A Dissertation

entitled

Molecular Insights into N-acetylaspartate Metabolism in

by

Yasanandana S. Wijayasinghe

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the

Doctor of Philosophy Degree in Chemistry

______Dr. Ronald E. Viola, Committee Chair

______Dr. Max O. Funk, Committee Member

______Dr. Donald Ronning, Committee Member

______Dr. Paul Erhardt, Committee Member

______Dr. Patricia R. Komuniecki, Dean College of Graduate Studies

The University of Toledo

December 2014

Copyright 2014, Yasanandana S. Wijayasinghe

This document is copyrighted material. Under copyright law, no parts of this document may be reproduced without the expressed permission of the author. An abstract of

Molecular Insights into N-acetylaspartate Metabolism in Canavan Disease

by

Yasanandana S. Wijayasinghe

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the Doctor of Philosophy Degree in Chemistry The University of Toledo December 2014

Canavan disease (CD) is a fatal, childhood with altered N- acetylaspartate (NAA) metabolism. This disease is caused by in the that encodes the (ASPA) in the . The deficiency of aspartoacylase activity leads to the hallmark symptoms of CD. Sixteen different

ASPA clinical mutations have been examined for their biophysical properties. Each of these recombinant mutant was found to have measureable catalytic activity, ranging from 0.3% to 35% of the native enzyme. Many of these mutants show either decreased thermal stability or decreased conformational stability. These results suggest that the loss of catalytic activity of ASPA is at least partially a consequence of decreased stability.

To study the possible structural defects caused by non-conservative changes in ASPA, four aspartoacylase mutations have been structurally characterized.

The mutant enzymes each have similar overall structures to the native ASPA enzyme, but with varying degrees of alterations that offer explanations for the respective loss of catalytic activity. The loss of van der Waals contacts in the F295S , and the loss of both hydrophobic and hydrogen bonding interactions in the Y231C mutation lead to a local collapse of the hydrophobic core structure contributing to a

iii decrease in protein stability. The E285A mutant structure shows that loss of interactions with the side chain carboxyl of Glu285 disturbs the architecture, leading to altered binding and diminished enzymatic activity.

Since many of the ASPA clinical missense mutations were found to have compromised stability, screening of a small library was performed to identify compounds that can potentially bind and stabilize these defective enzymes.

Three compounds (Glutathione, Patulin, and Ellagic acid) show significant protection against thermal denaturation of E285A, but they have negligible effect on the native enzyme. Optimization of these compounds, guided by kinetics and structural studies, could lead to drugs that have the potential to reverse the effect of these mutations in

Canavan patients.

N-acetylaspartate is an enigmatic metabolite whose biological function is yet to be completely defined. In order to understand the physiological role of NAA in the brain and to find a treatment therapy for CD, the NAA biosynthetic enzyme aspartate

N-acetyltransferase (AspNAT) was selected for examination. A codon optimized gene of AspNAT (NAT8L) was cloned with different fusion tags to enhance the .

The thioredoxin-tagged AspNAT was found to express as a soluble protein with low yield. This enzyme construct was partially purified using Ni affinity chromatography and was found to have N-acetyltransferase activity. Further optimizations are being carried out to obtain sufficient enzyme for biochemical characterization.

iv

To my parents who made me who I am today by giving me the best education they could think of. This dissertation is the harvest of your endless sacrifices, blessings, and love.

Acknowledgements

First and foremost I wish to thank my research adviser and mentor Prof. Ronald

Viola for his invaluable guidance and support throughout my PhD training. His motivation and commitment to science inspired me and enriched my passion for research. He gave me the freedom to explore science and made me learn from the failures. I was fortunate enough to have a great person like him as my advisor.

I would also like to thank past and present members of Viola research group, in particular Dr. Alexander Pavlovsky for his support in determination, and Drs. Stephen Zano, Nitesh Poddar, Mojun Zaho, and Pravin Bhansali for their help in various aspects of my research.

I thank all the professors who taught me and thank the Department of Chemistry and Biochemistry at the University of Toledo, USA for many years of support and assistantship.

Thank you to Dr. L.M.V. Tillekeratne, and all my Sri Lankan friends in and around Toledo, Ohio for their support that made my stay in USA a pleasant and memorable one.

Last but not least, I wish to express my heartfelt gratitude to my loving wife

Thanuja, son Sumiyuru, and daughter Isora for their consistent love, encouragement, and understanding. Without their strength and this much of success would not have been accomplished in such a short period of time.

vi

Contents

Abstract iii

Acknowledgements vi

Contents vii

List of Tables x

List of Figures xi

List of Abbreviations xiii

1 Introduction 1

1.1 Canavan Disease……………………………………………………… 1

1.2 N-acetylaspartate Metabolism………………………………………... 3

1.3 Aspartoacylase Deficiency…………………………………………… 6

2 Biophysical Characterization of Aspartoacylase Clinical Mutations 9

2.1 Selection of Aspartoacylase Mutants………………………………... 13

2.2 Cloning and Site Directed Mutagenesis……………………………... 13

2.3 Expression and Purification………………………………………….. 15

2.4 Enzymatic Activities of Aspartoacylase Mutants…………………… 17

vii

2.5 Protein Stability Studies……………………………………………... 21

2.5.1 Thermodynamic Stability…………………………………. 21

2.5.2 Conformational Stability………………………………….. 24

2.6 Summary…………………………………………………………….. 26

3 Structural Characterization of Aspartoacylase Mutants 27

3.1 Protein Purification and Crystallization…………………………….. 29

3.2 Data Collection and Structure Determination………………………. 32

3.3 Comparison of Oligomeric States…………………………………... 36

3.4 Structural Roles of the Mutation Sites……………………………… 38

3.5 Structural Differences between the Aspartoacylase Mutants……….. 40

3.6 Active Site Alterations………………………………………………. 46

3.7 Structural Effects on Enzyme ……………………………... 51

3.8 Summary……………………………………………………………... 56

4 Pharmacological Chaperone Approach for the Treatment of Canavan Disease 57

4.1 High-Throughput Thermal Stability Assay………………………….. 61

4.2 Assay Optimization………………………………………………….. 62

4.3 Compound Library Screening and Hit Validation…………………... 64

4.4 Effect of Hit Compounds on the Catalytic Activity………………... 69

4.5 Virtual Screening of Compound Libraries………………………….. 70

4.6 Summary and Future Directions…………………………………….. 72

viii

5 Biochemical Characterization of Aspartate N-acetyltransferase 73

5.1 Cloning of Aspartate N-acetyltransferase………………………….... 77

5.2 Expression of AspNAT…………………………………………….... 79

5.3 Purification of AspNAT…………………………………………….. 82

5.4 Kinetic Characterization of AspNAT……………………………….. 84

5.5 Summary and Future Directions…………………………………….. 86

References 87

A Supplemental Data for Chapter 2 100

B Supplemental Data for Chapter 3 103

C Supplemental Data for Chapter 4 106

D Supplemental Data for Chapter 5 109

ix

List of Tables

2.1 Sixteen aspartoacylase clinical mutants selected in the present study……… 13

2.2 Biochemical properties of aspartoacylase clinical mutants…………………. 20

3.1 X-ray data collection and refinement statistics……………………………… 34

4.1 Oligonucleotide primers used in amplification of NAT8L ……………. 77

A.1 DNA primers used in site directed mutagenesis of ASPA gene…………….. 104

x

List of Figures

1-1 Schematic presentation of N-acetylaspartate metabolism in the brain…….... 4

1-2 Cartoon representation of the ASPA crystal structure……………………… 7

2-1 Mapping of the missense CD mutations to the native ASPA enzyme…….... 11

2-2 Evaluation of PCR amplified mutant ASPA-pPICZ A on 1% agarose gel….. 14

2-3 Aspartoacylase purification profile………………………………………….. 16

2-4 Schematic representation of L-aspartase coupled ASPA activity assay…….. 18

2-5 A typical denaturation curve obtained in the thermal stability study……….. 22

2-6 Thermal stability of the clinical mutants of ASPA………………………….. 22

2-7 A typical conformational stability curve…………………………………….. 24

2-8 Conformational stability of the clinical mutants of ASPA………………….. 25

3-1 ASPA-intermediate analog binary complex crystals……………………….... 31

3-2 Diffraction pattern of aspartoacylase clinical mutants………………………. 32

3-3 2Fo-Fc difference electron density map for the mutated residues…………... 35

3-4 Structural comparison of the ASPA E285A with the native enzyme...... 37

3-5 Structural comparison of the ASPA E285A with the native enzyme...... 39 xi

3-6 The overall structures of aspartoacylase native and mutant enzymes……….. 41

3-7 Comparison of the K213E structure with ASPA wild type structure……….. 42

3-8 Comparison of the Y231C structure with the native ASPA enzyme………... 42

3-9 Structural comparison of ASPA F295S with the native enzyme…………….. 45

3-10 Comparison of the active site geometry of ASPA…………………………. 48

3-11 Conformational changes of Arg71 and Lys291……………………………. 50

3-12 Comparison of the catalytic center of the ASPA clinical mutants…………. 52

4-1 Schematic representation of the cellular protein homeostasis network……... 58

4-2 Thermal unfolding obtains in differential scanning fluorimetry….. 62

4-3 Thermal denaturation curves obtained for the native ASPA enzyme………... 64

4-4 The melting temperature shift observed in high-throughput screening……… 65

4-5 Concentration dependent stabilization of ASPA native and E285A enzymes.. 67

4-6 The binding of probe to the native and E285A enzymes…………. 71

5-1 Schematic representation of AspNAT protein and the proposed model……... 75

5-2 PCR amplification of NAT8L genes as seen on 1% agarose gel……………... 78

5-3 Expression profiles of AspNAT as seen on SDS-PAGE and Western blot….. 81

5-4 The chromatogram and SDS-PAGE for purification of Trx-AspNAT………. 83

5-5 A schematic representation of AspNAT coupled enzyme activity assay…….. 85

xii

List of Abbreviations

1,8-ANS…………. 1-anilinonapththalene-8-sulfonic acid AOX…………….. Alcohol oxidase ASPA…… ……… Aspartoacylase AspNAT…………. Aspartate N-acetyltransferase ATP……………… Adenosine Triphosphate CCD……………... Charge-Coupled Device CD…….………… Canavan disease CNS……. ………. Central Nervous System DLS……………... Dynamic Light Scattering DMSO …………... Dimethyl Sulfoxide DSC……………... Differential Scanning Calorimetry GABA…………… Gamma-amino butyric acid GC/MS…..………. Gas Chromatography – Mass Spectrometry IMAC……………. Immobilized Metal Affinity Chromatography IPTG……………... Isopropyl β-D-1-thiogalactopyranoside mGluR3…….……. Metabotropic glutamate receptor type 3 MRS……….…….. Magnetic Resonance Spectroscopy NAA…….………. N-acetylaspartate NAAG…………… N-acetylaspartylglutamate NaDC3…..………. Na+ dependent dicarboxylate transporter 3 NAG……………... N-acetylglutamate Ni-NTA………….. Ni-Nitrilotriacetate NMR…… ……….. Nuclear Magnetic Resonance OMIM……..…….. Online Mendelian Inheritance in Man PCR……………… Polymerase Chain Reaction PCs………………. Pharmacological Chaperones PDB ID………….. Identification rAAVs …………… recombinant adeno-associated viral vectors RMSD…………… Root-mean-square deviation SDS-PAGE……… Sodium Dodecyl Sulphate-Polyacrylamide Gel Electrophoresis Trx……………….. Thioredoxin

xiii

Chapter 1

Introduction

1.1 Canavan Disease

Neurological disorders are a spectrum of devastating diseases that can have an impact on the normal functions of brain and nervous system in people of all ages.

Among them Canavan disease (CD; OMIM #271900) is a tragic neurodegenerative disorder that affects the developing brain of newborns.1-3 This disease was first described as a spongy degeneration in the of the brain by Myrtelle May

Moore Canavan in 1931, and was later identified as a genetic disorder that follows the autosomal recessive inheritance by van Bogaert and Bertrand in 1949.4, 5 The affected children usually appear healthy at birth, but the clinical symptoms progressively develop starting at about 3 months of age (infantile form).6 The congenital form

(diseased at or shortly after birth) and juvenile form (disease appears after 4 – 5 year of age) of CD are relatively uncommon.7, 8 The clinical symptoms of Canavan disease include development delay, mental retardation, increased head circumference

(macrocephaly), decreased muscle tone (hypotonia), poor head control, and seizures.

The failure of the brain to develop properly often results in death in the first decade of life.6, 9 Fortunately, incidences of CD are rare, but Canavan patients have been

1 identified all over the world, with the highest frequency among the people of

Ashkenazi Jewish descent.5, 10-12

The irreversible collapse of brain structure in Canavan disease is associated with loss of neuronal sheath (demyelination/dysmyelination), swelling of astrocytes, and vacuolization in the central nervous system (CNS).7, 13 These pathological changes can be observed in brain biopsies and were used to confirm the diagnosis of CD. Apart from the above mentioned anatomical changes, Canavan disease is characterized biochemically by the unusually elevated acidic metabolite N- acetylaspartate in the body fluids, and its excessive (10 – 100 fold higher) excretion in the urine (N-acetylaspartic aciduria).14-16 Estimation of NAA levels by nuclear magnetic resonance (NMR) spectroscopy in the brain and by gas chromatography – mass spectrometry (GC/MS) in urine are now reliably used in the diagnosis of CD.17

In 1988, Matalon et al. identified the co-occurrence of the elevated urinary NAA and a deficiency of the metabolic enzyme aspartoacylase (ASPA, EC 3.5.1.15) in

Canavan patients.10 This enzyme is responsible for the hydrolysis of N- acetylaspartate,15 and was found to be localized in the white matter of the brain.18

Thus, the inability of aspartoacylase to efficiently breakdown NAA leads to the depletion of the reaction products L-aspartate and acetate in the brain and the concomitant accumulation of its substrate N-acetylaspartate in the body.

2

1.2 N-acetylaspartate Metabolism

N-acetylaspartate is the second most concentrated amino acid metabolite in the mammalian central nervous system after L-glutamate.19 In the brain NAA is present at millimolar levels and its concentration can exceed 10 mM in the .20-22 This very high abundance of brain NAA leads to a strong proton nuclear magnetic resonance signal of the acetyl group, allowing the in vivo estimation of NAA levels by non-invasive brain magnetic resonance spectroscopy (MRS).23 Pathological variations of brain NAA levels have been observed in several neurological disorders such as multiple sclerosis, Alzheimer’s disease, brain tumors, brain ischemia, and

Canavan disease.24 Therefore, N-acetylaspartate is used as a reliable biological marker for neuronal health and integrity.25, 26

N-acetylaspartate is an amino acid derivative whose metabolism is confined to two cellular compartments in the brain. NAA is synthesized through the N-acetylation of L-aspartate by a membrane bound enzyme, aspartate N-acetyltransferase

(AspNAT, EC 2.3.1.17) found in the mitochondria27, 28 and endoplasmic reticulum29 of the neurons. De novo synthesized anionic NAA is then released from the neurons and is taken up by the glial cells through a specific, high-affinity sodium coupled dicarboxylate transporter (NaDC3).30, 31 Subsequently, N-acetylaspartate is hydrolyzed by aspartoacylase to produce L-aspartate and acetate in the , the myelin forming cells of the CNS (Figure 1-1).

3

L-aspartate N-acetylaspartate Acetyl-CoA CoA O O O O Aspartate N-acetyltransferase O O (AspNAT) O HN O NH3

O Neurons

NaDC3 transporter Oligodendrocytes O

O Aspartoacylase O

O + (ASPA) O NH3

O

Acetate L-aspartate N-acetylaspartate

Figure 1-1: Schematic presentation of N-acetylaspartate metabolism in the brain. N-acetylaspartate (NAA) is biosynthesized from L-aspartate and acetyl-CoA by aspartate N-acetyltransferase (AspNAT) in the neurons. After transport into the oligodendrocytes through a Na+ dependent dicarboxylate transporter (NaDC3), NAA is hydrolyzed by aspartoacylase (ASPA) to regenerate L-aspartate and release acetate.

4

Despite the very high abundance of N-acetylaspartate in the brain, its definitive biological functions in the CNS and its role in the pathogenesis of Canavan disease are still poorly understood. A number of theories have been proposed to explain the physiological role of this enigmatic molecule. Three are described herein. Hypothesis

1: NAA as a source of acetate for myelin lipid .32-34 According to this concept, NAA serves as a carrier for the acetyl group that is shuttled from to . The acetate derived from NAA in oligodendroctes has been shown to participate in the biosynthesis of fatty acids which are then incorporated into myelin lipids. Myelin is the insulator covering of the axons of the neurons, which allows the efficient transduction of nerve impulse along the neuron. Hypothesis 2:

NAA as a precursor for the synthesis of N-acetylaspartylglutamate (NAAG).35-37

NAAG has been found to be synthesized from NAA and L-glutamate by an ATP dependent NAAG synthetase. NAAG is a peptide neurotransmitter that activates the type 3 metabotropic glutamate receptor (mGluR3) on the presynaptic membranes and thereby inhibits the release of other signaling molecules (ex: L-glutamate and

GABA).38 Hypothesis 3: NAA as molecular pump.39 NAA is an osmotically active metabolite, which contributes to 7% of the total neuronal osmolarity. NAA helps to remove the excess metabolic water from neurons, hence regulating the neuronal osmotic balance. In addition to these hypotheses, NAA has been identified as an acetate donor in posttranslational protein acetylation reactions40 and as a regulator of neuronal signal transmission.41

5

1.3 Aspartoacylase Deficiency

Aspartoacylase (ASPA, EC 3.5.1.15) is one of three enzymes that catalyze the deacetylation of N-acetylated amino acids.42, 43 ASPA also known as aminoacylase II, a brain-specific enzyme encoded by ASPA (acy2) gene, which is predominantly expressed in the oligodendrocytes.44 ASPA appears to be the only enzyme that efficiently catabolizes N-acetylaspartate in the brain with very high specificity.45 ASPA is structurally a dimeric enzyme that belongs to the carboxypeptidase family.46, 47

Over 70 different mutations have been identified to date in the ASPA gene, including various point mutations, deletions and insertions (http://www.hgmd.org/).

The majority of these are missense mutations resulting in non-conservative amino acid substitutions, and these mutations are found throughout the protein structure

(Figure 1-2). These mutated aspartoacylases are either not expressed or are expressed but are less stable than the native enzyme, with little or no catalytic activity observed.48, 49 Therefore, the biophysical properties of sixteen selected missense mutations associated with Canavan disease have been studied to determine why these mutated enzymes cause CD and how the mutations are related to disease severity. The answers to these questions are discussed in Chapter 2 of this dissertation. Often the catalytic deficiency of an enzyme resulted from a mutation located remote from the active site, with the loss of activity a consequence of the compromised stability of the structure. Therefore, how a distant mutation can affect the structural integrity of

ASPA leading to its catalytic deficiency is the focus of Chapter 3.

