to a MonoS column (Amersham Pharmacia Biotech) and eluted using a gradient

of 200-1000 mM NaCl in buffer D. Fractions containing Pol µ were pooled, dialyzed against buffer D, and concentrated using a Centriprep YM-30 (Millipore). The concentrated protein was ultimately dialyzed against buffer E (25 mM HEPES, 200 mM

19 NaCl, 10% glycerol, 1 mM EDTA, 1 mM DTT, 50% glycerol, pH 7.5 at 4 °C). Pol µ was

purified to >95% purity based on SDS/PAGE analysis (Figure 2.1).

Figure 2.1: Expression and purification of Pol µ. Coomassie Blue staining and SDS/PAGE analysis of purified human Pol µ are shown. Lane 1, protein size marker; lane 2, crude extracts of non-induced cells; lane 3, crude extracts of IPTG-induced cells; lane 4, purified full-length human Pol µ.

The concentration of the purified Pol µ was measured spectrophotometrically at

280 nm using the calculated extinction coefficient of 52,857 M-1cm-1. The yield was

approximately 0.5 mg per liter of initial E. coli culture. The sequence of the purified Pol

µ was confirmed by in-gel trypsin digestion and capillary-liquid chromatography-

nanospray tandem mass spectrometry analysis (Nano-LC/MS/MS) at the Campus

Chemical Instrument Center at The Ohio State University. The hexahistidine-tags of the

20 purified Pol µ were also detected by Western blot analysis using anti-hexahistidine tag antibody (data not shown).

2.2.2. Cloning and purification of full-length Pol β.

The previously constructed plasmid pET17b-Polβ (71) was transformed into E. coli strain BL21(DE3)pLysS (Stratagene) to express rat Pol β (335 amino acids, 39 kDa).

Transformed E. coli cells were grown at 37 °C in the presence of 100 ug/mL ampicillin and 50 ug/mL chloramphenicol for 2-3 hours until reaching an OD600 of 0.6-0.7. At this point, the cultures were induced with 0.4 mM IPTG and incubated at 37 °C for approximately 5 hours. Cells were harvested (5,000 rpm, 10 min) at 4 °C and resuspended in 150 mL buffer F (50 mM Tris, 300 mM KCl, 1 mM DTT, pH 8.0 at 4°C) containing a protease inhibitor cocktail tablet (Roche). The resuspended cells were lysed by sonication, and the cell debris removed by centrifugation (18,000 rpm, 45 min). The crude lysate was first passed over a weak anion exchange DEAE-Sepharose column

(Sigma) in buffer F for removal of DNA and acidic proteins. The flow-through was subsequently added to a P11 cation-exchange column (Whatman) pre-equilibrated in buffer F. After washing, the bound proteins were eluted through a linear gradient of 300-

1000 mM KCl in buffer F. Pol β-containing fractions were pooled and concentrated to 2 mL using an Amicon protein concentrator. The concentrated sample was loaded on an

FPLC Superdex 75 gel filtration column (GE Healthcare), eluted in buffer G (100 mM

Tris, 150 mM KCl, 2 mM DTT, pH 8.0 at 4 °C), and collected in 1 mL fractions. Peak fractions were pooled, concentrated, and diluted by 50% with glycerol. Protein aliquots were frozen by liquid nitrogen and stored at -80 °C. Pol β was purified to >95% purity as determined by SDS/PAGE followed by silver staining (Figure 2.2). The enzyme 21 concentration was determined by UV absorption at 280 nm, using an extinction

coefficient of 21,200 M-1 cm-1. The approximate yield was 5 mg per liter of initial E. coli

culture.

Figure 2.2: Expression and purification of Pol β. SDS/PAGE analysis followed by silver staining of purified rat Pol β are shown. Lane 1, crude extracts of non-induced cells; lane 2, crude extracts of IPTG-induced cells; lane 3, crude protein lysate; lane 4, DEAE flow-through; lane 5, P11 load flow-through; lane 6, P11 wash; lane 7, gel filtration load sample; lane 8, purified full-length rat Pol β.

2.2.3. Cloning and purification of Pol β R258A mutant.

To generate the R258A mutant, site-directed mutagenesis was carried out with primers:

5’-CCACACAGGAGAATCGATATCGCGTTGATCCCCAAAGATCAGTAC-3’ and

5’-GTACTGATCTTTGGGGATCAACGCGATATCGATTCTCCTGTGTGG-3’ using 22 the QuikChange method (Stratagene). The underlined region represents the site of

mutagenesis. Successful mutagenesis was verified by both sequencing of the plasmid

(OSU Plant Microbe Genomics Facility) and electrospray ionization mass spectrometry

of the mutant protein (OSU Campus Chemical Instrument Center). The R258A mutant

was overexpressed and purified from the BL21(DE3)pLysS E. coli strain as previously

described for wild-type (Section 2.2.2). All R258A assays used a larger enzyme excess

to account for the notable decrease in its DNA binding affinity compared to that of wild-

type (88).

2.2.4. Cloning and purification of Pol β I260Q mutant.

To generate the I260Q mutant, site-directed mutagenesis was carried out with

primers:

5′-GGAGAATCGATATCAGGTTGCAGCCCAAAGATCAGTACTACTGTGG-3′ and

5′-CCACAGTAGTACTGATCTTTGGGCTGCAACCTGATATCGATTCTCC-3′ using

the QuikChange method (Stratagene). The underlined region represents the site of

mutagenesis. Successful mutagenesis was verified by both plasmid sequencing (OSU

Plant Microbe Genomics Facility) and electrospray ionization mass spectrometry of the mutant protein (OSU Campus Chemical Instrument Center). The I260Q mutant was overexpressed and purified from the BL21(DE3)pLysS E. coli strain as previously described for wild-type (Section 2.2.2).

23 2.3 Pol µ Experimental Details.

2.3.1. DNA substrates.

To prepare DNA substrates for dNTP incorporation assays, a 21-mer primer was

annealed to each of four different 41-mer templates (Table 2.1). These oligomers were

purified by denaturing polyacrylamide gel electrophoresis (PAGE) (15-18% acrylamide,

7 M urea), and desalted using a reverse phase C18 cartridge. Their concentrations were

determined by UV absorbance at 260 nm with the following extinction coefficients (M-

1cm-1): 21-mer, ε = 194,100; D-1 41-mer, ε = 396,700; D-6 41-mer, ε = 394,200; D-7 41-

mer, ε = 392,200; and D-8 41-mer, ε = 389,500. The 21-mer primer was 5’-32P-labeled by

32 incubation with [γ- P]ATP and T4 polynucleotide kinase at 37 °C for an hour.

Subsequent centrifugation through a Biospin-6 column served to remove any remaining

32 [γ- P]ATP from the newly labeled 21-mer mixture. Both 21-mer and 41-mer (1:1.1 molar ratio) were annealed to form a 21/41-mer complex by heating the mixture at 95 °C for 10 min, and then slowly cooling it to room temperature. Single-nucleotide gapped

DNA (D-8g) was prepared in the same manner with 21-mer, 19-mer, and 41-mer annealed at a 1:2:2 molar ratio, respectively. The downstream primer utilized was 5’- phosphorylated.

D-1 5’-CGCAGCCGTCCAACCAACTCA 3’-GCGTCGGCAGGTTGGTTGAGTAGCAGCTAGGTTACGGCAGG-5’ D-6 5’-CGCAGCCGTCCAACCAACTCA 3’-GCGTCGGCAGGTTGGTTGAGTGGCAGCTAGGTTACGGCAGG-5’ D-7 5’-CGCAGCCGTCCAACCAACTCA 3’-GCGTCGGCAGGTTGGTTGAGTTGCAGCTAGGTTACGGCAGG-5’ D-8 5’-CGCAGCCGTCCAACCAACTCA 3’-GCGTCGGCAGGTTGGTTGAGTCGCAGCTAGGTTACGGCAGG-5’

D-8g 5’-CGCAGCCGTCCAACCAACTCA CGTCGATCCAATGCCGTCC-3’ 3’-GCGTCGGCAGGTTGGTTGAGTCGCAGCTAGGTTACGGCAGG-5’ P-21 5’-CGCAGCCGTCCAACCAACTCA-3’

Table 2.1: Pol µ DNA substrates. The underlined letter denotes the templating base.

2.3.2. Reaction buffer optimization.

Using a rapid chemical quench apparatus (KinTek, PA) (89), a preincubated solution of 40 nM 5’-radiolabeled D-1 and 200 nM Pol µ was mixed with 100 µM dTTP to initiate the reaction. The reactions were quenched by the addition of 0.37 M EDTA at various time intervals. For optimization reactions which were too slow to be efficiently performed using the rapid chemical quench, assays were carried out manually. In the latter case, an aliquot of the reaction mixture (10 µl) was withdrawn after various times and quenched with 0.37 M EDTA (40 µl). The NaCl optimization buffer consisted of 50 mM HEPES (pH 7.5 at 37 °C), 10 mM MgCl2, 0.2 mM EDTA, 5 mM DTT, 0.1 mg/mL

BSA, 10% glycerol, and varying concentrations of NaCl ranging from 0 to 420 mM. The

MgCl2 optimization buffer consisted of 50 mM HEPES (pH 7.5 at 37 °C), no additional

25 NaCl, 0.2 mM EDTA, 5 mM DTT, 0.1 mg/mL BSA, 10% glycerol, and varying

concentrations of MgCl2 ranging from 2.5 to 50 mM. The pH optimization buffer

consisted of 8.75 mM MgCl2, no additional NaCl, 0.2 mM EDTA, 5 mM DTT, 0.1

mg/ml BSA, 10% glycerol, and either 25 mM MES-NaOH for pH 6.0, 25 mM HEPES

for pH 6.8 to 9.0 at 37 °C, or 25 mM glycine-NaOH buffer for pH 10.0. The optimized

reaction buffer M consists of 50 mM HEPES (pH 7.8 at 37 °C and 8.0 at 25 °C), 12 mM

NaCl, 8.75 mM MgCl2, 0.2 mM EDTA, 5 mM DTT, 0.1 mg/mL BSA, and 10% glycerol.

Unless noted otherwise, all concentrations reported in this document refer to final

concentrations after mixing.

2.3.3. Enzyme stability assays.

In order to assess the stability of Pol µ, solutions containing 30 nM 5’-32P-labeled

D-1 and 150 nM Pol µ in buffer M were incubated at either 25 °C or 37 °C for various

periods of time. After times ranging from 10 minutes to 3 hours, an aliquot of the

incubated mixture was withdrawn and mixed with 100 µM dTTP. One minute after

reaction initiation with dTTP, each corresponding reaction mixture (10 µl) was quenched

with 0.37 M EDTA (40 µl). Quenched reaction products were then separated and

analyzed (see Section 2.3.6 below).

2.3.4. Substrate specificity measurements.

A solution of 150 nM Pol µ and 30 nM 5’-32P-labeled D-1 (or D-6, or D-7, or D-

8, or P-21) was preincubated in buffer M for 5 min at 25 °C. Nucleotide incorporation

was initiated by addition of increasing concentrations of dNTP (0.25 to 300 µM) to the reaction mixture. Each reaction (10 µl) was manually quenched with 0.37 M EDTA (40

26 µl) at specific time points ranging from 6 to 600 seconds for correct incoming nucleotide

and from 3 minutes to 3 hours for an incorrect incoming nucleotide. The reactions were

subsequently analyzed as described below (Section 2.3.6).

2.3.5. DNA trap experiment.

To qualitatively confirm distributive DNA polymerization by Pol µ, a

preincubated solution of 100 nM Pol µ and 100 nM D-8 (Table 1) in buffer M were

rapidly mixed with a solution containing both 100 µM dNTPs and 2.44 mg/ml activated calf thymus trap DNA. The reaction was quenched after 30 minutes in 0.37 M EDTA before polyacrylamide gel separation under conditions described below (Section 2.3.6).

2.3.6. Product and data analysis.

All reaction products were separated by denaturing polyacrylamide gel electrophoresis (PAGE) (17% acrylamide, 7 M urea) in 1x TBE running buffer. Products and unreacted primer were subsequently detected and quantitated using a

PhosphorImager 445 SI and ImageQuant 5.0 software (GE Healthcare, UK). All data were fit by nonlinear regression using KaleidaGraph software (Synergy Software, PA).

The concentration of products was plotted against reaction time and the resulting time

course fit to a single exponential equation: [product] = A[1 – exp(-kobst)], where A represents initial concentration of Pol µ•DNA binary complex, and kobs represents the

observed single turnover rate. The observed single turnover rates were subsequently plotted against their corresponding nucleotide concentrations, and the data were fit to the

hyperbolic equation: kobs = kp[dNTP]/(Kd + [dNTP]), to obtain the equilibrium

dissociation constant, Kd, and the maximum incorporation rate, kp.

27 32 Data from the [α- P]dNTP incorporation experiment in the absence of trap dNTP

(nonradioactive) were fit to a burst equation (90-92): [Product] = {[kpk1t/(kp + k1)] +

2 [kp/(kp + k1)] {1 – exp[-(kp + k1)t]}}[E0], to obtain the DNA dissociation rate constant, k1,

and the observed nucleotide incorporation rate constant, kp. [E0] is the initial enzyme

active site concentration.

2.4 Pol β Experimental Details.

2.4.1. DNA substrates.

The sequences of primer/template DNA substrates used in the Pol β studies are listed in

Table 2.2. In the text, the notation X:Y indicates [template base]:[incoming dNTP].

Custom synthesized oligomers were purified by denaturing PAGE (15-18% acrylamide, 7

M urea), and desalted using a reverse phase C18 cartridge. Oligomer concentrations were

determined by UV absorbance at 260 nm with the following extinction coefficients (M-

1cm-1): 19-mer primer, ε = 166,100; 18-mer primer, ε = 158,900; 36AP(T) template, ε =

335,200; 36AP(C) template, ε = 334,900; 36AP(A) template, ε = 341,100; 35AP(G)

template, ε = 327,100; 16-mer downstream primer, ε = 152,800; and 15-mer downstream

primer, ε = 144,800. Primer, template, and downstream primer (where applicable) were

annealed in a 1.1:1:1.2 molar ratio, respectively. Downstream primers utilized were 5’-

phosphorylated. DNA primers used in the chemical quench experiments were 32P-labeled

32 at 5’-end using T4 polynucleotide kinase and [γ- P]ATP (4500 Ci/mol) according to the

manufacturer’s protocol. The labeled primers were separated from unreacted ATP using

a G-25 microspin column.

19/36AP(T)/15 5’–GCCTCGCAGCCGTCCAACC AGTCACCTCAATCCA-3’ 3’–CGGAGCGTCGGCAGGTTGGTÃTCAGTGGAGTTAGGT–5’ 19/36AP(C)/15 5’–GCCTCGCAGCCGTCCAACC AGTCACCTCAATCCA-3’ 3’–CGGAGCGTCGGCAGGTTGGCÃTCAGTGGAGTTAGGT–5’ 19/36AP(T)/16 5’–GCCTCGCAGCCGTCCAACC TAGTCACCTCAATCCA-3’ 3’–CGGAGCGTCGGCAGGTTGGTÃTCAGTGGAGTTAGGT–5’ 19/36AP(T) 5’–GCCTCGCAGCCGTCCAACC 3’–CGGAGCGTCGGCAGGTTGGTÃTCAGTGGAGTTAGGT–5’ 19/36AP(C) 5’–GCCTCGCAGCCGTCCAACC 3’–CGGAGCGTCGGCAGGTTGGCÃTCAGTGGAGTTAGGT–5’ 19/36AP(A) 5’–GCCTCGCAGCCGTCCAACC 3’–CGGAGCGTCGGCAGGTTGGAÃTCAGTGGAGTTAGGT–5’ 18/35AP(G) 5’–GCCTCGCAGCCGTCCAAC 3’–CGGAGCGTCGGCAGGTTGGÃTCAGTGGAGTTAGGT–5’

Table 2.2: Pol β DNA substrates. The underlined letter denotes the templating base; Ã denotes 2-aminopurine; dd (in the text) refers to dideoxy-terminated.

2.4.2. Reaction buffer composition.

Assay buffers consisted of 20 mM BisTrisHCl, 50 mM TrisHCl, 2 mM DTT, 10-

35 % (w/v) glycerol, with the ionic strength adjusted to 81 mM with KCl, free Mg2+ adjusted to 5-10 mM with MgCl2 (as noted in the figure legends), at the indicated pH

ranging from 6.6 to 8.8 (adjusted at 37 ˚C). Standard Pol β assay buffer contained 10%

glycerol – all deviations from this are noted accordingly in the figure legends. Free Mg2+ concentrations were calculated using the WEBMAXC STANDARD program at:

29 http://www.stanford.edu/~cpatton/webmaxcS.htm, using the appropriate pH, temperature,

ionic strength, and nucleotide triphosphate concentration.

2.4.3. Stopped-flow fluorescence assays.

Stopped-flow fluorescence assays were performed on an SX 18MV Stopped-

Flow apparatus (Applied Photophysics Ltd., UK). The excitation wavelength for 2-

aminopurine (2-AP) was 312 nm with a spectral bandpass of 4 nm. Fluorescence

emission was monitored using a 360 nm high pass filter (Corion, MA) on a logarithmic

timescale for 1 s for correct dNTP incorporation and up to 50 s for mismatched dNTP

incorporation (depending on exact conditions). A typical forward dNTP incorporation

reaction was initiated by rapid mixing of two solutions in degassed Pol β assay buffer:

Solution A, containing 400 nM DNA substrate preincubated with 1 µM DNA polymerase, and Solution B, containing suitable correct or incorrect dNTP. In all cases, both syringes contained equal concentrations of MgCl2, since it is known that this can

have a significant effect on observed fluorescence (79). In experiments employing

altered buffer viscosity, both stopped-flow solutions A and B contained the indicated

concentration of glycerol. A typical reverse closing reaction was initiated by rapid

mixing of two solutions in degassed Pol β assay buffer: Solution A, containing 400 nM

dideoxy-terminated DNA substrate preincubated with 1 µM DNA polymerase, suitable correct or incorrect dNTP, and 5-10 mM MgCl2, and Solution B, containing 20-40 mM

K+EDTA (as noted in the figure legends). Typically, for all reactions a minimum of 7

runs were performed and averaged prior to data analysis.

30 2.4.4. Rapid chemical quench.

Follow-up of stopped-flow fluorescence experiments were performed on a

KinTek (State College, PA) rapid chemical quench instrument (89). The rapid chemical

quench reactions were initiated exactly as described in Section 2.4.3 for a typical forward

dNTP incorporation reaction, except that the upstream primer was 5’-radiolabeled.

Reactions were quenched at various time intervals with 0.6 M EDTA, and the products

visualized by denaturing PAGE followed by autoradiography using a STORM 840

PhosphorImager and quantitated using ImageQuant 5.0 software (GE Healthcare, UK).

2.4.5. Steady-state fluorescence.

Steady-state fluorescence experiments were performed on an Aminco-Bowman

Series 2 Luminescence Spectrometer. 2-AP was excited at 312nm, and the emission

spectra recorded from 330-500 nm with a step size of 1 nm. In Pol β assay buffer,

reactions consisted of 200 nM dideoxy-terminated DNA substrate, 500 nM DNA

polymerase, and either correct (300 µM ) or incorrect (5 mM) dNTP.

2.4.6. Product and data analysis.

Rapid chemical quench data were fit using Sigma Plot 9.0 (Jandel Scientific, CA)

−kquench t to a single exponential equation: [DNA n+1 ] = A(1− e ) , where A is amplitude and

kquench is the observed rate constant of single-nucleotide incorporation. Stopped-flow fluorescence traces (> 2 ms) were fit using Applied Photophysics software to a double

−k t −k t exponential equation: fluorescence = A1e fast + A2e slow + C , where A1 and A2

represent corresponding amplitudes, C is an offset constant, and kfast and kslow are the

observed rate constants for the fast and slow phases of the fluorescence transition,

respectively. For fast phase analysis, kfast was plotted as a function of dNTP 31 concentration and the data fit to a hyperbolic equation with a nonzero intercept:

kfast = k2 [dNTP]/(Kd + [dNTP]) + y0 . According to Scheme 2.1, k2 is the microscopic rate constant for the forward conformational closing step, Kd is the thermodynamic dissociation constant for dNTP (Kd = k-1 /k1), and y0 is a function of both microscopic

2+ constants k-2 and k4, as well as the binding constant for catalytic Mg as further elaborated in the appendix of Ref. (93). For slow phase analysis, kslow was plotted as a function of dNTP concentration and the data fit to hyperbolic equation:

kslow =k pol[dNTP]/(K d,app + [dNTP]) , where kpol represents the maximum rate of nucleotide incorporation and Kd,app represents the apparent dNTP dissociation constant.

Reverse closing traces (> 2 ms) were fit to a single exponential equation:

−k t fluorescence = Ae rc + C , where krc is the observed rate of fluorescence decay. The corresponding amplitude is represented by A and C is an offset constant.

NMPM k1 k2 k3 k4 k5 k6 k7 EDn EDnNE'DnNE'DnNM E'Dn+1PM E'Dn+1P EDn+1P EDn+1 k-1 k-2 k-3 k-4 k-5 k-6 k-7

Scheme 2.1: Pol β kinetic scheme of single-nucleotide incorporation. (Same as Figure 1.6, except microscopic rate constants are labeled here.) E = Pol β in open finger conformation; E’ = Pol β in closed finger conformation; Dn = DNA; Dn+1 = DNA elongated by addition of one nucleotide; N = M•dNTP; M = catalytic metal ion; P = M•PPi.

