MIAMI UNIVERSITY

The Graduate School

Certificate for Approving the Dissertation

We hereby approve the Dissertation

of

Meghan M Holdorf

Candidate for the Degree Doctor of Philosophy

Advisor______Dr. Chris A. Makaroff

Reader______Dr. Michael W. Crowder

Reader______Dr. Ann E. Hagerman

Reader______Dr. Gary A. Lorigan

Reader______Dr. John Z. Kiss

ABSTRACT

CHARACTERIZATION OF ARABIDOPSIS ETHE1, A ASSOCIATED WITH ETHYLMALONIC ENCEPHALOPATHY

by M.M.Holdorf

Mutations in the ETHE1 gene result in the complex metabolic disease ethylmalonic encephalopathy, which is characterized by symmetric brain lesions, lactic academia, elevated excretion of ethylmalonic acid, and death in the first decade of life. ETHE1-like are found in a wide range of organisms; however, to date, a detailed characterization of ETHE1 has not been performed. Therefore, neither the structure nor the function has been established for the in any organism. In this dissertation, a full structural characterization of the Arabidopsis homolog of ETHE1 as well as information on its functional role in plants is presented. We have obtained the first crystal structure of an ETHE1-like as well as performed and preliminary substrate analyses providing new insights into the possible role and substrate of ETHE1. In addition, we demonstrate that ETHE1 is essential for both plant growth and development.

Characterization of Arabidopsis ETHE1, a Gene Associated With Ethylmalonic Encephalopathy

A Dissertation

Submitted to the faculty of Miami University in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

Department of Chemistry and Biochemistry

By

Meghan Marie Holdorf

Miami University Oxford, Ohio 2008

Dissertation Director: Dr. Christopher A. Makaroff

Table of Contents

Chapter 1 1 Introduction 1 1.1 Ethylmalonic Encephalopathy (EE) 1 1.1.1 Symptoms of Ethylmalonic Encephalopathy 1 1.1.2 Biochemical Markers of EE 1 1.1.3 Ethylmalonic Acid 2 1.1.4 Role of EMA in metabolic pathways 2 1.1.5 Current treatments for EE 4 1.2 Ethylmalonic Encephalopathy Protein 1 (ETHE1) 4 1.3 Metallo-β-lactamase Fold Family of 6 1.4 The 6 1.4.1 The Glyoxalase System 6 1.4.2 Arabidopsis Glyoxalase II 8 1.5 Hepatoma Subtracted-cDNA Library Clone One 11 1.6 11 1.6.1 Arabidopsis thaliana as a Model system 11 1.6.2 Loss of Function Mutants of Arabidopsis thaliana 13 1.6.3 Embryo Lethal Mutants of Arabidopsis thaliana 13 1.6.4 Seed Development in Arabidopsis thaliana 13 1.7 Sections of the Dissertation 17 1.8 References 19 Chapter 2: Structure of an ETHE1-like Protein from Arabidopsis thaliana 26 2.1 Summary 28 2.2 Introduction 29 2.3 Material and Methods 30 2.3.1 Cloning, Expression, and Purification 30 2.3.2 Crystallization 31

ii 2.3.3 Data Collection 31 2.3.4 Structure Determination and Refinement 31 2.4 Results and Discussion 32 2.4.1 Overall Fold 32 2.4.2 Dimer 34 2.4.3 Metal 34 2.4.4 40 2.4.5 Structural Basis for Encephalopathy 41 2.4.6 Sequence Analysis 43 2.5 Conclusions 44 2.6 Acknowledgements 45 2.7 References 46 Chapter 3: Spectroscopic Studies of Arabidopsis ETHE1, a Glyoxalase II-like Protein 50 3.1 Summary 51 3.2 Introduction 52 3.3 Material and Methods 54 3.3.1 Over-expression and Purification 54 3.3.2 Extinction Coefficient Determination 54 3.3.3 Metal Analysis 55 3.3.4 Native Molecular Weight Determination 55 3.3.5 Substrate Analysis 55 3.3.6 EPR Spectroscopy 56 3.3.7 1H NMR Spectroscopy 56 3.4 Results 56 3.4.1 Over-expression, Purification, and Characterization of Arabidopsis ETHE1 56 3.4.2 ETHE1 Functions as a Dimer 57 3.4.3 Calculation of the ETHE1 Extinction Coefficient 57 3.4.4 Metal Analysis 60 3.4.5 ETHE1 Does Not Hydrolyze SLG 63

iii 3.4.6 Spectroscopic Studies on ETHE1 64 3.5 Discussion 68 3.6 References 73 Chapter 4: ETHE1, a Gene Associated with Human Ethylmalonic Encephalopathy, is Essential for Endosperm Development in A. thaliana 78 4.1 Summary 80 4.2 Introduction 81 4.3 Material and Methods 84 4.3.1 Plant Material 84 4.3.2 Phylogenetic Analysis of β-lactamase Proteins 84 4.3.3 Molecular Analysis of ETHE1 85 4.3.4 Microscopy 87 4.4 Results 88 4.4.1 Molecular Analysis of ETHE1 88 4.4.2 ETHE1 is Required for Early Seed Development 92 4.4.3 ETHE1 Expression and Localization 99 4.5 Discussion 103 4.5.1 ETHE1 is Essential for Early Endosperm Development 104 4.5.2 Potential Role(s) of ETHE1 106 4.6 Acknowledgements 109 4.7 References 110