6

Figure 1-2: Cartoon representation of the aspartoacylase crystal structure (PDB ID: 2O4H). ASPA is a dimeric enzyme with extensive subunit contacts (two subunits are colored differently). The active site residues are represented in green in both subunits and the residues at known missense clinical mutations are colored in red in one subunit.

The catalytic deficiency of aspartoacylase resulting from genetic mutations is the underlying cause leading to the hallmark symptoms of CD. Among the hypotheses that have been proposed to explain the brain pathology associated with Canavan disease, two major ideas have the most experimental support. (1) Metabolic hypothesis:34, 50, 51 The lack of NAA derived acetate in oligodendrocytes results in insufficient myelin lipid biosynthesis which leads to the characteristic dysmyelination of neuronal axons. (2) Osmotic-hydrostatic hypothesis:52 The accumulation of osmotically active NAA and its associated water in CNS results in increased extracellular hydrostatic pressure and the spongy form vacuolization. However, these hypotheses do not stand alone to account for the disease pathology, and there is a continuing debate about the significance of each of these mechanisms in the

7 progression of CD. Therefore, our approach assumes that the inhibition of NAA biosynthesis may allow us to identify the pathophysiology of NAA in CD, and perhaps help to ameliorate the possible deleterious effect of elevated NAA in the brain. Since little is known about the structural and mechanistic details of the NAA biosynthetic enzyme aspartate N-acetyltransferase, Chapter 5 of this dissertation describes our initial attempts toward the characterization of recombinant

AspNAT, with the ultimate goal of identifying selective inhibitors against this enzyme.

Despite many years of research, CD still remains a fatal and incurable disease.

Several therapeutic options have been explored, with gene therapy as the most promising treatment to date. Gene replacement using recombinant adeno-associated viral vectors carrying the ASPA gene (rAAVs-ASPA) was lead to the reduction of brain NAA levels and significant improvements of the clinical status in Canavan mouse models and in several diseased children.53, 54 Since the lack of brain acetate is thought to be a contributing factor to CD pathology, the use of acetate supplementation in CD has been investigated by administration of the acetate precursor glyceryltriacetate (GTA). Acetate therapy has ameliorated the clinical symptoms in a rat model of CD. 55 However, this therapy has not had any significant effect on human patients to date.56 Lithium as a therapy is widely used for the treatment of mood disorders.57 The use of lithium on Canavan patients has led to an observed reduction of brain NAA levels.58 In Chapter 4 of this dissertation we explore the possibility of a nonconventional therapeutic approach using small molecules as potential stabilizers or activators of ASPA to recover the stability and activity of defective aspartoacylase as a potential therapy for Canavan disease.

8

Chapter 2

Biophysical Characterization of Aspartoacylase Clinical Mutations

DNA contains the information about the synthesized in the cells that are essential for life. Anomalies in the DNA can result in a variety of inheritable genetic defects that could cause the loss of a vital biological function of the encoded protein, leading to potentially lethal outcomes. An amino acid substitution resulting from a may affect the structure and function, or may interfere with its proper cellular trafficking. Defects that occur at a catalytic site in an enzyme can be directly ascribed to the replacement of a critical functional group with a less suitable functionality, causing a loss of the biological activity. However, in many instances these mutations are found at sites that are remote from the active site of the proteins, yet can still dramatically reduce biological function. This loss of function is likely caused by the disruption of important structural elements, resulting in destabilization, misfolding, and shortening the lifespan of the protein.

Canavan disease (CD) is a fatal, autosomal-recessive neurodegenerative disorder that is caused by a catalytic deficiency in the enzyme aspartoacylase (aminoacylase

II, ASPA).39 This enzyme catalyzes the hydrolysis of N-acetylaspartate (NAA) in oligodendrocytes in the brain to produce acetate and L-aspartate, and appears to be 9 the only enzyme in the central nervous system that can effectively metabolize NAA.

The ASPA (acy2) gene that encodes for the human ASPA has been mapped to the human 17p13-ter region, and is composed of six exons which are highly conserved during evolution.59

More than 70 different mutations, including missense mutations, insertions, deletions, and nonsense mutations, have been reported in the ASPA gene in the

Human Gene Mutation Database (www.hgmd.org) (Figure 2-1). Although ethnic distributions of these known defects are widespread, CD is especially prevalent among the people of Eastern European Ashkenazi Jewish origin.11 Two key mutations, E285A (854 AC) and Y231X (693 CA) account for nearly 97% of the disease in this population.11 Approximately 1 in 40 Ashkenazi Jews carry these mutated alleles.12 The ASPA mutations among non-Jewish patients are more diverse, although the A305E (914 CA) mutation accounts for nearly half of mutant alleles found in non-Jewish patients.60 Most of the other mutations are confined to small clusters of patients across different ethnic groups, including cases among European,61, 62 Arab,63 and Asian64-66 patients. Even though several point mutations have been found in the active site of aspartoacylase, most of the clinical mutations are distributed throughout the protein structure, with many found at sites that are quite remote from the active site of this enzyme (Figure 2-1). Therefore, we hypothesize that the aspartoacylase functional deficiency seen in Canavan patients is primarily due to structural perturbations caused by these mutations, leading to improper protein folding and shortened enzyme lifetimes.

10

10 20 30 40 50 60 MTSCHIAEEH IQKVAIFGGT HGNELTGVFL VKHWLENGAE IQRTGLEVKP FITNPRAVKK T G T R P G R T 70 80 90 100 110 120 CTRYIDCDLN RIFDLENLGK KMSEDLPYEV RRAQEINHLF GPKDSEDSYD IIFDLHNTTS A S H E G Y Y 130 140 150 160 170 180 NMGCTLILED SRNNFLIQMF HYIKTSLAPL PCYVYLIEHP SLKYATTRSI AKYPVGIEVG I E T R I H F Y C W 190 200 210 220 230 240 PQPQGVLRAD ILDQMRKMIK HALDFIHHFN EGKEFPPCAI EVYKIIEKVD YPRDENGEIA L H F R E T C T 250 260 270 280 290 300 AIIHPNLQDQ DWKPLHPGDP MFLTLDGKTI PLGGDCTVYP VFVNEAAYYE KKEAFAKTTK R V P R L A TC S L S 310 LTLNAKSIRC CLH E

Figure 2-1: Mapping of the missense CD mutations to the primary structure of the native aspartoacylase enzyme. The enzyme active site residues are highlighted in yellow, and the missense mutants reported in human gene mutation database (public version) as of April 2014 are shown in red below the native amino acid sequence.

11

In addition to the diversity in the position of these point mutations, there are reports of variability in the rate of disease progression, in the severity of the symptoms and in the effect on patient lifetimes.6, 67 Numerous patients have been identified with delayed onset, milder symptoms or slower progressing forms of CD.68-

71 However, even for the mildest disease forms these CD patients show severe retardation, and never develop the ability to walk or talk. Given the distribution of enzyme mutations and the variability in CD phenotypes, it is essential to investigate the biochemical properties of these different pathogenic mutants of aspartoacylase to determine a possible correlation with the disease phenotype of CD patients.

12

2.1 Selection of Aspartoacylase Mutants

Mutants of aspartoacylase were selected to represent a wide range of clinical mutations found in CD patients that were reported to show different disease progressions and severities. In this regard, sixteen missense mutations were selected for this study (Table 2.1). These pathogenic mutations, except for ASPA R71H and

Y288C, are not directly involved in substrate binding, zinc coordination, or . The biochemical properties of the most common missense mutations

(ASPA E285A and A305E) were also selected for examination. The majority of these clinical mutations have been reported as the homozygous trait. Compound heterozygous mutants lead to variable disease symptoms dependent upon the identity of the paired mutation.

Table 2.1: Sixteen aspartoacylase clinical mutants selected in the present study.

Disease phenotype Mutations Mild G274R61, 67, P181T72, Y231C73, P257R, I143T64, K213E69, R71H71, Y288C68, 70 Severe I143F74, C152R63, C152W62 Variable D249V62, 75, A305E62, 67, 69, F295S61, 67, E285A11, 62, 76, N121I74

2.2 Cloning and Site Directed Mutagenesis

Human aspartoacylase wild type gene (ASPA) with an additional coding sequence for the C-terminal hexahistidine tag (His6) had been cloned into the pPICZ

A plasmid vector using EcoRI and XhoI restriction sites.46 Each of the aspartoacylase gene mutations was created using a site-directed mutagenesis kit (Stratagene) according to the manufacturer’s instructions to obtain the mutant enzymes. Briefly, an

ASPA-pPICZ A construct was amplified by polymerase chain reaction (PCR) using 13 the DNA primers (Integrated DNA Technologies) that contains the desired nucleotide variations (Appendix A1), and then the PCR reaction was treated with DpnI restriction enzyme (New England Bio Labs) to remove the parent template DNA.

Amplified mutated ASPA-pPICZ A is 4.3 kbp (Figure 2-2). The mutant construct was transformed into an XL1 blue E. coli competent cell line using zeocin as the selective antibiotic. The plasmids were isolated from XL1 cell line using a plasmid miniprep kit (Qiagen) and each of the mutant ASPA genes was sequenced to verify the presence of expected clinical mutation (Eurofins Genomics). The purified plasmids were linearized by SacI digestion (New England Bio Labs) to insert the gene into Pichia pastoris genomic DNA by homologous recombination using the easy select Pichia transformation kit (Invitrogen).

K213E F295S 0 ng 25 ng 50 ng 0 ng 25 ng 50 ng Template MW DNA

5 kb 4 kb 3 kb

Figure 2-2: Evaluation of PCR amplified mutant ASPA-pPICZ A after DpnI digestion on 1% agarose gel. The DNA bands near the 4 kbp marker indicate the successful amplification of ASPA-pPICZ A construct (Size ~ 4.3 kbp).

14

2.3 Expression and Purification

The human aspartoacylase mutant enzyme forms were expressed in P. pastoris

KM71H cell line under control of a methanol-induced (AOX 1) promoter. Cells were grown on YPDS (yeast extract, peptone, dextrose, and sorbitol) plates for 3 to 4 days at 30 oC under Zeocin selection, followed by growing of single colonies in YNB

o (yeast nitrogen base) – minimal glycerol (1% glycerol) media at 30 C until OD600 reaches to 12 – 15. The cells were then transferred to a YNB – minimal methanol (1% methanol) media to induce the expression of ASPA gene for 48 hrs at 30 oC. The cells were harvested by centrifugation and stored at –80 oC until further use.

Approximately 15 g of cell paste was resuspended in buffer A (20 mM potassium phosphate; pH 7.4, 5% glycerol, 20 mM imidazole, 5 mM β-mercaptoethanol, and 0.5

M NaCl), and was subjected to mechanical lysis using a bead beater in the presence of a protease inhibitor cocktail (P8340 from Sigma-Aldrich). Each mutant enzyme was purified by immobilized metal affinity chromatography (IMAC) using a linear imidazole gradient (20 to 400 mM) on a Ni-Nitrilotriacetate (Ni-NTA) column

(Qiagen), followed by anion-exchange chromatography using a linear Cl- gradient

(0 to 200 mM) from Source 15Q column (GE healthcare) to attain pure enzymes.

15

20120312 Supun ASPA R71H Source 15Q(1331600061)001:10_UV1_280nm 20120312 Supun ASPA R71H Source 15Q(1331600061)001:10_UV3_260nm 20120312 Supun ASPA R71H Source 15Q(1331600061)001:10_Conc 20120312 Supun ASPA R71H Source 15Q(1331600061)001:10_Fractions mAU A280 MW Fractions A260 50.0 16 17 18 19 20 21 22 23 24 25 26 28 [Cl-] ASPA

40.0

30.0 50 kDa 37 kDa 20.0 25 kDa ASPA

10.0

0.0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 100 150 200 250 300 ml

Figure 2-3: Aspartoacylase purification profile. The second peak corresponds to the protein of interest in the chromatogram from Source 15Q column. The purity of aspartoacylase was evaluated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) after the second purification step. The fractions that had aspartoacylase activity are circled in yellow on the gel image.

During the second (anion exchange) purification, two types of proteins with the similar molecular weight (~ 37 KDa) were eluted at different chloride ions concentrations (Figure 2-3). Only the protein that eluted at the higher Cl- ion concentration (2nd peak in Figure 2-3) represented the protein of interest and showed aspartoacylase activity. The inactive protein that eluted early from Source 15Q column was initially speculated to be a misfolded fraction of aspartoacylase.

However, peptide mass fingerprinting studies of tryptic digested proteins identified the inactive protein as the P. pastoris mitochondrial alcohol dehydrogenase isozyme

III (Appendix A2). This finding was further confirmed by the detection of alcohol dehydrogenase activity using ethanol as a substrate. In hindsight this is not surprising since the expression of aspartoacylase was controlled by an alcohol oxidase promoter

16 and was induced by the addition of methanol. The active enzyme was concentrated and dialyzed in a buffer containing 50 mM Hepes, pH 7.5, 100 mM NaCl, and 1 mM dithiothreitol (DTT). The yields of purified mutant enzymes ranged from 0.2 to 1.0 mg with > 95% purity (as assessed by Coomassie stained SDS-PAGE) per liter of growth medium.

2.4 Enzymatic Activities of Aspartoacylase Mutants

The catalytic activity of the purified enzyme forms was measured by using a previously developed continuous enzyme assay in which the L- produced from the aspartoacylase reaction is deaminated by the coupling enzyme L-aspartase to form fumarate (Figure 2-4).45 Assays were performed in a 96 well plate on a

Molecular Devices SpectraMax 190 spectrophotometer plate reader by monitoring the increasing concentration of fumarate at 240 nm (ε = 2.53 mM-1cm-1) in an assay buffer composed of 50 mM Hepes, pH 7.5, 1 mM Mg2+, 1 mM DTT, 1 mM N- acetylaspartate (NAA) and excess levels of the coupling enzyme. The reaction was initiated by the addition of aspartoacylase at a final concentration of 0.02 mg/ml.

Mutants with very low activity were tested at higher levels of the enzyme. The active site mutant enzymes (R71H and Y288C) and the E285A mutant enzyme each showed elevated KM values of 0.57 ± 0.03 mM, 0.86 ± 0.07 mM, and 0.93 ± 0.06 mM, respectively, for the substrate NAA. All of the other mutants had KM values comparable to that of the native enzyme (0.12 mM). Therefore, the mutants with elevated KM were tested at higher concentrations of substrate to ensure that the measured rates correspond to the maximum velocity. L-Aspartase, the coupling enzyme used in the assay, was purified using a previously published protocol.77

17

O NH O O NH O Aspartoacylase 3 Aspartase O O O O O O O NH O O O 4 N-acetylaspartate L-aspartate Fumarate O

Figure 2-4: Schematic representation of L-aspartase coupled aspartoacylase activity assay. (Adapted from Moore, 200345)

Previous studies on the aspartoacylase enzyme activity in cell cultures had reported little or no catalytic activity for these pathogenic mutants17, 48, even for the milder, more slowly progressing cases of CD68, 69, 71. In contrast, each of the purified enzyme forms demonstrates measureable activity levels in our more selective and more sensitive enzyme assay. The most active of the studied mutated enzymes have specific activities that range from 20% to 35% that of the native aspartoacylase activity, while the least active enzyme forms have less than 5% of the native enzyme activity (Table 2.2). E285A, the point mutation that is most commonly found in

Canavan patients leads to a mutated aspartoacylase with the lowest catalytic activity

(0.02 U/mg) found among the enzymes that were examined. Because the loss of aspartoacylase activity has been shown to be the cause of Canavan disease it is reasonable to expect that patients with the most severe and rapidly progressing forms of this disease should have mutant enzyme forms with very low catalytic activity. In fact, a general trend has been observed in which patients with the most severe and rapidly progressing forms of CD tend to have mutations in the ASPA gene that lead to enzyme forms with particularly low catalytic activities (Table 2.2).

However, there are some exceptions to this general trend, with two examples where a mutation associated with the mild form of Canavan disease produced an 18 enzyme form with very low catalytic activity. The mutations R71H and Y288C are found in patients with a slower onset form of CD69, 71, yet the specific activity of these mutant enzymes is fairly low, only 11% and 4% that of the native enzyme, respectively. These mutations are located within the active site of aspartoacylase that has been shown to function in substrate binding.47 Any non-conservative mutations at the active site would explain the dramatic loss of catalytic activity due to the loss of a functional residue, but this is not consistent with the mild disease phenotype in these patients.

19

Table 2.2: Biochemical properties of aspartoacylase clinical mutants78

a Thermal Conformational Aspartoacylase Specific activity Disease phenotype stability stability mutant (U/mg) % of native (oC) (mM)

Native* ---- 7.5 100% 59.9 1150

Mild disease phenotype G274R mild with K213E b 2.6 35% 58.3 650 P181T mild with E285A b 2.4 32% 58.3 600 Y231C* mild 1.8 24% 49.6 1050 P257R* mild with A305E b 1.6 21% 50.6 1000 I143T* mild 1.2 16% 51.9 1000 K213E mild with G274R b 1.16 15% 58.0 750 R71H mild 0.83 11% 57.2 650 mild Y288C mild with F295S & A305E b 0.29 4% 56.7 250

Variable/undetermined disease phenotype D249V variable c 2.2 29% 58.6 720 N121I undetermined d 1.2 16% 47.1 500 variable c A305E mild with P257R & Y288C b 0.78 10% 58.9 900 variable c F295S mild with Y288C b 0.75 10% 56.6 1050 E285A* variable c 0.02 0.3% 48.5 125

Severe disease phenotype I143F* severe 0.08 1% 51.2 600 C152W* severe 0.07 1% 58.5 1100 C152R* severe 0.04 0.5% 51.2 300 a standard errors on the specific activity measurements are ±5% b compound heterozygote patients c variable severity and disease progression reported in different patients d reported from a single patient of undetermined phenotype  * mutants characterized by Dr. Stephen Zano

20

2.5 Protein Stability Studies

The vast majority of the clinical mutations in aspartoacylase are located in regions that are remote from the active site of the enzyme, and the loss of catalytic function in Canavan patients is most likely a consequence of the destabilization of the aspartoacylase structure rather than a direct effect on enzyme activity. Therefore, two different approaches were employed to determine whether the stability of these aspartoacylase mutants is compromised relative to that of the native enzyme.

