Molecular Drivers of Specificity in Human Ribonucleotide

MOLECULAR DRIVERS OF SPECIFICITY IN HUMAN RIBONUCLEOTIDE

REDUCTASE

ANDREW JOHN KNAPPENBERGER

Submitted in partial fulfillment of the requirements

For the degree of Doctor of Philosophy

Dissertation advisor: Dr. Michael E. Harris

Department of Biochemistry

CASE WESTERN RESERVE UNIVERSITY

May, 2017

CASE WESTERN RESERVE UNIVERSITY SCHOOL OF GRADUATE STUDIES

We hereby approve the dissertation of Andrew John Knappenberger Candidate for the degree of Doctor of Philosophy*.

(signed) Committee Chair Hung-Ying Kao

Committee Member Chris Dealwis

Committee Member Michael Harris

Committee Member Arne Rietsch

Committee Member Martin Snider

(date) March 10, 2017

*We also certify that written approval has been obtained for any proprietary material contained therein.

Table of Contents Acknowledgements ...... ix

Chapter 1 : Introduction ...... 13

Molecular recognition by enzymes is central to life ...... 14

Recognition of multiple substrates and regulation by allostery are common features

of enzymes ...... 16

Ribonucleotide reductase is an essential enzyme for all known cellular organisms .... 18

Loop 2 is central to specificity regulation in ribonucleotide reductase ...... 21

Measuring multiple substrate kinetics through internal competition permits

comprehensive description of enzymatic specificity ...... 22

Chapter 2 : Nucleoside Analogue Triphosphates Allosterically Regulate Human

Ribonucleotide Reductase and Identify Chemical Determinants That Drive Substrate

Specificity ...... 28

Abstract ...... 29

Introduction ...... 30

Results and discussion ...... 42

Conclusions ...... 70

Materials and methods ...... 72

Chapter 3 : Phylogenetic Comparative Sequence Analysis and Functional Studies of

Mutant Enzymes Reveal Compensatory Amino Acid Substitutions in Loop 2 of

Human Ribonucleotide Reductase ...... 77

Abstract ...... 78

Introduction ...... 79

Results and discussion ...... 85

Conclusions ...... 119

Materials and methods ...... 122

Chapter 4 : Summary and Future Directions ...... 126

Summary ...... 126

Future directions ...... 131

Bibliography ...... 137

List of Tables

Table 3-1. Numerical values from activity assays in Figure 3-6...... 107

Table 3-2. Numerical values from activity assays in Figure 3-9...... 116

iii

List of Figures

Figure 1-1. Structure of eukaryotic R1...... 25

Figure 1-2. Conformational changes in loop 2...... 26

Figure 2-1. Structure of human ribonucleotide reductase and location

of the S-site and C-site...... 35

Figure 2-2. Comparison of RR loop 2 across species...... 37

Figure 2-3. Measurements of RR specificity from this and other studies...... 39

Figure 2-4. Comparison of eukaryotic and bacterial RRs with dGTP or dTTP

bound in the S-site...... 40

Figure 2-5. Application of internal competition to measurement of

hRR specificity...... 56

Figure 2-6. Measurement of native hRR and ScRR specificities...... 58

Figure 2-7. The specificities directed by a series of pyrimidine effector analogs...... 60

Figure 2-8. Structures of eukaryotic RR bound to S-site ligands...... 62

Figure 2-9. The specificities directed by a series of purine effector analogs...... 64

Figure 2-10. Competition between purine effector analogs and the natural S-site

effector dTTP...... 66

Figure 2-11. Effects of the D287A substitution on hRR substrate recognition...... 68

Figure 3-1. Three-dimensional structure of hRR...... 83

Figure 3-2. Sequence conservation in eukaryotic ribonucleotide reductase and

the extent of variation in loop 2...... 100

Figure 3-3. Loop 2 sequences from Figure 3-2C mapped onto a eukaryotic tree of life...... 102

Figure 3-4. Crystal structure of hR1 bound to dTTP and GDP...... 103

Figure 3-5. Circular dichroism (CD) of hRR variants...... 104

Figure 3-6. Activity and specificity of the hRR variants...... 105

Figure 3-7. Size exclusion chromatography (SEC) of hRR variants in the presence of dGTP, dTTP or ATP...... 109

Figure 3-8. SEC profiles of wild-type and P294K hRR large subunit in the presence of 3 mM ATP...... 111

Figure 3-9. Activity and specificity of N291G mutant hR1 in the presence of 2- aminopurine-drTP, dITP, and dZeb...... 112

Figure 3-10. Activity and specificity of the hRR variants in the presence of both

ATP and dGTP/dTTP...... 114

Figure 3-11. Correspondence between the present in vitro experiments and “natural experiments” of evolution...... 118

List of Abbreviations

• RR – ribonucleotide reductase.

• DNA – deoxyribonucleic acid.

• dNTP – deoxyribonucleotide diphosphate.

• ATP – adenosine triphosphate.

• dATP – deoxyadenosine triphosphate.

• dGTP – deoxyguanosine triphosphate.

• dTTP – deoxythymidine triphosphate.

• tRNA – transfer ribonucleic acid.

• ATC – aspartate transcarbamylase.

• CTP – cytidine triphosphate.

• SAMHD1 – SAM domain and HD domain-containing protein 1.

• HIV – human immunodeficiency virus.

• ADP – adenosine diphosphate.

• CDP – cytidine diphosphate.

• GDP – guanosine diphosphate.

• UDP – uridine diphosphate.

• R1 – ribonucleotide reductase large subunit.

• R2 – ribonucleotide reductase small subunit.

• C-site – catalytic site, the active site of ribonucleotide reductase.

• S-site – specificity site.

• A-site – activity site.

• dNDP – deoxynucleotide diphosphate.

• hRR – human ribonucleotide reductase.

• ScRR – S. cerevisiae ribonucleotide reductase.

• AMPPNP – adenylylimidodiphosphate.

• PDB – Protein Data Bank.

• HPLC – high-performance liquid chromatography.

• TmRR – Thermotoga maritima RR.

• EcRR – E. coli RR.

• aa – amino acid.

• hR1 – human ribonucleotide reductase large subunit.

• 5FdUTP – 5-fulorodeoxyuridine triphosphate.

• dUTP – deoxyuridine triphosphate.

• dZeb – deoxyzebularine triphosphate.

• 2-aminopurine-drTP – 2-aminopurine-deoxyribose triphosphate.

• N2dATP – 2-amino-deoxyadenosine triphosphate.

• dITP – deoxyinosine triphosphate.

• hRRM1 – human ribonucleotide reductase large subunit.

• hRRM2 – human ribonucleotide reductase small subunit.

• ScRR1 – S. cerevisiae ribonucleotide reductase large subunit.

• ScR2R4 – S. cerevisiae ribonucleotide reductase small subunit.

• DTT – dithiothreitol.

• CD – circular dichroism spectroscopy.

• SEC – size exclusion chromatography.

vii

• MALS – multi-angle light scattering.

• SD – standard deviation.

• cDNA – complementary deoxyribonucleic acid.

• mRR – M. musculus ribonucleotide reductase.

• SAXS – small-angle X-ray scattering.

• EM – electron microscopy.

viii

Acknowledgements

I thank my wife, Bethany Knappenberger. Her love, companionship, and support make

scientific pursuits both possible and worthwhile. Attempting to express my gratitude towards her in words can only fail, so I devote my life to her happiness instead.

I thank the rest of my family, especially my father, George Knappenberger. For helping me to work towards best possible version of myself, from the very beginning.

I thank my mentor, Michael Harris. He had enough faith in me to take me on as a student, and is so skilled as a mentor that I was able to justify that faith.

I thank the other members of my pre-thesis and thesis committees: Chris Dealwis, Pieter

deHaseth, Hung-Ying Kao, Arne Rietsch, and Martin Snider. Dr. Dealwis in particular

has been like a second mentor to me. The continuous intellectual and material support I

have received from my committee has been absolutely instrumental throughout my

graduate studies.

I thank my collaborators and the members of the Harris and Dealwis labs, especially Md.

Faiz Ahmad, Sarah Huff, Daniel Kellerman, Hsuan-Chun Lin, Tessianna Misko,

Courtney Niland, Mayra Pedraza, Edward Ollie, Kandice Simmons, Rajesh Viswanathan,

and Jing Zhao. For your solidarity, for your advice, for all of your help. You filled the space between the benches.

I thank everyone I have mentored as a graduate student: Da’Quan Bush-Pierce, Yuchen

Yang, Danni Liang, Alexandria Williams, Julia Hillenbrand, Emily Strong, Junye Wang,

Maya Smith, Marie-Louise Kloster, Han Hu, and especially Reena Sheth. I hope you learned half as much from me as I learned from you.

I thank countless other teachers, mentors, supporters, far too numerous to name, stretching from today back to the first people to look up at the stars with wonder. I am eternally grateful to stand on the shoulders of such giants.

Molecular Drivers of Specificity in Human Ribonucleotide Reductase

ABSTRACT

ANDREW JOHN KNAPPENBERGER

Ribonucleotide reductase (RR) enables organisms to grow and reproduce by

providing the deoxynucleotides necessary for DNA replication. Human RR can reduce

any of the four nucleoside diphosphates at the 2′ position, and binding of dNTP effectors

(ATP/dATP, dGTP, dTTP) modulates substrate specificity at an allosteric site near the

active site. This binding is known to change the conformation of a short loop (loop 2) that

bridges the two sites. Although the relationship between effector binding and enzymatic

specificity is well-established, the specific interactions through which RR recognizes the

effector and changes the conformation of loop 2 remain untested. The precise

interconnecting roles of the individual amino acid residues in loop 2 are also unclear.

Here, we systematically interrogate the contribution of each effector functional group

using a panel of dNTP analogues and multiple substrate kinetic assays, and confirm a key

prediction from crystal structures by showing that interactions with amino acid residue

D287 in loop 2 are essential for perturbing its conformational space. In addition, we examine natural variation in loop 2 among eukaryotes as an alternative to more

traditional alanine mutagenesis. We find that amino acid sequence space among eukaryotes is dominated by two major types of loop 2, and that they differ by just two substitutions, and these substitutions are highly nonconservative with respect to structure

(N291G and P294K). Simultaneously introducing both mutations into human RR partially rescues the effects of the single mutations. This work sheds light on a key molecular mechanism by which organisms generate balanced dNTP pools for timely and accurate DNA replication and repair.

xii

Chapter 1 : Introduction

Molecular recognition by enzymes is central to life

The chemistry of life is many orders of magnitude too slow to proceed unassisted

at biological timescales. For example, pyrimidine nucleotide synthesis requires

generating uridine monophosphate from orotidine monophosphate. This decarboxylation

reaction normally proceeds with a half-life of 78 million years1. The enzyme orotidine 5’-

phosphate decarboxylase enhances the rate of this process by a factor of 1017, allowing

organisms to grow and reproduce on timescales of minutes rather than eras1. However,

living things exist far from chemical equilibrium, and if all possible chemical processes

proceeded at enzymatic rates within the cell, it would swiftly dissolve into a simple

mixture of small molecules2. Therefore, the reactions that remain uncatalyzed are just as

important to life as the reactions that enjoy rate enhancement.

Like an orchestra’s conductor, the cell controls the expression levels of all its

enzymes so that some chemical processes may proceed quickly while other entities retain

their stability. Enzymes are often given names like “orotidine 5’-phosphate decarboxylase,” implying that they perform precisely one chemical reaction on precisely one chemical substrate. As demonstrated above, this exquisite specificity is as vital to biology as the degree of rate enhancement that enzymes provide. Enzymes typically assist chemistry via judicious positioning of amino acid or cofactor groups near functional groups on the substrate3. This changes the character of the substrate’s

environment such that the reaction’s transition state is more stable than it would be in

solution4. Different chemical changes proceed through different transition states, so it is

easy to imagine that different types of amino acids or cofactors placed in different

positions might facilitate different chemical transitions. It is therefore largely the position

and nature of catalytic functional groups that specifies the type of reaction an enzyme

catalyzes, which restricts its repertoire to just one chemical operation.

Although there are open questions remaining with respect to how enzymes

stabilize transition states, this is not the main focus of the current work. Rather, the

present study concerns itself chiefly with how the enzyme selects its substrate. Orotidine

5’-phosphate decarboxylase does not, for example, decarboxylate phosphatidylserine to phosphatidylethanolamine, even though the net chemical reactions are essentially identical5. How then does a decarboxylase assert its identity as an orotidine 5’-phosphate

decarboxylase and not a phosphatidylserine decarboxylase? How does a protein kinase

refrain from phosphorylating every free serine, threonine, and tyrosine residue on every

protein in the cell? How does the ribosome faithfully translate the genetic code into the

language of proteins and amino acids?

Ever since the discovery of enzymes’ incredible specificity, and long before the

advent of structural biology, biochemists have suspected that intimate physical

interactions between the enzyme and its substrate are an essential component of

enzymatic specificity. Emil Fischer wrote the following in 1894: “To give an illustration I

will say that enzyme and glucoside [substrate] must fit together like lock and key in order

to be able to exercise a chemical action on each other6-7.” This ‘lock-and-key’ hypothesis

was almost a necessity, as early researchers observed that enzymes often discriminate

even between optical isomers of a molecule. It also benefits from extreme parsimony:

potential alternative substrates remain unprocessed because they simply do not fit into the

enzyme’s active site. In the over 100 years that have followed Fischer’s postulation, the

lock-and-key hypothesis has suffered some refinements, but the preponderance of

experimental evidence has borne it out. Virtually every enzyme known has an active site

that exhibits extensive shape complementarity with its substrate8.

Recognition of multiple substrates and regulation by allostery are common features of enzymes

Even though enzymes are often presented in textbooks as accepting just one substrate, in truth this is rarely absolutely true. Nature is replete with enzymes that catalyze a chemical reaction on more than one canonical substrate, particularly within nucleic acid biology. In an example from our group, E. coli RNase P catalyzes the 3’ end

maturation of 87 different pre-tRNAs, and has been shown in vitro to catalyze the

cleavage of 4096 mutant substrates with comparable efficiency9. Enzymes such as RNase

P typically have a constant specificity; that is, substrate A is processed with higher

efficiency than substrate B, which in turn is processed more efficiently than substrate C,

and so on. Enzymes with relatively stringent preferences can be thought of as single-

substrate enzymes, while those with relatively lenient preferences can be thought of as

multi-substrate enzymes.

Regulation of overall enzyme activity is a central feature of biological catalysis.

Cells modulate enzyme activity through the obvious route of controlling how much

enzyme is made: if it would be advantageous to downregulate the expression of a given

enzyme, its mRNA can be targeted for degradation or a key transcription factor can be

sequestered10-11. But enzymes are also regulated through what is known as allostery12.

This word comes from Greek roots and refers to the fact that an allosterically regulated

enzyme has two ligand-binding sites with different (allo) stereo (steric) specificity13.

Therefore, allostery typically means that an enzyme’s activity is regulated through a binding interaction with some secondary ligand.

A common regulatory strategy in metabolic pathways is to use the end product of a synthetic pathway as an allosteric ligand that regulates the first committed step in its synthesis. For example, pyrimidine nucleotides are synthesized from simple precursors through a series of enzymatic steps. The first step in this synthetic pathway is catalyzed by the enzyme aspartate transcarbamylase (ATC)14. ATC has an active site which binds the substrates (aspartate and carbamyl phosphate), but it also has a second allosteric site which binds cytidine triphosphate (CTP). When pyrimidine nucleotides are abundant within the cell, the need for de novo synthesis is diminished. CTP binds to ATC, and this physical interaction changes the shape of the enzyme so that it can no longer process substrates. When nucleotides become less abundant in the cell, CTP is bound to the allosteric site of the enzyme less often, and the enzyme is able to process more molecules of precursor along the first committed step of nucleotide synthesis.

There are many examples of enzymes with activity regulated via allostery. But can an enzyme’s specificity be regulated in this way? In theory, it should be possible. An enzyme’s specificity is dictated by the shape of its active site, and binding of allosteric ligands is known to change a protein’s conformation. Therefore it is possible to imagine a protein constructed such that its active site and an allosteric site can communicate through some flexible region of the protein. When a ligand binds to the allosteric site of this imagined protein, it should change the shape of the linker region such that a substrate

A is favored. If the allosteric site can accept a second ligand, that other ligand might change the shape of the active site such that substrate B is favored. The enzyme could

then use the relative abundance of these two ligands as a barometer for the overall

metabolic state of the cell and respond accordingly.

As it turns out, such proteins do exist. However, they are much less common than

one might naïvely expect. The author is aware of only two proteins that exhibit this type

of regulation. The first is SAMHD1, a protein that hydrolyzes dNTPs to their

corresponding nucleosides15. It is involved in regulation of the cell’s nucleotide pools and

defense against viruses including HIV16. This enzyme binds dNTPs at three sites, and

changing the balance of dNTPs introduced into an in vitro reaction mixture changes the substrate preferences of the enzyme17-18.

Ribonucleotide reductase is an essential enzyme for all known cellular

organisms

The other protein bears structural similarity to SAMHD1 that raises the possibility that the two proteins descended from a common ancestor, and it arguably plays an even more important role in the chemistry of life17, 19. The enzyme is called ribonucleotide

reductase (RR), so named because it accepts ribonucleotides as substrates and reduces

them at the 2′ position to form deoxyribonucleotides20. The cell uses the resultant deoxynucleotides as the building blocks for all its DNA replication and repair, and RRs

are tightly regulated at the levels of transcription, localization, and allostery21-27. RR’s

position within metabolism gives it central importance to all life as we know it –

reduction by RR is the only known natural method for making deoxynucleotides, without

which no cell may reproduce or maintain its genome.

RRs are universal to cellular organisms, and lifestyle differences dictate that there exist three separate classes of RRs, classified based on the mechanisms by which they

generate free radicals for catalysis20, 28. Class I RRs require the presence of oxygen, and

generate a stable tyrosyl radical through the use of a di-iron cofactor. Class II RRs are insensitive to the presence of oxygen, and generate a thiyl radical with the aid of an adenosylcobalamin cofactor. Class III RRs require that oxygen be absent from the environment, and use an iron-sulfur cluster to generate a glycyl radical. Depending on lifestyle and ancestry, organisms may have genes for several RRs, but all known cellular organisms express at least one. Herpes simplex virus is an example of a biological entity that might be considered neither alive nor cellular, yet still encodes an RR within its genome29. Human RR (hRR) is of Class I and reduces NDPs. Like other RRs it is capable

of accepting any of the four NDPs (ADP, CDP, GDP, UDP) as substrates. This makes

hRR a multi-substrate enzyme, which is not altogether unusual – especially considering

the structural similarity among its potential substrates. Rather, what makes RR unique is

that its preferences among its substrates are modulated through binding of allosteric

ligands. The allosteric ligands are dNTPs, and they provide the enzyme with information

about the current dNTP balance so that it may produce an appropriate balance of

deoxynucleotide products.

hRR is composed of two subunits: a large subunit termed R1 and a small subunit

termed R2. The large subunit is active as a homodimer and contains all of the enzyme’s

nucleotide-binding sites, while the small subunit contains a tyrosyl radical that is

necessary to initiate catalysis19. The R2 subunit binds relatively distal from the active site

and must transport a free radical ca. 35 Å via proton-coupled electron transfer30. The R1

subunit is the chief focus of this work. R1 has three nucleotide-binding sites per

monomer: the Catalytic site or active site (C-site), the Specificity site (S-site), and the

Activity site (A-site) (Figure 1-1)19, 22. The C-site is the catalytically active site that binds the NDP substrate. The Activity site is capable of binding ATP and dATP, and both molecules cause R1 to hexamerize19, 31-37. ATP is activating to the enzyme while dATP is

deactivating. It has been shown in E. coli RR that dATP-bound RR holds the R2 subunit

in a position in which it cannot interact with its canonical binding site, and small-angle

X-ray scattering experiments on human RR produce results consistent with this model32,

38. This represents a classical negative feedback system: excess dATP will cause hRR to stop catalyzing formation of dNDP products, which in turn prevents dNTPs from building to potentially hazardous levels. The S-site is capable of binding ATP/dATP, dGTP, and dTTP, and all these molecules cause R1 to dimerize. Each ligand causes hRR to adopt different substrate preferences when bound in the S-site. A model in which

ATP/dATP binding causes hRR to favor CDP reduction, dGTP binding promotes ADP reduction, and dTTP promotes GDP reduction is well-supported by works including those that will be summarized in later chapters19, 21-22, 25, 27, 39-40.

While the substrate preferences of RR in the presence of its canonical allosteric

ligands are well-known, it is less clear how binding of an allosteric ligand induces a

conformational change in RR. Mechanistic insight into the allosteric regulation of hRR

specificity has the potential to generate design principles for drugs that target hRR.

Although RR’s catalytic (active) site and activity site are both targets of clinical drugs,

the specificity site remains an unexplored avenue of drug development for hRR inhibition or modulation37, 41-44. High-resolution crystal structures can provide important clues.

Crystal structures have been determined for RRs from T. maritima, E. coli, S. cerevisiae

(ScRR), and H. sapiens19, 22, 25, 27. Among these, the crystal structures from S. cerevisiae

are of particular interest for studies of hRR. This is because ScRR has extensive sequence

similarity with hRR and has been crystalized in the context of all its canonical

substrate/effector pairs: AMPPNP (an ATP analog)/CDP, AMPPNP/UDP, dGTP/ADP,

and dTTP/GDP (Figure 1-2).

Loop 2 is central to specificity regulation in ribonucleotide reductase

Crystal structures reveal that conformational change upon effector binding is not

extensive. Rather, it is largely confined to a short loop spanning the S-site and C-site.

