Comparative Thermodynamics of Partial Agonist Binding to An

Home , Drug development, Willardiine

COMPARATIVE THERMODYNAMICS OF PARTIAL AGONIST BINDING TO AN

AMPA RECEPTOR BINDING DOMAIN

A Dissertation

Presented to the Faculty of the Graduate School

of Cornell University

In Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

Madeline Martínez

May 2015

Comparative Thermodynamics of Partial Agonist Binding to an AMPA Receptor Binding Domain

Madeline Martínez, Ph.D.

Cornell University 2015

AMPA receptors, ligand-gated cation channels endogenously activated by glutamate, mediate the majority of rapid excitatory synaptic transmission in the vertebrate central nervous system and are crucial for information processing within neural networks. Their involvement in a variety of neuropathologies and injured states makes them important therapeutic targets, and understanding the forces that drive the interactions between the ligands and protein binding sites is a key element in the development and optimization of potential drug candidates.

Calorimetric studies yield quantitative data on the thermodynamic forces that drive binding reactions. This information can help elucidate mechanisms of binding and characterize ligand- induced conformational changes, as well as reduce the potential unwanted effects in drug candidates. The studies presented here characterize the thermodynamics of binding of a series of

5’substituted willardiine analogues, partial agonists to the GluA2 LBD under several conditions.

The moiety substitutions (F, Cl, I, H, and NO2) explore a range of electronegativity, pKa, radii, and h-bonding capability. Competitive binding of these ligands to glutamate-bound GluA2 LBD at different pH conditions demonstrates the effects of charge in the enthalpic and entropic components of the Gibb’s free energy of binding. They suggest that the structure of agonists can be changed to improve the enthalpic energetic component to binding by building additional charge interactions in the portion of the molecule interacting with lobe 2.

The differences in heat capacity change between the analogue-bound and glutamate-bound states highlights the effect of charge and additional hydrogen bond formation on the temperature dependence of thermodynamic parameters, and how this dependence affects the driving forces of the competitive binding reactions at physiological temperatures.

Competitive binding experiments in multiple buffers yield the contributions of buffer ionization to the measured enthalpy of binding, as well as the differences in protonation states between the protein bound to glutamate, willardiine, and 5-substituted analogues. The exchange of glutamate for willardiine entails additional proton uptake into the LBD-ligand complex. The addition of more electronegative substituents into the 5-position eliminates the protonation differences between the glutamate and analogue-bound complexes. Bulkier substituents result in a ligand exchange where more protons are released into the buffer.

Excision of buffer ionization contribution to the observed enthalpic parameters revealed that there is no statistical difference between the enthalpic components of binding to glutamate and fluorowillardiine. While there is no buffer-independent significant differences between the enthalpic parameters of binding of willardiine and the remaining halogenated analogues, their differences in protonation events suggests their intrinsic enthalpies of binding might be different from each other.

This information allowed the estimation of thermodynamic parameters of binding at physiological temperature, 37oC. At this temperature the competitive binding of willardiine and the halogenated analogues is driven by favorable entropic change. The additional hydrogen bonding and electrostatic interactions possible with the nitro 5-substituent result in a competitive binding driven by favorable enthalpic change that occurs at significant entropic penalty.

Biographical Sketch

Madeline Martínez was born in San Juan, Puerto Rico in 1981. She obtained a Bachelor of

Science from the University of Puerto Rico, Río Piedras, in 2005 where she majored in chemistry. As an undergraduate she acquired research experience in biochemistry and electrophysiology.

Her research projects in the University of Puerto Rico, under the supervision of Prof. Kai

Griebenow focused on the stabilization and preservation of protein structure and activity with the goal of increasing the range of available protein pharmaceuticals. She studied the effects of covalent modifications and exposure to solvents and excipients used in drug encapsulation on the enzymatic and structural properties of model enzymes through spectrophotometric techniques such as CD and FTIR spectroscopy.

As part of the Leadership Alliance Summer Research-Early Identification Program (SR-EIP) she conducted research under the supervision of Prof. George Hess at Cornell University. She measured the effects of cocaine exposure on glutamate-elicited whole cell currents in kainate receptors expressed in HEK 293 cells. During this time she became fascinated with neurotransmitter receptors and signaling.

In 2005 Madeline Martínez was admitted to Cornell University to pursue her PhD in

Pharmacology, where she joined the laboratory of Prof. Robert Oswald, and has focused on understanding the dynamics and thermodynamics of glutamate receptors.

Dedication

This work is dedicated to my spouse Jaime J. Benítez,

my parents Arnaldo and Madeline,

my sisters Ileana and Tania.

Everything is possible with your support and love.

Acknowledgements

First of all I thank my, complex, extended, and wonderful family. My parents, grandparents, stepmother Gladys, aunts and uncles who, along with my parents, ensured I had everything I needed to develop and learn. I thank Grandmother Elba who tutored and guided my studies and upbringing through all my school years, and Grandmother Olga who showed me firsthand how scientific knowledge can save lives and small kindnesses can transform communities.

I thank my advisor Dr. Robert Oswald for giving me the opportunity to perform my doctoral research in his laboratory. I am grateful for his mentorship, advice, and for encouraging me to persevere when the systems under study seemed incompatible with the techniques used to study them. I would also like to acknowledge all the members of the Oswald group, past and present, for all our conversations, help, and camaraderie. I’m especially thankful for the help and collaboration of Dr. Ahmed H. Ahmed, whose crystallographic and NMR studies of willardiine protonation and deprotonation allowed a greater understanding of their binding thermodynamics.

I would also like to express thanks to the members of my committee, Dr. Linda Nicholson,

Dr. Manfred Lindau, and Dr. Gregory Weiland, who provided me with advice, ideas, and guidance throughout this endeavor.

I am very thankful for the help, advice, and encouragement I have received from Dr. Linda

Nowak and Susan Coombs.

I am also grateful of Dr. Adrienne Loh for allowing me to use her calorimeter for such a long period of time, as well as for all her help and ideas. Few things are as pretty as the thermogram baselines in her calorimeter!

vii

Table of Contents

Biographical Sketch ...... v

Dedication ...... vi

Acknowledgements ...... vii

Table of Contents ...... viii

List of Figures ...... xi

List of Tables ...... xiii

List of Abbreviations ...... xiv

Chapter 1 Glutamate Receptors and Calorimetry ...... 1

L-Glutamate ...... 4 Glutamate Receptor Classification ...... 5 Ionotropic Glutamate Receptor Subtypes ...... 6 AMPA Receptors ...... 6 Kainate Receptors ...... 11 NMDA Receptors ...... 14 Delta Receptors ...... 18 AMPA Receptor Structure and Dynamics ...... 20 Receptor Structure ...... 20 Structure and Dynamics of the Ligand Binding Domain ...... 25 AMPA Receptor Roles in Injured States and CNS Diseases ...... 28 Amyotrophic Lateral Sclerosis...... 28 Alzheimer's Disease ...... 30 Epilepsy...... 31 Ischemia and Excitotoxicity ...... 32 Willardiine Analogues and Partial Agonism ...... 32 viii

Ligand Binding Thermodynamics ...... 35 Outline of Dissertation Research ...... 37 References ...... 39 Chapter 2 Thermodynamics and Mechanism of the Interaction of Willardiine Partial Agonists with a

Glutamate Receptor: Implications for Drug Development ...... 76

Introduction ...... 77 Methods ...... 79 Protein Purification ...... 79 Isothermal Titration Calorimetry (ITC) ...... 80 Implications of Competition Binding for Thermodynamic Parameters ...... 80 Crystallography ...... 81 Results ...... 82 Discussion ...... 89 Supplementary Material ...... 91 References ...... 93 Chapter 3 Determination of Ionization Effects and Protonation Events of Willardiine Partial Agonists with a Glutamate Receptor ...... 97

Introduction ...... 99 Methods ...... 101 Protein Purification ...... 101 Determination of the Extinction Coefficient of GluA2 S1S2 by Thiol Quantitation ...... 102 Isothermal Titration Calorimetry (ITC) ...... 103 Data Analysis ...... 104 Results and Discussion ...... 110 Protonation Events and Buffer Ionization Effects at 20oC ...... 114 Estimates of Binding Parameters at Physiological Temperatures ...... 117 References ...... 124 Chapter 4 Concluding Remarks and Future Outlook ...... 128

Research Summary ...... 129 Further Research ...... 131

References ...... 132

List of Figures

Figure 1.1 Subunit arrangement of AMPA receptors ...... 8

Figure 1.2 Subunit assembly of AMPA receptors...... 21

Figure 1.3. Crystal structure of the AMPAR GluA2 LBD bound to glutamate...... 26

Figure 1.4. Willardiine structure with a 5-X substitution...... 33

Figure 2.1 Thermograms showing raw (top) and integrated (bottom) data for NW (A) and IW (B) displacement of glutamate from the GluA2 LBD at 20 °C. The molar ratio is the ratio of ligand to protein.

Fits were performed as described by Sigurskjold et al.21 (C) Calculated ΔΔH, −TΔΔS°, and ΔΔG values for displacement of glutamate by the five willardiine derivatives. (D) Structures of the willardiine derivatives...... 83

(left) and HW (right) at 10 °C. For both compounds, the pH at which the ionized state is favored is shown on the left and that for which the un-ionized state is favored is shown on the right. (C) Structure of the binding site, showing the direct interactions between NW and the GluA2 LBD [Protein Data Bank (PDB) entry 3RTW32]. (D) Structure of the binding site for NW showing the hydrogen bonding network associated with the nitrogen at position 3 of the willardiine ring in the charged form (pH 6.5, PDB entry

3RTW32). On the left is the structure and on the right a schematic. In the schematic, NW is colored blue, the protein black, and water green and the H-bonds are colored red. (E) Structure of NW bound to the

GluA2 LBD obtained at pH 3.5 (PDB entry 4Q30). At this pH, the ring is largely uncharged. The formatting of this panel is similar to that of panel D. Note the change in the H-bonding network. The potential H-bond between the hydroxyl of S654 and the uracil carbonyl is shown with a question mark because the distance between the two oxygens is relatively long, 3.5 Å, and the angle between this H- bond and the H-bond with the water is not favorable (80°)...... 84

Figure 2.3 Dependence of ΔΔH on temperature...... 88

Figure 2.4 Thermograms showing raw (top) and integrated (bottom) data for displacement of glutamate ...... 91

Figure 3.1 Thermogram and binding isotherm for the displacement binding of (A) NW and (B) CW at

20oC, pH 7.2 in cacodylate buffer ...... 112

o Figure 3.4. Effects of buffer protonation on ΔΔH values at pH 7.2 and 20 C. ΔΔHobs (blue) and ΔΔHbind

(red) of competitive binding to willardiine analogues...... 117

Figure 3.5. Thermodynamic parameters of competitive binding of willardiine analogues as a function of temperature in phosphate buffer, pH 7.2, for: (A) HW, (B) NW, (C) FW, (D) CW, and (E) IW. Blue

(dashed) and red (solid) lines represent fits to the data. Data was fitted with equation 33 for ∆∆H and 34 for ∆∆S. From the fit, values for ∆∆H0 and ∆∆S0 were obtained. ∆∆H0 (kcal/mol): HW, -2.9 ± 0.1; NW, -

-2 17.4 ± 0.6; FW, -0.20 ± 0.03; CW, 1.2 ± 0.3; IW, 2.6 ± 0.3. ∆∆S0 (kcal/(mol K)): HW, (2.0 ± 0.2) x 10 ;

NW, (-3.0 ± 0.2) x 10-2; FW, (3.00 ± 0.02) x 10-2; CW, (3.0 ± 0.2) x 10-2; IW, (3.0 ± 0.1) x 10-2...... 118

Figure 3.6. Thermodynamic parameters of the competitive binding of willardiine analogues determined at

20oC (A) and estimated at 37oC (B) in phosphate buffer...... 121

xii

List of Tables

Table 2.1 Structural statistics for the crystal structure of NW bound to GluA2 LBD at pH 3...... 93

Table 3.1 Thermodynamic parameters of competitive binding for willardiine analogues at pH 7.2, 20oC in cacodylate buffer...... 111

Table 3.2. The observed enthalpy changes, ΔΔHobs, the change of enthalpy of competition binding ,

ΔΔHbind, and the number of protons exchanged with the buffering system, nH of the competition binding of willardiine analogues against glutamate...... 115

Table 3.3. Heat capacity change for displacement binding of substituted willardiines to GluA2 LBD ... 119

Table 3.4 Estimates of the observed thermodynamic parameters of competitive binding at 37oC ...... 121

Table 3.5. Estimates of the differences in enthalpic change upon competitive binding at 37oC...... 122

xiii

List of Abbreviations

1. AMPA: α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid 15, 18

2. AMPAR: AMPA receptors 20

3. ATD: Amino Terminal Domain 20

4. bp: base pair 22

5. Ca2+: divalent calcium ion 19

6. cDNA: complementary DNA 15

7. CNS: Central nervous system 15

8. CTD: Carboxy Terminal Domain 20

9. DAP: diaminopimelic acid 18

10. EPSC: AMPAR-mediated excitatory post-synaptic current 20

11. iGluRs: ionotropic glutamate receptors 15

12. K+ :monovalent potassium ion 19

13. KARs: Kainate receptors 23

14. LBD: Ligand binding domain 15, 20

15. LIVBP: leucine, isoleucine, valine binding protein 35

16. LTD: long-term depression 23

17. LTP: long-term potentiation 23

18. mGluRs: metabotropic glutamate receptors 19

19. Na+ : monovalent sodium ion 19

20. Neto 2: Neuropilin Tolloid-like 2 24

21. Neto1: Neuropilin Tolloid-like 1 24

22. NMDA: N-methyl D-aspartate 15, 18

xiv

23. NMR: Nuclear magnetic resonance 16

24. Q: Glutamine 23

25. R: Arginine 23

26. TARP: trans-membrane AMPA receptor regulatory proteins 22

27. TMD: Trans-Membrane Domain 20

Chapter 1

Glutamate Receptors and Calorimetry

L-glutamate is the major excitatory neurotransmitter in the central nervous system (CNS).

During neurotransmission, it is released in the synaptic cleft, inducing its effects by interacting with glutamate receptors in the post-synaptic density. These receptors can be metabotropic and activate g proteins, or ionotropic, forming integral ion channels. When ionotropic receptors

(iGluRs) are activated by glutamate they undergo conformational changes that trigger the opening of the ion channel, allowing cations like sodium and calcium into the cell. iGluRs are crucial to the development and homeostasis of the nervous system, and their dysfunction is implicated in many pathologies and injured states1–4.

Ionotropic glutamate receptors are the main mediators of excitatory synaptic transmission in the CNS and are classified into three families: those that are activated by N-methyl D-aspartate

(NMDA), α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), and kainate5–9.

While having similar topologies and sequence identity, their structure, function, and ligand affinity vary across subtypes10–12. NMDA receptors have coincidence detection functions, monitoring changes in membrane potential and the presence of glutamate in the synaptic cleft.

Kainate receptors modulate postsynaptic and presynaptic sites, while AMPA receptors mediate the majority of rapid excitatory synaptic current. Plenty of effort has been expended in understanding how these diverse functional characteristics arise from the distinct glutamate receptor structures. A crucial step was achieved in 1989, when Michael Hollmann, Stephen

Heinemann, and Peter Seeburg cloned individual mammalian iGluR cDNA, leading to the functional mapping of iGluRs. This allowed the structural characterization of isolated domains of various subtypes of iGluRs13–17, and later on to the X-ray diffraction analysis of an intact AMPA receptor18. Special attention was given to the isolated ligand binding domain (LBD) with the goal of understanding how it’s motion is translated into the opening and closing of the ion channel. 2

LBDs have been studied through biophysical techniques, such as nuclear magnetic resonance

(NMR), analytical ultracentrifugation, spectroscopy, optogenetics and calorimetry. These studies, performed in the apo state and in complexes with a variety of agonists, partial agonists, and antagonists, have helped elucidate key aspects of the function and dynamics of iGluR ligand binding14,15,19.

Calorimetric studies provide quantitative data on the driving thermodynamic forces of binding reactions. Understanding these forces is a key element in the optimization of drugs binding to their target receptor, and has been used to reduce unwanted effects of drug candidates20. These thermodynamic studies have played important roles in the elucidation of mechanisms of binding and the characterization of ligand-induced conformational changes.

The thermodynamic properties of the binding of glutamate to an isolated AMPAR LBD have been previously studied21. At 15oC the binding of glutamate to GluA4 LBD is an exothermic process, driven by favorable enthalpic and entropic forces. At this temperature the magnitude of the enthalpic component of the free energy is lower than that of the entropic contribution. The heat capacity of this binding reaction is both negative and larger than the entropic change at

15oC. Due to this property, the binding reaction of glutamate to the LBD of GluA4 at physiological temperatures is driven primarily by favorable enthalpy.

Competitive binding studies using a glutamate-bound GluA2 LBD construct has shed light on the thermodynamics of binding of partial agonists and antagonists. These showed that the displacement binding of the partial agonist Iodowillardiine at 15oC is an enthalpically unfavorable endothermic process driven by favorable entropy. At the same time, the competitive

3 binding of several antagonists was an exothermic process with favorable enthalpic contributions and unfavorable entropic contributions to the change in free energy of binding, ΔΔG.

L-Glutamate

L-glutamate is the main excitatory neurotransmitter in the brain, yet elucidating its role in neurotransmission was complicated with its role in intermediary metabolism. During the 1930s, studies recognized large concentrations of L-glutamate in the brain and speculated about the neurophysiological role of the amino acid. This led to a decade of studies and trials characterizing dietary glutamate and glutamine and their effects on learning disorders and epilepsy22. Meanwhile, the main role of glutamate was considered to be as an intermediary in energy metabolism.

Glutamate’s role in electrophysiology and neurotransmission was considered after observing that injections of L-glutamate into the brain or carotid arteries resulted in convulsions23. Later in

1960, the first observation of the excitatory action of L-glutamate in single cells was made24.

Evidence for a membrane-bound receptor for glutamate began to accrue with the study of the apparent nonstereoselectivity of the action of glutamate and aspartate. Bulkier analogues were used in structure-activity studies to make potential stereoselectivity more apparent. Among these analogues was NMDA, which had a higher potency than glutamate. This established that the sites of action of these two amino acids could exhibit stereoselectivity, and implied that there was a discrete membrane receptor site. Other substances discovered in this decade, like kainic acid, quisqualic acid, and the synthetic amino acid AMPA also had higher potencies than L- glutamate22.

The cellular uptake of L-glutamate and these substances was characterized in order to differentiate between it the phenomena of excitatory action. They showed different structure- 4 activity relationships, excluding the possibility that the excitatory effect was an indirect consequence of glutamate uptake. At the same time, uptake was established as a credible mechanism for the clearance of excitatory amino acids, providing a mechanism for the termination of transmitter action25.

At the same time, several laboratories were able to determine that electrical stimulation or potassium induced depolarization resulted in the calcium-dependent release of L-glutamate and

L-aspartate from synaptic terminals. Along with the establishment of the glutamate-glutamine cycle as the method of synthesis of L-glutamate in the synaptic terminal, these studies provided an initial platform for the role of glutamate in synaptic transmission26.

Glutamate Receptor Classification

The idea of multiple glutamate receptors became compelling after a series of studies found differences in the relative potency of various compounds across neuronal tissues27–29. In addition, various research groups determined that magnesium ions and compounds like HA-966 and diaminopimelic acid (DAP) reduced the responses of excitatory amino acid analogues like

NMDA, but not those of kainate or quisqualate30. Later on, C-fibers in dorsal roots were found to respond to kainate but not to NMDA or quisqualate31. Taken together, these results led to the initial classification of these receptors into NMDA, kainate, and quisqualate receptors.

We now classify ionotropic glutamate receptors into those sensitive to N-methyl D-aspartate

(NMDA; NR1, NR2A-D, NR3A-B), α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid

(AMPA; GluA1-4), and kainate(GluK5-7, KA 1,2)5–9. Each subtype has several molecular species, which were identified through the use of molecular cloning6–8.

Two additional subunits of iGluRs, the delta subunits (δ1 and δ2), have been identified through the use of homology screening. They have yet to show any ligand-gated ion channel function, and it is unclear if these subunits form ligand-gated ion channels or exert their physiological functions through protein-protein interactions32–35.

Even after the successful identification of iGluRs there were reports of receptor-mediated biochemical effects, induced by excitatory amino acids, not being antagonized by known glutamate receptor antagonists. In addition, certain compounds with glutamate-like structures could cause excitation without being susceptible to iGluR antagonists or depress synaptic excitation without having iGluR antagonist activity36. This activity was the result of another family of glutamate receptors. These receptors, called metabotropic glutamate receptors

(mGluRs) cause slower synaptic effects and are associated with biochemical changes through their interactions with G proteins. mGluRs are comprised of three groups: Group I (mGlu1, mGlu5), Group II (mGlu2, mGlu3), and Group III (mGlu4, mGlu6, mGlu7, mGlu8). Group I mGluR receptors are generally associated with excitatory synaptic responses, while groups II and

III are associated with depression of synaptic responses, via inhibition of glutamate release6,37,38.

Ionotropic Glutamate Receptor Subtypes

AMPA Receptors AMPA receptors bind α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) with high affinity and potency5. Activation of these receptors causes a rapid opening of channels that are permeable to Na+ , K+, and in certain instances Ca2+. These channels are responsible for most of the fast excitatory transmission in the vertebrate central nervous system. Their activation

6 produces a transient inward current that rises in a few hundred milliseconds and decays within a few milliseconds39,40.

The fast kinetics of AMPAR-mediated excitatory post-synaptic currents (EPSCs) provide a precise window for coincidence detection and can generate action potentials with a high degree of precision. Since information processing by neural networks requires precise timing and coincidence detection, AMPA receptors are a crucial part of these networks41–43.

In mammals, four genes are responsible for expressing AMPA subunits (GRIA1-4). The subunits expressed consist of GluA1-GluA4. While all iGluR subunits share 20-40% homology and a similar structure, the genes that encode for AMPARs have 70% homology 44.

AMPA receptors assemble as homomeric and heterometric tetramers. Their topology was determined through antibody targeting and domain mapping of glycosylation and phosphorylation sites, and later confirmed through high-resolution structural analysis6,19,45–47. It consists of a modular structure with four different domains, three of which are homologous to bacterial proteins46. Their subunits consist of an extracellular amino terminal domain (ATD), an extracellular ligand-binding domain (LBD), a transmembrane domain (TMD)consisting of three transmembrane-spanning regions (M1,M3-M4) and a reentrant pore loop (M2), and an intracellular carboxy terminal domain (CTD)(Figure 1.1).

Figure 1.1 Subunit arrangement of AMPA receptors

AMPARs are widely expressed in neurons and glia. They are generally located in the postsynaptic density of excitatory neurons, but they can also be present in presynaptic sites as well as in non-neuronal cell types48–50. Within the CNS, their subunit expression exhibits regional, developmental and cell-specific variation, which have profound effects on AMPAR function44,51,52.

The GluA2 subunit plays a key role in the determination of mammalian AMPAR function. It is expressed in the majority of AMPAR heteromers, and modulates receptor characteristics such as Ca2+ permeability, binding and desensitization kinetics, polyamine block, and single channel conductance. Within the adult hippocampus and cortex most receptors consist of GluA1/GluA2 and GluA2/GluA3 subunits. However, the low expression of the GluA3 subunit ensures that the majority of GluA2 associates with the GluA1 subunit53,54. The GluA4 subunit is expressed in the

8 immature hippocampus, and in other mature regions, where it co-assembles with GluA2 subunits to form functional receptors55.

The subunit composition of AMPARs determines receptor activation, desensitization kinetics, and other functional properties. Their functional diversity is amplified by the structural and functional differences arising from posttranscriptional and posttranslational modifications, co-assembly with trans-membrane AMPA receptor regulatory proteins (TARPs), and polyamine blocks44,56–60.

After transcription, subunit pre-mRNA is modified by splicing and editing, resulting in multiple subunit isoforms with their own structural and functional properties. AMPAR pre- mRNA is spliced in two distinct sites, preceding and following the last transmembrane region,

M4. The resulting splice variants have marked functional differences.

AMPAR pre-mRNA has an alternative splice that results in “flip” or “flop” isoforms. The flip/flop sequence consists of a 115bp region that encodes a 38 amino acid sequence. This sequence is located in the extracellular LBD, and precedes the final transmembrane, or M4, domain. Flip/flop isomers exhibit different expression patterns, channel kinetics and pharmacological profiles. The flip isoforms is expressed abundantly during early development, while the flop isoforms are expressed in low abundance early in development and become upregulated in adult animals61.

The pre-mRNA of GluA2 and GluA 4 subunits has an additional alternative splice in the C- terminal coding region. Differential splicing between exons 16 and 17 allow these subunits to express short and long C-terminal tails62. These C-terminal tails contain residues that can be

9 posttranslationally modified as well as sequence segments involved in protein-protein interactions. Since these mechanisms modulate receptor function and localization, this splicing plays an important role in receptor regulation63,64.

Pre-mRNA editing of AMPAR subunits also gives rise to functionally unique variants. The

Ca2+ permeability of GluA2-containing receptors is controlled by RNA editing of the Q/R site within the M2 region. During editing, an adenosine base within the codon of glutamine (Q) 607 undergoes enzymatic oxidative deamination, producing an inosine on the site. The codon is later translated into arginine (R) by the ribosomal machinery59,62,65. Most GluA2 subunits in adult mammal brains are Q/R edited. When incorporated, this subunit decreases the Ca2+ permeability, lower the single channel conductance, and abolishes the polyamine block of the receptors66.

Lower Q/R editing is observed during early development, in certain neurons, and in glial cells67–

69.