2.5.1 Thermodynamic Stability

The thermodynamic stability of different aspartoacylase enzyme forms were measured by differential scanning calorimetry (DSC) using a MicroCal VP-DSC instrument. Each of the enzymes was dialyzed overnight into 50 mM Hepes, pH 7.5 and 1 mM Tris(2-carboxyethyl)phosphine hydrochloride (TCEP-HCl). The buffer and enzyme solutions were centrifuged to remove dust particles and large aggregates, and then were degassed at 15 ºC for 20 – 30 minutes. Buffer was initially placed in both the sample and the reference cell of the DSC to achieve a stable baseline. Half a milliliter of 5 to 10 μM aspartoacylase solution was then introduced in the sample cell and the sample temperature was scanned from 15 ºC to 100 ºC at a rate of 1 ºC per minute (Appendix A3). The curve obtained from the sample run was subtracted from the baseline curve obtain in the buffer vs. buffer scan, followed by normalization of the resultant curve for the enzyme concentration. The calorimetric data were fitted to a two state model using the Origin (version 5.0) software package to obtain the melting temperature (Tm) for all forms of aspartoacylase (Figure 2-5).

21

o ASPA R71H; Tm 57.2 C

Figure 2-5: A typical thermal denaturation curve obtained in the DSC based thermal stability study.

60

58 Y288C R71H

56 C) º 54

52

Stability ( 50 E285A 48

46 0 0.5 1 1.5 2 2.5 Specific Activity (U/mg)

Figure 2-6: Thermal stability of the clinical mutants of aspartoacylase and correlation with the loss of catalytic activity and disease phenotype.78 Data are shown for mutant enzymes associated with severe (red), mild (green), and variable or undetermined (blue) disease phenotypes. The mutant enzymes are divided (dashed line at 56 oC) into high and low stability groups.

22

Differential scanning calorimetry measures the heat that is gained or lost upon significant protein conformational changes. Native aspartoacylase was found to undergo a thermal denaturation transition at about 60 ºC. Although the specific activities of the clinical mutants vary across two orders of magnitude, from 2.2 - 2.6

U/mg for the most active mutants to 0.02 U/mg for the least active E285A mutant, the thermal stability of these different enzyme forms are clearly organized into two distinct groups (Figure 2-6). Nine of the mutant enzyme forms have Tm values that are within three degrees of the native enzyme value. The remaining seven enzymes have

Tm values (Table 2.2, bold entries) ranging from 8 ºC to nearly 13 ºC lower than that of the native enzyme. This lower stability group includes three higher activity mutant enzymes that are associated with milder forms of CD (Figure 2-6, green diamonds).

23

2.5.2 Conformational stability

As an alternative measure of protein stability, the conformational stability of aspartoacylase mutants was measured by urea denaturation studies performed on a

Molecular Devices SpectraMax 190 spectrophotometer plate reader using the enzyme activity assay described in section 2.4. The catalytic activity of each purified enzyme form was measured at increasing concentrations of freshly prepared urea up to 3.0 M.

The conformational stability of the enzyme was then determined by plotting enzyme specific activity vs. urea concentration to determine the urea concentration where the enzyme activity had decreased to half of its original value in the absence of urea

(Figure 2-7).

2.5 ASPA P181T 2

1.5

1

Avg SA (U/mg) 0.5

0

0 500 1000 1500 2000 2500 3000 Urea (mM)

Figure 2-7: A typical conformational stability curve obtained plotting average specific activity of the enzyme vs. concentration of the denaturant.

24

1200

900

R71H 600

[urea]mM 300 Y288C E285A 0 0 0.5 1 1.5 2 2.5 Specific Activity (U/mg)

Figure 2-8: Conformational stability of the clinical mutants of aspartoacylase and correlation with the loss of catalytic activity and disease phenotype.78 Data are shown for mutant enzymes found in patients with severe (red), mild (green), and variable or undetermined (blue) disease phenotypes. The mutant enzymes divided (dashed lines) into high (>800 mM urea), moderate (400 – 800 mM urea) and low stability (<400 mM urea) groups.

Low molecular weight osmolytes such as urea and guanidine can be used to destabilize and denature proteins in vitro.79, 80 The various aspartoacylase enzyme forms were subjected to increasing concentration of urea to measure their conformational stability. The native enzyme requires exposure of greater than 1 M urea before losing 50% of its catalytic activity and six of the clinical mutants were equally resistant to urea denaturation (Figure 2-8). A second group of seven mutants

(Table 2.2, bold entries) denature at about half the urea concentration of the native enzyme, and three of the mutants (Table 2.2, bold italic entries) denatured at urea concentrations of 300 mM or below (Figure 2-8).

25

2.6 Summary

Sixteen aspartoacylase missense mutations found in Canavan patients were cloned, expressed and purified to examine their biophysical characteristics and the relationship between enzyme properties and disease phenotype. All of the mutant enzymes studied were found to have measureable catalytic activity. However, the activities of these mutants were diminished, by as little as 3-fold to greater than 100- fold when compared to the native enzyme. Many of the mutated aspartoacylases also show compromised enzyme stability as measured by thermal and conformational stability studies. Four mutants, including the most common E285A mutation, show both diminished thermal and diminished conformational stability. Significantly, each of these four low stability mutants was responsible for either a severe or, in one case, a variable form of Canavan disease. In contrast, seven enzyme forms exclusively associated with the milder forms of CD show either comparable thermal or conformational stability to that of the native enzyme.

Given the differences in conditions between these studies of the purified enzymes and the conditions found in their cellular environments, it is not clear how these in vitro stability measurements will reflect their in vivo stabilities. Hershfield et al.48 have shown that many of the aspartoacylase mutants are produced in very low levels in vivo in COS cells regardless of the comparable levels of the expressed ASPA gene when compared to the native enzyme. Therefore, the extent of protein destabilization found in our study would certainly be expected to dramatically shorten the cellular lifetime of these mutant enzymes and contribute to the overall loss of biological function.

26

Chapter 3

Structural Characterization of Aspartoacylase Mutants

Canavan disease is an inherited, progressive neurodegenerative disorder that occurs as a consequence of a defect in a single metabolic enzyme in the brain.2 The clinical manifestations of CD typically becomes apparent about 3 months after birth, including macrocephally, hypotonia, poor head control, developmental delay, motor, visual, and verbal retardation. The devastating impact on the developing brain usually leads to a premature death in childhood.6 CD has been reported in many populations but is most prevalent in the Ashkenazi Jewish community.11 Mutations in the ASPA gene that codes for aspartoacylase (ASPA, EC 3.5.1.15) lead to insufficient metabolism of N-acetylaspartate in the brain.1, 3 NAA acts as the acetyl group carrier, which is synthesized in the neuronal mitochondria and then transported to the oligodendrocytes where myelin formation takes place. In the oligodendrocytes, NAA is hydrolyzed by ASPA to generate L-aspartate and acetate, the precursor for myelin lipid biosynthesis.32 Deficiency in ASPA activity results in an inadequacy of acetate and elevation of NAA in the central nervous system leading to poor myelination of neurons and the spongy degeneration of the white matter in the developing brain of

Canavan patients.51, 81

27

The human ASPA gene consists of 6 exons and codes for a 313 amino acid protein with a two domain architecture. The first 212 residues of the enzyme form the amino-terminal domain, which is made up of a central 6-stranded β-bundle surrounded by eight α- helices. The carboxyl-terminal domain is composed mainly of

β-sheet and coil structures that wrap around the amino-terminal domain.82 The enzyme ASPA exists as a homodimer with the active site in each subunit formed primarily by residues in the N-terminal domain.82 More than 70 different mutations in the ASPA gene have been reported in the human gene mutation database

(www.hgmd.org), and the majority of these are missense mutations located remotely from the catalytic site of the enzyme. E285A is the most common among Ashkenazi Jews,11 comprising nearly 85% of the CD patients in this population. This clinical mutant has less than 1% residual activity when compared to the native enzyme.45, 78 The ASPA K213E, Y231C, and F295S mutants are found in mild or variable disease phenotypes 67, 69, 73 with relative residual activities of 15%,

24% and 10%, respectively.78 These alterations each occur in highly conserved regions of the aspartoacylase primary structure, but none of them result in the replacement of amino acids that are directly involved in the catalytic activity of

ASPA.47, 76 However, both in vivo and in vitro studies have found that the ASPA pathological missense mutations have lower protein stability and decreased cellular availability as compared to the native enzyme.48, 78

Because the loss of enzyme activity and stability appears to be dictated primarily by changes in protein structure, the three dimensional structure of aspartoacylase mutants can provide molecular insights into Canavan disease, and could guide the development of therapeutic approaches for the treatment of this devastating disease.

No structural information has been available for any of the ASPA clinical mutations. 28

Therefore, the effects of genetic mutations on the aspartoacylase structure are the focus of this study. Here we report and examine the first structures of four different aspartoacylase clinical mutants, including the predominant E285A mutation found in

Jewish Canavan patients.

3.1 Protein Purification and Crystallization

The human ASPA gene had been cloned into the pPICZ A plasmid vector.46 Each of the aspartoacylase clinical mutants was constructed by using the QuikChange II site-directed mutagenesis kit (Stratagene), with the presence of the correct mutations confirmed by DNA sequencing. Mutant enzymes were expressed and purified as described previously with only slight modifications.46 Obtaining a reasonable amount of protein for crystallization was a great challenge, as the clinical mutations yielded very low levels of protein. The ASPA expression in Pichia was performed in the presence of small molecule osmolytes such as glycerol, sorbitol, and trimethylamine

N-oxide (TMAO) to attain a higher amount of protein. Some mutant proteins produced significantly better yield in the presence of 0.4 – 0.5 M TMAO. However, the most severe forms of clinical mutations (ASPA C152R and I143F) did not show any improvement in yields irrespective of the presence of osmolytes. After each step of purification the purity of the enzyme samples was determined by SDS-PAGE using

4–12% Bis-Tris precast gels (Life Technologies) and MES running buffer.

Aspartoacylase activity was measured by our previously reported aspartase-coupled enzyme assay.45 The concentration of the protein was measured at 280 nm by using a

NanoDrop 2000 spectrophotometer (Thermo Fisher).

29

Binary complex crystals of the aspartoacylase clinical mutants with a bound intermediate analog of the ASPA catalyzed reaction were grown by the hanging-drop vapor diffusion method. The intermediate analog, N-phosphonomethyl-L-aspartate was synthesized by Dr. Amarjit Luniwal in the Viola research laboratory, University of Toledo as reported previously.47 Each of the mutant proteins (4 – 5 mg/ml) was incubated with 2 to 5 mM intermediate analog on ice for 1 hr. The protein – intermediate analog mixtures were centrifuged to remove dust particles and large protein aggregates before proceeding with the crystallization experiments. The conditions where aspartoacylase native enzyme crystallized were used as the starting point for crystallization of the pathological mutants. The crystallization drop was formed with two parts of protein-intermediate analog solution, and one part of well solution containing 50 mM sodium citrate (pH 6.0), 300 mM K2HPO4, 3% ethylene glycol, 10 mM dithiothreitol, 13 to 15% polyethylene glycol 3350. Showers of needle-shaped single crystals or crystal clusters appeared at 277 K within 24 h, but were not of sufficient quality for X-ray diffraction studies.

To obtain diffraction quality crystals of aspartoacylase clinical mutations, an additive screen (Hampton Research) was carried out. Somewhat improved crystals were obtained in the presence of the additives , 2-methyl-2,4-pentanediol, 1,2- butanediol, and EDTA. The final more promising rod-shaped robust crystals grew in

10 to 20 mM Na2EDTA at 277 K in 3 to 4 days (Figure 3-1). These crystals were transferred into a cryoprotecting solution containing 25% ethylene glycol in well solution, and were then flash frozen in liquid nitrogen for data collection.

30

Figure 3-1: Aspartoacylase – Intermediate analog binary complex crystals. Rod- shaped ASPA clinical mutant crystals were grown in 50 mM sodium

citrate (pH 6.0), 300 mM K2HPO4, 3% ethylene glycol, 10 mM dithiothreitol, 13 to 15% polyethylene glycol 3350, and 10 to 20 mM

Na2EDTA.

31

3.2 Data Collection and Structure Determination

X-ray diffraction data was collected at the Advanced Photon Source beam line

23-ID-D at Argonne National Laboratory, using an image plate MARmosaic detector.

Aspartoacylase crystals are prone to radiation damage, and diffraction signals became weakened before one third of the required data was collected. Diffraction data sets with reasonable completion were collected along the vector after identifying the highest diffraction region on the crystals by rastering (Figure 3-2). Diffraction images were indexed and integrated by HKL2000 program,83 and scaled by

Scalepack. The crystal structures were refined in REFMAC,84 initially with rigid- body refinement followed by restrained refinement using the atomic coordinates of human ASPA intermediate analog binary complex in the Protein Databank (PDB entry 2O4H) as the initial model. Subsequent refinement was performed by iterative cycles of manual model building with COOT.85 All figures were prepared by the

PyMOL molecular graphics system, version 1.3.86

2.8 Å

Figure 3-2: Diffraction pattern of aspartoacylase clinical mutants.

32

All four mutant enzyme forms crystallized as a binary complex in the same tetragonal space group (P42212) as the previously reported binary complex of ASPA native enzyme,47 and each of these crystal forms was found to diffract to medium resolution (Figure 3-2 and Table 3-1). The structures of these mutants were refined for residues 10 to 310 in each subunit of the enzyme by using the intermediate analog

– native ASPA binary complex structure (PDB ID: 2O4H) as the starting model.

There was no interpretable electron density for the first 9 residues at the amino- terminus, as well as for the last 3 residues at the carboxyl-terminus of the enzyme. In the case of the Y231C and E285A ASPA, the final structures were obtained by applying local non-crystallographic symmetry restrained refinement. The structures of all four mutants have crystallographic R factors less than 20% and the average differences between Rwork and Rfree is less than 5% (Table 3.1). The presence of the correct mutation in each structure was confirmed in the 2Fo-Fc difference maps

(Figure 3-3).

33

Table 3.1: X-ray data collection and refinement statistics. K213E Y231C E285A F295S ASPA ASPA ASPA ASPA Data collection statistics Wavelength (Å) 0.979 1.283 0.979 0.979

Space group P42212 P42212 P42212 P42212 Unit cell dimensions a = b, c (Å) 146.07, 147.96, 147.74, 146.17, 103.42 103.46 102.74 103.13 α = β = ɣ (o) 90 90 90 90 Resolution (Å) 46.25 – 2.60 48.88 – 2.90 46.76 – 3.00 46.27 – 2.80 (2.67 – (3.00 – (3.11 – (2.87 – a a 2.60) 2.90) 3.00) a 2.80) a No. of reflections 102690 123336 44136 84572 No. of unique reflections 28127 25970 17284 23349 Completeness (%) 80.1 (30.8) a 100.0 (99.9) 73.8 (16.3) a 83.3 (40.2) a a Redundancy 3.7 (1.5) a 4.7 (4.6) a 2.6 (1.3) a 3.6 (1.6) a a a a a Rmerge (%) 10.6 (20.6) 14.2 (74.4) 10.1 (18.8) 10.7 (25.1) a a a a I/σI 10.5 (2.1) 10.7 (2.1) 9.4 (3.0) 10.8 (2.4)

Refinement statistics No. of reflections, work/free 26625/1409 24598/1320 16360/886 22135/1191

Rwork/Rfree (%) 19.9/23.4 19.7/23.5 19.3/24.0 18.4/23.7 rmsd bonds (Å) 0.007 0.008 0.007 0.010 rmsd angles (o) 1.14 1.30 1.15 1.39

ESU from Rfree (Å) 0.287 0.309 0.445 0.340 No. of protein molecules/atoms 2/4842 2/4830 2/4834 2/4853 No. of atoms 2/26/67 2/26/24 2/26/33 2/26/72 (metal/ligand/water) Average B factor (Å2) 53.4 70.0 72.2 41.0 Ramachandran plot (%) 93.8/5.2/1.0 93.8/5.4/0.8 93.7/5.4/0.9 94.0/5.3/0.7 (favorable/allowed/generously allowed) PDB ID 4mxu 4tnu 4nfr 4mri a Statistics for the highest resolution shell are shown in parentheses.

34

A B

S295 A285 Y288

C D C231 E213

Figure 3-3: 2Fo-Fc difference electron density map for the mutated residues contoured at 1.5 σ. A. ASPA E285A in magenta. B. F295S in orange. C. Y231C in green. D. K213E in blue. The electron density for the perturbed Tyr288 residue is also shown in the E285A structure.

35

3.3 Comparison of the Oligomeric States

Since native aspartoacylase is a functional dimer, and because some of the clinical mutations are located close to the dimer interface (Figure 3-4), the effect of each mutation on subunit dimerization was evaluated to determine whether the observed loss of catalytic activity was a consequence of changes in the quaternary structure of the enzyme. The sample homogeneity and the oligomeric state of the clinical mutants were determined by light scattering using a DynaPro Titan DLS instrument (Wyatt Technology). Native polyacrylamide gel electrophoresis was performed using the dark blue cathode buffer and 3–12% Bis-Tris precast gels from

Life Technologies. Human ASPA was genetically engineered with a carboxyl- terminal hexa- tag, with a resulting calculated molecular weight of 36.4 kDa.

Dynamic light scattering of the ASPA clinical mutants K213E, Y231C, E285A, and

F295S indicated the presence of a dimer in solution for each enzyme form, with a hydrodynamic radius of approximately 3.8 nm, and a calculated molecular weight ranging from 73 to 78 kDa (Appendix B1). Native gel electrophoresis of ASPA also confirmed the dimeric states of these mutants in solution, and the dimer of ASPA

E285A is stable under conditions containing as high as 2 M urea (Appendix B2).

These mutants were also found to crystallize as homodimers, similarly to what was observed for the native enzyme.

36

K213

E285 IA

Y231

F295

>2.6 Å

>1.3 Å

Figure 3-4: Structural comparison of the aspartoacylase clinical mutant E285A with the native enzyme (PDB ID: 2O4H). Overall structure of E285A mutant (magenta) superimposed with native ASPA (gray), showing the locations of the four mutation sites, and the intermediate analog (IA) bound in the active site. Close up of the region near the active site showing the outward shift in the loop positions (green arrows) in the E285A structure. The zinc ion in the active site is represented by a cyan sphere.

37

3.4 Structural Roles of the Mutation Sites

Missense mutations in the Lys213, Tyr231, Phe295, and Glu285 positions of

ASPA have been reported in Canavan patients,11, 67, 69, 73 but these amino acid residues are not directly involved in catalysis.47 Lys213 is a surface residue found distal to the active site (Figure 3-4) and does not have any identifiable contacts within the protein structure. This positively-charged likely interacts with water molecules and helps to solvate the enzyme in its aqueous environment.