This loop is called loop 2, and it assumes a unique conformation in the presence of each

substrate/effector pair. Because loop 2 forms part of both the S-site and C-site, is

conformationally flexible, and is closely apposed with the Watson-Crick faces of both the

substrate and the effector, loop 2 is likely to be involved in the molecular information

transfer from the S-site to the C-site. However, crystal structures present a static picture of the protein, and they leave it impossible to deconvolute potential contributions from multiple functional groups. Therefore, important questions remain regarding the roles of individual nucleotide functional groups in RR function.

In addition, it is as yet unclear how all of loop 2’s amino acids work in concert to translate effector binding at the S-site to substrate preference at the C-site. Site-directed mutagenesis is a time-honored method for carrying out structure-function studies on both proteins and nucleic acids. However, loop 2 is highly conserved among eukaryotes, and so it is likely that its constituent amino acid residues work as part of a single system to carry out its function (see Chapter 3). As an alternative to more traditional alanine

mutagenesis, it is possible to explore the amino acid sequence variation that exists in

nature to pinpoint substitutions that are likely to provide insight into loop 2’s function.

Inspection of a sequence alignment of eukaryotic RR reveals that the loop 2 region is

dominated by two types of sequences: sequences identical to the human sequence, and

sequences that harbor two substitutions: N291G and P294K (human numbering). Both of

these substitutions are highly nonconservative, and the fact that they very often occur

together raises the possibility that they evolved in the context of other mutations to

produce a system which is functionally equivalent to hRR. As the present work will

demonstrate, exploring the results of “natural experiments” in RR function is fertile

ground for design of hRR mutants that reveal key insights into loop 2’s function.

Measuring multiple substrate kinetics through internal competition permits

comprehensive description of enzymatic specificity

Traditional enzyme biochemistry aims to describe the reaction of an enzyme with

a single substrate by determining the individual rate and equilibrium constants associated

45 with that process . Of particular interest are the parameters kcat, the second-order rate

constant for reaction of the enzyme-substrate complex to free enzyme and product, and

KM, the equilibrium constant that describes the saturation of the enzyme’s active site

(Scheme 1-1)46.

Scheme 1-1

The ratio of these two constants is kcat/KM, the second-order rate constant that describes

the enzyme’s catalytic efficiency for that substrate. This parameter can also be expressed

as V/K, and the meanings of the two terms are essentially equivalent for the purposes of

this study. If one measures the V/K for a given substrate and divides it by the V/K for another substrate, the resultant ratio is an expression of the enzyme’s specificity between the two. Therefore, it is possible to determine an enzyme’s specificity by measuring the kcat and KM for all substrates of interest. This type of inquiry has even been done as a method for measuring the native specificity of RR33, 47. However, this technique is

laborious and susceptible to variations in active enzyme concentration. A more rapid and

precise method for quantifying the specificity of an enzyme for any number of substrates

involves the use of internal competition48-51. In an internal competition model, all

substrates are included in the reaction and each substrate acts as a competitive inhibitor

for all other substrates present. This means that the concentration of active enzyme does

not affect the results, and if the substrates are < 10% reacted one can compare two

substrates’ V/K values in the following way (Equation 1-1) 52-57:

= Equation 1-1 푉 푣2 � �퐾�2 푆2 1 푉 1 푣 � �퐾�1 �푆 � Where v is the observed reaction velocity for a given substrate, S is its concentration, and

V/K is its second-order rate constant. In practice, the parameter of interest for these experiments is the ratio of V/K values, which can also be expressed using the term rk. By

comparing rk values, we can compare the specificity of an enzyme for any substrate

relative to a reference substrate. Because hRR specificity is central to this study,

measuring the formation of all four dNDP products by anion exchange high-performance

liquid chromatography (HPLC) and using those measurements to generate an rk value for

each substrate is a technique that will be employed throughout the following chapters58.

In the current work, we made targeted changes to the structures of loop 2 and the dNTP effector. By observing their effects on RR function, we have revealed key functional groups that make strong contributions to allosteric regulation in human RR.

Figure 1-1. Structure of eukaryotic R1.1 The homodimer of the S. cerevisiae R1 subunit is depicted. One monomer is shown in

yellow; the other is shown in green. The locations of the catalytic site (active site or C- site), activity site (A-site), and specificity site (S-site) are indicated. The substrate (ADP) is shown as blue spheres. The effector (dGTP) is shown as purple spheres.

1 Xu, Hai, et al. "Structures of eukaryotic ribonucleotide reductase I provide insights into dNTP regulation." Proceedings of the National Academy of Sciences of the United States of America 103.11 (2006): 4022-4027. Copyright 2006 National Academy of Sciences.

Figure 1-2. Conformational changes in loop 2.2 Each panel is a stereoview of the active site of S. cerevisiae RR in complex with a

substrate and an effector. Loop 2 is shown on the right. The effector is not pictured. Panel

2 Xu, Hai, et al. "Structures of eukaryotic ribonucleotide reductase I provide insights into dNTP regulation." Proceedings of the National Academy of Sciences of the United States of America 103.11 (2006): 4022-4027. Copyright 2006 National Academy of Sciences.

A depicts RR in complex with ADP and dGTP; B, dTTP and GDP; C, AMPPNP and

CDP; D, AMPPNP and UDP. AMPPNP is a structural analog of ATP.

Chapter 2 : Nucleoside Analogue Triphosphates Allosterically Regulate Human Ribonucleotide Reductase and Identify Chemical Determinants That Drive Substrate Specificity3

3 Knappenberger, Andrew J., et al. Biochemistry 55.41 (2016): 5884-5896.

Abstract

Class I Ribonucleotide reductase (RR) maintains balanced pools of deoxyribonucleotide substrates for DNA replication by converting ribonucleoside diphosphates (NDPs) to 2′-deoxyribonucleoside diphosphates (dNDPs). Binding of deoxynucleoside triphosphate (dNTP) effectors (ATP/dATP, dGTP and dTTP) modulates the specificity of Class I RR for CDP, UDP, ADP, and GDP substrates. Crystal structures of bacterial and eukaryotic RRs show that dNTP effectors and NDP substrates bind on either side of a flexible 9-amino acid loop (loop 2). Interactions with the effector nucleobase alter loop 2 geometry, resulting in changes in specificity among the four NDP substrates of RR. However, the functional groups proposed to drive specificity remain untested. Here, we use deoxynucleoside analog triphosphates to determine the nucleobase functional groups that drive human RR (hRR) specificity. The results demonstrate that the 5-methyl, O4 and N3 groups of dTTP contribute to specificity for GDP. The O6 and protonated N1 of dGTP direct specificity for ADP. In contrast, the unprotonated N1 of adenosine is the primary determinant of ATP/dATP-directed specificity for CDP.

Structure models from X-ray crystallography of eukaryotic RR suggest that the side chain of D287 in loop 2 is involved in binding of dGTP and dTTP, but not dATP/ATP. This feature is consistent with experimental results showing that a D287A mutant of hRR is deficient in allosteric regulation by dGTP and dTTP, but not ATP/dATP. Together, these data define the effector functional groups that are the drivers of human RR specificity and provide constraints for evaluating models of allosteric regulation.

Introduction

Regulation by allostery is a fundamental property of proteins and of enzymes in

particular59. Enzymes acting as key control points in metabolism are typically under tight

allosteric regulation60; therefore, functional insights into the nature of protein allostery

can provide a better understanding of biology as well as design principles for therapeutic

development. Decades of research beginning with inquiries into the structure and

function of hemoglobin have revealed numerous examples of allostery, and mechanisms

of site-to-site communication have been thoroughly explored in many systems61-63.

Providing experimental constraints useful for developing and benchmarking models of

protein allostery remains an important challenge in biochemistry. Ribonucleotide

reductases (RRs) present a key example of enzyme regulation by allostery in which

additional mechanistic detail at a chemical level would be highly valuable to both biology

and biomedicine.

RRs are a ubiquitous and highly conserved class of enzymes that catalyze the

reduction of ribonucleotides to produce 2′-deoxyribonucleotides. RR is essential for DNA synthesis and deoxynucleotide pool maintenance, and the activity of various RRs is tightly regulated at the levels of transcription, localization, and allostery21-27. The total

enzymatic activity of Class I RR and its specificity among the four NDP substrates are

under tight allosteric control by nucleoside and deoxynucleoside triphosphate effectors21.

ATP and dATP regulate total activity through binding at the Activity site (A-site). In

eukaryotic Class Ia RRs, this association results in the formation of active or inactive

hexamers of RR, respectively19, 31-37. Binding of dNTP effectors to the Specificity site (S- site) of Class I RR modulates the relative kcat/Km for ADP, CDP, GDP, and UDP at the

Catalytic site (C-site) (Figure 2-1A). A general model in which ATP/dATP binding directs reduction of CDP and UDP, dGTP directs ADP reduction, and dTTP directs GDP reduction is well-supported by biochemical and structural data19, 21-22, 25, 27, 39-40. However,

key information about the precise workings of RR allostery is still lacking. Areas in

which open questions still exist include the role of specific protein/ligand interactions in

altering RR structure, and a description of the energetic coupling between effector

binding, protein conformational changes and the resulting effects on substrate

discrimination.

Recent high-resolution crystal structure models of the large subunits of human RR

(hRR), Saccharomyces cerevisiae RR (ScRR), Thermotoga maritima RR (TmRR) and

Escherichia coli RR (EcRR) provide key insights into how allosteric communication is accomplished. They highlight the importance of a flexible 9-aa region called loop 2 that

forms part of both the C-site and the S-site19, 22, 25, 27 (Figure 2-1A). For both Bacteria and

Eukaryote Class I RRs, binding of dNTP effectors to the S-site biases the conformational

distribution of loop 2 such that it presents a binding pocket with more favorable potential

interactions for a particular NDP substrate in the C-site (Figure 2-1B). Amino acids in loop 2 that are proximal to the effector nucleobase are conserved at positions Q288, G289, and R293 (hRR numbering), consistent with a central role in allosteric communication

(Figure 2-2A). Residues G295 and A296 are also conserved and may play a more structural role involving loop flexibility. Several residues including Y285, D287 and

K292 are also conserved in eukaryotic RR enzymes and are variable between eukaryotes and Bacteria. The pattern of substrate specificity and regulation by effector binding is

highly similar among bacterial and eukaryotic RRs (Figure 2-3), consistent with common

mechanisms of allosteric communication among these enzymes.

Due to recent advances in RR structural biology and limited structure-function

studies, the roles of conserved residues as well as the potential consequences of

phylogenetic sequence variation in loop 2 are now coming into focus. The available X-

ray crystal structures of Class I and II RRs show that the effector nucleobase functional

groups form contacts with the N-terminal amino acids of loop 2. Structural models of both T. maritima and E. coli RR suggest that the N1 and N6 of dATP are Hydrogen- bonded to the main chain amide and carbonyl groups of the same homologous residue in both enzymes (K202 in TmRR (PDB ID: 1XJM)) and S293 in EcRR (PDB ID: 5CNS))25

(Figure 2-2B). Similar interactions between ScRR and ATP are inferred from data

obtained in the presence of AMPPNP (PDB ID: 2CVU)22. The use of AMPPNP as an

analog of ATP/dATP has the potential to produce adventitious interactions and binding

modes that are not representative of those that occur between the enzyme and native

effector in vivo. In spite of this, the ScRR structural model remains valuable for our

studies into the biomedically important human enzyme because it depicts a eukaryotic

RR in complex with an S-site ligand that is similar to the native ATP/dATP. This model

shows that the adenine N1 and N6 also contact the main chain amide and carbonyl of the

same homologous position (D287 in hRR). The overall geometry of loop 2, including the

positions of conserved residues, appears similar to the model of EcRR (Figure 2-2C).

However, the relative importance of adenosine functional groups in stabilizing the

appropriate conformation of loop 2 is not known.

The model of allosteric regulation by dGTP and dTTP is less clear, although some

common features can be observed. The N1 and N2 of dGTP appear to direct specificity

for ADP in both the TmRR model (PDB ID: 1XJK)27 and in the EcRR structure (PDB ID:

5CNU)25 by interaction with backbone functional groups. However, dGTP contacts the

same K202 position as ATP in the TmRR structure, while in EcRR, dGTP contacts the

universally conserved Q294 (PDB ID: 5CNU) (Q288 in hRR). In contrast, the O6 of

dGTP contacts the main chain amide groups of G289 and G290 and the N1 Hydrogen-

bonds to the side chain of D287 (PDB ID: 2CVX) in ScRR. As a consequence, the loop

geometries and positions of conserved residues are distinct between the bacterial and

eukaryotic RR models of the dGTP-ADP bound forms (Figure 2-4A and B). A similar

situation is observed for dTTP, which makes minimal contacts with loop 2 in the T. maritima ribonucleotide reductase structure model (PDB ID: 1XJJ). In the EcRR structure, the nucleobase N3 contacts the backbone carbonyl of C292, altering contacts involving Q294 (PDB ID: 5CNV). The dTTP nucleobase contacts residues outside loop 2 in models of ScRR and hRR. This allows a more compact geometry (PDB ID: 2CVW) compared to the bacterial enzyme (Figure 2-4C and D).

The available structural models make clear that non-covalent interactions between the effector nucleobase and the N-terminal side of loop 2 are likely to contribute to allosteric communication. However, there is still significant ambiguity regarding the chemical features of the effector that drive specificity and the interactions that stabilize the appropriate loop 2 conformations. Importantly, our current understanding of the interactions between loop 2 and the dNTP effector in the biomedically important human enzyme remains comparatively limited. Determining the precise contributions of effector

functional groups to substrate specificity is essential for evaluating any model of RR allosteric regulation. Therefore, we systematically tested the ability of a series of deoxynucleoside triphosphate analogs to allosterically regulate hRR specificity. The results identify the primary nucleobase functional groups that direct hRR specificity for its four native NDP substrates. This new information provides experimental support for current structural models of effector recognition by eukaryotic Class I RR. Achieving a detailed chemical picture of RR regulation by nucleotide binding also contributes to a more complete understanding of protein allostery in general. Although the present study is directed at understanding basic principles of RR regulation, because altered nucleotide pools inhibit cell growth, insight into the chemical basis for RR specificity is potentially useful for development of artificial effectors with therapeutic applications.

Figure 2-1. Structure of human ribonucleotide reductase and location of the S-site and C-site. (PDB ID: 3HND, 3.21 Å resolution). A. Three-dimensional structure of the large subunit dimer of hRR. The polypeptide backbone is shown in gray with loop 2 highlighted in red.

The locations of the S-site, C-site, and loop 2 are indicated. dTTP is bound in the S-site and GDP is bound in the C-site. B. Detail of the S-site, loop 2 and C-site motif involved in allosteric regulation. Effector (dTTP, blue sticks) binding alters loop 2 conformation

(purple sticks) to determine the optimal substrate (GDP, green sticks) in the C-site. Loop

2 amino acids are shown as purple sticks with the conserved residues Q288, G289, and

R293 shown as red sticks. Q288 and R293 have been implicated in substrate recognition through in vitro and in vivo studies (see text). G289 is also well-conserved and likely adds additional flexibility to loop 2, though its role has not been tested via structure- function studies. In this structural model, R293 forms an indirect contact with the phosphate groups of the NDP substrate (GDP).

Figure 2-2. Comparison of RR loop 2 across species.

A. Sequence alignment of the loop 2 region of RRs discussed in the present study19, 22, 25,

27, 64-66. Positions 288, 289, and 293 (human numbering) are in red. Positions 288, 289 and 293 are fully conserved within this sample, while considerable variation exists at position 287. hRR has an aspartic acid residue at position 287, and mutation of D287 to

alanine severely restricts the ability of hRR to modulate its specificity in response to effector binding (Figure 2-11). B. Crystal structure of EcRR bound to dATP and CDP

(not shown) (PDB ID: 5CNS, 2.97 Å resolution)25. C. Crystal structure of ScRR bound to

AMPPNP and CDP (not shown) (PDB ID: 2CVU, 2.9 Å resolution)22. In B and C, the S- site ligand is shown as purple sticks. Loop 2 amino acids are shown as white sticks, with the conserved glycine, glutamine, and arginine residues shown as red sticks. Key contacts described in the original references are shown.

Figure 2-3. Measurements of RR specificity from this and other studies. H. sapiens and S. cerevisiae measurements are taken from the present study. E. coli measurements are estimated by eye from Rofougaran et al., 2008. P. aeruginosa measurements are estimated by eye from Jonna et al., 2015. T. brucei measurements are taken from Hofer et al., 1998. M. musculus measurements are taken from Chimploy et al.,

2001. If kcat/KM values were available, these were normalized to that of the substrate with the highest kcat/KM value and reported as relative kcat/KM values. Relative activity values were converted into relative kcat/KM values using Equation 2-1. When numerical data were available, these were used directly. When data were presented graphically, their numerical values were estimated by eye.

Figure 2-4. Comparison of eukaryotic and bacterial RRs with dGTP or dTTP bound in the S-site. In each panel the S-site ligand is shown as purple sticks. The polypeptide backbone of loop 2 is shown as white sticks. The universally conserved glycine, glutamine, and

arginine residues are red with the side chain of the glutamine shown. Key contacts

described in the original references are shown. A. Crystal structure of EcRR bound to

dGTP and ADP (not shown) (PDB ID: 5CNU, 3.4 Å resolution). B. Crystal structure of

ScRR bound to dGTP and ADP (not shown) (PDB ID: 2CVX, 2.2 Å resolution). C.

Crystal structure of EcRR bound to dTTP and GDP (not shown) (PDB ID: 5CNV, 3.2 Å

resolution). D. Crystal structure of hRR bound to dTTP and GDP (not shown) (PDB ID:

3HND, 3.21 Å resolution).

Results and discussion

Quantification of the NDP substrate specificity of yeast and human RR by internal

competition: RR is intrinsically a multi-substrate enzyme. Thus, a means for determining

the rate constants for all four NDP substrates is required in order to dissect mechanisms

of allosteric regulation. Relatively few studies comprehensively interrogate RR

specificity, owing to the technical difficulties and low throughput of single-substrate

assays33, 47. Internal competition is an alternative method for quantifying enzyme rate

constants that involves analyzing the change in the ratios of concentrations of alternative

substrates or products as a function of reaction progress in reactions containing multiple

substrates48-51. This approach has potential advantages over direct fitting of kinetic data

by offering higher precision, increased throughput and less sensitivity to variation in

enzyme specific activity48-49. Internal competition kinetics have been used extensively to

measure kinetic isotope effects51, 67 and substrate and product ratios have been measured

using a wide range of analytical methods. Importantly, Mathews and colleagues

demonstrated simultaneous quantification of RR reduction of all four NDP substrates

using boronate chromatography to remove unreacted ribonucleotides followed by

separation of dNDP products using ion exchange HPLC58. This method has been used

previously to investigate RR enzymology, and permits direct, sensitive and quantitative

comparison of the effects of different allosteric effectors on specificity using internal

competition kinetics.

As shown in Figure 2-5A, a single RR enzyme can combine with one of four

NDP substrates and react to form dNDP products. Thus, multiple substrate enzymatic reactions of this kind follow internal competition kinetics in which differences in

56-57 observed reaction rates reflect differences in kcat/Km for individual NDP substrates .

Under steady-state conditions, the reaction velocity for a given substrate relative to a

reference substrate should be proportional to the relative concentrations of the two

substrates and their relative specificity constant or rk (Equation 2-1).

Therefore, quantification of the distribution of dNDP products in RR reactions

containing all four NDP substrates in steady state reactions permits the relative kcat/Km for

each substrate to be calculated. To validate this approach for hRR, we tested whether

r changes in the concentrations of two NDPs ([S1]/[Sref]) affect the observed k values. We

assayed hRR in the presence of varied concentrations of substrates ADP and CDP (0.3

mM ADP and 3 mM CDP; 1.8mM ADP and 1.8 mM CDP; 3 mM ADP and 0.3 mM CDP)

and effectors ATP and dGTP at 1 mM and 0.75 mM, respectively (Figure 2-5D-E).

Product accumulation and the initial substrate concentrations were quantified directly by

anion exchange HPLC, providing increased precision in the calculated rk values and ensuring that experimental reactions remained within steady-state conditions. The results show that vobsADP/vobsCDP is proportional to [ADP]/[CDP] and that the calculated relative

r specificity constant, kADP, varies less than 2-fold over the 100-fold range of substrate

concentrations examined.

To provide a baseline for comparison of the effects of nucleoside analog triphosphates on allosteric regulation, we first quantified the specificity of hRR and ScRR

bound to one of the three native allosteric effectors ATP, dGTP, or dTTP, included in the

reaction at 1 mM, 0.75 mM, and 1.6 mM, respectively (Figure 2-6). Consistent with

amino acid sequence conservation in loop 2, the two species’ enzymes have highly

similar specificity19, 22. The substrate specificity directed by the three native nucleotide

effectors is not absolute; each effector state will accept multiple substrates, albeit over a wide range of kcat/Km values. The data in Figure 2-6 show that all three effector-bound

states process an alternative substrate with an rk value within ca. 10-fold of the cognate

substrate. Multiple NDPs can serve as substrates for hRR and ScRR regardless of which effector is bound, although there can be a >1000-fold difference in the highest and lowest

35, 47, 58, 68-69 kcat/Km values . For example, we find that for ATP-bound hRR, CDP is the favored substrate as expected, however, the kcat/Km for ADP, GDP and UDP are within

100-fold of the value for CDP. The yeast enzyme is similar but has a 100-fold lower

kcat/Km for GDP compared to ADP and UDP. In the dGTP-bound state, the optimal ADP substrate is favored over CDP by only about 10-fold, while there is 100-fold

discrimination against UDP relative to ADP. The kcat/Km for GDP is at least 1000-fold lower than the value for ADP, although an accurate measurement was not possible due to a low level of hydrolysis of the dGTP effector to form dGDP. Analysis of the dTTP- bound state of RR demonstrates >10-fold discrimination against CDP relative to the optimal substrate GDP. For hRR bound to dTTP, both ADP and UDP have kcat/Km values

that are 100-fold lower than the optimal substrate, GDP. However, the dTTP bound state

of ScRR displays further discrimination over UDP, which was not detectable in our assay.