All these subunit variants undergo posttranslational modifications like phosphorylation, palmitoylation, and glycosylation. These modifications have profound effects on receptor trafficking, localization, and function 70–73. They also play important roles in the expression of long term potentiation (LTP) and long term depression (LTD)63,74–76.

During receptor assembly, AMPAR subunits assemble stoichiometrically with auxiliary proteins from the TARP family77. Assembly with these proteins has a profound effect on receptor function, trafficking, localization, and pharmacology77–81.

Kainate Receptors

Kainate receptors (KARs) have a strong preference for this agonist over AMPA and NMDA.

They play key roles in excitatory and inhibitory transmission, and are involved in the maturation of neural circuits during development. In addition, kainate receptors modulate the synaptic release of neurotransmitters like GABA and glutamate in different locations.

There are five genes encoding for kainate receptor subunits. These subunits are subdivided into two classes. GluK1, GluK2, GluK3 are part of the first class while GluK4 and GluK5 form the second4. These two classes have marked differences in primary structure and functional characteristics. Kainate receptors are thought to assemble with a similar transmembrane topology, stoichiometry, and subunit assembly as AMPA receptors. Similarly to AMPARs, the pre-mRNA of GluK1 and GluK2 can be edited in the Q/R site in the M2 reentrant loop region. In addition, the pre-mRNA of GluK1-Gluk3 subunits undergoes alternative splicing 82.

While GluK1-3 can be assembled as homomeric or heteromeric receptors, GluK4 and GluK5 are obligate heteromers and must co-assemble with GluK1-3 subunits83–87. The channels containing GluK4 and GluK5 have different properties, like a higher affinity for kainate, than

GluK1-3 homomeric channels36. While kainate receptors are widely expressed within the nervous system, each subunit has particular expression patterns4,88,89. Kainate receptors mediate postsynaptic depolarization, and can carry some of the synaptic current at some synapses. These kainate-mediated excitatory postsynaptic currents are slower and smaller than AMPAR-mediated ones90–92. The characteristics of these currents are partly due to their co-assembly with auxiliary subunits93–96.

These auxiliary subunits are integral membrane proteins known as Neuropilin Tolloid-like 1 and 2 (Neto1 and Neto2)97. Neto proteins radically alter the gating properties of KARs, and increase their equilibrium agonist affinity. When kainate receptors co-express with Neto1 and

Neto2, the onset of desensitization of kainate evoked responses is decelerated while the recovery from desensitization accelerates94,98,99 .

Unlike AMPARs and NMDA receptors, gating kainate channels requires external monovalent cations100,101. While all KAR subunits contain an ion binding pocket for these ions, at physiological salt concentrations a large fraction of KARs have an unoccupied cation binding site making them unable to activate by released glutamate102,103.

Another unique feature of kainate receptors is their non-canonical signaling through metabotropic mechanisms. This mechanism is independent of ion flux, and involves the activation of G proteins104–106. There is evidence of KAR non-canonical signaling in many cell types in different regions of the CNS, where they play key roles in the presynaptic control of neurotransmitter release and in the regulation of neuronal excitability107,108. This dual signaling, through ionotropic and metabotropic mechanisms, is responsible for many of the diverse actions of KARs.

Kainate receptors can be found in presynaptic and postsynaptic sites. While postsynaptic

KAR-mediated EPSCs have only been found in a few central synapses, their actions help regulate neuronal excitability and synaptic transmission82. KAR-mediated EPSCs are characterized by slow activation and deactivation kinetics109. In neurons that control the generation of network oscillations, their slower kinetics of can enable the generation of oscillations with different dominating frequencies than those mediated by AMPARs110.

Postsynaptic KARs can also modulate neuronal excitability through non-canonical signaling111. Postsynaptic KARs can regulate the actions of afterhyperpolarization currents by engaging Gi/o proteins and activation PKC112. This increases neuronal firing frequency, and results in circuit excitability113,114. While the ionotropic and metabotropic actions of kainate receptors can coexist in the same synapses, their actions are independent of each other106,114–116.

Presynaptic and postsynaptic kainate receptors coexist, but have different pharmacological profiles and subunit compositions. They also modulate neuronal activity through different signaling pathways117–120. When kainate receptors are located in presynaptic sites they are involved in the modulation of neurotransmitter release at excitatory and inhibitory synapses107.

Within inhibitory synapses KARs can both facilitate and depress GABA release91,104. Their modulatory action appears to occur through non-canonical signaling rather than through ion flux107.

Presynaptic KARs can also regulate excitatory transmission, and have the ability to both facilitate and suppress glutamate release. Activation of presynaptic kainate receptors in mossy fiber to CA3 synapses increases the depolarization of presynaptic terminals, facilitating glutamate release121. This augments the Ca2+ influx from action potentials122–124.

At the same time, long and strong trains of activity, as well as exogenous application of high concentrations of kainate, can depress synaptic transmission by inhibiting glutamate release91,121,125–127. This inhibition is mediated via non-canonical metabotropic signaling128,129.

The suppression of glutamate-release by KARs has a significant influence on the shaping of synaptic transmission during development130 .

Besides their functional roles, kainate receptors are involved in a variety of brain diseases, and thus are targets for drug action. SNPs in genes encoding for kainate subunits have been linked to schizoid and bipolar disorders131,132. In addition, kainate receptors are involved in several aspects of epilepsy.

Kainate is an epileptogenic compound. Its application to neuronal slices and injection to animal models reduces GABA release and can activate postsynaptic KARs. This process is involved in the generation of temporal lobe epilepsy and epileptic activity in the hippocampus104,105 After seizures, aberrant synapses containing KARs are formed, and their excitatory input contributes to the pathogenesis of epilepsy133–135. Evidence suggests that kainate receptor antagonists and agonists, in a subunit-dependent manner, can prevent the development of epileptic activity136–138.

Due to their roles in epilepsy and other pathologies, kainate receptors are important drug targets. The development of subunit-specific agonists and antagonists may result in drug candidates with anti-epileptic effects.

NMDA Receptors

NMDA receptors are obligate heterotetrameric cation channels, and their role as mediators of brain plasticity makes them indispensable for CNS function. Their dysfunction is involved in a host of neurological and psychiatric disorders such as schizophrenia, neurodegenerative diseases, pathological pain, and stroke139–141.

Seven genes are responsible for encoding NMDA receptor subunits. Subunits are classified as GluN1, GluN2 (A-D), and GluN3 (A,B)139. The number of amino acids of these subunits

14 varies from 900 to over 1,480. Most of the differences that result in the subunit length variation stem from differences in the C-terminal domains.

The structural and functional diversity of these receptors is increased with alternative splicing. The gene encoding the GluN1 subunit has eight isoforms GluN1-1a ̶ GluN1-4a and

GluN1-1b ̶ GluN1-4b). Incorporation of different GluN1 isoforms into the receptors changes the gating and pharmacological properties of the receptor142,143. Splicing also results in isoforms with different CTD lengths and trafficking properties144. GluN1 isoforms are differentially expressed in brain regions; however the functional significance of these patterns remains unclear 145.

After translation, NMDA receptor subunits can undergo modifications like phosphorylation, nitrosylation, ubiquitinization, and calpain cleavage. The single channel conductance can be modulated by the phosphorylation of serine or tyrosine residues in the c-terminal domain of the subunits146,147. Glycosylation of the extracellular N-terminal domain of the subunits can affect receptor assembly and trafficking73. NR2A subunits can undergo S-nitrosylation of cysteine residues, resulting in a downregulation of NMDAR activity and a reduction of Ca2+ entry into the cell148,149. The calcium-dependent protease calpain modulates NMDAR function and turnover by cleaving within the c-terminal domain of NR2 subunits, and is used, along with ubiquitination as a mechanism for degrading and removing the receptors from the cell surface150,151.

The GluN1 subunits are ubiquitously expressed across the CNS from embryonic stages throughout adulthood, but there are differences in the regions where they are expressed and in their presence and abundance across developmental stages152–155. GluN2 subunit composition provides NMDARs with an array of functional variety. The four different GluN2 subunits have large spatio-temporal differences in their expression profile154–156. For example, the GluN2

15 subunits present in the embryonic brain are mostly GluN2B and GluN2D, although GluN2D is mostly expressed in the caudal regions. The expression of GluN2A starts shortly after birth and continues to rise until it is ubiquitously expressed in the CNS. During this period, the expression of GluN2D drops sharply and localizes to the encephalon and mesoencephalon. While the high expression levels of the GluN2B subunits are maintained after birth, their expression drops sharply after the first postnatal week and expression is restricted to the forebrain. Lastly, expression of GluN2C subunits occurs late in development and restricted, to a large extent, to the cerebellum and olfactory bulb. GluN3A and GluN3B show similar differences in their spatiotemporal expression patterns. The expression levels of GluN3A peak in early postnatal life then decrease over time while the expression levels of GluN3B subunits increases throughout development, especially in motor neurons.

The GluN2A and GluN2B subunits are the predominant GluN2 subunits in the adult CNS.

Their expression in brain structures such as the cortex and hippocampus is indicative of their role in synaptic function and plasticity153–155. The expression of GluN2B, GluN2D, and GluN3A subunits occurs primarily early in development, and highlights their importance in synaptogenesis and synaptic maturation157,158.

The overlapping expression of many of NMDAR subunits gives the CNS a wide array of coexisting NMDA receptors. The subunit composition determines the biophysical and functional properties of the receptors155,159,160. Receptors can assemble with identical or different GluN1 isoforms and their non-GluN1 subunits can be identical or differ as well156. The resulting

NMDARs can be di-heteromeric or tri-heteromeric, with GluN1/GluN2B, GluN1/GluN2A,

GluN1/GlN2A /GluN2B representing a significant fraction of the adult NMDARs161–163. GluN3

16 subunits are thought to participate in tri-heteromeric GluN1/GluN2/GluN3 assemblies in vivo , while di-heteromeric GluN1/GluN3 receptors have also been shown to generate glycine- activated excitatory currents157,158,164,165.

NMDA receptors have similar transmembrane topology and stoichiometry to that of AMPA receptors. However, N-terminal domains of NMDARs have a unique twisted clamshell conformation that results in looser NTD dimer assemblies than those seen in other iGluRs166.

NMDAR NTDs are dynamic and can bind a host of subtype-specific allosteric modulators, while their structural rearrangement can be sensed by the downstream gating machinery141,166,167. These modulatory sites in the NTD make NMDARs very sensitive to extracellular microenvironments, and are key elements in the generation of the functional and pharmacological diversity of the receptors.

The basic gating properties of the iGluRs, involving the activation of the receptor through the agonist-induced closure of the ligand-binding domains, are conserved in NMDARs. However, activation of NMDARs requires the binding of glutamate in GluN2 subunits and of glycine (or

D-serine) in GluN1 and GluN3 subunits. In addition, NMDAR activation is dependent on the relief of a voltage-dependent block caused by extracellular Mg2+ 168. At resting membrane potential Mg2+ ions block NMDAR channel activation. Relief of this blockade occurs upon membrane depolarization and is usually mediated by AMPARs.

The opening of NMDAR ion channels allows the entry of divalent ions, including Ca2+ (7%-

18% of the inward current), into the cell, which leads to an excitatory postsynaptic potential146,169–171. The slower unbinding of glutamate from the NMDAR binding site results in channels with slower kinetics than that of AMPARs. These kinetics, as well as calcium

17 permeability, extent of magnesium blockade, and channel conductance of the receptors are modulated by subunit composition2,170,172.

Delta Receptors

Delta receptors consist of two subunits GluD1 (δ1) and GluD2 (δ2), which were cloned by homology screening 32–34. Their sequence homology placed them within the iGluR family even though no endogenous ligands and ion channel function had been observed 146. Delta receptors play important roles in high-frequency hearing and cerebellar function, and can serve structural functions at synapses.

The GluD1 and GluD2 subunits are encoded by grid1and grid2 genes, respectively. These genes are much larger than the genes encoding other iGluRs, although the sizes of the cDNA are similar. The polypeptide chains expressed by these genes share a sequence similarity of 76% with each other, and ~40% with the rest of the iGluRs33,34.

GluD1 receptors are highly expressed in hair cells of the auditory and vestibular systems, and gene deletion leads to deficits in high-frequency hearing. In addition, GluD1 mRNA is present within the hippocampus in pyramidal and dentate granule cell layers173.

GluD2 receptors are highly expressed in the post-synaptic sites of parallel fibre (PF)-Purkinje cell synapses, and they are expressed at low levels in places like the midbrain-spinal cord and interneurons in the cerebellum 33,34,95,174. Within Purkinje cells, GluD2 channels associate with transient receptor potential cation channel TRPC3 and mGluR receptors, and their ion channels can be triggered open by mGluR1 activity175.

While most neurons don’t coexpress GluD1 and GluD2, they do coexpress with other iGluRs. Delta subunits can coassemble with AMPAR and Kainate receptor subunits in vitro, which change their channel properties176,177. However, co-precipitation studies suggest that this is less likely in vivo178.

Sequence similarity, computational studies, and accessibility studies suggest that their membrane topology is similar to other iGluRs: extracellular N-terminal domains and ligand binding domain, three transmembrane domains and a reentrant loop segment that form the ion channel, and an intracellular C-terminal domain179.

The NTDs of delta receptors, similarly to that of other iGluRs, contribute to receptor assembly and surface transport146,173. The GluD2 NTD can induce synapse formation in vitro and is necessary and sufficient to fulfill GluD2 function in the regulation of PF-Purkinje cell synaptogenesis180. In addition, both GluD1 and GluD2 NTDs bind to secreted proteins involved in synaptogenesis and synapse maintenance 181,182.

The GluD2 LBD was crystallized and shown to bind to D-serine183. Mutagenesis studies have shown that abolishing ligand binding doesn’t impair GluD2 function in adult mice, but results in impaired LTD and motor dyscoordination during development, and studies seem to indicate that

D-serine serves as an endogenous ligand for GluD2 in the immature cerebellum173,184,185. There is evidence that D-serine binds to GluD1, but the physiological role of this binding remains elusive186. So far, there is no evidence that ligands shown to bind to the delta receptor LBDs activate channel function183,187.

The gating machinery and ion permeation pathways of Delta receptors are similar to those of

AMPARs and kainate receptors, but have key differences188. The grafting of LBDs of AMPARs and kainate receptors into GluD1 and GluD2 receptors result in functional ion channels that can be activated by AMPAR and kainate ligands188,189. At the same time, the introduction of receptor mutants with impaired channel activity into Lurcher mice and grid2-null Purkinje cells result in the rescue of LTD and other functional abnormalities, which suggests that, at least GluD2, is unlikely to serve as a Ca2+ permeable channel in vivo190,191.

The cytoplasmic domain binds to a host of delta receptors bind to a host of intracellular proteins, and contains PDZ binding motifs146,173. Signaling via de CTD of GluD2 is required for the induction of LTD and motor learning185,190.

AMPA Receptor Structure and Dynamics

Receptor Structure

Functional iGluRs are tetrameric assemblies consisting of a dimer of dimers surrounding a central ion channel192–194. They can assemble as homomer and heteromers from subunits within the same subtype, but are found in a heteromeric state in most organisms. All iGluRs share major structural features and transmembrane topology. Each receptor has an amino-terminal domain

(ATD), a ligand binding domain (LBD) and transmembrane domain (TMD) arranged in layers18.

The ATDs are splayed outwards in the extracellular space while the transmembrane domain forms the narrowest part of the receptor. iGluRs have a two-fold axis of symmetry, perpendicular to the membrane, that transitions into a four-fold axis of symmetry in the ion channel.

Figure 1.2 Subunit assembly of AMPA receptors.

Within the ATD, the subunits involved in local dimers have extensive subunit-subunit contacts. These dimers have a local axis of symmetry that is ~24o off the overall axis of symmetry. As the polypeptide chains reach the LBD layer, the subunits that interact as dimers are crossed over between the ATD and LBD layers. The LBDs are also organized as a pair of dimers, and their local two-fold axis is ~19o off the overall axis of symmetry. The crossover of interacting subunits across domain layers, combined with the symmetry mismatch between them, result in subunit non-equivalence.

Within a homomeric receptor all subunits are chemically identical, but their assembly results in two conformationally distinct subunit pairs related to the two-fold axis of symmetry, considered the proximal and distal pairs. In distal subunits there is a substantial interface between the ATD and LBD which is not present in the proximal subunits. At the same time the

LBD of proximal subunits is closer to the membrane plane than that of distal subunits. These

21 conformational differences are defined by the linker regions between each domain layer, whose flexibility allows for different conformations.

The ATD is homologous in sequence to the bacterial leucine, isoleucine, valine binding protein (LIVBP) and the mGluR ligand-binding domain. The ATD influences receptor desensitization, and contributes to receptor assembly and surface expression195–198. However, many of the major functions of iGluRs remain intact in the absence of the ATD, and bacterial iGluRs are expressed and trafficked to the cell surface while lacking an ATD199.

The ATD is the most structurally diverse region of iGluRs16. The ATDs of AMPA and kainate receptors have similar architecture and subunit arrangement, while the ATDs of NMDA receptors have distinct conformational features and arrangements within the ATD dimer structure166,200–206.

AMPAR ATDs have upper and lower lobes, domains R1 and R2, in a clamshell arrangement.

These domains don’t have the conformational range exhibited by homologous proteins that open and close their clamshell during ligand binding. Isolated LBDs assemble as dimers in crystals and in solution200,201.

The LBD is encoded by two polypeptide segments, S1 and S2, which are interrupted by the insertion of the channel pore domain207. The S1 segment precedes the first transmembrane segment, while the S2 segment is located between the second and third transmembrane segments. This domain is homologous, both in sequence and structure, to bacterial periplasmic amino acid binding proteins, specifically lysine, arginine, ornithine binding (LAOBP) and glutamine binding (QBP) proteins46,208–210

The linkers that mediate the transition between the ATD and LBD domains are considered the ATD-S1 linkers. The ATDs of distal subunits interact with the LBDs via an ATD-LBD interface that makes the ATD-S1 linker adopt a compact conformation. At the same time the

ATD-S1 linkers of the proximal subunits have an extended conformation that nearly spans an

LBD dimer.

The bilobed structure of the LBD binds the agonist in a cleft between its two lobes211. Lobe 1 consists of the N-terminal part of S1 and a short C-terminal segment of S2, while the lobe 2 is consists mostly of the S2 segment and a short C-terminal segment of the S1. Both lobes consist of a core of the β-pleated sheets surrounded by α-helices. Lobe 1 is formed of six α-helices, four belonging to the S1 segment and two belonging to S2, while lobe 2 consists of five α-helices212.

The transmembrane region of iGluRs has a large degree of sequence homology to the K+ channel pore, although its position with respect to the membrane is inverted and their ion selectivity is very different46,213. The core channel domain consists of two transmembrane α- helices (M1-M3) and a partially helical pore loop (M2), and is in between the two polypeptide chains that make up the agonist–binding core.

The crystallized structure of a GluA2 receptor provided the first structure of an iGluR ion channel18. Viewed from the extracellular side of the membrane, the four GluA2 subunits have their transmembrane domains around a 4-fold rotational axis of symmetry. As determined initially in topology studies each subunit has three transmembrane helices (M1, M3 and M4), a central pore-like helix (M2) and a polypeptide pore-lining loop45.

The transition from the two-fold symmetry of the ATD and LBD and the four-fold symmetry of the transmembrane domain is mediated by another set of polypeptide linkers, S1-M1. This polypeptide segment adopts an extended conformation until it reaches the TMD, where the polypeptide forms a 90o turn and starts a short pre-M1 helix oriented parallel to the membrane.

This pre-M1 helix acts like a cuff that surrounds the top of the ion channel domain. This cuff makes contacts with carboxyl and amino-terminal ends of the M3 and M4 transmembrane segments, respectively.

The transmembrane domain begins with the M1 segment, which resides in the exterior of the ion channel domain. Within the ion channel pore lies the M2 helix. The M3 helices line the inside of the ion channel pore. They are ~52Å in length, and in the antagonist-bound state they cross at the pre-M1 cuff helices, creating an occlusion of the ion permeation pathway. The M3 segment is followed by the N-terminal portion of the S2 polypeptide segment, which then connects to the M4 transmembrane helix. The M4 helix resides on the exterior of the ion channel domain and connects to the S2 segment of the LBD. It makes significant subunit-subunit interactions with the transmembrane domains of adjacent subunits. This may explain in part the crucial role of the M4 helix in receptor assembly and function214.

The intracellular carboxy-terminal domain has no homology to bacterial proteins. It is the most divergent region in the iGluR subunit, ranging in size from approximately 50 amino acids in NMDA NR1 and most AMPA and kainate receptors, to hundreds of amino acids in some subtypes of NMDA subunits. This domain is involved in cytoskeletal interactions, subunit trafficking, and modulation of channel conductance. It contains different sites for posttranslational modification as well as post-transcriptional modifications215–218.

Structure and Dynamics of the Ligand Binding Domain

The ligand-binding domain, as the region that can trigger channel opening, is an important point of receptor control. As such, knowledge of its structural and dynamic properties can help in the elucidation of the mechanisms of ligand-binding and channel-opening, and be used for the development of therapeutic compounds.

The ligand binding domains of multiple iGluRs have been expressed as recombinant, isolated, water- soluble protein constructs. The polynucleotide segments encoding for S1 and S2 were genetically excised and joint by a segment encoding a linker peptide (GT), replacing the membrane segments M1-M3 of the intact receptor. These constructs retained ligand binding ability and were suitable for crystallization219,220. Currently, there are more than 120 crystal structures of iGluR LBDs, and over 80 of these are of the GluA2 LBD15,221.

In these structures the soluble LBDs are in complexes with agonists, partial agonists, and antagonists. Comparisons of the crystal structures of the LBD in the full receptor and the soluble construct have shown that they are very similar, confirming that the soluble domain is a good model to study ligand binding in AMPA receptors14. However, one must keep in mind that the binding conformations of LBDs in the full-length receptor can be influenced by other receptor domains222–225.

The ligand binding domain is a bilobate structure, where the ligand binding in the cleft between the two lobes. It has a core of beta-pleated sheets surrounded by alpha helices, which can close down in a clamshell manner when an agonist binds. This backbone is shared by all iGluR LBDs.

Lobe 1 consists of the N-terminal part of S1 and a short C-terminal segment of S2, while the lobe 2 is consists mostly of the S2 segment and a short C-terminal segment of the S1. Both lobes are composed of a central core of the β-pleated sheets surrounded by α-helices. Lobe 1 is formed of six α-helices: A, B, C, and D from the S1 segment, and J and K helices, which are formed by the S2 segment. The lobe 2 forms α-helices E, F, G, H, and I212 (Figure 1.3).

Figure 1.3. Crystal structure of the AMPAR GluA2 LBD bound to glutamate.

The ligand binding domain structures are generally flexible and exhibit functionally important dynamics at different timescales, with internal movement both in a relatively rapid timescale (ps-ns) and at the timescale where concerted atomic movements occur (µs-ms)226.

Studies performed on the GluA2 S1S2 protein have shown that the residues in lobe 1 that contact the α-substituents of glutamate in the binding site have little or no movement in the µs- ms timescale, whereas the residues in lobe 2 that contact the γ-substituents of glutamate are flexible in the same timescale. In a similar manner, the β-sheets of Lobe 1 show very little movement in the µs-ms timescale, while the β-sheets of Lobe 2 are flexible. The binding cleft

26 has various subsites where important non-covalent ligand-protein interactions play key roles in the binding process. The occupancy of different cleft subsites creates changes in binding domain dynamics that translate into changes in affinity and function. However, the physico-chemical properties of the moieties occupying them also have a large impact on affinity and function227,228.

The LBDs can explore a range of conformations in the apo and ligand-bound states, and their dynamic properties within certain regions contribute significantly to receptor function 229–234. The range and type of conformations adopted can be modified by the binding of different types of agonists.

In the apo state, the LBD domains explore open clamshell conformations and have a wider range of motions. When a ligand approaches the binding domain, it docks between the two lobes, making intermolecular contacts. Binding to full and partial agonists leads to lobe closure while the binding of antagonists can result in lobe closure or hyperextension. Lobe closure allows for the formation of a hydrogen bonding network across the lobe interface235.

When the LBD binds to a full agonist the closed lobe state is explored more often, increasing the likelihood of activating the channel in the full receptor. The binding of partial agonists to the

LBD includes a distribution of more open states as well as fully closed states, resulting in less frequent activation of the gate for that subunit227,234.

The efficacy of these ligands depends on their ability to stabilize the maximally closed state of the LBD, and the extent in which they can activate the full receptors relies on how likely the

LBD is to occupy its maximally closed conformation93,223,231,233. The free energy associated with

27 the conformational transitions involved in the binding process is however, not accounted for exclusively by the process of lobe closure223.

While there have been many studies probing the structural and dynamic characteristics of this domain, less is known about the energetic consequences of those properties and how they relate to ligand binding. Calorimetric studies can determine differences in the forces that drive ligand binding, and how these relate to receptor function.

AMPA Receptor Roles in Injured States and CNS Diseases

Due to the roles they play in fast excitatory signaling, AMPARs are crucial components of all neuronal networks. Proper function of these networks relies on the precise control of synaptic development and connectivity, and deficiencies and disruption of these systems are present in many neuropathologies158, 159. Likewise, the tight control of each step of AMPAR transmission present in healthy tissue is altered in many pathologies and injured states. One of the best-studied roles of AMPARs in neuropathologies occurs in the motor neuron disease amyotrophic lateral sclerosis, ALS.

Amyotrophic Lateral Sclerosis

Patients of ALS suffer from selective degeneration of upper and lower motor neurons, where muscle weakness in their limbs progresses through their body resulting in loss of movement.