A at the Tyr231 position is one of the founder mutations present in Ashkenazim.11 Tyr231 is located in a loop structure in the carboxyl- terminal domain of aspartoacylase, and is in apparent hydrogen bond contact with the hydroxyl group of Tyr289 and the side chain carbonyl oxygen of Asn284. These residues are located in a hydrophobic region surrounded by Ile239, Ile243, Phe262,

Tyr289 and Ala294, as well as Leu187 from the adjacent subunit (Figure 3-4). The enzyme active site is found adjacent to this area and the substrate binding residue

Tyr288 is in the same helix as Tyr289. Phe295 is found in a hydrophobic core at the carboxyl-terminal domain of ASPA, and van der Waals interactions of Phe295 with the neighboring Val229, Ile243, Leu247, and Pro280 residues, along with the hydrocarbon chain of Lys297 (Figure 3-4) help to stabilize the structure of the enzyme. Replacement of this large hydrophobic residue with a small polar would create a void in the core structure of this domain.

Similar to Tyr231, Glu285 is also located fairly close to the enzyme active site

(Figure 3-4). The γ-carboxyl of Glu285 forms hydrogen bonding interactions with the side chain hydroxyl of Thr118 and the backbone of Thr119 (Figure 3-5), and is located in a loop that connects two β-strands in the central β-bundle of the native enzyme. This γ-carboxyl group also interacts with the backbone amide group of

38

Ala287 in an adjacent loop. In addition, the backbone carbonyl of Glu285 is in hydrogen bonding contact with the backbone amide of Tyr288, an amino acid located within the active site cleft (Figure 3-5).

T118 T119

2.8 2.6 A286 2.8

E285 A287 3.1 4.5 Å

Y288

Figure 3-5: Structural comparison of the aspartoacylase clinical mutant E285A with the native enzyme. Participation of Glu285 carboxylate group in a hydrogen-bonding network in native ASPA enzyme (gray) and the structural changes in ASPA E285A (magenta) resulting from an absence of these Glu285 interactions.

39

3.5 Structural Differences between the Aspartoacylase Mutants

The overall structures of these mutant enzyme forms are similar to the structure of the native enzyme (Figure 3-6). Among the mutants examined in this study, the

K213E mutant has the least structural perturbation, and the E285A mutant showed the greatest variances as compared to the native enzyme structure. The non- of a positively-charged amino acid with a negatively-charged one would be expected to cause significant structural perturbations. However, in the case of the

K213E mutation, substituting this lysine with a glutamate at position 213 does not have a substantial effect on the overall enzyme structure because there are no intra- protein interactions with this side chain -amino group (Figure 3-7). The Lys213 side chain is found to be disordered in both subunits of native ASPA, suggesting a failure to make stabilizing interactions with adjacent functional groups. In the K213E mutant the position of this introduced carboxyl side chain is reasonably well defined, but is still not involved in any significant interactions with other protein atoms.

40

Figure 3-6: The overall structures of aspartoacylase native and mutant enzymes. All the enzymes have similar structural fold and exist as homo dimers. Native (gray), K213E (blue), Y231C (green), F295S (orange), and E285A (Magenta)

41

K213E K213E

Figure 3-7: Comparison of the K213E mutant binary complex structure (blue) with ASPA wild type complex structure (gray). Superimposition of the ASPA K213E structure with wild type enzyme gives an rmsd of 0.26 Å for 4134 atoms. The intermediate state analog bound to the active site of K213E (cyan) and wild type (green) is also shown.

The mutation at Tyr231 causes several changes in the aspartoacylase structure.

The Cα of the mutated Cys231 residue moves by about 0.5 Å compared to its position in the native structure, but this introduced side chain does not participate in any new binding interactions (Figure 3-8). One of the adjacent amino acids, Asn284, appears to play an important structural role, as it is involved in a number of hydrogen bonding interactions with the surrounding residues. Due to the loss of balancing interactions and disruption of this hydrogen bonding network in the Y231C mutant,

Asn284 is shifted slightly towards Tyr289 thereby weakening the hydrogen bond interaction of Asn284 with the backbone carbonyl oxygen of Leu187 from the other subunit. This slight movement of Asn284 causes additional structural perturbations 42 around this residue. For example, the loop bearing the Tyr231 residue is solvent exposed and is mainly stabilized by polar and charged residues. This loop is fairly flexible in both the ASPA native and mutant structures as indicated by the higher average temperature factors for these backbone atoms. However, in the Y231C mutant this loop becomes even more disordered, with temperature factors for the Cα atoms greater than 100 Å2 in some areas of the loop. The position of this loop is stabilized in the native enzyme by the van der Waals interaction between Tyr231 and

Ile239, and also by an electrostatic interaction (2.8 Å) between Asp230 and Arg233.

As a result of the Y231C mutation and the loss of this van der Waals contact the

Ile239 residue shifts by about 1 Å away from its original position, and the salt bridge interaction to Arg233 is subsequently weakened with the interatomic distance increasing by about 0.4 Å.

43

L187 (B) 2.8 Q248 3.2 (2.7) 2.9 (3.4) (2.9) 2.7 I243 (2.4) N284 3.2 3.2

F262 Y289 A294

C231 I239 Y231

Figure 3-8: Comparison of the aspartoacylase Y231C structure with the native ASPA enzyme (PDB ID: 2O4H). The Y231C (green) structure is superimposed with the native enzyme structure (gray) with a rmsd of 0.33 Å for 4235 atoms. The hydrogen bond contact distances are shown for the native enzyme, with the new hydrogen bond lengths in the Y231C structure given in parentheses.

The phenylalanine to serine mutation at position-295 of ASPA causes a localized structural perturbation. Replacing this bulky phenyl group with a small polar residue creates a void in the core of the carboxyl-domain structure which eliminates any possible stabilizing interactions with this new side chain unless these are substantial structural rearrangements in this mutated enzyme as compared to the native enzyme.

Instead, the side chain of Met261 moves into this void space in the F295S mutant to occupy the gap as a consequence of a change in the torsion angle around the - bond, by about 160º in subunit A (Figure 3-9) and about 100º in subunit B.

The side chain of Lys297 also shifts to occupy this free space through a 1 Å 44

movement in subunit A and 0.5 Å in subunit B. However, the introduced Ser295

hydroxyl side chain does not appear to make any new interactions with its

surrounding environment that would help to stabilize this core structure.

L272 L272

M261 M261 F262 F262 L255 L255

I243 I243 V229 V229

L247 L247

S295 S295 K297 K297 F295 F295 P280 P280 V283 V283

Figure 3-9: Structural comparison of the aspartoacylase mutant F295S with the native enzyme (PDB ID: 2O4H). Stereo view of the overlaid ASPA F295S (orange) and native ASPA (gray) with a rmsd of 0.31 Å for 4179 atoms, showing the changes in the hydrophobic core around Phe295, including the rotation of Met261 (arrow).

Replacing the glutamate side chain at position 285 with an causes

substantial local structural perturbations at this site. The original position of γ-

carboxyl group of Glu285 is now occupied by a water molecule in the E285A mutant

which is hydrogen bonded (3.1 Å) to the hydroxyl of Thr118 in both subunits. The

localized nature of the structural differences caused by glutamate to alanine mutation 45 at position 285 is best seen when the E285A mutant structure is overlaid with the native enzyme, with an rmsd of only 0.37 Å for 4004 atoms despite this non- conservative (Figure 3-4). To understand how the enzyme activity of this mutant is significantly diminished, the immediate region around this mutation has been examined. Both the main and side chains of the loop formed by residues 156 to 166 are quite disordered, with relatively high temperature factors in the ASPA E285A binary complex structure as compared to its position in the native enzyme binary complex. This loop was also found to be disordered in the native

ASPA (PDB ID: 2O53), but only in the apoenzyme structure,47, 82 suggesting a role for this structural element in response to substrate binding.

3.6 Active Site Alterations

The active site of ASPA is found in a deep cavity within a channel, with the substrate binding pocket composed of several lysyl and arginyl residues. These positively-charged residues help to direct and orient the negatively-charged NAA substrate into the active site. In the native apo-enzyme the side chains of Arg71 and

Lys291 are not well ordered, leading to an open active site channel. Upon substrate or intermediate analog binding conformational changes close the active site and position the substrate for catalysis. The Nz of the Lys291 side chain now makes a stabilizing interaction with the carboxyl oxygen of Asp68 (Figure 3-11), while the guanidine group of Arg71 interacts with α-carboxyl group of the bound intermediate analog in the active site of the native enzyme (Figure 3-10A; F295S enzyme). These new interactions are needed to orient the substrate in the active site and set up the position of the catalytic functional groups.

46

The structural differences that are observed in the ASPA E285A mutant are most prominent in the active site of the enzyme. The loops surrounding the active site contain two conserved amino acids, Tyr164 and Tyr288, which have moved in the

E285A structure from their original positions of Cα by 1.9 Å in subunit A and by 1.3

Å in subunit B for Tyr164, and by 2.6 Å for Tyr288 (Figure 3-4). These structural shifts make the active site channel much wider and more accessible even in the presence of the bound intermediate analog.

The disruption of a stabilizing hydrogen-bonding network in E285A caused by the absence of the carboxylate side chain of Glu285, leads to the displacement of the helix containing residues 285 to 290 and the loop containing the residues from 290 to

294. The active site residue Tyr288 in this helix is moved about 4.5Å away from the active site in both of the enzyme subunits (Figure 3-5). The phenyl oxygen of Tyr288 interacts with both the amide and the α-carboxyl groups of the substrate in the native

ASPA enzyme (Figure 3-10A; F295S enzyme). As a consequence of this altered geometry Tyr288 is no longer able to form these hydrogen-bonding interactions with the substrate in the E285A mutant (Figure 3-10B). The loop containing Tyr164 has also been displaced in the E285A structure, and Tyr164 has become less well ordered.

Thus, the side chain hydroxyl group of Tyr164 is no longer in position to interact with the β-carboxyl group of the substrate (Figure 3-10B).

47

A R63 E178 3.3

2.8 2.8 3.1 Y288 2.9

2.9 2.7 2.9 2.6 3.0 2.6 3.1

N70 R168

R71 Y164

B R63

E178 3.1 2.8 Y288 3.1 2.6

2.9 3.1

N70

R168 R71

Y164

Figure 3-10: Comparison of the active site geometry of ASPA with the bound intermediate analog, N-phosphonomethyl-L-aspartate (shown in cyan). A. ASPA F295S (orange). B. ASPA E285A (magenta).

48

However, in the ASPA E285A structure, the active site remains open even after the intermediate analog (and presumably the substrate) enters into the active site

(Figure 3-11). This failure to close the active site leaves the Arg71 side chain disordered in this mutant, along with Lys291, which is now situated in a dislocated loop. In addition, a hydrogen bond contact between the side chains of Tyr64 and

Glu290, which is important to maintain active site channel architecture, is no longer present in the E285A mutant (Figure 3-11). In contrast, the active site channel of both the ASPA F295S and K213E mutants appear to function similarly to the native enzyme and the active site is observed to be closed in the presence of the intermediate analog. This substrate-induced conformational change of the protein is critical for the proper function of aspartoacylase, and offers a likely explanation for the dramatic loss of activity in the most prevalent E285A clinical mutation.

49

Y64 D68

R71

3.1 2.5

Y164 2.8

E290

K291

Figure 3-11: Conformational changes of Arg71 and Lys291 that occurs upon binding of the intermediate analog in the active site. Overlay of active site cavity of ASPA, with the native apo-enzyme (PDB ID: 2O53) in green, native binary intermediate analog complex enzyme (PDB ID: 2O4H) in gray and the bound intermediate analog in blue, and the E285A binary complex enzyme in magenta, surrounded by positively-charged Arg71 and Lys291. Lower panels: The active site of the native apo-enzyme is accessible and the presence of phosphate ions can be seen in the active site (left). Active site of binary complex of native enzyme is closed by Arg71 and Lys291 and the intermediate analog is trapped in the active site (middle). However, even after the intermediate analog binds in the active site in E285A, the active site remains open and accessible (right).

50

3.7 Structural Effects on Enzyme Catalysis

Aspartoacylase is a metallo-enzyme that belongs to the carboxypeptidase family, and is active as a homodimer.46, 47 Each subunit of ASPA contains one zinc ion, which acts as the catalytic center of the enzyme. The zinc ion in all four mutants

(K213E, Y231C, F295S, and E285A) remains coordinated to the His21, Glu24, and

His116 ligands, similar to what was observed in the native enzyme (Figure 3-12). The active site Glu178 acts as the general base that activates the nucleophilic water molecule coordinated to zinc ion, and later donates a proton to assist the departure of the L-aspartate from the enzyme.47 This catalytic residue remains intact and properly positioned in each of the mutant structures. Arg63, which stabilizes the tetrahedral intermediate formed during the reaction, is also well placed in each of these mutants to support catalysis (Figure 3-10). The phosphonate oxygen of the intermediate analog that mimics the tetrahedral intermediate is coordinated to the zinc center of the mutants similarly to the native enzyme. Therefore, the catalytic machinery of aspartoacylase remains well conserved and poised to carry out the catalytic reaction in these mutant enzyme forms. Nevertheless, each of these mutants have been found to be associated with the symptoms observed in Canavan disease, with each leading to somewhat different clinical outcomes.

51

E24 E24

H116 H116 E178 H21 E178

H21 2.3 2.1 2.3 2.5 Zn 2.4 2.1 Zn 2.9 1.8 3.2 2.4

E24 E24

H116 H116 E178 E178 H21 H21

1.9 2.4 2.3 2.1 Zn 2.3 2.5 2.3 Zn 2.8 2.8 2.3

Figure 3-12: Comparison of the catalytic center of the aspartoacylase clinical mutants (E285A in magenta, F295S in orange, Y231C in green) and native enzyme (gray). The intermediate analog that is coordinated to the zinc ion is shown in green and cyan. The active site Glu178 acts as the general base during the aspartoacylase reaction.

52

The intermediate analog, N-phosphonomethyl-L-aspartate, that mimics the transition state of aspartoacylase was utilized to study the mechanistic details of these mutant enzyme forms. The intermediate analog in the native enzyme forms multiple electrostatic and hydrogen bonding interactions with multiple active site residues; among them Asn70, Arg71, Tyr164, Arg168, and Tyr288 are each important to position the substrate properly in the active site. Based on the orientation of the intermediate analog in the active site, substrate binding in F295S (Figure 3-10A),

Y231C, and K213E are likely the same as that observed in the native enzyme.

K213E is associated with mild disease phenotype. Any defects in the K213E mutant are not apparent from the structure, and this enzyme form has comparable thermal stability to the native enzyme.78

The Y231C mutation is reported to cause a mild disease phenotype, and this mutant enzyme shows fairly high catalytic activity.73, 78 While this mutation is found close to the active site, substrate binding is not significantly affected. Most importantly, Tyr289 can still make the same hydrogen bond interactions with side chain carbonyl of Asn284 and the backbone carbonyl of Leu187 from the adjacent subunit that serves to orient Tyr288 in the active site. However, structural anomalies that arise from the Y281C mutation results in compromised thermal stability.78 In addition, since Tyr231 is present at the origin of a coil structure in the carboxyl- terminal domain, this appears to make critical interactions with the surrounding residues that may facilitate proper folding of the carboxyl domain around the rest of the protein structure to form a functional enzyme.

The diminished activity of F295S mutant most likely results from the decreased stability of the protein78 as a consequence of the perturbations in the core of the carboxyl-domain structure. Immunoblotting studies of ASPA enzyme expressed in

53

COS-7 cells have shown that the F295S mutant yields very low protein levels despite the similar level of mRNA when compared to the native enzyme.48 In the same study, the K213E and E285A mutants were found to express similar amounts of enzyme when compared to native ASPA. Among these four mutations F295S and Y231C yield somewhat reduced levels of protein in the Pichia pastoris expression system employed in our laboratory (data not shown). Therefore, both Y231C and F295S mutants appear to suffer from protein stability and folding issues that would likely limit their lifetime in the brain of CD patients.

ASPA E285A is one of the lowest activity mutants examined, and this enzyme form also has very low thermal stability when compared to the native enzyme.78 In addition, the substrate binding residues in the active site of this mutant are not correctly positioned to interact with and orient the substrate (Figure 3-10B). The

Tyr288 that forms a hydrogen bond interaction with the α-carboxyl and the amide nitrogen of the intermediate analog in the native enzyme has moved away from the active site. The Arg71 and Tyr164 side chains are each not well ordered in the E285A mutant structure (Appendix B3). Tyr164 does appear to be capable of making a hydrogen bonding contact with α-carboxyl of the substrate in one of the monomers in the structure. Similarly, in one of the monomers Asn70 and Arg168 are positioned to form stabilizing interactions with the β-carboxyl group of the intermediate analog.

Thus, the orientation of the substrate in each monomer of the E285A mutant has been altered. The loss of activity in ASPA E285A is a direct result of the distorted architecture in the active site. These differences in substrate binding to the active site of E285A mutant is also supported by the five-fold higher Ki value observed for the intermediate analog for this mutant (1.48 ± 0.13 µM) compared to that for the native enzyme (0.33 ± 0.05 µM). In addition, N-acetylglutamate (NAG), a one carbon longer

54 homolog of NAA, and a dipeptide neurotransmitter N-acetylaspartylglutamate

(NAAG) were found to selectively interact with the E285A mutant. This mutant enzyme is inhibited by NAG with a Ki of 0.84 ± 0.08 mM, and by NAAG with a Ki of

0.47 ± 0.03 mM, whereas these compounds do not show any measureable inhibition of the native enzyme when examined at concentrations up to 5 mM. The inhibition of

E285A by NAG and NAAG can be overcome by increasing the concentration of

NAA, confirming its competitive inhibition vs. the substrate. This selective binding of these NAA homologs to the active site cavity of ASPA E285A is a direct consequence of the altered active site structure of this mutant. Introduction of a conserved mutation at this position (E285D) has been found to produce an enzyme form with significantly higher activity and stability when compared to E285A mutation.45 We can conclude that the negatively-charged glutamate residue at position-285 is clearly essential for maintaining the aspartoacylase structure and supporting its catalytic function, even though this side chain functional group does not directly participate in substrate binding or in catalysis.

55

3.8 Summary

Canavan disease is caused by the mutations in the ASPA gene that results in aspartoacylase enzyme forms with impaired catalytic efficiency. The first structures of the four ASPA pathological mutants have been determined to study the structural defects resulted from these genetic mutations. The non-conservative replacement of

Lys213 with glutamate has minimal catalytic and structural consequences because of the surface location and lack of intra-protein interactions. Replacement of bulky

Tyr231 and Phe295 in the hydrophobic core of the enzyme results in decreased protein stability. The negatively charged glutamate residue at position-285 is clearly essential for maintaining aspartoacylase structure and supporting its catalytic function even though this side chain functional group does not directly participate in substrate binding or in catalysis. The understanding of structural anomalies provides a basis for the development of treatment therapies for this fatal disease.