Several groups previously used internal competition or individual substrate assays

to quantify specificity for RR enzymes from diverse species. In some studies the data

were interpreted quantitatively in terms of relative specificity constants, while others

report the relative activities when substrates are present in equimolar concentrations35, 47,

58, 68-69. These inquiries provide important context and calibration points for these new

data, and they are summarized in Figure 2-3. When compared across RR enzymes from

different species and methods of analysis, the data confirm the universal core selection

rules, but also reveal some variation in the relative rates for non-optimal substrates. For

example, the present data show good qualitative agreement with the allosteric rules

derived from studies of EcRR, though ATP-bound EcRR processes UDP more efficiently

and dGTP-bound EcRR processes CDP less efficiently than hRR or ScRR35.

As described above, current models of eukaryotic and bacterial RRs include key

interactions between effector nucleobases and amino acids proximal to D287 and the

conserved Q288 (hRR numbering) in loop 2. However, functional tests of the proposed

chemical interactions between loop 2 and nucleotide effectors remain comparatively

limited, especially for the biomedically important hRR. Therefore, we systematically

varied effector nucleobase functional groups and quantitatively compared the resulting

effects on hRR substrate specificity.

Identification of the chemical groups responsible for allosteric regulation of hRR specificity by dTTP: To determine the molecular features of dTTP that contribute to substrate discrimination, we tested the specificities directed by a set of pyrimidine nucleotide effector analogs (2-thio-dTTP, 5FdUTP, dCTP, 5-methyl-dCTP, dUTP, dZeb)

(Figure 2-7). When assayed individually at 100 µM concentration, all of the analogs tested except dCTP and 5-methyl-dCTP result in significant RR activity that is within 3-5

fold of that observed in the presence of dTTP. The concentration of 100 µM is ~200-fold

greater than the measured KD for dTTP binding. For comparison, this concentration is also 2-fold higher than the concentration of dTTP in control reactions. The ability to induce activity similar to the native effector is consistent with functional binding of the analogs to the S-site. Nonetheless, it is possible that the S-site may not be fully saturated

even though this concentration is significantly greater than the dissociation constant for dTTP (see below, Figure 2-10B). However, for most of the analogs the total rate of product accumulation is within ca. 2-fold of the control reaction containing 50 µM dTTP.

Amounts of products formed are sufficient to accurately identify the optimal substrate and observe reproducible effects on specificity, except when dCTP or 5-methyl- dCTP is used as an effector. Consistent with previous biochemical studies of RR, hRR discriminates exquisitely against the N4 group of dCTP70-71. This observation is also consistent with a crystal structure of hRR in complex with dTTP and GDP, which features the O4 of dTTP proximal to the backbone amide groups of D287, Q288 and

G289 (PDB ID: 3HND) (Figure 2-8)19. Removal of the 5-methyl group from dTTP reduces the discrimination between GDP and CDP by ca. 2-fold (compare dTTP and dUTP). This functional group packs against a side chain methylene group in the S-site of the hRR structure model, and replacement with a fluorine atom increases the rk for GDP, consistent with a nonpolar contact (compare dUTP and 5FdUTP) (Figure 2-7).

Substitution of the O2 with a sulfur atom also increases specificity for GDP, indicating that the larger size and reduced electronegativity can be accommodated by the S-site at this position.

Importantly, removal of the O4 atom and the concomitant deprotonation of the N3 atom in dZeb are sufficient to eliminate discrimination between GDP and CDP (compare dUTP and dZeb). This result is consistent with interactions involving these functional groups contributing to specificity for GDP over CDP; however, the overall shape of the residual pyrimidine base still enforces some degree of native specificity. The structural model of the hRR-dTTP-GDP complex predicts interactions between the O4 and N3 of

dTTP and the backbone of loop 2 and nearby N270, while the O2 group does not appear

to be involved in a contact (PDB ID: 3HND) (Figure 2-8A). Thus, the results reveal a pattern of sensitivity to chemical modification consistent with interactions involving primarily the N3 and O4 with hydrophobic packing interactions with the 5-methyl group contributing incrementally to discrimination between CDP and GDP.

The N1 of the purine nucleobase is a primary determinant that differentiates allosteric regulation by dGTP versus ATP: Structures of eukaryotic RR in which dGTP/ADP and

AMPPNP/CDP are the effector/substrate ligands bound to the S-site and C-site suggest differential interactions with the purine nucleobase at the S-site (Figure 2-8B-C, PBD ID:

2CVU, 2CVX). We identified the chemical features of the purine effectors ATP and dGTP that drive specificity for CDP and ADP, respectively by testing a series of purine effector analogs (2-aminopurine-drTP, N2dATP, 7-deaza-dATP, 7-deaza-dGTP, dITP) for their ability to allosterically regulate hRR.

Since both the S-site and A-site bind dATP, we first considered whether confounding effects due to A-site binding could lead to inaccuracies in observed specificity. Importantly, we observed that all effectors tested stimulated RR activity to some degree at concentrations that are fully inhibitory for dATP33. This observation

suggests that the analogs do not inhibit hRR to the same extent as dATP. Moreover, the

presence or absence of A-site ligands is not known to have significant effects on substrate

specificity (compare Figure 2-6 with Figures 2-7 and 2-9). Also noteworthy is that none of the dATP analogs abrogated substrate processing. It is therefore possible that they were unable to bind in the A-site. Such binding would likely lead to inhibition because the analogs all lack a 2′ hydroxyl group and dATP A-site binding inhibits eukaryotic

Class I RR32. Thus, the observed specificity in the presence of nucleotide analogs is most

likely due to S-site binding. This assumption is tested directly for dITP and 2-

aminopurine-drTP, below (Figure 2-10). Nonetheless, modulation of overall activity by

A-site binding cannot be entirely excluded even though there is no current evidence that nucleotide binding to the A-site influences substrate specificity at the C-site.

As shown in Figure 2-9, the specificities observed with purine analogs strongly suggest that the protonation state of the N1 atom on the purine ring and the presence of a

guanosine O6 group determine whether hRR recognizes it as ATP or dGTP. Most purine nucleobase modifications have relatively small effects on the total activity

(v(obs)T(analog)/v(obs)T(dGTP)) at the concentration of effector tested (100 µM). The purine N7

is dispensable for specificity determination since 7-deaza-dATP and 7-deaza-dGTP have specificity comparable to ATP and dGTP, respectively. Furthermore, N2dATP and 2- aminopurine-drTP both direct C-site specificity for CDP, while dITP directs specificity for ADP. This result argues strongly against any model in which either the guanosine N2 or adenosine N6 are the main determinants of specificity. These positions are proximal to the loop 2 peptide backbone, but they do not appear to influence C-site specificity

(Figure 2-8B-C, PDB IDs 2CVU, 2CVX). The experimental results strongly implicate the N1 atom of the purine nucleobase as making a large binary contribution towards determining whether the enzyme has an optimal kcat/Km for ADP or CDP. The identity of

the group at the 6 position is also likely to contribute to the specificity directed by dGTP;

however, removal of the O6 group necessarily results in deprotonation of the N1 group,

making these variables difficult to deconvolute.

Evidence that dITP and 2-aminopurine-drTP direct specificity by binding at the

Specificity site: When dNTP analogs are added to in vitro hRR specificity assays as the only effector, they stimulate NDP reduction and cause shifts in specificity, consistent with S-site binding. However, mammalian RR has three known nucleotide binding sites, and P. aeruginosa RR is known to bind two ATP molecules in its N-terminal ATP cone domain32-33, 64. Given the complexity in RR regulation and nucleotide binding, we tested

whether nucleotide analog triphosphates and native dNTP effectors compete for S-site

binding. The possibility of alternative binding at the A-site is strongest for the purine

analogs because they are chemically most similar to dATP, a ligand that is known to bind

to both the S-site and A-site of hRR. However, the native effector dTTP is only known to

bind at the S-site, and the specificity it directs is different from specificity directed by any

purine effector or effector analog19, 72. Therefore, competition between dTTP binding and

analog binding at the S-site should necessarily result in both a concentration-dependent

change in the proportions of the products formed as well as weaker apparent affinity for

dTTP.

As described above, data gathered in the presence of dITP and 2-aminopurine- drTP as effectors support the interpretation that the N1 atom directs specificity for purines. Accordingly, we tested whether dTTP competes with dITP or 2-aminopurine- drTP for binding at the S-site (Figure 2-10). When a mixture of hRR and 50 μM dITP or

2-aminopurine-drTP is challenged with dTTP, the specificity directed by the analog is suppressed in a concentration-dependent manner (Figure 2-10A). The specificity of hRR for GDP over CDP can be used to estimate the degree to which the S-site is occupied by dTTP. The data fit to hyperbolic binding isotherms (Figure 2-10B-C) that in the presence

of dITP or 2-aminopurine-drTP yield apparent dissociation constants for dTTP that are

approximately two orders of magnitude higher than the observed value of 0.46 μM.

Assuming a simple competitive binding model, the dissociation constants for dITP and 2-

aminopurine-drTP are in the range of 100-200 nM. These values are comparable to the

dissociation constant for dGTP binding to murine RR (700 nM)72. While binding of some

of the tested analogs at additional sites on hRR cannot be entirely excluded, the results

are most consistent with a model in which dITP and 2-aminopurine-drTP exert their effects on specificity through S-site binding.

The side chain of D287 in loop 2 of hRR is essential for specificity regulation by dGTP and dTTP, but not ATP: Although a unifying model of RR allostery has yet to be fully realized, the available ScRR structures with effectors bound in the S-site provide a context for identifying potential interactions that may interpret the signals directed by the

N1 and O6 groups of purine effectors (Figure 2-8). In the structure of ScRR containing

AMPPNP in the S-site, the N1 atom of the adenine nucleobase accepts a Hydrogen-bond from the backbone amide group of D287 in loop 2 while the side chain is pointed away from the S-site (PDB ID: 2CVU)73-74. In contrast, the protonated N1 group of dGTP acts

as a Hydrogen-bond donor and contacts the carboxylic acid moiety of D287 in the

dGTP/ADP structure of ScRR (PDB ID: 2CVX). In a crystal structure of hRR with dTTP

bound in the S-site, the nucleobase does not contact D287. Instead, D287 forms a

Hydrogen-bond with N270, which in turn contacts the N3 group of dTTP (Figure 2-8,

PDB ID: 3HND)73, 75-76. These different interactions with D287 appear to favor

alternative conformations of loop 2. Consistent with this notion, previous studies of RR

function in vivo showed that a D287A mutation in ScRR results in elevated cellular dCTP

and dTTP pools77. Together, the structure modeling and available mutagenesis data

suggest that the D287 sidechain may play a role in recognition of dGTP and dTTP, but is

not apparently involved in contacting ATP (Figure 2-8).

To address the potential role of D287 in hRR function, we constructed the D287A mutant of hRR and assayed its specificity in the presence of the native effectors ATP, dGTP and dTTP (Figure 2-11). This mutant shows essentially identical specificity when

ATP is used as the effector. In contrast, this mutant fails to efficiently discriminate among any specificity drivers intrinsic to either dGTP or dTTP, and favors CDP as the optimal substrate for all three effectors. When dGTP is used as the effector, the D287A mutant has the highest kcat/Km for CDP while wild-type hRR prefers ADP. When dTTP is

used as the effector, only GDP and CDP are processed detectably, which is somewhat

similar to the native enzyme. However, the kcat/Km for CDP is greater than GDP by 10- fold. Thus, the primary effect of deleting the D287 side chain is to silence allosteric information provided by the dGTP and dTTP nucleobase functional groups and shift hRR substrate selection toward ATP-driven specificity.

In sum, the results provide a chemically detailed picture of the effector functional groups responsible for directing the allosteric regulation of hRR substrate specificity. The data provide experimental evidence that the N1 of the purine nucleobase is a primary chemical signal that differentiates specificity directed by the effector dGTP versus

ATP/dATP. The guanosine O6 is also likely to contribute to specificity, as suggested by

structural models of both eukaryotic RR and EcRR. However, effects of individual

functional group modifications at this position also affect the conjugation of N1,

complicating interpretation of the effects on specificity. Although the adenosine N6

figures prominently in structure models, this functional group does not appear play a

major role in allosteric communication. This interpretation is based primarily on the

observation that 2-aminopurine-drTP directs CDP specificity like ATP despite that fact

that it has an exocyclic N2 like dGTP and lacks an N6 like that characteristic of

adenosine.

The data suggest that the exocyclic amine of dCTP acts as an anti-determinant for overall effector function. This interpretation is based on the ability of dZeb to act as an effector. The dZeb nucleobase is essentially identical to cytosine but lacks an exocyclic amine at position 4 on the pyrimidine ring. Nonetheless, at equivalent concentrations

(100 μM) dZeb is able to direct hRR catalytic activity while dCTP fails to do so.

Importantly, dZeb has greatly reduced specificity for GDP over CDP when compared to

the specificity directed by the native dTTP effector. This effect may be due to the

deprotonated N3 of the zebularine nucleobase, which could function similarly to the

deprotonated N1 of adenosine and form interactions that drive specificity for CDP.

Surprisingly, dZeb drives GDP reduction in spite of its lack of a 5-methyl group, lack of

an O4 group and deprotonated N3 relative to the native dTTP effector. Thus, none of

these characteristic functional groups is absolutely required to favor the loop 2

conformation that is specific for GDP. Interestingly, none of the pyrimidine

modifications caused hRR to adopt a new preferred substrate other than GDP. Together,

the results reveal an S-site that is highly sensitive to the presence of an N4 amine, but

otherwise remarkably accommodating of variation in the size and electronic properties of

the pyrimidine nucleobase.

The observation that D287A mutant hRR primarily reduces CDP in the presence

of any of the three effectors is consistent with the presence of functional interactions

between dGTP/dTTP and the side chain of D287. In current structural models of

eukaryotic RR, the adenosine N1 contacts the backbone amide of D287, while the

protonated N1 of dGTP can Hydrogen-bond to the carboxylic acid side chain (PDB IDs:

2CVU, 2CVX). Loss of this interaction could weaken the ability of dGTP to induce the loop 2 conformation that is optimal for ADP binding. While the D287 side chain does not directly contact dTTP, it is part of a network of Hydrogen-bonding interactions including

N270 (PDB ID: 3HND). Disruption of this network could diminish the ability of dTTP to

induce loop 2 conformations that favor interactions with GDP in the C-site.

As illustrated in Figure 2-2A, the conservation of loop 2 sequence is very strong and mutations at these positions can generate aberrant substrate recognition phenotypes in ScRR74, 77-78, consistent with its central role in allosteric communication. The major

features of substrate recognition are conserved among RRs from various species and the

roles of conserved amino acids in loop 2 are likely to be analogous (Figure 2-3). It is less

clear how phylogenetic variation in loop 2 affects structure-function relationships

involved in allostery. Comparisons of structural models and inferences from functional

experiments are consistent with a role for R293 in interacting with the phosphate groups

and nucleobase of the substrate74. However, structures of ScRR do not show direct

contacts between R293 and the phosphate groups of the substrate, raising a potential issue

with interpretation of these structural models22. Q288 occupies a similar position in the

EcRR and ScRR structural models of the dATP/AMPPNP bound states (Figure 2-2),

consistent with a conserved role in stabilizing loop 2 structure. The corresponding

positions of Q288 differ more significantly between the bacterial and eukaryotic RR

models with dGTP or dTTP bound in the S-site (Figure 2-4). Phylogenetic comparative

sequence analyses show that the identity of D287 is not conserved among all species,

though it is strongly conserved among eukaryotes. While the D287 sidechain of ScRR

contacts the N1 of dGTP, a similar interaction is not observed for the corresponding

serine residue in EcRR (PDB IDs: 2CVX, 5CNU). The N1 is instead contacted by the

conserved Q294 (EcRR numbering), and it is possible that this contact is key for EcRR

effector recognition. Additional structure-function studies of RR loop 2 interactions and conformations are required to test proposed functional interactions. Importantly, although structural models are vital tools for model building and hypothesis generation,

the present biochemical data represent a complementary yet independent line of inquiry

which is foundational for any model of hRR allostery.

The ability to quantify specificity for all four NDP substrates in the presence of a

range of nucleotide analogs reveals several important general features of RR substrate

discrimination. A well-known characteristic of RRs that is further documented here is the

ability of the enzyme to accept multiple alternative substrates even in the presence of a

single dNTP effector. Interestingly, CDP has the highest kcat/KM or is within ca. 10-fold of the optimal NDP substrate for both hRR and ScRR regardless of which effector is bound. Additionally, the D287A hRR mutant is defective in allosteric communication induced by dGTP and dTTP, and reduces primarily CDP regardless of effector nucleobase identity. These results together suggest that the CDP-reducing conformation of loop 2 may be a default state with respect to specificity. It has previously been suggested that loop 2 conformations exist in dynamic equilibrium and key interactions

with the effector nucleobase serve to perturb that equilibrium. In this model, binding of

dGTP or dTTP acts to shift the conformation or conformational ensemble away from this

default state27.

Indeed, interactions between protein and ligand are typically accompanied by

redistribution of thermally accessible conformations79-80. Mutations can cause multiple direct and indirect changes in allosteric communication81-82, and pinning down the

individual roles of particular interactions is difficult. Importantly, Cooperman and

colleagues have developed foundational equilibrium thermodynamic schemes that

describe levels of activity based on the population of effector- and substrate-bound states33-34, 83. Current challenges now include the need to incorporate specific structural

and functional detail, and to account for the contributions from both optimal and non-

optimal substrates for a given effector. Moreover, a complete comprehension of RR

allostery necessarily requires understanding the linkages among effector and substrate

binding thermodynamics, the dynamic behavior of loop 2 conformations, and other

elements of RR protein structure.

Figure 2-5. Application of internal competition to measurement of hRR specificity.

A. Simplified kinetic scheme showing reaction of substrate (S) with enzyme-effector complex (E•F) to produce product (P). Substrate Sn reacts with second-order rate constant

(kcat/KM)n to produce Pn. B. Representative chromatogram of t=0 aliquot from an assay of

hRR specificity containing all four NDP substrates. This aliquot experienced all aspects

of experimental workup except boronate chromatography, so substrate NDPs are present.

C. Representative chromatogram of t=30min aliquot with substrates removed by boronate

chromatography, showing only dNDP products. Insert: detail of the region of the

chromatogram containing dCDP, dUDP, and dADP. D. Plot of observed reaction velocity

(vobs) vs. [S1]/[S2]. Each symbol represents one independent trial. E. Data from D were

r used to calculate the relative kcat/Km ( k) using Equation 2-1 with ADP as the

experimental substrate and CDP as the reference. The calculated value is plotted as a

function of [S1]/[S2] that was measured directly for each reaction by integration of the

NDP peaks from the t=0 chromatogram as shown in A.

Figure 2-6. Measurement of native hRR and ScRR specificities.

Data are shown as dimensionless rk values. rk values may be considered equal to the

proportion of each product formed relative to the reference product when the two

substrates are present in equimolar concentration. A. Specificity of hRR (orange bars) and ScRR (blue bars) in the presence of 1 mM ATP. B. As in A, but in the presence of 1

mM ATP and 0.75 mM dGTP. C. As in A, but in the presence of 1 mM ATP and 1.6 mM

dTTP. ‡ indicates a product that was not present in sufficient quantity to accurately

measure its formation. † indicates a product that coelutes with a hydrolysis product from an effector. Specific activity for ScRR across all substrates in the presence of 1 mM ATP was 0.11+0.017 mol/(s*mol ScRR1)19. See Figures 2-7 and 2-9 for hRR specific activity values19. Similar values were obtained when ATP was present in the presence or absence of dGTP or dTTP.

Figure 2-7. The specificities directed by a series of pyrimidine effector analogs.

“R” denotes the deoxyribose triphosphate moiety. Functional groups on the nucleobase of dTTP are numbered. Altered functional groups are highlighted. Red denotes a functional group that has been replaced by a hydrogen atom. Blue denotes a hydrogen atom that has been replaced by a functional group. Specificity is defined as in Figure 2-6.

‡ indicates a product that was not present in sufficient quantity to accurately measure its formation. † indicates a product that coelutes with a hydrolysis product from an effector.

ADP was not included (*). Velocity ratios represent the total observed reaction velocity for the indicated analog relative to the total velocity for dTTP. Specific activity across all substrates in the presence of 50 μM dTTP was 0.0066+0.00052 mol/(s*mol hRRM1)19.

Figure 2-8. Structures of eukaryotic RR bound to S-site ligands.

A. Crystal structure of hRR bound to dTTP and GDP (not shown) (PDB ID: 3HND, 3.21

Å resolution)19. B. Crystal structure of ScRR bound to dGTP and ADP (not shown) (PDB

ID: 2CVX, 2.2 Å resolution). C. Crystal structure of ScRR bound to AMPPNP and CDP

(not shown) (PDB ID: 2CVU, 2.9 Å resolution)22. S-site ligands are shown as purple sticks. Atoms which perturb specificity when modified are shown as small spheres. Loop

2 amino acids are shown as white sticks. Q288, G289, and R293 are shown as red sticks.

Potential contacts involving D287 are shown as yellow dashes.

Figure 2-9. The specificities directed by a series of purine effector analogs.