Within the spinal cord, the neurodegeneration that leads to motor neuron death is partly due to

AMPAR-mediated excitotoxicity236,238. This occurs when disruptions in AMPAR function allow the entry of excess Ca2+ into the neurons that can trigger an excitotoxic cascade that leads to neuron death239. Most of the Ca2+ elevations in spinal motor neurons are mediated by

AMPARs240–242. These neurons lack mechanisms to cope with the high intracellular Ca2+ concentrations resulting from intense AMPAR activation238. While most AMPA receptors in the brain contain Q/R edited GluA2 subunits, rendering them Ca2+ impermeable, spinal neurons contain lower amounts of GluA2 mRNA and are more Ca2+ permeable243,244.

The amount of GluR2 mRNA between ALS patients and control subjects are comparable, but

ALS patients exhibit lower editing of the GluA2 Q/R mRNA editing site245,246. Some patients also have decreased expression of ADAR2, the adenosine deaminase that converts adenosine to inosine in the Q/R site of GluA2 pre-mRNA247,248. This leads to the expression of unedited

GluA2 subunits, increased Ca2+-permeable AMPAR levels, and thus increased likelihood of excitotoxicity. In addition, some evidence indicates that ADAR2 degradation increases with increased glutaminergic activity, resulting in a positive feedback loop249. This suggests that

ADAR2 is a potential drug target for the treatment of ALS and other diseases where AMPARs become more Ca2+ permeable.

To date, riluzole, a glutamate release inhibitor and Na+ channel blocker, is the only drug that has proven effective in prolonging patient survival and slowing the progress of the disease239,250.

Recent attempts to target the exocytotic process focused on the AMPAR non-competitive antagonist, telampanel (LY300164). In studies conducted on the mouse ALS model, treatment with telampanel reduced Ca2+ elevation in motor neurons only at the presymptomatic stage of the disease, and didn’t rescue neuron death251. ALS patients tolerated telampanel and reported some recovery of motor function during phase II clinical trials252,253. However a phase III clinical trial showed no improvement with telampanel174. These studies highlight the importance of early

29 intervention to reduce excitotoxicity, and the need of other methods of affecting AMPA receptor without blocking them globally.

Alzheimer's Disease

Alzheimer’s disease, AD, is a neurodegenerative illness that manifests as progressive cognitive and behavioral impairment, and the most common form of dementia in the US254. Early in the disease patients exhibit short-term memory loss, but as synapse loss and neuronal death progress, and different zones of the brain become atrophied, other cognitive abilities are affected and bodily functions progressively worsen255. AD is also characterized by the presence of Aβ plaques and neurofibrillary tangles, but there is disagreement on the relationship between these features and cognitive decline236,255. Currently, there are no curative or preventative therapies that can stop its progressive neuronal progressive damage, although some can improve symptoms temporarily.

In animal models, the first signs of AD involve the alteration of synapse morphology and

AMPA-mediated transmission, which occur well before cell death and plaque accumulation occurs256,257.

Soluble Aβ oligomers negatively regulate synaptic properties258–261. These molecules can bind to AMPARs and trigger clathrin-mediated endocytosis which weakens the synapses and leads to the shrinking of the dendritic spines. This dampens the overall excitatory transmission258,262.

AMPA modulators have undergone clinical trials for the treatment of mild to moderate cognitive decline263. LY451395, a positive AMPAR modulator, showed promising results in rat

30 studies. However, patients given the drug during clinical trials showed no improvement compared to the placebo group264. Currently, TARPs are being considered as drug targets to enhance AMPA receptors at the synapse236,265. In addition, clinical trials are underway in which

Aβ immunotherapy is used to remove Aβ oligomers in patients266.

Epilepsy

Epilepsy occurs in about 0.5% of the global population. Although more than a dozen drugs are available, 20-30% of sufferers don’t achieve proper seizure control using these therapies, thus the development of new drugs with different mechanisms of action is still needed267,268.

In epilepsy excitatory synaptic transmission is involved in many aspects of synchronization and seizure generation. AMPA receptors play critical roles in the initiation of the discharges characteristic of seizures 269,270. Calcium permeability is critical in initiating a seizure discharge, and can lead to long-term changes in cell excitability through the activation of NMDARs. Since

NMDARs are less essential for epileptic discharge initiation, yet are key players in the strengthening of synaptic connections necessary for learning processes, drugs targeting

AMPARs can result in less learning impairment than those that target NMDARs271.

Several compounds that inhibit AMPAR-mediated glutaminergic transmission have been synthesized and undergone clinical trials 272. Compounds like telampanel and NS1209 underwent clinical trials but currently perampanel, a noncompetitive AMPAR antagonist is approved for use within the US and Europe273. A promising compound, selurampanel (BGG492), is a competitive agonist of AMPARs and kainate receptors that has undergone Phase II clinical trials267.

Ischemia and Excitotoxicity

Ischemic stroke is the third leading cause of death in developed countries. In this disease a pattern of neuronal loss is observed due to Ca2+/Zn2+ entry into vulnerable neurons through

GluA2-containing AMPARs. Neuronal susceptibility to cerebral ischemia is correlated with a reduction of the adenosine deaminase ADAR2 and reduced Q/R RNA editing of the GluA2 subunit. Both changes lead to an increase in the amount of Ca2+- permeable AMPARs274,275.

Regions where Ca2+-permeable AMPARs are highly expressed, like the CA1 pyramidal neurons within the hippocampus, are more vulnerable to cell death after an ischemic insult than other regions236,276–278. AMPAR antagonists like NBQX and YM872 showed promise in animal model studies, but failed to treat stroke in patients due to side effects and lack of efficacy279–281 .

Willardiine Analogues and Partial Agonism

Willardiine, S(−)-α-Amino-3,4-dihydro-2,4-dioxo-1(2H)-pyrimidinepropanoic acid, is a naturally occurring excitatory amino acid first discovered in the seeds of Acacia and Mimosa plants282. It’s selectivity for non-NMDA receptors led to the development and of a series of 5- substituted willardiines with the purpose of characterizing the function of different iGluR subtypes283. Willardiine and its 5-substituted derivates act as partial agonists for AMPA and kainate receptors, and have a range of potency, efficacy, and receptor subtype specificity, thus acting as a subtle tool to study the relationship between the structural, dynamic, and functional properties of AMPARs284–286.

Figure 1.4. Willardiine structure with a 5-X substitution.

Willardiine and its analogues bind to AMPAR LBDs with a similar mode of binding to glutamate287. Their α-amino and α-carboxyl groups form bonds with residues in Lobe 1 of the

LBD, the charged moieties in within the uracil group attract regions in Lobe 2, making the cleft close upon the ligand and allowing the formation of additional inter-lobe hydrogen bonds in the

‘flip region’ that stabilize the closed conformations.

While the binding of glutamate to the LBD involves the interaction of its γ-carboxyl group with Lobe 2 willardiine analogues have a uracil moiety in its place, leading to dynamic differences of the bound complexes, and in functional differences in the full receptor232. In addition, the willardiine uracil moiety has two tautomeric forms, stemming from the deprotonation of the 3-nitrogen. The substitution of the hydrogen in the 5-position of the uracil with substituents or moieties with greater electron-withdrawing ability lowers the pKa of the compound, and shifts the equilibrium of the tautomers to the deprotonated state. The deprotonated form can interact with subsites in Lobe 2 in a way that better mimics interactions with the γ-carboxyl of glutamate than the protonated form, but these interactions are still weaker and less stable than those of a full agonist235.

The complexes formed by willardiine partial agonists and the LBD also differ from those formed by full agonists in the dynamics of the flip region, where the formation of two inter-lobe hydrogen bonds can stabilize the closed conformation287. Complexes with willardiine analogues exist in equilibrium between this conformation, and one that results in the rotation of a peptide bond of 180o and the loss of the inter-lobe hydrogen bonds232.

The differences that arise during the binding process result in differences in the stability of the closed conformations of the LBD, as well as the spread of conformations that a particular bound complex can explore. Both of these characteristics play key roles in translating cleft closure to activation232–234,288.

The physical properties of the 5-substituent affect the current responses evoked by the willardiine analogue232,283,286,289. Analogues with more electronegative 5-substituents have increased ligand potency (EC50) and affinity. Increasing the size of the 5-substituent, on the other hand, can lower peak currents and the extent of receptor desensitization. All willardiine partial agonists can activate the same subconductance states as full agonists, but they do so with different probabilities 227,228,290

Single-site substitution between willardiine analogues results in structural and dynamic differences of their LBD-bound complexes, which are reflected in functional differences when bound to AMPARs. By obtaining the thermodynamic signature of the binding process for a series of 5-substituted willardiine partial agonists we can gain a quantitative understanding of the forces responsible for their functional differences.

Ligand Binding Thermodynamics

Deepening the understanding of the molecular events occurring in biological systems through the quantification of their thermodynamic properties can reveal the nature of the energetic forces driving these events, the relative contributions of specific molecular interactions to the total free energy of binding, and how the events are affected by different conditions.

Characterizing the thermodynamic properties protein-ligand interactions is a crucial part of developing a rigorous understanding of many biological systems. By integrating thermodynamic information with detailed information of the structure and dynamics of the system, it is possible to obtain a clear vision of the mechanism of ligand binding291. This information is important not only to increase our fundamental understanding of biological systems, but for the development of biotechnology and therapeutic compounds. The development of therapeutic compounds relies on knowledge of the binding affinity of the putative drugs, as well as how processes such as ligand desolvation and the formation of hydrogen bonds, van der Waals interactions, etc. between the ligand and target contribute to the free energy of binding (ΔG)292. The binding affinity at equilibrium is related to the free energy of binding as well. This is described by the Gibb’s equation:

∆푮 = −푹푻 퐥퐧 푲푨 (Eq. 1.1)

In this equation, R is the universal gas constant (1.9872 cal/mol∙K), T is the absolute temperature in Kelvin, and KA is the equilibrium association constant.

The free energy of binding can be further separated into enthalpic (ΔH) and entropic (ΔS) contributions, either one of which can be the dominant driving force of the binding reaction.

∆퐺표 = ∆퐻표 − 푇∆푆표 (Eq. 1.2)

The contributions of enthalpic and entropic components to the free energy of binding are associated with structural and dynamics changes of the protein and ligand molecules291.

Contributions to these two forces are influenced mostly by hydrogen bonds, hydrophobic interactions, conformational changes and changes in molecular flexibility and electrostatics of the binding partners. Binding interactions with a similar free energy of binding can have different thermodynamic properties, giving them different thermodynamic signatures.

Entropy and enthalpy changes can be temperature-dependent, and their change as a function of temperature is dictated by the change in constant pressure heat capacity, ΔCp, upon binding.

Within a narrow temperature range, the temperature dependence of ΔH is given by:

∆퐻 = ∆퐻표 + ∆퐶푃(푇 − 푇표) (Eq. 1.3)

푇 ∆∆푆 = ∆∆푆표 + 훥Δ퐶푝 ln ( ) (Eq. 1.4) 푇표

Where ΔHo is the binding enthalpy at temperature To, and ΔCp is the heat capacity change of binding. The temperature dependence of the change in entropy is given by equation 1.4. Changes in the heat capacity have been related to changes in solvent accessible surface area, such as those arising from the dehydration of protein and ligand surface291.

Isothermal titration calorimetry (ITC) can be used to to quantify the specific enthalpy and entropy contributions to the free energy of binding reactions. In this technique a cell containing the protein solution is kept at the same temperature as a control cell, and the power necessary to maintain the cells in isothermal conditions is plotted over time. A concentrated ligand solution is

36 injected at specific time intervals, and the energy change resulting from the binding process is reflected in the change of power necessary to maintain the cells in isothermal conditions. The power necessary to do this is directly proportional to the heat of the reaction, and can be observed as a peak in the power vs. time plot.

As the experiment progresses, more protein binding sites become saturated, and the magnitude of these peaks decreases. A binding isotherm can be obtained through the plotting of the area of the injection peaks as a function of the molar ratio of the protein and ligand in the cell. This can be fit to different binding models to obtain the enthalpy, binding constant, and stoichiometry of the binding reaction. The binding model used for data analysis is discussed in detail in Chapter 3.

The binding of different ligands to AMPAR LBDs creates bound complexes with different dynamic and conformational properties, which have significant effect of the free energy of the binding process and the functional properties of the full receptor. The knowledge of the thermodynamic effects of physicochemical characteristics of the ligand can help further our understanding of internal dynamics of the binding cleft, and how these relate to AMPAR function. These thermodynamic patterns can also be used in the development of new probes and therapeutic drugs.

Outline of Dissertation Research

In this project we used Isothermal Titration Calorimetry to perform a full thermodynamic characterization of the GluA2 ligand binding domain, in its glutamate-bound state, binding to a series of 5-substituted willardiine analogues.

In chapter 2, we explored the role of uracil deprotonation on the thermodynamic properties of competitive binding to GluA2 LBD. The substitution of the 5-H of willardiine with a nitro moiety results in a large decrease in the pKa of the ligand, shifting the equilibrium of the uracil tautomers to one more favorable to the 3-deprotonated form. The competitive binding thermodynamics of these ligands was determined at pH levels above and below their respective pKa, and showed that the charged state of the ligand determines the enthalpic contribution to binding. The binding of the charged forms of the ligands is largely driven by enthalpic change, while that of the uncharged forms is mostly driven by entropic forces. Crystallographic work is used to demonstrate that these properties arise in part from changes in the hydrogen bonding network within the binding site involving one water molecule. The pH dependence of willardiine binding illustrates the importance of charges to the thermodynamic properties of AMPAR ligands.

The following chapter focuses on the determination of the temperature dependence of the the competitive binding of willardiine analogues to the GluA2 LBD, as well as the determination of the protonation differences between the glutamate-bound and willardiine-bound LBD complexes at physiological pH. The thermodynamic parameters of competition binding to all the studied ligands varied across temperatures, although there is little difference between the changes in free energy of binding. The change in heat capacity determined by these measurements was negative for all ligands, and while the magnitude of this change decreases with the addition of halogenated 5-substituents, it is increased significantly by the nitro 5-substitution. The difference in magnitude of the heat capacity of the ligand series is reflected by the estimated thermodynamic parameters of binding at 37oC. While the binding of willardiine and all halogenated analogues is driven by favorable entropic change, compared to the glutamate-bound 38 state, the binding of nitrowillardiine at physiological temperature is expected to be an enthalpically driven process.

References

1. Lees, G. J. Pharmacology of AMPA/Kainate Receptor Ligands and Their Therapeutic

Potential in Neurological and Psychiatric Disorders. Drugs. Jan2000 59, (2000).

2. Cull-Candy, S., Brickley, S. & Farrant, M. NMDA receptor subunits: diversity,

development and disease. Curr Opin Neurobiol 11, 327–335 (2001).

3. Franciosi, S. AMPA receptors: potential implications in development and disease. Cell.

Mol. Life Sci. 58, 921–30 (2001).

4. Lerma, J., Paternain, A. V, Rodríguez-Moreno, A. & López-García, J. C. Molecular

physiology of kainate receptors. Physiol. Rev. 81, 971–98 (2001).

5. Dingledine, R., Borges, K., Bowie, D. & Traynelis, S. F. The glutamate receptor ion

channels. Pharmacol. Rev. 51, 7–61 (1999).

6. Hollmann, M. & Heinemann, S. Cloned glutamate receptors. Annu. Rev. Neurosci. 17, 31–

108 (1994).

7. Nakanishi, S. & Masu, M. Molecular Diversity and Functions of Glutamate Receptors.

Annu. Rev. Biophys. Biomol. Struct. 23, 319–348 (1994).

8. Seeburg, P. H. The TINS/TiPS lecture: The molecular biology of mammalian glutamate

receptor channels. in Trends in Neurosciences 16, 359–365 (1993).

9. Armstrong, N. & Gouaux, E. Mechanisms for activation and antagonism of an AMPA-

sensitive glutamate receptor: crystal structures of the GluR2 ligand binding core. Neuron

28, 165–181 (2000).

10. Mayer, M. L. & Armstrong, N. Structure and function of glutamate receptor ion channels.

Annu. Rev. Physiol. 66, 161–181 (2004).

11. Mayer, M. L. Glutamate receptor ion channels. Current Opinion in Neurobiology 15, 282–

288 (2005).

12. Keinänen, K. et al. A family of AMPA-selective glutamate receptors. Science 249, 556–

560 (1990).

13. Mayer, M. L. Structure and mechanism of glutamate receptor ion channel assembly,

activation and modulation. Current Opinion in Neurobiology 21, 283–290 (2011).

14. Stawski, P., Janovjak, H. & Trauner, D. Pharmacology of ionotropic glutamate receptors:

A structural perspective. Bioorg. Med. Chem. 18, 7759–7772 (2010).

15. Pøhlsgaard, J., Frydenvang, K., Madsen, U. & Kastrup, J. S. Lessons from more than 80

structures of the GluA2 ligand-binding domain in complex with agonists, antagonists and

allosteric modulators. Neuropharmacology 60, 135–50 (2011).

16. Furukawa, H. Structure and function of glutamate receptor amino terminal domains. J.

Physiol. 590, 63–72 (2012).

17. Mayer, M. L. Emerging models of glutamate receptor ion channel structure and function.

Structure 19, 1370–80 (2011).

18. Sobolevsky, A. I., Rosconi, M. P. & Gouaux, E. X-ray structure, symmetry and

mechanism of an AMPA-subtype glutamate receptor. Nature 462, 745–756 (2009).

19. Mayer, M. L. Glutamate receptors at atomic resolution. Nature 440, 456–462 (2006).

20. Schön, A., Madani, N., Smith, A. B., Lalonde, J. M. & Freire, E. Some binding-related

drug properties are dependent on thermodynamic signature. Chem. Biol. Drug Des. 77,

161–5 (2011).

21. Madden, D. R., Abele, R., Andersson, a & Keinänen, K. Large-scale expression and

thermodynamic characterization of a glutamate receptor agonist-binding domain. Eur. J.

Biochem. 267, 4281–9 (2000).

22. Watkins, J. C. & Jane, D. E. The glutamate story. Br. J. Pharmacol. 147 Suppl , S100–8

(2006).

23. HAYASHI, T. A physiological study of epileptic seizures following cortical stimulation in

animals and its application to human clinics. Jpn. J. Physiol. 3, 46–64 (1952).

24. CURTIS, D. R., PHILLIS, J. W. & WATKINS, J. C. The chemical excitation of spinal

neurones by certain acidic amino acids. J. Physiol. 150, 656–82 (1960).

25. Curtis, D. R. & Johnston, G. A. Amino acid transmitters in the mammalian central

nervous system. Ergeb. Physiol. 69, 97–188 (1974).

26. Watkins, J. C. Metabolic regulation in the release and action of excitatory and inhibitory

amino acids in the central nervous system. Biochem. Soc. Symp. 33–47 (1972). at

27. McLennan, H., Huffman, R. D. & Marshall, K. C. Patterns of excitation of thalamic

neurones by amino-acids and by acetylcholine. Nature 219, 387–8 (1968).

28. Duggan, A. W. The differential sensitivity to L-glutamate and L-aspartate of spinal

interneurones and Renshaw cells. Exp. brain Res. 19, 522–8 (1974).

29. McCulloch, R. M., Johnston, G. A., Game, C. J. & Curtis, D. R. The differential

sensitivity of spinal interneurones and Renshaw cells to Kainate and N-methyl-D-

aspartate. Exp. brain Res. 21, 515–8 (1974).

30. Evans, R. H., Francis, A. A., Hunt, K., Oakes, D. J. & Watkins, J. C. Antagonism of

excitatory amino acid-induced responses and of synaptic excitation in the isolated spinal

cord of the frog. Br. J. Pharmacol. 67, 591–603 (1979).

31. Evans, R. H., Francis, A. A., Jones, A. W., Smith, D. A. & Watkins, J. C. The effects of a

series of omega-phosphonic alpha-carboxylic amino acids on electrically evoked and

excitant amino acid-induced responses in isolated spinal cord preparations. Br. J.

Pharmacol. 75, 65–75 (1982).

32. Yamazaki, M., Araki, K., Shibata, A. & Mishina, M. Molecular cloning of a cDNA

encoding a novel member of the mouse glutamate receptor channel family. Biochem.

Biophys. Res. Commun. 183, 886–92 (1992).

33. Araki, K. et al. Selective expression of the glutamate receptor channel delta 2 subunit in

cerebellar Purkinje cells. Biochem. Biophys. Res. Commun. 197, 1267–76 (1993).

34. Lomeli, H. et al. The rat delta-1 and delta-2 subunits extend the excitatory amino acid

receptor family. FEBS Lett. 315, 318–22 (1993).

35. Schmid, S. M. & Hollmann, M. To gate or not to gate: are the delta subunits in the

glutamate receptor family functional ion channels? Mol. Neurobiol. 37, 126–41

36. Monaghan, D. T., Bridges, R. J. & Cotman, C. W. The excitatory amino acid receptors:

their classes, pharmacology, and distinct properties in the function of the central nervous

system. Annu. Rev. Pharmacol. Toxicol. 29, 365–402 (1989).

37. Nakanishi, S. Molecular diversity of glutamate receptors and implications for brain

function. Science 258, 597–603 (1992).

38. Schoepp, D. D., Jane, D. E. & Monn, J. A. Pharmacological agents acting at subtypes of

metabotropic glutamate receptors. Neuropharmacology 38, 1431–76 (1999).

39. Trussell, L. Control of time course of glutamatergic synaptic currents. Prog. Brain Res.

116, 59–69 (1998).

40. Jonas, P. & Spruston, N. Mechanisms shaping glutamate-mediated excitatory postsynaptic

currents in the CNS. Curr. Opin. Neurobiol. 4, 366–72 (1994).

41. Galarreta, M. & Hestrin, S. Spike transmission and synchrony detection in networks of

GABAergic interneurons. Science 292, 2295–9 (2001).

42. Fricker, D. & Miles, R. EPSP amplification and the precision of spike timing in

hippocampal neurons. Neuron 28, 559–69 (2000).

43. Geiger, J. R., Lübke, J., Roth, A., Frotscher, M. & Jonas, P. Submillisecond AMPA

receptor-mediated signaling at a principal neuron-interneuron synapse. Neuron 18, 1009–

23 (1997).

44. Borges, K. & Dingledine, R. AMPA receptors: molecular and functional diversity. Prog.

Brain Res. 116, 153–70 (1998).

45. Wo, Z. G. & Oswald, R. E. Transmembrane topology of two kainate receptor subunits

revealed by N-glycosylation. Proc. Natl. Acad. Sci. U. S. A. 91, 7154–8 (1994).

46. Wo, Z. G. & Oswald, R. E. Unraveling the modular design of glutamate-gated ion

channels. Trends Neurosci. 18, 161–8 (1995).

47. Molnár, E. et al. Biochemical and immunocytochemical characterization of antipeptide

antibodies to a cloned GluR1 glutamate receptor subunit: cellular and subcellular

distribution in the rat forebrain. Neuroscience 53, 307–26 (1993).

48. Belachew, S. & Gallo, V. Synaptic and extrasynaptic neurotransmitter receptors in glial

precursors’ quest for identity. Glia 48, 185–96 (2004).

49. Schenk, U. & Matteoli, M. Presynaptic AMPA receptors: more than just ion channels?

Biol. Cell 96, 257–60 (2004).

50. Nicoll, R. A., Frerking, M. & Schmitz, D. AMPA receptors jump the synaptic cleft. Nat.

Neurosci. 3, 527–9 (2000).

51. Wisden, W. & Seeburg, P. H. Mammalian ionotropic glutamate receptors. Curr. Opin.

Neurobiol. 3, 291–8 (1993).

52. Bettler, B. & Mulle, C. Review: neurotransmitter receptors. II. AMPA and kainate

receptors. Neuropharmacology 34, 123–39 (1995).

53. Craig, A. M., Blackstone, C. D., Huganir, R. L. & Banker, G. The distribution of

glutamate receptors in cultured rat hippocampal neurons: postsynaptic clustering of

AMPA-selective subunits. Neuron 10, 1055–68 (1993).

54. Wenthold, R. J., Petralia, R. S., Blahos J, I. I. & Niedzielski, A. S. Evidence for multiple

AMPA receptor complexes in hippocampal CA1/CA2 neurons. J. Neurosci. 16, 1982–9

(1996).

55. Zhu, J. J., Esteban, J. A., Hayashi, Y. & Malinow, R. Postnatal synaptic potentiation:

delivery of GluR4-containing AMPA receptors by spontaneous activity. Nat. Neurosci. 3,

1098–106 (2000).

56. Vandenberghe, W., Nicoll, R. A. & Bredt, D. S. Stargazin is an AMPA receptor auxiliary

subunit. Proc. Natl. Acad. Sci. U. S. A. 102, 485–90 (2005).

57. Jackson, A. C. & Nicoll, R. a. Stargazing from a new vantage--TARP modulation of

AMPA receptor pharmacology. J. Physiol. 589, 5909–10 (2011).

58. Sommer, B. et al. Flip and flop: a cell-specific functional switch in glutamate-operated

channels of the CNS. Science 249, 1580–5 (1990).

59. Sommer, B., Köhler, M., Sprengel, R. & Seeburg, P. H. RNA editing in brain controls a

determinant of ion flow in glutamate-gated channels. Cell 67, 11–9 (1991).

60. Gereau, R. G. S. The Glutamate Receptors: AMPA Receptors. Humana Press, Totowa,

N.J. (2008). at

2F978-1-59745-055-

3_1.pdf?auth66=1405631203_3aa65422ecdfc37273a4370f33908473&ext=.pdf>

61. Monyer, H., Seeburg, P. H. & Wisden, W. Glutamate-operated channels: developmentally

early and mature forms arise by alternative splicing. Neuron 6, 799–810 (1991).

62. Köhler, M., Kornau, H. C. & Seeburg, P. H. The organization of the gene for the

functionally dominant alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid

receptor subunit GluR-B. J. Biol. Chem. 269, 17367–70 (1994).