56

Chapter 4

Pharmacological Chaperone Approach for the Treatment of Canavan Disease

Proteins are structurally diverse biological macromolecules that participate in many complex cellular processes. Each protein must have a correct native three- dimensional conformation in order to fulfill its biological function. Attaining and maintaining this functional protein fold is somewhat challenging, as proteins are in an extremely crowded environment within the cell and have only marginal stability under physiological conditions.87, 88 However, the biological system is equipped with a protein quality control machinery that maintains the proper homeostasis of proteins

(proteostasis) in the cell. A special class of proteins known as molecular chaperones

(ex: heat shock proteins) assist the folding and refolding of proteins into the correct conformation. Misfolded proteins are cleared from the system by the cellular degradation machinery (Figure 4-1).89, 90

Genetic mutations can disrupt critical structural interactions that are important to maintain the proper functional conformation of the protein. These alterations can perturb the equilibrium between the native and unfolded states of the protein, shifting it towards less stable non-native conformations. Such mutated proteins may be partially folded or misfolded, and may have only residual activity or no activity at all. 57

In some instances, partially folded proteins can also be removed from the cell even though they possess some residual activity, leading to a deficiency in the functional protein levels that can result in disease.91, 92 Misfolded proteins are sometimes diverted into alternative stable structures called amyloids, which are toxic fibrous aggregates. Cellular accumulation of amyloids is associated with a number of diseases, particularly Huntington disease, Alzheimer’s disease, and Parkinson’s disease.93, 94

Transcription

Misfolded PC stabilized protein nonfunctional protein

Translation

Trafficking

Biological Transport degradation destination

Trafficking Unfolded peptide

Functional protein Pharmacological chaperone (PC) Aggregate

Figure 4-1: Schematic representation of the cellular protein homeostasis network. (Adapted and modified from Balch, 200895)

58

Interactions of proteins with low-molecular weight ligands typically result in a more stable protein that has improved resistance to thermal denaturation.96 Applying this phenomenon, a number of small molecules that stabilize defective proteins have been identified in several human protein conformational disorders such as lysosomal storage diseases,97 phenylketonuria,98, 99 and cystic fibrosis100. These small molecule ligands that can selectively stabilize a target protein and mitigate the effects of a less stable protein resulting from a genetic mutation are called pharmacological chaperones (PCs). PCs are low molecular weight compounds that selectively bind to the protein of interest and stabilize the folded state, thereby enabling its proper trafficking to the final destination (Figure 4-1).101 Binding of PCs can also stabilize folding intermediate states that would enhance the protein folding rate and lower the misfolded protein population in the cell. Pharmacological chaperones ideally should bind to an allosteric site that is not associated with a biological function and should enhance both the stability and the activity of the target protein.

Potential pharmacological chaperones are usually identified by high-throughput experimental and virtual screening of compound libraries, or by rational design of active site-directed ligands.102 The PCs that have been identified and used in enzyme enhancement therapy for protein folding diseases so far are competitive inhibitors that reversibly bind to the active site of a target enzyme and enhance the stability, preventing its premature degradation and promoting its proper trafficking.99, 103, 104

This enzyme enhancement treatment for lysosomal storage disease is promising and a pharmacological chaperone identified for Fabry disease is currently in phase 3 clinical trials.105, 106 However, these identified PCs are less than ideal since they compete with the substrate for binding and do not increase the residual activity of the

59 defective enzyme. Rather their effect is to increase the cellular pool of the less active enzyme to alleviate the detrimental outcome of the mutation.

Here we explore the feasibility of pharmacological chaperone therapy as a potential avenue for the treatment of Canavan diseases. CD is a rare neurological disorder, caused by the inherited defects in ASPA gene that encodes the enzyme aspartoacylase.17, 76 These mutations result in loss of structural integrity, catalytic activity, and shorted lifetime.48, 78 As a consequence of this deficiency of catalytic activity of ASPA, the substrate N-acetylaspartate accumulates, and a decrease in the production of acetate leads to demyelination and spongy degeneration of brain. Both experimental and computational screening methods have been used to identify potential lead compounds and potential binding sites for PCs on the aspartoacylase clinical mutants. Success with these approaches would give us a new direction to identify pharmacological chaperones for use in enzyme enhancement therapy for the treatment of Canavan disease.

60

4.1 High-Throughput Thermal Stability Assay

A high-throughput screening approach was used to identify initial lead compounds that improve the stability of some less stable aspartoacylase pathological mutations. Binding of small molecule ligands to ASPA was determined by a fluorescence-based thermal shift assay using a ThermoFluor instrument (Johnson &

Johnson Pharmaceutical) at the Life Sciences Institute in the University of Michigan.

The ThermoFluor instrument employs the principle of differential scanning fluorimetry and is a variation of the real time PCR instrument.107 Differential scanning fluorimetry monitors the thermal unfolding of proteins in the presence of a fluorescent probe.108 The solvent exposed surfaces of soluble proteins are composed predominantly of polar and charged residues, while the inner core is mainly hydrophobic in nature. When a soluble protein undergoes thermal unfolding, more hydrophobic regions get exposed to the solvent which favors the aggregation of these unfolded proteins. The fluorescent probes are small organic compounds that have minimal fluorescence in the aqueous medium and preferentially bind to the hydrophobic patches that become exposed during protein denaturation.109 The obsereved increase in fluorescence that occurs upon binding is monitored as a function of temperature to obtain the thermal denaturation curve (Figure 4-2). The temperature at the midpoint of the unfolding transition is the melting temperature

(Tm) of the protein. ThermoFluor is a sensitive high-throughput assay that uses small amounts of sample and requires very little time compared to differential scanning calorimetry of individual compounds.

61

Unfolded

Aggregate d

Fluorescence

hμ

Native

Tm Temperature (oC)

Figure 4-2: Thermal unfolding transition obtain in differential scanning fluorimetry. Insert: fluorescence emission by florescence probe upon binding to the hydrophobic areas of unfolded protein. (Adapted and modified from Niesen, 2007109 and from thermofluo.org)

4.2 Assay optimization

Preliminary experiments were carried out to identify the optimal assay conditions for the thermal denaturation of aspartoacylase in the ThermoFluor screening platform.

ASPA native and the mutant enzyme forms were expressed in Pichia pastoris cell line and were purified by metal affinity and ion exchange chromatography as described in the section 2.3 in chapter 2. Each of the enzymes was dialyzed into 50 mM Hepes at pH 7.5, containing 1.0 mM TCEP. The thermal stability assay was

62 performed on a 384 well black PCR plate using a napthylamine sulfonic acid dye (1- anilinonapththalene-8-sulfonic acid; 1,8-ANS) (Sigma-Aldrich) in a 10 µl final volume, overlaid with silicon oil (2 µl) to protect the solution from evaporation. All the samples in the plate were scanned from 25 oC to 85 oC in continuous ramp mode at a rate of 1 oC/min. Fluorescence emission of the entire plate was captured at 500 nm using a CCD (charge-coupled device) imaging camera after excitation at 380 nm.

The images were collected at every increment degree of temperature and the fluorescence intensity of each well in the sample plate was plotted with respect to temperature to obtain the thermal denaturation curve and the melting temperature.

Initial validation of the ThermoFluor assay was performed to study the effect of the concentartion of 1,8-ANS dye and ASPA native enzyme on the thermal unfolding transition. Enzyme was tested at three different levels (0.05, 0.1, 0.2 mg/ml) and the fluorescence dye was tested at 50 μM and 100 μM. The optimal thermal denaturation transition was obtained with a better signal to noise ratio at an enzyme and dye combination of 0.1 mg/ml and 100 µM, respectively (Figure 4-3). To check reproducibility and obtain reliable statistics this experiment was carried out in four replicates. The observed melting temperatures of ASPA in the ThermoFluor assay were found to be quite reproducible (56.75 ± 0.05 oC) and were comparable with previous calorimetric data obtained by differential scanning calorimetry (59.9 oC).

63

Figure 4-3: Thermal denaturation curves obtained for native aspartoacylase enzyme in ThermoFluor thermal stability assay at 0.1 mg/ml enzyme and 100 µM 1,8-ANS dye. The melting temperature obtained from each replicate is also shown.

4.3 Compound Library Screening and Hit Validation

To identify lead compounds that can stabilize the mutant forms of ASPA, a focused library of small molecules was screened by ThermoFluor high-throughput thermal shift assay using the optimized experimental protocol. This compound library was available at the Center for Chemical Genomics, University of Michigan and consisted of approximately 1000 compounds belonging to 3 categories: 1. natural products, 2. epigenetics/autophagy/redox library, and 3. protease inhibitors/Wnt pathway modulators/cannabinoid agonists. The primary screening was performed at

64

40 µM compound concentration against aspartoacylase clinical mutant E285A to find the initial hits. The reference sample consisted of the protein only control. Titration with the intermediate analog of the aspartoacylase-catalyzed reaction at 300 µM served as the positive control. The native enzyme served as a benchmark for comparison of the degree of stabilization of the mutant enzymes.

∆Tm

Figure 4-4: The melting temperature shift observed in high-throughput screening for a representative compound identified as a hit.

Compounds that significantly stabilize E285A were identified by comparing the melting temperature (Tm) values obtained in the presence and absence of each compound (Figure 4-4). Preliminary screening identified 32 compounds that increased the Tm of ASPA E285A, indicating the presence of enzyme - ligand 65 interactions that led to stabilization of the mutant enzyme structure. These compounds

o were chosen for further analysis if they exhibited a ∆Tm of +0.5 C or greater.

Because the ligand-induced melting temperature shift is proportional to both binding affinity and the ligand concentration,110 follow up dose response studies of the most successful compounds were carried out against several mutant enzyme forms

(C152R, I143F, E285A, A305E), as well as the native enzyme control, from 0 to 100

µM compound concentrations. This experiment serves to exclude any artifacts from the ANS fluorophore and any false positive hits identified in the primary screening.

Nine compounds (Ellagic acid, Fumonisin B2, 6-Hydroxytropinone, Glutathione,

Tamoxifen citrate, UNC0638, Patulin, Loperamide HCl, U82836EA 2HCl) showed the necessary dose-dependent stabilization for more than one mutant enzyme tested.

Among them glutathione, patulin, and ellagic acid had the highest stabilization potential (Appendix C1). Interestingly, these three compounds stabilize the most prevalent E285A mutant, but had no effect on the stability of the native enzyme

(Figure 4-5).

66

ASPA E285A Native ASPA

Figure 4-5: The concentration dependent stabilization of aspartoacylase native (right) and E285A (left) enzymes by glutathione, patulin, and ellagic acid.

67

Differential scanning calorimetry (DSC) was performed to further verify the effect of these hit compounds on the aspartoacylase clinical mutants. DSC was carried out as described previously (Section 2.5.1) with ASPA native and E285A in the presence of glutathione, pattulin, and elagic acid at 200 μM concentrations.

Ellagic acid had limited solubility in both aqueous and DMSO (dimethyl sulfoxide), and DMSO interfered the DSC signal making it unable to test ellagic acid by DSC.

The melting temperatures of the native enzyme in the presence of glutathione and patulin were tested, and were similar to that of the ASPA native apo-enzyme. These findings are in agreement with the results obtained in ThermoFluor assay. Therefore, we could assume that neither glutathione nor patulin has binding affinity to the native enzyme. In the presence of these hit compounds, the equilibrium of E285A is shifted to a more stable conformation. However, these compounds are not being examined at saturating conditions, which would maximize the shift to the more stable conformation. In order to understand the effect of saturation conditions of the hit compounds, the DSC scans of E285A was performed in the presence of 2 mM glutathione and 2 mM patulin. Surprisingly, the thermal stability of E285A is greatly

o o increased by ∆Tm of +10 C and +13 C, respectively in the presence of the higher levels of glutathione and patulin (data obtained by Dr. Nitesh Poddar, Viola Research

Laboratory). This indicates that E285A has a unique binding pocket available for these compounds which is not available when compared to the native aspartoacylase structure.

68

4.4 Effect of Hit Compounds on the Catalytic Activity

Although the stability of E285A mutant is enhanced by glutathione, patulin, and ellagic acid, it is also important to study the effect of these compounds on the catalytic activity of the enzyme. The enzyme activity was measured in the presence of these lead candidates as described in section 2.4 in chapter 2, by using the coupled enzyme assay in which the aspartic acid produced from the aspartoacylase reaction is converted by L-aspartase to produce fumarate that absorbs light at 240 nm. Among the three compounds, ellagic acid was not tested because it has strong absorbance at the wavelength at which the kinetic assay is operated. The activity of native ASPA is not significantly altered by the presence of either glutathione or patulin. This confirms that these compounds do not interact with the native enzyme. This finding is analogous to the finding of the thermal stability studies. Kinetics analyses indicate that there is considerable inhibition of the E285A enzyme by glutathione and patulin.

Inhibitory constants were determined by varying the compound concentration and keeping the substrate concentration at KM level with a Dixon plot analysis.

Glutathione has a Ki value of 11 ± 2 µM (Appendix C2) and patulin has a Ki of 490 ±

70 µM (Appendix C3). This clearly shows the high affinity of these compounds to the enzyme binding pocket, leading to both inhibition and stability of the E285A enzyme.

Interestingly, as a control oxidized glutathione did not show any significant inhibition of the E285A mutant enzyme.

69

4.5 Virtual Screening of Compound Libraries

Since the crystal structures of different aspartoacylase enzyme forms are now available, a small library of fragment molecules was screened to identify druggable binding pockets on the enzyme surface. This approach has been successfully used in other proteins to identify small molecules that can bind and potentially stabilize the structure of a target protein.111 This in silico screening approach involves docking of probe molecules onto the protein of interest using the freely available FTMap server

(www.ftmap.bu.edu). The FTMap algorithm uses 16 small organic molecules

(molecular weight less than 100) as probes to identify binding hot spots that have low ligand binding free energy on the surface of proteins.112 The binding sites of these probe molecules are then ranked according to the number of molecules bound to each pocket.

ASPA native (PDB ID: 2O4H) and the most common E285A mutant (PDB ID:

4NFR) enzyme forms were submitted to the FTMap server. Virtual screening of 16 probe molecules resulted in 10 – and 12 – molecule binding clusters for native and mutant enzyme, respectively. Interestingly, the highest number of probes is bound to the dimer interface of the native enzyme and to the active site cavity in the E285A enzyme (Figure 4-6). Several molecules also showed binding deep in the active site cleft, where Thr118 and Thr119 are located in mutant enzyme. These residues make hydrogen bond contacts with the side chain of Glu285 in the native enzyme to maintain the correct active site geometry. If the binding of molecules at this site could restore the missing interactions in E285A that would result in an active site similar to that in the native enzyme, which would potentially enhance the activity of the mutant. On the other hand, the fact that many fragment molecules bound to the dimer interface of both native and mutant enzyme, could allow us to use rational drug

70 design to find a molecule that binds to the majority of the ASPA clinical mutants and improve both their stability and lifetime.

A

B

E285A

E285A

Figure 4-6: The binding of probe molecules to the native (A) and E285A (B) enzymes. The location of the active site is marked by the red square. The mutated residue is shown in red in the E285A structure.

71

4.6 Summary and Future Directions

Several different strategies have been employed to identify pharmacological chaperones for Canavan disease. High-throughput screening of compound libraries identified glutathione and patulin as promising lead compounds that could serve as a starting point for the development of potent pharmacological chaperones. Glutathione and patulin selectively bind to the E285A mutant enzyme. The integration of a

ThermoFluor assay, DSC studies, and kinetic studies have validated our findings for these lead compounds. Co-crystallization experiments of the aspartoacylase clinical mutants with these hit compounds will be carried out to identify the actual binding pocket and the specific binding interactions. Further chemical modification of these lead compounds will be conducted to improve the binding affinity, an approach that should result in the identification of molecules with greater potency. Overall our results show that pharmacological chaperones could be employed as a viable approach to overcome the catalytic deficiency of E285A mutation. These PCs thus could restore the activity by extending the lifetime of the enzyme. However, these results need to be further validated by in vivo cell culture experiments and animal studies.

72

Chapter 5

Biochemical Characterization of Aspartate N- acetyltransferase

N-acetylaspartate (NAA) is an abundant, brain-specific small molecule derived from the amino acid, L-aspartate. The concentration of NAA typically exceeds ten millimolar, which is second only to the levels of L-glutamate in the central nervous system.113 Due to its high abundance and its characteristic proton NMR signal, NAA has been used as a non-invasive diagnostic marker for neuronal health as determined by magnetic resonance imaging.23 This brain-specific molecule is synthesized in the neurons from acetyl coenzyme A and L-aspartate, catalyzed by an enzyme called aspartate N-acetyltransferase (AspNAT; EC 2.3.1.17). N-acetylaspartate serves as an acetyl group carrier from the neurons to the oligodendrocytes where it is hydrolyzed to regenerate acetate by the enzyme aspartoacylase. This regenerated acetate is subsequently incorporated into lipids in order to make the myelin sheath in the brain.114

The NAA biosynthetic enzyme, AspNAT, was also identified as “Shati” whose expression has been found to be induced by methamphetamine, a potent CNS stimulating drug, indicating a possible role for NAA in neuronal signal

73 transmission.115, 116 AspNAT is a membrane bound protein and is bimodally distributed in both mitochondria28, 117 and endoplasmic reticulum29. The gene coding

AspNAT was recently identified and has been annotated as NAT8L (N- acetyltransferase 8-like protein). Until recently AspNAT was thought to be a brain specific enzyme. However, in 2013 Pessenheiner A. R. et al. reported a higher level of extra-neuronal NAT8L gene expression in mitochondria of brown adipose cells, suggesting the existence of an additional NAA dependent lipid biosynthetic pathway in the adipose tissue.118

The molecular details of aspartate N-acetyltransferase structure and function still remains unknown. Since AspNAT is a membrane bound enzyme and is sensitive to high levels of detergents, purification of AspNAT in the active form is very challenging.119 Nevertheless, several attempts have been made to characterize the biochemical properties of aspartate N-acetyltransferase. Madhavarao and coworkers purified AspNAT from rat brain and reported that this enzyme existed as a multi- protein assembly with a molecular weight of around 650 kDa, and was found to be inactivated above 10 mM of the detergent, CHAPS (3-[(3-cholamidopropyl) dimethylammonia]-1-propanesulfonate).27 Wiame and colleagues identified the gene that encodes AspNAT in 2010 and expressed the human recombinant aspartate N- acetyltransferase in active form in HEK (human embryonic ) -293 cell line.119

Aspartate N-acetyltransferase is composed of 302 amino acids. Based on modeling five regions of AspNAT have been described.29 Region one is highly conserved among the N-acetyltransferase family of enzymes and consists of the N-terminal 40 residues. This region is followed by the and alanine rich region 2 which has a variable number of residues. Regions 3 and 5 are separated by a

74 hydrophobic region of about 30 residues (region 4) and is each composed of highly conserved sequences ~30 and ~150 residues in length, respectively (Figure 5-1A).