“R” denotes the deoxyribose triphosphate moiety, except that ATP has a ribose

triphosphate moiety. Functional groups on the nucleobase of ATP are numbered. Altered

functional groups are highlighted. Red denotes a functional group that has been replaced by a hydrogen atom. Blue denotes a functional group that has been replaced. Green denotes a hydrogen atom that has been replaced with a functional group. Specificity is defined as in Figure 2-6. ‡ indicates a product that was not present in sufficient quantity

to accurately measure its formation. † indicates a product that coelutes with a hydrolysis product from an effector. Velocity ratios represent the total observed reaction velocity for the indicated analog relative to the total velocity for dGTP. Specific activity across all substrates in the presence of 50 μM dGTP was 0.0055+0.0017 mol/(s*mol hRRM1)19.

Figure 2-10. Competition between purine effector analogs and the natural S-site effector dTTP. Competition between purine effector analogs and the natural S-site effector dTTP. 50 μM dITP or 2-aminopurine-drTP was challenged with dTTP in the presence of all four NDP substrates. Substrates were present at 600 μM. A. Representative chromatograms from a

competition assay. hRR in the presence of 50 μM dITP was challenged with 10 μM dTTP

(black trace), 50 μM dTTP (red trace), 100 μM dTTP (blue trace), or 1000 μM dTTP

(green trace). The specificity directed by dITP (dADP favored) is suppressed by

specificity directed by dTTP (dGDP favored). B. Measurement of the dissociation constant between hRRM1 and dTTP. GDP reductase activity in the presence of 1.2 mM

GDP was measured in the presence of 0 μM dTTP, 0.1 μM dTTP, 0.5 μM dTTP, 1 μM dTTP, 2 μM dTTP, 5 μM dTTP, and 10 μM dTTP. C. Plot of [dTTP] versus the relative

second-order rate constant for GDP over CDP in the presence of 50 μM dITP (black

circles, dashed line) or 50 μM 2-aminopurine-drTP (black squares, solid line). In B and C, data can be fit to a hyperbolic binding isotherm to derive an apparent dissociation constant for dTTP.

Figure 2-11. Effects of the D287A substitution on hRR substrate recognition.

A. Specificity of D287A hRR under the indicated conditions. Specificity is defined as in

Figure 2-6. ‡ indicates a product that was not present in sufficient quantity to accurately

measure its formation. † indicates a product that coelutes with a hydrolysis product from an effector. Specific activity for D287A hRR in the presence of 1 mM ATP was 0.053 +

0.002 mol/(s*mol hRRM1); 50 μM dGTP, 0.0028 + 0.00048 mol/(s*mol hRRM1); 50

μM dTTP, 0.0072 + 0.0025 mol/(s*mol hRRM1). B. For each substrate, the difference

between its rk in the presence and absence of the D287A mutation is shown. Negative values indicate that the D287A mutant prefers the substrate less than wild-type; positive

values indicate that the opposite is true. For example, if a substrate is not appreciably

processed by wild-type hRR but is the favored substrate for D287A hRR, the difference has a value of 1.

Conclusions

Ribonucleotide reductases make use of a unique form of allostery to maintain proper dNTP balance in cells. Balance in dNTP pools is essential for DNA replication and repair, making the action of RRs vital to the existence of life as we know it and a key target for diseases of cell proliferation such as cancers. Binding of NTP/dNTP effectors at the specificity site changes the conformation of loop 2 such that RR presents an active site pocket that is most favorable for binding and reduction of one of the four NDPs.

Through X-ray crystallography we know that the effectors are extensively recognized by amino acid residues in the specificity site. However, the static nature of the picture this technique reveals does not differentiate between functional and adventitious contacts, i.e., contacts that drive specificity and those that are present but do not contribute to conformational changes in loop 2. In this study, we systematically modified the functional groups of the dNTP effectors and observed the effects on human RR specificity. Alternative substrate kinetics provided a complete picture of specificity that revealed the functional groups on the effector that are read out into conformational changes in loop 2. This technique revealed that it is the protonation state of the N1 group that allows human RR to differentiate between the purine effectors ATP and dGTP.

Conversely, we found that the 3, 4, and 5 groups of the pyrimidine effector dTTP are all functionally recognized by human RR, while the 2 group is likely dispensable.

Examining available crystal structures of human and yeast RR, we noticed that amino acid residue D287 in loop 2 is involved in contacts with all three effectors. Consistent with structural models and the present biochemical information, mutation of residue

D287 to alanine resulted in an RR variant that is incapable of differentially recognizing

the three effectors. Thus, the present results suggest that although RR’s specificity site extensively recognizes the dNTP effectors, only a limited subset of protein-ligand contacts functionally direct allostery to modulate specificity at the active site.

Materials and methods

Purification of human and yeast ribonucleotide reductases: The two human RR

subunits hRRM1 and hRRM2 were purified according to procedures described in

Fairman et al19. Briefly, hRRM1 was recombinantly expressed in BL-21 (DE3) RIL E. coli cells, while hRRM2 was expressed in BL-21 DE3 E. coli cells. hRRM1 was purified via peptide affinity chromatography. Wild-type hRR specific activity across all substrates in the presence of 50 μM dTTP was 0.0066+0.00052 mol/(s*mol hRRM1)19. Specific

activity across all substrates in the presence of 50 μM dGTP was 0.0055+0.0017

mol/(s*mol hRRM1)19. hRRM2 was purified by addition of an N-terminal 6x His-tag and

subsequent Ni-NTA affinity chromatography. The hRRM2 subunit was assembled into

the active cofactor via iron loading under anoxic conditions using ferrous ammonium

sulfate. The two RR subunits from S. cerevisiae, ScRR1 and ScR2R4, were purified by a

method similar to hRRM1 and hRRM2 except that ScRR1 was recombinantly expressed

in BL-21 (DE3) pLysS E. coli cells and purified to homogeneity using peptide affinity

chromatography. Specific activity for ScRR across all substrates in the presence of 1 mM

ATP was 0.11+0.017 mol/(s*mol ScRR1)19. Similar values were obtained when ATP was

present in the presence or absence of dGTP or dTTP. After one freeze-thaw cycle, active

ScR2R4 still contained ca. 0.15Y•/ββ´ (0.15 tyrosyl radical per small subunit heterodimer)

(data not shown).

In vitro measurement of ribonucleotide reductase multiple turnover kinetics: Kinetic

assays were conducted at 37 °C in 50 mM gly-gly pH 7.7, 15 mM MgCl2, 20 mM DTT.

Reactions used to validate internal competition kinetics as shown in Figure 2-5 contained

0.5 μM R1, 5 μM R2, 1 mM ATP, and 0.75 mM dGTP. Substrates ADP and CDP were

present in concentrations which range from 300 μM to 3 mM. Prior to the start of the

reaction, all components except R1 were mixed in 250 μL of reaction buffer and

incubated 2 mins at 37 °C. Two 30 μL aliquots were removed prior to initiating the reaction. These samples were used to measure the substrate concentrations and verify the efficacy of subsequent boronate chromatography. The reactions were started by adding hRRM1, and aliquots were removed at specific times after mixing (2-30 min) and

r quenched (see below). Reaction times were chosen to permit k (relative kcat/KM)

measurements for slow substrates while maintaining product accumulation in the linear

phase of the reaction under steady-state conditions (< 10% reacted).

Upon removal, all aliquots were quenched by rapid freezing to -80 ºC. Substrate and product concentrations were quantified by UV absorbance, as below. Measuring substrate and product concentrations independently permits greater precision in rk

(relative kcat/KM) measurements and allows for confirmation that steady-state assumptions are satisfied. Accumulation of product concentration as a function of time yielded the observed velocity (vobs) for each substrate. Relative velocity is measured via product

accumulation rather than substrate depletion because of the low fraction of reaction in

experimental aliquots. Substrate concentration and vobs data were combined via Equation

52-55 r 2-1 (see below) to yield the k, or the ratio of each substrate’s kcat/Km relative to that

of the reference substrate. hRRM2 or ScR2R4 were added in ten-fold excess of hRRM1

or ScRR1.

[ ] [ ] = 푘푐푐푐 = Equation 2-1 , � � � 푣표표표1 퐾푚 1 푆1 푟 푆1 푣표표표 푟푟푟 푘푐푐푐 �푆푟푟푟� 1 �푆푟푟푟� �� 퐾푚�푟푟푟� 푘

Quantification of ribonucleotide reductase substrate specificity by internal competition:

Experiments measuring native hRR and ScRR specificity were performed essentially as

above. ATP experiments contained 1 mM ATP as the only allosteric effector. This

concentration was chosen for inquiries in both Chapters 2 and 3 because it has been

previously observed that 1 mM ATP is sufficient for full oligomerization of the hR1

subunit, and we observe the 5- to 10-fold expected activation under these conditions33.

dGTP experiments also contained 0.75 mM dGTP, and dTTP experiments also contained

1.6 mM dTTP. These concentrations ensure that ATP is excluded from the S-site33, 72.

Substrates ADP, CDP, GDP, and UDP were present in concentrations ranging from 50

μM to 1.5 mM. Typically, substrates were present at 0.5 mM. However, substrates with

high rk values were given concentrations as low as 50 μM and substrates with low rk

values were given concentrations as high as 1.5 mM to facilitate vobs measurement.

Quantitative analysis of the specificity induced by deoxynucleotide analogs was

performed essentially as above. All deoxynucleotide analogs were purchased from

TriLink in stock concentrations of 100 mM in H2O. 1 mM ATP was omitted to prevent

competition with effectors of interest. Substrates were present in equimolar

concentrations (0.6 mM). In experiments with pyrimidine analogs, ADP was omitted to

prevent adventitious binding of ATP present as a minor contaminant. Nucleoside

triphosphate effector analogs were present in the concentrations indicated in the text. All

nucleotide effector analogs tested were present at 100 μM. This concentration is ~100-

fold greater than the dissociation constants for the natural effectors dGTP and dTTP, and

stimulated enzymatic activity for all analogs except dCTP derivatives (Figure 2-7)72.

Measurement of the apparent dissociation constant between hRRM1 and dTTP in the

presence of 50 μM 2-aminopurine-drTP or dITP was performed by varying the

concentration of dTTP and measuring the rk value for GDP with all four NDP substrates

present at 600 μM (Figure 2-10C). Measurement of the dissociation constant between hRRM1 and dTTP was performed by measuring the specific GDP reductase activity in the presence of 1.2 mM GDP, no other NDP substrates, and varying concentrations of dTTP (Figure 2-10B).

Reaction mixtures of RR substrates and products were processed according to procedures described in Hendricks et al58. Briefly, frozen aliquots were thawed by

addition of 169.4 μL boronate chromatography buffer (150 mM ammonium acetate, 15

mM MgCl2, pH 9) and 1M MgCl2 to a final MgCl2 concentration of 18 mM. For

experiments described in Figure 2-10B, 1 μL dATP was added as an internal standard.

Diluted aliquots were immediately processed by application onto a tuberculin syringe

packed with >250 μL Affi-Gel 601 (BIO-RAD). Aliquots were pushed through the

syringe at a flow rate of ~1 mL/min. The resin was then washed with 400 μL of boronate

chromatography buffer at the same flow rate. The resulting mixture was acidified with ~5

μL of 85% phosphoric acid to pH 3. 100 μL was immediately injected onto a

Phenomenex SphereClone anion exchange column with 150 mM sodium phosphate pH

3.7 as the mobile phase. Analytes were eluted via a gradient of 150-800 mM sodium

phosphate pH 3.7 on a Shimadzu LC 20-AB chromatograph. The chromatograph’s dual-

wavelength detector was set to 259 nm and 271 nm. Peaks were identified by comparing

their retention time and A259/A271 with those of standards. ADP, GDP and UDP were

quantified by their A259; CDP was quantified by its A271. A final aliquot was treated

identically but not subjected to boronate chromatography so that the concentration of

each substrate in the reaction could be measured independently of expected concentration.

Site-directed mutagenesis of human ribonucleotide reductase: Site-directed mutagenesis of hRRM1 was conducted using the Thermo Phusion Site-Directed

Mutagenesis Kit. The D287A mutant was generated using primers 5′-

ACAGCTAGATATGTGGCTCAAGGTGGGAACAAG-3′ and 5′-

GTTGTTATATACTCTCAGCATCGGTACAAGGC-3′. Plasmid DNA was purified using the QIAgen Miniprep and Midiprep plasmid purification kits. Plasmid sequences were verified by Sanger sequencing from primers 5′-TTCGGCTTTAAGACGCTAGA-3′,

5′-CTTGGCATTTAGACATCTTTGA-3′, 5′-TTGGCTGAAGTCACTAAAGTCG-3′,

5′-CGCAGAGTCTTGTCAGGAGA-3′, and the T7 promoter. D287A hR1 was purified via the same method as wild-type hR1. Specific activity for D287A hRR in the presence of 1 mM ATP was 0.053 + 0.002 mol/(s*mol hRRM1); 50 μM dGTP, 0.0028 + 0.00048

mol/(s*mol hRRM1); 50 μM dTTP, 0.0072 + 0.0025 mol/(s*mol hRRM1).

Sequence alignment: Amino acid sequence alignment was carried out using the Multiple

Sequence Alignment feature of Clustal Omega84-85. cDNA sequences of ribonucleotide reductase large subunit were retrieved from GenBank86.

Chapter 3 : Phylogenetic Comparative Sequence Analysis and

Functional Studies of Mutant Enzymes Reveal Compensatory

Amino Acid Substitutions in Loop 2 of Human Ribonucleotide

Reductase4

4 In collaboration with Sneha Grandhi, Reena Sheth, Md. Faiz Ahmad, Rajesh Viswanathan, Chris Dealwis, and Michael Harris.

Abstract

Eukaryotic class I ribonucleotide reductases (RRs) generate deoxyribonucleotides for DNA synthesis. Binding of dNTP effectors is coupled to the formation of active dimers and induces conformational changes in a short loop (loop 2) to regulate RR’s specificity among its nucleoside diphosphate (NDP) substrates. In addition, ATP and dATP bind at an additional allosteric site 40 Å away from loop 2 and drive formation of activated or inactive hexamers, respectively. To better understand how dNTP binding influences the specificity, activity, and oligomerization of human RR, we examined natural sequence variation in loop 2 using an alignment of >300 eukaryotic RR sequences.

Most amino acids in eukaryotic loop 2 are nearly invariant in this sample; however, two positions covary by undergoing non-conservative substitutions (N291G and P294K; human numbering). Individual N291G and P294K mutations in human RR have additive effects on substrate specificity. The P294K substitution causes a significant defect in effector-induced oligomerization required for enzyme activity that is rescued in the

N291G+P294K enzyme. None of the other mutants show altered ATP-mediated hexamerization; however, certain combinations of loop 2 mutations and dNTP effectors perturb ATP’s role as an allosteric activator. Thus, the results demonstrate that the observed covariation of amino acids in eukaryotic loop 2 is essential for its role in dNTP- induced dimerization. In contrast, defects in substrate specificity are not rescued in the double mutant, implying that functional sequence variation elsewhere in the protein is necessary. The present results grant insight into loop 2’s roles in regulation of specificity, allostery, and oligomerization. However, our observations of human ribonucleotide reductase’s oligomeric states are only partially consistent with those from other groups.

Therefore, further experiments are required before a complete understanding of

nucleotide-induced oligomerization in human ribonucleotide reductase is attained.

Introduction

Ribonucleotide reductase (RR) catalyzes the reduction of ribonucleotides to

generate the 2′-deoxynucleotides required for DNA synthesis, and at least one type of RR

is found in all cellular organisms. RR is responsible for maintaining balanced pools of

dNTPs and its activity is regulated at the levels of transcription, localization, and

allostery26, 87. Eukaryotic Class I RR is composed of a large subunit which binds

nucleotides (R1) and a small subunit that carries a tyrosyl radical essential for catalysis

(R2). The R1 subunit contains the active site for NDP reduction (the catalytic, or C-site)

and two allosteric sites: the Activity site (A-site) and the Specificity site (S-site) (Figure

3-1A). Binding of an NTP/dNTP effector to the S-site results in formation of active R1 dimers and controls enzymatic specificity, i.e., the relative kcat/KM values for the four

NDP substrates83. ATP/dATP binding promotes reduction of primarily CDP and UDP,

dGTP binding favors ADP reduction, and dTTP binding favors GDP reduction33-34. The

A-site is approximately 40 Å from the active site and binds ATP or dATP. Nucleotide binding induces formation of activated or inactive hexamers, respectively, and the ATP- activated form is generally accepted as the physiologically relevant species19, 32, 36, 88-89.

Together, the actions of ligand binding at the A-site and S-site ensure that total dNTP

pools remain at appropriate levels 33. dNTP pools are so finely tuned that they vary significantly even among cellular compartments90.

RR is a key target for human disease treatment including cancer chemotherapy, 91.

The nucleotide analog gemcitabine diphosphate binds in the enzyme active site (C-site)

and acts as a suicide inhibitor41. Interactions between small molecules and hRR can be

measured through observing changes in oligomeric state. Recently, it was shown that the

triphosphate form of clofarabine, a deoxyadenosine analog, induces formation of inactive

R1 hexamers. Adenosine analogs cladribine and fludarabine can also act as inhibitors in

diphosphate forms by inducing the formation of inactive hexamers91. Although the

functional connection between effector S-site binding, dimerization, and RR activity

upregulation is well-understood, there are fewer studies examining the mechanistic link

between dATP binding, hexamerization, and inhibition 19, 22, 25, 27, 32. Moreover, current

models do not yet account for ATP’s activating effect on activity. Achieving a better

understanding of nucleotide binding and associated enzyme conformations responsible

for allosteric regulation of specificity and activity is therefore essential, both to our basic

understanding of RR function and to development of further therapeutics.

Current X-ray crystal structures of bacterial and eukaryotic RR together with in

vitro functional studies provide valuable and detailed insight into the potential

mechanisms by which nucleotide binding regulates eukaryotic Class I RR activity and

substrate specificity (Figure 3-1A)19, 22, 25, 40, 92-93. Binding of dNTP effectors results in formation of active R1 dimers in which the C-site and S-site of opposite monomers are closely apposed at the dimer interface. Structure models of the R1 subunit from S. cerevisiae show that the active site is only large enough to accommodate the substrate after dNTP effector binding22. Kinetic studies of murine RR also support the idea that

effector-dependent dimerization is necessary for activity, preventing substrate reduction

from taking place without allosteric information transfer from the S-site94. Structures of

RRs from Bacteria and eukaryotes further reveal that the C-site and S-site are joined by a short loop called loop 2 (Figure 3-1B). Structural models of RRs bound to different substrate/effector pairs show that binding of a dNTP effector in the S-site directs C-site specificity by changing the conformation of loop 2. The general features of this model are supported by the available biochemical data and contribute to the overall paradigm for enzyme regulation by allostery.

While some features of loop 2 appear to be conserved, others vary between species, but how such differences influence function is not known. Furthermore, crystal structures of RRs from various species show differences in loop 2 conformations in the presence of common ligands 19, 22, 25, 27, which can leave the precise roles of individual

amino acids unclear. Mutation of R293 or Q288 in S. cerevisiae RR to alanine results in

impaired substrate binding, reduced catalytic activity, and altered loop 2 conformations95.

Mutation of R293 or Q288 (human numbering) to alanine in E. coli RR results in

inactivation and decreased activation by dATP, respectively. Mutation of the less

conserved D287 to alanine in hRR leads to an inability to tune substrate specificity to

effector identity96. Mutating any of residues Y285, D287, Q288, or R293 in loop 2 results

in a mutator phenotype in S. cerevisiae77-78. Thus, conserved residues in loop 2 are clearly

important for function, but additional information is needed to further our understanding

of how conserved and variable residues in loop 2 work together to transmit allosteric

information between the S-site and C-site.

To further this goal, we systematically investigated patterns of phylogenetic sequence variation in eukaryotic RR enzymes and used this information to guide

structure-function studies of loop 2 function in vitro. A sequence alignment of 310 R1 subunits from Eukarya reveals two amino acids in loop 2 that consistently covary

(N291G and P294K, human numbering). Individual mutations at these positions in hRR

affect NDP discrimination, consistent with the canonical role of loop 2 in directing

substrate specificity. However, the results also reveal that these positions are essential for

dNTP-induced RR dimerization, and suggest a surprising new role for loop 2 in

mediating the long-range effects of ATP on activity.

Figure 3-1. Three-dimensional structure of hRR. A: Crystal structure model of the hRR large subunit homodimer (PDB ID: 3HND, 3.21 Å

resolution. ATP was modeled into the figure by aligning this structure to PDB ID:

3HNE.). The protein backbone is shown as gray ribbons. Loop 2 is shown as purple

ribbons. The S-site ligands (dTTP) are shown as blue spheres. The C-site ligands (GDP)

are shown as green spheres. The A-site ligand (ATP) is shown as red spheres. The distance between the 2′ carbon atoms of ATP and dTTP is 42.7 Å (calculated in PyMol).

B: Stereo diagram of the region including the C-site, S-site, and loop 2. The C-site ligand

(GDP) is shown as green sticks. The S-site ligand (dTTP) is shown as blue sticks. Loop 2 is shown as purple sticks. Residues N291 and P294 are shown as red sticks.

Results and discussion

Phylogenetic comparative sequence analysis reveals loop 2 positions that covary

in eukaryotic RR: To better understand the extent of sequence variation in eukaryotic RR

and the loop 2 region in particular, we conducted an alignment of 310 eukaryotic RR

sequences retrieved from GenBank (Figure 3-2, Figure 3-3; Supplemental data).

Putative or predicted proteins, as well as those under 500 amino acids, were excluded.

With respect to overall conservation of eukaryotic RR sequence, the hydrophobic core,

the active site (C-site), and the ligand-binding sites involved in allosteric regulation (S- site and A-site) show strong conservation as expected, while most surface-exposed residues are less conserved (Figure 3-2A-B). These patterns are consistent with the expectation of strong selection pressure to maintain the overall three-dimensional structure and functional ligand-binding properties of the protein.