CHAPTER 3

PRE-STEADY-STATE KINETIC ANALYSIS OF THE FIDELITY OF HUMAN DNA POLYMERASE µ

3.1 Reaction Buffer Optimization.

In an in vitro system, evaluation of enzymatic activity in buffer is largely sensitive

to factors including ionic strength, pH, and the presence, identities, and concentrations of

additional buffer components such as enzyme cofactors, metals, or stabilizers. For a

polymerase, an optimal reaction buffer will allow incorporation of nucleotides with a

maximal polymerization rate and full substrate turnover. To determine the conditions

under which Pol µ carries out optimal catalysis, the MgCl2 concentration, NaCl

concentration, and pH of the buffer were varied individually while the concentrations of

all other buffer components remained constant. Each optimization reaction was initiated

by mixing a preincubated solution of 40 nM 5’-32P-labeled D-1 primer/template (Table

2.1) and 200 nM Pol µ with 100 µM dTTP in buffer at 37 °C. The reactions were quenched by 0.37 M EDTA at various time intervals. Experiments were carried out either manually or by using a rapid chemical quench apparatus (see Chapter 2, Section

2.3.2). To cover a wide pH range throughout the pH optimization experiments, three different buffers were used. The final optimized reaction buffer M consisted of 12 mM

33 NaCl (Figure 3.1A), 8.75 mM MgCl2 (Figure 3.1B), and a pH of 7.8 at 37 °C (Figure

3.1C). For additional stabilization of the enzyme in these reactions, 0.1 mg/ml BSA and

10% glycerol were also included in the optimized reaction buffer M.

Figure 3.1: Pol µ reaction buffer optimization. The optimal levels of MgCl2 concentration, NaCl concentration, and buffer pH were examined under single turnover conditions at 37 °C by rapidly mixing 100 µM correct nucleotide dTTP with a preincubated solution of 40 nM D-1 and 200 nM Pol µ. (A) The NaCl concentration was varied from 12 to 432 mM. (B) The MgCl2 concentration was varied from 2.5 to 50 mM. (C) The pH of the reaction buffer was varied from 6.0 to 10.0. Single turnover rates were measured in 25 mM MES-NaOH buffer for pH 6.0, 25 mM HEPES buffer for pH 7.0 to 9.0, and 25 mM glycine-NaOH buffer for pH 10.0.

Figure 3.1 36 3.2 Enzyme Stability Assays.

A relatively slow rate of nucleotide incorporation observed in the optimization

assays indicated that in order to accurately ascertain the fidelity of Pol µ, we needed this

enzyme to retain its full activity over a span of several hours. In order to determine

whether the recombinant Pol µ was fully active at 37 °C for the desired length of time in

buffer M, enzyme stability assays were conducted. A solution of 30 nM D-1 and 150 nM

Pol µ in buffer M was incubated at 37 °C. After incubation time intervals ranging from

0-200 min, 10 µl of the enzyme-DNA solution was withdrawn and mixed with 100 µM

dTTP to initiate the reaction. Each reaction was stopped after exactly 1 min (~7t1/2) at 37

°C which was sufficient time to allow a single turnover of dTTP incorporation to be completed (see below). If the enzyme remains stable during incubation, the product concentration would be expected to reach full reaction amplitude (~30 nM) for all

incubation periods. Surprisingly, the product formation began to drop dramatically after only 20 minutes of incubation at 37 °C (Figure 3.2). This suggested that the purified Pol

µ would not be stable long enough to complete the slow incorrect dNTP incorporation

assays at 37 °C.

37 In order to achieve greater stability of this enzyme over a greater period of time, we lowered the incubation temperature and conducted the same enzyme stability assay at

25 °C. Interestingly, the product concentration remained close to full amplitude (equal to the initial E•DNA concentration) over three hours at this temperature (Figure 3.2). In addition, we performed the similar enzyme stability assay with 5-fold molar excess of D-

1 over Pol µ (data not shown) and the product formation patterns were similar as shown in Figure 3.2. Therefore, all subsequent nucleotide incorporation assays were performed at 25 °C.

Figure 3.2: Pol µ stability assay. Solutions contained 30 nM D-1 preincubated with 150 nM Pol µ in optimized reaction buffer M was incubated at either 25 °C ({) or 37 °C (z). Aliquots were withdrawn after various incubation times ranging from 10 to 200 minutes and reacted with 100 µM dTTP for 1 minute at 37 °C before quenching by 0.37 M EDTA. The product concentration was plotted against the incubation time.

3.3 Rapid Equilibrium of Nucleotide Binding to the Binary E•DNA Complex.

Scheme 3.1 is a proposed simplified kinetic mechanism for single nucleotide

incorporation into a synthetic primer/template substrate catalyzed by Pol µ. This minimal

mechanism is shared by all DNA polymerases which have been studied so far (92-100)

including two other X-family members Pol β (93) and Pol λ (92). In Scheme 3.1, the

initially formed binary complex of enzyme and DNA (E•DNAn) binds an incoming

nucleotide (dNTP) to form a ternary complex E•DNAn•dNTP. During catalysis, the

DNA primer is elongated by one nucleotide, pyrophosphate (PPi) is produced, and the

product-containing ternary complex, E•DNAn+1•PPi, is then formed. After catalysis, the

enzyme dissociates from E•DNAn+1•PPi to initiate subsequent turnovers. Scheme 3.1 sets

a kinetic basis for evaluating substrate selection in a template-directed polymerization

event catalyzed by Pol µ.

kon kp k1 E•DNAn + dNTP E•DNAn•dNTP E•DNAn+1•PPi E + DNAn+1 + PPi koff

Scheme 3.1. Minimal kinetic mechanism for a DNA polymerase.

To measure the binding affinity of an incoming nucleotide in Scheme 3.1, the binding

(kon) and dissociation (koff) steps have to be much faster than the nucleotide incorporation

(kp) such that the reversible binding of dNTP is at rapid equilibrium. To verify this rapid equilibrium, a preincubated solution of Pol µ, 5-fold unlabeled D-8, and [α-32P]dGTP (20

µM) in the optimized buffer M lacking Mg2+ and containing EDTA (1 mM) was mixed

39 with a solution of excess unlabeled dGTP (2 mM) and Mg2+ (8.75 mM) for various

reaction times prior to being quenched with EDTA. Although Pol µ was purified and

stored in the absence of divalent metal ions, we used additional 1 mM EDTA present in the preincubated E•DNA•[α-32P]dNTP solution to chelate any contaminant divalent

cations carried over from the protein purification. This concentration of EDTA was

shown to be sufficient to prevent any product formation in this preincubated solution

(data not shown). If the dissociation of [α-32P]dNTP from the E•DNA•[α-32P]dNTP

ternary complex was much faster than the polymerization we would expect to see very

little [α-32P]-labeled product formation due to the unfavorable kinetic partitioning and

large molar excess of unlabeled dGTP that would trap dissociated [α-32P]dNTP. This

assay, performed by labmate Kevin Fiala, resulted in a time course (z) showing background-level radio-labeled product (less then 0.01 nM at the longest reaction time in

Figure 3.3). This suggested the nucleotide dissociation rate constant was indeed much faster than the polymerization rate constant (koff >> kp). In addition, the product concentration was significantly less when compared to a similar experiment performed

-1 with truncated Pol λ in which koff is determined to be 300 s (92). We therefore placed a

lower limit on the dissociation rate constant of dGTP from Pol µ•DNA to be >300 s-1. In

addition, the affinity of dGTP (Kd) was measured to be 1.8 µM (Table 3.1). The association rate constant of the binding of dGTP to the Pol µ•D-8 binary complex kon =

8 -1 -1 koff/Kd > 1.7 x 10 M s was thus calculated from the values of koff and Kd. This

suggested the binding of a nucleotide to the binary complex Pol µ•DNA was under

diffusion-control or at rapid equilibrium.

40 3.4 Rapid DNA Dissociation and Slow Polymerization.

Lack of product formation shown above could be due to preincubation of the

E•DNA•dNTP ternary complex in the absence of Mg2+. To ensure this was not the case,

another control experiment was performed. In this experiment, a preincubated solution of

Pol µ (60 nM), unlabeled D-8 (300 nM), [α-32P]dGTP (20 µM), and EDTA (1 mM) in the

optimized buffer lacking Mg2+ was mixed with a solution containing Mg2+, and no

additional unlabeled trap dGTP for various reaction times. This experiment yielded a time course of product formation () (Figure 3.3) that was similar to that obtained when

a preincubated Pol µ•DNA binary complex in the presence of Mg2+ reacted with

Mg•dNTP•Mg2+ (data not shown). This suggested the nucleotide incorporation was not

affected by preincubating the E•DNA•dNTP ternary complex in the absence of Mg2+.

Moreover, lack of a burst phase in the data () suggested the k1 value was significantly

larger than the kp value (90, 91). We attempted to fit the data () to a burst equation (see

Chapter 2, Section 2.3.6) to obtain observed kp and k1 (Scheme 3.1) as we have done

previously with Pol λ (92), but failed due to lack of convergence of non-linear regression

analysis.

Figure 3.3: Incorporation of [α-32P]dGTP into D-8. In the first time course, a preincubated solution of Pol µ (60 nM), unlabeled D-8 (300 nM), [α-32P]dGTP (20 µM), and 1 mM EDTA in the absence of Mg2+ was reacted with a solution containing a large molar excess of unlabeled dGTP (2 mM) and Mg2+ (9.75 mM) for various reaction times (z). In the second time course, a preincubated solution of Pol µ (60 nM), unlabeled D-8 (300 nM), [α-32P]dGTP (20 µM), and 1 mM EDTA in the absence of Mg2+ was reacted with a solution containing Mg2+ (9.75 mM) with no additional unlabeled dGTP for various reaction times ().

3.5 Substrate Specificity of Incoming Correct dNTP.

The lack of a burst phase in the time course of product formation () in Figure

3.3 suggested the observed reaction rate was not the true nucleotide incorporation rate in the first turnover and was complicated by the fast dissociation of the enzyme and DNA binary complex (E•DNA) (91). To eliminate the complication from subsequent turnovers, we measured the true nucleotide incorporation rate under single-turnover conditions. These single-turnover experiments were performed with the enzyme in molar

42 excess over DNA to allow the direct observation of the conversion of 21/41-mer to 22-

mer/41-mer in a single pass of D-1 through the enzymatic pathway (Figure 3.4) (91).

Figure 3.4: Time course of product formation for correct dTTP incorporation into D-1. A preincubated solution of 30 nM D-1 and 150 nM Pol µ was rapidly reacted with increasing concentrations of 28 µM dTTP for various time intervals. The reactions were stopped by 0.37 M EDTA and analyzed (as described in Chapter 2, Section 2.3.6).

Solutions of 30 nM radiolabeled D-1 and 150 nM Pol µ were incubated at 25 °C in buffer

M, and then reacted with increasing concentrations of the correct incoming nucleotide,

dTTP. Aliquots of the reactions were quenched with EDTA at various time intervals.

The products were separated from unextended 21-mer primer by gel electrophoresis, and

the results quantitated using PhosphorImaging technology. The concentration of product

was plotted against time and the data were fit to a single exponential equation to obtain an observed single turnover rate, kobs, for each concentration of dTTP (Figure 3.5A). The

reaction amplitudes achieved were close to 100% in all time courses. These reaction

amplitudes were not improved with a larger fold excess of Pol µ over DNA (data not

shown). This suggested that the 5-fold excess of enzyme over DNA was enough to 43 ensure that nearly all DNA 21/41-mer molecules were bound by the enzyme. The single

turnover rates were then plotted against dTTP concentration and the curve was fit to a

hyperbolic equation (see Chapter 2, Section 2.3.6), to obtain a dissociation constant (Kd)

-1 of 1.4 ± 0.1 µM and a maximum rate of polymerization (kp) of 0.076 ± 0.001 s at 25 °C

(Figure 3.5B).

Similar single-turnover experiments for the remaining correct nucleotide incorporations were performed with dCTP incorporation against D-6, dATP incorporation against D-7, and dGTP incorporation against D-8 at 25 °C. The measured values of Kd and kp are listed in Table 3.1. The substrate specificity of the incorporation

of each correct nucleotide, defined by pre-steady-state kinetics as kp/Kd (91), was

calculated and listed in Table 3.1.

Figure 3.5: Substrate specificity of incoming correct nucleotide. (A) A preincubated solution of 30 nM D-1 and 150 nM Pol µ was rapidly reacted with increasing concentrations of dTTP (0.25 µM, z; 0.50 µM, {; 1 µM, ; 2 µM, ; 8 µM, S; 18 µM, U; 28 µM, ¡; 50 µM, ; 80 µM, T) for various time intervals. The reactions were stopped by 0.37 M EDTA. A single exponential fit yields a kobs for each dTTP concentration. (B) A hyperbolic fit of the dTTP concentration versus kobs yields a Kd of -1 1.4 ± 0.1 and a kp of 0.076 ± 0.001 s .

Sugar Nucleotide K (µM) k (s-1) k /K (µM-1s-1) Fidelitya d p p d Selectivityb Template A (D-1) dTTP 1.4 ± 0.1 7.6 x 10-2 ± 1 x 10-3 5.6 x 10-2 1 2074 rUTP 86 ± 10 2.3 x 10-3 ± 9 x 10-5 2.7 x 10-5 4.8 x 10-4 dATP 80 ± 8 9.7 x 10-5 ± 4 x 10-6 1.2 x 10-6 2.2 x 10-5 dCTP 98 ± 24 3.0 x 10-4 ± 3 x 10-5 3.0 x 10-6 5.5 x 10-5 dGTP 81 ± 21 7.4 x 10-5 ± 7 x 10-6 9.1 x 10-7 1.6 x 10-5

Template G (D-6) dCTP 0.35 ± 0.03 2.2 x 10-2 ± 4 x 10-4 6.4 x 10-2 1 492 rCTP 56 ± 13 7.3 x 10-3 ± 5 x 10-4 1.3 x 10-4 2.0 x 10-3 dATP 89 ± 11 6.1 x 10-5 ± 3 x 10-6 6.9 x 10-7 1.1 x 10-5 dGTP 12 ± 3 4.0 x 10-5 ± 2 x 10-6 3.3 x 10-6 5.2 x 10-5 dTTP 51 ± 11 3.7 x 10-5 ± 3 x 10-6 7.1 x 10-7 1.1 x 10-5

Template T (D-7) dATP 0.76 ± 0.34 6.1 x 10-3 ± 5 x 10-4 8.0 x 10-3 1 10959 rATP 302 ± 24 2.2 x 10-4 ± 7 x 10-6 7.3 x 10-7 9.1 x 10-5 dCTP 24 ± 10 1.5 x 10-4 ± 1 x 10-5 6.1 x 10-6 7.6 x 10-4 dGTP 7.3 ± 1.5 3.0 x 10-5 ± 1 x 10-6 4.1 x 10-6 5.1 x 10-4 dTTP 45 ± 7 2.3 x 10-5 ± 1 x 10-6 5.0 x 10-7 6.3 x 10-5

Template C (D-8) dGTP 1.8 ± 0.5 5.5 x 10-2 ± 3 x 10-3 3.0 x 10-2 1 6122 rGTP 45 ± 7 2.2 x 10-4 ± 8 x 10-6 4.9 x 10-6 1.6 x 10-4 dATP 90 ± 30 5.3 x 10-5 ± 8 x 10-6 6.0 x 10-7 2.0 x 10-5 dCTP 135 ± 28 2.8 x 10-4 ± 3 x 10-5 2.1 x 10-6 6.7 x 10-5 dTTP 97 ± 14 3.6 x 10-5 ± 2 x 10-6 3.7 x 10-7 1.2 x 10-5 a Calculated as (kp/Kd)incorrect/[(kp/Kd)correct + (kp/Kd)incorrect]. b Calculated as (kp/Kd)dNTP/(kp/Kd)rNTP.

Table 3.1: Pre-steady-state kinetic parameters of human DNA polymerase µ.

46 3.6 Substrate Specificity of Incoming Incorrect dNTP.

To evaluate if Pol µ also incorporated incorrect nucleotides, we tested all 12 possible mismatched dNTP incorporations into D-1, D-6, D-7, and D-8 (Table 2.1). With

a reaction time of 3 hours, all possible misincorporations did occur (Figure 3.6). To

determine the incorporation fidelity of Pol µ, we used the same assay described above to

determine the substrate specificity of each misincorporation (data not shown). The

corresponding kinetic parameters were recorded in Table 3.1. Due to both weak binding

affinity and slow misincorporation, the substrate specificity for an incorrect dNTP was

three to five orders of magnitude lower than the value of a correct nucleotide. The

observed tighter binding of mismatched dGTP to the enzyme•D-7 complex (Kd = 7.3 µM) when compared to other mismatched base pairs is likely due to the formation of relatively stable G:T wobble base pair. In addition, the misincorporation of dCTP into D-1, D-7, and D-8 was more efficient than other misincorporations in Figure 3.6, and possessed slightly higher substrate specificity (Table 3.1). These observations were possibly because dCTP skipped the template thymine base and base paired with the next template base guanine. This hypothesis is supported by the fact that Pol µ causes frequent frameshift mutation in vitro (35).

47 The fidelity, defined as (kp/Kd)incorrect/[(kp/Kd)correct + (kp/Kd)incorrect], was calculated

for all 16 possible nucleotide incorporations (Table 3.1). Thus, the fidelity of dNTP

polymerization catalyzed by human DNA Pol µ was determined to be 10-4-10-5 with

normal primer/template DNA substrates. Since Pol µ does not possess 3’→5’

exonuclease activity, this enzyme is thereby considered to be a polymerase with moderate

fidelity.

Figure 3.6: Comparison of matched and mismatched dNTP incorporation by Pol µ. DNA substrates D-1, D-6, D-7, and D-8 (30 nM each) preincubated with Pol µ (150 nM) were mixed with 150 µM dATP (lanes A), 150 µM dCTP (lanes C), 150 µM dGTP (lanes G), 150 µM dTTP (lanes T), 4x150 µM dNTPs (lanes N), and no dNTP (lane B) as indicated for 3 hours. 48 3.7 Substrate Specificity of Incoming Matched rNTP.

TdT, the homolog of Pol µ, has been previously shown to incorporate both dNTPs

and rNTPs (101, 102). To evaluate if Pol µ possesses similar activity, we performed

single matched rNTP incorporation into D-8 and the elongation of the DNA primer

indicated Pol µ has RNA polymerase activity (Figure 3.7). Similar results were obtained

with D-1, D-6, and D-7(data not shown). However, consecutive incorporations of rNTPs

were very slow in the presence of all rNTPS (Figure 3.7). These results have been

observed previously by others (36, 37). Additionally, misincorporations of rNTPs into

D-8 (Figure 3.7) were much less efficient than the misincorporations of dNTPs (Figure

3.6). Thus, only matched rNTPs were kinetically competent enough to compete against

dNTPs in the elongation of a DNA primer/template substrate by Pol µ. The substrate

specificity of each mismatched rNTP was thereby not measured.

Figure 3.7: Pol µ template-dependent ribonucleotide incorporation. A preincubated solution of Pol µ (150 nM) and D-8 (30 nM) was mixed with 150 µM rATP, 150 µM rCTP, 150 µM rGTP, 150 µM rUTP, 4x150 µM rNTPs (lane N), and no rNTPs (lane B) as indicated for 3 hours.

49 To determine the sugar selectivity of Pol µ, we measured substrate specificity

(Table 3.1) of each of the four matched rNTPs using similar assays as described above.

For example, Figure 3.8 shows Pol µ incorporation of rGTP into D-8 with an equilibrium dissociation constant (Kd) of 45 ± 7 µM and a maximum rate of polymerization (kp) of

0.000220 ± 0.000008 s-1. The fidelity of misincorporations of matched rNTPs into DNA

primer/template substrates was calculated to be in the range of 10-3-10-5 (Table 3.1).

Notably, the pyrimidines especially CTP had higher substrate specificity than purines

(Table 3.1). The preference was probably due to the next template base Guanine which

base paired with the pyrimidines to form the second basepairs G:C or G:U (wobble).

Interestingly, the incorporation efficiency (kp/Kd) of rNTPs was higher or similar to that

of mismatched dNTPs (Table 3.1). This suggests Pol µ misincorporates matched rNTPs at least as frequently as it misincorporates unmatched dNTPs.

Figure 3.8: Substrate specificity of incoming correct ribonucleotide. (A) Increasing concentrations of rGTP (2 µM, z; 8 µM, {; 25 µM, ; 75 µM, ; 125 µM, S; 200 µM, U; 350 µM, ¡; 500 µM, ; 800 µM, T) were rapidly mixed with a solution of preincubated D-8 (30 nM) and Pol µ (150 nM). Reactions were quenched at various time increments with 0.37 M EDTA. A single exponential fit yields a kobs for each rGTP concentration. (B) A hyperbolic fit of each rGTP concentration versus the corresponding -1 kobs yields a Kd of 45 ± 7 and a kp of 0.000220 ± 0.000008 s .

51 Surprisingly, the sugar selectivity, defined as the ratio of substrate specificity of a matched dNTP and a matched rNTP (kp/Kd)dNTP/(kp/Kd)rNTP, was calculated to be 492,

2074, 6122, and 10959 for the incorporation of dCTP/rCTP, dTTP/UTP, dGTP/rGTP, and dATP/rATP (Table 3.1). The sugar selectivity is significantly higher than the previously estimated values 1.4-11 (36) and 1.34-11.04 (37) obtained with single-

nucleotide gapped DNA substrates.

3.8 Weak Strand-Displacement Activity.

To probe if Pol µ possesses strand-displacement activity, we performed the

incorporation of single dNTP or all four dNTPs into a single nucleotide gapped DNA D-

8g (Table 2.1). The multiple incorporations in the presence of all dNTPs (Figure 3.9)

suggested Pol µ partially displaced the downstream primer 19-mer after filling the gap.

This activity was not due to contamination of the primer/template 21/41-mer since the

DNA was annealed with a ratio of 1:2:2 for 21-mer:19-mer:41-mer. If there was

significant amount of unwanted 21/41-mer in the reaction mixture, the product bands of

22-mer to 24-mer should not be more intense than longer product bands after 3 hour

polymerization in the presence of dNTPs (Figure 3.9). Small amount of products longer

than 25-mer suggested the strand-displacement activity of Pol µ was relatively weak.

This strand-displacement activity of Pol µ was not detected by Ruiz et al. (37). The

reason why we detected this weak activity is probably due to long reaction time and high

enzyme concentration used in our studies. The energy source for this weak strand-

displacement activity is probably derived from the net favorable free energy of nucleotide

incorporation in addition to thermal breathing of the basepairs. The slow nucleotide 52 incorporation by Pol µ and fast DNA dissociation result in inefficient polymerization and

little favorable free energy which leads to weak strand-displacement activity.