63. Soderling, T. R. & Derkach, V. A. Postsynaptic protein phosphorylation and LTP. Trends

Neurosci. 23, 75–80 (2000). 46

64. Collingridge, G. L., Isaac, J. T. R. & Wang, Y. T. Receptor trafficking and synaptic

plasticity. Nat. Rev. Neurosci. 5, 952–62 (2004).

65. Higuchi, M. et al. RNA editing of AMPA receptor subunit GluR-B: A base-paired intron-

exon structure determines position and efficiency. Cell 75, 1361–1370 (1993).

66. Hollmann, M., Hartley, M. & Heinemann, S. Ca2+ permeability of KA-AMPA--gated

glutamate receptor channels depends on subunit composition. Science 252, 851–3 (1991).

67. Burnashev, N., Monyer, H., Seeburg, P. H. & Sakmann, B. Divalent ion permeability of

AMPA receptor channels is dominated by the edited form of a single subunit. Neuron 8,

189–98 (1992).

68. Nutt, S. L. & Kamboj, R. K. Differential RNA editing efficiency of AMPA receptor

subunit GluR-2 in human brain. Neuroreport 5, 1679–83 (1994).

69. Kim, D. Y., Kim, S. H., Choi, H. B., Min, C. & Gwag, B. J. High abundance of GluR1

mRNA and reduced Q/R editing of GluR2 mRNA in individual NADPH-diaphorase

neurons. Mol. Cell. Neurosci. 17, 1025–33 (2001).

70. Banke, T. G. et al. Control of GluR1 AMPA receptor function by cAMP-dependent

protein kinase. J. Neurosci. 20, 89–102 (2000).

71. Derkach, V., Barria, A. & Soderling, T. R. Ca2+/calmodulin-kinase II enhances channel

conductance of alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate type glutamate

receptors. Proc. Natl. Acad. Sci. U. S. A. 96, 3269–74 (1999).

72. Hayashi, T., Rumbaugh, G. & Huganir, R. L. Differential regulation of AMPA receptor

subunit trafficking by palmitoylation of two distinct sites. Neuron 47, 709–23 (2005).

73. Standley, S. & Baudry, M. The role of glycosylation in ionotropic glutamate receptor

ligand binding, function, and trafficking. Cell. Mol. Life Sci. 57, 1508–16 (2000).

74. Malenka, R. C. & Bear, M. F. LTP and LTD: an embarrassment of riches. Neuron 44, 5–

21 (2004).

75. Malinow, R. & Malenka, R. C. AMPA receptor trafficking and synaptic plasticity. Annu.

Rev. Neurosci. 25, 103–26 (2002).

76. Lee, H.-K. et al. Phosphorylation of the AMPA receptor GluR1 subunit is required for

synaptic plasticity and retention of spatial memory. Cell 112, 631–43 (2003).

77. Nicoll, R. A., Tomita, S. & Bredt, D. S. Auxiliary subunits assist AMPA-type glutamate

receptors. Science 311, 1253–6 (2006).

78. Tomita, S. et al. Stargazin modulates AMPA receptor gating and trafficking by distinct

domains. Nature 435, 1052–8 (2005).

79. Tomita, S., Stein, V., Stocker, T. J., Nicoll, R. A. & Bredt, D. S. Bidirectional synaptic

plasticity regulated by phosphorylation of stargazin-like TARPs. Neuron 45, 269–77

(2005).

80. Tomita, S. et al. Functional studies and distribution define a family of transmembrane

AMPA receptor regulatory proteins. J. Cell Biol. 161, 805–16 (2003).

81. Schnell, E. et al. Direct interactions between PSD-95 and stargazin control synaptic

AMPA receptor number. Proc. Natl. Acad. Sci. U. S. A. 99, 13902–7 (2002).

82. Lerma, J. & Marques, J. M. Kainate receptors in health and disease. Neuron 80, 292–311

(2013).

83. Egebjerg, J., Bettler, B., Hermans-Borgmeyer, I. & Heinemann, S. Cloning of a cDNA for

a glutamate receptor subunit activated by kainate but not AMPA. Nature 351, 745–8

(1991).

84. Gu, Z.-Q. et al. Synthesis, Resolution, and Biological Evaluation of the Four

Stereoisomers of 4-Methylglutamic Acid: Selective Probes of Kainate Receptors. J. Med.

Chem. 38, 2518–2520 (1995).

85. Schiffer, H. H., Swanson, G. T. & Heinemann, S. F. Rat GluR7 and a carboxy-terminal

splice variant, GluR7b, are functional kainate receptor subunits with a low sensitivity to

glutamate. Neuron 19, 1141–6 (1997).

86. Herb, A. et al. The KA-2 subunit of excitatory amino acid receptors shows widespread

expression in brain and forms ion channels with distantly related subunits. Neuron 8, 775–

785 (1992).

87. Wilding, T. & Huettner, J. Differential antagonism of alpha-amino-3-hydroxy-5-methyl-4-

isoxazolepropionic acid-preferring and kainate-preferring receptors by 2,3-

benzodiazepines. Mol. Pharmacol. 47, 582–587 (1995).

88. Bahn, S., Volk, B. & Wisden, W. Kainate receptor gene expression in the developing rat

brain. J. Neurosci. 14, 5525–5547 (1994).

89. Wisden, W. & Seeburg, P. A complex mosaic of high-affinity kainate receptors in rat

brain. J. Neurosci. 13, 3582–3598 (1993).

90. Castillo, P. E., Malenka, R. C. & Nicoll, R. A. Kainate receptors mediate a slow

postsynaptic current in hippocampal CA3 neurons. Nature 388, 182–6 (1997).

91. Vignes, M. et al. The GluR5 subtype of kainate receptor regulates excitatory synaptic

transmission in areas CA1 and CA3 of the rat hippocampus. Neuropharmacology 37,

1269–77

92. Frerking, M., Malenka, R. C. & Nicoll, R. A. Synaptic activation of kainate receptors on

hippocampal interneurons. Nat. Neurosci. 1, 479–86 (1998).

93. Zhang, W. et al. A transmembrane accessory subunit that modulates kainate-type

glutamate receptors. Neuron 61, 385–96 (2009).

94. Straub, C., Zhang, W. & Howe, J. R. Neto2 modulation of kainate receptors with different

subunit compositions. J. Neurosci. 31, 8078–82 (2011).

95. Straub, C. et al. Distinct functions of kainate receptors in the brain are determined by the

auxiliary subunit Neto1. Nat. Neurosci. 14, 866–73 (2011).

96. Tang, M. et al. Neto1 is an auxiliary subunit of native synaptic kainate receptors. J.

Neurosci. 31, 10009–18 (2011).

97. Ng, D. et al. Neto1 is a novel CUB-domain NMDA receptor-interacting protein required

for synaptic plasticity and learning. PLoS Biol. 7, e41 (2009).

98. Copits, B. A., Robbins, J. S., Frausto, S. & Swanson, G. T. Synaptic targeting and

functional modulation of GluK1 kainate receptors by the auxiliary neuropilin and tolloid-

like (NETO) proteins. J. Neurosci. 31, 7334–40 (2011).

99. Fisher, J. L. & Mott, D. D. Modulation of homomeric and heteromeric kainate receptors

by the auxiliary subunit Neto1. J. Physiol. 591, 4711–24 (2013).

100. Paternain, A. V, Cohen, A., Stern-Bach, Y. & Lerma, J. A role for extracellular Na+ in the

channel gating of native and recombinant kainate receptors. J. Neurosci. 23, 8641–8

(2003).

101. Bowie, D. External anions and cations distinguish between AMPA and kainate receptor

gating mechanisms. J. Physiol. 539, 725–733 (2002).

102. Plested, A. J. R. & Mayer, M. L. Structure and mechanism of kainate receptor modulation

by anions. Neuron 53, 829–41 (2007).

103. Plested, A. J. R., Vijayan, R., Biggin, P. C. & Mayer, M. L. Molecular basis of kainate

receptor modulation by sodium. Neuron 58, 720–35 (2008).

104. Rodríguez-Moreno, A., Herreras, O. & Lerma, J. Kainate receptors presynaptically

downregulate GABAergic inhibition in the rat hippocampus. Neuron 19, 893–901 (1997).

105. Clarke, V. R. et al. A hippocampal GluR5 kainate receptor regulating inhibitory synaptic

transmission. Nature 389, 599–603 (1997).

106. Rozas, J. L., Paternain, A. V & Lerma, J. Noncanonical signaling by ionotropic kainate

receptors. Neuron 39, 543–53 (2003).

107. Lerma, J. Roles and rules of kainate receptors in synaptic transmission. Nat. Rev.

Neurosci. 4, 481–95 (2003).

108. Rodrigues, R. J. & Lerma, J. Metabotropic signaling by kainate receptors. Wiley

Interdiscip. Rev. Membr. Transp. Signal. 1, 399–410 (2012).

109. Castillo, P. E., Malenka, R. C. & Nicoll, R. A. Kainate receptors mediate a slow

postsynaptic current in hippocampal CA3 neurons. 388, 182–186 (1997).

110. Goldin, M. et al. Synaptic kainate receptors tune oriens-lacunosum moleculare

interneurons to operate at theta frequency. J. Neurosci. 27, 9560–72 (2007).

111. Melyan, Z., Lancaster, B. & Wheal, H. V. Metabotropic regulation of intrinsic excitability

by synaptic activation of kainate receptors. J. Neurosci. 24, 4530–4 (2004).

112. Melyan, Z., Wheal, H. V & Lancaster, B. Metabotropic-mediated kainate receptor

regulation of IsAHP and excitability in pyramidal cells. Neuron 34, 107–14 (2002).

113. Fisahn, A., Heinemann, S. F. & McBain, C. J. The kainate receptor subunit GluR6

mediates metabotropic regulation of the slow and medium AHP currents in mouse

hippocampal neurones. J. Physiol. 562, 199–203 (2005).

114. Ruiz, A., Sachidhanandam, S., Utvik, J. K., Coussen, F. & Mulle, C. Distinct subunits in

heteromeric kainate receptors mediate ionotropic and metabotropic function at

hippocampal mossy fiber synapses. J. Neurosci. 25, 11710–8 (2005).

115. Fernandes, H. B. et al. High-affinity kainate receptor subunits are necessary for ionotropic

but not metabotropic signaling. Neuron 63, 818–29 (2009).

116. Pinheiro, P. S. et al. Selective block of postsynaptic kainate receptors reveals their

function at hippocampal mossy fiber synapses. Cereb. Cortex 23, 323–31 (2013).

117. Rodríguez-Moreno, A., López-García, J. C. & Lerma, J. Two populations of kainate

receptors with separate signaling mechanisms in hippocampal interneurons. Proc. Natl.

Acad. Sci. U. S. A. 97, 1293–8 (2000).

118. Mulle, C. et al. Subunit composition of kainate receptors in hippocampal interneurons.

Neuron 28, 475–84 (2000).

119. Christensen, J. K., Paternain, A. V, Selak, S., Ahring, P. K. & Lerma, J. A mosaic of

functional kainate receptors in hippocampal interneurons. J. Neurosci. 24, 8986–93

(2004).

120. Maingret, F., Lauri, S. E., Taira, T. & Isaac, J. T. R. Profound regulation of neonatal CA1

rat hippocampal GABAergic transmission by functionally distinct kainate receptor

populations. J. Physiol. 567, 131–42 (2005).

121. Schmitz, D., Mellor, J. & Nicoll, R. A. Presynaptic kainate receptor mediation of

frequency facilitation at hippocampal mossy fiber synapses. Science 291, 1972–6 (2001). 53

122. Lauri, S. E. et al. A critical role of a facilitatory presynaptic kainate receptor in mossy

fiber LTP. Neuron 32, 697–709 (2001).

123. Contractor, A., Swanson, G. & Heinemann, S. F. Kainate receptors are involved in short-

and long-term plasticity at mossy fiber synapses in the hippocampus. Neuron 29, 209–16

(2001).

124. Pinheiro, P. S. et al. GluR7 is an essential subunit of presynaptic kainate autoreceptors at

hippocampal mossy fiber synapses. Proc. Natl. Acad. Sci. U. S. A. 104, 12181–6 (2007).

125. Chittajallu, R. et al. Regulation of glutamate release by presynaptic kainate receptors in

the hippocampus. Nature 379, 78–81 (1996).

126. Kamiya, H. & Ozawa, S. Kainate receptor-mediated inhibition of presynaptic Ca2+ influx

and EPSP in area CA1 of the rat hippocampus. J. Physiol. 509 ( Pt 3, 833–45 (1998).

127. Frerking, M., Schmitz, D., Zhou, Q., Johansen, J. & Nicoll, R. A. Kainate receptors

depress excitatory synaptic transmission at CA3-->CA1 synapses in the hippocampus via

a direct presynaptic action. J. Neurosci. 21, 2958–66 (2001).

128. De Waard, M. et al. Direct binding of G-protein betagamma complex to voltage-

dependent calcium channels. Nature 385, 446–50 (1997).

129. Jin, X.-T. & Smith, Y. Activation of presynaptic kainate receptors suppresses GABAergic

synaptic transmission in the rat globus pallidus. Neuroscience 149, 338–49 (2007).

130. Lauri, S. E. et al. Functional maturation of CA1 synapses involves activity-dependent loss

of tonic kainate receptor-mediated inhibition of glutamate release. Neuron 50, 415–29

(2006).

131. Pickard, B. S. et al. A common variant in the 3’UTR of the GRIK4 glutamate receptor

gene affects transcript abundance and protects against bipolar disorder. Proc. Natl. Acad.

Sci. U. S. A. 105, 14940–5 (2008).

132. Pickard, B. S. et al. Cytogenetic and genetic evidence supports a role for the kainate-type

glutamate receptor gene, GRIK4, in schizophrenia and bipolar disorder. Mol. Psychiatry

11, 847–57 (2006).

133. Epsztein, J., Represa, A., Jorquera, I., Ben-Ari, Y. & Crépel, V. Recurrent mossy fibers

establish aberrant kainate receptor-operated synapses on granule cells from epileptic rats.

J. Neurosci. 25, 8229–39 (2005).

134. Artinian, J., Peret, A., Marti, G., Epsztein, J. & Crépel, V. Synaptic kainate receptors in

interplay with INaP shift the sparse firing of dentate granule cells to a sustained rhythmic

mode in temporal lobe epilepsy. J. Neurosci. 31, 10811–8 (2011).

135. Li, J.-M. et al. Aberrant glutamate receptor 5 expression in temporal lobe epilepsy lesions.

Brain Res. 1311, 166–74 (2010).

136. Mulle, C. et al. Altered synaptic physiology and reduced susceptibility to kainate-induced

seizures in GluR6-deficient mice. Nature 392, 601–5 (1998).

137. Smolders, I. et al. Antagonists of GLU(K5)-containing kainate receptors prevent

pilocarpine-induced limbic seizures. Nat. Neurosci. 5, 796–804 (2002).

138. Khalilov, I., Hirsch, J., Cossart, R. & Ben-Ari, Y. Paradoxical anti-epileptic effects of a

GluR5 agonist of kainate receptors. J. Neurophysiol. 88, 523–7 (2002).

139. Traynelis, S. F. et al. Glutamate Receptor Ion Channels : Structure , Regulation , and

Function. Pharmacol. Rev. 62, 405–496 (2010).

140. Lau, A. Y. & Roux, B. The free energy landscapes governing conformational changes in a

glutamate receptor ligand-binding domain. Structure 15, 1203–14 (2007).

141. Gielen, M., Siegler Retchless, B., Mony, L., Johnson, J. W. & Paoletti, P. Mechanism of

differential control of NMDA receptor activity by NR2 subunits. Nature 459, 703–7

(2009).

142. Rumbaugh, G., Prybylowski, K., Wang, J. F. & Vicini, S. Exon 5 and spermine regulate

deactivation of NMDA receptor subtypes. J. Neurophysiol. 83, 1300–6 (2000).

143. Vance, K. M., Hansen, K. B. & Traynelis, S. F. GluN1 splice variant control of

GluN1/GluN2D NMDA receptors. J. Physiol. 590, 3857–75 (2012).

144. Horak, M. & Wenthold, R. J. Different roles of C-terminal cassettes in the trafficking of

full-length NR1 subunits to the cell surface. J. Biol. Chem. 284, 9683–91 (2009).

145. Laurie, D. J. & Seeburg, P. H. Regional and developmental heterogeneity in splicing of

the rat brain NMDAR1 mRNA. J. Neurosci. 14, 3180–94 (1994).

146. The Glutamate Receptors. (Humana Press, 2008). doi:10.1007/978-1-59745-055-3

147. Leonard, A. S., Lim, I. A., Hemsworth, D. E., Horne, M. C. & Hell, J. W.

Calcium/calmodulin-dependent protein kinase II is associated with the N-methyl-D-

aspartate receptor. Proc. Natl. Acad. Sci. U. S. A. 96, 3239–44 (1999).

148. Kim, W. K. et al. Attenuation of NMDA receptor activity and neurotoxicity by nitroxyl

anion, NO-. Neuron 24, 461–9 (1999).

149. Choi, Y. B. et al. Molecular basis of NMDA receptor-coupled ion channel modulation by

S-nitrosylation. Nat. Neurosci. 3, 15–21 (2000).

150. Dong, Y. N., Waxman, E. A. & Lynch, D. R. Interactions of postsynaptic density-95 and

the NMDA receptor 2 subunit control calpain-mediated cleavage of the NMDA receptor.

J. Neurosci. 24, 11035–45 (2004).

151. Dong, Y. N., Wu, H.-Y., Hsu, F.-C., Coulter, D. A. & Lynch, D. R. Developmental and

cell-selective variations in N-methyl-D-aspartate receptor degradation by calpain. J.

Neurochem. 99, 206–17 (2006).

152. Paoletti, P. Molecular basis of NMDA receptor functional diversity. Eur. J. Neurosci. 33,

1351–65 (2011).

153. Watanabe, M., Inoue, Y., Sakimura, K. & Mishina, M. Developmental changes in

distribution of NMDA receptor channel subunit mRNAs. Neuroreport 3, 1138–40 (1992).

154. Akazawa, C., Shigemoto, R., Bessho, Y., Nakanishi, S. & Mizuno, N. Differential

expression of five N-methyl-D-aspartate receptor subunit mRNAs in the cerebellum of

developing and adult rats. J. Comp. Neurol. 347, 150–60 (1994).

155. Monyer, H., Burnashev, N., Laurie, D. J., Sakmann, B. & Seeburg, P. H. Developmental

and regional expression in the rat brain and functional properties of four NMDA receptors.

Neuron 12, 529–40 (1994).

156. Sheng, M., Cummings, J., Roldan, L. A., Jan, Y. N. & Jan, L. Y. Changing subunit

composition of heteromeric NMDA receptors during development of rat cortex. Nature

368, 144–7 (1994).

157. Henson, M. A., Roberts, A. C., Pérez-Otaño, I. & Philpot, B. D. Influence of the NR3A

subunit on NMDA receptor functions. Prog. Neurobiol. 91, 23–37 (2010).

158. Pachernegg, S., Strutz-Seebohm, N. & Hollmann, M. GluN3 subunit-containing NMDA

receptors: not just one-trick ponies. Trends Neurosci. 35, 240–9 (2012).

159. Vicini, S. et al. Functional and pharmacological differences between recombinant N-

methyl-D-aspartate receptors. J. Neurophysiol. 79, 555–66 (1998).

160. Yuan, H., Hansen, K. B., Vance, K. M., Ogden, K. K. & Traynelis, S. F. Control of

NMDA receptor function by the NR2 subunit amino-terminal domain. J. Neurosci. 29,

12045–58 (2009).

161. Al-Hallaq, R. A., Conrads, T. P., Veenstra, T. D. & Wenthold, R. J. NMDA di-

heteromeric receptor populations and associated proteins in rat hippocampus. J. Neurosci.

27, 8334–43 (2007).

162. Rauner, C. & Köhr, G. Triheteromeric NR1/NR2A/NR2B receptors constitute the major

N-methyl-D-aspartate receptor population in adult hippocampal synapses. J. Biol. Chem.

286, 7558–66 (2011).

163. Gray, J. A. et al. Distinct modes of AMPA receptor suppression at developing synapses by

GluN2A and GluN2B: single-cell NMDA receptor subunit deletion in vivo. Neuron 71,

1085–101 (2011).

164. Chatterton, J. E. et al. Excitatory glycine receptors containing the NR3 family of NMDA

receptor subunits. Nature 415, 793–8 (2002).

165. Piña-Crespo, J. C. et al. Excitatory glycine responses of CNS myelin mediated by

NR1/NR3 ‘NMDA’ receptor subunits. J. Neurosci. 30, 11501–5 (2010).

166. Karakas, E., Simorowski, N. & Furukawa, H. Subunit arrangement and

phenylethanolamine binding in GluN1/GluN2B NMDA receptors. Nature 475, 249–53

(2011).

167. Gielen, M. et al. Structural rearrangements of NR1/NR2A NMDA receptors during

allosteric inhibition. Neuron 57, 80–93 (2008).

168. Nowak, L., Bregestovski, P., Ascher, P., Herbet, A. & Prochiantz, A. Magnesium gates

glutamate-activated channels in mouse central neurones. Nature 307, 462–5 59

169. Schneggenburger, R., Zhou, Z., Konnerth, A. & Neher, E. Fractional contribution of

calcium to the cation current through glutamate receptor channels. Neuron 11, 133–43

(1993).

170. Skeberdis, V. A. et al. Protein kinase A regulates calcium permeability of NMDA

receptors. Nat. Neurosci. 9, 501–10 (2006).

171. MacDermott, A. B., Mayer, M. L., Westbrook, G. L., Smith, S. J. & Barker, J. L. NMDA-

receptor activation increases cytoplasmic calcium concentration in cultured spinal cord

neurones. Nature 321, 519–22

172. Hollmann, M. & Heinemann, S. Cloned glutamate receptors. Annu. Rev. Neurosci. 17, 31–

108 (1994).

173. Encyclopedia of Signaling Molecules. (Springer New York, 2012). doi:10.1007/978-1-

4419-0461-4

174. Habib, A. A. & Mitsumoto, H. Emerging drugs for amyotrophic lateral sclerosis. Expert

Opin. Emerg. Drugs 16, 537–58 (2011).

175. Ady, V. et al. Type 1 metabotropic glutamate receptors (mGlu1) trigger the gating of

GluD2 delta glutamate receptors. EMBO Rep. 15, 103–9 (2014).

176. Kohda, K. et al. Heteromer formation of delta2 glutamate receptors with AMPA or

kainate receptors. Brain Res. Mol. Brain Res. 110, 27–37 (2003).

177. Landsend, A. S. et al. Differential localization of delta glutamate receptors in the rat

cerebellum: coexpression with AMPA receptors in parallel fiber-spine synapses and

absence from climbing fiber-spine synapses. J. Neurosci. 17, 834–42 (1997).

178. Mayat, E., Petralia, R. S., Wang, Y. X. & Wenthold, R. J. Immunoprecipitation,

immunoblotting, and immunocytochemistry studies suggest that glutamate receptor delta

subunits form novel postsynaptic receptor complexes. J. Neurosci. 15, 2533–46 (1995).

179. Matsuda, S. et al. The C-terminal juxtamembrane region of the delta 2 glutamate receptor

controls its export from the endoplasmic reticulum. Eur. J. Neurosci. 19, 1683–90 (2004).

180. Kakegawa, W. et al. The N-terminal domain of GluD2 (GluRdelta2) recruits presynaptic

terminals and regulates synaptogenesis in the cerebellum in vivo. J. Neurosci. 29, 5738–

48 (2009).

181. Matsuda, K. et al. Cbln1 is a ligand for an orphan glutamate receptor delta2, a

bidirectional synapse organizer. Science 328, 363–8 (2010).

182. Matsuda, K. & Yuzaki, M. Cbln family proteins promote synapse formation by regulating

distinct neurexin signaling pathways in various brain regions. Eur. J. Neurosci. 33, 1447–

61 (2011).

183. Naur, P. et al. Ionotropic glutamate-like receptor delta2 binds D-serine and glycine. Proc.

Natl. Acad. Sci. U. S. A. 104, 14116–21 (2007).

184. Hirai, H. et al. Rescue of abnormal phenotypes of the delta2 glutamate receptor-null mice

by mutant delta2 transgenes. EMBO Rep. 6, 90–5 (2005). 61

185. Kakegawa, W. et al. D-serine regulates cerebellar LTD and motor coordination through

the δ2 glutamate receptor. Nat. Neurosci. 14, 603–11 (2011).

186. Yadav, R., Rimerman, R., Scofield, M. A. & Dravid, S. M. Mutations in the

transmembrane domain M3 generate spontaneously open orphan glutamate δ1 receptor.

Brain Res. 1382, 1–8 (2011).

187. Hansen, K. B. et al. Modulation of the dimer interface at ionotropic glutamate-like

receptor delta2 by D-serine and extracellular calcium. J. Neurosci. 29, 907–17 (2009).

188. Orth, A., Tapken, D. & Hollmann, M. The delta subfamily of glutamate receptors:

characterization of receptor chimeras and mutants. Eur. J. Neurosci. 37, 1620–30 (2013).

189. Schmid, S. M., Kott, S., Sager, C., Huelsken, T. & Hollmann, M. The glutamate receptor

subunit delta2 is capable of gating its intrinsic ion channel as revealed by ligand binding

domain transplantation. Proc. Natl. Acad. Sci. U. S. A. 106, 10320–5 (2009).

190. Kakegawa, W., Kohda, K. & Yuzaki, M. The delta2 ‘ionotropic’ glutamate receptor

functions as a non-ionotropic receptor to control cerebellar synaptic plasticity. J. Physiol.

584, 89–96 (2007).