Previous alignments of RRs have shown that loop 2 is well-conserved, consistent with its known role in allosteric regulation 25, 40, 64, 68, 97. However, the current analysis

reveals important differences among enzymes from different groups of eukaryotes. The

alignment shows that three major types of eukaryotic loop 2 sequences exist. We

designate them as Type I, Type II, and Type III (Figure 3-2C-D). Type I and Type II loop 2 sequences contain both the highly conserved Q288 and R293 residues and D287, which is conserved in eukaryotes but variable in RR enzymes from Bacteria96. In contrast,

Type III sequences do not show significant homology to Type I or Type II RR, and in

particular do not include the hallmark Q288 or R293 residues. The lack of significant

homology between Type III and I/II loop 2 sequences makes it difficult to draw

functional correlations, particularly because no structural information exists for these enzymes. Therefore, more detailed inquiry was directed at Type I and Type II sequences.

Importantly, Type I and II sequences are distinguished by two sites of surprising non-conservative amino acids substitutions in loop 2. hRR is representative of Type I sequences, which have an asparagine at position 291 and a proline at position 294

(human numbering). Type II sequences have a glycine at position 291 and a lysine at position 294 (human numbering). These changes are highly non-conservative, and the fact that the observed variation takes place in glycine and proline residues is particularly surprising. The biophysical properties of these two residue types are markedly different from those of the other 18 canonical amino acid residues, and can be considered to be

“punctuation marks” in protein structure98.

To facilitate comparison between loop 2 diversity and the eukaryotic tree of life, we divided the sampled organisms into animals, plants, fungi, and protists. We recognize that Protista is an obsolete phylogenetic group, and when mapping organisms onto a eukaryotic tree of life we make use of the most current information (Figure 3-3).

However, in Figure 3-2 we make use of the term Protista for convenience because of the relative undersampling of microbial eukaryotes. Most animals and fungi have Type I loop

2 sequences, while most plants and protists have Type II loop 2 sequences (Figure 3-2C-

D) 86. A notable exception to the covariation at positions 291 and 294 is C. albicans, whose RR has a glycine at position 291 but a proline at position 294 (the so-called “GP” loop 2). Furthermore, almost all flies in this data set have “GP” loop 2 sequences, suggesting that this variant arose independently in the common ancestor of flies (Figure

3-3). Conversely and importantly, loop 2 sequences that harbor an asparagine at position

291 and a lysine at position 294 (human numbering) are not observed. These data suggest

that the common ancestor of eukaryotes may have had a Type II loop 2 sequence. The

data also raise the possibility that mutation of K294 to proline occurred relatively early in

the common ancestor of Unikonts (the group that includes animals and fungi), which then

allowed for G291 to mutate to asparagine (human numbering).

The metabolic properties and genomic GC contents of the organisms included in

the present analysis do not appear to correlate with loop 2 sequence variation a way that

suggests an adaptation to an environmental pressure on evolutionary fitness. P.

falciparum, the malarial parasite, is a member of Apicomplexa and is known for having

an unusually low GC content in its DNA99. T. brucei, the cause of African sleeping sickness, is another parasitic protist whose RR reduces a comparatively large amount of

UDP in the presence of dATP47. It is possible that a Type II loop 2 allows for adaptations

in RR specificity regulation that are advantageous for the lifestyles of particular

organisms. However, many plants have GC contents that are closer to average, and would

thus likely require RR enzymes with specificity similar to those present in animals and

fungi100.

Available structural data offer clues into the potential roles of amino acid residues

N291 and P294. For example, N291 interacts with neither the substrate nor the effector in

crystallo, but appears to participate in an H-bonding network with the other amino acid residues in the loop. However, N291 also has the potential to participate in a crystal contact in the structural model of hRR, potentially confounding interpretation of its role in enzyme function (Figure 3-4). In S. cerevisiae RR, P294 forms part of the active site and appears to hold it in a conformational ensemble which is permissible to substrate

ingress22. Yet, it is not obvious from inspection of current structures how the observed

variation at these two positions would affect RR activity, nor how covariation would

result in a new and compensatory structure/function relationship40.

Because positions 291 and 294 are phylogenetically linked, determining the

properties of single and double mutants at these positions in the background of the

biomedically important and comparatively well-studied human enzyme is likely to

provide insight into these amino acids’ contributions to RR activity, substrate specificity

and allosteric regulation by oligomerization. Accordingly, we generated N291G, P294K,

and N291G+P294K mutants of hRR. Because of the highly nonconservative nature of the

substitutions involved in the present study, we considered the possibility that one or more

of the mutant human proteins is misfolded and/or thermally unstable. However, at 20 °C,

wild-type hRR has a circular dichroism (CD) spectrum consistent with significant α-

helical content and similar to the CD spectrum of murine RR (Figure 3-5)101. The three hRR loop 2 mutants have essentially identical CD spectra and Tm values within ca. 5 °C

of the wild type enzyme (55-60 °C). Thus, the functional effects of these mutations are likely to reveal only the results of local changes in loop 2 geometry, as opposed to gross effects on overall protein structure.

N291G and P294K point mutations in human RR loop 2 disrupt specificity but have complementary effects on activity: We determined the effects of loop 2 mutation on

catalytic activity and NDP substrate specificity using alternative substrate kinetics as

96 r described previously . Briefly, the relative kcat/KM value ( k) for each NDP substrate is

measured in reactions containing all four substrates by using boronate chromatography to

remove unreacted NDP substrates, and anion exchange HPLC to separate and quantify

the resulting dNDP products. Using this approach, we find that N291G hRR is ca. 5-fold

more active than wild-type hRR in the presence of dGTP or dTTP alone (Figure 3-6A,

C). In contrast, the P294K mutation reduces catalytic activity by ca. 10-fold when either

dGTP or dTTP is used as the effector. In the N291G+P294K double mutant, these effects

are offset and its activity is comparable to the wild-type enzyme. Strikingly, both the

N291G and P294K enzymes are inactive in the presence of 1 mM ATP, a concentration

of S-site ligand that is sufficient to drive robust activity in the native enzyme. Under the

same conditions the N291G+P294K double mutation rescues catalytic activity to ca. 10%

of wild-type levels, with essentially the same NDP substrate specificity as the wild-type

enzyme.

Comparison of the relative kcat/KM values for the four NDP substrates shows that

N291G hRR processes significantly more CDP than wild-type when dGTP is the effector.

The P294K mutation also results in a higher proportion of CDP reduction than wild-type

hRR under these conditions. However, compared to the effect of the P294K mutation on

RR activity, the effect on specificity is relatively modest. The effects of each single mutation on dGTP-directed activity and specificity are additive in the double mutant.

N291G also has decreased specificity for GDP when dTTP is used as the effector. P294K

hRR processes primarily GDP like the wild-type enzyme, although quantitative analysis

of specificity is limited due to effects of this mutation on activity. Similar to the results observed in the presence of dGTP, the effects of each mutation on dTTP-directed activity and specificity are additive in the N291G+P294K double mutant. Thus, N291G, and to a lesser extent P294K, result in effects on specificity that are additive in the

N291G+P294K double mutant. In addition, the P294K mutant causes a defect in overall

activity that is rescued in the N291G+P294K enzyme.

The P294K mutation causes defects in oligomerization that are rescued by the

N291G mutation: To further investigate the basis for the defect in catalytic activity

induced by P294K mutation in hRR, and the robust compensatory rescue in the double

mutant protein, we tested the effects of loop 2 mutations on effector-induced dimerization.

It is known that the minimal active form of RR is a dimer, so observed dimerization

defects could potentially explain variations in activity83. Alternatively, the P294K mutant

could be capable of dNTP binding and dimerization, but have perturbed local structure

resulting in loss of NDP binding or catalysis. In this case, the N291G mutation could

allow for compensatory interactions that correct local geometry at the active site. To

distinguish between these possibilities, we used size exclusion chromatography (SEC) to

determine the oligomeric state of hRR and the three loop 2 mutants in the presence of 10

μM dGTP or dTTP, and in the presence of 1 mM ATP (Figure 3-7). dGTP and dTTP were included at a concentration of 10 μM because this concentration is approximately one order of magnitude greater than the dissociation constants for these effectors at the S- site72. ATP was included at a concentration of 1 mM for the sake of consistency with

enzyme activity assay conditions, and because resolution suffered at higher

concentrations. The wild-type elution profile is essentially unchanged at 3 mM ATP

(Figure 3-8).

In the absence of effectors, all enzyme variants exhibit retention times consistent

with the monomeric form of the enzyme, as expected from previous studies 19, 33. In the presence of 10 μM dGTP or dTTP, all four enzymes dimerize. However, the extent of

P294K dimerization is reduced ca. 3-fold relative to wild-type. Importantly, the results

demonstrate that the N291G mutation fully rescues dimerization in the N291G+P294K

double mutant, restoring the extent of its dimerization to that of wild-type. This

observation is consistent with the kinetic results in the presence of dGTP or dTTP, in

which all enzyme forms show activity near wild-type levels except for P294K (Figure 3-

6A, C). The P294K mutant also has deficient oligomerization in the presence of 1 mM

ATP: the extent of hexamerization and the formation of intermediate molecular weight species are reduced ca. 2-fold. The N291G mutation restores P294K oligomerization to wild-type levels in the presence of ATP, but this result does not correlate with activity levels in the presence of 1 mM ATP (Figure 3-6E). The correlation between SEC and activity results in the presence of dGTP or dTTP suggests deficient dimerization as a plausible contributor to the P294K variant’s deficient activity under those conditions. In contrast, the lack of correlation between the two techniques in the presence of 1 mM ATP excludes oligomerization as a potential explanation for the two single mutants’ deficient activity. Instead, the data point to potential local perturbations in the crosstalk between the S-site and the active site. Thus, these results show that multiple systems within hRR are perturbed when these two positions in loop 2 are mutated.

N291G hRR recognizes the same effector nucleobase functional groups as wild- type hRR: Only a limited subset of effector nucleobase functional groups are recognized by the S-site in order to drive specificity at the C-site96. To test whether the altered

specificity of the N291G mutant arises due to new adventitious contacts with the dNTP

effector, we assayed the specificity of this protein in the presence of three effector analogs: 2-aminopurine deoxyribonucleotide triphosphate (2-aminopurine-drTP),

deoxyinosine triphosphate (dITP), and deoxyzebularine triphosphate (dZeb) (Figure 3-9).

The N1 of ATP drives specificity, but the N6 is dispensable in wild-type. In the presence

of 2-aminopurine-drTP, N291G hRR has specificity similar to wild-type hRR with ATP

or 2-aminopurine-drTP. The N2 of dGTP is also dispensable for allosteric regulation, and

wild-type hRR recognizes dITP as if it were dGTP. In the presence of dITP, N291G hRR

directs reduction of CDP followed closely by ADP, similar to its specificity in the

presence of GDP. In the presence of dZeb, wild-type hRR reduces GDP and CDP with

approximately equal efficiency. The N291G variant gives a similar result, but with a

greater preference for CDP than wild-type. N291G hRR also reduces more CDP than

wild-type in the presence of dTTP. Thus, the effects of introducing the N291G mutation

and introducing the chemical mutation of dTTP to dZeb are roughly additive. This

additivity suggests that specificity perturbations do not derive from effector recognition

via new adventitious contacts, but through impaired information transfer once canonical

contacts are established.

dGTP binding to N291G+P294K hRR and dTTP binding to P294K hRR convert

the allosteric activator ATP into a dATP-like negative allosteric regulator: To

investigate the effects of loop 2 mutations on allosteric activation by ATP, we measured

the activity and substrate specificity of the mutants in the presence of 1 mM ATP and

either 0.75 mM dGTP or 1.6 mM dTTP (Figure 3-10)96. Although hRR can bind ATP at

both the S-site and the A-site, under these conditions the higher affinity of the S-site for deoxynucleotides (KD values of ca. 1 µM versus ca. 150 µM) results in the dNTPs

dominating occupancy of the S-site33. This conclusion is also supported by the present observation that the presence of 1 mM ATP does not result in increased specificity for

CDP when compared to reactions containing either dGTP or dTTP alone (compare

Figures 3-6 and 3-10).

As expected, the overall activity of wild-type hRR is stimulated 5-10 fold by ATP

binding to the A-site. In contrast, the N291G mutant is inhibited ca. 5-fold. The P294K

mutant is stimulated by the presence of ATP when dGTP is the S-site effector, which is similar to wild-type hRR. Unexpectedly, although the N291G+P294K double mutant has robust catalytic activity in the presence of dGTP or ATP alone, the reaction is quenched when both are present. This effect is clearly not due to competition for ligand binding at the S-site, since the N291G+P294K double mutant has activity within ca. 10-fold of wild-type hRR in the presence of 1mM ATP or 50 µM dGTP. Thus, the presence of both

N291G and P294K mutations in hRR convert ATP, which stimulates enzyme activity in wild-type hRR, into an inhibitor. P294K mutant hRR is inactive in the presence of ATP and dTTP, thus revealing a second set of conditions in which ATP binding is interpreted as a negative allosteric signal. In contrast to the results in the presence of 1 mM ATP alone (Figure 3-6E), the present group of experiments is best explained by a mechanism in which long-range communication between the A-site and the active site is disrupted by mutations in loop 2.

Despite the wide range of lifestyles and genomic DNA GC content among eukaryotes, RR enzymes are well-conserved across this domain of life (Figure 3-2A, 3-

2B)100. Analysis of the alignment of eukaryotic RRs reported here (Figure 3-2, Figure 3-

3, Supplemental data) reveals that eukaryotic loop 2 sequences primarily occur as one

of two major types. Type I sequences are exemplified by hRR loop 2 and have an

asparagine at position 291 and a proline at position 294 (human numbering). Type II

loop 2 sequences are identical at almost other positions, yet harbor the highly nonconservative substitutions N291G and P294K. Most plants and protists have Type II loop

2s, while most animals and fungi have Type I loop 2s. The present bioinformatics and biochemical data together suggest that Type II was the ancestral form of RR loop 2 in eukaryotes (Figure 3-3). The glycine to asparagine and lysine to proline mutations likely occurred in the common ancestor of Unikonta before the group differentiated into its earliest members. A small number of loop 2 sequences of an intermediate genotype, exemplified by most members of the Candida genus, have a glycine at position 291, but retain a proline at position 294 (human numbering) (Figure 3-2C, Figure 3-3). The fact that C. albicans has an intermediate type of loop 2 sequence is of particular interest because of the clinical significance of the organism102-103. Almost all species of flies that

were sampled also have the single N291G mutation in the absence of a corresponding

P294K substitution (Figure 3-3). A small group of algae including N. gaditana also evolved this intermediate loop 2 independently from flies or Candida. Because the loop 2 variant with an asparagine at position 291 and a lysine at position 294 (P294K) is less

favorable to hRR enzyme function, thus it is likely that the Type I loop 2 arose from

mutation of the lysine residue followed by mutation of glycine residue, rather than the

converse (Figure 3-11A).

The bioinformatic, biochemical, and biophysical results reported here provide

insight into the functional roles of two key residues in hRR loop 2. N291 is not proposed

to contact either the substrate or the effector in current structure models of eukaryotic RR.

Yet, the N291G mutant of hRR is impaired in its ability to recognize purine substrates

(Figure 3-6). Similarly, P294 apparently plays a limited role in substrate recognition: its

only contact with a substrate is a second-sphere water-mediated contact with ADP in S.

cerevisiae22. Nonetheless, the conformation of P294 is altered by binding of different effector-substrate pairs. The substrate specificity observed in the presence of nucleotide

effector analogs rules out the possibility that N291G is recognizing different effector

functional groups to drive specificity. In particular, results in the presence of dZeb

exclude that possibility because the N291G substitution and the dZeb substitution

independently perturb specificity. If the protein and effector substitutions perturbed the

same set of contacts, it would be expected that introducing both substitutions would

produce similar specificity to either when present individually. The results reported here

are therefore consistent with previous proposals that N291 is involved in a network of contacts among loop 2 residues that helps transmit information from the S-site to the C- site19, 22, 25, 27.

Loop 2 is clearly the primary structural motif involved in differential substrate

recognition 19, 22, 25, 27. However, the results presented here make clear that its function

depends on the surrounding protein context. If loop 2 were perfectly modular with respect to allosteric communication between the S-site and C-site, the Type 2 sequence

(i.e. the N291G+P294K mutant) should necessarily serve as a perfect functional

equivalent of the Type I sequence (i.e., wild-type). Because the defects in the

N291G+P294K double mutant are additive, loop 2 function must also depend on other regions of the protein that work in concert to specify the correct substrate in the context of S-site ligand binding. Examination of phylogenetic data make clear that functional equivalency in loop 2 is achieved with a variety of sequences, and that Type I loop 2’s asparagine and proline are not obligate requirements. Complementary biochemical data

reveal the apparent necessity for additional mutations elsewhere in the protein that allow

for variation within loop 2. Importantly, whenever hRR structure is changed through site-

directed mutagenesis, or effector structure is changed through chemical mutagenesis, the

predominant effect on specificity is increased reduction of CDP96. It follows that the most

thermodynamically stable conformation of loop 2 favors reduction of CDP, and that a key

role of dGTP or dTTP binding is to contribute binding energy that RR harnesses to

perturb loop 2’s conformation away from this “default” state.

Both loop 2 mutants retain some catalytic activity and are capable of

oligomerization to some extent in the presence of S-site ligands. Although the effects of

N291G and P294K mutation on NDP specificity are additive, several key aspects of enzyme function disrupted by these point mutations are rescued in the double mutant

(Figure 3-11B). First, the catalytic activities of the single mutants in the presence of dGTP or dTTP offset one another in the double mutant, yielding activity similar to that of wild-type (Figure 3-6A, C). Second, P294K’s decreased ability to oligomerize in the

presence of allosteric ligands is rescued in the double mutant (Figure 3-7). Third, both single mutants are catalytically inactive in the presence of 1 mM ATP, while the double mutant shows activity rescued to ca. 10% of wild-type levels (Figure 3-6E). The fact that rescue is observed for some, but not all, effects of RR functionality suggests that loop 2 plays a variety of roles in RR. Those aspects that show rescued functionality are likely governed within loop 2; for aspects that fail to show rescue it is likely that other regions of the protein are involved.

An important negative result is the observation that for the most part, the loop 2 mutations described here do not appear to have large effects on ATP-induced

oligomerization. Several groups have independently measured the oligomeric states of

RRs from several species under a variety of conditions using a diverse array of

techniques including SEC as employed in the current study. Ando et al. employed small- angle X-ray scattering (SAXS) and electron microscopy (EM) to study the human enzyme32. They found that both dATP and ATP cause formation of hexamers in the

presence of substrates and that, notably, R2 can change the structure of dATP hexamers

but not ATP hexamers. ATP can even induce formation of R1 filaments at high

concentrations. Aye et al. and Wang et al. showed using SEC that hRR can form

hexamers, dimers, or a mix of oligomeric states depending on the precise reaction and

chromatographic conditions used43, 104. Fairman et al. employed SEC/MALS to observe

that hRR forms dimers and hexamers in the presence of dATP19. Kashlan et al. examined

murine RR via dynamic light scattering and sedimentation velocity experiments33. They

observe that murine RR may form tetramers or hexamers. Rofougaran et al. examined the

oligomeric state of human RR using SEC in a method that is similar to our own36. They

found that hRR forms a mixture of hexamers, dimers, and monomers in the presence of 3

mM ATP, which is only partially consistent with the results in the present study. There

are several potential explanations that can account for this apparent discrepancy. The two

sets of experiments make use of different methods of protein purification and different

divalent ion concentrations. They may also differ in other variables such as flow rate and

temperature. Part of the hexamer population in this study may be dissociating during the

course of the SEC run, leading to the formation of a peak with the apparent molecular

weight of a tetramer. However, the two studies agree on the point that ATP binding

triggers formation of an ensemble of oligomers in hRR. Importantly, all comparisons

made in this study use wild-type protein under identical conditions as a reference, and

interpretation is restricted to examination of differences in the behavior of wild-type and

mutant proteins. Although potential dissociation during the chromatographic run

precludes quantification, the fact that no significant observable differences are induced in the N291G or N291G+P294K mutants is consistent with minimal perturbation of ATP-

induced oligomerization under steady state reaction conditions. In contrast, the P294K

mutation slightly reduces the accumulation of higher-order oligomeric states induced by

ATP. This attenuation is rescued when the ATP concentration is increased to 3 mM

(Figure 3-8), suggesting that the dissociation constant for ATP at the S-site or A-site is slightly increased.

In sum, this diverse set of bioinformatic and in vitro experiments leads to the following conclusions. Loop 2 is well-conserved, yet variation is tolerated at positions

291 and 294. This variation tends to take the form of two major types of loop 2: Type I and Type II. Introducing mutations into hRR loop 2 to convert it from Type I to Type II does not disrupt hRR’s overall secondary structure. In the presence of dGTP or dTTP alone, all enzyme variants are active, yet the P294K mutant has deficient activity. This is likely due to a partial failure to dimerize under similar conditions. In the presence of ATP alone, neither single mutant is active, and this may be due to aberrant conformational ensembles in loop 2 that are partially restored in the double mutant. The presence of dGTP or dTTP in the S-site and ATP in the A-site has complex effects on the mutant enzymes, which may be due to disruption in long-range communication between the two allosteric sites. In all activity assays, the mutants showed small detrimental effects on specificity, suggesting that loop 2 works together with other regions of the enzyme to

dictate substrate specificity. Throughout the biochemical experiments, P294K was more

detrimental than N291G or N291G+P294K mutations, which correlates well with the

observed bioinformatics data. Loop 2 is clearly involved in a wide range of processes in

hRR, and mutations in loop 2 can have effects on both short-range and long-range allostery within the protein.