Figure 3.9: Pol µ deoxynucleotide incorporation into a single-nucleotide gapped DNA substrate. Strand displacement activity of Pol µ was assessed by reacting 100 µM dATP, 100 µM dCTP, 100 µM dGTP, 100 µM dTTP and 4x100 µM dNTPs (lane N1) with a preincubated solution containing D-8g (30 nM) and Pol µ (150 nM). As a control, a non-gapped D-8 substrate reacted with 4x100 µM dNTPs was indicated in lane N2. All reactions were quenched after 3 hours.

3.9 Distributive Elongation of a Primer/Template Substrate.

The kinetic studies shown in Figure 3.3 suggest that Pol µ has a higher DNA dissociation rate constant than single nucleotide incorporation rate constant, supporting 53 that Pol µ is a distributive polymerase as observed previously (7, 8, 37). To provide

further evidence for this conclusion, a pre-incubated solution of radio-labeled D-8 and

Pol µ in buffer M was reacted with a solution of dNTPs in buffer M containing high

concentration of activated calf thymus DNA (see Chapter 2, Section 2.3.5). If D-8

dissociated rapidly from Pol µ, any dissociated Pol µ molecules would be trapped by large molar excess of calf thymus DNA molecules, leading to no product formation.

Conversely, in the case of a slow DNA dissociation, the primer 21-mer should be elongated to longer products. The results shown in Figure 3.10 qualitatively demonstrated that DNA dissociation from Pol µ is relatively fast. In fact, product 22-mer was barely seen after 30 minutes incubation in the presence of the trap, in contrast to the extensive product formation pattern in the absence of the trap.

Figure 3.10: Pol µ DNA trap assay. Preincubated Pol µ (100 nM) and D-8 (100 nM) were mixed with dNTPs (100 µM each) in the presence (+) or absence (-) of 2.44 mg/mL activated calf thymus trap DNA for 30 minutes.

54 3.10 Pol µ Displays Template-independent Terminal Transferase Activity in the

Presence of Mg2+.

Previous studies have reported the terminal transferase activity of Pol µ, yet have

predominantly utilized Mn2+ as the divalent metal (8, 103). In order to better understand

the template-independent activity of this enzyme, Pol µ’s ability to incorporate dNTPs

and rNTPs in the presence of Mg2+ was examined. A comparison of the relative

processivity of Pol µ incorporation of dNTPs versus rNTPs into a single-stranded DNA

substrate, P-21 (Table 2.1), demonstrated that while Pol µ can, in fact, incorporate ribonucleotides, although the rNTP incorporation causes eventual premature chain termination (Figure 3.11). This is consistent with the reported inability of many DNA or

RNA polymerases, including TdT (101), to catalyze multiple insertions of the “wrong” sugar (104-108). It is anticipated that multiple rNTP incorporation alters the nucleic acid structure to the extent that Pol µ can no longer accommodate the substrate terminus (108).

A qualitative comparison of the ability of Pol µ and TdT to incorporate both dNTPs

(Figure 3.12) and rNTPs (Figure 3.13) revealed that under identical conditions Pol µ is

much less efficient than TdT in extending a single-stranded substrate with either. Note

the short reaction times for TdT (3.5 to 15 minutes) in comparison with Pol µ (180

minutes), as described in Figure legends 3.12 and 3.13. Since Pol β does not extend

single-stranded DNA substrates, it was used as a negative control (Figure 3.12, far right

lane) to ensure that the P-21 substrate was not forming a pseudo-template/primer through

intra- or inter- molecular base pairing. Since this control reaction shows no product

formation, we are confident that we are truly monitoring the template-independent

activity of Pol µ. Furthermore, the template-independent activity of Pol µ has also been

55 recently reported on homopolymeric single-stranded DNA (109) and single-strand DNA immobilized on streptavidin-agarose beads (110).

Figure 3.11: Processivity evaluation of Pol µ template-independent incorporation. A preincubated solution of Pol µ (100 nM) and single-stranded P-21 DNA (100 nM) were mixed with 4x100 µM dNTPs or 4x100 µM rNTPs for a total of 3 hours in the presence of 8.75 mM Mg2+. Reactions were quenched at time points of 0, 4, 8, 20, 30, 50, 70, 90, 110, 140, 180 minutes (with the 30 minute time point omitted for rNTP incorporation).

Figure 3.12: Template-independent dNTP incorporation by Pol µ. A preincubated solution of Pol µ (150 nM) and single-stranded P-21 DNA (30 nM) was mixed with 100 µM dATP, 100 µM dCTP, 100 µM dGTP, 100 µM dTTP, 4x100 µM dNTPs (lane N), or 0 µM dNTP (lane B) for 3 hours in the presence of 8.75 mM Mg2+. The corresponding reactions were carried out for TdT and Pol β, except that reactions were quenched after 3.5 minutes and 30 minutes, respectively. 57

Figure 3.13: Template-independent rNTP incorporation by Pol µ. A preincubated solution of Pol µ (150 nM) and single-stranded P-21 DNA (30 nM) was mixed with 300 µM rATP, 300 µM rCTP, 300 µM rGTP, 300 µM rUTP, 4x100 µM rNTPs (lane rN), or 0 µM rNTP (lane B) for 3 hours in the presence of 8.75 mM Mg2+. The corresponding reactions were carried out for TdT, except that reactions were quenched after 15 minutes.

3.11 Substrate Specificity of dNTP and rNTP Incorporation into Single-strand DNA.

To further characterize the template-independent activity of Pol µ in the presence

of Mg2+ we measured the substrate specificity of each of four incoming dNTPs (Figure

3.12) and four incoming rNTPs (Figure 3.13) into the single-stranded DNA substrate P-

21 (Table 2.1) using single-turnover conditions as described in previous sections. The corresponding parameters along with the sugar selectivity are reported in Table 3.2. The substrate specificity values for incorporation of both dNTP and rNTP into single-strand

DNA by Pol µ fell within the same range as those determined for mismatched incorporation into primer/template DNA (Table 3.1). The specificity for dNTP

58 incorporation into a single-stranded DNA substrate was about one order of magnitude

higher than that for rNTP incorporation, yielding a sugar selectivity range of 2.8-4.9. As

noted above (Figure 3.11 and Figure 3.13), multiple rNTP incorporations were

unfavorable for this enzyme. While all four dNTPs could be incorporated, Pol µ showed

a slight preference for incorporation of dCTP and dGTP into this particular substrate.

-1 -1 -1 b Nucleotide Kd (µM) kp (s ) kp/Kd (µM s ) Sugar Selectivity

dATP 89 ± 12 4.4 x 10-5 ± 2 x 10-6 4.9 x 10-7 ND

dCTP 102 ± 11 1.8 x 10-3 ± 8 x 10-5 1.8 x 10-5 3.1

dGTP 3.2 ± 0.6 1.4 x 10-4 ± 3 x 10-6 4.4 x 10-5 2.8

dTTP 45 ± 11 4.7 x 10-5 ± 3 x 10-6 1.0 x 10-6 4.9

rATPa ND ND ND -----

rCTP 451 ± 127 2.6 x 10-4 ± 4 x 10-5 5.8 x 10-7 -----

rGTP 36 ± 14 5.7 x 10-5 ± 4 x 10-6 1.6 x 10-6 -----

rUTP 217 ± 58 4.6 x 10-5 ± 5 x 10-6 2.1 x 10-7 -----

a ND indicates that there was not sufficient detectable product for quantitation. b Calculated as (kp/Kd)dNTP/(kp/Kd)rNTP.

Table 3.2: Pre-steady-state kinetic parameters of Pol µ incorporation into single- strand DNA.

3.12 Discussion and Conclusion.

To evaluate the proposition that human DNA polymerase µ acts as a mutase involved in somatic hypermutation (8), we overexpressed, purified, and determined the 59 fidelity of this polymerase for incorporations of both deoxynucleotides and

ribonucleotides into primer/template DNA using pre-steady-state kinetic methods under

single-turnover conditions. In the same capacity, we examined the ability of Pol µ to

extend single-stranded DNA and provided a qualitative comparison with TdT. Pol µ was

found to be sensitive to concentrations of Mg2+, salt, pH, and temperature, suggesting

that precautions should be taken when studying Pol µ, or any polymerase, in vitro.

Inefficient Polymerization on Primer/Template Substrates.

The substrate specificity (0.008 to 0.064 µM-1s-1) for correct dNTP incorporations by Pol µ (Table 3.1) is more than 1000-fold, 100-fold, and 10-fold lower than that of T7

DNA polymerase (a replicative polymerase) (78), Pol β (a repair polymerase) (111), and

yeast DNA polymerase η (a DNA lesion bypass polymerase) (95), respectively.

Moreover, the correct nucleotide incorporation rates (0.006-0.076 s-1) of Pol µ fall into the range of misincorporation rates by other DNA polymerases. These results suggest that Pol µ is a very inefficient DNA polymerase in the elongation of a primer/template substrate. However, it is possible that Pol µ has higher polymerase activity with different types of DNA substrates. Nick McElhinny and Ramsden have found that Pol µ has more than a 100-fold higher polymerase activity with single-nucleotide gapped DNA substrates, than with DNA primer/template substrates (36). Alternatively, in vivo Pol µ may interact with other proteins through its N-terminal BRCT domain resulting in enhanced polymerase activity and stability. This is further supported by the fact that deletion of the N-terminal BRCT domain of Pol µ completely abolishes its polymerase

60 activity (S. Sompalli and Z. Suo, unpublished results). Pol µ has also been found to associate with Ku and Ligase IV in vitro (34).

Fidelity of Deoxynucleotide Incorporation.

The deoxynucleotide incorporation fidelity of Pol µ was determined to be in the range of 10-4 to 10-5 with DNA primer/template substrate. This range was different from

the fidelity (10-2 to 10-5) estimated by Zhang et al. using DNA primer/template substrates

and steady-state kinetic methods (35). The difference in the fidelity is predominantly

due to the difficulty in the implementation of steady-state kinetic methods by Zhang et

al. (35), although different DNA substrates and different reaction temperatures (25 °C in our studies versus 30 °C in their studies (35)) may have contributed. Zhang et al. (35) use 50 fmol DNA substrate and a relatively significant amount of Pol µ (14 fmol) in order to observe sufficient product formation. The enzyme/DNA substrate ratio (28%) used in these studies is too high to maintain bona fide steady-state reaction conditions, therefore the constants they have used to define the fidelity are not true steady-state kinetic parameters. The fidelity Zhang et al. estimated is thus inaccurate. Moreover, the

“steady-state” kinetic parameters for several misincorporations are not reported by Zhang et al. (35), probably due to extremely slow product formation even under compromised steady-state reaction conditions. Therefore, the single-turnover method used in our studies has a clear advantage over steady-state kinetic methods, particularly in the studies of a slow enzyme like Pol µ (91). In addition, the fidelity and substrate specificity of Pol

µ (Table 3.1) are not sequence-dependent. The asymmetry of misincorporation

61 efficiency, e.g. G:A and A:G basepairs, previously observed with Pol β (111) was not found (Table 3.1).

Overall Fidelity, Sugar Selectivity, and Ribonucleotide Incorporation.

During the course of this study, two other groups (36, 37) also reported the templated incorporation of rNTPs by Pol µ which is consistent with our results shown in

Figure 3.7. Surprisingly, Pol µ incorporates matched rCTP, rUTP, and rGTP more efficiently than mismatched dNTPs (Table 3.1). Moreover, higher cellular concentrations of rNTPs over dNTPs in vivo will further increase the misincorporation frequency of matched rNTPs (112). Thus, for any DNA polymerase which is capable of incorporating rNTPs, the fidelity evaluation should be inclusive of both dNTPs and rNTPs. On the other hand, the incorporations of mismatched rNTPs are much less efficient than the matched rNTPs (Figure 3.7), suggesting these types of misincorporations are rare and may not need to be considered. Together, the data in

Table 3.1 suggest that the overall fidelity of nucleotide incorporation into a DNA primer/template by Pol µ is in the range of 10-3 to 10-5. However, further extension of an incorporated rNTP with either dNTPs (37) or rNTPs (Figure 3.7 and (37)) is very slow, leading to arrest of DNA synthesis.

The Tyr-Phe motif found in the α-helix M of the thumb subdomain of Pol β

(Tyr271-Phe272) (see Figure 1.2) and Pol λ (Tyr505-Phe506) provide a preference for strong ribonucleotide discrimination (55, 113). The corresponding motif in Pol µ is replaced with a Gly435-Trp436. Sequence alignment analysis coupled with site-directed mutagenesis have identified the amino acid residue Gly435 of Pol µ as the residue critically responsible for this enzyme’s lack of sugar discrimination during incorporation 62 into single-nucleotide gapped DNA. This is likely because the small side chain of

Gly435, unlike its bulkier counterpart Tyr271 in Pol β and Tyr505 in Pol λ, does not act

as a steric barrier for the 2’-OH of an incoming rNTP (37). Surprisingly, the sugar

selectivity of Pol µ with a DNA primer/template substrate, 492-10959 (Table 3.1), is

similar to E. coli DNA polymerase I (105, 114), but much larger than the selection

factors previously obtained with single-nucleotide gapped DNA substrates, 1.4-11 (36)

and 1.34-11.04 (37). Therefore, our results suggest that sugar selection by Pol µ utilizes

different structural mechanisms with gapped and non-gapped DNA substrates. With a

primer/template substrate, it is possible that a residue other than Gly433 uses its bulky

side chain to block the binding of the 2’-hydroxyl of an incoming rNTP. This further

suggests the active site structure of Pol µ is different in the presence of gapped or non-

gapped DNA substrates. To date, the ternary structure of Pol µ with DNA and dNTP has

been solved only for gapped DNA, therefore we can speculate that the interaction

between the DNA downstream primer and Pol µ might contribute to a structural and functional difference (115). Interestingly, Pol µ showed significantly relaxed sugar specificity, ranging from 2.8-4.9, in the context of single-strand elongation (Table 3.2).

This is very similar to that observed for Pol µ’s close homolog TdT, which also demonstrates similar indiscretion (101).

Comparison to Pol β.

Pol β, an X-family DNA polymerase, has been well characterized by pre-steady- state kinetics (71, 73, 111, 116). In comparison to Pol β, Pol µ has a 200-500 fold slower nucleotide incorporation rate (kp), a 20-60 fold higher nucleotide binding affinity, and an

63 8-25 fold lower substrate specificity (kp/Kd) for correct dNTP incorporation into

primer/template DNA substrates (Table 3). Pol µ has a 100-400 fold slower polymerization rate, a 7-25 fold higher affinity, and a 16-20 fold lower efficiency than

Pol β in the incorporations of mismatched incoming dNTPs (Table 3.3). Thus, Pol β is more efficient than Pol µ in the incorporations of both correct and incorrect nucleotides.

As a consequence of the aforementioned factors, the fidelity of Pol µ and Pol β with

DNA primer/template substrates is in the same range of 10-4 to 10-5 (Table 3.3).

Similarly, Pol β has higher substrate specificity in the incorporations of both correct and

incorrect dNTPs with single-nucleotide gapped DNA substrates than with normal DNA

substrates, but the fidelity in comparison is unchanged (Table 3.3). The fact that Pol µ

binds both correct and incorrect nucleotides with higher affinity than Pol β (Table 3.3),

may be reconciled by the additional putative interactions of Lys325 and His329 with the

γ-phosphate of an incoming nucleotide, which are not conserved in Pol β (115).

Pol µ Pol β Pol β (primer/template DNA) (primer/templateDNA)a (1 nt-gapped DNA)b Correct Incorrect Correct Incorrect Correct Incorrect

Kd (µM) 0.35-1.8 7.3-135 6.7-66 181-980 1.9-8.5 190-1600

-1 -5 kp (s ) 0.006-0.076 (2-30)x10 3-17 0.008-0.03 12-36 0.019-1.3

-1 -1 -7 -6 -5 kp/Kd (µM s ) 0.008-0.064 (4-61)x10 0.2-0.5 (8-100)x10 4-6 (7-250)x10

Fidelityc 10-4 to 10-5 10-4 to 10-5 10-4 to 10-5

a b c Ref. (111). Ref. (73). Calculated as (kp/Kd)incorrect/[(kp/Kd)correct + (kp/Kd)incorrect].

Table 3.3: Comparison of kinetic parameters of human Pol µ and rat Pol β for dNTP incorporation. 64 Biological Functions of Human DNA Polymerase µ in Ig Development.

V(D)J recombination and somatic hypermutation are two essential mechanisms

for generating antibody diversity during human Ig development. The latter mechanism

introduces an estimated 10-3 to 10-4 point mutations (per base pair per generation) into

the V domain of Ig genes and is essential for the affinity maturation of antibodies (117-

119). The somatic hypermutation rate is about 106–fold higher than the spontaneous

mutation rate of the rest of human genome. The dominant point mutations in the V

domain are transition mutations at the G:C base pairs embedded in the motif of Purine-

G-Pyrimidine-(A/T) and short palindromes or hairpin loops (117-119). The error-prone

mutase involved in somatic hypermutation has not been identified in spite of extensive

searches. Human DNA Pol µ has been proposed to be the mutase based on the following

characteristics: shares similar sequence (41% identity) and domain organization as TdT,

one of the key proteins in V(D)J recombination; catalyzes a random insertion of

nucleotides in the presence of Mn2+; and is preferentially expressed in peripheral

lymphoid tissues (8). However, this hypothesis is not supported by the inefficiency of

Pol µ as a polymerase (see above discussion) and its relative high dNTP incorporation fidelity (10-4 to 10-5) (Table 3.1). The sequence-independent fidelity of Pol µ is not

consistent with the distinctive pattern of somatic hypermutation, nor is the fact that

human Pol µ catalyzes frequent frameshift errors, rather than base substitution errors

(35). If Pol µ is not a mutase and does not act as TdT in the presence of Mg2+ (35, 40), it

has been proposed that this enzyme could simply bind and protect coding ends from

degradation by exonucleases during V(D)J recombination (32).

65 However, more studies are required to evaluate the mutase hypothesis since both

fidelity and activity of Pol µ can be affected by the identity of the metal ion cofactor and

the type of DNA substrate (ie: non-gapped versus gapped). Although the metal ion

cofactor for Pol µ is most likely to be Mg2+ in vivo as observed with other polymerases,

this is not unequivocally confirmed. If the preferred metal ion were Mn2+ or Co+, the fidelity of Pol µ would be lower since these transition metal ions are known to promote mutagenic consequences (120-122). For example, human Pol µ has been found to incorporate more non-canonical ribonucleotides with Mn2+ than with Mg2+ (37). Thus, it is important to identify the preferred metal ion cofactor for Pol µ. Additionally, if the mismatch extension fidelity of Pol µ is low, it could decrease the overall fidelity of this polymerase and thereby substantiate the mutase hypothesis.

The largest body of evidence currently points to a role of Pol µ in V(D)J recombination and/or NHEJ DNA end-processing. In vivo analysis of Pol µ-deficient mice suggests Pol µ is a key element contributing to the relative homogeneity in size of light chain CDR3 and taking part in Ig κ light chain rearrangement at a stage where TdT is no longer expressed (32). Pol µ has also been shown to associate with Ku and

XRCC4-ligase IV proteins, which play key roles in both V(D)J recombination and NHEJ

(34). These results clearly demonstrate that Pol µ is involved in Ig development. Recent reports have demonstrated that the template-independent activity of Pol µ provides microhomology for ligation of incompatible DNA ends by XRCC4-ligase IV (110). In addition, as discussed above, this enzyme possesses more efficient polymerization activity on single-nucleotide gapped substrates (36), and may experience potential

66 activity modification through interactions with other proteins. Overall, the wide variety of DNA substrates that Pol µ can accommodate (ie: primer/template, gapped, single- strand, lesions) as well as variance in incorporating dNTPs and rNTPs, provide unique potential that may prove useful in the active processing of DNA ends during V(D)J recombination and NHEJ.

In summary, to evaluate the potential role of Pol µ as a mutase involved in

generation of high-affinity Ig variants, Pol µ was overexpressed in Escherichia coli and

purified to homogeneity. The purified enzyme had a lifetime shorter than 20 minutes at

37 °C, but was stable for over three hours at 25 °C in an optimized reaction buffer. The fidelity of human Pol µ was thus determined using pre-steady-state kinetic analysis of the incorporation of single nucleotides into undamaged DNA 21/41-mer substrates at 25 °C.

Single turnover saturation kinetics for all sixteen possible deoxynucleotide (dNTP) incorporations and for four matched ribonucleotide (rNTP) incorporations were measured under conditions where Pol µ was in molar excess over primer/template DNA. The polymerization rate (kp), binding affinity (Kd), and substrate specificity (kp/Kd) are 0.006-

0.076 s-1, 0.35-1.8 µM, and (8-64)x10-3 µM-1s-1, respectively, for matched incoming dNTPs, (2-30)x10-5 s-1, 7.3-135 µM, and (4-61)x10-7 µM-1s-1, respectively, for

mismatched incoming dNTPs, and (2-73)x10-4 s-1, 45-302 µM, and (7-1300)x10-7 µM-1s-

1, respectively, for matched incoming rNTPs. The overall fidelity of Pol µ was estimated

to be in the range of 10-3-10-5 for both dNTP and rNTP incorporations and was sequence- independent. Furthermore, to evaluate the template-independent polymerization of this enzyme, the kinetics of dNTP and rNTP incorporation into a single-stranded DNA

67 substrate were measured and qualitatively compared to terminal deoxynucleotidyl transferase (TdT), a known template-independent enzyme from the same family. The sugar selectivity, defined as the substrate specificity ratio of a matched dNTP versus a matched rNTP, was measured to be in the range of 492-10959 for primer/template substrates, while markedly reduced to 2.8-4.9 for single-stranded DNA substrates. In addition to a slow and distributive DNA polymerase activity, Pol µ was identified to possess a weak strand-displacement activity. The in vitro characterization presented here has shed some light on the potential biological functions of this enzyme in adding to the body of evidence which negates Pol µ as a mutase, while further examining the template- independent activity consistent with the proposed role of this enzyme in DNA end- processing during V(D)J recombination and NHEJ. Clearly continued studies are further required to definitively assess the physiological function Pol µ.