191. Kakegawa, W. et al. Ca2+ permeability of the channel pore is not essential for the delta2

glutamate receptor to regulate synaptic plasticity and motor coordination. J. Physiol. 579,

729–35 (2007).

192. Laube, B., Kuhse, J. & Betz, H. Evidence for a tetrameric structure of recombinant

NMDA receptors. J. Neurosci. 18, 2954–61 (1998). 62

193. Rosenmund, C., Stern-Bach, Y. & Stevens, C. F. The tetrameric structure of a glutamate

receptor channel. Science 280, 1596–9 (1998).

194. Safferling, M. et al. First images of a glutamate receptor ion channel: oligomeric state and

molecular dimensions of GluRB homomers. Biochemistry 40, 13948–53 (2001).

195. Ayalon, G. & Stern-Bach, Y. Functional assembly of AMPA and kainate receptors is

mediated by several discrete protein-protein interactions. Neuron 31, 103–13 (2001).

196. Kuusinen, A., Abele, R., Madden, D. R. & Keinänen, K. Oligomerization and ligand-

binding properties of the ectodomain of the alpha-amino-3-hydroxy-5-methyl-4-isoxazole

propionic acid receptor subunit GluRD. J. Biol. Chem. 274, 28937–43 (1999).

197. Leuschner, W. D. & Hoch, W. Subtype-specific assembly of alpha-amino-3-hydroxy-5-

methyl-4-isoxazole propionic acid receptor subunits is mediated by their n-terminal

domains. J. Biol. Chem. 274, 16907–16 (1999).

198. Wells, G. B., Lin, L., Jeanclos, E. M. & Anand, R. Assembly and ligand binding

properties of the water-soluble extracellular domains of the glutamate receptor 1 subunit.

J. Biol. Chem. 276, 3031–6 (2001).

199. Chen, G. Q., Cui, C., Mayer, M. L. & Gouaux, E. Functional characterization of a

potassium-selective prokaryotic glutamate receptor. Nature 402, 817–21 (1999).

200. Clayton, A. et al. Crystal structure of the GluR2 amino-terminal domain provides insights

into the architecture and assembly of ionotropic glutamate receptors. J. Mol. Biol. 392,

1125–32 (2009). 63

201. Jin, R. et al. Crystal structure and association behaviour of the GluR2 amino-terminal

domain. EMBO J. 28, 1812–23 (2009).

202. Kumar, J., Schuck, P., Jin, R. & Mayer, M. L. The N-terminal domain of GluR6-subtype

glutamate receptor ion channels. Nat. Struct. Mol. Biol. 16, 631–8 (2009).

203. Kumar, J. & Mayer, M. L. Crystal structures of the glutamate receptor ion channel GluK3

and GluK5 amino-terminal domains. J. Mol. Biol. 404, 680–96 (2010).

204. Sukumaran, M. et al. Dynamics and allosteric potential of the AMPA receptor N-terminal

domain. EMBO J. 30, 972–82 (2011).

205. Karakas, E., Simorowski, N. & Furukawa, H. Structure of the zinc-bound amino-terminal

domain of the NMDA receptor NR2B subunit. EMBO J. 28, 3910–20 (2009).

206. Farina, A. N. et al. Separation of domain contacts is required for heterotetrameric

assembly of functional NMDA receptors. J. Neurosci. 31, 3565–79 (2011).

207. Stern-Bach, Y. et al. Agonist selectivity of glutamate receptors is specified by two

domains structurally related to bacterial amino acid-binding proteins. Neuron 13, 1345–57

(1994).

208. Sutcliffe, M. J., Wo, Z. G. & Oswald, R. E. Three-dimensional models of non-NMDA

glutamate receptors. Biophys. J. 70, 1575–89 (1996).

209. O’Hara, P. J. et al. The ligand-binding domain in metabotropic glutamate receptors is

related to bacterial periplasmic binding proteins. Neuron 11, 41–52 (1993).

210. Nakanishi, N., Shneider, N. A. & Axel, R. A family of glutamate receptor genes: evidence

for the formation of heteromultimeric receptors with distinct channel properties. Neuron 5,

569–81 (1990).

211. Paas, Y., Devillers-Thiéry, A., Teichberg, V. I., Changeux, J. P. & Eisenstein, M. How

well can molecular modelling predict the crystal structure: the case of the ligand-binding

domain of glutamate receptors. Trends Pharmacol. Sci. 21, 87–92 (2000).

212. Armstrong, N., Sun, Y., Chen, G. Q. & Gouaux, E. Structure of a glutamate-receptor

ligand-binding core in complex with kainate. Nature 395, 913–7 (1998).

213. Wood, M. W., VanDongen, H. M. & VanDongen, A. M. Structural conservation of ion

conduction pathways in K channels and glutamate receptors. Proc. Natl. Acad. Sci. U. S.

A. 92, 4882–6 (1995).

214. Wollmuth, L. P. & Sobolevsky, A. I. Structure and gating of the glutamate receptor ion

channel. Trends Neurosci. 27, 321–8 (2004).

215. Henley, J. M., Nishimune, A., Nash, S. R. & Nakanishi, S. Use of the two-hybrid system

to find novel proteins that interact with alpha-amino-3-hydroxy-5-methyl-4-isoxazole

propionate (AMPA) receptor subunits. Biochem. Soc. Trans. 25, 838–42 (1997).

216. Hirbec, H. et al. The PDZ proteins PICK1, GRIP, and syntenin bind multiple glutamate

receptor subtypes. Analysis of PDZ binding motifs. J. Biol. Chem. 277, 15221–4 (2002).

217. Leonard, A. S., Davare, M. A., Horne, M. C., Garner, C. C. & Hell, J. W. SAP97 is

associated with the alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor

GluR1 subunit. J. Biol. Chem. 273, 19518–24 (1998).

218. Sheng, M. & Pak, D. T. Ligand-gated ion channel interactions with cytoskeletal and

signaling proteins. Annu. Rev. Physiol. 62, 755–78 (2000).

219. Kuusinen, A., Arvola, M. & Keinänen, K. Molecular dissection of the agonist binding site

of an AMPA receptor. EMBO J. 14, 6327–32 (1995).

220. Chen, G. Q., Sun, Y., Jin, R. & Gouaux, E. Probing the ligand binding domain of the

GluR2 receptor by proteolysis and deletion mutagenesis defines domain boundaries and

yields a crystallizable construct. Protein Sci. 7, 2623–30 (1998).

221. Kumar, J. & Mayer, M. L. Functional Insights from Glutamate Receptor Ion Channel

Structures*. (2013). at

030212-183711>

222. Sobolevsky, A. I. Structure and Gating of Tetrameric Glutamate Receptors. J. Physiol.

593, 29–38 (2013).

223. Lau, A. Y. & Roux, B. The hidden energetics of ligand binding and activation in a

glutamate receptor. Nat. Struct. Mol. Biol. 18, 283–7 (2011).

224. Ylilauri, M. & Pentikäinen, O. T. Structural mechanism of N-methyl-D-aspartate receptor

type 1 partial agonism. PLoS One 7, e47604 (2012).

225. Hansen, K. B. et al. Structural determinants of agonist efficacy at the glutamate binding

site of N-methyl-D-aspartate receptors. Mol. Pharmacol. 84, 114–27 (2013).

226. McFeeters, R. L. & Oswald, R. E. Structural mobility of the extracellular ligand-binding

core of an ionotropic glutamate receptor. Analysis of NMR relaxation dynamics.

Biochemistry 41, 10472–81 (2002).

227. Poon, K., Ahmed, A. H., Nowak, L. M. & Oswald, R. E. Mechanisms of Modal

Activation of GluA3 Receptors □. 1, (2011).

228. Poon, K., Nowak, L. M. & Oswald, R. E. Characterizing single-channel behavior of

GluA3 receptors. Biophys. J. 99, 1437–1446 (2010).

229. Madden, D. R., Armstrong, N., Svergun, D., Pérez, J. & Vachette, P. Solution X-ray

scattering evidence for agonist- and antagonist-induced modulation of cleft closure in a

glutamate receptor ligand-binding domain. J. Biol. Chem. 280, 23637–42 (2005).

230. Ahmed, A. H., Loh, A. P., Jane, D. E. & Oswald, R. E. Dynamics of the S1S2 glutamate

binding domain of GluR2 measured using 19F NMR spectroscopy. J. Biol. Chem. 282,

12773–84 (2007).

231. Ahmed, A. H., Wang, S., Chuang, H.-H. & Oswald, R. E. Mechanism of AMPA receptor

activation by partial agonists: disulfide trapping of closed lobe conformations. J. Biol.

Chem. 286, 35257–66 (2011).

232. Fenwick, M. K. & Oswald, R. E. NMR spectroscopy of the ligand-binding core of

ionotropic glutamate receptor 2 bound to 5-substituted willardiine partial agonists. J. Mol.

Biol. 378, 673–85 (2008).

233. Landes, C. F., Rambhadran, A., Taylor, J. N., Salatan, F. & Jayaraman, V. Structural

landscape of isolated agonist-binding domains from single AMPA receptors. Nat. Chem.

Biol. 7, 168–73 (2011).

234. Ramaswamy, S. et al. Role of conformational dynamics in α-amino-3-hydroxy-5-

methylisoxazole-4-propionic acid (AMPA) receptor partial agonism. J. Biol. Chem. 287,

43557–64 (2012).

235. Ahmed, A. H. et al. Dynamics of cleft closure of the GluA2 ligand-binding domain in the

presence of full and partial agonists revealed by hydrogen-deuterium exchange. J. Biol.

Chem. 288, 27658–66 (2013).

236. Chang, P. K.-Y., Verbich, D. & McKinney, R. A. AMPA receptors as drug targets in

neurological disease--advantages, caveats, and future outlook. Eur. J. Neurosci. 35, 1908–

16 (2012).

237. Südhof, T. C. & Malenka, R. C. Understanding synapses: past, present, and future. Neuron

60, 469–76 (2008).

238. Corona, J. C., Tovar-y-Romo, L. B. & Tapia, R. Glutamate excitotoxicity and therapeutic

targets for amyotrophic lateral sclerosis. Expert Opin. Ther. Targets 11, 1415–28 (2007).

239. Lau, A. & Tymianski, M. Glutamate receptors, neurotoxicity and neurodegeneration.

Pflugers Arch. 460, 525–42 (2010).

240. Carriedo, S. G., Yin, H. Z. & Weiss, J. H. Motor neurons are selectively vulnerable to

AMPA/kainate receptor-mediated injury in vitro. J. Neurosci. 16, 4069–79 (1996).

241. Greig, A., Donevan, S. D., Mujtaba, T. J., Parks, T. N. & Rao, M. S. Characterization of

the AMPA-activated receptors present on motoneurons. J. Neurochem. 74, 179–91 (2000).

242. Van Damme, P., Van Den Bosch, L., Van Houtte, E., Callewaert, G. & Robberecht, W.

GluR2-dependent properties of AMPA receptors determine the selective vulnerability of

motor neurons to excitotoxicity. J. Neurophysiol. 88, 1279–87 (2002).

243. Kawahara, Y. et al. Human spinal motoneurons express low relative abundance of GluR2

mRNA: an implication for excitotoxicity in ALS. J. Neurochem. 85, 680–9 (2003).

244. Seeburg, P. H. & Hartner, J. Regulation of ion channel/neurotransmitter receptor function

by RNA editing. Curr. Opin. Neurobiol. 13, 279–83 (2003).

245. Kawahara, Y. et al. Glutamate receptors: RNA editing and death of motor neurons. Nature

427, 801 (2004).

246. Kwak, S. & Kawahara, Y. Deficient RNA editing of GluR2 and neuronal death in

amyotropic lateral sclerosis. J. Mol. Med. (Berl). 83, 110–20 (2005).

247. Hideyama, T. & Kwak, S. When Does ALS Start? ADAR2-GluA2 Hypothesis for the

Etiology of Sporadic ALS. Front. Mol. Neurosci. 4, 33 (2011).

248. Kwak, S., Hideyama, T., Yamashita, T. & Aizawa, H. AMPA receptor-mediated neuronal

death in sporadic ALS. Neuropathology 30, 182–8 (2010).

249. Mahajan, S. S., Thai, K. H., Chen, K. & Ziff, E. Exposure of neurons to excitotoxic levels

of glutamate induces cleavage of the RNA editing enzyme, adenosine deaminase acting on

RNA 2, and loss of GLUR2 editing. Neuroscience 189, 305–15 (2011).

250. Siciliano, G. et al. Clinical trials for neuroprotection in ALS. CNS Neurol. Disord. Drug

Targets 9, 305–13 (2010).

251. Paizs, M., Tortarolo, M., Bendotti, C., Engelhardt, J. I. & Siklós, L. Talampanel reduces

the level of motoneuronal calcium in transgenic mutant SOD1 mice only if applied

presymptomatically. Amyotroph. Lateral Scler. 12, 340–4 (2011).

252. Howes, J. F. & Bell, C. Talampanel. Neurotherapeutics 4, 126–9 (2007).

253. Pascuzzi, R. M. et al. A phase II trial of talampanel in subjects with amyotrophic lateral

sclerosis. Amyotroph. Lateral Scler. 11, 266–71 (2010).

254. Burns, A. & Iliffe, S. Alzheimer’s disease. BMJ 338, b158 (2009).

255. Selkoe, D. J. Alzheimer’s disease is a synaptic failure. Science 298, 789–91 (2002).

256. DeKosky, S. T. & Scheff, S. W. Synapse loss in frontal cortex biopsies in Alzheimer’s

disease: correlation with cognitive severity. Ann. Neurol. 27, 457–64 (1990).

257. Penzes, P. & Vanleeuwen, J.-E. Impaired regulation of synaptic actin cytoskeleton in

Alzheimer’s disease. Brain Res. Rev. 67, 184–92 (2011). 70

258. Hsieh, H. et al. AMPAR removal underlies Abeta-induced synaptic depression and

dendritic spine loss. Neuron 52, 831–43 (2006).

259. Walsh, D. M. et al. Naturally secreted oligomers of amyloid beta protein potently inhibit

hippocampal long-term potentiation in vivo. Nature 416, 535–9 (2002).

260. Lacor, P. N. et al. Synaptic targeting by Alzheimer’s-related amyloid beta oligomers. J.

Neurosci. 24, 10191–200 (2004).

261. Shankar, G. M. et al. Amyloid-beta protein dimers isolated directly from Alzheimer’s

brains impair synaptic plasticity and memory. Nat. Med. 14, 837–42 (2008).

262. Lau, C. G. & Zukin, R. S. NMDA receptor trafficking in synaptic plasticity and

neuropsychiatric disorders. Nat. Rev. Neurosci. 8, 413–26 (2007).

263. Chappell, A. S. et al. AMPA potentiator treatment of cognitive deficits in Alzheimer

disease. Neurology 68, 1008–12 (2007).

264. Shepherd, T. A. et al. Design and synthesis of a novel series of 1,2-disubstituted

cyclopentanes as small, potent potentiators of 2-amino-3-(3-hydroxy-5-methyl-isoxazol-4-

yl)propanoic acid (AMPA) receptors. J. Med. Chem. 45, 2101–11 (2002).

265. Shepardson, N. E., Shankar, G. M. & Selkoe, D. J. Cholesterol level and statin use in

Alzheimer disease: II. Review of human trials and recommendations. Arch. Neurol. 68,

1385–92 (2011).

266. Nitsch, R. M. & Hock, C. Targeting beta-amyloid pathology in Alzheimer’s disease with

Abeta immunotherapy. Neurotherapeutics 5, 415–20 (2008).

267. Faught, E. BGG492 (selurampanel)an AMPA / kainate receptor antagonist drug for

epilepsy. 492, 107–113 (2014).

268. Kwan, P. et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc

Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia 51, 1069–77

(2010).

269. Rogawski, M. A. & Hanada, T. Preclinical pharmacology of perampanel, a selective non-

competitive AMPA receptor antagonist. Acta Neurol. Scand. Suppl. 19–24 (2013).

doi:10.1111/ane.12100

270. Rogawski, M. A. AMPA receptors as a molecular target in epilepsy therapy. Acta Neurol.

Scand. Suppl. 9–18 (2013). doi:10.1111/ane.12099

271. Li, H. B., Matsumoto, K., Tohda, M., Yamamoto, M. & Watanabe, H. NMDA-but not

AMPA-receptor antagonists augment scopolamine-induced spatial cognitive deficit of rats

in a radial maze task. Brain Res. 725, 268–71 (1996).

272. Rogawski, M. a. Revisiting AMPA receptors as an antiepileptic drug target. Epilepsy

Curr. 11, 56–63 (2011).

273. Steinhoff, B. J. et al. Efficacy and safety of adjunctive perampanel for the treatment of

refractory partial seizures: a pooled analysis of three phase III studies. Epilepsia 54, 1481–

9 (2013). 72

274. Peng, P. L. et al. ADAR2-dependent RNA editing of AMPA receptor subunit GluR2

determines vulnerability of neurons in forebrain ischemia. Neuron 49, 719–33 (2006).

275. Soundarapandian, M. M., Tu, W. H., Peng, P. L., Zervos, A. S. & Lu, Y. AMPA receptor

subunit GluR2 gates injurious signals in ischemic stroke. Mol. Neurobiol. 32, 145–55

(2005).

276. Yin, H. Z., Sensi, S. L., Carriedo, S. G. & Weiss, J. H. Dendritic localization of Ca(2+)-

permeable AMPA/kainate channels in hippocampal pyramidal neurons. J. Comp. Neurol.

409, 250–60 (1999).

277. Ogoshi, F. & Weiss, J. H. Heterogeneity of Ca2+-permeable AMPA/kainate channel

expression in hippocampal pyramidal neurons: fluorescence imaging and

immunocytochemical assessment. J. Neurosci. 23, 10521–30 (2003).

278. Plant, K. et al. Transient incorporation of native GluR2-lacking AMPA receptors during

hippocampal long-term potentiation. Nat. Neurosci. 9, 602–4 (2006).

279. Sheardown, M. J., Nielsen, E. O., Hansen, A. J., Jacobsen, P. & Honoré, T. 2,3-

Dihydroxy-6-nitro-7-sulfamoyl-benzo(F)quinoxaline: a neuroprotectant for cerebral

ischemia. Science 247, 571–4 (1990).

280. Takahashi, M. et al. YM872, a novel selective alpha-amino-3-hydroxy-5-methylisoxazole-

4-propionic acid receptor antagonist, reduces brain damage after permanent focal cerebral

ischemia in cats. J. Pharmacol. Exp. Ther. 284, 467–73 (1998).

281. Furukawa, T. et al. The glutamate AMPA receptor antagonist, YM872, attenuates cortical

tissue loss, regional cerebral edema, and neurological motor deficits after experimental

brain injury in rats. J. Neurotrauma 20, 269–78 (2003).

282. Gmelin, R. Mimosaceen-Aminosäuren. Teil VII. Isolierung von Willardiin (3(l-Uracyl)-L-

Alanin) aus den Samen von Acacia millefolia, Acacia lemmoni und Mimosa asperata.

Acta Chem.Scand. 15 (1961) No. 5 1188–1189 (1961).

doi:DOI:10.3891/acta.chem.scand.15-1188

283. Patneau, D. K., Mayer, M. L., Jane, D. E. & Watkins, J. C. Activation and desensitization

of AMPA/kainate receptors by novel derivatives of willardiine. J. Neurosci. 12, 595–606

(1992).

284. Patneau, D. K. & Mayer, M. L. Kinetic analysis of interactions between kainate and

AMPA: evidence for activation of a single receptor in mouse hippocampal neurons.

Neuron 6, 785–98 (1991).

285. Armstrong, N., Mayer, M. & Gouaux, E. Tuning activation of the AMPA-sensitive GluR2

ion channel by genetic adjustment of agonist-induced conformational changes. Proc. Natl.

Acad. Sci. U. S. A. 100, 5736–41 (2003).

286. Jin, R., Banke, T. G., Mayer, M. L., Traynelis, S. F. & Gouaux, E. Structural basis for

partial agonist action at ionotropic glutamate receptors. Nat. Neurosci. 6, 803–10 (2003).

287. Jin, R. & Gouaux, E. Probing the function, conformational plasticity, and dimer-dimer

contacts of the GluR2 ligand-binding core: studies of 5-substituted willardiines and GluR2

S1S2 in the crystal. Biochemistry 42, 5201–13 (2003).

288. Maltsev, A. S., Ahmed, A. H., Fenwick, M. K., Jane, D. E. & Oswald, R. E. Mechanism

of partial agonism at the GluR2 AMPA receptor: Measurements of lobe orientation in

solution. Biochemistry 47, 10600–10 (2008).

289. Wong, L. A., Mayer, M. L., Jane, D. E. & Watkins, J. C. Willardiines differentiate agonist

binding sites for kainate- versus AMPA-preferring glutamate receptors in DRG and

hippocampal neurons. J. Neurosci. 14, 3881–97 (1994).

290. Prieto, M. L. & Wollmuth, L. P. Gating modes in AMPA receptors. J. Neurosci. 30,

4449–59 (2010).

291. Biocalorimetry 2. (John Wiley & Sons, Ltd, 2004). doi:10.1002/0470011122

292. Chaires, J. B. Calorimetry and thermodynamics in drug design. Annu. Rev. Biophys. 37,

135–51 (2008).

Chapter 2

Thermodynamics and Mechanism of the Interaction of

Willardiine Partial Agonists with a Glutamate Receptor:

Implications for Drug Development

Reproduced with permission from Biochemisty, 53(23), Madeline Martínez, Ahmed H. Ahmed, Adrienne P. Loh, and Robert E. Oswald, Thermodynamics and Mechanism of the Interaction of Willardiine Partial Agonists with a Glutamate Receptor: Implications for Drug Design, 3790-3795, Copyright 2014, Americal Chemical Society Press.

Understanding the thermodynamics of binding of a lead compound to a receptor can provide valuable information for drug design. The binding of compounds, particularly partial agonists, to subtypes of the α-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid (AMPA) receptor is, in some cases, driven by increases in entropy. Using a series of partial agonists based on the structure of the natural product, willardiine, we show that the charged state of the ligand determines the enthalpic contribution to binding. Willardiines have uracil rings with pKa values ranging from 5.5 to 10. The binding of the charged form is largely driven by enthalpy, while that of the uncharged form is largely driven by entropy. This is due at least in part to changes in the hydrogen bonding network within the binding site involving one water molecule. This work illustrates the importance of charge to the thermodynamics of binding of agonists and antagonists to AMPA receptors and provides clues for further drug discovery.

Introduction

Understanding the thermodynamics of binding can be an important element in the optimization of the binding of a drug to its receptor.1-3 Ionotropic glutamate receptors are ligand-gated ion channels that are important drug targets because of their roles in a variety of neurological diseases, and in learning and memory.4 This group of neurotransmitter receptors consists of three major subtypes that include (1) α-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid receptors (AMPA; GluA1–4), (2) kainate receptors (GluK1–5), and (3) N-methyl-d-aspartic acid receptors (NMDA; GluN1, GluN2A–D, and GluN3A and -B). AMPA receptors are responsible for most of the fast excitatory transmission in the vertebrate central nervous system (CNS). They assemble as homo- or heterotetrameric channels, and each subunit is modularly arranged with an

N-terminal domain, a ligand-binding domain (LBD), a transmembrane region, and a C-terminal

77 domain.5 Both extracellular domains are arranged as dimers of dimers, with subunit crossover occurring between the N-terminal and ligand-binding domains.6 The LBD can be produced in isolation by bacterial expression7 and has been shown to be an effective model system for understanding the details of effector binding.8

The LBD is a bilobed structure, and glutamate binds in the cleft between the two lobes. The mechanism of activation follows a three-step process, which includes the binding of glutamate to lobe 1, followed by closure of the lobes and interaction with lobe 2, and finally the stabilization of the closed lobe form with the formation of two hydrogen bonds.9 A range of compounds can bind to the agonist site, all of which bind in the same manner to lobe 1. The size of the compound, the characteristics of the interaction with lobe 2, and the stability of the fully closed form (including the presence of two stabilizing hydrogen bonds) determine if a compound is a full agonist, partial agonist, or antagonist. This is illustrated by derivatives of the natural product, willardiine, which has served as a scaffold for a range of glutamateric agonists and antagonists.10,11 Substituting the willardiine ring at position 3 with a carboxybenzyl or carboxyethyl substituent results in an antagonist with binding driven largely by a favorable enthalpy decrease.12 Alternatively, 5-iodowillardiine (IW) is a partial agonist with binding driven largely by an entropy increase at physiological pH.12 We describe here the thermodynamics of the binding of a series of willardiine partial agonists13-14 to the GluA2 LBD using isothermal titration calorimetry (ITC). These analogues, modified at position 5 (F, Cl, I, H, and NO2), illustrate the effect of ligand charge on the entropic and enthalpic contributions to the Gibbs free energy of binding and suggest how the structure of the agonist can be modified to optimize the enthalpic component of binding.

Methods

Protein Purification

The plasmid for the GluA2 ligand-binding domain, GluA2 LBD, was provided by E. Gouaux. It consists of residues N392–K506 and P632–S775 of the full rat GluA2-flop subunit with a “GT” linker connecting K506 and P632.15 The plasmid was transformed in the Origami B (DE3)

Escherichia coli strain, and the GluA2 LBD was expressed and purified as described previously.16 Because of the necessity of obtaining accurate protein concentrations and the difficulties with standard methods for protein concentration, thiol quantitation was used (Thiol and Sulfide Quantitation Kit, Molecular Probes), based on a method described by Singh et al.17,18

In this assay, the thiols in the protein reduce a disulfide-inhibited derivative of papain, which releases the active enzyme in a stoichiometric manner. A chromogenic papain substrate is then used to measure the activity of the enzyme colorimetrically. As a standard, Ellman’s Reagent was used to determine the thiol concentration of l-cysteine standard solutions. Because the number of thiols in each protein molecule is known (four cysteines), we can determine the number of protein molecules within the protein sample. At the same time, we determined the absorbance at 280 nm of an identical protein aliquot, which allowed us to determine the extinction coefficient for our preparation of GluA2 LBD (58363 M–1). The protein concentrations reflect the GluA2 LBD monomer. The dimerization constant for the GluA2 LBD is on the order of 6–40 mM,19,20 so that no significant concentration of dimer was present at the concentrations of protein used in these experiments.