The results of this study implicate two amino acids in loop 2 as having important roles in diverse areas of enzyme function including overall activity, substrate specificity, allosteric regulation, and oligomerization. Some of these aspects were anticipated from prior studies, while others could not have been predicted from examination of previous biochemical or structural data. Because the entirety of loop 2 is well-conserved, it is likely that each individual residue plays a unique and key role in some aspect of enzyme function that is vital to biological fitness. The roles of Q288 and R293 have been well- examined, and studies from our group have shed some light on residues D287, N291, and

P294. Future studies should examine the remaining positions of loop 2; e.g., K292 and the almost universally conserved G289. Further bioinformatics analysis of RR may also search for positions that covary similarly to positions 291 and 294. Continued thorough phylogenetic and structure-function analysis will likely reveal many surprising roles for individual amino acid residues in regulation of eukaryotic RR enzymes.

Figure 3-2. Sequence conservation in eukaryotic ribonucleotide reductase and the extent of variation in loop 2. A: Crystal structure model of the hRR large subunit homodimer (PDB ID: 3HND, 3.21 Å resolution. ATP was modeled into the figure by aligning this structure to PDB ID:

3HNE.). The protein backbone is shown as ribbons colored by conservation score (see key). Yellow is used to indicate that insufficient data exist for that position. The S-site ligands (dTTP) are shown as blue spheres. The C-site ligands (GDP) are shown as green spheres. The A-site ligand (ATP) is shown as red spheres. B: As in panel A, but with the protein represented as a van der Waals surface. C: Representative examples of loop 2 diversity in eukaryotes. The group to which each organism belongs is listed to the right of its species name (A = Animalia; F = Fungi; Pl = Plantae; Pr = Protista). The loop 2 sequence of each organism is shown to the right of its group name. Positions 291 and 294 are indicated and in bold (human numbering). Deviations from the human loop 2

100

sequence are shown in red. The species name and loop 2 sequence are surrounded by a blue box if the loop 2 sequence is of Type I (identical to hRR), a green box if the loop 2 sequence is of Type II (N291G and P294K), and an orange box if the loop 2 sequence has the N291G substitution only (human numbering). D: Pie charts showing the percentage of organisms from each group with each major type of loop 2 sequence. Type I and Type

II loop 2 sequences are as described in C. “GP” denotes loop 2 sequences with the

N291G substitution only. “Other” denotes sequences with the conserved glutamine and arginine residues that have some other sequence at the remaining residues. “Type III” denotes putative loop 2 sequences that lack the conserved glutamine and arginine residues.

101

Figure 3-3. Loop 2 sequences from Figure 3-2C mapped onto a eukaryotic tree of life. Loop 2 sequences from Figure 3-2C mapped onto a eukaryotic tree of life. The names of the organisms are arranged according to their evolutionary relationships. The names of plant species are in black; animals, burnt orange; fungi, purple; protists, red. The loop 2 sequence for each species is listed to the right of its name. Type I loop 2s are in blue;

Type II, green; the intermediate “GP” loop 2 is in orange. Note that “GP” loop 2s evolved independently in three different lineages of eukaryotes, while converse “NK” loops are not observed.

102

Figure 3-4. Crystal structure of hR1 bound to dTTP and GDP. PDB ID: 3HND, 3.21 Å resolution. The protein backbone is shown as gray sticks. The C- site ligand (GDP) is shown as green sticks. The S-site ligand (dTTP) is shown as blue sticks. This image includes a neighboring asymmetric unit in close proximity to loop 2

(compare with Figure 3-1B). N291 and the D129 residue of another monomer are closely apposed in this structure; the distance in Ångstroms is indicated. Hydrogen-bonding between these two residues cannot be excluded, particularly if the true bond angle between the β-carbon and γ-carbon of N291 differs by 180º from the assigned bond angle.

This in turn potentially complicates interpretation of the role of N291 in loop 2 function.

103

Figure 3-5. Circular dichroism (CD) of hRR variants. A: Overlaid CD spectra for all enzyme variants used in this study. These spectra were taken at 20ºC. B: Thermal melting of hRR mutants. All variants of hRR were subjected to thermal melting in 2ºC increments and θ, the mean residue ellipticity at 210nm, was monitored. Tm values are displayed with each transition.

104

Figure 3-6. Activity and specificity of the hRR variants. Activity and specificity of the hRR variants. Wild-type data are reprised from

Knappenberger et al. (2016)96. A. Activity of the hRR variants in the presence of 50 μM

dGTP. The specific activity of wild-type hRR is 0.0055+0.0017 mol/(s*mol hRRM1)19.

Data are reported as the ratio of the specific activity of a given enzyme variant to that of

wild-type. B. Specificity of the hRR variants in the presence of 50 μM dGTP. Specificity

is defined as in Knappenberger et al., and is the ratio of the kcat/KM for a given substrate

105

to that of a reference substrate. ‡ indicates a product that was not present in sufficient

quantity to accurately measure its formation. † indicates a product that coelutes with a

hydrolysis product from an effector. C. Activity of the hRR variants in the presence of 50

μM dTTP. The specific activity of wild-type hRR is 0.0066+0.00052 mol/(s*mol

hRRM1)19. Data are reported as the ratio of the specific activity of a given enzyme

variant to that of wild-type. D. Specificity of the hRR variants in the presence of 50 μM

dTTP. Specificity is defined as in Knappenberger et al., and is the ratio of the kcat/KM for

a given substrate to that of a reference substrate. ‡ indicates a product that was not present in sufficient quantity to accurately measure its formation. E. Activity of the hRR variants in the presence of 1 mM ATP. The specific activity of wild-type hRR is

0.070+0.008 mol/(s*mol hRRM1)19. Data are reported as the ratio of the specific activity

of a given enzyme variant to that of wild-type. Wild-type hRR prefers CDP over UDP by

>10-fold96. N291G+P2924K hRR only appreciably reduces CDP. The limit of detection

for this assay is ~1 x 10-4 mol/(s*mol hRRM1). Results are also available in numerical form (Table 3-1).

106

Condition Enzyme Total activity (mol/(s*mol rk (ADP) rk (CDP) rk rk (UDP)

variant hRRM1) (GDP)

wild-type 0.0055 + 0.0018 1 + 0 0.56 + 0.14 ND 0.023 + 0 (one

determination)

50 µM N291G 0.023 + 0.009 0.28 + 1 + 0 ND 0.23 + 0.02

dGTP 0.04

P294K 0.00050 + 0.00016 0.59 + 1 + 0 ND ND

0.15

N291G+ 0.0039 + 0.0011 0.13 + 1 + 0 ND ND

P294K 0.03

Condition Enzyme Total activity (mol/(s*mol rk (ADP) rk (CDP) rk rk (UDP)

variant hRRM1) (GDP)

wild-type 0.0066 + 0.0005 ND 0.080 + 1 + 0 ND

0.029

50 µM N291G 0.022 + 0.008 ND 0.50 + 0.24 1 + 0 0.020 (one determination)

dTTP

P294K 0.00039 + 0.00014 ND ND 1 + 0 ND

N291G+ 0.0022 + 0.0002 ND 0.54 + 0.16 1 + 0 ND

P294K

Condition Enzyme Total activity (mol/(s*mol

variant hRRM1)

wild-type 0.070 + 0.012

1 mM ATP N291G ND

P294K ND

N291G+ 0.012 + 0.008

P294K

Table 3-1. Numerical values from activity assays in Figure 3-6.

107

Wild-type data are reprised from Knappenberger et al.96. Reported error values indicate 1

SD. The fastest substrate’s rk always has a value of 1 + 0 because slower substrates’ rk values are normalized to this substrate.

108

Figure 3-7. Size exclusion chromatography (SEC) of hRR variants in the presence of dGTP, dTTP or ATP. Size exclusion chromatography (SEC) of hRR variants in the presence of dGTP, dTTP or

ATP. A: A simple model depicting S-site ligand binding and oligomerization in hRR.

The protein is represented by large circles; the ligand is represented by small black circles. hRR converts from a monomeric form (red circles) to a dimeric form (green circles) in the presence of dGTP, dTTP, or ATP. ATP also binds at the A-site to trigger formation of hexamers (purple) and intermediate species that migrate as apparent tetramers (blue)83.

The species are shown in the order in which they elute from a size exclusion column. B:

109

Representative SEC chromatograms of hRR variants. Each square represents a representative SEC run. The rows represent the protein variants, while the columns represent the running conditions. The raw data are shown as black dots. The fitted peaks are shown in colors that correspond with oligomeric states in A. The extent of partitioning into each non-monomer state, if appreciable, is indicated in each square. The

P294K mutant is indicated by a dotted gray outline.

110

Figure 3-8. SEC profiles of wild-type and P294K hRR large subunit in the presence of 3 mM ATP. One replicate each. These experiments were conducted essentially as in Figure 3-7. The chromatograph was a GE AKTA Purifier 10 and the flow rate was 0.5 mL/min. Under these conditions, wild-type hRR has an elution profile similar to that obtained at 1 mM

ATP. P294K forms ca. 2-fold more hexamer in the presence of 3 mM ATP than 1 mM

ATP, suggesting that the dissociation constant between the P294K mutant and one of the nucleotide-binding sites may be slightly attenuated.

111

Figure 3-9. Activity and specificity of N291G mutant hR1 in the presence of 2- aminopurine-drTP, dITP, and dZeb. Wild-type data are reprised from Knappenberger et al. (2016)96. The chemical structures

for each effector analog are listed. “R” denotes the deoxyribose triphosphate moiety.

Specificity is defined as in Figure 3-6. ‡ indicates a product that was not present in

sufficient quantity to accurately measure its formation. † indicates a product that coelutes

with a hydrolysis product from an effector. * indicates a substrate that was not included.

Δrk denotes the difference between the specificity of wild-type and N291G hRR under

identical conditions. Negative values indicate that the N291G mutant prefers the substrate

less than wild-type; positive values indicate that the opposite is true. For example, if a

substrate is not appreciably processed by wild-type hRR but is the favored substrate for

N291G hRR, the difference has a value of 1. Activity data are reported as the ratio of the

specific activity of N291G hRR to that of wild-type. The specific activity of N291G hR1 in the presence of 100 μM 2-aminopurine-drTP is 0.061+0.011 mol/(s*mol hRRM1); 100

μM dITP, 0.034+0.0073 mol/(s*mol hRRM1); dZeb, 0.0074+0.0023 mol/(s*mol

hRRM1).

112

113

Figure 3-10. Activity and specificity of the hRR variants in the presence of both ATP and dGTP/dTTP. Wild-type data are reprised from Knappenberger et al. (2016)96. A. Activity of the hRR variants in the presence of 1 mM ATP and 0.75 mM dGTP. The specific activity of wild- type hRR is 0.030+0.0041 mol/(s*mol hRRM1)19. Data are reported as the ratio of the

specific activity of a given enzyme variant to that of wild-type. The limit of detection for

this assay is ~1 x 10-4 mol/(s*mol hRRM1). B. Activity of the hRR variants in the

presence of 1 mM ATP and 1.6 mM dTTP. The specific activity of wild-type hRR is

0.021+0.0040 mol/(s*mol hRRM1)19. Data are reported as the ratio of the specific

activity of a given enzyme variant to that of wild-type. C. The extent of activation or

inhibition of dGTP-bound hRR by ATP. Data represent the ratio of the activity in the

presence of ATP (panel A) to the activity in the presence of dGTP but not ATP (Figure

3-6A). D. The extent of activation or inhibition of dTTP-bound hRR by ATP. Data represent the ratio of the activity in the presence of ATP (panel B) to the activity in the presence of dTTP but not ATP (Figure 3-6C). In panels C and D, numbers greater than unity represent activation, while numbers less than unity represent inhibition. E.

Specificity of the hRR variants in the presence of 1 mM ATP and 0.75 mM dGTP.

114

Specificity is defined as in Figure 3-6, and is the ratio of the kcat/KM for a given substrate to that of a reference substrate. ‡ indicates a product that was not present in sufficient quantity to accurately measure its formation. † indicates a product that coelutes with a hydrolysis product from an effector. F. Specificity of the hRR variants in the presence of

1 mM ATP and 1.6 mM dTTP. Specificity is defined as in Figure 3-6, and is the ratio of the kcat/KM for a given substrate to that of a reference substrate. ‡ indicates a product that was not present in sufficient quantity to accurately measure its formation. Results are also available in numerical form (Table 3-2).

115

Condition Enzyme variant Total activity (mol/(s*mol Activity +/- rk (ADP) rk (CDP) rk rk (UDP)

hRRM1) ATP (GDP)

wild-type 0.030 + 0.004 5.3 + 1.8 1 + 0 0.21 + ND 0.014 +

0.05 0.004

0.75 mM N291G 0.0075 + 0.0050 0.32 + 0.24 0.8 + 1 + 0 ND ND

dGTP + 0.5

1 mM ATP

P294K 0.0023 + 0.0009 4.2 + 2.8 1 + 0 0.53 + ND ND

0.22

N291G+ ND ND ND ND ND ND

P294K

Condition Enzyme Total activity Activity +/- ATP rk (ADP) rk (CDP) rk rk (UDP)

variant (mol/(s*mol hRRM1) (GDP)

wild-type 0.039 + 0.002 6.0 + 1.9 0.021 + 0.062 + 1 + 0 0.0097 +

0.004 0.037 0.0039

1.6 mM N291G 0.0063 + 0.0003 0.28 + 0.09 ND 0.46 + 1 + 0 ND

dTTP + 0.14

1 mM ATP

P294K ND ND ND ND ND ND

N291G+ 0.0046 + 0.0011 2.0 + 0.5 ND 0.83 + 1 + 0 ND

P294K 0.25

Table 3-2. Numerical values from activity assays in Figure 3-9.

116

Reported error values indicate 1 SD. “Activity +/- ATP” refers to the activity of the enzyme variant in the presence of ATP and dGTP/dTTP divided by the activity in the presence of dGTP/dTTP alone. Wild-type data are reprised from Knappenberger et al.96.

The fastest substrate’s rk always has a value of 1 + 0 because slower substrates’ rk values

are normalized to this substrate.

117

Figure 3-11. Correspondence between the present in vitro experiments and “natural experiments” of evolution. A. Naturally occurring patterns of variation in eukaryotic RR loop 2. Type I loop 2

dominates Unikonta (animals, fungi, and some protists), while Type II loop 2 dominates

the rest of Eukarya (Figure 3-3). GP loop 2 evolved independently in representatives of

animals, fungi, and protists. NK loop 2 was not observed in this study of 310 eukaryotic

RRs. B. Compensatory effects of N291G and P294K mutations in hRR loop 2. The two mutations affect specificity in an additive way, but features of the enzyme including oligomerization, specific activity, and ATP stimulation are fully or partially rescued by addition of both substitutions in the same enzyme. N291G has a less severe effect on enzyme function when present individually, which corresponds with the patterns of loop

2 evolution observed in A.

118

Conclusions

In Chapter 3 we bioinformatically examined natural variation in eukaryotic RR loop 2 as an inroad to structure-function studies in that region. The results showed that while loop 2 is well-conserved across Eukarya, conservation is imperfect. In particular, there are two amino acid residues that differ nonconservatively between groups of eukaryotes, and these two substitutions have a high tendency to occur together. Namely,

animals and fungi (collectively Unikonta) have loop 2 regions with an N291 residue and

a P294 residue (human numbering used throughout this section). All remaining

eukaryotes tend to have loop 2s with G291 and K294 residues. We chose to refer to these

two major loop 2 phenotypes as Type I and Type II, respectively. Although many

organisms have a Type I or Type II loop, an intermediate “GP” loop containing G291 and

P294 residues evolved independently in animals, fungi, and algae. From this

bioinformatic analysis we concluded that Type II was the most likely ancestral loop in

Eukarya, and that the most likely point of evolutionary interconversion between these

phenotypes is the “GP” loop, because the converse “NK” loop was not observed in the

310 organisms we sampled.

In order to learn how variants of loop 2 achieve functional equivalency between

organisms, we constructed three mutants of hRR: N291G, giving the protein a “GP” loop;

P294K, giving the protein an “NK” loop, and N291G+P294K, giving the protein a Type

II loop. We found that introducing any of these mutations is not disruptive to overall

protein structure, but either single mutation causes defects in protein function. We found

that the P294K variant was the most deficient with respect to activity, specificity, and

ligand-induced oligomerization. This is consistent with the fact that we did not observe

119

an “NK” loop in nature, though we observed several instances of “GP” loops.

Surprisingly, we found that many, but not all, defects associated with the P294K mutant

were rescued in the double mutant. This is consistent with our observation that Type II is

a highly prevalent types of loop 2 in Eukarya. However, the fact that loop 2 is not a

perfect functional module suggests that there are other loci on the protein that covary

with loop 2 and form a functional ensemble with it to bring about normal function.

Because this chapter has not been peer-reviewed as of March 2017, its

conclusions must be interpreted in a more open-ended way than those of Chapter 2. In particular, the gel filtration results from the present study are only partially consistent with a study the Hofer lab conducted under similar conditions36. Furthermore, a small- angle X-ray scattering study conducted by the Drennan lab and unpublished gel filtration studies from the Dealwis lab show different populations of oligomers when compared to the present study32. The aforementioned studies generally show that 3 mM ATP causes

hR1 to form predominantly hexamers, with small populations of dimers and monomers

and without species that could be interpreted as tetramers. In contrast, the present study

shows that hR1 forms primarily dimers in the presence of 1 mM ATP, with smaller

populations of hexamers and apparent tetramers. It is important to note that methods of protein purification and running conditions vary among the studies mentioned above. A key exception is that the method of protein purification from the present study is identical to that used to gather the unpublished Dealwis lab data. Possible dissociation of hexamers during gel filtration in the present experiment is a potential explanation for the formation of peaks that are of lower molecular weight than expected. Other potential explanations include effects from variations in protein concentration or temperature, the presence of an

120

impurity that prevents proper oligomerization, and a structural defect in some population

of the protein such as loss of the N-terminal hexamerization domain. At the time of approval of this dissertation, it is important to note that the discrepancies between our oligomerization data and that presented in the literature have not been definitively accounted for. Future experiments in this study and others will need to be conducted before a complete understanding of ATP-mediated hexamerization in hRR can be reached, and a survey of potential techniques is included in Chapter 4.

121

Materials and methods

Phylogenetic comparative sequence analysis- cDNA sequences of eukaryotic RR

large subunits were retrieved from GenBank. GenBank entries of putative and predicted

proteins as well as proteins under 500 amino acid residues were excluded 86. In the two

instances when duplicate entries existed for a single organism, the protein sequence that

was most conserved within the organism’s kingdom was chosen for further analysis.

Amino acid sequence alignment was carried out using the Multiple Sequence Alignment

feature of Clustal Omega84-85. This multiple sequence alignment was used to produce

conservation scores for hRR in ConSurf105 and visualized using the PyMOL Molecular

Graphics System, Version 1.8, Schrödinger, LLC. A phylogenetic tree of all 310 organisms queried was generated using phyloT and Interactive Tree of Life based on their taxonomy data from the National Center for Biotechnology Information (NCBI)106.

Mutagenesis and purification of human ribonucleotide reductase- Protein purification was conducted as described previously96. Site-directed mutagenesis of hR1

was conducted essentially as described previously96. The N291G mutant was generated

using primers 5’-GATCAAGGTGGGGGCAAGCGTCCTGG-3’ and 5’-

CACATATCTAGCTGTGTTGTTATATACTCTCAGC-3’. The P294K mutant was

generated using primers 5’-GAACAAGCGTAAGGGGGCATTTG-3’ and 5’-

CCACCTTGATCCACATATCTAGCTGTGTT-3’. The N291G+P294K mutant was generated by using primers 5’-GGGCAAGCGTAAGGGGGCATTTG-3’ and 5’-

CCACCTTGATCCACATATCTAGCTGTGTT-3’ on N291G mutant DNA. Plasmid

sequences were verified by Sanger sequencing from primers 5’-

TTCGGCTTTAAGACGCTAGA-3’, 5’-CTTGGCATTTAGACATCTTTGA-3’, 5’-

122

TTGGCTGAAGTCACTAAAGTCG-3’, 5’-CGCAGAGTCTTGTCAGGAGA-3’, and

the T7 promoter. All mutants were purified via the same method as wild-type hR1.

Circular dichroism (CD) spectroscopy- CD spectroscopy was carried out

essentially as described in Davis et al101. Briefly, stocks of hR1 in 50 mM Tris pH 8.0, 5

mM MgCl2, 5% glycerol, 10 mM DTT were diluted into >10-fold excess of 25 mM

potassium phosphate dibasic, pH 7.5 for a final volume of 800 µL and a final protein

concentration of 0.1 mg/mL. Initial CD spectra were recorded at 20 ºC on an Applied

Photophysics PiStar 180 spectrophotometer in a quartz cuvette with a path length of 0.5 mm. The sample was heated in 2 ºC increments and held at each temperature for 30 s before recording each spectrum. Samples were heated to at least 86 ºC to ensure that thermal melting was complete. Fitting was performed using Origin 8. Melting curves were fit to Equation 3-1:

( ) ( ) 훥퐻푚 = 푛 푛 푑 푑 1 1 Equation 3-1 푦 +푚 푇 + 푦 +푚 푇 exp�푅� − �� 푇푚 푇

훥퐻푚 푌 1 1 1+exp�푅� − �� 푇푚 푇

Where Y is the measured ellipticity, ΔHm is the enthalpy at the unfolding transition, Tm is

the melting temperature in degrees Kelvin, and R is the universal gas constant107. The

baseline and slope before the transition are represented by mn and yn, respectively.

Likewise, md and yd represent the baseline and slope after the transition.