CHAPTER 4

REOPENING STUDIES OF DNA POLYMERASE β AND R258A MUTANT (CONTRIBUTION OF R258 TO SUBDOMAIN MOTIONS)

4.1 Arg258 Background and Importance.

In the open binary complex of Pol β with gapped DNA, Arg258 is engaged in a salt-bridge with Asp192 of the catalytic aspartate triad. Upon subdomain closure represented in the ternary “closed” complex, the salt bridge between Asp192 and Arg258 is disrupted by the phenyl ring of Phe272 (Figure 4.1). This activates the enzyme by freeing Asp192 to coordinate the two metal ions required for catalysis (57). Recent molecular dynamics (MD) simulations and transition path sampling of Pol β conformational closing/opening have highlighted the possible importance of the Arg258 residue in fidelity discrimination. These simulations suggest that following a fast subdomain closing upon nucleotide binding, there exists a slow and possibly rate-limiting

Arg258 side chain rotation prior to chemistry, which serves to help properly poise the active site geometry for chemistry. It is also suggested that after chemistry, the rotation of Arg258 back to re-engage in the salt bridge with Asp192 is kinetically significant in subdomain reopening. It is further postulated that binding of an incorrect nucleotide will hamper such subtle conformational adjustments, leading to improper catalytic geometry

69 and thus acting as a means of fidelity discrimination (123-125). Furthermore, extrapolation of information from MD simulations of Pol β mutants predict a facilitated closing of the fingers subdomain before chemistry for R258A, whereas MD simulations of other mutants, such as R258K, predict a tendency toward a more “open” state, though less “open” than observed in the Pol β•DNA binary complex structure, which indicates a possible delay in conformational closing before chemistry (81).

Figure 4.1: Motion of Arg258 upon subdomain closure. Green residues denote the position assumed in the Pol β open state, whereas red residues denote the position assumed in the closed catalytic state. Figure modified from Ref. (81).

Based upon the obvious changes in Arg258 during subdomain closing and reopening, it was of interest to apply stopped-flow fluorescence to examine if there are

global or kinetic perturbations during such events when this bulky residue is removed and replaced with another residue. In addition, the notion that the rotation of this residue prior to chemistry is rate-limiting is inconsistent with the results of all of our previous 70 mechanism studies on Pol β. For these two motivations, R258A was purified and

characterized using novel approaches developed in our lab.

4.2 Dissection of the Chemistry and Conformational Subdomain-reopening Steps of

WT Pol β Using Differential pH and Viscosity.

Although the rate of the slow phase of fluorescence transition observed in

stopped-flow upon rapid mixing of E•DNA with dNTP matches that of chemistry, as determined by rapid chemical quench, it is not expected that this fluorescence transition is produced by chemistry itself. Rather, it likely results from a conformational step after

chemistry that is rate-limited by chemistry (see Section 1.5 for Pol β mechanism

description). Namely, we hypothesize that the slow fluorescence transition originates

from the subdomain-reopening step (Step 6, Figure 1.6) following chemistry. By

applying differential pH and viscosity, Marina Bakhtina from our lab was able to

generate reaction conditions in which chemistry was facilitated to the point at which it

was no longer rate-limiting in the single-turnover dNTP incorporation pathway. This

ultimately allowed the dissection of the chemistry and the post-chemistry conformational

steps in Pol β’s reaction pathway, thus providing the necessary evidence to authenticate

this hypothesis (84).

The effects of pH on Pol β dNTP incorporation were examined, and it was

observed that as pH increased, the rate of the fast fluorescence phase (kfast) remained moderately unchanged, while the rate of the slow fluorescence phase (kslow) increased

considerably (Figure 4.2). Furthermore, the rate of dNTP incorporation (kquench) obtained

from rapid chemical quench matched kslow throughout the pH range studied (Figure 4.2).

71 Therefore, under conditions of high pH (Figure 4.3, green trace), the energy barrier of the

chemical step is lowered relative to that at neutral pH (Figure 4.3, blue trace), becoming

significantly more comparable to the energy barriers of the conformational steps.

kslow 20 kquench

15 ) -1

(s 10 obs k

6.4 6.8 7.2 7.6 8.0 8.4 8.8 pH

Figure 4.2: pH dependence of kslow and kquench in stopped-flow and chemical quench by WT Pol β. Reactions monitored correct dCTP incorporation into 18/35AP(G) at 25 °C in standard Pol β assay buffer. Analogous results have also been obtained for dATP incorporation into 19/36AP(T). Figure from M.B. and (84).

Figure 4.3: Qualitative free energy profile of correct dNTP incorporation by Pol β. This figure depicts Pol β mechanism at neutral pH (blue), high pH (green), and high pH and high viscosity (red). E = DNA polymerase in open conformation; E' = closed 2+ conformation; Dn = DNA; N = MgdNTP; M = catalytic Mg ; P = MgPPi. Figure modified from M.B. and (84).

Viscogens, such as glycerol, can be used to perturb conformational steps in an

enzymatic pathway which involve large spatial motions (126-131). Our lab has

previously established that upon an increasing concentrations of glycerol present in the

reaction buffer during catalysis at neutral pH 7.0, the rate of the fast fluorescence phase

(kfast), which corresponds to Pol β’s conformational closing (Step 2, Figure 1.6), is notably slowed, while the rate of the slow fluorescence phase (kslow), pertaining to the

chemical step, remains virtually unaffected (Figure 4.4) (80).

73 1.2 Fast Phase Slow Phase 1.0

0.8

0.6

0.4

Normalized Rate Constant Rate Normalized 0.2

0.0 10 15 20 25 30 35 % Glycerol

Figure 4.4: Viscosity dependence of kfast and kslow in stopped-flow during correct dNTP incorporation by WT Pol β at pH 7.0. Reactions monitored correct dCTP (200 µM) incorporation into 18/35AP(G) at 25 °C in Pol β standard assay buffer with differential glycerol. Analogous results have also been obtained for dATP incorporation into 19/36AP(T). Figure generated with data from M.B. as reported in (80).

From analysis of all the aforementioned results, it was predicted that a selective

slowing of the conformational steps (via increased buffer viscosity) in combination with a

selective increase in chemistry (via increased pH), could yield conditions in which the

post-chemistry conformational change, rather than chemistry, is the rate-limiting step

under single-turnover conditions. Under such conditions (high pH and high viscosity),

kquench would be expected to remain faster than kslow.

As predicted, at high pH both fast and slow fluorescence transitions were sensitive to altered buffer viscosity. As the glycerol concentration was increased from 10 to 40% at pH 8.3, both kfast and kslow were slowed (Table 4.1 and Figure 4.5), while the 74 rate of dNTP incorporation (kquench) was constant throughout (Table 4.1). Therefore,

under conditions of high pH and high viscosity (Figure 4.3, red trace), the energy barrier

of the chemical step is lowered and the energy barrier of the post-chemistry

conformational step is increased with proportion significant enough to allow a tangible

separation of the two events.

-1 -1 -1 a Glycerol kfast (s ) kslow (s ) kquench (s ) 10% 73.6 ± 0.8 27.9 ± 0.2 26.9 ± 0.8 15% 51.3 ± 0.7 24.6 ± 0.3 ----- 20% 48.1 ± 0.2 20.8 ± 0.1 ----- 25% 45.5 ± 0.4 16.5 ± 0.1 ----- 30% 36.2 ± 0.2 15.3 ± 0.1 ----- 35% 29.9 ± 0.4 12.8 ± 0.2 24.0 ± 1.0 40% 28.4 ± 0.3 8.73 ± 0.10 -----

a Data obtained by M.B. and reported in Figure 3B of Ref. (84). b Each reaction was initiated by rapid mixing of solution A (containing 1 µM Pol β and 400 nM 19/36AP DNA substrate) with solution B (containing 800 µM dATP) at 37 °C.

Table 4.1: Effect of viscosity on kfast, kslow, and kquench in stopped-flow during correct dNTP incorporation by WT Pol β at pH 8.3.b

75 1.0 Fast Phase Slow Phase

0.8

0.6

0.4

0.2 Normalized Rate Constant Rate Normalized

0.0 10 15 20 25 30 35 40 % Glycerol

Figure 4.5: Viscosity dependence of kfast and kslow in stopped-flow during correct dNTP incorporation by WT Pol β at pH 8.3. Reactions monitored correct dATP (400 µM) incorporation into 19/36AP(T) at 37 °C in Pol β assay buffer with differential glycerol.

4.3 The Slow Phase of Fluorescence Transition does not Originate from Enzyme

Translocation.

To test the possibility that the slow fluorescence change originates from translocation of DNA rather than subdomain reopening, we performed a stopped-flow assay using a single-nucleotide gapped substrate (Table 2.2). We found that the slow fluorescence transition is also present under these conditions (Figure 4.6). Since the product of dNTP incorporation into single-nucleotide gapped DNA is nicked DNA, and in the crystal structure of Pol β with a nicked DNA substrate the enzyme has not translocated, it is unlikely that the slow phase of fluorescence change originates from the

76 translocation of DNA to the next templating position. Furthermore, given the in vivo

involvement of Pol β in BER, the results obtained on the more physiologically pertinent

single-nucleotide gapped substrate were analogous to those observed for primer/template,

further validating the relevance of previous studies from our lab using the latter substrate.

0.4 19/36AP/16 19dd/36AP/16 0.3

0.2

0.1

2-AP Fluorescence 2-AP 0.0

-0.1

0.00.10.20.30.40.5 Time (s)

Figure 4.6: Pol β stopped-flow fluorescence assay of correct dNTP incorporation into single-nucleotide gapped DNA substrate. The black trace shows dATP -1 incorporation into 19/36AP(T)/16; a double exponential fit yielded kfast = 86.9 ± 4.7 s -1 and kslow = 13.2 ± 0.2 s . The gray trace shows the corresponding experiment with -1 dideoxy-terminated primer; a single exponential fit yielded kfast = 49.1 ± 3.4 s . Reactions were initiated by rapid mixing of solution A (containing 2 µM Pol β and 800 nM DNA substrate) with solution B (containing 600 µM dATP). Reactions were performed at pH 7.7 and 37 °C in Pol β standard assay buffer.

4.4 R258A Characterization.

Computational studies suggested that reorientation of R258 side chain is rate- limiting, and that the R258A mutant shows facilitated subdomain closing, consistent with 77 a reported increased rate of nucleotide insertion (81, 124). Since our aforementioned

results allowed the dissection of the chemical and the subdomain-reopening steps of wild-

type Pol β, it was of further interest to employ similar techniques to examine the specific

role of Arg258 in the conformational closing and reopening steps of the enzyme. To

accomplish this, we performed kinetic analyses of the R258A mutant. As shown in

Figure 4.7, a biphasic stopped-flow trace for dATP incorporation into 19/36AP by

R258A resembles that of wild-type. It was also confirmed by rapid chemical quench that

-1 the rate of single-nucleotide incorporation catalyzed by R258A (kquench = 13.6 s , Figure

-1 4.7) is identical to that of WT (kquench = 13.7 s , data not shown) and matches the rate of

the slow fluorescence phase. If rotation of Arg258 is truly rate-limiting, a perturbation in

kquench would be expected for this mutant.

78 50

2 [DNA

100 n+1 ]

1 (nM)

2-AP Fluorescence 150

0.0 0.2 0.4 0.6 0.8 1.0 Time (s)

Figure 4.7: R258A mutant of Pol β stopped-flow and chemical quench overlay of dATP incorporation into 19/36AP (T). Reactions were initiated by rapid mixing of solution A (containing 4 µM R258A and 400 nM 19/36AP DNA substrate) with solution B (containing 800 µM dATP). Reactions were performed at pH 7.7 and 37 °C in Pol β assay buffer containing 10 % glycerol and 10 mM free MgCl2. A double exponential fit -1 -1 of the stopped-flow trace (magenta) yielded kfast = 167 ± 0.9 s and kslow = 12.3 ± 0.1 s . Chemical quench data (open circles) were fit to a single exponential and yielded kquench = 13.6 ± 0.5 s-1.

4.5 R258A Mutant of Pol β has a Facilitated Subdomain-reopening Step.

To compare the rates of conformational changes between R258A and WT Pol β,

both differential pH and viscosity were employed. Results show that at pH 7.5 kslow for

R258A remained virtually unaltered upon increasing viscosity, whereas WT

demonstrated a systematic decrease in its rate of slow fluorescence transition (Figure

4.8). This result indicates that for nearly the entire viscosity range examined at pH 7.5, the putative subdomain reopening step of R258A remains faster than chemistry, whereas

79 for WT it becomes slower than chemistry, suggesting that the R258A mutant of Pol β has

a facilitated subdomain-reopening step (as depicted in the magenta trace of Figure 4.9).

Due to much faster chemistry at pH 8.3, the differences in the relative rates of subdomain

reopening between the two enzymes diminished at this pH, and the fluorescence phases

became comparably sensitive to viscosity (Figure 4.8).

1.0

0.5

pH 7.5 R258A

Normalized Rate Constant Normalized pH 7.5 WT pH 8.3 R258A pH 8.3 WT 0.0 10 15 20 25 30 35 % Glycerol (w/v)

Figure 4.8: Viscosity dependence of kslow in stopped-flow during correct dNTP incorporation by WT Pol β and R258A at pH 7.5 versus pH 8.3. The figure demonstrates the relative effect of buffer viscosity in stopped-flow for R258A (magenta) vs. WT (cyan) catalyzed dATP incorporation into 19/36AP(T) at pH 7.5 (shaded bars) and pH 8.3 (solid bars). Bars show the normalized rates of the slow fluorescence transition (kslow) vs. % glycerol in the reaction. Reactions were initiated as described in Figure 4.7, with solution A containing either 4 µM R258A or 1 µM Pol β. Both solutions A and B contained equal indicated concentrations of glycerol in the Pol β assay buffer.

Figure 4.9: Qualitative free energy profile of correct dNTP incorporation by Pol β versus R258A. This figure depicts relative differences in the conformational motions between Pol β (cyan) and R258A (magenta). E = DNA polymerase in open 2+ conformation; E' = closed conformation; Dn = DNA; N = MgdNTP; M = catalytic Mg ; P = MgPPi.

4.6 Discussion and Conclusion.

Previous studies in our lab have demonstrated that Pol β’s dNTP incorporation

pathway can be altered such that either the chemical step is selectively facilitated, or the

conformational steps selectively slowed. Through combinatorial exploitation of both, we

have demonstrated that under conditions of high pH and high viscosity, the rate of single-

nucleotide incorporation can be decoupled from the rate of the slow fluorescence

transition, which most likely represents subdomain reopening. This methodology has

been extended to the characterization of an R258A mutant predicted by computational studies to have perturbed conformational motions.

On the basis of crystal structures of Pol β binary (open) and ternary (closed)

complexes, it is known that upon subdomain closure the Asp192···Arg258 salt bridge is

81 disrupted, and the Arg258 side-chain rotates to form a hydrogen bond with Tyr296. This

activates the enzyme by freeing Asp192 to coordinate the two metal ions required for catalysis (57). Recent computational studies have suggested that Arg258 side-chain reorientation is likely the rate-limiting microscopic event during the course of Pol β’s subdomain closing before chemistry and reopening after chemistry (123, 124).

Furthermore, on this basis it has been proposed that an R258A mutant may have a lower energy barrier for the subdomain-closing step, and thus a faster rate of single dNTP incorporation (81, 124).

Our kinetic analysis indicates that the R258A mutation does not perturb the rate of single-nucleotide incorporation as measured by chemical quench, which further reinforces our proposed kinetic scheme in which conformational closing is not the rate- limiting step in Pol β’s nucleotide incorporation pathway. On the other hand, our results clearly demonstrate that the R258A mutant does possess facilitated subdomain reopening compared to that of WT. Overall, these results significantly highlight the importance of the Arg258 residue in isolated subdomain conformational events during the course of Pol

β nucleotide incorporation, while also suggesting that rotation of this residue is not rate- limiting in the overall kinetic pathway. Most significantly, these results demonstrate the usefulness of our kinetic scheme and experimental approaches in examining alterations in the mechanism of viable site-specific mutants.

CHAPTER 5

MECHANISM OF MISMATCHED NUCLEOTIDE INCORPORATION BY DNA POLYMERASE β AND I260Q “MUTATOR” MUTANT

5.1 Mismatched dNTP Incorporation.

Polymerase fidelity, at its most basic level, is determined by the free energy

difference between the highest energy barriers along the correct and incorrect dNTP

incorporation pathways. Since fidelity requires a such a comparison, an ever-increasing

gap between the structural and mechanistic information available for matched dNTP

incorporation compared with that concerning mismatched dNTP incorporation, renders

the molecular mechanism of DNA polymerase fidelity unclear (133, 134). The thermodynamic instability inherent to polymerase mismatched ternary complexes makes them difficult to crystallize. For this reason, we do not yet have available a structure representing a truly functional pre-chemistry closed mismatched ternary complex

(E’•Dn•N, Figure 1.6). Although, complexes depicting mispairs within the active site (62,

135) and mismatch extension (61, 136) for various polymerases have been resolved.

Likewise, kinetic data for mismatched dNTP incorporations by many DNA polymerases

have been reported (73, 78, 92, 137-140), yet commonly used kinetic approaches prevent

characterization of individual steps in the mismatched dNTP incorporation pathway.

83 As described in Chapter 1 and Chapter 4, stopped-flow analyses of Pol β, using either 2-aminopurine (2-AP) or Pol β’s sole Trp325 residue, have been very important in advancing Pol β mechanism studies (59, 79, 80, 84). Both probes serve as independent reporters of global enzyme-DNA conformational changes which occur during the catalysis of dNTP incorporation into DNA. Intricate studies involving a variety of chemical probes have been conducted in order to determine the nature of the two fluorescence transitions observed in stopped-flow fluorescence assays examining correct dNTP incorporation by Pol β. The collective findings of these studies support a model in which the fast fluorescence transition is attributed to dNTP-induced subdomain conformational change (Step 2, Figure 1.6) to form the ternary closed complex prior to chemistry (Step 4, Figure 1.6), and the slow fluorescence transition is attributed to subdomain reopening (Step 6, Figure 1.6) after the rate-limiting chemical step.

This chapter describes the application of our previously established stopped-flow fluorescence methods, as well as steady-state fluorescence spectroscopy to explore two specific aspects of polymerase fidelity: (i) the mechanism by which WT Pol β

discriminates against incorrect dNTP substrate to maintain a moderately high fidelity, and

(ii) the deviation of this mechanism in generation of a lower fidelity variant by a single

I260Q mutation. Such topics are significant in developing a full understanding of the

unique means by which DNA polymerases are able to choose a correct dNTP substrate,

whose identity is ever-changing with each round of catalysis, from a pool four

nucleotides. Moreover, while it is interesting to comprehend how a polymerase

accomplishes this feat to achieve a certain fidelity in DNA replication, it is equally

interesting to investigate how a single residue mutation can alter its fidelity. The results

84 of our kinetic analyses suggest that Pol β incorporates both correct and mismatched dNTP via analogous kinetic mechanisms differing at the transition state of the chemical step. Furthermore, the divergence in fidelity between WT and I260Q is consequent of differences in ability to stabilize the mismatched ternary complex and chemical transition state.

5.2 Ile260 Background.

The Ile260 residue of Pol β is located in the hydrophobic hinge region between the palm and the C-terminal fingers subdomain (56, 141). Several Pol β variants with alterations in residues lining this hinge region have demonstrated compromised fidelity

(142-146). The mutator activity of I260Q was first identified by a genetic screen (147), and subsequent kinetic characterization showed that while this mutant possessed parameters of DNA binding similar to that of wild-type, it interestingly showed loose binding discrimination of mismatched dNTP substrates, and thus a low fidelity (148).

These characteristics make the I260Q mutant of specific interest in fidelity studies of Pol

β, in order to further understand how one mutation can so drastically alter the mismatch discrimination profile of this enzyme.

5.3 Mechanism of Mismatched dNTP Incorporation by WT Pol β.

Stopped-flow fluorescence assays and rapid quench experiments have been previously performed in our lab under a wide variety of conditions in order to elucidate the mechanism for correct dNTP incorporation by WT Pol β (79, 80, 84). Here we report systematic pre-steady-state kinetic analyses of mismatched dNTP incorporation by WT

85 Pol β. As shown in Figure 5.1, the stopped-flow trace of T:G mismatched incorporation

catalyzed by WT follows the same biphasic pattern of fluorescence change observed for

correct dNTP incorporation, albeit with lower amplitude. Follow-up of mismatched

dNTP incorporation in rapid chemical quench yields rates which correspond precisely to

the slow phase rates obtained by stopped-flow (Figure 5.1B and Table 5.1). This suggests that the slow phase is attributed to a conformational step limited by the rate of chemistry. Use of dideoxy-terminated primer, which eliminates the chemistry step, abolishes the slow phase completely (Figure 5.1B inset). This implies that the fast fluorescence transition results from a step before chemistry, while the slow phase originates from a step after chemistry. Using the same approaches previously employed for matches (80), we showed that the rate of the fast conformational change is slowed as a function of increasing viscosity for mismatches as well (Figure 5.2, red bars).

Sensitivity to viscosity has been previously interpreted to be indicatory of large conformational movements within the ternary complex during the step responsible for the fast fluorescence transition.