According to this putative model regions 3 and 5 together constitute the catalytic domain of the enzyme. The hydrophobic region 4 makes the membrane domain and anchors the AspNAT enzyme to the biological membranes (Figure 5-1B). This enzyme has been found to be highly specific towards L-aspartate with negligible activity towards L-glutamate, which is only one carbon longer amino acid homolog of

L-aspartate.27, 119, 120

A 1 ~40 ~90 ~120 ~150 302

N C

B Membrane

Figure 5-1: A. Schematic representation of AspNAT protein and B. the proposed model for AspNAT. The five regions proposed in AspNAT primary structure with the approximate lengths. The putative catalytic domain formed by regions 3 and 5 in blue and green, and the membrane domain in yellow are shown in the suggested model. The regions 1 and 2 are not modelled. (Adapted from Tahay, 201229)

75

Despite the known difficulties of purifying aspartate N-acetyltransferase, the biochemical studies of AspNAT are still essential as it would allow us to understand the precise biological roles of N-acetylaspartate in the brain. In addition, it would also help our understanding of the pathology of Canavan disease (CD), a rare neurological disorder with impaired NAA metabolism.3, 10, 17 The catalytic deficiency of the NAA catabolizing enzyme, aspartoacylase, leading to the accumulation of NAA is seen in

CD. However, the influence of abnormally elevated NAA to CD pathology is still poorly understood. The inhibition of the biosynthetic enzyme has successfully been used as a promising therapeutic strategy in Gaucher disease that has a deficiency in a catabolic enzyme.121 The same substrate reduction therapeutic approach could be employed in CD. Therefore, AspNAT would also be an interesting target to pursue for the development of a treatment therapy for CD to mitigate any deleterious effects that arose from elevated NAA. Keeping the importance of NAA biology in mind, we have started molecular cloning to characterize the enzyme aspartate N- acetyltransferase. Several AspNAT constructs have been made in our laboratory to express this enzyme in E. coli cells. Thioredoxin fused AspNAT protein was expressed in E. coli and has been partially purified by metal affinity chromatography with detectable activity. Further optimizations are being carried out to obtain the pure

AspNAT enzyme for subsequent biochemical studies.

76

5.1 Cloning of Aspartate N-acetyltransferase

The open reading frame of aspartate N-acetyltransferase (NAT8L) gene codon optimized to a mammalian expression system (hNAT8L) was provided by Dr. Aryan

Namboodiri of Uniformed Services University of Health Sciences, Bethesda, MD,

USA. A codon optimized NAT8L gene for E. coli expression was also purchased from

GenScript USA, Inc. These open reading frames were amplified using the oligonucleotide primers (Integrated DNA Technologies) that contained NcoI and

XhoI restriction sites for eNAT8L, and EcoRI and XhoI restriction sites for the hNAT8L gene (Table 5.1). Polymerase chain reaction was carried out in a 50 µl reaction consisted of 25 ng NAT8L template, 0.2 to 0.4 µM forward and reverse primers, 0.2 mM dNTPs, and 1 U of vent DNA polymerase in Thermopol buffer in the presence or absence of 2% DMSO (dimethyl sulfoxide). Thermal cycles were performed following the initial denaturation at 95 oC for 2 mins, and then 30 cycles of denaturation step 95 oC for 15 sec, annealing at 55 oC for 15 sec, and the extension at

72 oC for 60 sec. A final extension was done at 72 oC for 5 minutes to ensure the complete amplification.

Table 5.1: Oligonucleotide primers used in amplification of NAT8L genes

Primer Sequence Tm eNAT8L_For 5’- TAT ACC ATG GCT CAC TGT GGT CCG CCG GAT A -3’ 66.3 eNAT8L_Rev 5’- GTG TGG TGC TCG AGT TCT TCA CGC AGT TGC -3’ 65.8 hNAT8L_For 5’- ACG AGG AAT TCA TGC ACT GCG GCC CCC CTG A -3’ 70.5 hNAT8L_Rev 5’- TGG TGC TCG AGT CCT CCC TCA GCT GCA GTC T -3’ 69.3

Restriction sites are in bold

77

The NAT8L gene (~ 900 bp) was successfully amplified using the above

protocol only in the presence of 2% DMSO for eNAT8L, and irrespective of the

presence of DMSO for hNAT8L (Figure 5-2). The PCR amplified eNAT8L gene was

restriction digested by NcoI/XhoI and was cloned into pET28a (Novagen) and

pDR32b (generously provided by Dr. Donald Ronning, The University of Toledo)

expression vectors using a quick DNA kit (New England BioLabs Inc.). The

ligation reaction was transformed into E. coli DH5α cloning host and positive

colonies were identified by restriction mapping and gene sequencing at Eurofins

Genomics (Appendix D1). The sequences confirmed eNAT8L-pET28 and eNAT8L-

pDR32 were then transformed into an E. coli BL21 (DE3) expression cell line and

were preserved in 20% glycerol at – 80 oC for future studies.

Temp PCR3 PCR4 PCR5 PCR1 PCR2 PCR3 PCR4 MW PCR1 PCR2 MW only -DMSO +DMSO -Temp -DMSO + DMSO - Primer -Temp -DMSO +DMSO

1.5 kb

1.5 kb 1 kb 900 bp 1 kb 900 bp

MW= marker; Temp = template; - = in the absence; + = in the presence

Figure 5-2: PCR amplification of NAT8L genes as seen on 1% agarose gel. The amplification eNAT8L gene was observed only in the presence of 2% DMSO (PCR 2 and 5 in the left figure). The amplification of hNAT8L gene did not depend on DMSO (PCR 3 and 4 in the right figure).

78

5.2 Expression of Aspartate N-acetyltransferase

The recombinant AspNAT has a C-terminal hexa-histidine (His6) tag in the eNAT8L-pET28 construct and both N-terminal thioredoxin (Trx) and C-terminal His6 tag in the eNAT8L-pDR32 construct. A small scale pilot expression experiment was performed to identify optimum conditions for the aspartate N-acetyltransferase protein expression. An overnight starter culture of eNAT8L construct in BL21 E. coli cells (~100 μl) was inoculated into a 10 ml fresh LB media containing the appropriate antibiotic (30 μg/ml kanamycin for eNAT8L-pET28 and 100 μg/ml ampicillin for eNAT8L-pDR32) and was grow in an orbital shaker incubator (250

o rpm) at 37 C until OD600 reached 0.8. Then the expression of AspNAT was induced by 1 mM IPTG (isopropyl β-D-1-thiogalactopyranoside) for 6 hrs at 30 oC. One milliliter samples were withdrawn from the culture at different time intervals and cells were collected by centrifugation.

The total soluble and insoluble proteins were isolated by chemical cell lysis using the Bugbuster reagent (Novagen). The time dependent expression of AspNAT was analyzed by SDS-PAGE using 4 – 12% Bis-Tris gels (Life technologies) and

MES running buffer and by Western blotting. The Western blots were performed by transferring the pre-separated proteins by SDS-PAGE onto a nitrocellulose membrane

(Life Technologies). After blocking the membrane with 5% (w/v) bovine serum albumin (BSA) in Tris-buffered saline and Tween 20 solution (TBST), blotting was done with diluted monoclonal anti-polyhistidine antibodies conjugated with alkaline phosphatase (Sigma-Aldrich) diluted 1:4000 in 3% (w/v) BSA – TBST solution. The signals were developed by adding the chromogenic BCIP/NBT substrate (Sigma-

Aldrich).

79

The over-expression of the protein of interest over time can be followed by SDS-

PAGE. The recombinant human AspNAT expressed by eNAT8L-pET28 construct has a molecular weight of approximately 33 kDa, and the eNAT8L-pDR32 expressed Trx-

AspNAT fusion construct has a molecular weight around 45 kDa. The small scale pilot expression of both the constructs did not yield detectable AspNAT protein that could be distinguished among other proteins on an SDS gel (Figure 5-3A). However, the Western blot analysis showed the presence of several anti-His antibody reactive protein species from eNAT8L-pDR32 construct (Figure 5-3B). All of these proteins only appeared sometime after the addition of IPTG, while the absence of the protein of interest right after induction is seen in lane 1 next to the molecular weight marker of Figure 5-3B. Among these immune reactive proteins, a band corresponding to intact Trx-AspNAT fusion protein (expected size is ~ 45 kDa) between 42 and 57 kDa markers could also be identified. This AspNAT was expressed as a soluble protein as confirmed by the absence of similar bands in insoluble fractions (inclusion bodies) on the Western blot. While the recombinant poly-histidine tagged AspNAT has been found to be expressed in mammalian cell lines,119 our eNAT8L-pET28 construct did not express any AspNAT in E. coli BL21 (DE3) cell line that could be detected by Western blot analysis. The optimization trials of IPTG level and induction cell density revealed that induction of protein expression at OD600 around

0.8 gives higher yield, with negligible correlation of the level of IPTG (0.25 to 1.0 mM) added to the medium (Figure 5-3 C and D). For each induction time four distinguishable immune reactive protein bands were visible on the Western blot after the protein was induced by IPTG. Among these proteins, the highest molecular weight band (between 42 kDa and 57 kDa) was believed to be the intact Trx-AspNAT

80 protein (Figure 5-3D). However, its yield was low compared to the other bands observed on the Western blot.

A B 1 2 3 4 5 6 7 MW 1 2 3 4 5 6 7 1 2 3 4 5 6 7 MW 1 2 3 4 5 6 7

57 kDa 57 42 kDa 42

31 kDa 31 24

eNAT8L-pET28 eNAT8L-pDR32 eNAT8L-pET28 eNAT8L-pDR32

C D MW 1 2 3 4 5 6 7 8 9 10 11 12 13 14 MW 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 15

57 57 kDa 42 42 kDa 31 31 kDa 24 kDa

eNAT8L-pDR32 eNAT8L-pDR32

Figure 5-3: Expression profiles of AspNAT as seen on SDS-PAGE (A and C) and Western blot (B and D). Figures A&B show the expression of soluble proteins in lanes 1 to 4 at 0, 1, 3, 6 hours, respectively and insoluble proteins in lanes 5 to 7 at same time intervals. Figures C&D show the expression of soluble Trx-AspNAT fusion construct at

OD600 4 - 5 (lanes 1 to 7) and OD600 8 - 9 (lanes 8 to 14). The induction of protein with different IPTG levels: 0.25 mM (lanes 2&3), 0.5 mM (lanes 4&5), and 1 mM (lanes 6&7), and at different time intervals: 0 hr (lane 1 & 8), 3 hr (lanes 2,4, 6, 9, 11, and 13), and 6 hr (lanes 3, 5, 7, 10, 12, 14).

81

5.3 Purification of AspNAT

A poly-histidine tagged Trx-AspNAT fusion protein was expressed in E. coli

BL21 (DE3) cells in one liter cultures, and was partially purified using Ni immobilized metal affinity chromatography (Ni IMAC). Five grams of cells induced with 1 mM IPTG at 30 oC for 6 hrs were re-suspended in Ni-NTA buffer A (50 mM sodium phosphate pH 7.4, 300 mM NaCl, 10% glycerol, 20 mM imidazole, and 5 mM β-mercaptoethanol) and cells were lysed by sonication on ice using 30 second pulses for 8 minutes with 3 minute time intervals between each pulse. After centrifugation at 15,000 X g at 4 oC for 30 minutes and subsequent filtration through a 0.45 µm syringe filter, the soluble lysate was applied onto a 20 ml Ni-NTA column

(Qiagen) equilibrated with buffer A. The bound AspNAT protein was eluted using a linear gradient of 20 – 500 mM imidazole over 10 column volumes and collected into

6 ml fractions. The presence of AspNAT protein in the collected fractions was confirmed by SDS PAGE, Western blotting, and the activity assay (see the following section). The chromatogram of Ni affinity purification shows a single peak that eluted around 250 mM imidazole in fractions 11 through 16 (Figure 5-4A). SDS-PAGE shows that theses fractions also contain many non-specific proteins (Figure 5-4B).

However, among these proteins, a band that corresponds to the intact Trx-AspNAT fusion protein could be seen between 42 and 57 kDa markers, which was subsequently confirmed by Western blotting.

82

A

20140526 Supun eANAT pET32 OD08 IPTG 08 30C 6hr(1401119027)001:10_UV1_280nm 20140526 Supun eANAT pET32 OD08 IPTG 08 30C 6hr(1401119027)001:10_UV3_260nm A280 20140526 Supun eANAT pET32 OD08 IPTG 08 30C 6hr(1401119027)001:10_Fractions 20140526 Supun eANAT pET32 OD08 IPTG 08 30C 6hr(1401119027)001:10_Logbook A260 mAU 20140526 Supun eANAT pET32 OD08 IPTG 08 30C 6hr(1401119027)001:10_UV1_280nm 20140526 Supun eANAT pET32 OD08 IPTG 08 30C 6hr(1401119027)001:10_UV3_260nm 20140526 Supun eANAT pET32 OD08 IPTG 08 30C 6hr(1401119027)001:10_Fractions 20140526 Supun eANAT pET32 OD08 IPTG 08 30C 6hr(1401119027)001:10_Logbook

mAU

3000

200

2500 150

100 2000

50

1500 0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 120 140 160 180 200 220 240 ml 1000

500

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 50 100 150 200 ml

B MW 1 2 3 4 5 6 7 C MW 1 2 3 4 5 6 7

57 kDa 57 42 kDa 42 31 kDa 31

24 kDa 24

Figure 5-4: The chromatogram (A), SDS-PAGE (B), and western blot (C) profiles for preliminary purification of Trx-AspNAT using Ni IMAC. The proteins eluted at the peak observed in chromatographic separation are collected into fractions 11 through 17 and are loaded onto wells from 1 to 7, respectively in SDS-PAGE and Western blotting.

83

5.4 Kinetic Characterization of AspNAT

Aspartate N-acetyltransferase enzymatic activity was assessed by a continuous enzyme assay. In this assay, the formation of coenzyme A during AspNAT catalyzed reaction was coupled to a pyruvate dehydrogenase (PDH) – catalyzed reaction that reduces NAD+ to yield NADH (Figure 5-5). The assay was performed on a

SpectraMax 190 UV-visible spectrophotometer by monitoring the increasing

-1 -1 production of NADH at 340 nm (Ɛ340 = 6.22 mM cm ). The enzymatic reaction was comprised of assay buffer (20 mM sodium phosphate pH 7.4, 100 mM NaCl, 1 mM

CHAPS, 10% glycerol, and 1 mM DTT), 1 mM L-aspartate, 200 µM acetyl coenzyme A (acetyl-CoA), 0.05 U pyruvate dehydrogenase, 2 mM pyruvate, 0.2 mM

NAD+, and 0.2 mM thiamine pyrophosphate (TPP). The reaction was carried out in

200 μl volumes in 96 well clear plates and was initiated by addition of AspNAT enzyme.

The AspNAT activity of the purified fractions from Ni affinity chromatography was measured to determine the presence of the enzyme activity of interest in the fractions. To eliminate any interference from other enzymes that co-eluted with

AspNAT that could potentially reduce NAD+ and lead to false positive results, a control reaction without L-aspartate substrate was included. The chromatographic fractions 11 through 15 found to have measurable AspNAT activity (Appendix D2).

The kinetic parameters of partially purified aspartate N-acetyltransferase were determined by varying the concentration of the substrate, L-aspartate from 0 to 5 mM.

The calculated KM for L-aspartate was 0.46 ± 0.11 mM and the Vmax was 6.39 ± 0.4

-1 -1 mM s . This KM is comparable to the KM reported before for AspNAT purified from

27 rat brain. However, this value is five times higher than the KM reported for the

119 recombinant AspNAT expressed in HEK-293 cells. The KM for acetyl-CoA could 84 not be evaluated, because it is being recycled by the coupled reaction in this assay

(Figure 5-5). AspNAT did not show measurable activity towards D-aspartate or L- glutamate at the highest concentration (5 mM) tested. These findings reconfirmed the specificity of aspartate N-acetyltransferase towards L-aspartate. This level of specificity of AspNAT for L-aspartate is essential to overcome any side reactions or inhibition that could result due to the very high natural abundance of L-glutamate in the brain.

Aspartate N-acetyltransferase L-aspartate N-acetylaspartate

Pyruvate Acetyl-CoA CoA + NAD

Pyruvate TPP dehydrogenase

Acetyl-CoA NADH

CO 2

340 nm

Figure 5-5: A schematic representation of pyruvate dehydrogenase coupled AspNAT enzyme activity assay.

85

5. 5 Summary and Future Directions

The codon-optimized open reading frame of the N-acetylaspartate biosynthetic enzyme, aspartate N-acetyltransferase (NAT8L) was successfully cloned into pET28 and pDR32 bacterial plasmid vectors. The recombinant human AspNAT fused with the thioredoxin soluble partner was satisfactorily expressed in soluble fraction in

E.coli cells. This soluble Trx-AspNAT construct was partially purified by Ni affinity chromatography with measurable activity. The kinetic parameters of recombinant aspartate N-acetyltransferase enzyme were found to be analogous to those reported in the literature. However, further experiments are needed to optimize the yield and purity of this enzyme. The effect of the thioredoxin tag on AspNAT activity and stability also needs to be evaluated. After obtaining AspNAT in better yield and higher purity, enzyme inhibition studies will be initiated to find potential inhibitors for this enzyme that could help to understand the biological role of N-acetylaspartate in the brain, and to find potential therapies for Canavan disease. At the same time crystallization experiments will be carried out to solve the structures of the apo- enzyme and enzyme-substrate or enzyme-inhibitor complexes. Theses structural studies will provide mechanistic insights into the reaction catalyzed by the N- acetyltransferase enzyme.

86

References

1. Matalon, R., and Michals-Matalon, K. (1999) Biochemistry and molecular biology of Canavan disease, Neurochem Res 24, 507-513.

2. Kumar, S., Mattan, N. S., and de Vellis, J. (2006) Canavan disease: a white matter disorder, Ment Retard Dev Disabil Res Rev 12, 157-165.

3. Matalon, R., Kaul, R., Casanova, J., Michals, K., Johnson, A., Rapin, I., Gashkoff, P., and Deanching, M. (1989) SSIEM Award. Aspartoacylase deficiency: the enzyme defect in Canavan disease, J Inherit Metab Dis 12 Suppl 2, 329-331.

4. Canavan, M. M. (1931) Schindler's Encephalitis Periaxialis Diffusa, Arch Neurol Psychiatr 25, 299-308.

5. van Bogaert, L., Bertrand, I. (1949) Sur une idiotie familiale avec degenerescence spongieuse du nevraxe, Acta Neur Psychiatr Belg 49, 572- 585.