Analyses of hRR catalytic activity and NDP substrate specificity- Kinetic assays

were conducted as described previously96. Assays were performed at 37 °C in 50 mM

gly-gly pH 7.7, 15 mM MgCl2, 20 mM DTT. Protein concentrations were 0.5 μM R1 and

123

5 μM R2. Substrates ADP, CDP, GDP, and UDP were typically present at 0.6 mM. Two

aliquots were removed prior to initiating the reaction. These samples were used to empirically measure the substrate concentrations by high-performance liquid chromatography (HPLC), and to verify the efficacy of subsequent boronate chromatography to remove unreacted substrate. Reactions were incubated for 30 min. All

aliquots were quenched by rapid freezing to -80 ºC. Substrates were removed by boronate

chromatography essentially as described in Hendricks et al.58. Substrate and product

concentrations were quantified by anion exchange HPLC as in Hendricks et al.58.

Reactions were kept in steady state conditions (fraction of reaction < 10%). Substrate

52-55 r concentration and vobs data were combined via Equation 3-2 to yield the k, or the

ratio of each substrate’s kcat/KM relative to that of the reference substrate.

[ ] [ ] = 푘푐푐푐 = Equation 3-2 , � � � 푣표표표1 퐾푚 1 푆1 푟 푆1 푣표표표 푟푟푟 푘푐푐푐 �푆푟푟푟� 1 �푆푟푟푟� �� 퐾푚�푟푟푟� 푘 The specificity of hRR is thus quantified by comparison of rk values that represent the

ratio of each substrate’s kcat/KM relative to that of a reference substrate. In all cases, specific activity is reported as the enzymatic activity of the R1 subunit across all four

NDP substrates.

Analysis of hRR oligomerization by size exclusion chromatography (SEC)- SEC

was carried out essentially as in Fairman et al19. Briefly, 100 µL of 4 µM hR1 in 50 mM

Tris pH 7.6, 5 mM MgCl2, 100 mM KCl was mixed with 0-4 µL of NTP/dNTP solution.

The mixture was centrifuged 10 min at 20,000 g and 25+3 ºC. The mixture was injected

onto a Superdex 200 10/300 size exclusion column on a Shimadzu LC-20AB

chromatograph equilibrated with 50mM Tris pH 7.6, 5 mM MgCl2, 100 mM KCl and a

124

concentration of ligand equal to that in the sample. SEC was carried out at 25+3 ºC. Flow rate was 0.25 mL/min and absorbance was recorded at 290 nm. Peak retention time was correlated with apparent molecular weight using the retention times of a set of standards

(Sigma Molecular Weight Marker Kit). Peak fitting to multiple Lorentzian peaks was performed using Origin 8.

125

Chapter 4 : Summary and Future Directions

Summary

The use of biological macromolecules to accomplish chemical catalysis is a

fundamental feature of life. The enzymes that catalyze the chemistry of life enhance the

pace of desirable reactions by astronomical factors, allowing organisms to grow and

reproduce on a time scale of minutes rather than eras1. Just as important as enzymes’

catalytic capacity is their ability to differentiate among alternative substrates, so that

reactions that are deleterious or maladaptive essentially do not occur on biological time

scales. Since the discovery of enzymes, scientists have marveled at their exquisite

substrate specificity, and we have devoted over 50 years of research to studies of enzyme

allostery13. Allostery is a common mode of enzyme regulation, but is most commonly used to regulate overall enzyme activity. However, ribonucleotide reductases (RRs) use allostery to regulate their specificity and thereby maintain balance in dNTP pools19, 21-22,

25, 27, 39-40. Because of its unique mechanism of regulation and central importance for

health and disease, human RR (hRR) is a desirable target for research into the

mechanisms and effects of protein allostery.

Situated between RR’s active site and an allosteric site called the specificity site

(S-site), a short loop called loop 2 plays an important role in the transfer of biological

information into specificity. From extensive crystallographic study we know that loop 2

assumes a unique conformation in the presence of each substrate/effector pair19, 22, 25, 27.

However, X-ray crystallography presents only a static picture of the fully formed protein-

ligand complexes. This means that it is not possible to use this technique to shed light on

126

the features of the ligand that the protein functionally recognizes, as opposed to those that

are only incidentally present. The experiments summarized in Chapter 2 complement the

existing structural biology with chemical biology and multiple substrate kinetics. In

Chapter 2, we developed and validated an assay for determining the specificity of human

RR in the presence of a given effector. Using this assay, we determined hRR specificity

in the presence of a panel of modified dNTPs. The modified dNTPs bind in the S-site, yet

have systematic modifications to the nucleobase moiety. By correlating the alterations in

nucleotide structure with changes in observed specificity, we were able to deduce the

functional groups on the effector that drive specificity.

hRR has three native effectors: ATP/dATP, dGTP, and dTTP. Although the

purine effectors ATP and dGTP have different functionality at the 1, 2, and 6 positions,

we found that hRR only uses the protonation state of the N1 group to differentiate

between the two. In contrast, we found that hRR uses several features of dTTP to drive

specificity – chiefly the 3, 4, and 5 positions. Examining available X-ray crystal

structures showed that amino acid residue D287 appears to be involved in key contacts

with all three effectors. Mutating this residue to alanine resulted in an enzyme variant that

always prefers to reduce CDP – that is, it treats all three effectors as if they were ATP.

This is consistent with a model in which the CDP-reducing conformation of hRR is the most thermodynamically stable of the three. ATP binding causes the enzyme to dimerize but does not change loop 2’s conformation. Binding energy from interactions between dGTP or dTTP and the side chain of D287 perturbs loop 2’s conformational space. These new conformations favor reduction of ADP and GDP, respectively.

127

Although crystal structures implicate amino acid residues Q288 and R293 as important for loop 2 function, all of loop 2 is well-conserved among eukaryotes. This means that its amino acids likely work together to perform loop 2’s functions. However, conservation among amino acids in loop 2 is imperfect among Eukarya. This raises the question of how different loop 2 phenotypes might achieve functional equivalency with respect to one another. In Chapter 3, we explored natural loop 2 sequence variation in eukaryotes as a way of bioinformatically and biochemically interrogating this amino acid sequence space. This bioinformatic inquiry revealed a surprising trend in loop 2 sequences. Most amino acids in loop 2 are strongly conserved, but two of them show variation. N291 and P294 are mutated to a glycine and a lysine, respectively, in most eukaryotes not belonging to the group Unikonta (animals and fungi). We termed these types of loop 2 Type I and Type II, respectively. These two substitutions are highly nonconservative, and they almost always covary. In addition, organisms exist which have a glycine and a proline, giving them an intermediate “GP” loop 2. Conversely, we did not observe any organisms that have the other intermediate “NK” loop 2. From this analysis we can draw two major conclusions: that Type II was the ancestral eukaryotic loop 2, and that the most likely point of interconversion between these two phenotypes is through a

“GP” loop.

Having observed apparent functional equivalency between Type I and Type II loop 2s, we created three mutants of hRR in order to observe the effects of individual substitutions in a wild-type background. They were: N291G, giving hRR a “GP” loop 2;

P294K, giving hRR an “NK” loop 2, and N291G+P294K, giving hRR a Type II loop 2.

We found that none of these mutations was broadly destabilizing to hRR structure or

128

catalytic activity, in spite of the fact that we introduced nonconservative mutations in a well-conserved region of the protein. Although crystal structures paint loop 2 as the sole mediator of specificity, the mutants all showed a greater preference for CDP than wild- type hRR. We also observed that the N291G mutant had overall less deleterious effects than the P294K mutant. Further, we found that defects in the P924K mutant’s activity and ability to oligomerize were often rescued in the double mutant. Kinetic results in the presence of ATP showed that ATP’s normal role as an allosteric activator through binding at the activity site (A-site) is perturbed in the presence of these loop 2 mutations.

From in vitro experiments we were able to draw several key conclusions. First, loop 2 is not perfectly modular with respect to specificity regulation, because “transplanting” a

Type II loop into an animal RR did not produce an enzyme with normal specificity.

Second, the fact that the mutants prefer CDP more than wild-type reinforces the conclusion from Chapter 2: that the CDP-reducing conformation of hRR is the most stable, and binding energy from dGTP or dTTP binding primarily serves to perturb that conformation. Finally, although loop 2 is canonically only implicated in regulation of specificity, it must also play a role in the transfer of allosteric information from the A-site that normally results in catalytic activation. Therefore, loop 2 is more important for activity regulation, and less important for specificity regulation, than we had expected.

In conclusion, because of the work contained herein it is possible to enhance the model of how loop 2 in hRR works to effect enzyme function. Loop 2 likely has a conformation that is lowest in free energy and presents an active site that most prefers to bind and reduce CDP. Contacts between the N1 of dGTP or the N3 of dTTP and the carboxylic acid group of D287 pay an energetic cost to changing the conformation of the

129

loop so that the active site prefers purine substrates (see Chapter 2). Crystal structures

and biochemical data indicate that R293 and Q288 play prominent roles in this process.

We have also shown loop 2 to play key roles in oligomerization and in allosteric regulation by ATP binding at the A-site (see Chapter 3). Continued biochemical and structural inquiries will help to further enhance the molecular details of the features depicted in this model.

130

Future directions

A good scientific inquiry both answers old questions and raises new ones. While

the present experiments answered some essential questions about hRR regulation, they

also invite new inquiry at the emerging frontiers of research into enzyme specificity and

allostery. Experiments from Chapters 2 and 3 invite structural biology as a technique for

complementing their biochemical observations, just as they complement previous

structural observations. It would be beneficial to crystallize D287A hRR in the presence

of the three effectors, to determine the new conformations of loop 2 and to see how

effector recognition is perturbed in the context of this mutation. For similar reasons, it

would also be interesting to learn how the structure of the enzyme adapts to the chemical

mutations introduced by using nucleotide analogs as effectors. In particular, it would be

beneficial to observe how hRR adapts to the presence of dZeb in the S-site, and

potentially learn why it prefers CDP and GDP equally when this effector analog is bound.

In each case, the biochemical data raise clear predictions regarding loop 2’s

conformational space, and learning whether these predictions are borne out in crystallo

can only contribute to a fuller understanding of allostery in hRR.

The experiments summarized in Chapter 2 showed us that in spite of the

existence of an abundance of differential protein-ligand contacts in the S-site, only a key subset of those contacts drive specificity at the active site. In particular, a Hydrogen-bond

between the carboxylic acid side chain of amino acid residue D287 and the protonated N1

of dGTP is predicted to drive specificity. In this model, ATP binding does not direct

reduction of ADP because its N1 is unprotonated and therefore does not form an H-bond

with the side chain of D287. Therefore, if D287 is mutated to asparagine, ATP should

131

direct reduction of ADP because the primary amine of glutamine can serve as an H-bond

donor and interact with the unprotonated N1 of ATP. D287N hRR should also reduce

primarily ADP in the presence of dGTP, because its amide carbonyl group can still

function as an H-bond acceptor. However, specificity for ADP may be attenuated in this

case because the partial negative charge on the carbonyl oxygen should be less than the

predicted -1/2 charge on a carboxyl oxygen.

Site-directed mutagenesis has proved to be a powerful tool for experimentation

within the present work, and promises to continue to be essential in the future. In pursuit

of the mechanism by which hRR excludes ribonucleoside triphosphates from the S-site, the author has constructed the Y285A mutant of hRR. The hypothesis was that because the hydroxyl group of this tyrosine residue is proximal to the 2′ carbon of the effector in crystallo, it may be working to exclude ribonucleotides by steric hindrance. However, experimentation has not borne this out. Rather, Y285A hRR has perturbed specificity for purines on par with that of the N291G+P294K mutant (data not shown). This is consistent with its previously observed effect on S. cerevisiae in vivo77-78. In light of these

results, it is possible that Y285 is responsible for sterically packing the effector such that

its N1 or N3 groups may interact with D287. However, this model has not been tested.

Intriguingly, loop 2 is not modular with respect to specificity. Because D.

melanogaster has an intermediate “GP” loop 2 yet is a relatively close relative of humans,

it may be enlightening to mutagenize the human enzyme incrementally towards the fly

enzyme until specificity is restored. Along that line of thinking, we have conducted a

smaller, yet more stringent, bioinformatics analysis of some enzymes that have Type I

and Type II loop 2s (data not shown). We found that there are only two amino acid

132

residues that “invariantly vary” in this context; i.e., they always take on one identity in

the context of a Type I loop, and another in the context of a Type II loop. The first is

M115, which is a leucine in the context of a Type II loop. This methionine is positioned

to form an S-π interaction with nearby Y104. With a conformational change, it may also

be capable of forming a similar interaction with Y261, an amino acid residue close to the

S-site. The second is Q749, which is a lysine in the context of a Type II loop.

Unfortunately Q749 is in the disordered C-terminus of the protein and therefore is not

visible in crystal structures. However, it is possible that this amino acid residue also

exists near the active site/loop 2/S-site axis. The fact that the only two amino acid

residues that fit our criteria of covariance with loop 2 are also proximal to its region of

the protein is intriguing, to say the least. It is likely that construction of triple and

quadruple mutants at these positions would help to shed light on why loop 2 cannot

regulate specificity alone, and how it participates in transferring chemical information

from the A-site.

The experiments in Chapter 3 have not been peer-reviewed as of March 2017, and discrepancies between some of its results and those present in the literature invite further experimentation. Remaining within the confines of gel filtration, increasing the protein concentration in the gel filtration experiments from Chapter 3 will help resolve any signal-to-noise issues, and increasing the concentration of ATP to 3 mM will reduce the possibility that the nucleotide-binding sites are unsaturated under the running conditions. In any case, it will be important to ensure consistency between the present results and the unpublished data from the Dealwis lab.

133

While it will be important to continue to use SEC (gel filtration) as a tool to

determine the oligomeric state of RR, use of multiple techniques will lead to stronger

conclusions than if only one technique is employed. A wide variety of techniques has been used to assay RR’s oligomeric state. Here, I include a catalog of potential techniques with citations of selected studies that employ them. This catalog is not intended to be exhaustive, and in particular is generally restricted to studies of the human enzyme when possible. SEC has of course been used to measure RR’s oligomeric state, and a publication from the Dealwis group provides a representative reference19. The

Stubbe group has also made use of this technique to determine the mechanism of action

for chemotherapeutic agents37, 104. Disadvantages of SEC include that it is a non-

equilibrium technique, that the presence of high concentrations of nucleotides interferes

with measuring absorbance from protein molecules, and that it requires the use of large

amounts of nucleotides if the concentration of effector is to remain constant. There is a

fluorescence-based method developed by the Aye group. It has been used to interrogate

the mechanism of action for chemotherapeutic agents, though one can note several

important deficiencies with this method91. Namely, the authors did not test the

hexamerization efficiency of one of the tagged proteins (Figure 2), the observed

quenching effect can be explained by the formation of either dimers or hexamers (Figure

4), there is an unexplained fast phase in association (Figure 4C), and there is a further

unexplained change in fluorescence that is attributed to a postulated conformational

change (Figure 5) (figure references are all to Fu et al.). The Cooperman group has

studied murine RR (mRR) extensively, and in one publication uses both analytical

centrifugation and light scattering to assay mRR oligomeric state72. Light scattering

134

reports only on the average molecular weight of the particles present, and observations

with this technique led to the conclusion that mRR can form primarily tetramers, a

conclusion that has not been borne out in other studies cited in this section33. Analytical

centrifugation has not been used by other groups to the author’s knowledge, perhaps

because it is also a non-equilibrium technique or because it returns results in Svedbergs

rather than reporting on oligomeric state directly.

The Drennan group has made use of small-angle X-ray scattering (SAXS) and

electron microscopy (EM) to measure the oligomeric state of hRR32. This study is

noteworthy because it uses two excellent and complementary techniques to measure the same phenomenon, and therefore is an exemplar among the studies cited in this section.

SAXS is an equilibrium technique that measures the average particle size in a sample. It

has the disadvantage that it cannot discern the sizes of individual particles, but this is

complemented by the use of EM. In EM, the experimenter directly observes the electron

density of individual particles as in a photograph, allowing for quantitative descriptions

of the ensembles of particles present. For future inquiries into hRR oligomeric state, the

most viable method is likely to be the approach of Ando et al. The chief disadvantage of

this approach is that it requires the use of specialized equipment, but it is likely that the

advantages will far outweigh this drawback. Conducting these experiments will help

ensure the accuracy of any conclusions regarding the oligomeric states of the loop 2

mutants.

The present work will not be the last doctoral dissertation on regulation of

ribonucleotide reductases. Among the already stated potential future directions, a more

extensive bioinformatics analysis of RRs from across all domains of life should be

135

conducted. In addition, a chemical interrogation of the active site analogous to that

conducted in Chapter 2 on the S-site would provide key information. Finally, using the

information gleaned from this work as an inroad to development of therapeutics that

target the S-site is an excellent way to use this basic knowledge to improve and lengthen

people’s lives. Such therapy may in particular be synergistic with cancer mutations that

disrupt mismatch repair, as is often the case in colon, ovarian and endometrial cancers108-

111. In theory, such therapy may disrupt dNTP pools such that cancerous cells cannot cope

with the additional mismatch errors while normal cells are less affected. In any case, new

surprises surely await those who carry the torch forward and continue research into these

vital and fascinating enzymes.

136

Bibliography

1. Wolfenden, R.; Snider, M. J., The depth of chemical time and the power of enzymes as catalysts. Accounts of chemical research 2001, 34 (12), 938-945. 2. Lineweaver, C. H.; Egan, C. A., Life, gravity and the second law of thermodynamics. Physics of Life Reviews 2008, 5 (4), 225-242. 3. Glasner, M. E.; Gerlt, J. A.; Babbitt, P. C., Evolution of enzyme superfamilies. Current opinion in chemical biology 2006, 10 (5), 492-497. 4. Lienhard, G. E., Enzymatic catalysis and transition-state theory. Science 1973, 180 (4082), 149-154. 5. Harris, P.; Navarro Poulsen, J.-C.; Jensen, K. F.; Larsen, S., Structural basis for the catalytic mechanism of a proficient enzyme: orotidine 5'-monophosphate decarboxylase. Biochemistry 2000, 39 (15), 4217-4224. 6. Fischer, E., Einfluss der Configuration auf die Wirkung der Enzyme. Berichte der deutschen chemischen Gesellschaft 1894, 27 (3), 2985-2993. 7. Cummings, R. D., The repertoire of glycan determinants in the human glycome. Molecular BioSystems 2009, 5 (10), 1087-1104. 8. Pauling, L., Chemical achievement and hope for the future. Am. Sci. 1948, 36, 51-58. 9. Guenther, U.-P.; Yandek, L. E.; Niland, C. N.; Campbell, F. E.; Anderson, D.; Anderson, V. E.; Harris, M. E.; Jankowsky, E., Hidden specificity in an apparently nonspecific RNA-binding protein. Nature 2013, 502 (7471), 385-388. 10. Bevilacqua, A.; Ceriani, M. C.; Capaccioli, S.; Nicolin, A., Post‐transcriptional regulation of gene expression by degradation of messenger RNAs. Journal of cellular physiology 2003, 195 (3), 356-372. 11. Hao, N.; O'Shea, E. K., Signal-dependent dynamics of transcription factor translocation controls gene expression. Nature structural & molecular biology 2012, 19 (1), 31-39. 12. Motlagh, H. N.; Wrabl, J. O.; Li, J.; Hilser, V. J., The ensemble nature of allostery. Nature 2014, 508 (7496), 331-339. 13. Changeux, J. P., 50th anniversary of the word “allosteric”. Protein Science 2011, 20 (7), 1119-1124. 14. Gerhart, J. C.; Pardee, A. B., The enzymology of control by feedback inhibition. The Journal of biological chemistry 1962, 237, 891-896. 15. Goldstone, D. C.; Ennis-Adeniran, V.; Hedden, J. J.; Groom, H. C.; Rice, G. I.; Christodoulou, E.; Walker, P. A.; Kelly, G.; Haire, L. F.; Yap, M. W., HIV-1 restriction factor SAMHD1 is a deoxynucleoside triphosphate triphosphohydrolase. Nature 2011, 480 (7377), 379- 382. 16. Lahouassa, H.; Daddacha, W.; Hofmann, H.; Ayinde, D.; Logue, E. C.; Dragin, L.; Bloch, N.; Maudet, C.; Bertrand, M.; Gramberg, T., SAMHD1 restricts the replication of human immunodeficiency virus type 1 by depleting the intracellular pool of deoxynucleoside triphosphates. Nature immunology 2012, 13 (3), 223-228. 17. Ji, X.; Tang, C.; Zhao, Q.; Wang, W.; Xiong, Y., Structural basis of cellular dNTP regulation by SAMHD1. Proceedings of the National Academy of Sciences 2014, 111 (41), E4305-E4314. 18. Jang, S.; Zhou, X.; Ahn, J., Substrate specificity of SAMHD1 triphosphohydrolase activity is controlled by deoxyribonucleoside triphosphates and phosphorylation at Thr592. Biochemistry 2016, 55 (39), 5635-5646. 19. Fairman, J. W.; Wijerathna, S. R.; Ahmad, M. F.; Xu, H.; Nakano, R.; Jha, S.; Prendergast, J.; Welin, R. M.; Flodin, S.; Roos, A., Structural basis for allosteric regulation of human