86 A T:A Match T:G Mismatch 3

1 2-AP Fluorescence 2-AP

0.0 0.2 0.4 0.6 0.8 Time (s) B 2.5 0 0.90 20 [DNA 2.0 0.85 40 0.80

60 n+1 1.5 0.75

80 ] (nM)

0.0 0.2 0.4 0.6 0.8 1.0 1.2 100 1.0 120 2-AP Fluorescence 2-AP Stopped-Flow 140 0.5 Rapid Chemical Quench 160 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Time (s)

Figure 5.1: Pol β mismatched dNTP incorporation in stopped-flow. Reactions were initiated by rapid mixing of solution A (containing 1 µM Pol β and 400 nM DNA substrate) with solution B (containing either 110 µM dATP or 4 mM dGTP). Reactions were performed at 37 °C in Pol β assay buffer at pH 8.5 containing 10 % glycerol and 5 mM free Mg2+. (A) Stopped-flow traces of WT T:A matched (black) and T:G mismatched incorporation (red) into 19/36AP(T)/15 DNA, yielding slow phase rates of 37.0 ± 0.2 s-1 and 4.36 ± 0.01 s-1, respectively. (B) Stopped-flow trace of T:G mismatch (red) incorporation with rapid quench overlay (open circles) yielding a rate of 4.75 ± 0.34 s-1. The inset shows the corresponding T:G mismatch reaction using dideoxy-terminated primer (gray trace). 87

T:A (matched) T:G (mismatched)

Wild-Type I260Q Wild-Type I260Q

-1 -1 -1 -1 k2 116 ± 5 s 108 ± 7 s 256 ± 8 s 485 ± 25 s

Kd 29.5 ± 3.6 µM 6.57 ± 2.12 µM 488 ± 103 µM 232 ± 37 µM

-1 -1 -1 -1 kpol 42.9 ± 0.6 s 43.6 ± 1.4 s 5.70 ± 0.11 s 13.6 ± 0.2 s

Kd,app 6.79 ± 0.48 µM 7.07 ± 1.20 µM 489 ± 26 µM 48.8 ± 2.5 µM

-1 -1 -1 -1 kquench 45.7 ± 2.8 s 47.6 ± 2.5 s 4.75 ± 0.34 s 14.8 ± 0.5 s

a kpol / Kd,app 6.32 6.17 0.0117 0.279 Fidelity b ------541 23.1

C:G (matched) C:A (mismatched)

Wild-Type I260Q Wild-Type I260Q

-1 -1 -1 -1 kpol 20.8 ± 0.5 s 18.3 ± 0.4 s 1.86 ± 0.05 s 9.40 ± 0.15 s

Kd,app 2.15 ± 0.27 µM 1.20 ± 0.14 µM 227 ± 22 µM 44.5 ± 3.4 µM

a kpol / Kd,app 9.67 15.3 0.00819 0.211 Fidelity b ------1180 73.5

a Catalytic efficiency as measured in units of µM-1s-1. b Fidelity defined as [(kpol/Kd,app)cor+(kpol/Kd,app)inc]/(kpol/Kd,app)inc - where the subscripts “cor” and “inc” indicate the correct (matched) and incorrect (mismatched) nucleotide incorporation, respectively.

Table 5.1: Kinetic rate and binding constants for WT versus I260Q nucleotide incorporation into 19/36AP(T)/15 and 19/36AP(C)/15 DNA substrates at pH 8.5. The k2, Kd, kpol and Kd,app values were obtained from hyperbolic fit of the dNTP concentration dependence of the observed rates of the fast and slow fluorescence phases (kfast and kslow) (as described in Chapter 2, Section 2.4.6). The rate of dNTP incorporation (kquench) was obtained from single exponential fit of rapid chemical quench data at a saturating dNTP concentration. The k2 value represents the rate constant of forward conformational closing, while Kd reflects the stability of the ternary complex before closing. Therefore, WT and I260Q show little difference during the initial dNTP binding before conformational change. The kpol value represents the maximum rate of dNTP incorporation, while the Kd,app value possesses a contribution from all steps up to the rate- limiting step and can be thought as the dissociation constant of the closed ternary complex. 88 400 WT I260Q 300 ) -1

(s 200 fast k

100

0 10% 15% 20% 25% 30% % Glycerol

Figure 5.2: Viscosity dependence of kfast in stopped-flow during mismatched dNTP incorporation by WT Pol β and I260Q. Reactions were initiated by rapid mixing of solution A (containing 1 µM Pol β and 400 nM 19/36AP(T)/15) with solution B (containing 4 mM dGTP). Reactions were performed at 37 °C in Pol β assay buffer at pH 8.5 containing 5 mM free Mg2+ and differential glycerol in both syringes.

5.4 dNTP Concentration Dependence of the Fast and Slow Fluorescence Transitions

during Mismatch Incorporation.

Previous stopped-flow fluorescence assays investigating matched dNTP

incorporation showed that both the fast and slow fluorescence transitions demonstrated a

hyperbolic dependence on dNTP concentration (79, 93). Here, we analyze the dNTP

dependence of both the fast and slow fluorescence phases during mismatched dNTP incorporation in stopped-flow. The observed rate constants for the fast and slow phases

(kfast and kslow), obtained from a double exponential fit of each stopped-flow trace (see

Chapter 2, Section 2.4.6), individually plotted as a function of dNTP concentration,

89 revealed that both phases demonstrate a hyperbolic dependence on dNTP concentration.

In the case of the fast phase, a plot of the observed rate constant as a function of dNTP

was fit to a hyperbola with a nonzero intercept to obtain values for k2 (see Scheme 2.1),

the rate constant of forward conformational closing, and Kd, the thermodynamic dNTP

dissociation constant (see Chapter 2, Section 2.4.6). If the fast phase were resultant from

the bimolecular binding of mismatched dNTP, one would expect a linear dependence on

dNTP concentration. However, the observed hyperbolic dependence of the fast phase on

mismatched dNTP largely indicates that this phase originates from a conformational

change induced by mismatched dNTP binding. Similarly, for slow phase analysis, the observed rate constant was plotted against dNTP concentration and fit to a hyperbola to obtain values for kpol, the maximum rate of nucleotide incorporation, and Kd,app, the

apparent dNTP dissociation constant (see Chapter 2, Section 2.4.6). These parameters

correspond to the pre-steady-state values of kpol and Kd,app traditionally obtained by rapid

chemical quench. The parameters for k2, Kd, kpol, and Kd,app obtained for T:A match and

T:G mismatch incorporation by WT Pol β are recorded in Table 5.1.

Overall, all the aforementioned results support that, analogous to our prior assignments for correct dNTP incorporation, the fast and slow fluorescence changes observed for mismatched incorporation (Figure 5.1) can be assigned to the dNTP-induced subdomain-closing conformational change and the chemical step (which likely limits the reopening step), respectively. An important observation is that the forward rate of conformational closing (k2, Scheme 2.1) for mismatched dNTP incorporation is

comparable with that for correct dNTP incorporation (though with significant increase in

Kd), while the maximum rate of nucleotide incorporation (kpol) is substantially slower

90 (also with significant increase in Kd,app) (Table 5.1). This suggests that overall

mismatched incorporation follows a similar pathway, though both the Kd and Kd,app values are higher and the rate of the chemical step is slower.

5.5 A Biphasic Trace is Observed for Multiple Mismatches.

To further substantiate the conclusions drawn from the investigation of T:G mismatch incorporation in stopped-flow, several additional mismatches were investigated

(T:C, C:A, A:A, A:G, G:A, G:G). A biphasic trace was observed for all mismatches examined (Figure 5.3), and only the fast phase was present when dideoxy-terminated primer was employed (Figure 5.3, insets). These results lend credence to the notion that while Pol β incorporates mismatches via the same overall mechanism, involving fast subdomain closure upon dNTP binding, each mismatch is unique in its approach to the transition state, as exemplified by the variance in slow phase rates among the mismatches examined.

91 -0.2 0.8 0.0 -0.4 0.5 -0.1 0.8 -0.3 -0.6 -0.2 -0.4 0.6 T:C 0.4 C:A A:G -0.3 -0.8 0.6 -0.5 -0.4 0.3 -1.0 0.4 0.00.20.40.60.81.0 0.0 0.2 0.4 0.6 0.8 1.0 0.4 0 5 10 15 20 0.2 0.2 0.2 0.1 2-AP Fluorescence 2-AP 0.0

0.0 0.5 1.0 1.5 2.0 0.0 0.2 0.4 0.6 0.8 1.0 0 5 10 15 20 Time (s) Time (s) Time (s)

-0.75 0.4 -0.75 1.2 0.5 -0.80 0.2 -0.85 -0.80 0.4 1.0 0.0 -0.90 G:A -0.85 G:G A:A 0.4 -0.2 -0.95 0.8 -0.90 -0.4 0.3 0 5 10 15 20 0.3 0 10203040500.6 012345

0.2 0.4

2-AP Fluorescence 2-AP 0.2 0.2 0.1

0 5 10 15 20 0 1020304050012345 Time (s) Time (s) Time (s)

Figure 5.3: Biphasic fluorescence trace observed for multiple mismatched dNTP incorporation by WT Pol β. Reactions were initiated by rapid mixing of solution A (containing 1 µM Pol β and 400 nM 19/36AP(T,C,A) or 18/35AP(G)), with solution B (containing 6 mM mismatch dNTP). Reactions were performed at 37 °C in Pol β assay 2+ buffer at pH 8.5 containing 5 mM free Mg and 35% glycerol. The outside traces demonstrate the biphasic trace observed for regular primer, while the insets depict the corresponding reaction with dideoxy-terminated primer. Fast and slow phase misincorporation rates are as follows: T:C mismatch (dark green), 122 ± 1 s-1 and 3.22 ± 0.02 s-1; C:A mismatch (dark yellow), 114 ± 1 s-1 and 14.0 ± 0.1 s-1; A:G mismatch (dark red), 113 ± 2 s-1 and 0.226 ± 0.001 s-1; G:A mismatch (dark cyan), 115 ± 2 s-1 and 0.189 ± 0.001 s-1; G:G mismatch (dark gray), 96.9 ± 2.5 s-1 and 0.0401 ± 0.001 s-1; A:A mismatch (dark blue), 92.3 ± 1.1 s-1 and 1.34 ± 0.01 s-1.

5.6 Kinetic Properties of I260Q and WT are Similar for Correct dNTP

Incorporation.

As an extension of the WT studies delineated above, further studies on the I260Q

mutant were conducted in the interest of kinetic comparison. Stopped-flow analysis of

correct dNTP incorporation (Figure 5.4 and Table 5.1) indicates that I260Q behaves

92 similarly to WT in that: (i) it possesses a biphasic trace with similar rate constants

compared to WT; (ii) use of dideoxy-terminated primer abolishes the slow phase; (iii) the

rapid chemical quench rate matches that of the slow phase; and (iv) the viscosity

dependence pattern of the fast and slow phases are identical to that of WT (Figure 5.5).

Furthermore, dNTP concentration dependence analysis revealed that for T:A correct

-1 -1 incorporation both enzymes have similar k2 values (116 ± 5 s and 108 ± 7 s for WT and

-1 -1 I260Q, respectively), kpol values (42.9 ± 0.6 s and 43.6 ± 1.4 s for WT and I260Q,

respectively) and Kd,app values (6.79 ± 0.48 µM and 7.07 ± 1.20 µM for WT and I260Q,

respectively), while demonstrating a 4.5-fold difference in Kd (29.5 ± 3.6 µM and 6.57 ±

2.12 µM for WT and I260Q, respectively) (Table 5.1). Likewise, analysis of C:G

matched incorporation yielded similar values of kpol and Kd,app between the two enzymes

(Table 5.1).

93 2.5 80 2.2 2.0 2.0 100 [DNA

1.5 1.8 120 n+1

1.0 ] (nM) 0.0 0.1 0.2 0.3 140

0.5

2-AP Fluorescence 2-AP 160

0.0 180 0.00 0.05 0.10 0.15 0.20 Time (s)

Figure 5.4: I260Q correct dNTP incorporation in stopped-flow. I260Q T:A matched incorporation into 19/36AP(T)/15 (teal trace). Reactions were initiated by rapid mixing of solution A (containing 1 µM Pol β and 400 nM DNA substrate) with solution B (600 µM dATP) at pH 8.5. The slow phase fits to a rate of 48.5 ± 0.4 s-1, while rapid quench overlay (open circles) fits to a rate of 47.6 ± 2.5 s-1. The inset shows the corresponding T:A match reaction using dideoxy-terminated primer (gray trace).

94 A 140 WT 120 I260Q

100 )

-1 80 (s 60 fast k 40

0 10% 15% 20% 25% 30% % Glycerol B 14

10 )

-1 8 (s 6 slow k 4

0 10% 15% 20% 25% 30% % Glycerol

Figure 5.5: Viscosity dependence of kfast and kslow in stopped-flow during correct dNTP incorporation by WT Pol β versus I260Q. Reactions were initiated by rapid mixing of solution A (containing 1 µM Pol β or I260Q and 400 nM 19/36AP(T)/15) with solution B (containing 600 µM dATP). Reactions were performed at 37 °C in Pol β assay buffer at pH 7.6 containing 5 mM free Mg2+ and differential glycerol in both syringes.

95 5.7 Kinetic Properties of I260Q and WT are Different for Mismatched dNTP

Incorporation.

There are clear kinetic differences between I260Q and WT during mismatched

incorporations. The similarities and differences are summarized here: (i) A comparison

in Figure 5.6 demonstrates that the fluorescence change of I260Q mismatch incorporation

follows the same pattern as that of WT, though the slow phase is faster by a factor of 2-3

(Table 5.1). The assignments of the two phases have again been verified by viscosity

dependence (Figure 5.2) and rapid quench (Table 5.1). (ii) Note that the rate of the

conformational change, k2, is not slowed down for I260Q relative to WT (Figure 5.7A and B and Table 5.1), suggesting that the dNTP-induced conformational closing is not disrupted during mismatched dNTP incorporation by I260Q. This is in contrast to what has been suggested previously (148). (iii) The main difference between I260Q and WT lies in the ability of I260Q to more efficiently stabilize the mismatched ternary complex, as suggested by the 10-fold decrease in Kd,app of I260Q (48.8 ± 2.5 µM, see Figure 5.7A

and C) relative to WT (489 ± 26 µM) for T:G mismatch incorporation and the 5-fold

decrease in Kd,app of I260Q (44.5 ± 3.4 µM) relative to WT (227 ± 22 µM) for C:A

mismatch incorporation (Table 5.1). This is in agreement with the tight mismatch dNTP

binding of this mutant formerly reported (148). The combination of the lower Kd,app and

faster kpol values observed for I260Q mismatched incorporation yields catalytic

efficiencies of 0.279 µM-1s-1 for T:G mismatch and 0.211 µM-1s-1 for C:A mismatch,

which are significantly larger than those recorded for WT (Table 5.1). Note that the

catalytic efficiencies for both enzymes during matched incorporation show little difference. Selective enhancement of mismatch incorporation by I260Q translates into a

96 fidelity 23-fold lower for T:G mismatch incorporation and 16-fold lower for C:A mismatch.

WT 3 I260Q

1 2-AP Fluorescence 2-AP

0 0.0 0.2 0.4 0.6 0.8 Time (s)

Figure 5.6: Comparison of WT Pol β and I260Q T:G mismatched incorporation at pH 8.5. Reactions were initiated by rapid mixing of solution A (containing 1 µM Pol β or I260Q and 400 nM 19/36AP(T)/15) with solution B (containing 4 mM dGTP) and 2+ performed at 37 °C in Pol β assay buffer at pH 8.5 containing 5 mM free Mg and 10% glycerol. Fast and slow phase misincorporation rates were 323 ± 6 s-1 and 4.49 ± 0.01 s-1 correspondingly for WT (red trace), and 383 ± 6 s-1 and 12.2 ± 0.03 s-1 correspondingly for I260Q (blue trace).

Figure 5.7: I260Q T:G mismatch titration at pH 8.5. Reactions were initiated by rapid mixing of solution A (containing 1 µM I260Q and 400 nM 19/36AP(T)/15) with solution B (containing varied dGTP concentration). Reactions were performed at 37 °C in Pol β assay buffer containing 5 mM free Mg2+. (A) Stopped-flow traces as a function of increasing dGTP concentration are pictured. (B) A hyperbolic fit of the fast phase rate -1 (kfast) versus dGTP concentration yields a k2 of 485 ± 25 s , a Kd, of 232 ± 37 µM, and a -1 y0 of 106 ± 8 s . (C) A hyperbolic fit of the slow phase rate (kslow) versus dGTP -1 concentration yields a Kd, app of 48.8 ± 2.5 µM and a kpol of 13.6 ± 0.2 s .

98 A 3.5 10 µM 3.0 25 µM 50 M 2.5 µ 100 µM 2.0 200 µM 350 µM 1.5 500 µM 1.0 2-AP Fluorescence 2-AP 0.5

0.0 0.0 0.2 0.4 0.6 0.8 Time (s) B 500

400

) 300 -1 (s

fast 200 k

100

0 0 100 200 300 400 500 [dGTP] (µM) C 14

10 )

-1 8 (s 6 slow k 4

0 0 100 200 300 400 500 [dGTP] (µM) Figure 5.7

5.8 The Fast Fluorescence Transition is also Present at Physiological pH.

To further validate the aforementioned results, it must be reported that at

physiological pH, a biphasic trace of fluorescence change is still observed in stopped-

flow for both WT and I260Q during T:G misincorporation (Figure 5.8). Similar to that

observed for pH 8.5, the observed rate of the dNTP induced conformational change is not

slowed down for I260Q relative to WT at pH 7.6, again pointing out that the dNTP-

induced conformational closing of I260Q is not disrupted during matched or mismatched

dNTP incorporation. A rigorous dNTP titration analysis of the fast phase was not possible for mismatches at this pH due to a combination of the low amplitude of this

phase at dNTP concentrations below 2 mM and a noted pattern of inhibition at high

dNTP concentration. Slow phase analyses for WT versus I260Q nucleotide incorporation

at pH 7.6 are reported in Table 5.2. Again, for matched incorporation, little difference is

observed for kpol and Kd,app between WT and I260Q at this pH, leading to similar catalytic

efficiencies (Table 5.2). However, a 7.4-fold decrease in the Kd,app of I260Q (88.1 ± 6.4

µM) relative to WT (654 ± 42 µM) for T:G mismatch incorporation, and a 4.3-fold

decrease in the Kd,app of I260Q (69.4 ± 2.7 µM) relative to WT (298 ± 23 µM) for C:A

mismatch incorporation at pH 7.6 further supports that the main difference between

I260Q and WT lies in the enhanced ability of I260Q to stabilize the mismatched ternary

complex. Again, the combination of the lower Kd,app and faster kpol values observed for

I260Q mismatched incorporation at this pH translates into a higher catalytic efficiency

for mismatch incorporation by I260Q compared to WT, and thus a lower fidelity (Table

5.2).

100 2.8 2.25 2.5 2.7 2.20 2.15 2.6 2.10 2.0 0.00 0.05 0.10 0.00.20.40.6 WT 1.5 I260Q

1.0 2-AP Fluorescence 2-AP

0.5

012345 Time (s)

Figure 5.8: Comparison of WT Pol β and I260Q T:G mismatched incorporation at pH 7.6. Reactions were initiated by rapid mixing of solution A (containing 1 µM Pol β or I260Q and 400 nM 19/36AP(T)/15) with solution B (containing 4 mM dGTP) and 2+ performed at 37 °C in Pol β assay buffer at pH 7.6 containing 5 mM free Mg and 10% glycerol. Fast and slow phase misincorporation rates were 69.9 ± 4.5 s-1 and 0.245 ± 0.0004 s-1 respectively for WT, and 314 ± 22 s-1 and 1.68 ± 0.003 s-1 respectively for I260Q. The insets show a close-up of the fast phases for WT (red trace) and I260Q (blue trace).

101

T:A (matched) T:G (mismatched)

Wild-Type I260Q Wild-Type I260Q

-1 -1 -1 -1 kpol 13.4 ± 0.5 s 13.4 ± 0.4 s 0.261 ± 0.007 s 1.82 ± 0.03 s

Kd,app 2.22 ± 0.33 µM 1.10 ± 0.15 µM 654 ± 42 µM 88.1 ± 6.4 µM

a kpol / Kd,app 6.04 12.2 0.000399 0.0207

Fidelity b ------15100 590

C:G (matched) C:A (mismatched)

Wild-Type I260Q Wild-Type I260Q

-1 -1 -1 -1 kpol 7.21 ± 0.12 s 6.61 ± 0.15 s 0.191 ± 0.005 s 1.13 ± 0.01 s

Kd,app 1.03 ± 0.1 µM 0.456 ± 0.074 µM 298 ± 23 µM 69.4 ± 2.7 µM

a kpol / Kd,app 7.00 14.5 0.000641 0.0163

Fidelity b ------10900 891

Table 5.2: Kinetic rate and binding constants for WT versus I260Q nucleotide incorporation into 19/36AP(T)/15 and 19/36AP(C)/15 DNA substrates at pH 7.6. The kpol and Kd,app values were obtained from hyperbolic fit of the dNTP concentration dependence of the observed rate of slow fluorescence phase (kfast and kslow) (as described in Chapter 2, Section 2.4.6). The kpol value represents the maximum rate of dNTP incorporation, while the Kd,app value possesses a contribution from all steps up to the rate- limiting step and can be thought as the dissociation constant of the closed ternary complex.

102 5.9 Steady-state Fluorescence Studies.

Steady-state fluorescence emission spectra of WT versus I260Q demonstrate identical binary complex traces. For both enzymes there is an equivalent increase in 2-

AP fluorescence intensity at the emission maximum (360 nm) upon addition of saturated concentrations of correct dNTP (Figure 5.9). Addition of mismatched dNTP also results

in a fluorescence change in the same direction, though with lower amplitude.

Importantly, I260Q demonstrates higher fluorescence intensity (Figure 5.9) and lower

saturating mismatched dNTP concentration compared to WT (data not shown). The data correlate well to the observed amplitude differences in the stopped-flow fluorescence assays (Figure 5.1 and Figure 5.6). Overall, these results are consistent with the proposal that both matches and mismatches are incorporated via a similar mechanism, as both correct and mismatched dNTPs elicit the same direction of fluorescence change. In combination with conclusions from recent SAXS structural studies (149), they also qualitatively support the hypothesis that the enhanced ability of I260Q to incorporate mismatched dNTP may arise from a more “partially closed” mismatched ternary complex structure than WT (see Discussion).

103 WT Binary WT Matched Ternary 6 WT Mismatched Ternary I260Q Mismatched Ternary

2 2-AP Fluorescence 2-AP

0 320 340 360 380 400 420 440 460 480 500 Emission Wavelength (nm)

Figure 5.9: Steady-state fluorescence spectra of matched and mismatched ternary complexes for WT Pol β and I260Q. Reactions contained 200 nM 19dd/36AP(T)/15 DNA, 500 nM enzyme in the presence of either no dNTP, 300 µM matched dNTP, or 5 mM mismatched dNTP in Pol β assay buffer at pH 8.5. Emission spectra shown include WT binary complex (cyan), WT T:A matched ternary complex (black), WT T:G mismatched ternary complex (red), I260Q T:G mismatched ternary complex (blue). Omitted for simplification are the I260Q binary and T:A matched ternary complexes, which overlay accurately with the corresponding WT traces.

5.10 Discussion and Conclusion.

As an extension of previous methodology used to delineate Pol β’s correct dNTP incorporation mechanism (79, 80), similar techniques were employed to examine the mechanism of mismatched dNTP incorporation by Pol β and its I260Q “mutator” mutant.