6. Traeger, E. C., and Rapin, I. (1998) The clinical course of Canavan disease, Pediatr Neurol 18, 207-212.

7. Adachi, M., Schneck, L., Cara, J., and Volk, B. W. (1973) Spongy degeneration of the central nervous system (van Bogaert and Bertrand type; Canavan's disease). A review, Hum Pathol 4, 331-347.

8. Goodhue, W. W., Jr., Couch, R. D., and Namiki, H. (1979) Spongy degeneration of the CNS: an instance of the rare juvenile form, Arch Neurol 36, 481-484.

9. Matalon, R., Michals, K., and Kaul, R. (1995) Canavan disease: from spongy degeneration to molecular analysis, J Pediatr 127, 511-517.

87

10. Matalon, R., Michals, K., Sebesta, D., Deanching, M., Gashkoff, P., and Casanova, J. (1988) Aspartoacylase deficiency and N-acetylaspartic aciduria in patients with Canavan disease, Am J Med Genet 29, 463-471.

11. Kaul, R., Gao, G. P., Aloya, M., Balamurugan, K., Petrosky, A., Michals, K., and Matalon, R. (1994) Canavan disease: mutations among Jewish and non- Jewish patients, Am J Hum Genet 55, 34-41.

12. Surendran, S., Michals-Matalon, K., Quast, M. J., Tyring, S. K., Wei, J., Ezell, E. L., and Matalon, R. (2003) Canavan disease: a monogenic trait with complex genomic interaction, Mol Genet Metab 80, 74-80.

13. Adachi, M., Wallace, B. J., Schneck, L., and Volk, B. W. (1966) Fine structure of spongy degeneration of the central nervous system (van Bogaert and Bertrand type), J Neuropathol Exp Neurol 25, 598-616.

14. Kvittingen, E. A., Guldal, G., Borsting, S., Skalpe, I. O., Stokke, O., and Jellum, E. (1986) N-acetylaspartic aciduria in a child with a progressive cerebral atrophy, Clin Chim Acta 158, 217-227.

15. Hagenfeldt, L., Bollgren, I., and Venizelos, N. (1987) N-acetylaspartic aciduria due to aspartoacylase deficiency--a new aetiology of childhood , J Inherit Metab Dis 10, 135-141.

16. Divry, P., Vianey-Liaud, C., Gay, C., Macabeo, V., Rapin, F., and Echenne, B. (1988) N-acetylaspartic aciduria: report of three new cases in children with a neurological syndrome associating macrocephaly and leukodystrophy, J Inherit Metab Dis 11, 307-308.

17. Matalon, R., Kaul, R., and Michals, K. (1993) Canavan disease: biochemical and molecular studies, J Inherit Metab Dis 16, 744-752.

18. Kaul, R., Casanova, J., Johnson, A. B., Tang, P., and Matalon, R. (1991) Purification, characterization, and localization of aspartoacylase from bovine brain, J Neurochem 56, 129-135.

19. Bluml, S. (1999) In vivo quantitation of cerebral metabolite concentrations using natural abundance 13C MRS at 1.5 T, J Magn Reson 136, 219-225.

20. Tallan, H. H., Moore, S., and Stein, W. H. (1956) N-Acetyl-L-aspartic acid in brain, J Biol Chem 219, 257-264.

88

21. Tallan, H. H. (1957) Studies on the distribution of N-acetyl-L-aspartic acid in brain, J Biol Chem 224, 41-45.

22. Rigotti, D. J., Kirov, II, Djavadi, B., Perry, N., Babb, J. S., and Gonen, O. (2011) Longitudinal whole-brain N-acetylaspartate concentration in healthy adults, AJNR Am J Neuroradiol 32, 1011-1015.

23. Rigotti, D. J., Inglese, M., and Gonen, O. (2007) Whole-brain N- acetylaspartate as a surrogate marker of neuronal damage in diffuse neurologic disorders, AJNR Am J Neuroradiol 28, 1843-1849.

24. Lin, A., Ross, B. D., Harris, K., and Wong, W. (2005) Efficacy of proton magnetic resonance spectroscopy in neurological diagnosis and neurotherapeutic decision making, NeuroRx 2, 197-214.

25. Demougeot, C., Garnier, P., Mossiat, C., Bertrand, N., Giroud, M., Beley, A., and Marie, C. (2001) N-Acetylaspartate, a marker of both cellular dysfunction and neuronal loss: its relevance to studies of acute brain injury, J Neurochem 77, 408-415.

26. Block, W., Traber, F., Flacke, S., Jessen, F., Pohl, C., and Schild, H. (2002) In-vivo proton MR-spectroscopy of the human brain: assessment of N- acetylaspartate (NAA) reduction as a marker for , Amino Acids 23, 317-323.

27. Madhavarao, C. N., Chinopoulos, C., Chandrasekaran, K., and Namboodiri, M. A. (2003) Characterization of the N-acetylaspartate biosynthetic enzyme from rat brain, J Neurochem 86, 824-835.

28. Ariyannur, P. S., Madhavarao, C. N., and Namboodiri, A. M. (2008) N- acetylaspartate synthesis in the brain: mitochondria vs. microsomes, Brain Res 1227, 34-41.

29. Tahay, G., Wiame, E., Tyteca, D., Courtoy, P. J., and Van Schaftingen, E. (2012) Determinants of the enzymatic activity and the subcellular localization of aspartate N-acetyltransferase, Biochem J 441, 105-112.

30. Sager, T. N., Thomsen, C., Valsborg, J. S., Laursen, H., and Hansen, A. J. (1999) Astroglia contain a specific transport mechanism for N-acetyl-L- aspartate, J Neurochem 73, 807-811.

89

31. Huang, W., Wang, H., Kekuda, R., Fei, Y. J., Friedrich, A., Wang, J., Conway, S. J., Cameron, R. S., Leibach, F. H., and Ganapathy, V. (2000) Transport of N-acetylaspartate by the Na(+)-dependent high-affinity dicarboxylate transporter NaDC3 and its relevance to the expression of the transporter in the brain, J Pharmacol Exp Ther 295, 392-403.

32. Chakraborty, G., Mekala, P., Yahya, D., Wu, G., and Ledeen, R. W. (2001) Intraneuronal N-acetylaspartate supplies acetyl groups for myelin lipid synthesis: evidence for myelin-associated aspartoacylase, J Neurochem 78, 736-745.

33. Namboodiri, A. M., Peethambaran, A., Mathew, R., Sambhu, P. A., Hershfield, J., Moffett, J. R., and Madhavarao, C. N. (2006) Canavan disease and the role of N-acetylaspartate in myelin synthesis, Mol Cell Endocrinol 252, 216-223.

34. Nordengen, K., Heuser, C., Rinholm, J. E., Matalon, R., and Gundersen, V. (2013) Localisation of N-acetylaspartate in oligodendrocytes/myelin, Brain Struct Funct.

35. Gehl, L. M., Saab, O. H., Bzdega, T., Wroblewska, B., and Neale, J. H. (2004) Biosynthesis of NAAG by an enzyme-mediated process in rat central nervous system neurons and glia, J Neurochem 90, 989-997.

36. Arun, P., Madhavarao, C. N., Moffett, J. R., and Namboodiri, M. A. (2006) Regulation of N-acetylaspartate and N-acetylaspartylglutamate biosynthesis by protein kinase activators, J Neurochem 98, 2034-2042.

37. Becker, I., Lodder, J., Gieselmann, V., and Eckhardt, M. (2010) Molecular characterization of N-acetylaspartylglutamate synthetase, J Biol Chem 285, 29156-29164.

38. Zuo, D., Bzdega, T., Olszewski, R. T., Moffett, J. R., and Neale, J. H. (2012) Effects of N-acetylaspartylglutamate (NAAG) peptidase inhibition on release of glutamate and dopamine in prefrontal cortex and nucleus accumbens in phencyclidine model of schizophrenia, J Biol Chem 287, 21773-21782.

39. Baslow, M. H. (2002) Evidence supporting a role for N-acetyl-L-aspartate as a molecular water pump in myelinated neurons in the central nervous system. An analytical review, Neurochem Int 40, 295-300.

40. Long, P. M., Moffett, J. R., Namboodiri, A. M., Viapiano, M. S., Lawler, S. E., and Jaworski, D. M. (2013) N-acetylaspartate (NAA) and N- 90

acetylaspartylglutamate (NAAG) promote growth and inhibit differentiation of glioma stem-like cells, J Biol Chem 288, 26188-26200.

41. Ariyannur, P. S., Arun, P., Barry, E. S., Andrews-Shigaki, B., Bosomtwi, A., Tang, H., Selwyn, R., Grunberg, N. E., Moffett, J. R., and Namboodiri, A. M. (2013) Do reductions in brain N-acetylaspartate levels contribute to the etiology of some neuropsychiatric disorders?, J Neurosci Res 91, 934-942.

42. Baslow, M. H. (2003) N-acetylaspartate in the vertebrate brain: metabolism and function, Neurochem Res 28, 941-953.

43. Moffett, J. R., Arun, P., Ariyannur, P. S., Garbern, J. Y., Jacobowitz, D. M., and Namboodiri, A. M. (2011) Extensive aspartoacylase expression in the rat central nervous system, Glia 59, 1414-1434.

44. Baslow, M. H., Suckow, R. F., Sapirstein, V., and Hungund, B. L. (1999) Expression of aspartoacylase activity in cultured rat macroglial cells is limited to oligodendrocytes, J Mol Neurosci 13, 47-53.

45. Moore, R. A., Le Coq, J., Faehnle, C. R., and Viola, R. E. (2003) Purification and preliminary characterization of brain aspartoacylase, Arch Biochem Biophys 413, 1-8.

46. Le Coq, J., An, H. J., Lebrilla, C., and Viola, R. E. (2006) Characterization of human aspartoacylase: the brain enzyme responsible for Canavan disease, Biochemistry 45, 5878-5884.

47. Le Coq, J., Pavlovsky, A., Malik, R., Sanishvili, R., Xu, C., and Viola, R. E. (2008) Examination of the mechanism of human brain aspartoacylase through the binding of an intermediate analogue, Biochemistry 47, 3484-3492.

48. Hershfield, J. R., Pattabiraman, N., Madhavarao, C. N., and Namboodiri, M. A. (2007) Mutational analysis of aspartoacylase: implications for Canavan disease, Brain Res 1148, 1-14.

49. Sommer, A., and Sass, J. O. (2012) Expression of aspartoacylase (ASPA) and Canavan disease, Gene 505, 206-210.

50. Namboodiri, A. M., Moffett, J. R., Arun, P., Mathew, R., Namboodiri, S., Potti, A., Hershfield, J., Kirmani, B., Jacobowitz, D. M., and Madhavarao, C. N. (2006) Defective myelin lipid synthesis as a pathogenic mechanism of Canavan disease, Adv Exp Med Biol 576, 145-163; discussion 361-143. 91

51. Madhavarao, C. N., Arun, P., Moffett, J. R., Szucs, S., Surendran, S., Matalon, R., Garbern, J., Hristova, D., Johnson, A., Jiang, W., and Namboodiri, M. A. (2005) Defective N-acetylaspartate reduces brain acetate levels and myelin lipid synthesis in Canavan's disease, Proc Natl Acad Sci U S A 102, 5221-5226.

52. Baslow, M. H. (2003) Brain N-acetylaspartate as a molecular water pump and its role in the etiology of Canavan disease: a mechanistic explanation, J Mol Neurosci 21, 185-190.

53. Ahmed, S. S., Li, H., Cao, C., Sikoglu, E. M., Denninger, A. R., Su, Q., Eaton, S., Liso Navarro, A. A., Xie, J., Szucs, S., Zhang, H., Moore, C., Kirschner, D. A., Seyfried, T. N., Flotte, T. R., Matalon, R., and Gao, G. (2013) A single intravenous rAAV injection as late as P20 achieves efficacious and sustained CNS Gene therapy in Canavan mice, Mol Ther 21, 2136-2147.

54. Leone, P., Shera, D., McPhee, S. W., Francis, J. S., Kolodny, E. H., Bilaniuk, L. T., Wang, D. J., Assadi, M., Goldfarb, O., Goldman, H. W., Freese, A., Young, D., During, M. J., Samulski, R. J., and Janson, C. G. (2012) Long- term follow-up after gene therapy for canavan disease, Sci Transl Med 4, 165ra163.

55. Arun, P., Madhavarao, C. N., Moffett, J. R., Hamilton, K., Grunberg, N. E., Ariyannur, P. S., Gahl, W. A., Anikster, Y., Mog, S., Hallows, W. C., Denu, J. M., and Namboodiri, A. M. (2010) Metabolic acetate therapy improves phenotype in the tremor rat model of Canavan disease, J Inherit Metab Dis 33, 195-210.

56. Segel, R., Anikster, Y., Zevin, S., Steinberg, A., Gahl, W. A., Fisher, D., Staretz-Chacham, O., Zimran, A., and Altarescu, G. (2011) A safety trial of high dose glyceryl triacetate for Canavan disease, Mol Genet Metab 103, 203- 206.

57. Malhi, G. S., Tanious, M., Das, P., and Berk, M. (2012) The science and practice of lithium therapy, Aust N Z J Psychiatry 46, 192-211.

58. Assadi, M., Janson, C., Wang, D. J., Goldfarb, O., Suri, N., Bilaniuk, L., and Leone, P. (2010) Lithium citrate reduces excessive intra-cerebral N-acetyl aspartate in Canavan disease, Eur J Paediatr Neurol 14, 354-359.

92

59. Kaul, R., Balamurugan, K., Gao, G. P., and Matalon, R. (1994) Canavan disease: genomic organization and localization of human ASPA to 17p13-ter and conservation of the ASPA gene during evolution, Genomics 21, 364-370.

60. Kaul, R., Gao, G. P., Matalon, R., Aloya, M., Su, Q., Jin, M., Johnson, A. B., Schutgens, R. B., and Clarke, J. T. (1996) Identification and expression of eight novel mutations among non-Jewish patients with Canavan disease, Am J Hum Genet 59, 95-102.

61. Elpeleg, O. N., and Shaag, A. (1999) The spectrum of mutations of the aspartoacylase gene in Canavan disease in non-Jewish patients, J Inherit Metab Dis 22, 531-534.

62. Zeng, B. J., Wang, Z. H., Ribeiro, L. A., Leone, P., De Gasperi, R., Kim, S. J., Raghavan, S., Ong, E., Pastores, G. M., and Kolodny, E. H. (2002) Identification and characterization of novel mutations of the aspartoacylase gene in non-Jewish patients with Canavan disease, J Inherit Metab Dis 25, 557-570.

63. Kaul, R., Gao, G. P., Michals, K., Whelan, D. T., Levin, S., and Matalon, R. (1995) Novel (cys152 > arg) missense mutation in an Arab patient with Canavan disease, Hum Mutat 5, 269-271.

64. Kobayashi, K., Tsujino, S., Ezoe, T., Hamaguchi, H., Nihei, K., and Sakuragawa, N. (1998) Missense mutation (I143T) in a Japanese patient with Canavan disease, Hum Mutat Suppl 1, S308-309.

65. Zhang, H., Liu, X., and Gu, X. (2010) Two novel missense mutations in the aspartoacylase gene in a Chinese patient with congenital Canavan disease, Brain Dev 32, 879-882.

66. Hussain, R., Daud, S., Kakar, N., Ahmad, A., Baloch, A. H., Tareen, A. M., Kakar, M. A., and Ahmad, J. (2012) A missense mutation (p.G274R) in gene ASPA causes Canavan disease in a Pakistani family, Mol Biol Rep 39, 6197- 6201.

67. Shaag, A., Anikster, Y., Christensen, E., Glustein, J. Z., Fois, A., Michelakakis, H., Nigro, F., Pronicka, E., Ribes, A., Zabot, M. T., and et al. (1995) The molecular basis of canavan (aspartoacylase deficiency) disease in European non-Jewish patients, Am J Hum Genet 57, 572-580.

93

68. Surendran, S., Bamforth, F. J., Chan, A., Tyring, S. K., Goodman, S. I., and Matalon, R. (2003) Mild elevation of N-acetylaspartic acid and macrocephaly: diagnostic problem, J Child Neurol 18, 809-812.

69. Tacke, U., Olbrich, H., Sass, J. O., Fekete, A., Horvath, J., Ziyeh, S., Kleijer, W. J., Rolland, M. O., Fisher, S., Payne, S., Vargiami, E., Zafeiriou, D. I., and Omran, H. (2005) Possible genotype-phenotype correlations in children with mild clinical course of Canavan disease, Neuropediatrics 36, 252-255.

70. Yalcinkaya, C., Benbir, G., Salomons, G. S., Karaarslan, E., Rolland, M. O., Jakobs, C., and van der Knaap, M. S. (2005) Atypical MRI findings in Canavan disease: a patient with a mild course, Neuropediatrics 36, 336-339.

71. Janson, C. G., Kolodny, E. H., Zeng, B. J., Raghavan, S., Pastores, G., Torres, P., Assadi, M., McPhee, S., Goldfarb, O., Saslow, B., Freese, A., Wang, D. J., Bilaniuk, L., Shera, D., and Leone, P. (2006) Mild-onset presentation of Canavan's disease associated with novel G212A point mutation in aspartoacylase gene, Ann Neurol 59, 428-431.

72. Sistermans, E. A., de Coo, R. F., van Beerendonk, H. M., Poll-The, B. T., Kleijer, W. J., and van Oost, B. A. (2000) Mutation detection in the aspartoacylase gene in 17 patients with Canavan disease: four new mutations in the non-Jewish population, Eur J Hum Genet 8, 557-560.

73. Rady, P. L., Vargas, T., Tyring, S. K., Matalon, R., and Langenbeck, U. (1999) Novel missense mutation (Y231C) in a turkish patient with canavan disease, Am J Med Genet 87, 273-275.

74. Zeng, B. J., Wang, Z. H., Torres, P. A., Pastores, G. M., Leone, P., Raghavan, S. S., and Kolodny, E. H. (2006) Rapid detection of three large novel deletions of the aspartoacylase gene in non-Jewish patients with Canavan disease, Mol Genet Metab 89, 156-163.

75. Olsen, T. R., Tranebjaerg, L., Kvittingen, E. A., Hagenfeldt, L., Moller, C., and Nilssen, O. (2002) Two novel aspartoacylase gene (ASPA) missense mutations specific to Norwegian and Swedish patients with Canavan disease, J Med Genet 39, e55.

76. Kaul, R., Gao, G. P., Balamurugan, K., and Matalon, R. (1993) Cloning of the human aspartoacylase cDNA and a common missense mutation in Canavan disease, Nat Genet 5, 118-123.