Isothermal Titration Calorimetry (ITC)

ITC experiments were conducted on a Microcal VP-ITC calorimeter at 10, 15, and 20 °C.

Willardiine derivatives were obtained from Tocris and Abcam Biochemicals. The final protein concentration ranged from 8 to 40 μM. Titrations were conducted in the same buffer over a range of temperatures (5–20 °C) using a syringe speed of 300 rpm and a reference power of 10 μcal/s.

Typical titration experiments consisted of 35 injections in which the individual injections were 6

μL (0.4 mM ligand) and were made every 240 s. The calibrated cell feedback signal

(microcalories per second) was collected at 2 s intervals. The areas derived from the first injection were not used in the analysis. Experimental data were corrected for buffer mismatch by subtracting control titrations of the ligand solution into the ITC buffer (phosphate or cacodylate).

The thermograms were analyzed using the competitive binding approach described by

Sigurskjold et al., with the glutamate-bound state as the reference state21. Data were fit using a

Ka for the binding of glutamate of 1.0 × 106 M–1. The binding stoichiometry (n), the apparent enthalpy of binding (ΔΔH), and Ka were obtained by fitting. The value of n is the molar ratio of the ligand to the protein, and ΔΔH is expressed in kilocalories per mole of injectant. The values of n, Ka, and ΔΔH were used to compute ΔΔS° and ΔΔG° at the experimental temperature (T).

Whenever possible, the binding reactions were conducted at multiple temperatures, so that the change in heat capacity (ΔΔCp) of the binding reaction could be determined.

Implications of Competition Binding for Thermodynamic Parameters

Because of the low stability of the apo state in combination with the difficulty of completely removing glutamate from the GluA2 LBD sample, more consistent results were obtained using competition experiments, that is, the competition of a given ligand with glutamate. The ground 80 state is, in this case, defined as the glutamate-bound state rather than the apo state. For a related

AMPA receptor LBD (GluA4), Madden and collaborators22 were able to obtain thermodynamic parameters for the binding of glutamate to the apo state (ΔH values of −2.49 ± 0.03 and −4.21 ±

0.03 kcal/mol at 15 and 20 °C, respectively; −TΔS° values of −5.39 and −3.96 kcal/mol at 15 and

20 °C, respectively; ΔCp of −171 cal mol–1 K–1). Despite the fact that the binding sites of all

AMPA receptors are very similar, the thermodynamic parameters for GluA4 LBD cannot necessarily be compared directly to the quantities determined for the GluA2 LBD23-25. However, referencing the glutamate-bound state of GluA2 at physiological pH does reveal the quantitative and qualitative differences between the thermodynamic parameters of the binding of willardiine ligands, as shown in Figure 2.1,C. Likewise, in the calculation of ΔΔCp (slope of ΔΔH vs temperature), the reference state is constant for each temperature (Figure 2.3). ΔCp (heat capacity relative to the apo state) would be determined by the difference between ΔΔCp and the ΔCp for glutamate relative to the apo state.

Crystallography

Crystals were grown using the hanging drop technique at 4°C. Each drop contained a 1:1 (v/v) ratio of protein solution to reservoir solution [16–18% PEG 8K, 0.1 M sodium cacodylate, and

0.1–0.15 M zinc acetate (pH 3.5)]. Data were collected at Cornell High Energy Synchrotron

Source (CHESS) beamline A1 (wavelength of 0.987 Å, temperature of 100 K) using a Quantum-

210 Area Detector Systems charge-coupled device detector. Data were indexed and scaled using

HKL-3000.26 Structures were determined with molecular replacement and refined with Phenix.27

Model building was conducted with Coot version 0.7.28

Results

The apo form of the GluA2 LBD is difficult to prepare and is much less stable than ligand-bound forms, so the experiments described here were conducted as competition binding relative to the glutamate-bound form. For this reason, thermodynamic parameters are expressed with the glutamate-bound form as the reference state (ΔΔH, −TΔΔS°, and ΔΔCp). Panels A and B of

Figure 2.1 show thermograms (pH 7.2 and 20 °C) of the displacement of glutamate by 5- nitrowillardiine (NW) and IW. The most striking difference is that the reaction is exothermic for

NW and endothermic for IW. The thermograms for 5-chlorowillardiine (ClW), willardiine (HW), and 5-fluorowillardiine (FW) indicate endothermic processes (Figure 2.4 of the Supporting

Information). The enthalpy change becomes progressively less favorable with decreasing electronegativity of the substituent [increasing ΔΔH (Figure 2.1,C)], with only NW having a favorable ΔΔH (−9 ± 1 kcal/mol). Conversely, at 20 °C, the entropic component of the displacement (−TΔΔS°) is favorable for all of the derivatives except NW. The favorability of the entropy of binding decreases as a function of electronegativity, with IW and HW having the largest increases.

Figure 2.1 Thermograms showing raw (top) and integrated (bottom) data for NW (A) and IW (B) displacement of glutamate from the GluA2 LBD at 20 °C. The molar ratio is the ratio of ligand to protein.

The influence of substituent electronegativity on the thermodynamic parameters suggested that the differing pKa of the uracil ring, due to the different analogues in position 5, may affect the

83 thermodynamics of binding. While the pKa of all the halogenated uracils is 8, those of HW and

NW are approximately 10 and 6, respectively.29,30 This means that willardiine and the halogenated analogues were, at physiological pH, in mostly uncharged (protonated) states

(Figure 2A). NW, however, was in a mostly charged (deprotonated) state. The GluA2 LBD is well-behaved between pH 4 and 10, with minor changes in the NMR 1H–15N HSQC spectrum.9

We tested the possibility that the charged state of the ligand could affect the enthalpy versus entropy distribution by changing the pH. Thus, binding of largely uncharged (protonated uracil ring) NW was examined at pH 4, and binding of largely charged (deprotonated uracil ring) HW was examined at pH 10.

Figure 2.2(A) Ionization states of the uracil ring of NW. Similar ionization states are possible for the other willardiine derivatives. (B) Effect of a change in pH on the thermodynamic parameters for NW (left) and HW (right) at 10 °C. For both compounds, the pH at which the ionized state is favored is shown on the left and that for which the un-ionized state is favored is shown on the right. (C) Structure of the binding site, showing the direct interactions between NW and the

GluA2 LBD [Protein Data Bank (PDB) entry 3RTW32]. (D) Structure of the binding site for NW showing the hydrogen

84 bonding network associated with the nitrogen at position 3 of the willardiine ring in the charged form (pH 6.5, PDB entry

3.5 (PDB entry 4Q30). At this pH, the ring is largely uncharged. The formatting of this panel is similar to that of panel D.

Note the change in the H-bonding network. The potential H-bond between the hydroxyl of S654 and the uracil carbonyl is shown with a question mark because the distance between the two oxygens is relatively long, 3.5 Å, and the angle between this H-bond and the H-bond with the water is not favorable (80°).

Unlike the case at pH 7.2, at pH 10, the displacement of glutamate by HW (approximately 80% deprotonated) was exothermic, with a favorable binding enthalpy of −5 ± 1 kcal/mol (Figure 2.2

B, 10 °C). The ΔΔH at pH 10 is 14 ± 3 kcal/mol more favorable than at pH 7.2. Deprotonation, however, results in a −TΔΔS° that is less favorable by 14 ± 3 kcal/mol than at pH 7.2. This is a clear example of enthalpy–entropy compensation, where the interactions that result in a favorable enthalpy change result in an increased order of the system, and thus in an equivalent entropic cost (decrease). The compensation between enthalpic and entropic factors results in a negligible difference in ΔΔG° between the binding of protonated and deprotonated forms of willardiine.

At pH 4, the uracil ring in NW is more than 97% protonated (uncharged). At this pH, the binding reaction is endothermic, unlike binding at pH 7.2, with a ΔΔH of 7 ± 1 kcal/mol (Figure 2.2 B).

This is 10 kcal/mol less favorable than the binding at pH 7.2. However, at pH 4, the −TΔΔS° is 7

± 2 kcal/mol more favorable than in the deprotonated form. Although there is also an enthalpy– entropy compensation observed between the two pH values for NW, a decrease in the free energy of binding by 0.7 ± 0.3 kcal/mol is observed, with a lower binding affinity at pH 4.

Because of the role of the side chain of E705 in the binding site (Figure 2.2 C–E), the decrease in affinity at pH 4 could be, in part, related to a partial protonation at this site. Nevertheless, the

85 data strongly indicate that, although pH is a minor determinant of ΔΔG° (and, thus, the KD of binding) of willardiine analogues, the distribution of entropy and enthalpy changes varies with the charged state of the uracil ring in a predictable way. In particular, a negatively charged state correlates with an exothermic binding enthalpy, while an uncharged state correlates with an endothermic binding enthalpy. This would suggest that for enthalpy optimization, the presence of a negative charge on the ligand as it interacts with lobe 2 would be preferred. This is consistent with the observed binding of other charged willardiines.12

The temperature profile of the thermodynamic parameters can provide more detailed insight into the binding modes of these ligands. Within a narrow temperature range, the temperature dependence of ΔΔH is given by

ΔΔH= ΔΔHo + ΔΔCp(T-To)

o where ΔΔH is the binding enthalpy at an arbitrary reference temperature, To, and ΔΔCp is the heat capacity change of binding. The values of ΔΔCp are determined from the slope of a plot of

ΔΔH versus temperature as shown in Figure 2.3. The interpretation of ΔΔCp is somewhat complex because it is based on a competition measurement. However, Madden and collaborators described the temperature dependence of ΔH for glutamate binding to the apo form in a related

22 LBD (GluA4), and the ΔΔCp was −171 cal mol–1 K–1 . In comparison, the ΔΔCp for NW binding relative to glutamate is more negative (−671 cal mol–1 K–1), suggesting that the ΔCp for

NW binding relative to that of the apo form would be considerably more negative than for glutamate binding. This is consistent with burying polar groups upon binding. In addition, in comparison to the other willardiine derivatives used, NW has a negatively charged nitro group at position 5 and, as discussed above, the uracil ring is largely charged at physiological pH. Thus, 86 the uracil ring of NW is more charged than the other willardiines used or glutamate, consistent with its more negative change in heat capacity. For the other willardiines, which have largely uncharged uracil rings at physiological pH and have halogens or a proton at position 5, the temperature dependence of ΔΔH is similar to or less than that of glutamate alone, possibly suggesting that their ΔCp (relative to apo) is either near zero or slightly negative. Given the polarity of the binding site and the H-bonds formed by the uracil ring with lobe 2 of GluA2, the binding is unlikely to be driven by the hydrophobic effect, which predicts a positive ΔCp.

Overall, because the binding of agonists and antagonists shows very consistent interactions with lobe 1, it is the interaction with lobe 2 that contributes to the thermodynamics of binding and the efficacy of a ligand. Like the pH effects described above, these results suggest that charged interactions with lobe 2 contribute to the optimization of the enthalpy of the reaction, particularly at physiological temperatures (extrapolating from the temperature dependence shown in Figure

2.3.

Figure 2.3 Dependence of ΔΔH on temperature.

While crystal structures of NW bound to GluA2 LBD differ little when the pH of the crystallization medium is changed from 6.5 (Figure 2.2 D) to 3.5 (Figure 2.2 E and Table 2.1 of the Supporting Information), the ITC results suggest that the binding interactions in the charged form differ significantly from those in the uncharged form. An extensive H-bonding network is present in which both the nitrogen at position 3 on the uracil ring and the α-amide are involved.

A similar network is observed for all of the willardiine derivatives modified at position 5 and bound to GluA2or GluA314,31,32. A change in the configuration of the H-bond donors and acceptors could accommodate the charged and uncharged forms of the uracil ring (Figure 2.2, D vs Figure 2.2, E), maintaining a similar network. In the charged form (Figure 2.2, D), the H-bond between the willardiine and the hydroxyl of T655 may be stronger, perhaps giving rise to the more favorable enthalpy change. Additionally, the number of H-bonds formed by the hydroxyl

88 of S654 is expected to change with the charge state. If the uracil ring is charged, the hydroxyl of

S654 forms two favorable H-bonds. However, when the uracil ring is uncharged, the H-bond with water seems favorable, but the H-bond shown with the ring carbonyl does not have a favorable angle or distance and may or may not be present, possibly reducing the enthalpy of binding at lower pH values. Although this is a complex experiment because it compares bound willardiine to bound glutamate, the glutamate reference state remains largely constant (for a given condition). Thus, the binding of the charged state of the willardiines may be expected to have a change in enthalpy more favorable than that of the uncharged state. Interaction with solvent may also affect the thermodynamics such that a portion of the enthalpic and entropic differences may well be associated with differences in the enthalpy and entropy of the interactions with water.

Nevertheless, these results strongly suggest that the willardiine partial agonists can bind in either the charged or uncharged state of the uracil ring and provide a tool for showing how the enthalpy of binding can be optimized. The differences between the binding of the charged form of the uracil ring and the uncharged form are negligible in the crystal structure. Therefore, the difference is likely to arise, at least in part, from the increased strength of the H-bonding in the charged form that would stabilize the closed lobe form of the protein. This increases the enthalpy of binding at an entropic cost.

Discussion

A great deal of consideration has been given to strategies for the development of lead compounds for drug development. Although large hydrophobic drugs can be optimized to relatively high free energies of binding, the driving force is often a change in the entropy due to 89 the increase in the degrees of freedom of water when the hydrophobic surface is buried in the protein. This sometimes results in compounds with poor solubility and selectivity. In a number of cases (e.g., HIV protease inhibitors33 and statins34), optimizing binding enthalpy improved the affinity and selectivity. In the case of glutamate receptors, drugs targeting the agonist-binding site and the several allosteric sites have significant therapeutic potential. Of those drugs that act at the agonist-binding site, subtype selective antagonists may be more likely to have therapeutic efficacy for AMPA and kainate receptors, whereas partial agonists seem to have better therapeutic profiles for NMDA receptors.35 Our studies of willardiine partial agonists suggested that even when bound, the receptor exhibits considerable dynamics on the microsecond to millisecond time scale.36 The electronegativity of the substituent has been correlated with the affinity of the willardiine derivative, and the size of the substituent has been correlated with efficacy.14 Measurements of HD exchange,37 disulfide trapping experiments,38 and residual dipolar coupling measurements39 suggested that it is the stability of the fully closed lobe form that determines the efficacy of the partial agonist. For the halogenated willardiines, all of the previous studies were conducted below the uracil pKa, so that the uncharged form predominated, and the binding was driven largely by entropy.12 While the interaction with lobe 2 is clearly not hydrophobic, in the uncharged state, the H-bonds are not formed through charged interactions and are thus weaker than what one would see with, for example, the γ-carboxyl of glutamate.

The work described here suggests that interactions with lobe 2 that are driven by charge lead to higher enthalpies of binding and lower entropy (possibly accompanied by lower dynamics in the ligand–receptor complex). This would suggest that in the future design of partial agonists for

NMDA receptors or antagonists for AMPA and kainate receptors, an important consideration would be to build one or more charge interactions into the portion of the molecule that interacts

90 with lobe 2, taking into consideration transfer through the blood–brain barrier in the case of drugs targeted to the CNS. These results illustrate the power of ITC in combination with X-ray crystallography and NMR spectroscopy in improving our understanding of the important thermodynamic characteristics of ligand binding to this essential neurotransmitter receptor.

Supplementary Material

The main article examines the thermodynamics of the interaction of partial agonists with the

GluA2 ligand-binding domain (LBD). In the supporting information, we provide thermograms and integrated data for the binding of derivatives not shown in the body of the paper at pH 7.2 and 20°C (Figures 2.41), and provide the structural statistics for the crystal structure of GluA2

LBD bound to nitrowillardiine (NW) at pH 3.5 (Table 2.1).

Figure 2.4 Thermograms showing raw (top) and integrated (bottom) data for displacement of glutamate from GluA2 LBD at pH 7.2 and 20°C.

STRUCTURE

Space Group P22121

Unit Cell (Å) a=47.8 b=113.3 c=164.3

X-ray source CHESS (A1)

Wavelength (Å) 0.977

Resolution (Å) 50-2.03 (2.07-2.03)

Measured reflections (#) 412588

Unique reflections (#) 56635

Data redundancy 7.1 (6.1)

Completeness (%) 99.6 (97.7)

Rsym (%) 13.5 (75.1)

I/σi 25.3 (4.0)

PDB ID 4Q30

MODEL REFINEMENT STATISTICS

Phasing MR

Molecules/AU 3

Rwork/Rfree (%) 18.6/23.0

Free R test set size (#/%) 1944 (3.4)

Number of protein atoms 6002

Number of heteroatoms 51

Rmsd bond length (Å) . 007

Rmsd bond angles (°) 1.06

Average B factors:

Protein 23.3 ± 0.12

Nitrowillardiine 18.3 ± 0.88

Zinc 32.4 ± 3.7

Water 32.4 ± 0.29

Table 2.1 Structural statistics for the crystal structure of NW bound to GluA2 LBD at pH 3.

REFERENCES

1. Freire, E. (2008) Do enthalpy and entropy distinguish first in class from best in class?, Drug Discov Today 13, 869-874. 2. Freire, E. (2009) A thermodynamic approach to the affinity optimization of drug candidates, Chem Biol Drug Des 74, 468-472. 3. Ferenczy, G. G., and Keseru, G. M. (2010) Thermodynamics guided lead discovery and optimization, Drug Discov Today 15, 919-932. 4. Traynelis, S. F., Wollmuth, L. P., McBain, C. J., Menniti, F. S., Vance, K. M., Ogden, K. K., Hansen, K. B., Yuan, H., Myers, S. J., Dingledine, R., and Sibley, D. (2010) Glutamate receptor ion channels: structure, regulation, and function, Pharmacol Rev 62, 405-496. 5. Wo, Z. G., and Oswald, R. E. (1995) Unraveling the modular design of glutamate-gated ion channels, Trends Neurosciences 18, 161-168. 6. Sobolevsky, A. I., Rosconi, M. P., and Gouaux, E. (2009) X-ray structure, symmetry and mechanism of an AMPA-subtype glutamate receptor, Nature 462, 745-756. 7. Chen, G. Q., and Gouaux, E. (1997) Overexpression of a glutamate receptor (GluR2) ligand binding domain in Escherichia coli: application of a novel protein folding screen, Proc Natl Acad Sci USA 94, 13431-13436. 8. Du, M., Reid, S. A., and Jayaraman, V. (2005) Conformational changes in the ligand binding domain of a functional ionotropic glutamate receptor, J Biol Chem 280, 8633- 3636.

9. Fenwick, M. K., and Oswald, R. E. (2010) On the mechanisms of alpha-amino-3- hydroxy-5- methylisoxazole-4-propionic acid (AMPA) receptor binding to glutamate and kainate, J Biol Chem 285, 12334-12343. 10. Jane, D. E., Hoo, K., Kamboj, R., Deverill, M., Bleakman, D., and Mandelzys, A. (1997) Synthesis of willardiine and 6-azawillardiine analogs: pharmacological characterization on cloned homomeric human AMPA and kainate receptor subtypes, J Med Chem 40, 3645- 3650. 11. Dolman, N. P., More, J. C., Alt, A., Knauss, J. L., Troop, H. M., Bleakman, D., Collingridge, G. L., and Jane, D. E. (2006) Structure-activity relationship studies on N3- substituted willardiine derivatives acting as AMPA or kainate receptor antagonists, J Med Chem 49, 2579-2592. 12. Ahmed, A., Thompson, M., Fenwick, M., Romero, B., Loh, A., Jane, D., Sondermann, H., and Oswald, R. (2009) Mechanisms of antagonism of the GluR2 AMPA receptor: Structure and dynamics of the complex of two willardiine antagonists with the glutamate binding domain, Biochemistry 48, 3894-3903. 13. Patneau, D. K., Mayer, M. L., Jane, D. E., and Watkins, J. C. (1992) Activation and desensitization of AMPA/kainate receptors by novel derivatives of willardiine., J. Neurosci. 12, 595-606. 14. Jin, R., Banke, T. G., Mayer, M. L., Traynelis, S. F., and Gouaux, E. (2003) Structural basis for partial agonist action at ionotropic glutamate receptors, Nat Neurosci 6, 803-810. 15. Chen, G. Q., Sun, Y., Jin, R., and Gouaux, E. (1998) Probing the ligand binding domain of the GluR2 receptor by proteolysis and deletion mutagenesis defines domain boundaries and yields a crystallizable construct, Protein Sci 7, 2623-2630. 16. Ptak, C. P., Ahmed, A. H., and Oswald, R. E. (2009) Probing the allosteric modulator binding site of GluR2 with thiazide derivatives, Biochemistry 48, 8594-8602. 17. Singh, R., Blattler, W. A., and Collinson, A. R. (1993) An amplified assay for thiols based on reactivation of papain, Anal Biochem 213, 49-56. 18. Singh, R., Blattler, W. A., and Collinson, A. R. (1995) Assay for thiols based on reactivation of papain, Methods Enzymol 251, 229-237. 14 19. Ptak, C. P., Hsieh, C. L., Weiland, G. A., and Oswald, R. E. (2014) Role of stoichiometry in the dimer-stabilizing effect of AMPA receptor allosteric modulators, ACS Chem Biol 9, 128-133. 20. Sun, Y., Olson, R., Horning, M., Armstrong, N., Mayer, M., and Gouaux, E. (2002) Mechanism of glutamate receptor desensitization, Nature 417, 245-253. 21. Sigurskjold, B. W. (2000) Exact analysis of competition ligand binding by displacement isothermal titration calorimetry, Anal Biochem 277, 260-266.

22. Madden, D. R., Abele, R., Andersson, A., and Keinanen, K. (2000) Large-scale expression and thermodynamic characterization of a glutamate receptor agonist-binding domain, Eur J Biochem 267, 4281-4289. 23. Ahmed, A. H., Wang, Q., Sondermann, H., and Oswald, R. E. (2009) Structure of the S1S2 glutamate binding domain of GluR3, Proteins 75, 628-637. 24. Armstrong, N., and Gouaux, E. (2000) Mechanisms for activation and antagonism of an AMPAsensitive glutamate receptor: crystal structures of the GluR2 ligand binding core, Neuron 28, 165- 181. 25. Gill, A., Birdsey-Benson, A., Jones, B. L., Henderson, L. P., and Madden, D. R. (2008) Correlating AMPA receptor activation and cleft closure across subunits: crystal structures of the GluR4 ligandbinding domain in complex with full and partial agonists, Biochemistry 47, 13831-13841. 26. Otwinowski, Z., and Minor, W. (1997) Processing of X-ray diffraction data collected in oscillation mode, In Methods in Enzymology, Vol. 276, Macromolecular Crystallography, part A. (Carter, C. W., and Sweet, R. M., Eds.), pp 307-326, Academic Press, New York. 27. Adams, P. D., Grosse-Kunstleve, R. W., Hung, L. W., Ioerger, T. R., McCoy, A. J., Moriarty, N. W., Read, R. J., Sacchettini, J. C., Sauter, N. K., and Terwilliger, T. C. (2002) PHENIX: building new software for automated crystallographic structure determination, Acta Crystallogr D Biol Crystallogr 58, 1948-1954. 28. Emsley, P., and Cowtan, K. (2004) Coot: model-building tools for molecular graphics, Acta Crystallogr D Biol Crystallogr 60, 2126-2132. 29. Lidak, M. Y., Dipan, I. V., Paégle, R. A., and Stradyn, Y. P. (1972) Protolysis of some alphaamino-beta-(1-pyrimidyl)propionic acids and their analogs., Translated from: Khimiya Geterotsiklicheskikh Soedinenii 5, 708-712. 30. Wempen, I., and Fox, J. J. (1964) Spectrometric studies of nucleic acid derivatives and related compounds. VI. On the structure of certain 5- and 6-halogenouracils and - cytosines., J. Am. Chem. Soc. 86, 2474-2477. 31. Jin, R., and Gouaux, E. (2003) Probing the function, conformational plasticity, and dimer- dimer contacts of the GluR2 ligand-binding core: Studies of 5-substituted willardiines and GluR2 S1S2 in the crystal, Biochemistry 42, 5201-5213. 32. Poon, K., Ahmed, A. H., Nowak, L. M., and Oswald, R. E. (2011) Mechanisms of modal activation of GluA3 receptors, Mol Pharmacol 80, 49-59. 33. Ohtaka, H., and Freire, E. (2005) Adaptive inhibitors of the HIV-1 protease, Prog Biophys Mol Biol 88, 193-208. 34. Carbonell, T., and Freire, E. (2005) Binding thermodynamics of statins to HMG-CoA reductase, Biochemistry 44, 11741-11748.