Detailed studies of these two enzymes led to the following important conclusions: (i)

Stopped-flow fluorescence analyses indicate that there is, in fact, a conformational closing event that occurs for mismatched dNTP incorporation, as evidenced by the

104 existence of a fast fluorescence phase preceding chemistry. Furthermore, the rate of

conformational change induced by mismatched dNTP is comparable to that of subdomain

closing induced by correct dNTP. (ii) Comparison studies of WT and I260Q show that

I260Q possesses fast conformational closing steps during both matched and mismatched dNTP incorporation as well. This argues against a hindered dNTP closing for this enzyme previously hypothesized to account for its altered fidelity (148). (iii) The main

difference between the two enzymes lies in more effective stabilization of the

mismatched ternary complex by I260Q. (iv) Steady-state fluorescence studies

demonstrate that while both matched and mismatched dNTP elicit the same direction of fluorescence change, the I260Q mismatched ternary complex fluorescence emission

spectrum bears more resemblance to that of the matched ternary complex. Recent SAXS

and molecular modeling studies from our lab suggest that for both enzymes the mismatched ternary complex lies in-between the open and the closed forms (149). The

combination of these conclusions with the steady-state fluorescence studies presented

here provides grounds for speculation that on the global conformational closing pathway

ranging from fully open (binary complex) to fully closed (matched ternary complex),

mismatches may induce a less “partially closed” (more open) conformation for WT than

for I260Q.

Although our observation of a incorrect dNTP-induced conformational change by

Pol β may appear to contradict the findings of two FRET-base studies monitoring the

conformational motions of Klentaq (150) and Klenow (151), the differences could be

reconciled. Both studies report an increase in FRET signal, upon addition of correct

dNTP, yet do not observe any significant change upon addition of incorrect dNTP. In

105 contrast to moderate and low fidelity polymerases, including Pol β, since Klentaq and

Klenow are higher fidelity enzymes, it is likely that they more effectively destabilize the mismatched ternary complex to the extent that no fluorescence change is observable for mismatch binding. Our observation of differences in the amplitude of stopped-flow fluorescence traces (Figure 5.6) and steady-state fluorescence spectra (Figure 5.9) between WT Pol β and the lower fidelity I260Q variant also illustrate an extension of this correlation between fidelity and mismatch destabilization.

Our observation that the conformational change occurs with a similar rate for both correct and mismatched incorporation by Pol β differs from the conclusion of the recent

single molecule kinetic analysis of T7 DNA polymerase (152), which reported a

significantly reduced rate of conformational closing induced by mismatched dNTP binding. It remains to be established whether such discrepancy reflects differences in the mismatched discrimination mechanism employed by two polymerases, or whether it

results from different experimental systems and conditions. However, another study on

T7 reported little difference in the forward rates of conformational closing between matches and mismatches, while noting a large difference in the reverse rates of conformational closing (153).

On the other hand, our results are in full agreement with recent reports on the linear free energy relationship (LFER) of pyrophosphate leaving group elimination for both correct and mismatched dGTP incorporations utilizing dNTP analogs in which the

β,γ-bridging oxygen is substituted with various halomethylene moieties (154, 155).

Based on observation that the Brønsted correlation between log kpol and the leaving group

pKa is very similar between correct and mismatched incorporations, the authors conclude 106 that their results support our earlier prediction that the rate-limiting step is the chemical step, not the conformational change (133). Our observation of a fast conformational change for mismatched dNTP incorporations supports the interpretation of this work, and additionally argues against the notion that conformational closing is a major contributor to fidelity.

Figure 5.10 illustrates the current knowledge of the kinetic scheme of Pol β and its free energy diagram for matched dNTP incorporation (black profile) (84). What remain to be established are the free energy diagram (i.e., kinetic mechanism) and the intermediate structures (i.e., structural mechanism) for mismatched incorporations.

Multiple and recent proposals contemplate a perturbed conformation of the ternary complex and/or a different reaction pathway may be engaged by a mismatched dNTP incorporation in DNA polymerases (62, 153, 156-159). Our steady-state and stopped- flow fluorescence studies strongly support that both correct and mismatches are incorporated via a similar mechanism, and that the fidelity of Pol β is controlled, at least partly, by destabilization of the mismatched ternary complex (higher Kd,app values) and the chemical transition state (smaller kpol and kquench values) in the same reaction pathway

(Figure 5.10, red profile).

The loss of fidelity of I260Q has previously been explained by a hindrance of dNTP-induced subdomain closure (147, 148). It was suggested that upon the substitution of Ile with Gln in I260Q, rotamers are generated which cause a reduction in an important cavity space near residue 260, thus perturbing subdomain motion. However, our results indicate the contrary – that the kinetic properties of I260Q, including the rate of the dNTP-induced subdomain closure, are largely unperturbed from WT Pol β for correct 107 dNTP incorporations. Thus, for correct dNTP incorporation its free energy profile is

assumed to be the same as that of WT for the purpose of discussion here. The main

difference leading to the low fidelity of I260Q appears to be a loss in ability to destabilize

the mismatched ternary complex (Figure 5.10, blue profile). This conclusion is based on

the observation that the low fidelity of I260Q mutant is primarily determined by

enhanced efficiency of mismatched incorporation, whereas for most low fidelity DNA

polymerase variants, reduced fidelity correlates with low efficiency in correct dNTP

incorporation (160). Based upon the two mismatches examined, I260Q incorporates mismatches 24- to 54-fold more efficiently than WT (see Tables 5.1 and 5.2). The

observed change in kpol/Kd,app for mismatched incorporations by I260Q suggests that the

“over stabilization” of mismatched ternary complex by I260Q is further enhanced in the transition state. Molecular modeling of the I260Q mismatched ternary complex predicts the presence of additional hydrogen bonds, involving Gln260, Tyr296, Glu295, and

Arg258, which may a plausible source of this stabilization (149).

Overall, kinetic behavioral comparisons between WT Pol β and I260Q allow us to

conclude that infidelity of I260Q originates from enhanced stabilization of the

mismatched ternary complex and the chemical transition state. Regarding the mechanism

of mismatched discrimination, our lab previously suggested that DNA polymerase

fidelity originates primarily from differential substrate binding at the highest energy

transition state and suggested a most likely model (Figure 1D in (133)), which is highly

consistent with the results and conclusions of this work. Of future interest to enhance our

knowledge of the mechanism of polymerase fidelity, would be crystallization of the

I260Q mismatched ternary complex. This would provide insight into the extent of this

108 variant’s conformational closing profile and delineate a structural basis for understanding

which specific residues might contribute to this variant’s unique capability to support mismatched dNTP incorporation more efficiently.

Figure 5.10: Qualitative free energy profile of matched and mismatched dNTP incorporation by Pol β versus I260Q. E = DNA polymerase in open conformation; E' 2+ = closed conformation; Dn = DNA; N = Mg•dNTP; M = catalytic Mg ; P = Mg•PPi.

The insights into the mechanism of misincorporation presented here add to the

growing body of knowledge ultimately desired to consist of the microscopic rates of all

steps in Pol β’s matched and mismatched incorporation mechanism, in order to fully and

confidently assess which step is the most important contributor to this enzyme’s fidelity.

Recent illumination of the importance of the reverse rate of the dNTP-induced

conformational closing step and its effect on enzyme specificity further stresses the

necessity for complete evaluation of all microscopic rate constants (153). Ongoing

109 studies in our lab are examining this step for Pol β during both matched and mismatched incorporation (see Chapter 6). On a broader scale, through the use of Pol β we continue to develop a more precise understanding of the basis of DNA polymerase fidelity by establishing our techniques as a systematic framework of investigation that can be applied to multiple polymerases and additional site-specific mutants during both matched and mismatched nucleotide incorporations.

110

CHAPTER 6

DNA POLYMERASE β REVERSE CLOSING

6.1 Significance of the Reverse of the Conformational Closing Step.

Recently, the significance of the reverse of the conformational closing step has

been examined for T7 DNA polymerase (153). In this particular work, an analysis of the

microscopic constants used to define a polymerase’s specificity constant (kcat/Km) using the kinetic scheme depicted in Scheme 6.1 points out that during rigorous evaluation of the mechanism of fidelity discrimination for any given polymerase it is critical to evaluate the relative magnitudes of rate constants for chemistry (k3) and the reverse of the conformational closing step (k-2), here termed reverse closing for simplification for the remainder of the chapter.

k1 k2 k3 E•DNAn + dNTP E•DNAn•dNTP E’•DNAn•dNTP E•DNAn+1 + PPi k-1 k-2

Scheme 6.1. Minimal kinetic scheme used to derive specificity constant kcat/Km.

111 Using the minimal model picture in Scheme 6.1, the general specificity constant for

dNTP incorporation can be derived in which kcat/Km = k1k2k3/[k2k3+k-1(k-2+k3)]. In the

case that chemistry is much faster than reverse closing (k3 >> k-2), the specificity constant

can be reduced to k1k2/(k2+k-1), in which case it is evident that the chemistry step (k3,

Scheme 6.1) has no bearing on specificity. However, in the case that reverse closing is much faster than chemistry (k-2>> k3), the specificity constant is reduced to k1k2k3/(k2k3+k-

1k-2) in which chemistry is still significant. For T7 the relative magnitudes of chemistry

and reverse closing during matched and mismatched dNTP incorporation were examined

using a modified enzyme containing a covalently attached fluorophore on the nucleotide

binding subdomain (153). Here, we analyze the reverse closing of wild-type Pol β using

our established stopped-flow fluorescence system, which monitors the emission of 2-AP

modified DNA.

6.2 Monitoring Reverse Closing in Stopped-flow.

Since the closing of the Pol β fingers subdomain upon M•dNTP binding (not dNTP alone) is a dynamic process involving a conformational closing/reverse closing equilibrium (Step 2, Figure 6.1), it follows that starting from a preformed ternary

E’•Dn•N complex the reverse closing step could theoretically be isolated by stripping the

incoming dNTP of its metal ion. Removal of the divalent metal from the M•dNTP

complex would be expected to ensure that substrate rebinding cannot occur, and thus

would force the equilibrium strictly towards the enzyme open conformation. As a side

note, we speculate that stripping the magnesium from the catalytic binding site is not

sufficient to isolate the reverse closing step, as a pre-chemistry intermediate crystal

112 structure of Pol β (59) indicates that the enzyme resides in the closed conformation in the

presence of M•dNTP and in the absence of the catalytic magnesium.

Figure 6.1: Kinetic scheme of single-nucleotide incorporation with emphasis on reverse closing. E = Pol β in open finger conformation; E’ = Pol β in closed finger conformation; Dn = DNA; Dn+1 = DNA elongated by addition of one nucleotide; N = M•dNTP; M = catalytic metal ion; P = M•PPi . Boxed portion represents the reaction steps involved in putative reverse closing. Figure modified from Ref. (80).

To initiate the putative reverse closing reaction, a preincubated solution

containing Pol β, dideoxy-terminated DNA, and dNTP in the presence of Mg2+ was

mixed with EDTA, a known Mg2+ chelator (Figure 6.2, dark cyan trace). Upon mixing,

fluorescence decay with a single exponential rate of 82.4 ± 0.4 s-1 was observed.

Providing support that the observed fluorescence decay actually represents reverse

closing, the direction of fluorescence change was opposite to that elicited by the forward

reaction using dideoxy-terminated primer (Figure 6.2, black trace). Note that control 113 reactions in which either EDTA, dNTP, or Mg2+ were omitted from the reaction did not

exhibit any significant fluorescence change (see Section 6.3).

Forward Closing Reverse Closing 2.0

A: E•19dd/36AP(T) + 5.3 mM Mg2+ 1.5 B: 600 µM dATP(T) + 5.3 mM Mg2+

1.0

0.5 2+ 2-AP Fluorescence 2-AP A: E•19dd/36AP(T) + 5.6 mM Mg + 600 µM dATP B: 20 mM K+EDTA

0.0

0.00 0.05 0.10 0.15 0.20 Time (s)

Figure 6.2: Comparison of forward and reverse closing as monitored in stopped- flow. The reverse reaction (dark cyan trace) was initiated by rapid mixing of solution A (containing 1 µM Pol β, 400 nM 19dd/36AP(T), 600 µM dATP, and 5.6 mM Mg2+) with solution B (containing 20 mM K+EDTA). Single exponential fit yielded a rate of 82.4 ± 0.4 s-1. The corresponding forward closing reaction (black trace) was initiated by rapid mixing of solution A (containing 1 µM Pol β, 400 nM 19dd/36AP(T), and 5.3 mM Mg2+) with solution B (containing 600 µM dATP, and 5.3 mM Mg2). Double exponential fit yielded rates of 137.9 ± 4.6 s-1 and 67.9 ± 1.0 s-1. Reactions were performed at 37 °C and pH 7.6.

6.3 Dependence of Reverse Closing on Altered Reaction Parameters.

Based upon the kinetic scheme depicted in Figure 6.1, it is expected that the rate

of reverse closing reaction would not depend on EDTA, dNTP, DNA, or enzyme

concentration. In order to provide further evidence that the fluorescence decay observed

114 in Figure 6.2 is representative of the reverse closing step, the dependence of the reverse

closing reaction was thus investigated under multiple reaction conditions including varied

EDTA concentration, free Mg2+ concentration, matched dNTP concentration, enzyme

concentration, and DNA concentration. First, the effect of different EDTA

concentrations on reverse closing was examined (Figure 6.3A). A control reaction with 0

mM EDTA did not demonstrate any fluorescence change (Figure 6.3A, black trace). The

observed rate of reverse closing (krc) did not show significant change over a range of 20

to 100 mM EDTA (Table 6.1).

Next, the reverse closing reaction was also examined as a function of increasing

free Mg2+ concentration in the presence of constant matched Mg•dATP concentration (1

mM). The rate of reverse closing showed a slight decrease as the free Mg2+ concentration

was increased over a range of 0.03 mM to 12 mM (Table 6.1). It is possible that as the

2+ Mg2+ catalytic metal binding site (see Figure 1.4) becomes saturated with Mg (Kd, app =

1.0 mM (79)), the ternary complex is increasingly stabilized leading to a slower observed rate for the rate of the reverse conformational step, krc. Note that the amplitude of the

fluorescence change is not dependent on Mg2+ concentration, indicating that the fraction

of ternary complex in the closed conformation (E’•Dn•N, Figure 6.1) remains the same

regardless of occupation of the catalytic metal site. A control reaction in which MgCl2 was omitted from the reaction, did not demonstrate any fluorescence change (Figure

6.3B, black trace).

The reverse closing reaction in stopped-flow was also examined as a function of increasing matched dNTP concentration. A control reaction with 0 mM dATP did not demonstrate any fluorescence change (Figure 6.3C, black trace). While the rate of

115 reverse closing did not show significant change over a range of 5 to 500 µM dATP (Table

6.1), the amplitude did show a steady increase (Figure 6.3C). The increase in amplitude

is likely observed due to the fact that the initial amount of ternary complex formed prior to mixing increases as the concentration of dATP is increased. In addition, krc was not

altered by changes in DNA or enzyme concentration (data not shown).

116

Figure 6.3: Effect of altered reaction conditions on reverse closing. (A) EDTA concentration dependence. The reverse closing reaction was initiated by rapid mixing of solution A (containing 1 µM Pol β, 400 nM 19dd/36AP(T), 600 µM dATP, and 10 mM Mg2+) with solution B (containing varied K+EDTA from 0 – 100 mM). (B) Free Mg2+ concentration dependence. The reverse closing reaction was initiated by rapid mixing of solution A (containing 1 µM Pol β, 400 nM 19dd/36AP(T), 1 mM dATP, and free Mg2+ varied from 0 – 12 mM) with solution B (containing 40 mM K+EDTA). (C) Matched dATP concentration dependence. The reverse closing reaction was initiated by rapid mixing of solution A (containing 1 µM Pol β, 400 nM 19dd/36AP(T), 5 mM free Mg2+, and dATP varied from 0 – 500 µM) with solution B (containing 40 mM K+EDTA). All reactions were performed at 37 °C in Pol β assay buffer at pH 7.6. Corresponding rates are recorded in Table 6.1.

117 A

0.0 0 mM -0.5 10 mM 40 mM 70 mM -1.0 100 mM

-1.5 2-AP Fluorescence 2-AP

-2.0

0.00 0.02 0.04 0.06 0.08 0.10 0.12 Time (s) B 0 mM 3 0.03 mM 0.1 mM 2 0.5 mM 1 mM 3 mM 1 7 mM 12 mM 2-AP Fluorescence 2-AP 0

0.00 0.02 0.04 0.06 0.08 0.10 0.12 Time (s) C 2.5 0 µM 5 µM 2.0 10 µM 20 µM 1.5 50 µM 1.0 100 µM 300 µM 0.5 500 µM 2-AP Fluorescence 2-AP

0.0

0.00 0.02 0.04 0.06 0.08 0.10 0.12 Time (s)

Figure 6.3 118

EDTA Free Mg2+ dATP

2+ [EDTA] krc [free Mg ] krc [dATP] krc 0 mM ----- 0 mM ----- 0 µM ----- 10 mM 76.7 ± 0.2 s-1 0.03 mM 99.1 ± 0.4 s-1 5 µM 103 ± 2 s-1 40 mM 72.2 ± 0.2 s-1 0.1 mM 99.2 ± 0.5 s-1 10 µM 89.7 ± 1.0 s-1 70 mM 69.1 ± 0.2 s-1 0.5 mM 93.0 ± 0.4 s-1 20 µM 91.0 ± 0.7 s-1 100 mM 63.6 ± 0.2 s-1 1 mM 92.6 ± 0.4 s-1 50 µM 90.3 ± 0.6 s-1 3 mM 88.2 ± 0.4 s-1 100 µM 89.6 ± 0.6 s-1 7 mM 78.2 ± 0.4 s-1 300 µM 89.0 ± 0.5 s-1 12 mM 67.9 ± 0.3 s-1 500 µM 89.5 ± 0.5 s-1

Table 6.1: Dependence of the rate of reverse closing on EDTA, free Mg2+, and matched dATP. Reaction conditions under which reported rates were obtained are described in the legend of Figure 6.3.

6.4 Dependence of Reverse Closing on Differential Viscosity.

Since reverse closing should involve a large conformational motion, it is expected that this motion would be altered with viscosity, similar to that observed for the conformational closing during the forward mechanism of dNTP incorporation (as described in Chapter 4, Section 4.2). Evaluation of the effect of viscosity on the reverse closing reaction was conducted by applying variable concentrations of glycerol in the reaction buffer (Figure 6.4). The observed rate of reverse closing reaction (krc) demonstrated a marked dependence on glycerol concentration (Table 6.2). As the concentration of glycerol increased, krc decreased. This is similar to the same trend

observed for forward conformational closing (ie: kfast in Figures 4.4 and 4.5). 119 3.5 10% Glycerol 15% Glycerol 3.0 20% Glycerol 30% Glycerol 2.5

2.0

1.5

1.0 2-AP Fluorescence

0.5

0.0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Time (s)

Figure 6.4: Viscosity dependence of reverse closing. The reverse closing reaction was initiated by rapid mixing of solution A (containing 1 µM Pol β, 400 nM 19dd/36AP(T), 400 µM dATP, and 10 mM Mg2+) with solution B (containing 20 mM K+EDTA). Reactions were performed at 37 °C in Pol β assay buffer at pH 7.6 containing and differential glycerol in both syringes. Corresponding rates are reported in Table 6.2.

[Glycerol] 10% 15% 20% 30% 40%

-1 -1 -1 -1 -1 krc 75.2 ± 0.2 s 55.3 ± 0.1 s 38.5 ± 0.1 s 19.0 ± 0.1 s 8.54 ± 0.03 s

Table 6.2: Viscosity dependence of the rate of reverse closing. Conditions under which rates were obtained as described in Figure 6.4.

120 6.5 Reverse Closing in the Presence of a Mismatched dNTP.

Reverse closing starting from a T:G mismatched ternary complex was examined

analogous to that described previously starting from a T:A matched ternary complex

(Section 6.2-6.4). For the mismatched reverse closing reaction, a fluorescence decay

with a single exponential rate of 887 ± 15 s-1 was observed, while the rate for matched reverse closing under identical conditions was 78.2 ± 0.3 s-1 (Figure 6.5A). The

significantly faster krc observed for mismatched reverse closing likely results from the

instability of the mismatched ternary complex, thus perturbing the equilibrium of the

conformational closing upon dNTP binding. Further examination of the same

mismatched reverse closing reaction under conditions of higher reaction buffer viscosity

-1 (35% glycerol versus 10% glycerol), yielded a krc of 418 ± 3 s (Figure 6.5B). This

noted decrease in rate is similar to that observed for matched reverse closing under conditions of differential viscosity (Figure 6.4 and Table 6.2), indicating the fluorescence decay observed for both matched and mismatched reverse closing likely represents the same conformational process. Additionally, analysis of the mismatched dNTP concentration dependence of krc (Figure 6.5C) revealed similar results to those observed

for matched dNTP concentration dependence (Figure 6.3C), in which the traces

demonstrate increased amplitude of fluorescence change upon increasing mismatch

dNTP concentration.

We also examined reverse closing starting from several other matched and

mismatched ternary complexes, including T:A versus T:C, C:G versus C:A, and A:T

versus A:G (Figure 6.6). In all cases the rate of fluorescence change was significantly

faster for mismatched reverse closing reactions.

121

Figure 6.5: Mismatched reverse closing. (A) Comparison of matched and mismatched reverse closing. Rapid mixing of solution A (containing 1 µM Pol β, 400 nM 19dd/36AP(T), 5 mM free Mg2+, and 2 mM dATP or dGTP) with solution B (containing 40 mM K+EDTA), yielded traces with single exponential fits of 78.2 ± 0.3 s-1 and 887 ± 15 s-1 for match and mismatch, respectively. (B) Viscosity dependence of mismatched reverse closing. The reverse closing reaction was initiated by rapid mixing of solution A (containing 1 µM Pol β, 400 nM 19dd/36AP(T), 2 mM dGTP, and 5 mM Mg2+) with solution B (containing 40 mM K+EDTA). Corresponding reverse closing rates for 10% and 35% glycerol are 887 ± 15 s-1 and 418 ± 3 s-1, respectively. (C) Mismatched reverse closing dNTP concentration dependence. Reactions were initiated by rapid mixing of solution A (containing 1 µM Pol β, 400 nM 19dd/36AP(T), 5 mM free Mg2+, and dGTP varied) with solution B (containing 20 mM K+EDTA). Single exponential fit yielded rates of 833 ± 17 s-1, 743 ± 13 s-1, 645 ± 8 s-1, 591 ± 7 s-1, for 1 mM, 2 mM, 4 mM, and 6 mM dGTP, respectively. All reactions were performed at 37 °C in Pol β assay buffer at pH 7.6.