94

77. Karsten, W. E., Hunsley, J. R., and Viola, R. E. (1985) Purification of aspartase and aspartokinase-homoserine dehydrogenase I from Escherichia coli by dye-ligand chromatography, Anal Biochem 147, 336-341.

78. Zano, S., Wijayasinghe, Y. S., Malik, R., Smith, J., and Viola, R. E. (2013) Relationship between enzyme properties and disease progression in Canavan disease, J Inherit Metab Dis 36, 1-6.

79. Lim, W. K., Rosgen, J., and Englander, S. W. (2009) Urea, but not guanidinium, destabilizes proteins by forming hydrogen bonds to the peptide group, Proc Natl Acad Sci U S A 106, 2595-2600.

80. Almarza, J., Rincon, L., Bahsas, A., and Brito, F. (2009) Molecular mechanism for the denaturation of proteins by urea, Biochemistry 48, 7608- 7613.

81. Surendran, S., Matalon, K. M., Tyring, S. K., and Matalon, R. (2003) Molecular basis of Canavan's disease: from human to mouse, J Child Neurol 18, 604-610.

82. Bitto, E., Bingman, C. A., Wesenberg, G. E., McCoy, J. G., and Phillips, G. N., Jr. (2007) Structure of aspartoacylase, the brain enzyme impaired in Canavan disease, Proc Natl Acad Sci U S A 104, 456-461.

83. Otwinowski, Z., and Minor, W. (1997) Processing of X-ray Diffraction Data Collected in Oscillation Mode, Methods in Enzymology Volume 276, 307-326.

84. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997) Refinement of macromolecular structures by the maximum-likelihood method, Acta Crystallogr D Biol Crystallogr 53, 240-255.

85. Emsley, P., and Cowtan, K. (2004) Coot: model-building tools for molecular graphics, Acta Crystallogr D Biol Crystallogr 60, 2126-2132.

86. DeLano, W. L. (2002) The PyMOL Molecular Graphics System,, DeLano Scientific, San Carlos, CA.

87. Rivas, G., Ferrone, F., and Herzfeld, J. (2004) Life in a crowded world, EMBO Rep 5, 23-27.

95

88. Taverna, D. M., and Goldstein, R. A. (2002) Why are proteins marginally stable?, Proteins 46, 105-109.

89. Hartl, F. U., Bracher, A., and Hayer-Hartl, M. (2011) Molecular chaperones in protein folding and proteostasis, Nature 475, 324-332.

90. Kim, Y. E., Hipp, M. S., Bracher, A., Hayer-Hartl, M., and Hartl, F. U. (2013) Molecular chaperone functions in protein folding and proteostasis, Annu Rev Biochem 82, 323-355.

91. Meacham, G. C., Patterson, C., Zhang, W., Younger, J. M., and Cyr, D. M. (2001) The Hsc70 co-chaperone CHIP targets immature CFTR for proteasomal degradation, Nat Cell Biol 3, 100-105.

92. Ron, I., and Horowitz, M. (2005) ER retention and degradation as the molecular basis underlying Gaucher disease heterogeneity, Hum Mol Genet 14, 2387-2398.

93. Aguzzi, A., and O'Connor, T. (2010) Protein aggregation diseases: pathogenicity and therapeutic perspectives, Nat Rev Drug Discov 9, 237-248.

94. Arosio, P., Vendruscolo, M., Dobson, C. M., and Knowles, T. P. (2014) Chemical kinetics for drug discovery to combat protein aggregation diseases, Trends Pharmacol Sci 35, 127-135.

95. Balch, W. E., Morimoto, R. I., Dillin, A., and Kelly, J. W. (2008) Adapting proteostasis for disease intervention, Science 319, 916-919.

96. Sanchez-Ruiz, J. M. (2007) Ligand effects on protein thermodynamic stability, Biophys Chem 126, 43-49.

97. Boyd, R. E., Lee, G., Rybczynski, P., Benjamin, E. R., Khanna, R., Wustman, B. A., and Valenzano, K. J. (2013) Pharmacological chaperones as therapeutics for lysosomal storage diseases, J Med Chem 56, 2705-2725.

98. Pey, A. L., Ying, M., Cremades, N., Velazquez-Campoy, A., Scherer, T., Thony, B., Sancho, J., and Martinez, A. (2008) Identification of pharmacological chaperones as potential therapeutic agents to treat phenylketonuria, J Clin Invest 118, 2858-2867.

96

99. Santos-Sierra, S., Kirchmair, J., Perna, A. M., Reiss, D., Kemter, K., Roschinger, W., Glossmann, H., Gersting, S. W., Muntau, A. C., Wolber, G., and Lagler, F. B. (2012) Novel pharmacological chaperones that correct phenylketonuria in mice, Hum Mol Genet 21, 1877-1887.

100. Hanrahan, J. W., Sampson, H. M., and Thomas, D. Y. (2013) Novel pharmacological strategies to treat cystic fibrosis, Trends Pharmacol Sci 34, 119-125.

101. Bernier, V., Lagace, M., Bichet, D. G., and Bouvier, M. (2004) Pharmacological chaperones: potential treatment for conformational diseases, Trends Endocrinol Metab 15, 222-228.

102. Valenzano, K. J., Khanna, R., Powe, A. C., Boyd, R., Lee, G., Flanagan, J. J., and Benjamin, E. R. (2011) Identification and characterization of pharmacological chaperones to correct enzyme deficiencies in lysosomal storage disorders, Assay Drug Dev Technol 9, 213-235.

103. Guce, A. I., Clark, N. E., Rogich, J. J., and Garman, S. C. (2011) The molecular basis of pharmacological chaperoning in human alpha- galactosidase, Chem Biol 18, 1521-1526.

104. Clark, N. E., Metcalf, M. C., Best, D., Fleet, G. W., and Garman, S. C. (2012) Pharmacological chaperones for human alpha-N-acetylgalactosaminidase, Proc Natl Acad Sci U S A 109, 17400-17405.

105. Wu, X., Katz, E., Della Valle, M. C., Mascioli, K., Flanagan, J. J., Castelli, J. P., Schiffmann, R., Boudes, P., Lockhart, D. J., Valenzano, K. J., and Benjamin, E. R. (2011) A pharmacogenetic approach to identify mutant forms of alpha-galactosidase A that respond to a pharmacological chaperone for Fabry disease, Hum Mutat 32, 965-977.

106. Ishii, S. (2012) Pharmacological chaperone therapy for Fabry disease, Proc Jpn Acad Ser B Phys Biol Sci 88, 18-30.

107. Pantoliano, M. W., Petrella, E. C., Kwasnoski, J. D., Lobanov, V. S., Myslik, J., Graf, E., Carver, T., Asel, E., Springer, B. A., Lane, P., and Salemme, F. R. (2001) High-density miniaturized thermal shift assays as a general strategy for drug discovery, J Biomol Screen 6, 429-440.

108. Poklar, N., Lah, J., Salobir, M., Macek, P., and Vesnaver, G. (1997) pH and temperature-induced molten globule-like denatured states of equinatoxin II: a

97

study by UV-melting, DSC, far- and near-UV CD spectroscopy, and ANS fluorescence, Biochemistry 36, 14345-14352.

109. Niesen, F. H., Berglund, H., and Vedadi, M. (2007) The use of differential scanning fluorimetry to detect ligand interactions that promote protein stability, Nat Protoc 2, 2212-2221.

110. Matulis, D., Kranz, J. K., Salemme, F. R., and Todd, M. J. (2005) Thermodynamic stability of carbonic anhydrase: measurements of binding affinity and stoichiometry using ThermoFluor, Biochemistry 44, 5258-5266.

111. Landon, M. R., Lieberman, R. L., Hoang, Q. Q., Ju, S., Caaveiro, J. M., Orwig, S. D., Kozakov, D., Brenke, R., Chuang, G. Y., Beglov, D., Vajda, S., Petsko, G. A., and Ringe, D. (2009) Detection of ligand binding hot spots on protein surfaces via fragment-based methods: application to DJ-1 and glucocerebrosidase, J Comput Aided Mol Des 23, 491-500.

112. Hall, D. H., Grove, L. E., Yueh, C., Ngan, C. H., Kozakov, D., and Vajda, S. (2011) Robust identification of binding hot spots using continuum electrostatics: application to hen egg-white lysozyme, J Am Chem Soc 133, 20668-20671.

113. Tsai, G., and Coyle, J. T. (1995) N-acetylaspartate in neuropsychiatric disorders, Prog Neurobiol 46, 531-540.

114. Mehta, V., and Namboodiri, M. A. (1995) N-acetylaspartate as an acetyl source in the nervous system, Brain Res Mol Brain Res 31, 151-157.

115. Niwa, M., Nitta, A., Mizoguchi, H., Ito, Y., Noda, Y., Nagai, T., and Nabeshima, T. (2007) A novel molecule "shati" is involved in methamphetamine-induced hyperlocomotion, sensitization, and conditioned place preference, J Neurosci 27, 7604-7615.

116. Miyamoto, Y., Ishikawa, Y., Iegaki, N., Sumi, K., Fu, K., Sato, K., Furukawa- Hibi, Y., Muramatsu, S., Nabeshima, T., Uno, K., and Nitta, A. (2014) Overexpression of Shati/Nat8l, an N-acetyltransferase, in the nucleus accumbens attenuates the response to methamphetamine via activation of group II mGluRs in mice, Int J Neuropsychopharmacol 17, 1283-1294.

117. Lu, Z. H., Chakraborty, G., Ledeen, R. W., Yahya, D., and Wu, G. (2004) N- Acetylaspartate synthase is bimodally expressed in microsomes and mitochondria of brain, Brain Res Mol Brain Res 122, 71-78.

98

118. Pessentheiner, A. R., Pelzmann, H. J., Walenta, E., Schweiger, M., Groschner, L. N., Graier, W. F., Kolb, D., Uno, K., Miyazaki, T., Nitta, A., Rieder, D., Prokesch, A., and Bogner-Strauss, J. G. (2013) NAT8L (N-acetyltransferase 8-like) accelerates lipid turnover and increases energy expenditure in brown adipocytes, J Biol Chem 288, 36040-36051.

119. Wiame, E., Tyteca, D., Pierrot, N., Collard, F., Amyere, M., Noel, G., Desmedt, J., Nassogne, M. C., Vikkula, M., Octave, J. N., Vincent, M. F., Courtoy, P. J., Boltshauser, E., and van Schaftingen, E. (2010) Molecular identification of aspartate N-acetyltransferase and its mutation in hypoacetylaspartia, Biochem J 425, 127-136.

120. Ariyannur, P. S., Moffett, J. R., Manickam, P., Pattabiraman, N., Arun, P., Nitta, A., Nabeshima, T., Madhavarao, C. N., and Namboodiri, A. M. (2010) Methamphetamine-induced neuronal protein NAT8L is the NAA biosynthetic enzyme: implications for specialized acetyl coenzyme A metabolism in the CNS, Brain Res 1335, 1-13.

121. Cox, T. M., Aerts, J. M., Andria, G., Beck, M., Belmatoug, N., Bembi, B., Chertkoff, R., Vom Dahl, S., Elstein, D., Erikson, A., Giralt, M., Heitner, R., Hollak, C., Hrebicek, M., Lewis, S., Mehta, A., Pastores, G. M., Rolfs, A., Miranda, M. C., Zimran, A., and Advisory Council to the European Working Group on Gaucher, D. (2003) The role of the iminosugar N- butyldeoxynojirimycin (miglustat) in the management of type I (non- neuronopathic) Gaucher disease: a position statement, J Inherit Metab Dis 26, 513-526.

99

Appendix A

Appendix A1

Mutant Primer For 5’-ACG ATC CCA CTG GGC AGA GAC TGT ACC GTG TAC-3’ G274R2 Rev 5’-GTA CAC GGT ACA GTC TCT GCC CAG TGG GAT CGT-3’ For 5’-CCT GTG GGT ATA GAA GTT GGT ACT CAG CCT C-3’ P181T2 Rev 5’-G AGG CTG AGT ACC AAC TTC TAT ACC CAC AGG-3’ For 5’-GAG AAA GTT GAT TGC CCC CGG GAT GAA AAT GGA G-3’ Y231C2 Rev 5’-C TCC ATT TTC ATC CCG GGG GCA ATC AAC TTT CTC-3’ For 5’-CCA CTG CAT CGT GGG GAT CCC ATG TTT TTA ACT -3’ P257R2 Rev 5’-AGT TAA AAA CAT GGG ATC CCC ACG ATG CAG TGG-3’ For 5’-CAG ATG TTT CAT TAC ACT AAG ACT TCT CTG GCT C-3’ I143T1 Rev 5’-G AGC CAG AGA AGT CTT AGT GTA ATG AAA CAT CTG-3’ For 5’-CAG ATG TTT CAT TAC TTT AAG ACT TCT CTG GCT C-3’ I143F2 Rev 5’-G AGC CAG AGA AGT CTT AAA GTA ATG AAA CAT CTG-3’ For 5’-CAT TTC AAT GAA GGA GAA GAA TTT CCT CCC TGC -3’ K213E Rev 5’-GCA GGG AGG AAA TTC TTC TCC TTC ATT GAA ATG-3’ For 5’-GAC CTG AAT CAC ATT TTT GAC CTT GAA AAT CTT GGC -3’ R71H Rev 5’-GCC AAG ATT TTC AAG GTC AAA AAT GTG ATT CAG GTC-3’ For 5’-GAG GCC GCA TGT TAC GAA AAG AAA GAA GCT -3’ Y288C Rev 5’-AGC TTC TTT CTT TTC GTA ACA TGC GGC CTC-3’ For 5’-CAT CCT AAT CTG CAG GTT CAA GAC TGG AAA CCA C-3’ D249V Rev 5’-G TGG TTT CCA GTC TTG AAC CTG CAG ATT AGG ATG-3’ For 5’-CAC AAC ACC ACC TCT ATC ATG GGG TGC ACT C-3’ N121I Rev 5’-G AGT GCA CCC CAT GAT AGA GGT GGT GTT GTG-3’ For 5’-CG CTC AAT GAA AAA AGT ATT CGC TGC TG-3’ A305E2 Rev 5’-CA GCA GCG AAT ACT TTT TTC ATT GAG CG-3’ For 5’-C GAA AAG AAA GAA GCT TCT GCA AAG ACA ACT AAA CTA ACG -3’ F295S Rev 5’-CGT TAG TTT AGT TGT CTT TGC AGA AGC TTC TTT CTT TTC G-3’ For 5’-G TAC CCC CTG TTT GTG AAT GCG GCC GCA TAT TAC G-3’ E285A2 Rev 5’-C GTA ATA TGC GGC CGC ATT CAC AAA CAG GGG GTA C-3’ For 5’-CT CTG GCT CCA CTA CCC TGG TAC GTT TAT CTG -3’ C152W1 Rev 5’-CAG ATA AAC GTA CCA GGG TAG TGG AGC CAG AG-3’ For 5’-CT CTG GCT CCA CTA CCC CGC TAC GTT TAT CTG -3’ C152R2 Rev 5’-CAG ATA AAC GTA GCG GGG TAG TGG AGC CAG AG-3’

DNA primers used in site directed mutagenesis of ASPA gene. The mutated nucleotide is shown in bold. Several clinical mutants were constructed by Dr. Radhika Malik (1) and by Dr. Stephen Zano (2).

100

Appendix A2

Protein score is -10*Log(P), where P is the probability that the observed match is a random event. Protein scores greater than 86 are significant (p<0.05).

gi|254568544 Mass: 36976 Score: 239 Expect: 2.7e-17 Matches: 16

Mitochondrial alcohol dehydrogenase isozyme III [Komagataella pastoris GS115]

gi|4557335 Mass: 36225 Score: 186 Expect: 6.7e-11 Matches: 14 aspartoacylase [Homo sapiens]

Mascot search results for the tryptic digested inactive (top panel) and active (bottom panel) proteins eluted from Source 15Q column.

101

Appendix A3

K213E N121I

Y288C

Examples of some thermal denaturation curves obtained in the differential scanning calorimetry experiment. Top panel shows the thermal transitions in two different mutants, and the bottom panel shows the reproducibility of data using different concentrations of the same mutant.

102

Appendix B

Appendix B1

ASPA E285A Item R (nm) %Pd MW-R (kDa) Peak 1 3.8 0.0 78 Peak 2 34.6 18.6 13468

ASPA Y231C Item R (nm) %Pd MW-R (kDa) Peak 1 0.5 15.0 1 Peak 2 3.7 10.3 73 Peak 3 34.7 9.0 13494

Representative DLS profiles of two ASPA clinical mutants, showing the presence of the dimer state of the ASPA mutants in solution.

103

Appendix B2

Urea (M)

MW 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

1048 kDa

480 kDa

146 kDa E285A dimer form

66 kDa E285A monomer form

Native gel electrophoresis profile of ASPA E285A mutant incubated with increasing amounts of urea. At 2.0 M urea and above the enzyme dimer dissociates into monomers and unfolded aggregates.

104

Appendix B3

A R63

E178

Y288

R168 N70

R71 Y164

B R63

E178 Y288

N70

R71 R168 Y164

2Fo-Fc difference electron density map for the active site residues and the intermediate analog of ASPA E285A mutant subunit A (A) and subunit B (B) contoured at 1.5 σ.

105

Appendix C

Appendix C1

A

0 μM

100 μM

B Glutathione

Ellagic acid Patulin

(A) A representative example of thermal denaturation curves obtained in dose response study. (B) Molecular structures of three hit compounds. 106

Appendix C2

A

0 μM

60 μM

240 A

Time (Sec)

B

0 μM

100 μM

240 A

Time (Sec)

Enzyme kinetics for glutathione with ASPA E285A (A) and native (B) enzymes.

107

Appendix C3

A

0 mM

240 A 1.6 mM

Time (Sec)

B

0 mM

1.6 mM

240 A

Time (Sec)

Enzyme kinetics for patulin with ASPA E285A (A) and native (B) enzymes.

108

Appendix D

Appendix D1

A

B

The amino acid sequence alignments for AspNAT-pET28 (A) and AspNAT-pDR32 (B) constructs. 109

Appendix D2

2 mM Asp 0.08

0.07

0.06 1 mM Asp

0.05

40 3

A 0.04

0.03

0.02 0 mM Asp 0.01

0 0 500 1000 1500 2000 2500 TimeTime (secs) (Sec) Vmax Points = 721 Well A8 D8 F8 Vmax 0.291 1.171 1.793 R^2 0.937 0.978 0.974

The activity of Trx-AspNAT fusion construct determined by the pyruvate dehydrogenase coupled aspartate N-acetyltransferase assay. The activity of AspNAT is dependent on the level of the substrate, L-aspartate.

110