35. Stanton, P. K., Potter, P. E., Aguilar, J., Decandia, M., and Moskal, J. R. (2009) Neuroprotection by a novel NMDAR functional glycine site partial agonist, GLYX-13, Neuroreport 20, 1193-1197. 36. Fenwick, M. K., and Oswald, R. E. (2008) NMR spectroscopy of the ligand-binding core of ionotropic glutamate receptor 2 bound to 5-substituted willardiine partial agonists., J. Mol. Biol. 378, 673-685. 15 37. Ahmed, A. H., Ptak, C. P., Fenwick, M. K., Hsieh, C. L., Weiland, G. A., and Oswald, R. E. (2013) Dynamics of cleft closure of the GluA2 ligand-binding domain in the presence of full and partial agonists revealed by hydrogen-deuterium exchange, J Biol Chem 288, 27658-27666. 38. Ahmed, A. H., Wang, S., Chuang, H. H., and Oswald, R. E. (2011) Mechanism of AMPA receptor activation by partial agonists: disulfide trapping of closed lobe conformations, J Biol Chem 286, 35257-35266. 39. Maltsev, A. S., Ahmed, A. H., Fenwick, M. K., Jane, D. E., and Oswald, R. E. (2008) Mechanism of partial agonism at the GluR2 AMPA receptor: Measurements of lobe orientation in solution., Biochemistry 47, 10600-10610.

Chapter 3

Determination of Ionization Effects and Protonation Events of

Willardiine Partial Agonists with a Glutamate Receptor

AMPA receptors are important therapeutic targets1. They mediate the majority of rapid excitatory synaptic transmission in the vertebrate CNS, and their dysfunction is involved in many neuropathologies2. The AMPAR ligand binding domain is where agonist binding sends information that can be translated into channel opening. Thus, understanding how the forces that drive its complex formation with ligands is therapeutically relevant.

The competitive binding of a series of 5-substituted willardiine analogues to the glutamate- bound GluA2 LBD in cacodylate buffer, (CH3)2AsO2Na was used in conjunction with the thermodynamic characterization in chapter 2 to determine how differences in the protonation state of the LBD-ligand complexes affect the observed thermodynamic parameters of binding.

There are no protonation differences between the LBD complexes with glutamate and analogues with highly electronegative substituents in their 5-position. However, the binding of willardiine and analogues with less electronegative substituents results in differences in the number of protonation and deprotonation events occurring within the protein-ligand complexes.

Excision of the enthalpic contributions of buffer protonation from the observed enthalpy changes for binding shows that at 20oC there are no significant differences between the enthalpy of binding to glutamate and FW, while the binding of willardiine and the remaining halogenated analogues occurs at a similar enthalpic cost.

Estimates suggest that at physiological temperature, 37oC, the competitive binding of willardiine and halogenated analogues is driven by favorable entropic change. However, the additional electrostatic interactions possible by the nitro substituent result in a competitive binding driven by favorable enthalpic change and occurring at significant entropic penalty.

Introduction

AMPA receptors are ligand-gated cation channels, and are responsible for most of the fast excitatory transmission in the vertebrate central nervous system3. They play multiple roles in neurotransmission, and their expression and trafficking is tightly regulated4. At the same time,

AMPAR function, expression and trafficking are altered in many neurophathologies and diseased states5. Increasing our understanding of the functional properties and structural dynamics of

AMPA receptors is necessary for the development of novel therapeutic compounds.

Drug design relies on understanding the forces that drive the ligand-binding process. In addition to details about the structural dynamics of the interaction partners, this requires a detailed description of the thermodynamics of the binding process6. This information can be experimentally determined through the use of calorimetric techniques, like isothermal titration calorimetry (ITC)7. Through them it has been determined that the binding of glutamate to an

AMPAR binding domain is an exothermic process, where both enthalpic and entropic components make significant contributions to the free energy of binding8. Further insight on the energetics of binding has been obtained through the use of willardiine derivatives9.

Willardiine, S(−)-α-Amino-3,4-dihydro-2,4-dioxo-1(2H)-pyrimidinepropanoic acid, is a natural product that acts as a partial agonist for glutamate receptors. A range of agonists and antagonists have been synthesized through substitutions to its pyrimidine moiety10,11. Substituting the hydrogen in the 3-position of the pyrimidine ring with carboxybenzyl and carboxyethyl groups result in antagonists who’s binding to AMPARs is largely driven by favorable enthalpy changes12.

At the same time, substitutions in the 5-position with halogens and other groups results in compounds with partial agonist activity10. Changes in the 5-position substituent of these partial

99 agonists lead to differences in the properties of the protein-ligand complex13–15. The stability of the bound complex, as well as the affinity, potency and efficacy of the ligand, vary with the electronegativity and size of the substituent13,16. This feature makes willardiine and its 5’- substituted analogues useful probes for studying partial agonist action.

The thermodynamic properties of these ligands are dependent on the protonation state of their uracil moiety, and thus on ligand pKa9. Deprotonation of the nitrogen position 3 on the uracil ring results in a protein-ligand complex with a more energetically favorable hydrogen bond complex and a more favorable enthalpic contribution to binding.

We characterized the effects of 5-position willardiine substituents (F, Cl, I, NO2, H) on the thermodynamics characteristics of competitive binding in cacodylate buffer, (CH3)2AsO2Na to glutamate-bound GluA2 through ITC.

Results were combined to those of binding in phosphate buffer under the same conditions, described in chapter 2, to determine the differences in protonation events between the LBD complexes bound to glutamate and the partial agonists, and the change in enthalpy of the competition binding process, ΔΔHbind.

The elimination of buffer ionization contributions to the changes in enthalpy demonstrates that there is no enthalpic difference between the glutamate and FW-bound states, and little enthalpic difference between the LBD complexes with HW, IW, and CW. Each protonation difference from the glutamate-bound state contributed or subtracted 1 kcal/mol to the ΔΔHobs.

There were no protonation differences between the glutamate, NW, and FW bound states, while there are differences of one proton in CW and IW bound states, and two protons for the HW bound form.

100

The thermodynamic parameters and heat capacity changes described in chapter 2 were used to estimate the thermodynamic properties of binding at physiological temperature. At 37oC, the competition binding of HW and the halogenated willardiines is driven by favorable entropy. The heat capacity differences between the glutamate-bound complex and those of HW and the halogenated analogues lead to higher binding affinities at 37oC than those measured at 20oC. The competitive binding of NW at 37oC is driven by large enthalpic change while incurring a large entropic cost. This may result in the free energy of binding that, unlike the other analogues studies, can be lower than that of binding at 20oC. Thus, the binding of NW at physiological conditions might occur with lower affinities than those determined at 20oC.

Methods

Protein Purification

The gluA2 LBD construct consisted of residues N392-K506 and P632-S775 of the full rat

GluA2-flop subunit with a glycine-threonine (GT) linker connecting K506 and P632, and a polyhistidine tag17. The DNA construct was inserted into a pet-22(+) plasmid and transformed into Origami B (DE3) E. coli. The protein was over-expressed by growing cultures of Origami B

o (DE3) E. coli at 37 C until they reached an OD600 of 0.7-1.0. At this point their temperature was lowered to 20oC and the culture flasks were supplemented with isopropyl-β-D-thiogalactoside

(IPTG) to a final concentration of 0.5mM, and kept at this temperature for 22 hours. Cells were collected by centrifuging the cultures 4,000 rpm for 30 min. and were frozen at -20o C for storage.

The frozen cells were thawed and resuspended in Tris/Mg buffer (20mM Tris, 5mM MgSO4,

150mM NaCl, 1mM glutamate, pH 8) containing 1.25 mM PMSF, 25µg/mL DNAse, 50 µg/mL

101 lysozyme, and 125 µg/mL deoxycholate, then lysed by stirring at 4oC. The lysis supernatant was clarified by centrifugation for 30 mins. at 35,000 rpm and injected into a His-Trap nickel column

(GE Healthcare) to separate the poly-histidine tagged GluA2 S1S2 from the supernatant. The protein was eluted using an imidazole gradient, and the protein was exchanged into a sodium acetate buffer (30mM NaOAc, 1mM EDTA, 1mM glutamate, pH 8, 2mM CaCl2). The polyhistidine tag was cleaved using thrombin digestion (2.5U/mg), and the protein purified using a HiTrap Sp HP ion exchange sepharose column (GE Healthcare). After purification, the protein was concentrated and dialyzed against phosphate (20mM sodium phosphate, 50 mM NaCl, 1mM

EDTA, 1mM NaN3, pH 7.2) or cacodylate (20mM sodium cacodylate, 50 mM NaCl, 1mM

EDTA, 1mM NaN3, pH 7.2) buffers, supplemented with 2-50µM glutamate.

Determination of the Extinction Coefficient of GluA2 S1S2 by Thiol Quantitation

Thiol quantitation (Thiol and Sulfide Quantitation Kit T-6060, Molecular Probes, and a methodology developed by Singh et al18,19, were used to determine the extinction coefficient at

280nm, ε280, of GluA2 LBD, a protein with 4 cysteine residues. Quantitating the thiols of a protein solution and simultaneously determining its A280 allows the use of A= 휀[퐶], where A is the absorbance of the sample, is the extinction coefficient, and [C] is the protein concentration, to obtain ε280.

An L-cysteine thiol standard solution, calibrated using Ellman’s assay, was used to create a thiol standard curve. In the assay, the thiol groups in the L-cysteine solution reacted chromogenically with with 5,5’-dithiobis-(2-nitrobenzoic acid), DTNB to yield mixed disulfides and 2-nitro-5-thiobenzoic acid (TNB). The apparent A412 of the solutions was measured

-1 -1 background corrected, then coupled to the TNB ε412, 13,600M cm , to determine the standard solution concentration. 102

The L-cysteine standard curve, GluA2 LBD, and control samples were exposed to a disulfide-inhibited derivative of papain. Thiol groups in the samples reduced the papain derivative and released the active enzyme stoichiometrically. This enzyme was used to hydrolyze an added substrate, Nα-Benzoyl-L-arginine 4-nitroanilide hydrochloride (L-BAPNA), at the bond between the arginine and p-nitroaniline groups.

The A410 of the chromophore p-nitroaniline was measured for all samples, and the control samples were used to correct experimental values. The A410 of the L-cysteine standard curve samples were plotted as a function of L-cysteine concentration and a linear regression of the plot was used to obtain the thiol concentration of the GluA2 LBD sample from its A410. The thiol concentration was used to determine the protein concentration in the solution and the A280 from

-1 the protein solution were used to calculate the GluA2 LBD ε280, 58363.2 ± 0.1 M .

Isothermal Titration Calorimetry (ITC)

Competition binding experiments were conducted on a Microcal VP-ITC calorimeter. The competitive binding approach increases the consistency of the data by improving the stability of

GluA2 LBD and avoiding errors due to incomplete glutamate removal. In this system, the glutamate-bound state of GluA2 LBD is considered the ground state, rather than the apo state.

Phosphate and cacodylate were selected as experimental buffers for their low ionization enthalpies, and pKa. Willardiine and its derivatives were obtained from Tocris and Abcam

Biochemicals. Stock solutions were made by solubilizing the ligand in cacodylate buffer and adjusting the pH. The experimental ligand solutions used the buffer from the protein dialysis or buffer exchange processes, reducing mismatch between the protein and ligand solutions.

The GluA2 S1S2 concentration in the experimental cell ranged from 8-40 µM, and ligand concentrations were kept at 15-20 times the protein concentration. The experiments were

103 conducted using a reference power of 10 µcal/sec while the syringe rotated at 300 rpm. Each experiment typically consisted of 35 injections of 6 µL every 240 seconds where the calibrated cell feedback signal (µcal/s) was collected at 2s intervals. Controls were performed for each experimental condition by injecting the ligand under study into an experimental cell containing

ITC buffer, using the same rates and volumes as the GluA2 LBD experiment. These controls minimize the effects of buffer mismatch and take into account any heat effect that arises from ligand dilution.

Data Analysis

Thermograms were analyzed using specialty Origin software. Control thermograms were subtracted from the experimental ones, and the resulting thermogram was used to determine the area under each injection. The resulting areas were plotted as a function of the molar ratio between the injected ligand and GluA2 LBD to obtain the binding isotherm. The areas derived from the first injection were not used during analysis, since they might have distortions due to diffusive mixing of the ligand in the tip of the syringe, or due to backlash error from the injector drive screw20.

The binding isotherms were analyzed using a previously described competitive binding method21. It utilizes the mathematical description of competitive binding, where two ligands A and B compete for the binding site in protein P 22.

Their binding equilibrium is described by:

푃 + 퐴 ⇌ 푃퐴 and 푃 + 퐵 ⇋ 푃퐵, (1)

where ligands A and B have affinity constants of

퐾퐴 = [푃퐴]⁄([푃][퐴]) and 퐾퐵 = [푃퐵]⁄([푃][퐵]). (2)

The conservation of mass for this system is given by: 104

[푨]풐 = [푨] + [푷푨] (3)

[푩]풐 = [푩] + [푷푩] (4)

[푷]풐 = [푷] + [푷푨] + [푷푩] (5)

Substituting equation 2 into equations 3 and 4, one can obtain:

[푷][푨] [푷][푩] [푷푨] = 풐 and [푷푩] = 풐 (6) ퟏ⁄푲푨+[푷] ퟏ⁄푲푩+[푷]

The mole of species containing [P] is defined as 푥푃 = [푃]⁄[푃]푂, 푥푃퐴 = [푃퐴] ∕ [푃]표, and

푥푃퐵 = [푃퐵] ∕ [푃]표 , while the molar ratios between the stoichiometric concentrations of the ligands and the protein are defined as 푟퐴 = [퐴]표⁄[푃]표 and 푟퐵 = [퐵]표⁄[푃]표. The product of the stoichiometric protein concentration and the ligand binding constants are given by 푐퐴 = 퐾퐴[푃]표

푐퐵=퐾퐵푃표. When these relations are substituted into equations 푷풐=푷+푷푨+푷푩

=푷푨풐ퟏ푲푨+푷 and 푷푩=푷푩풐ퟏ푲푩+푷 (6) result in:

풙푷 + 풙푷푨 + 풙푷푩 = ퟏ, (7)

풓푨풙푷 풙푷푨 = (8) ퟏ⁄풄푨+풙푷

풓푩풙푷 풙푷푩 = (9) ퟏ⁄풄푩+풙푷

풙푷푨=풓푨풙푷ퟏ풄푨+풙푷 (8 and

풙푷푩=풓푩풙푷ퟏ풄푩+풙푷 (9 can be integrated into 풙푷+풙푷푨 + 풙푷푩 = ퟏ,

(7 to obtain

ퟎ = [푷]ퟑ + 풂[푷]ퟐ + 풃[푷] + 푪 (10)

where:

풂 = 푲푨 + 푲푩 + [푨]ퟎ + [푩]ퟎ − [푷]ퟎ (11)

105

풃 = 푲푩([푨]ퟎ − [푷]ퟎ) + 푲푨([푩]ퟎ − [푷]ퟎ) + 푲푨푲푩 (12)

풄 = −푲푨푲푩[푷]ퟎ (13)

This cubic equation can be reduced by the substitution [P] = u - (푎⁄3). This leads to the equation

ퟎ = 풖ퟑ − 풒풖 − 풓, (14)

풂ퟐ 풒 = ( − 풃) (15) ퟑ

ퟐ ퟏ 풓 = − 풂ퟑ + 풂풃 − 풄 (16) ퟐퟕ ퟑ

To solve this equation we consider two numbers v and m where 푢 = 푚 + 푣 to obtain

(푚 + 푣)3 − 푞(푚 + 푣) − 푟 = 0. This equation can be rearranged and turned into a quadratic

푞 3 푟 푟 2 푞 3 function 푣6 − 푟푣3 + ( ) = 0 whose solution is 푣3 = ± √( ) − ( ) . The discriminant of 3 2 2 3 the solution, Δ, is given by:

풓 ퟐ 풒 ퟑ 횫 = ( ) − ( ) (17) ퟐ ퟑ

푟 2 푞 3 Since ( ) < ( ) the discriminant is less than zero, Δ < 0. When this occurs within a cubic 2 3

formula all the roots of the equations are real and the discriminant Δ is a complex

푟 number. The solution to the quadratic can be expressed as 푣3 = ± √훥 , and since √∆= 2

푟 √(−∆) 푖, it can be expressed as 푣3 = + √(−∆)푖. These are complex numbers of the form 푎 + 2

푏푖 and can be represented in trigonometric form as:

풋(퐜퐨퐬 흋 + 풊 퐬퐢퐧 흋). (18)

106

풓 풂 풃 In this equation 풋 = √풂ퟐ + 풃ퟐ, = , 풃 = √(−∆), 퐜퐨퐬 흋 = , and 퐬퐢퐧 흋 = . Therefore, 풋 = ퟐ 풋 ퟒ

풒 ퟑ 풓 √( ) , 퐜퐨퐬 흋 = , and ퟑ 풒 ퟑ ퟐ√( ) ퟑ

풓 흋 = 퐜퐨퐬−ퟏ ( ). (19) 풒 ퟑ ퟐ√( ) ퟑ

Using de Moivre’s formula on equation 풋(퐜퐨퐬 흋 + 풊 퐬퐢퐧 흋).

1 ( ) 휑 휑 (18 results in 푣 = 푗 3 (cos ( ) + 푖 sin ( )). Since the real roots 3 3 of v and m are equal, the three real roots of the equation are given by:

ퟐ 흋 풖 = √(풂ퟐ − ퟑ풃) 퐜퐨퐬 ( ) (20) ퟏ ퟑ ퟑ

ퟐ ퟐ흅−흋 풖 = √(풂ퟐ − ퟑ풃) 퐜퐨퐬 ( ) (21) ퟐ ퟑ ퟑ

ퟐ ퟐ흅+흋 풖 = √(풂ퟐ − ퟑ풃) 퐜퐨퐬 ( ) (22) ퟑ ퟑ ퟑ

The roots of equation ퟎ= 풖ퟑ − 풒풖 − 풓,

(14 need to be examined to determine which are physically meaningful, taking into account the physical conditions of the system under study. Since 푢 = (푎⁄3) + [푃] and [P] is the free protein concentration, [푃] = 푢 − (푎⁄3) > 0, and roots must satisfy 푢 > 푚푎푥{푎⁄3 , 0}.

When the angle φ is between 0 and π/2 the cosine of u1 is positive while the cosine of u3 is negative. This is also the case when π/2 < φ < π. Therefore we can discard u3 as a physically relevant root. The root u2 has 휋⁄3 < (2휋 − 휑)⁄3 < 2휋⁄3 for all values of φ. This means that the value of cos{(2휋 − 휑)⁄3} lies between -0.5 and 0.5. Whether the root is positive or negative depends on the value of a22. Since the root doesn’t satisfy 푢 > 푚푎푥{푎⁄3 , 0} is not physically

107 meaningful and can be discarded from consideration. Therefore u1 is the only root that references the system under study.

2 휑 Since 푢 = 푚 + 푣 = √(푎2 − 3푏)cos ( ), 3 3

풂 ퟐ√풂ퟐ−ퟑ풃 퐜퐨퐬(흋⁄ퟑ)−풂 풙 = 풖 − = (23) 푷 ퟑ ퟑ

−ퟐ풂ퟑ+ퟗ풂풃−ퟐퟕ풄 흋 = 퐜퐨퐬−ퟏ (24) ퟑ ퟐ√(풂ퟐ−ퟑ풃)

These are incorporated into equations 6 to obtain:

휑 휑 [퐴] {2√(푎2−3푏) cos( )−푎} [퐵] {2√(푎2−3푏) cos( )−푎} 푂 3 푂 3 [푃퐴] = 휑 and [푃퐵] = 휑 . 3퐾 +{2√(푎2−3푏) cos( )−푎} 3퐾 +{2√(푎2−3푏) cos( )−푎} 퐴 3 퐵 3

These relationships are used to fit the competitive binding isotherm. During a competitive binding ITC experiment, ligand A is injected into an experimental cell of volume Vo containing ligand B and protein P. The heat evolved after each injection is proportional to the changes in

[PA] and [PB] and their own molar enthalpies, ∆HA and ∆HB. The observed heat, ΔQ, can be defined as

∆푄 = 푉표(∆퐻퐴∆[푃퐴] + ∆퐻퐵∆[푃퐵]) =

∆푸 = 푽풐[푷]풐(∆풙푷푨∆푯푨 + ∆풙푷푩∆푯푩) (25)

Since the volume of the cell is constant, each injection causes a small volume of the solution in the experimental cell to be displaced. This results in the dilution of the species in the experimental cell. It is necessary to take this effect into account and incorporate it into the equation.

For an injection volume Vi, the change in concentration of a molecular species X within the experimental cell upon an infinitesimal change in volume is given by 푑[푋] = (− 푑푉푖⁄푉푂)[푋].

108

Integrating the equation with [X] between the limits [X]i-1 and [X]i and Vi from 0 to Vi results in

푉푖 [푋]푖 = [푋]푖−1푒푥푝 (− ). Within this function, the exponential factor is the dilution factor for a 푉푂 single injection:

푽풊 풇풊 = 풆풙풑 (− ). (26) 푽풐

After i injections, the total dilution factor is given by

ퟏ 풊 풇풕,풊 = 풆풙풑 (− ∑풋=ퟏ 푽풋). (27) 푽푶

For an A concentration in the syringe of [A]s, the stoichiometric concentration of A in the experimental cell after i injection is given by:

[푨]ퟎ,풊 = [푨]풔(ퟏ − 풇풕,풊), (28)

while the corresponding stoichiometric concentrations of B and P are

[푷]ퟎ,풊 = 풇풕,풊[푷]푶 (29)

and [푩]ퟎ,풊 = 풇풕,풊[푩]푶 (30)

These corrections is incorporated into ∆푸 = 푽풐[푷]풐(∆풙푷푨∆푯푨 + ∆풙푷푩∆푯푩)

(25 to yield

∆푸풊 = 푽푶[푷]푨(∆푯푨{풙푷푨,풊 − 풇풊풙푷푨,풊−ퟏ} + ∆푩{풙푷푩,풊 − 풇풊풙푷푩,풊−ퟏ}) + 풒풅, (31)

the regression function for least squares fitting of ITC competitive binding isotherms. The term qd refers to the heat of dilution. This parameter is also determined by blank experiments where ligand A is injected into a buffer solution. This dual collection of the heat of dilution helps identify when there are differences between the interactions in the blank buffer solution and a concentrated solution of P, so it is used as an additional parameter in the regression analysis.

109

Equations 8, 9, 11-13, 23, 24, 26-31 were integrated into the Origin Function Definition File obtained from Sigurskjold et al, and are used in the regression analysis of the competitive binding data 21.

For the experimental analysis, ligand A represents the willardiine analogue under study, while B represents the glutamate within the cell bound to GluA2 S1S2. The regression analysis uses the ligand and protein concentrations determined spectrophotometrically, and an association

6 -1 constant, KA, for glutamate of 1.0x10 M . Fitting of the binding isotherm yields the apparent enthalpy of binding ΔΔH, the KA of the willardiine analogue, and the binding stoichiometry n.

These three values are used to determine the ΔΔGo and ΔΔSo of the experiment. In order to obtain the change in heat capacity, ΔΔCp, of the binding reactions at physiological pH the binding reactions were carried out at multiple temperatures.

Results and Discussion

Competition binding of a series of willardiine analogues to the glutamate-bound GluA2 LBD was examined through the use of isothermal titration calorimetry in cacodylate buffer at pH 7.2 and 20oC. Since the experiments studied binding relative to the glutamate-bound form the thermodynamic parameters are expressed with the glutamate-bound form as the reference state

(ΔΔH, -TΔΔS, ΔΔG, and ΔΔCp).

The competitive binding of willardiine analogues in cacodylate buffer is similar to that in phosphate buffer 9,12. The binding of HW, FW, CW, and IW are endothermic processes white the binding of NW is exothermic (Figure 3.1).

110

ΔΔGo ΔΔHo Δ(-TΔS)o Ligand Kd (µM) n (kcal/mol) (kcal/mol) (kcal/mol)

NW -8.6±0.3 0.3±0.2 0.96±0.02 -10.7±0.2 2.1±0.4

FW -10.48±0.07 0.015±0.002 0.91±0.03 1.2±0.9 -12.1±0.1

CW -7.08±0.04 5.3±0.3 1.06±0.04 5.6±0.6 -12.7±0.6

IW -6.6±0.2 11±3 1.0±0.1 7.1±0.5 -13.8±0.6

HW -7.1±0.3 5±2 0.9 3±1 -10.4±0.7

Table 3.1 Thermodynamic parameters of competitive binding for willardiine analogues at pH 7.2, 20oC in cacodylate buffer.

In the thermograms shown, initial injections of CW into the experimental cell required 0.3-

0.35 µcals/sec of power to maintain the cell at 20oC and injections of NW reduced the amount of power necessary by 0.8-0.9 µcals/sec. The thermodynamic parameters of the competition binding of these analogues are shown in Table 3.1.

While the binding of HW, IW, CW, and FW is endothermic, there are large differences in their binding heats and other thermodynamic characteristics.

The thermodynamic parameters of binding to glutamate-bound GluA2 LBD are shown in

Table 3.. The binding affinity of the willardiine analogues varies with the electronegativity of the substituents with FW binding with the highest affinity and willardiine binding with the lowest.

This trend is reflected in the free energy change (ΔΔGo) of the displacement of glutamate. The

ΔΔGo of binding is always negative, meaning that the displacement reaction is energetically favored, and changes with the electronegativity of the substituents.

111

Figure 3.1 Thermogram and binding isotherm for the displacement binding of (A) NW and (B) CW at 20oC, pH 7.2 in cacodylate buffer

The free energy of binding is most favorable for FW, -10.48 kcal/mol, and lowest for IW, -

6.6 kcal/mol, With FW and NW having a more favorable change in free energy of binding than

HW and the rest of the halogenated series. The differences between their individual thermodynamic profiles are more pronounced.

With the exception of NW the displacement of glutamate by willardiine and its analogues comes at an enthalpic cost. The binding of HW is 3 kcal/mol less enthalpically favorable than the glutamate-bound complex. The fluoro substituent reduces the enthalpic cost of binding, but larger substituents like chloro and iodo increase the enthalpic cost (Figure 3.2).