122 A 3.0 T:A Match 2.5 T:G Mismatch

2.0

1.5

1.0

0.5 2-AP Fluorescence 2-AP 0.0

0.00 0.02 0.04 0.06 0.08 0.10 0.12 Time (s) B 2.0 10% Glycerol 1.5 35% Glycerol

1.0

0.5

0.0 2-AP Fluorescence 2-AP

-0.5 0.00 0.02 0.04 0.06 0.08 0.10 0.12 Time (s) C 0.8 1 mM 2 mM 0.6 4 mM 6 mM 0.4

0.2

0.0

-0.2 2-AP Fluorescence 2-AP

-0.4

0.00 0.01 0.02 0.03 0.04 0.05 Time (s) Figure 6.5 123

Figure 6.6: Effect of alternate matches/mismatches on reverse closing. The reverse reaction was initiated by rapid mixing of solution A (containing 1 µM Pol β, 400 nM 19dd/36AP(T,C,A), 5 mM free Mg2+, and either 600 µM match or 4 mM mismatch dNTP) with solution B (containing 20 mM K+EDTA). Reactions were performed at 37 °C in Pol β assay buffer at pH 7.6. (A) Reverse conformational closing for T:A match versus T:C mismatch. Single exponential fit yielded rates of 82.4 ± 0.4 s-1 (T:A), 708 ± 11 s-1 (T:C). (B) Reverse conformational closing for C:G match versus C:A mismatch. Single exponential fit yielded rates of 30.2 ± 0.1 s-1 (C:G), and 462 ± 5 s-1 (C:A). (C) Reverse conformational closing for A:T match versus A:G mismatch. Single exponential fit yielded rates of 110 ± 1 s-1 (A:T), 753 ± 15 s-1 (A:G).

124 A 2.0 T:A Match T:C Mismatch 1.5

1.0

0.5

0.0 2-AP Fluorescence 2-AP -0.5

0.00 0.05 0.10 0.15 0.20 Time (s) B 2.0 C:G Match C:A Mismatch 1.5

1.0

0.5

0.0 2-AP Fluorescence 2-AP -0.5

0.00 0.05 0.10 0.15 0.20 Time (s) C 1.5 A:T Match A:G Mismatch

1.0

0.5

0.0 2-AP Fluorescence 2-AP -0.5

0.00 0.05 0.10 0.15 0.20 Time (s) Figure 6.6 125 6.6 R258A Demonstrates Facilitated Reverse Closing.

We have previously demonstrated that the R258A mutant has facilitated

reopening after correct dNTP incorporation (Chapter 4 and (84)). Because of this, it is

anticipated that this mutation could also have a similar effect on the reverse rate of the

conformational closing step before chemistry. For this reason, we compared the rate of

fluorescence change induced by addition of EDTA to a preformed matched ternary

complex for both WT and R258A for multiple correct base pairs. The rates of the reverse

closing are recorded in Table 6.3, along with the corresponding rates of forward

chemistry, as determined from the rate of the slow fluorescence transition (kslow) during

the forward reaction with non-dideoxyterminated primer in stopped-flow (see Chapter 2,

Section 2.4.3 and 2.4.6). As we expected, we found that for all correct base pairs examined, krc was slightly faster for R258A compared to WT, again highlighting the

importance of R258 in subdomain conformational motions. Note also, that for both

enzymes and for all matches reported, the rate of chemistry remains slower than the rate

of reverse closing (k-2> k3, according to Scheme 6.1) – the magnitude of difference

ranging from 3.7 to 6.5-fold slower for WT and 4.5 to 9.1-fold slower for R258A.

126

WT R258A

krc kslow krc kslow T:A 84.7 ± 0.4 s-1 13.0 ± 0.05 s-1 122 ± 1 s-1 13.4 ± 0.05 s-1 C:G 30.8 ± 0.09 s-1 8.27 ± 0.04 s-1 35.5 ± 0.1 s-1 7.82 ± 0.03 s-1 A:T 108 ± 0.6 s-1 17.7 ± 0.06 s-1 140 ± 1 s-1 18.3 ± 0.07 s-1

Table 6.3: Comparison of WT versus R258A reverse closing and forward chemistry rates. Reverse reactions were initiated by rapid mixing of solution A (containing 1 µM Pol β or 3 µM R258A, 400 nM 19dd/36AP(T,C,A), 5 mM free Mg2+, and 600 µM matched dNTP) with solution B (containing 20 mM K+EDTA). Forward reactions were initiated by rapid mixing of solution A (containing 1 µM Pol β or 3 µM R258A and 400 nM 19/36AP(T,C,A)) with solution B (containing 600 µM matched dNTP) in Pol β assay buffer containing 5 mM free Mg2+ at pH 7.6 and 37°C.

6.7 Discussion and Conclusion.

All of the aforementioned studies have provided supporting evidence that the

fluorescence decay observed for our putative reverse closing reaction in stopped-flow

actually corresponds to reverse closing. First and foremost, the direction of fluorescence

decay for reverse conformational closing is opposite that observed for forward

conformational closing initiated by Mg•dNTP binding. Studies demonstrated that the

fluorescence transition is sensitive to differential viscosity, as expected for a step

corresponding to a conformational change, while also demonstrating that the rate of the

reaction is not significantly affected by EDTA, dNTP, DNA, or enzyme concentration.

Saturating versus sub-saturating catalytic Mg2+ concentrations may serve to stabilize the

2+ preformed ternary complex as indicated by a slight reduction in krc as free Mg is

127 increased. Furthermore, mismatched dNTP ternary complexes are significantly

destabilized as evidenced by the significantly faster krc observed for mismatched reverse closing compared to matched reverse closing. Evaluation of the reverse closing for

R258A indicates that it possesses a facilitated reverse conformational step before chemistry, which coincides well with our previous work indicating that it possesses facilitated subdomain reopening after chemistry (Chapter 4).

In light of the entire Pol β mechanism, based upon the aforementioned results under multiple reaction conditions, the rate of reverse closing for both matches (~ 80 s-1) and mismatches (~ 900 s-1) remains significantly faster than chemistry (~ 13 s-1), greater than 5-fold under identical reaction conditions (using T:A match and T:G mismatch, for example). This corresponds to the case according to Scheme 6.1, in which reverse closing is faster than chemistry (k-2> k3). Under such conditions, in the case of Pol β, the rate-limiting step (chemistry) for both matched and mismatched dNTP incorporation plays a significant role in defining the specificity constant (kcat/Km ≈ kpol/Kd) which is

used to define fidelity. Therefore, by analysis of the relative magnitudes of chemistry

and reverse closing for both matched and mismatched dNTP incorporation, in contrast to

what has recently been reported for T7 DNA polymerase (153), this work supports our

long-standing hypothesis the fidelity of Pol β is determined by the difference in free energy between matched and mismatched dNTP incorporation pathways at the chemical transition state. However, since these studies utilized a dideoxy-terminated primer, the loss of the 3’-OH could affect the stability of the preformed ternary complex from which the reverse closing reaction is initiated, as this group is likely essential for proper catalytic Mg2+ binding. For this reason, future reverse closing studies utilizing a normal

128 primer and a non-hydrolyzable dNTP analog, in which the α-β bridging oxygen is

substituted with a methylene group, would lend further insights into the contribution of the reverse of the conformational closing step to Pol β’s fidelity.

129

BIBLIOGRAPHY

(1) Kunkel, T. A. (2003) Considering the cancer consequences of altered DNA polymerase function. Cancer Cell 3, 105-10.

(2) Hubscher, U., Maga, G., and Spadari, S. (2002) Eukaryotic DNA polymerases. Annu Rev Biochem 71, 133-63.

(3) Kunkel, T. A. (2004) DNA replication fidelity. J Biol Chem 279, 16895-8.

(4) Raszka, M., and Kaplan, N. O. (1972) Association by hydrogen bonding of mononucleotides in aqueous solution. Proc Natl Acad Sci U S A 69, 2025-9.

(5) Aboul-ela, F., Koh, D., Tinoco, I., and Martin, F. H. (1985) Base-base mismatches. Thermodynamics of double helix formation for dCA3XA3G + dCT3YT3G (X, Y = A,C,G,D). Nucleic Acids Res 13, 4811.

(6) Loeb, L. A., and Kunkel, T. A. (1982) Fidelity of DNA synthesis. Annu Rev Biochem 51, 429-57.

(7) Aoufouchi, S., Flatter, E., Dahan, A., Faili, A., Bertocci, B., Storck, S., Delbos, F., Cocea, L., Gupta, N., Weill, J. C., and Reynaud, C. A. (2000) Two novel human and mouse DNA polymerases of the polX family. Nucleic Acids Res 28, 3684-93.

(8) Dominguez, O., Ruiz, J. F., Lain de Lera, T., Garcia-Diaz, M., Gonzalez, M. A., Kirchhoff, T., Martinez, A. C., Bernad, A., and Blanco, L. (2000) DNA polymerase mu (Pol mu), homologous to TdT, could act as a DNA mutator in eukaryotic cells. Embo J 19, 1731-42.

130 (9) Reynaud, C. A., Frey, S., Aoufouchi, S., Faili, A., Bertocci, B., Dahan, A., Flatter, E., Delbos, F., Storck, S., Zober, C., and Weill, J. C. (2001) Transcription, beta- like DNA polymerases and hypermutation. Philos Trans R Soc Lond B Biol Sci 356, 91-7.

(10) Ruiz, J. F., Dominguez, O., Lain de Lera, T., Garcia-Diaz, M., Bernad, A., and Blanco, L. (2001) DNA polymerase mu, a candidate hypermutase? Philos Trans R Soc Lond B Biol Sci 356, 99-109.

(11) Ramadan, K., Shevelev, I., and Hubscher, U. (2004) The DNA-polymerase-X family: controllers of DNA quality? Nat Rev Mol Cell Biol 5, 1038-43.

(12) Callebaut, I., and Mornon, J. P. (1997) From BRCA1 to RAP1: a widespread BRCT module closely associated with DNA repair. FEBS Lett 400, 25-30.

(13) Bork, P., Hofmann, K., Bucher, P., Neuwald, A. F., Altschul, S. F., and Koonin, E. V. (1997) A superfamily of conserved domains in DNA damage-responsive cell cycle checkpoint proteins. Faseb J 11, 68-76.

(14) Yu, X., Chini, C. C., He, M., Mer, G., and Chen, J. (2003) The BRCT domain is a phospho-protein binding domain. Science 302, 639-42.

(15) Manke, I. A., Lowery, D. M., Nguyen, A., and Yaffe, M. B. (2003) BRCT repeats as phosphopeptide-binding modules involved in protein targeting. Science 302, 636-9.

(16) Rodriguez, M., Yu, X., Chen, J., and Songyang, Z. (2003) Phosphopeptide binding specificities of BRCA1 COOH-terminal (BRCT) domains. J Biol Chem 278, 52914-8.

(17) Nick McElhinny, S. A., and Ramsden, D. A. (2004) Sibling rivalry: competition between Pol X family members in V(D)J recombination and general double strand break repair. Immunol Rev 200, 156-64.

(18) Doherty, A. J., Serpell, L. C., and Ponting, C. P. (1996) The helix-hairpin-helix DNA-binding motif: a structural basis for non-sequence-specific recognition of DNA. Nucleic Acids Res 24, 2488-97.

131 (19) Wilson, S. H. (1998) Mammalian base excision repair and DNA polymerase beta. Mutat Res 407, 203-15.

(20) Wood, R. D., and Shivji, M. K. (1997) Which DNA polymerases are used for DNA-repair in eukaryotes? Carcinogenesis 18, 605-10.

(21) Gilfillan, S., Benoist, C., and Mathis, D. (1995) Mice lacking terminal deoxynucleotidyl transferase: adult mice with a fetal antigen receptor repertoire. Immunol Rev 148, 201-19.

(22) Komori, T., Okada, A., Stewart, V., and Alt, F. W. (1993) Lack of N regions in antigen receptor variable region genes of TdT-deficient lymphocytes. Science 261, 1171-5.

(23) Gilfillan, S., Dierich, A., Lemeur, M., Benoist, C., and Mathis, D. (1993) Mice lacking TdT: mature animals with an immature lymphocyte repertoire. Science 261, 1175-8.

(24) Papavasiliou, F. N., and Schatz, D. G. (2002) Somatic hypermutation of immunoglobulin genes: merging mechanisms for genetic diversity. Cell 109 Suppl, S35-44.

(25) Li, Z., Woo, C. J., Iglesias-Ussel, M. D., Ronai, D., and Scharff, M. D. (2004) The generation of antibody diversity through somatic hypermutation and class switch recombination. Genes Dev 18, 1-11.

(26) Di Noia, J. M., and Neuberger, M. S. (2007) Molecular mechanisms of antibody somatic hypermutation. Annu Rev Biochem 76, 1-22.

(27) Ruiz, J. F., Lucas, D., Garcia-Palomero, E., Saez, A. I., Gonzalez, M. A., Piris, M. A., Bernad, A., and Blanco, L. (2004) Overexpression of human DNA polymerase mu (Pol mu) in a Burkitt's lymphoma cell line affects the somatic hypermutation rate. Nucleic Acids Res 32, 5861-73.

(28) Schatz, D. G. (2004) V(D)J recombination. Immunol Rev 200, 5-11.

(29) Gellert, M. (2002) V(D)J recombination: RAG proteins, repair factors, and regulation. Annu Rev Biochem 71, 101-32. 132 (30) Lieber, M. R., Ma, Y., Pannicke, U., and Schwarz, K. (2004) The mechanism of vertebrate nonhomologous DNA end joining and its role in V(D)J recombination. DNA Repair (Amst) 3, 817-26.

(31) Rooney, S., Chaudhuri, J., and Alt, F. W. (2004) The role of the non-homologous end-joining pathway in lymphocyte development. Immunol Rev 200, 115-31.

(32) Bertocci, B., De Smet, A., Berek, C., Weill, J. C., and Reynaud, C. A. (2003) Immunoglobulin kappa light chain gene rearrangement is impaired in mice deficient for DNA polymerase mu. Immunity 19, 203-11.

(33) Bertocci, B., De Smet, A., Flatter, E., Dahan, A., Bories, J. C., Landreau, C., Weill, J. C., and Reynaud, C. A. (2002) Cutting edge: DNA polymerases mu and lambda are dispensable for Ig gene hypermutation. J Immunol 168, 3702-6.

(34) Mahajan, K. N., Nick McElhinny, S. A., Mitchell, B. S., and Ramsden, D. A. (2002) Association of DNA polymerase mu (pol mu) with Ku and ligase IV: role for pol mu in end-joining double-strand break repair. Mol Cell Biol 22, 5194-202.

(35) Zhang, Y., Wu, X., Yuan, F., Xie, Z., and Wang, Z. (2001) Highly frequent frameshift DNA synthesis by human DNA polymerase mu. Mol Cell Biol 21, 7995-8006.

(36) Nick McElhinny, S. A., and Ramsden, D. A. (2003) Polymerase mu is a DNA- directed DNA/RNA polymerase. Mol Cell Biol 23, 2309-15.

(37) Ruiz, J. F., Juarez, R., Garcia-Diaz, M., Terrados, G., Picher, A. J., Gonzalez- Barrera, S., Fernandez de Henestrosa, A. R., and Blanco, L. (2003) Lack of sugar discrimination by human Pol mu requires a single glycine residue. Nucleic Acids Res 31, 4441-9.

(38) Zhang, Y., Wu, X., Guo, D., Rechkoblit, O., Taylor, J. S., Geacintov, N. E., and Wang, Z. (2002) Lesion bypass activities of human DNA polymerase mu. J Biol Chem 277, 44582-7.

(39) Havener, J. M., McElhinny, S. A., Bassett, E., Gauger, M., Ramsden, D. A., and Chaney, S. G. (2003) Translesion synthesis past platinum DNA adducts by human DNA polymerase mu. Biochemistry 42, 1777-88.

133 (40) Covo, S., Blanco, L., and Livneh, Z. (2004) Lesion bypass by human DNA polymerase mu reveals a template-dependent, sequence-independent nucleotidyl transferase activity. J Biol Chem 279, 859-65.

(41) Duvauchelle, J. B., Blanco, L., Fuchs, R. P., and Cordonnier, A. M. (2002) Human DNA polymerase mu (Pol mu) exhibits an unusual replication slippage ability at AAF lesion. Nucleic Acids Res 30, 2061-7.

(42) Hirose, F., Hotta, Y., Yamaguchi, M., and Matsukage, A. (1989) Difference in the expression level of DNA polymerase beta among mouse tissues: high expression in the pachytene spermatocyte. Exp Cell Res 181, 169-80.

(43) Gu, H., Marth, J. D., Orban, P. C., Mossmann, H., and Rajewsky, K. (1994) Deletion of a DNA polymerase beta gene segment in T cells using cell type- specific gene targeting. Science 265, 103-6.

(44) Sugo, N., Aratani, Y., Nagashima, Y., Kubota, Y., and Koyama, H. (2000) Neonatal lethality with abnormal neurogenesis in mice deficient in DNA polymerase beta. Embo J 19, 1397-404.

(45) Srivastava, D. K., Berg, B. J., Prasad, R., Molina, J. T., Beard, W. A., Tomkinson, A. E., and Wilson, S. H. (1998) Mammalian abasic site base excision repair. Identification of the reaction sequence and rate-determining steps. J Biol Chem 273, 21203-9.

(46) Kubota, Y., Nash, R. A., Klungland, A., Schar, P., Barnes, D. E., and Lindahl, T. (1996) Reconstitution of DNA base excision-repair with purified human proteins: interaction between DNA polymerase beta and the XRCC1 protein. Embo J 15, 6662-70.

(47) Nicholl, I. D., Nealon, K., and Kenny, M. K. (1997) Reconstitution of human base excision repair with purified proteins. Biochemistry 36, 7557-66.

(48) Sobol, R. W., Horton, J. K., Kuhn, R., Gu, H., Singhal, R. K., Prasad, R., Rajewsky, K., and Wilson, S. H. (1996) Requirement of mammalian DNA polymerase beta in base excision repair. Nature 379, 183-6.

134 (49) Parikh, S. S., Mol, C. D., and Tainer, J. A. (1997) Base excision repair enzyme family portrait: integrating the structure and chemistry of an entire DNA repair pathway. Structure 5, 1543-50.

(50) Lindahl, T., and Wood, R. D. (1999) Quality control by DNA repair. Science 286, 1897-905.

(51) Piersen, C. E., Prasad, R., Wilson, S. H., and Lloyd, R. S. (1996) Evidence for an imino intermediate in the DNA polymerase beta deoxyribose phosphate excision reaction. J Biol Chem 271, 17811-5.

(52) Prasad, R., Singhal, R. K., Srivastava, D. K., Molina, J. T., Tomkinson, A. E., and Wilson, S. H. (1996) Specific interaction of DNA polymerase beta and DNA ligase I in a multiprotein base excision repair complex from bovine testis. J Biol Chem 271, 16000-7.

(53) Nash, R. A., Caldecott, K. W., Barnes, D. E., and Lindahl, T. (1997) XRCC1 protein interacts with one of two distinct forms of DNA ligase III. Biochemistry 36, 5207-11.

(54) Davies, J. F., 2nd, Almassy, R. J., Hostomska, Z., Ferre, R. A., and Hostomsky, Z. (1994) 2.3 A crystal structure of the catalytic domain of DNA polymerase beta. Cell 76, 1123-33.

(55) Pelletier, H., Sawaya, M. R., Kumar, A., Wilson, S. H., and Kraut, J. (1994) Structures of ternary complexes of rat DNA polymerase beta, a DNA template- primer, and ddCTP. Science 264, 1891-903.

(56) Pelletier, H., Sawaya, M. R., Wolfle, W., Wilson, S. H., and Kraut, J. (1996) Crystal structures of human DNA polymerase beta complexed with DNA: implications for catalytic mechanism, processivity, and fidelity. Biochemistry 35, 12742-12761.

(57) Sawaya, M. R., Prasad, R., Wilson, S. H., Kraut, J., and Pelletier, H. (1997) Crystal structures of human DNA polymerase beta complexed with gapped and nicked DNA: evidence for an induced fit mechanism. Biochemistry 36, 11205-15.

135 (58) Pelletier, H., Sawaya, M. R., Wolfle, W., Wilson, S. H., and Kraut, J. (1996) A Structural basis for metal ion mutagenicity and nucleotide selectivity in human DNA polymerase beta. Biochemistry 35, 12762-12777.

(59) Arndt, J. W., Gong, W., Zhong, X., Showalter, A. K., Liu, J., Dunlap, C. A., Lin, Z., Paxson, C., Tsai, M.-D., and Chan, M. K. (2001) Insight into the catalytic mechanism of DNA polymerase beta: structures of intermediate complexes. Biochemistry 40, 5368-5375.

(60) Krahn, J. M., Beard, W. A., Miller, H., Grollman, A. P., and Wilson, S. H. (2003) Structure of DNA polymerase beta with the mutagenic DNA lesion 8- oxodeoxyguanine reveals structural insights into its coding potential. Structure 11, 121-127.

(61) Batra, V. K., Beard, W. A., Shock, D. D., Pedersen, L. C., and Wilson, S. H. (2005) Nucleotide-induced DNA polymerase active site motions accommodating a mutagenic DNA intermediate. Structure 13, 1225-33.

(62) Krahn, J. M., Beard, W. A., and Wilson, S. H. (2004) Structural insights into DNA polymerase beta deterrents for misincorporation support an induced-fit mechanism for fidelity. Structure 12, 1823-32.

(63) Prasad, R., Beard, W. A., and Wilson, S. H. (1994) Studies of gapped DNA substrate binding by mammalian DNA polymerase beta. Dependence on 5'- phosphate group. J Biol Chem 269, 18096-101.

(64) Matsumoto, Y., and Kim, K. (1995) Excision of deoxyribose phosphate residues by DNA polymerase beta during DNA repair. Science 269, 699-702.