137

ribonucleotide reductase by nucleotide-induced oligomerization. Nature structural & molecular biology 2011, 18 (3), 316-322. 20. Nordlund, P.; Reichard, P., Ribonucleotide reductases. Annu. Rev. Biochem. 2006, 75, 681-706. 21. Brown, N. C.; Reichard, P., Role of effector binding in allosteric control of ribonucleoside diphosphate reductase. Journal of molecular biology 1969, 46 (1), 39-55. 22. Xu, H.; Faber, C.; Uchiki, T.; Fairman, J. W.; Racca, J.; Dealwis, C., Structures of eukaryotic ribonucleotide reductase I provide insights into dNTP regulation. Proceedings of the National Academy of Sciences of the United States of America 2006, 103 (11), 4022-4027. 23. Huang, M.; Zhou, Z.; Elledge, S. J., The DNA replication and damage checkpoint pathways induce transcription by inhibition of the Crt1 repressor. Cell 1998, 94 (5), 595-605. 24. Zhang, Z.; An, X.; Yang, K.; Perlstein, D. L.; Hicks, L.; Kelleher, N.; Stubbe, J.; Huang, M., Nuclear localization of the Saccharomyces cerevisiae ribonucleotide reductase small subunit requires a karyopherin and a WD40 repeat protein. Proceedings of the National Academy of Sciences 2006, 103 (5), 1422-1427. 25. Zimanyi, C. M.; Chen, P. Y.-T.; Kang, G.; Funk, M. A.; Drennan, C. L., Molecular basis for allosteric specificity regulation in class Ia ribonucleotide reductase from Escherichia coli. eLife 2016, 5, e07141. 26. Yao, R.; Zhang, Z.; An, X.; Bucci, B.; Perlstein, D. L.; Stubbe, J.; Huang, M., Subcellular localization of yeast ribonucleotide reductase regulated by the DNA replication and damage checkpoint pathways. Proceedings of the National Academy of Sciences 2003, 100 (11), 6628- 6633. 27. Larsson, K.-M.; Jordan, A.; Eliasson, R.; Reichard, P.; Logan, D. T.; Nordlund, P., Structural mechanism of allosteric substrate specificity regulation in a ribonucleotide reductase. Nature structural & molecular biology 2004, 11 (11), 1142-1149. 28. Stubbe, J.; van der Donk, W. A., Ribonucleotide reductases: radical enzymes with suicidal tendencies. Chemistry & biology 1995, 2 (12), 793-801. 29. Dutia, B. M., Ribonucleotide reductase induced by herpes simplex virus has a virus- specified constituent. Journal of General Virology 1983, 64 (3), 513-521. 30. Stubbe, J.; Nocera, D. G.; Yee, C. S.; Chang, M. C., Radical initiation in the class I ribonucleotide reductase: long-range proton-coupled electron transfer? Chemical reviews 2003, 103 (6), 2167-2202. 31. Thelander, L.; Reichard, P., Reduction of ribonucleotides. Annual review of biochemistry 1979, 48 (1), 133-158. 32. Ando, N.; Li, H.; Brignole, E. J.; Thompson, S.; McLaughlin, M. I.; Page, J. E.; Asturias, F. J.; Stubbe, J.; Drennan, C. L., Allosteric inhibition of human ribonucleotide reductase by dATP entails the stabilization of a hexamer. Biochemistry 2015, (55 (2)), 373–381. 33. Kashlan, O. B.; Scott, C. P.; Lear, J. D.; Cooperman, B. S., A comprehensive model for the allosteric regulation of mammalian ribonucleotide reductase. Functional consequences of ATP- and dATP-induced oligomerization of the large subunit. Biochemistry 2002, 41 (2), 462-474. 34. Cooperman, B. S.; Kashlan, O. B., A comprehensive model for the allosteric regulation of Class Ia ribonucleotide reductases. Advances in enzyme regulation 2003, 43 (1), 167-182. 35. Rofougaran, R.; Crona, M.; Vodnala, M.; Sjöberg, B.-M.; Hofer, A., Oligomerization status directs overall activity regulation of the Escherichia coli class Ia ribonucleotide reductase. Journal of Biological Chemistry 2008, 283 (51), 35310-35318. 36. Rofougaran, R.; Vodnala, M.; Hofer, A., Enzymatically active mammalian ribonucleotide reductase exists primarily as an α6β2 octamer. Journal of Biological Chemistry 2006, 281 (38), 27705-27711.

138

37. Wang, J.; Lohman, G. J.; Stubbe, J., Enhanced subunit interactions with gemcitabine-5′- diphosphate inhibit ribonucleotide reductases. Proceedings of the National Academy of Sciences 2007, 104 (36), 14324-14329. 38. Ando, N.; Brignole, E. J.; Zimanyi, C. M.; Funk, M. A.; Yokoyama, K.; Asturias, F. J.; Stubbe, J.; Drennan, C. L., Structural interconversions modulate activity of Escherichia coli ribonucleotide reductase. Proceedings of the National Academy of Sciences 2011, 108 (52), 21046-21051. 39. Reichard, P., Ribonucleotide reductases: the evolution of allosteric regulation. Archives of biochemistry and biophysics 2002, 397 (2), 149-155. 40. Eriksson, M.; Uhlin, U.; Ramaswamy, S.; Ekberg, M.; Regnström, K.; Sjöberg, B.-M.; Eklund, H., Binding of allosteric effectors to ribonucleotide reductase protein R1: reduction of active-site cysteines promotes substrate binding. Structure 1997, 5 (8), 1077-1092. 41. Artin, E.; Wang, J.; Lohman, G. J.; Yokoyama, K.; Yu, G.; Griffin, R. G.; Bar, G.; Stubbe, J., Insight into the mechanism of inactivation of ribonucleotide reductase by gemcitabine 5′- diphosphate in the presence or absence of reductant. Biochemistry 2009, 48 (49), 11622-11629. 42. Cerqueira, N. M.; Fernandes, P. A.; Ramos, M. J., Understanding ribonucleotide reductase inactivation by gemcitabine. Chemistry-A European Journal 2007, 13 (30), 8507-8515. 43. Aye, Y.; Stubbe, J., Clofarabine 5′-di and-triphosphates inhibit human ribonucleotide reductase by altering the quaternary structure of its large subunit. Proceedings of the National Academy of Sciences 2011, 108 (24), 9815-9820. 44. Bonate, P. L.; Arthaud, L.; Cantrell, W. R.; Stephenson, K.; Secrist, J. A.; Weitman, S., Discovery and development of clofarabine: a nucleoside analogue for treating cancer. Nature reviews Drug discovery 2006, 5 (10), 855-863. 45. Johnson, K. A.; Goody, R. S., The original Michaelis constant: translation of the 1913 Michaelis–Menten paper. Biochemistry 2011, 50 (39), 8264-8269. 46. Northrop, D. B., On the meaning of Km and V/K in enzyme kinetics. J. Chem. Educ 1998, 75 (9), 1153. 47. Hofer, A.; Ekanem, J. T.; Thelander, L., Allosteric regulation of Trypanosoma bruceiribonucleotide reductase studied in vitro and in vivo. Journal of Biological Chemistry 1998, 273 (51), 34098-34104. 48. Cha, S., Kinetics of enzyme reactions with competing alternative substrates. Molecular pharmacology 1968, 4 (6), 621-9. 49. Schellenberger, V.; Siegel, R. A.; Rutter, W. J., Analysis of enzyme specificity by multiple substrate kinetics. Biochemistry 1993, 32 (16), 4344-8. 50. Cleland, W. W., The use of isotope effects to determine enzyme mechanisms. Arch Biochem Biophys 2005, 433 (1), 2-12. 51. Melander, L.; Saunders, W. H., Reaction Rates of Isotopic Moleculaes. Wiley: New York, 1980. 52. Schellenberger, V.; Siegel, R. A.; Rutter, W. J., Analysis of enzyme specificity by multiple substrate kinetics. Biochemistry 1993, 32 (16), 4344-4348. 53. Northrop, D. B., Fitting enzyme-kinetic data to VK. Analytical biochemistry 1983, 132 (2), 457-461. 54. Northrop, D. B., Rethinking fundamentals of enzyme action. Advances in enzymology and related areas of molecular biology 1999, 73, 25-56. 55. Cornish-Bowden, A., Enzyme specificity: its meaning in the general case. Journal of theoretical biology 1984, 108 (3), 451-457. 56. Anderson, V. E., Multiple alternative substrate kinetics. Biochimica et Biophysica Acta (BBA)-Proteins and Proteomics 2015.

139

57. Cha, S., Kinetics of enzyme reactions with competing alternative substrates. Molecular pharmacology 1968, 4 (6), 621-629. 58. Hendricks, S. P.; Mathews, C. K., Regulation of T4 Phage Aerobic Ribonucleotide Reductase SIMULTANEOUS ASSAY OF THE FOUR ACTIVITIES. Journal of Biological Chemistry 1997, 272 (5), 2861-2865. 59. Changeux, J.-P., Allostery and the Monod-Wyman-Changeux model after 50 years. Annual review of biophysics 2012, 41, 103-133. 60. Wegner, A.; Meiser, J.; Weindl, D.; Hiller, K., How metabolites modulate metabolic flux. Current opinion in biotechnology 2015, 34, 16-22. 61. Benesch, R.; Benesch, R. E., Some relations between structure and function in hemoglobin. Journal of Molecular Biology 1963, 6 (6), 498-505. 62. Gerhart, J. C.; Schachman, H. K., Allosteric interactions in aspartate transcarbamylase. II. Evidence for different conformational states of the protein in the presence and absence of specific ligands. Biochemistry 1968, 7 (2), 538-552. 63. Srinivasan, B.; Forouhar, F.; Shukla, A.; Sampangi, C.; Kulkarni, S.; Abashidze, M.; Seetharaman, J.; Lew, S.; Mao, L.; Acton, T. B., Allosteric regulation and substrate activation in cytosolic nucleotidase II from Legionella pneumophila. FEBS Journal 2014, 281 (6), 1613-1628. 64. Johansson, R.; Jonna, V. R.; Kumar, R.; Nayeri, N.; Lundin, D.; Sjöberg, B.-M.; Hofer, A.; Logan, D. T., Structural Mechanism of Allosteric Activity Regulation in a Ribonucleotide Reductase with Double ATP Cones. Structure 2016, 24 (6), 906-917. 65. Hofer, A.; Schmidt, P. P.; Gräslund, A.; Thelander, L., Cloning and characterization of the R1 and R2 subunits of ribonucleotide reductase from Trypanosoma brucei. Proceedings of the National Academy of Sciences 1997, 94 (13), 6959-6964. 66. Nowakowski, S. G.; Kolwicz, S. C.; Korte, F. S.; Luo, Z.; Robinson-Hamm, J. N.; Page, J. L.; Brozovich, F.; Weiss, R. S.; Tian, R.; Murry, C. E., Transgenic overexpression of ribonucleotide reductase improves cardiac performance. Proceedings of the National Academy of Sciences 2013, 110 (15), 6187-6192. 67. Kohen, A.; Limbach, H. H., Isotope Effects In Chemistry and Biology. Taylor & Francis: 2005. 68. Jonna, V. R.; Crona, M.; Rofougaran, R.; Lundin, D.; Johansson, S.; Brännström, K.; Sjöberg, B.-M.; Hofer, A., Diversity in Overall Activity Regulation of Ribonucleotide Reductase. Journal of Biological Chemistry 2015, jbc. M115. 649624. 69. Chimploy, K.; Mathews, C. K., Mouse Ribonucleotide Reductase Control INFLUENCE OF SUBSTRATE BINDING UPON INTERACTIONS WITH ALLOSTERIC EFFECTORS. Journal of Biological Chemistry 2001, 276 (10), 7093-7100. 70. Larsson, A.; Reichard, P., Enzymatic synthesis of deoxyribonucleotides X. Reduction of purine ribonucleotides; allosteric behavior and substrate specificity of the enzyme system from Escherichia coli B. Journal of Biological Chemistry 1966, 241 (11), 2540-2549. 71. Eriksson, S.; Thelander, L.; Akerman, M., Allosteric regulation of calf thymus ribonucleoside diphosphate reductase. Biochemistry 1979, 18 (14), 2948-2952. 72. Scott, C. P.; Kashlan, O. B.; Lear, J. D.; Cooperman, B. S., A quantitative model for allosteric control of purine reduction by murine ribonucleotide reductase. Biochemistry 2001, 40 (6), 1651-1661. 73. Xu, H.; Faber, C.; Uchiki, T.; Fairman, J. W.; Racca, J.; Dealwis, C., Structures of eukaryotic ribonucleotide reductase I provide insights into dNTP regulation. Proc Natl Acad Sci U S A 2006, 103 (11), 4022-7.

140

74. Ahmad, M. F.; Kaushal, P. S.; Wan, Q.; Wijerathna, S. R.; An, X.; Huang, M.; Dealwis, C. G., Role of arginine 293 and glutamine 288 in communication between catalytic and allosteric sites in yeast ribonucleotide reductase. Journal of molecular biology 2012, 419 (5), 315-329. 75. Ahmad, M. F.; Dealwis, C. G., The structural basis for the allosteric regulation of ribonucleotide reductase. Progress in molecular biology and translational science 2013, 117, 389-410. 76. Xu, H.; Faber, C.; Uchiki, T.; Racca, J.; Dealwis, C., Structures of eukaryotic ribonucleotide reductase I define gemcitabine diphosphate binding and subunit assembly. Proc Natl Acad Sci U S A 2006, 103 (11), 4028-33. 77. Kumar, D.; Viberg, J.; Nilsson, A. K.; Chabes, A., Highly mutagenic and severely imbalanced dNTP pools can escape detection by the S-phase checkpoint. Nucleic acids research 2010, gkq128. 78. Kumar, D.; Abdulovic, A. L.; Viberg, J.; Nilsson, A. K.; Kunkel, T. A.; Chabes, A., Mechanisms of mutagenesis in vivo due to imbalanced dNTP pools. Nucleic acids research 2011, 39 (4), 1360-1371. 79. Goodey, N. M.; Benkovic, S. J., Allosteric regulation and catalysis emerge via a common route. Nat Chem Biol 2008, 4 (8), 474-82. 80. Tsai, C. J.; Nussinov, R., A unified view of "how allostery works". PLoS computational biology 2014, 10 (2), e1003394. 81. Carlson, G. M.; Fenton, A. W., What Mutagenesis Can and Cannot Reveal About Allostery. Biophys J 2016, 110 (9), 1912-23. 82. Weinkam, P.; Chen, Y. C.; Pons, J.; Sali, A., Impact of mutations on the allosteric conformational equilibrium. J Mol Biol 2013, 425 (3), 647-61. 83. Kashlan, O. B.; Cooperman, B. S., Comprehensive model for allosteric regulation of mammalian ribonucleotide reductase: refinements and consequences. Biochemistry 2003, 42 (6), 1696-1706. 84. Sievers, F.; Wilm, A.; Dineen, D.; Gibson, T. J.; Karplus, K.; Li, W.; Lopez, R.; McWilliam, H.; Remmert, M.; Söding, J., Fast, scalable generation of high‐quality protein multiple sequence alignments using Clustal Omega. Molecular systems biology 2011, 7 (1). 85. Goujon, M.; McWilliam, H.; Li, W.; Valentin, F.; Squizzato, S.; Paern, J.; Lopez, R., A new bioinformatics analysis tools framework at EMBL–EBI. Nucleic acids research 2010, 38 (suppl 2), W695-W699. 86. Benson, D. A.; Clark, K.; Karsch-Mizrachi, I.; Lipman, D. J.; Ostell, J.; Sayers, E. W., GenBank. Nucleic acids research 2015, 43 (Database issue), D30. 87. Gupta, V.; Peterson, C. B.; Dice, L. T.; Uchiki, T.; Racca, J.; Guo, J. T.; Xu, Y.; Hettich, R.; Zhao, X.; Rothstein, R.; Dealwis, C. G., Sml1p is a dimer in solution: characterization of denaturation and renaturation of recombinant Sml1p. Biochemistry 2004, 43 (26), 8568-78. 88. Ahmad, M. F.; Dealwis, C. G., The structural basis for the allosteric regulation of ribonucleotide reductase. Progress in molecular biology and translational science 2013, 117, 389. 89. Zimanyi, C. M.; Ando, N.; Brignole, E. J.; Asturias, F. J.; Stubbe, J.; Drennan, C. L., Tangled up in knots: structures of inactivated forms of E. coli class Ia ribonucleotide reductase. Structure 2012, 20 (8), 1374-1383. 90. Bestwick, R.; Moffett, G.; Mathews, C., Selective expansion of mitochondrial nucleoside triphosphate pools in antimetabolite-treated HeLa cells. Journal of Biological Chemistry 1982, 257 (16), 9300-9304.

141

91. Wisitpitthaya, S.; Zhao, Y.; Long, M. J.; Li, M.; Fletcher, E. A.; Blessing, W. A.; Weiss, R. S.; Aye, Y., Cladribine and Fludarabine Nucleotides Induce Distinct Hexamers Defining a Common Mode of Reversible RNR Inhibition. ACS chemical biology 2016. 92. Larsson, K. M.; Jordan, A.; Eliasson, R.; Reichard, P.; Logan, D. T.; Nordlund, P., Structural mechanism of allosteric substrate specificity regulation in a ribonucleotide reductase. Nat Struct Mol Biol 2004, 11 (11), 1142-9. 93. Uhlin, U.; Eklund, H., Structure of ribonucleotide reductase protein R1. 1994. 94. Ingemarson, R.; Thelander, L., A kinetic study on the influence of nucleoside triphosphate effectors on subunit interaction in mouse ribonucleotide reductase. Biochemistry 1996, 35 (26), 8603-8609. 95. Ahmad, M. F.; Wan, Q.; Jha, S.; Motea, E.; Berdis, A.; Dealwis, C., Evaluating the therapeutic potential of a non-natural nucleotide that inhibits human ribonucleotide reductase. Mol Cancer Ther 2012, 11 (10), 2077-86. 96. Knappenberger, A. J.; Ahmad, M. F.; Viswanathan, R.; Dealwis, C. G.; Harris, M. E., Nucleoside Analogue Triphosphates Allosterically Regulate Human Ribonucleotide Reductase and Identify Chemical Determinants That Drive Substrate Specificity. Biochemistry 2016, 55 (41), 5884-5896. 97. Lundin, D.; Berggren, G.; Logan, D. T.; Sjöberg, B.-M., The origin and evolution of ribonucleotide reduction. Life 2015, 5 (1), 604-636. 98. Gunasekaran, K.; Nagarajaram, H.; Ramakrishnan, C.; Balaram, P., Stereochemical punctuation marks in protein structures: glycine and proline containing helix stop signals. Journal of molecular biology 1998, 275 (5), 917-932. 99. Gardner, M. J.; Hall, N.; Fung, E.; White, O.; Berriman, M.; Hyman, R. W.; Carlton, J. M.; Pain, A.; Nelson, K. E.; Bowman, S., Genome sequence of the human malaria parasite Plasmodium falciparum. Nature 2002, 419 (6906), 498-511. 100. Pessia, E.; Popa, A.; Mousset, S.; Rezvoy, C.; Duret, L.; Marais, G. A., Evidence for widespread GC-biased gene conversion in eukaryotes. Genome Biology and Evolution 2012, 4 (7), 787-794. 101. Davis, R.; Thelander, M.; Mann, G. J.; Behravan, G.; Soucy, F.; Beaulieu, P.; Lavallée, P.; Gräslund, A.; Thelander, L., Purification, characterization, and localization of subunit interaction area of recombinant mouse ribonucleotide reductase R1 subunit. Journal of Biological Chemistry 1994, 269 (37), 23171-23176. 102. Sardi, J.; Scorzoni, L.; Bernardi, T.; Fusco-Almeida, A.; Giannini, M. M., Candida species: current epidemiology, pathogenicity, biofilm formation, natural antifungal products and new therapeutic options. Journal of medical microbiology 2013, 62 (1), 10-24. 103. McCullough, M.; Ross, B.; Reade, P., Candida albicans: a review of its history, taxonomy, epidemiology, virulence attributes, and methods of strain differentiation. International journal of oral and maxillofacial surgery 1996, 25 (2), 136-144. 104. Wang, J.; Lohman, G. J.; Stubbe, J., Mechanism of inactivation of human ribonucleotide reductase with p53R2 by gemcitabine 5′-diphosphate. Biochemistry 2009, 48 (49), 11612-11621. 105. Landau, M.; Mayrose, I.; Rosenberg, Y.; Glaser, F.; Martz, E.; Pupko, T.; Ben-Tal, N., ConSurf 2005: the projection of evolutionary conservation scores of residues on protein structures. Nucleic acids research 2005, 33 (Web Server issue), W299-302. 106. Letunic, I.; Bork, P., Interactive tree of life (iTOL) v3: an online tool for the display and annotation of phylogenetic and other trees. Nucleic acids research 2016, 44 (W1), W242-5. 107. Koepf, E. K.; Petrassi, H. M.; Sudol, M.; Kelly, J. W., WW: An isolated three-stranded antiparallel [beta]-sheet domain that unfolds and refolds reversibly; evidence for a structured

142

hydrophobic cluster in urea and GdnHCl and a disordered thermal unfolded state. Protein Science 1999, 8 (04), 841-853. 108. Peltomäki, P., Deficient DNA mismatch repair: a common etiologic factor for colon cancer. Human Molecular Genetics 2001, 10 (7), 735-740. 109. Jiricny, J.; Nyström-Lahti, M., Mismatch repair defects in cancer. Current opinion in genetics & development 2000, 10 (2), 157-161. 110. Goodfellow, P. J.; Buttin, B. M.; Herzog, T. J.; Rader, J. S.; Gibb, R. K.; Swisher, E.; Look, K.; Walls, K. C.; Fan, M.-Y.; Mutch, D. G., Prevalence of defective DNA mismatch repair and MSH6 mutation in an unselected series of endometrial cancers. Proceedings of the National Academy of Sciences 2003, 100 (10), 5908-5913. 111. Song, H.; Ramus, S. J.; Quaye, L.; DiCioccio, R. A.; Tyrer, J.; Lomas, E.; Shadforth, D.; Hogdall, E.; Hogdall, C.; McGuire, V., Common variants in mismatch repair genes and risk of invasive ovarian cancer. Carcinogenesis 2006, 27 (11), 2235-2242.

143