112

The change in entropy, -TΔΔS, is favorable for the binding of willardiine and its analogues, with the exception of NW, whose competitive binding results in a –TΔΔS of 2.1 kcal/mol. The binding of HW is more entropically favorable than the glutamate-bound state by -10.4 kcal/mol.

The entropic changes of FW and CW are similar and favorable -12.1 and -12.7 kcal/mol, respectively, while the presence of the larger iodo substituent increases the favorable enthalpic change by -13.8 kcal/mol.

These parameters can be compared to those obtained in phosphate buffer, which were described in chapter 2. The competitive binding of HW has a different ratio of enthalpic and

o -1 entropic contributions to the free energy of binding. The ΔΔH app is 4 ± 2 kcal∙mol more

o - favorable than when binding in phosphate buffer (Table 3.2), but the -TΔΔS app is 3 ± 1 kcal∙mol

1 -1 o less favorable, resulting in no significant difference, 0.3 ± 0.5 kcal∙mol , between the ΔΔG app in the different buffers. The binding of NW in this buffer also results in little difference between the measured free energies of binding, but there is no significant difference between the enthalpic and entropic parameters of binding in cacodylate and phosphate buffers.

Figure 3.2. Thermodynamic parameters of competition binding in phosphate (A) cacodylate (B) buffers.

113

The binding of halogenated analogues in cacodylate buffer results in different thermodynamic properties compared to the binding in phosphate buffer (Figure 3.2)9. . The binding of FW has an enthalpy change similar to that of binding in phosphate buffer and a -

o -1 o TΔΔS app that is 1.4 ± 0.1 kcal∙mol more favorable, resulting in a ΔΔG app that is 0.4 ± 0.2 kcal∙mol-1 more favorable than that of binding in phosphate buffer. The binding of CW is 2.5 ±

0.9 kcal∙mol-1 less enthalpically favorable and 2.3 ± 0.7 kcal∙mol-1 more entropically favorable, resulting in no statistical difference between the free energy of binding in each buffer. This is not

o -1 the case for the binding of IW, where the ΔΔG app is 1.1 ± 0.4 kcal∙mol less favorable than the one from binding in phosphate buffer. This arises from differences in entropic and enthalpic

o -1 change, with as the binding if IW in cacodylate buffer has a ΔΔH app the 2.0 ± 0.2 kcal∙mol less

o -1 favorable and a -TΔΔS app 1.02 ± 0.3 kcal∙mol more favorable than binding in phosphate buffer.

Protonation Events and Buffer Ionization Effects at 20oC

Binding reactions are often coupled to changes in the protonation state of the system. When this happens, the measured heats contain the effects of the ionization of the buffer and any protonation and deprotonation events of the molecules involved in the binding interaction. This means that the contributions of these processes to the observed enthalpy change, ΔΔHobs depends on the ionization enthalpy of the buffer through:

∆∆푯풐풃풔 = ∆∆푯풃풊풏풅 + 풏푯∆푯풊풐풏 (32)

where ΔΔHbind is the change in enthalpy arising from the binding reaction, nH is the number of protons that are released (nH > 0) or absorbed (nH < 0) by the buffer, and ΔHion is the ionization enthalpy of the buffer. Performing binding experiments using multiple buffers allows the determination of ΔΔHbind and the number of protons released or absorbed during the binding

114 process. ΔΔHbind includes the enthalpic contribution of the protonation and deprotonation processes of the protein complex while excluding the contribution of buffer ionization. This allows a clearer look at the effect of substituting the glutamate γ-carboxyl group with a uracil moiety, and how modifying the properties of its 5-position affect the energetics of ligand binding.

To this effect, the thermodynamic parameters of binding in cacodylate and phosphate buffers

o at pH 7.2 and 20 C were used to determine the differences in nH and the changes in binding enthalpy in the absence of buffer ionization through the use of equation 32. The uncertainties of these results were determined through the propagation of error of all the experimental values.

-1 The phosphate buffer used in previous experiments has a ΔHion of 860 cal∙mol while the ΔHion of cacodylate is -717 cal∙mol-1. The results are shown in Table 3.2

ΔΔHobs ΔΔHbind ligand nH+ (kcal/mol) (kcal/mol)

NW -10.7 ± 0.2 -10.1 ± 0.6 0.8 ± 0.8

FW 1.2 ± 0.9 1 ± 1 -0.3 ± 0.6

CW 5.6 ± 0.6 4.5 ± 0.7 -1.6 ± 0.6

IW 7.1 ± 0.5 6.3 ± 0.5 -1.1 ± 0.3

HW 3 ± 1 5 ± 1 2 ± 1

Table 3.2. The observed enthalpy changes, ΔΔHobs, the change of enthalpy of competition binding , ΔΔHbind, and the number of protons exchanged with the buffering system, nH of the competition binding of willardiine analogues against glutamate.

115

The exchange of glutamate for HW results in the absorption of 2 protons from the protein- ligand complex into the buffer. Buffer protonation contributes 2.1 kcal/mol to the ΔΔHapp of competitive binding in phosphate buffer, resulting in a ΔΔHbind of 5 kcal/mol. The nH for the competitive binding of NW is 0.8 kcal/mol, which is the same magnitude as the uncertainty of the calculation. While the competitive binding of NW might result in the absorption of one proton by the buffer, it is also likely no proton exchange takes place. The calculated ΔΔHbind is -

10.1 kcal/mol, which is within error of the ΔΔHapp.

With a nH of -0.3 kcal/mol, the competition binding of FW results in no difference in protonation events than the glutamate-bound complex. The ΔΔHbind of FW results in no difference in protonation events than the glutamate-bound complex. The ΔΔHbind of FW is 1 kcal/mol, making its enthalpy of binding the most similar to that of the glutamate bound state.

CW and IW have an nH of -1.6 and -1.1 kcal/mol, respectively, indicating that the buffer absorbs a proton when the ligand binds to CW and IW. This absorption results in ΔHion of phosphate buffer ionization of ~ 1 kcal/mol for both binding reactions, and makes the ΔΔHapp of binding in this buffer more favorable than ΔΔHbind. Overall, each protonation or deprotonation event results in an enthalpy difference of ~1kcal/mol.

116

kcal/mol -5

-10

-15 NW FW CW IW HW

o Figure 3.2. Effects of buffer protonation on ΔΔH values at pH 7.2 and 20 C. ΔΔHobs (blue) and ΔΔHbind (red) of competitive binding to willardiine analogues.

The determination of protonation differences arising from the competitive binding of willardiine partial agonists shows that these events do not have a significant impact on the observed enthalpic changes. This means that the heat capacity changes determined in Chapter 2,

ΔΔCp (obs), can be used to calculate the thermodynamic parameters of binding in other temperatures.

Estimates of Binding Parameters at Physiological Temperatures

As discussed in chapter 2, the thermodynamic properties of willardiine partial agonists are temperature dependent9. The effect temperature has on binding parameter varies by agonist

(Figure 3.5). While NW binding is driven by favorable enthalpy, at 10 oC the binding is less enthalpically favorable and exhibits favorable entropic change. At 20 oC, the binding reaction is driven by favorable enthalpy and exhibits unfavorable entropic change. The temperature dependence of the binding parameters of the other willardiine analogues is less pronounced, but

117 still shows an increase in enthalpic favorability at higher temperatures, as well as a reduction in the magnitude of the entropic forces.

A B C

D E

Figure 3.3. Thermodynamic parameters of competitive binding of willardiine analogues as a function of temperature in phosphate buffer, pH 7.2, for: (A) HW, (B) NW, (C) FW, (D) CW, and (E) IW. Blue (dashed) and red (solid) lines represent fits to the data. Data was fitted with equation 33 for ∆∆H and 34 for ∆∆S. From the fit, values for ∆∆H0 and

∆∆S0 were obtained. ∆∆H0 (kcal/mol): HW, -2.9 ± 0.1; NW, -17.4 ± 0.6; FW, -0.20 ± 0.03; CW, 1.2 ± 0.3; IW, 2.6 ± 0.3.

-2 -2 -2 -2 ∆∆S0 (kcal/(mol K)): HW, (2.0 ± 0.2) x 10 ; NW, (-3.0 ± 0.2) x 10 ; FW, (3.00 ± 0.02) x 10 ; CW, (3.0 ± 0.2) x 10 ; IW, (3.0

± 0.1) x 10-2.

These calculations use apparent enthalpies of binding and thus result in apparent heat capacity changes, which include effects stemming from buffer ionization and protonation. The results are listed in Table 3.3. All heat capacity changes were negative, which is common in systems where there is hydrophobic burial.

118

Ligand NW FW CW IW HW

ΔΔCp -670 ± 80 -50 ± 10 -100 ± 20 -100 ± 10 -200 ± 80 cal∙mol-1∙K-1

Table 3.3. Heat capacity change for displacement binding of substituted willardiines to GluA2 LBD

The binding of HW results in a larger ΔΔCp than any of the halogenated willardiines. The

-1 -1 binding of FW results in a ΔΔCp of -50 cal∙mol ∙K while the binding of ligands with larger,

-1 -1 less electronegative halogens results in twice the ΔΔCp. The largest ΔΔCp, -670 cal∙mol ∙K , occurs upon binding to NW, and the large enthalpy-entropy compensation results in very similar

ΔΔG across temperatures.

The change in heat capacity can be used to estimate the ΔΔH of binding at other temperatures.

Within a narrow temperature range, the temperature dependence of ΔΔH is given by:

휟휟푯 = 횫횫퐇퐨 + 횫횫퐂퐩(퐓 − 퐓퐨) 33

while the temperature dependence of ΔΔS is given by:

푻 ∆∆푺 = ∆∆푺풐 + 휟횫푪풑 퐥퐧 ( ) 34 푻풐

where ΔΔHo and ΔΔSo are the binding enthalpy and entropy, respectively, at an arbitrary reference temperature To and ΔΔCp is the change in heat capacity of competitive binding. The uncertainty of the estimates was determined through the propagation of error of the uncertainty of the collected data.

If protein denaturation occurs at temperature T there will be an additional heat associated with protein refolding affecting the heat of binding such that: 119

∆∆푯풐풃풔 = ∆∆푯풃풊풏풅 + ∆∆푯풇풐풍풅 35

Where ΔΔHbind is the change in reaction enthalpy and ΔΔHfold is the change in the enthalpy of refolding between the glutamate- and willardiine analogue-bound LBD complexes. As the amount of unfolded protein increases with temperature, the ΔΔHfold contribution to ΔΔHobs increases. For this system, the contribution of the folding component to the observed enthalpy change is dependent on the difference between the melting point, Tm, of the glutamate-bound and analogue-bound complexes.

Thermal stability studies have shown that GluA2 LBD complexes with glutamate, HW, and

o 5-halogenated willardiine analogues have Tm > 37 C (Dr. Ahmed H. Ahmed, unpublished data), where the melting temperatures of the willardiine-bound complexes increase with increased electronegativity of the 5-substituent. These thermal properties suggest that any contributions from protein unfolding to the observed changes in the enthalpy of binding at physiological temperature, 37oC, are unlikely to be significant, and can be safely ignored in the determination of thermodynamic values at physiological temperatures.

Estimates of ΔΔH and -TΔΔS at 37oC were calculated to better understand the thermodynamic properties of binding to GluA2 LBD at physiological temperatures (Table 3.4).

At 37oC, the competition binding of HW is driven by favorable entropy. The heat capacity difference between the glutamate- and HW-bound complexes can lead to higher binding affinities at 37oC than those measured at 20oC (Figure 3.6). This is because the magnitude of the entropic penalty incurred at this temperature is more than compensated by a reduction of the enthalpic unfavorability of the reaction.

120

ΔΔHo -TΔΔSo ΔΔGo Ligand 37 37 37 (kcal/mol) (kcal/mol) (kcal/mol) NW -21 ± 2 13 ± 1 -8 ± 2 FW -0.2 ± 0.3 -10.4 ± 0.3 -10.6 ± 0.4 CW 1 ± 1 -9 ± 1 -8 ± 1 IW 3.6 ± 0.2 -11.9 ± 0.3 -8.3 ± 0.4 HW 4 ± 2 -11 ± 2 -7 ± 2 Table 3.4 Estimates of the observed thermodynamic parameters of competitive binding at 37oC

Binding to the halogenated willardiine analogues at 37oC would result in similar changes in the thermodynamic binding properties, but their smaller ΔCp leads to smaller changes in their enthalpic and entropic change parameters across temperatures.

Figure 3.4. Thermodynamic parameters of the competitive binding of willardiine analogues determined at 20oC (A) and estimated at 37oC (B) in phosphate buffer.

The binding of NW produces the largest change in thermodynamic parameters across temperatures. While the binding at 10oC is driven mainly by favorable changes in entropic forces, relative to the glutamate-bound state, these changes become less favorable at 15oC, and at

20oC the binding reaction is driven exclusively by enthalpic forces while incurring a small entropic cost. At 37oC the binding process will be driven by large enthalpic change, but will incur a large entropic cost. This may result in the free energy of binding that, unlike the other

121 analogues studies, can be lower than that of binding at 20oC. Thus, the binding of NW at physiological conditions might occur with lower affinities than those determined at 20oC.

ΔΔH37 (obs) ΔΔH37 (bind) ligand (kcal/mol) (kcal/mol)

NW -21±2 -21 ± 2

FW -0.2±0.3 -0.2 ± 0.3

CW 1±1 1.3 ± 0.7

IW 3.6±0.2 3.5 ± 0.2

HW 4±2 4 ± 2

Table 3.5. Estimates of the differences in enthalpic change upon competitive binding at 37oC.

Assuming that the protonation differences that occur during competition binding are not temperature dependent, one can use the heat capacity of the buffer system to determine the contribution of buffer ionization to the enthalpic changes of binding at physiological temperatures through the use of equation 33. This value can be incorporated into equation 32 to

37 37 determine ∆∆퐻푏푖푛푑 from ∆∆퐻표푏푠. The comparison of these values is shown in Table 3.5, where the uncertainties of the estimates were determined through the propagation of the uncertainties of

o ΔΔHobs at 20 C, ΔΔCp, and nH. These calculations confirm that at physiological temperatures buffer ionization effects have a negligible contribution to the enthalpy changes of competitive binding.

The displacement of glutamate by most willardiine partial agonists is driven, at RT and physiological pH, by favorable entropic change and occurs at enthalpic cost. The magnitude of

122 the differences with the glutamate-bound state is influenced by the hydrogen bonding properties and electronegativity of the 5-substituent.

The competitive binding to willardiine analogues from a glutamate-bound state can involve changes in the protonation state of the protein-bound complexes and free ligands in solution.

While the buffer gains protons after willardiine competitive binding, the addition of 5- substituents to the ligand can result in the release of buffer protons into the protein-ligand complex or in no statistical difference between the protonation of the buffer in the glutamate and analogue-bound states.

These protonation processes contribute to the enthalpic differences of the individual ligand- bound states, as well as their changes in apparent heat capacity. However, the magnitude of the contribution of buffer ionization to these parameters is small, indicating that the bulk of the energetic differences that arise from protonation or deprotonation events are included in ΔΔHbind parameters. The characterization of heat capacity changes and protonation effects across different pH values would allow the determination of the energetics of protein-ligand protonation and deprotonation. These energetic contributions can be subtracted from ΔΔHbind to determine the intrinsic thermodynamic parameters of competition binding.

The thermodynamic parameters of competition binding vary across temperatures, although there is little difference between the changes in free energy of binding. This variability is described by the change in heat capacity, ΔΔCp. ΔΔCp is negative for the competitive binding of all analogues. Halogen substituents decreased the magnitude of ΔΔCp, while substitution with a nitro moiety increased the magnitude by three times.

Extrapolation of the thermodynamic parameters determined through the use of the change in heat capacity that competitive binding at 37oC may result in a decrease of enthalpic penalties for

123 the binding of willardiine and its halogenated analogues while decreasing the entropic favorability of binding to a lesser extent. The addition of a nitro moiety in the 5-position results in a ligand whose competitive binding at physiological temperatures may be driven by enthalpic forces and occur at significant entropic cost.

References

1. Franciosi, S. AMPA receptors: potential implications in development and disease. Cell.

Mol. Life Sci. 58, 921–30 (2001).

2. Gereau, R. G. S. The Glutamate Receptors: AMPA Receptors. Humana Press, Totowa,

N.J. (2008).

3. Mayer, M. L. & Armstrong, N. Structure and function of glutamate receptor ion channels.

Annu. Rev. Physiol. 66, 161–81 (2004).

4. Bassani, S., Valnegri, P., Beretta, F. & Passafaro, M. The GLUR2 subunit of AMPA

receptors: synaptic role. Neuroscience 158, 55–61 (2009).

5. Chang, P. K.-Y., Verbich, D. & McKinney, R. A. AMPA receptors as drug targets in

neurological disease--advantages, caveats, and future outlook. Eur. J. Neurosci. 35, 1908–

16 (2012).

6. Chaires, J. B. Calorimetry and thermodynamics in drug design. Annu. Rev. Biophys. 37,

135–51 (2008).

124

7. Hansen, L. D., Russell, D. J. & Choma, C. T. From Biochemistry to Physiology: The

Calorimetry Connection. Cell Biochem. Biophys. 49, 125–140 (2007).

8. Madden, D. R., Abele, R., Andersson, a & Keinänen, K. Large-scale expression and

thermodynamic characterization of a glutamate receptor agonist-binding domain. Eur. J.

Biochem. 267, 4281–9 (2000).

9. Martinez, M., Ahmed, A. H., Loh, A. P. & Oswald, R. E. Thermodynamics and

mechanism of the interaction of willardiine partial agonists with a glutamate receptor:

implications for drug development. Biochemistry 53, 3790–5 (2014).

10. Jane, D. E. et al. Synthesis of willardiine and 6-azawillardiine analogs: pharmacological

characterization on cloned homomeric human AMPA and kainate receptor subtypes. J.

Med. Chem. 40, 3645–50 (1997).

11. Dolman, N. P. et al. Structure-activity relationship studies on N3-substituted willardiine

derivatives acting as AMPA or kainate receptor antagonists. J. Med. Chem. 49, 2579–92

(2006).

12. Ahmed, A. H. et al. Mechanisms of antagonism of the GluR2 AMPA receptor: structure

and dynamics of the complex of two willardiine antagonists with the glutamate binding

domain. Biochemistry 48, 3894–903 (2009).

13. Jin, R. & Gouaux, E. Probing the function, conformational plasticity, and dimer-dimer

contacts of the GluR2 ligand-binding core: studies of 5-substituted willardiines and GluR2

S1S2 in the crystal. Biochemistry 42, 5201–13 (2003).

125

14. Fenwick, M. K. & Oswald, R. E. NMR spectroscopy of the ligand-binding core of

ionotropic glutamate receptor 2 bound to 5-substituted willardiine partial agonists. J. Mol.

Biol. 378, 673–85 (2008).

15. Ramaswamy, S. et al. Role of conformational dynamics in α-amino-3-hydroxy-5-

methylisoxazole-4-propionic acid (AMPA) receptor partial agonism. J. Biol. Chem. 287,

43557–64 (2012).

16. Poon, K., Nowak, L. M. & Oswald, R. E. Characterizing single-channel behavior of

GluA3 receptors. Biophys. J. 99, 1437–1446 (2010).

17. Armstrong, N., Sun, Y., Chen, G. Q. & Gouaux, E. Structure of a glutamate-receptor

ligand-binding core in complex with kainate. Nature 395, 913–7 (1998).

18. Singh, R., Blättler, W. A. & Collinson, A. R. Assay for thiols based on reactivation of

papain. Methods Enzymol. 251, 229–37 (1995).

19. Singh, R., Blättler, W. A. & Collinson, A. R. An amplified assay for thiols based on

reactivation of papain. Anal. Biochem. 213, 49–56 (1993).

20. Mizoue, L. S. & Tellinghuisen, J. The role of backlash in the ‘first injection anomaly’ in

isothermal titration calorimetry. Anal. Biochem. 326, 125–7 (2004).

21. Sigurskjold, B. W. Exact analysis of competition ligand binding by displacement

isothermal titration calorimetry. Anal. Biochem. 277, 260–266 (2000).

126

22. Wang, Z.-X. An exact mathematical expression for describing competitive binding of two

different ligands to a protein molecule. FEBS Lett. 360, 111–114 (1995).

127

Chapter 4

Concluding Remarks and Future Outlook

128

Research Summary

This dissertation focused on the determination of the thermodynamic properties of competitive

binding of a series of partial agonists to the ligand binding domain (LBD) of the AMPA receptor

subunit GluA2 through the use of Isothermal Titration Calorimetry (ITC). This technique plays

an important role in rational drug design, as it helps identify useful lead compounds and optimize

their structure to improve their pharmacological properties.

The thermodynamic information collected reflects the difference between the glutamate-GluA2

LBD complex and those formed by the binding of willardiine analogues. The combination of the

calorimetric information described in this dissertation with 3D structural information of the

protein-ligand complexes has yielded insight into the energetic consequences of binding to

AMPARs under a variety of conditions.

The complexes formed by willardiine analogues at 20oC are more entropically favorable than

those formed by glutamate. The pKa of the 3-position of the willardiine uracil moiety determines

whether the binding reaction is exothermic or endothermic, and the favorability of the enthalpic

contributions to the free energy of binding. When willardiine analogues have an h-bond acceptor

in their 3-position they create a more energetically favorable H-bond network with lobe 2 within

the binding cleft. These stronger interactions with lobe 2 could provide increased stability to the

closed lobe forms of the protein, while increasing the enthalpy of binding at entropic cost.

The complexes formed by willardiine binding have a more negative heat capacity than those of

glutamate-LBD complexes. This is consistent with their burial of larger polar surfaces. While

there are differences in the enthalpic and entropic parameters of binding across temperatures for

129

willardiine analogues with uncharged uracil moieties, binding of the analogue with a charged

uracil moiety increases the magnitude of this difference in heat capacity, and leads to

thermodynamic parameters of binding that change dramatically across temperatures. This results

in a binding process driven at 10oC by favorable entropic and enthalpic contributions and driven

at 37oC by large, favorable enthalpic forces while occurring at a large entropic cost.

The excision of the buffer ionization effects from the enthalpic components of binding confirmed

that the trends seen for the enthalpic contributions to the free energy of binding are due to

changes in the energetic properties of the protein-ligand complexes, and not due to the

contributions of the buffering system to the thermodynamics of binding.

Determining the contribution of buffer ionization to observed enthalpic parameters of binding

also allowed the determination of protonation differences between the LBD complexes with

glutamate and the willardiine analogues. While binding of willardiine and its analogues with less

electronegative substituents results in differences in protonation, ligands with more

electronegative substituents yield complexes with no differences in protonation to those of the

glutamate-bound complexes.

Overall, these studies highlight the importance of hydrogen bonding arrangement on the

thermodynamic properties of binding to AMPARs. It also illustrates how differences in the

characteristics of LBD-ligand interactions are masked by enthalpy-entropy compensation, but

can become apparent through the measurement of the enthalpic properties of binding.

130

Further Research

Our understanding of the the energetic properties of binding to AMPAR LBDs could be further improved by further experimentation. Among these is studying the thermodynamic properties of binding to the apo LBD.

These studies have been hampered by the lower stability of the apo protein, and the difficulty of removing all glutamate from the protein solution. The elimination of glutamate from the solution could be improved through digestion using immobilized enzymes. Proteins such as glutamate decarboxylase (GAD) have been immobilized and used to generate gamma- aminobutyric acid (GABA)1. Using this technique would require assaying for conditions which can maintain GluA2 stability, the catalytic activity and efficiency of GAD, and ensure that there is no non-immobilized GAD before and after the digestion process. A possible drawback could be loss of activity after GAD reuse.

Recent efforts to improve apo protein preparations increased construct stability by purifying the GluA2 LBD construct in L-aspartate2. Crystallization of the purified protein yielded a mixed dimer, with a glutamate-bound monomer, at lower aspartate concentrations. Similar purification schemes could be assayed to determine if they can improve the yield of apo preparations without selecting for LBD populations that have retained glutamate from the expression system.

The thermodynamic parameters of the binding of glutamate to the apo state could provide information of the thermodynamics of the docking processes shared by glutamate and the ligands under study. They could also help determine the overall energetics of willardiine analogues binding to an empty binding domain.

131

Further analysis of the enthalpic properties of binding, and the differences between the protonation or deprotonation events between the glutamate and willardiine complexes is necessary in order to dissect the energetics of electrostatic interactions with the GluA2 LBD.

The determination of the thermodynamic characteristics of binding across a range (> 3) of pH values in cacodylate and phosphate buffers would allow the determination of the pKa of protonation and deprotonation events responsible for the differences in nH at pH 7.2. The coupling of this information to structural data could pinpoint the molecular sources of these protonation differences3.

Through the determination of nH at various temperatures one can also obtain the intrinsic difference in binding, ΔΔHint, and the proportion of ΔΔHbind contributed by protonation or deprotonation events. The determination of these parameters allows in turn the calculation of the enthalpic change of protonation or deprotonation of the glutamate-LBD and willardiine-LBD complexes.

A deeper study of the effects of pH in the thermodynamic properties of binding may allow the dissect the energetics of electrostatic interactions behind the differences in protonation and deprotonation events, and shed light on how these differences might alter the h-bonding networks between the ligand and the residues within the binding cleft.

References

1. Lee, S. et al. Gamma-aminobutyric Acid production using immobilized glutamate

decarboxylase followed by downstream processing with cation exchange chromatography.

Int. J. Mol. Sci. 14, 1728–39 (2013). 132

2. Krintel, C. et al. L -Asp is a useful tool in the purification of the ionotropic glutamate

receptor A2 ligand-binding domain. FEBS J. 281, 2422–2430 (2014).

3. Baker, B. M. & Murphy, K. P. Evaluation of linked protonation effects in protein binding

reactions using isothermal titration calorimetry. Biophys. J. 71, 2049–55 (1996).

133