(65) Prasad, R., Beard, W. A., Strauss, P. R., and Wilson, S. H. (1998) Human DNA polymerase beta deoxyribose phosphate lyase. Substrate specificity and catalytic mechanism. J Biol Chem 273, 15263-70.

(66) Huang, H., Chopra, R., Verdine, G. L., and Harrison, S. C. (1998) Structure of a covalently trapped catalytic complex of HIV-1 reverse transcriptase: implications for drug resistance. Science 282, 1669-75.

136 (67) Doublie, S., Tabor, S., Long, A. M., Richardson, C. C., and Ellenberger, T. (1998) Crystal structure of a bacteriophage T7 DNA replication complex at 2.2 A resolution. Nature 391, 251-8.

(68) Ollis, D. L., Brick, P., Hamlin, R., Xuong, N. G., and Steitz, T. A. (1985) Structure of large fragment of Escherichia coli DNA polymerase I complexed with dTMP. Nature 313, 762-6.

(69) Steitz, T. A., Smerdon, S. J., Jager, J., and Joyce, C. M. (1994) A unified polymerase mechanism for nonhomologous DNA and RNA polymerases. Science 266, 2022-2025.

(70) Beard, W. A., and Wilson, S. H. (2006) Structure and mechanism of DNA polymerase beta. Chem Rev 106, 361-382.

(71) Werneburg, B. G., Ahn, J., Zhong, X., Hondal, R. J., Kraynov, V. S., and Tsai, M. D. (1996) DNA polymerase beta: pre-steady-state kinetic analysis and roles of arginine-283 in catalysis and fidelity. Biochemistry 35, 7041-50.

(72) Ahn, J., Kraynov, V. S., Zhong, X., Werneburg, B. G., and Tsai, M. D. (1998) DNA polymerase beta: effects of gapped DNA substrates on dNTP specificity, fidelity, processivity, and conformational changes. Biochem J 831, 79-87.

(73) Ahn, J., Werneburg, B. G., and Tsai, M. D. (1997) DNA polymerase beta: structure-fidelity relationship from pre-steady-state kinetic analyses of all possible correct and incorrect base pairs for wild type and R283A mutant. Biochemistry 36, 1100-7.

(74) Aboul-ela, F., Koh, D., Tinoco, I., and Martin, F. H. (1985) Base-base mismatches. Thermodynamics of double helix formation for dCA3XA3G + dCT3YT3G (X, Y = A,C,G,D). Nucleic Acids Res 13, 4811.

(75) Doublie, S., Sawaya, M. R., and Ellenberger, T. (1999) An open and closed case for all polymerases. Structure 7, R31-R35.

(76) Kunkel, T. A., and Bebenek, K. (2000) DNA replication fidelity. Ann Rev Biochem 69, 497-529.

137 (77) Johnson, K. A. (1993) Conformational coupling in DNA polymerase fidelity. Ann Rev Biochem 62, 685-713.

(78) Wong, I., Patel, S. S., and Johnson, K. A. (1991) An induced-fit kinetic mechanism for DNA replication fidelity: direct measurement by single-turnover kinetics. Biochemistry 30, 526-37.

(79) Dunlap, C. A., and Tsai, M. D. (2002) Use of 2-aminopurine and tryptophan fluorescence as probes in kinetic analyses of DNA polymerase beta. Biochemistry 41, 11226-35.

(80) Bakhtina, M., Lee, S., Wang, Y., Dunlap, C., Lamarche, B., and Tsai, M. D. (2005) Use of viscogens, dNTPalphaS, and rhodium(III) as probes in stopped- flow experiments to obtain new evidence for the mechanism of catalysis by DNA polymerase beta. Biochemistry 44, 5177-87.

(81) Yang, L., Beard, W. A., Wilson, S. H., Broyde, S., and Schlick, T. (2004) Highly organized but pliant active site of DNA polymerase beta: compensatory mechanisms in mutant enzymes revealed by dynamics simulations and energy analyses. Biophys J 86, 3392-3408.

(82) Steitz, T. A. (1993) DNA- and RNA-dependent DNA polymerases. Curr Opin Struct Biol 3, 31-38.

(83) Steitz, T. A. (1999) DNA polymerases: structural diversity and common mechanisms. J Biol Chem 274, 17395-17398.

(84) Bakhtina, M., Roettger, M. P., Kumar, S., and Tsai, M. D. (2007) A unified kinetic mechanism applicable to multiple DNA polymerases. Biochemistry 46, 5463-72.

(85) Zhong, X., Patel, S. S., and Tsai, M.-D. (1998) DNA polymerase beta: dissecting the functional roles of the two metal ions with Cr(III)dTTP. J Am Chem Soc 120, 235-236.

(86) Wada, K., Wada, Y., Ishibashi, F., Gojobori, T., and Ikemura, T. (1992) Codon usage tabulated from the GenBank genetic sequence data. Nucleic Acids Res 20 Suppl, 2111-8.

138 (87) Boule, J. B., Johnson, E., Rougeon, F., and Papanicolaou, C. (1998) High-level expression of murine terminal deoxynucleotidyl transferase in Escherichia coli grown at low temperature and overexpressing argU tRNA. Mol Biotechnol 10, 199-208.

(88) Menge, K. L., Hostomsky, Z., Nodes, B. R., Hudson, G. O., Rahmati, S., Moomaw, E. W., Almassy, R. J., and Hostomska, Z. (1995) Structure-function analysis of the mammalian DNA polymerase beta active site: role of aspartic acid 256, arginine 254, and arginine 258 in nucleotidyl transfer. Biochemistry 34, 15934-42.

(89) Johnson, K. A. (1986) Rapid kinetic analysis of mechanochemical adenosinetriphosphatases. Methods Enzymol 134, 677-705.

(90) Gutfreund, H. (1972) Enzymes: Physical Principles, Wiley-Interscience, New York.

(91) Johnson, K. A. (1992) Transient-state kinetic analysis of enzyme reaction pathways. The Enzymes 20, 1-61.

(92) Fiala, K. A., Abdel-Gawad, W., and Suo, Z. (2004) Pre-steady-state kinetic studies of the fidelity and mechanism of polymerization catalyzed by truncated human DNA polymerase lambda. Biochemistry 43, 6751-62.

(93) Zhong, X., Patel, S. S., Werneburg, B. G., and Tsai, M. D. (1997) DNA polymerase beta: multiple conformational changes in the mechanism of catalysis. Biochemistry 36, 11891-900.

(94) Fiala, K. A., and Suo, Z. (2004) Mechanism of DNA polymerization catalyzed by Sulfolobus solfataricus P2 DNA Polymerase IV. Biochemistry 43, 2116-25.

(95) Washington, M. T., Prakash, L., and Prakash, S. (2001) Yeast DNA polymerase eta utilizes an induced-fit mechanism of nucleotide incorporation. Cell 107, 917- 27.

(96) Patel, S. S., Wong, I., and Johnson, K. A. (1991) Pre-steady-state kinetic analysis of processive DNA replication including complete characterization of an exonuclease-deficient mutant. Biochemistry 30, 511-25.

139 (97) Dahlberg, M. E., and Benkovic, S. J. (1991) Kinetic mechanism of DNA polymerase I (Klenow fragment): identification of a second conformational change and evaluation of the internal equilibrium constant. Biochemistry 30, 4835-43.

(98) Bryant, F. R., Johnson, K. A., and Benkovic, S. J. (1983) Elementary steps in the DNA polymerase I reaction pathway. Biochemistry 22, 3537-46.

(99) Kuchta, R. D., Benkovic, P., and Benkovic, S. J. (1988) Kinetic mechanism whereby DNA polymerase I (Klenow) replicates DNA with high fidelity. Biochemistry 27, 6716-25.

(100) Johnson, A. A., Tsai, Y., Graves, S. W., and Johnson, K. A. (2000) Human mitochondrial DNA polymerase holoenzyme: reconstitution and characterization. Biochemistry 39, 1702-8.

(101) Boule, J. B., Rougeon, F., and Papanicolaou, C. (2001) Terminal deoxynucleotidyl transferase indiscriminately incorporates ribonucleotides and deoxyribonucleotides. J Biol Chem 276, 31388-93.

(102) Kato, K. I., Goncalves, J. M., Houts, G. E., and Bollum, F. J. (1967) Deoxynucleotide-polymerizing enzymes of calf thymus gland. II. Properties of the terminal deoxynucleotidyltransferase. J Biol Chem 242, 2780-9.

(103) Ramadan, K., Shevelev, I. V., Maga, G., and Hubscher, U. (2004) De novo DNA synthesis by human DNA polymerase lambda, DNA polymerase mu and terminal deoxyribonucleotidyl transferase. J Mol Biol 339, 395-404.

(104) Cases-Gonzalez, C. E., Gutierrez-Rivas, M., and Menendez-Arias, L. (2000) Coupling ribose selection to fidelity of DNA synthesis. The role of Tyr-115 of human immunodeficiency virus type 1 reverse transcriptase. J Biol Chem 275, 19759-67.

(105) Astatke, M., Ng, K., Grindley, N. D., and Joyce, C. M. (1998) A single side chain prevents Escherichia coli DNA polymerase I (Klenow fragment) from incorporating ribonucleotides. Proc Natl Acad Sci U S A 95, 3402-7.

140 (106) Gao, G., Orlova, M., Georgiadis, M. M., Hendrickson, W. A., and Goff, S. P. (1997) Conferring RNA polymerase activity to a DNA polymerase: a single residue in reverse transcriptase controls substrate selection. Proc Natl Acad Sci U S A 94, 407-11.

(107) Sousa, R., and Padilla, R. (1995) A mutant T7 RNA polymerase as a DNA polymerase. Embo J 14, 4609-21.

(108) Huang, Y., Beaudry, A., McSwiggen, J., and Sousa, R. (1997) Determinants of ribose specificity in RNA polymerization: effects of Mn2+ and deoxynucleoside monophosphate incorporation into transcripts. Biochemistry 36, 13718-28.

(109) Juarez, R., Ruiz, J. F., Nick McElhinny, S. A., Ramsden, D., and Blanco, L. (2006) A specific loop in human DNA polymerase mu allows switching between creative and DNA-instructed synthesis. Nucleic Acids Res 34, 4572-82.

(110) Gu, J., Lu, H., Tippin, B., Shimazaki, N., Goodman, M. F., and Lieber, M. R. (2007) XRCC4:DNA ligase IV can ligate incompatible DNA ends and can ligate across gaps. Embo J 26, 1010-23.

(111) Kraynov, V. S., Werneburg, B. G., Zhong, X., Lee, H., Ahn, J., and Tsai, M. D. (1997) DNA polymerase beta: analysis of the contributions of tyrosine-271 and asparagine-279 to substrate specificity and fidelity of DNA replication by pre- steady-state kinetics. Biochem J 323 (Pt 1), 103-11.

(112) Meuth, M. (1984) The genetic consequences of nucleotide precursor pool imbalance in mammalian cells. Mutat Res 126, 107-12.

(113) Garcia-Diaz, M., Bebenek, K., Krahn, J. M., Kunkel, T. A., and Pedersen, L. C. (2005) A closed conformation for the Pol lambda catalytic cycle. Nat Struct Mol Biol 12, 97-8.

(114) Joyce, C. M. (1997) Choosing the right sugar: how polymerases select a nucleotide substrate. Proc Natl Acad Sci U S A 94, 1619-22.

(115) Moon, A. F., Garcia-Diaz, M., Bebenek, K., Davis, B. J., Zhong, X., Ramsden, D. A., Kunkel, T. A., and Pedersen, L. C. (2007) Structural insight into the substrate specificity of DNA polymerase mu. Nat Struct Mol Biol 14, 45-53.

141 (116) Liu, J., and Tsai, M. D. (2001) DNA polymerase beta: pre-steady-state kinetic analyses of dATP alpha S stereoselectivity and alteration of the stereoselectivity by various metal ions and by site-directed mutagenesis. Biochemistry 40, 9014- 22.

(117) Wagner, S. D., and Neuberger, M. S. (1996) Somatic hypermutation of immunoglobulin genes. Annu Rev Immunol 14, 441-57.

(118) Neuberger, M. S., and Milstein, C. (1995) Somatic hypermutation. Curr Opin Immunol 7, 248-54.

(119) Gonzalez-Fernandez, A., Gupta, S. K., Pannell, R., Neuberger, M. S., and Milstein, C. (1994) Somatic mutation of immunoglobulin lambda chains: a segment of the major intron hypermutates as much as the complementarity- determining regions. Proc Natl Acad Sci U S A 91, 12614-8.

(120) El-Deiry, W. S., Downey, K. M., and So, A. G. (1984) Molecular mechanisms of manganese mutagenesis. Proc Natl Acad Sci U S A 81, 7378-82.

(121) Goodman, M. F., Keener, S., Guidotti, S., and Branscomb, E. W. (1983) On the enzymatic basis for mutagenesis by manganese. J Biol Chem 258, 3469-75.

(122) Weymouth, L. A., and Loeb, L. A. (1978) Mutagenesis during in vitro DNA synthesis. Proc Natl Acad Sci U S A 75, 1924-8.

(123) Yang, L., Beard, W. A., Wilson, S. H., Broyde, S., and Schlick, T. (2002) Polymerase beta simulations suggest that Arg258 rotation is a slow step rather than large subdomain motions per se. J Mol Biol 317, 651-671.

(124) Radhakrishnan, R., and Schlick, T. (2004) Orchestration of cooperative events in DNA synthesis and repair mechanism unraveled by transition path sampling of DNA polymerase beta's closing. Proc Natl Acad Sci U S A 101, 5970-5975.

(125) Yang, L., Arora, K., Beard, W. A., Wilson, S. H., and Schlick, T. (2004) Critical role of magnesium ions in DNA polymerase beta's closing and active site assembly. J Am Chem Soc 126, 8441-8453.

142 (126) Gavish, B., and Werber, M. M. (1979) Viscosity-dependent structural fluctuations in enzyme catalysis. Biochemistry 18, 1269-75.

(127) Shaffer, J., Sun, G., and Adams, J. A. (2001) Nucleotide release and associated conformational changes regulate function in the COOH-terminal Src kinase, Csk. Biochemistry 40, 11149-11155.

(128) Lew, J., Taylor, S. S., and Adams, J. A. (1997) Identification of a partially rate- determining step in the catalytic mechanism of cAMP-dependent protein kinase: A transient kinetic study using stopped-flow fluorescence spectroscopy. Biochemistry 36, 6717-6724.

(129) Kurz, L. C., Weitkamp, E., and Frieden, C. (1987) Adenosine deaminase: viscosity studies and the mechanism of binding of substrate and of ground- and transition-state analog inhibitors. Biochemistry 26, 3027-32.

(130) Kanosue, Y., Kojima, S., and Ohkata, K. (2004) Influence of solvent viscosity on the rate of hydrolysis of dipeptides by carboxypeptidase Y. J Phys Org Chem 17, 448-457.

(131) Cole, P. A., Burn, P., Takacs, B., and Walsh, C. T. (1994) Evaluation of the catalytic mechanism of recombinant human Csk (C-terminal Src kinase) using nucleotide analogs and viscosity effects. J Biol Chem 269, 30880-7.

(132) Yang, L., Beard, W. A., Wilson, S. H., Broyde, S., and Schlick, T. (2002) Polymerase beta simulations suggest that Arg258 rotation is a slow step rather than large subdomain motions per se. J Mol Biol 317, 651-671.

(133) Showalter, A. K., and Tsai, M. D. (2002) A reexamination of the nucleotide incorporation fidelity of DNA polymerases. Biochemistry 41, 10571-6.

(134) Showalter, A. K., Lamarche, B. J., Bakhtina, M., Su, M. I., Tang, K. H., and Tsai, M. D. (2006) Mechanistic comparison of high-fidelity and error-prone DNA polymerases and ligases involved in DNA repair. Chem Rev 106, 340-60.

(135) Johnson, S. J., and Beese, L. S. (2004) Structures of mismatch replication errors observed in a DNA polymerase. Cell 116, 803-16.

143 (136) Trincao, J., Johnson, R. E., Wolfle, W. T., Escalante, C. R., Prakash, S., Prakash, L., and Aggarwal, A. K. (2004) Dpo4 is hindered in extending a G.T mismatch by a reverse wobble. Nat Struct Mol Biol 11, 457-62.

(137) Showalter, A. K., and Tsai, M. D. (2001) A DNA polymerase with specificity for five base pairs. J Am Chem Soc 123, 1776-7.

(138) Roettger, M. P., Fiala, K. A., Sompalli, S., Dong, Y., and Suo, Z. (2004) Pre- steady-state kinetic studies of the fidelity of human DNA polymerase mu. Biochemistry 43, 13827-38.

(139) Kati, W. M., Johnson, K. A., Jerva, L. F., and Anderson, K. S. (1992) Mechanism and fidelity of HIV reverse transcriptase. J Biol Chem 267, 25988-97.

(140) Eger, B. T., and Benkovic, S. J. (1992) Minimal kinetic mechanism for misincorporation by DNA polymerase I (Klenow fragment). Biochemistry 31, 9227-36.

(141) Sawaya, M. R., Pelletier, H., Kumar, A., Wilson, S. H., and Kraut, J. (1994) Crystal structure of rat DNA polymerase beta: evidence for a common polymerase mechanism. Science 264, 1930-5.

(142) Shah, A. M., Li, S. X., Anderson, K. S., and Sweasy, J. B. (2001) Y265H mutator mutant of DNA polymerase beta. Proper teometric alignment is critical for fidelity. J Biol Chem 276, 10824-31.

(143) Opresko, P. L., Sweasy, J. B., and Eckert, K. A. (1998) The mutator form of polymerase beta with amino acid substitution at tyrosine 265 in the hinge region displays an increase in both base substitution and frame shift errors. Biochemistry 37, 2111-9.

(144) Clairmont, C. A., Narayanan, L., Sun, K. W., Glazer, P. M., and Sweasy, J. B. (1999) The Tyr-265-to-Cys mutator mutant of DNA polymerase beta induces a mutator phenotype in mouse LN12 cells. Proc Natl Acad Sci U S A 96, 9580-5.

(145) Li, S. X., Vaccaro, J. A., and Sweasy, J. B. (1999) Involvement of phenylalanine 272 of DNA polymerase beta in discriminating between correct and incorrect deoxynucleoside triphosphates. Biochemistry 38, 4800-8.

144 (146) Shah, A. M., Maitra, M., and Sweasy, J. B. (2003) Variants of DNA polymerase beta extend mispaired DNA due to increased affinity for nucleotide substrate. Biochemistry 42, 10709-17.

(147) Starcevic, D., Dalal, S., and Sweasy, J. (2005) Hinge residue Ile260 of DNA polymerase beta is important for enzyme activity and fidelity. Biochemistry 44, 3775-84.

(148) Starcevic, D., Dalal, S., Jaeger, J., and Sweasy, J. B. (2005) The hydrophobic hinge region of rat DNA polymerase beta is critical for substrate binding pocket geometry. J Biol Chem 280, 28388-93.

(149) Tang, K. H., Niebuhr, M., Tung, C. S., Chan, H. C., Chou, C. C., and Tsai, M. D. (2008) Mismatched dNTP incorporation by DNA polymerase beta does not proceed via globally different conformational pathways. Nucleic Acids Res E- published.

(150) Rothwell, P. J., Mitaksov, V., and Waksman, G. (2005) Motions of the fingers subdomain of klentaq1 are fast and not rate limiting: implications for the molecular basis of fidelity in DNA polymerases. Mol Cell 19, 345-55.

(151) Stengel, G., Gill, J. P., Sandin, P., Wilhelmsson, L. M., Albinsson, B., Norden, B., and Millar, D. (2007) Conformational dynamics of DNA polymerase probed with a novel fluorescent DNA base analogue. Biochemistry 46, 12289-97.

(152) Luo, G., Wang, M., Konigsberg, W. H., and Xie, X. S. (2007) Single-molecule and ensemble fluorescence assays for a functionally important conformational change in T7 DNA polymerase. Proc Natl Acad Sci U S A 104, 12610-5.

(153) Tsai, Y. C., and Johnson, K. A. (2006) A new paradigm for DNA polymerase specificity. Biochemistry 45, 9675-87.

(154) Sucato, C. A., Upton, T. G., Kashemirov, B. A., Osuna, J., Oertell, K., Beard, W. A., Wilson, S. H., Florian, J., Warshel, A., McKenna, C. E., and Goodman, M. F. (2008) DNA polymerase beta fidelity: halomethylene-modified leaving groups in pre-steady-state kinetic analysis reveal differences at the chemical transition state. Biochemistry 47, 870-879.

145 (155) Sucato, C. A., Upton, T. G., Kashemirov, B. A., Batra, V. K., Martinek, V., Xiang, Y., Beard, W. A., Pedersen, L. C., Wilson, S. H., McKenna, C. E., Florian, J., Warshel, A., and Goodman, M. F. (2007) Modifying the beta,gamma leaving- group bridging oxygen alters nucleotide incorporation efficiency, fidelity, and the catalytic mechanism of DNA polymerase beta. Biochemistry 46, 461-71.

(156) Joyce, C. M., and Benkovic, S. J. (2004) DNA polymerase fidelity: kinetics, structure, and checkpoints. Biochemistry 43, 14317-24.

(157) Purohit, V., Grindley, N. D., and Joyce, C. M. (2003) Use of 2-aminopurine fluorescence to examine conformational changes during nucleotide incorporation by DNA polymerase I (Klenow fragment). Biochemistry 42, 10200-11.

(158) Florian, J., Goodman, M. F., and Warshel, A. (2005) Computer simulations of protein functions: searching for the molecular origin of the replication fidelity of DNA polymerases. Proc Natl Acad Sci U S A 102, 6819-24.

(159) Xiang, Y., Oelschlaeger, P., Florian, J., Goodman, M. F., and Warshel, A. (2006) Simulating the effect of DNA polymerase mutations on transition-state energetics and fidelity: evaluating amino acid group contribution and allosteric coupling for ionized residues in human pol beta. Biochemistry 45, 7036-48.

(160) Beard, W. A., Shock, D. D., Vande Berg, B. J., and Wilson, S. H. (2002) Efficiency of correct nucleotide insertion governs DNA polymerase fidelity. J Biol Chem 277, 47393-8.

146