Chapter 5: Functional Studies on Arabidopsis Plants and Cell Cultures Expressing an Inducible ETHE1 RNAi 117 5.1 Summary 118 5.2 Introduction 119 5.3 Material and Methods 120 5.3.1 Plant Material 120 5.3.2 Cell Culture 121

iv 5.3.3 Generation of Inducible ETHE1-RNAi Arabidopsis Plants 121 5.3.4 Generation of Inducible ETHE1-RNAi Arabidopsis Suspension cell cultures 121 5.3.5 Molecular Analysis 122 5.3.6 EMA Toxicity studies in Arabidopsis cell culture 122 5.3.7 Microscopy 123 5.4 Results 123 5.4.1 Generation of Transgenic Plants and Suspension Cell Cultures Expressing the Inducible pX7-ETHE1-RNAi Construct 123 5.4.2 Time Course Expression Analysis of px7-ETHE1- RNAi in Cell Cultures 123 5.4.3 ETHE1 is Critical for Arabidopsis Cell Suspension Growth 127 5.4.4 Inducible Expression of ETHE1-RNAi Results in Filamentous Growth and Arrest of Normal Division in Arabidopsis Cell Culture 129 5.4.5 Incubation of Arabidopsis Cell Culture with EMA Undergoes Autophagy 129 5.4.6 ETHE1 is Essential for Plant Germination 132 5.5 Discussion 136 5.5.1ETHE1 is Critical for Cell Survival in Arabidopsis 140 Cell Culture 5.5.2 ETHE1 is Essential for Seed Germination 141 5.6 References 143 Chapter 6: Over-expression of Arabidopsis ETHE1 Results in Enhanced Growth Properties of Plants 147 6.1 Summary 148 6.2 Introduction 149 6.3 Material and Methods 150

v 6.3.1 Plant Material and Growth Conditions 150 6.3.2 Generation of ETHE1-OE plants and cell cultures 152 6.3.3 Microscopy 156 6.3.4 N-terminal Analysis and Protein Localization 156 6.4 Results and Discussion 157 6.4.1 Localization and N-terminal Determination Studies 157 6.4.2 Generation of ETHE1-OE Arabidopsis plants 165

6.4.3 ETHE1256-OE Plants Show Resistance to High of Valine 165

6.4.4 Over-expression of ETHE1256 Leads to Earlier Bolting and Flowering in Arabidopsis 170

6.4.5 ETHE1256 Expression Enhances Seed Yield in Arabidopsis 172

6.4.6 Arabidopsis ETHE1256-OE Plants have Thicker Primary Inflorescence Stems 175

6.4.7 ETHE1256-OE in Nicotiana tabacum Results in Enhanced Growth Properties 177 6.4.8 Concluding remarks 180 6.5 References 182 Chapter 7: Conclusions 190 7.1 Scientific Problems Addressed in this Dissertation 190 7.2 Structural Analysis of ETHE1 in Arabidopsis 190 7.3 Functional Analysis of ETHE1 in Arabidopsis 191 7.4 Potential Roles of ETHE1 193 7.5 Future Aspects and Directions of ETHE1 Analysis 194 7.5.1 Microarray 195 7.5.2 NMR Metabolomics and NMR Metabolite Tracing 195 7.5.3 Inducible pX7-ETHE1-OE Constructs 196 7.5.4 Characterization of Human ETHE1 197 7.6 References 198

vi List of Tables 2.1 Data-Collection and Refinement 33 3.1 Determination of ETHE1’s Extinction Coefficient 62

6.1 Peptide Matches of ETHE1294 Purified Protein 164

vii

List of Figures

1.1 Metabolic Routes of EMA Production 3 1.2 Pairwise Alignment of ETHE1 Homologs 5 1.3 Sequence Alignment of Human ETHE1 and Glyoxalase 7 1.4 The Glyoxalase System 9 1.5 Sequence Alignment of GLX2-3(ETHE1) and ETHE1 10 1.6 Sequence Alignment of the Arabidopsis Glyoxalase II isozymes 12 1.7 Life Cycle of Arabidopsis 14 2.1 The At1g53580 monomer 2.2 Sequence Alignment of AtETHE1, ETHE1, GLX2-5, and Human 35 Glyoxalase II (Human GLX2) 36 2.3 The AtETHE1 Dimer 37 2.4 Overlay of the Metal-binding Residues in the AtETHE1 enzyme (Magenta) and the GLX2-5 enzyme (Cyan). 39 2.5 Overlay of the Substrate Binding Residues from the Human Glyoxalase II (Cyan) with the Equivalent Residues in the AtETHE1 Enzyme (Magenta). 42 3.1 Sequence Alignment of ETHE1 and Select Metallo-β-lactamase Fold Proteins 58 3.2 Gel Filtration Elution Profile of Internal Protein Standards and ETHE1 59 3.3 Protein Quantification of ETHE1 Based off of Known Amounts of GLX2-5 61 3.4 1H NMR Spectrum of 3.5 mM Bound ETHE1 at pH 7.2 65 3.5 EPR Spectrum of 4.1 mM ETHE1 under Different Conditions 67 3.6 Dinuclear Metal Binding Center Model for A. thaliana ETHE1 69 4.1 Molecular Characterization of ETHE1 89 4.2 Inactivation of ETHE1 Disrupts Seed Development 93 4.3 Embryo Development in ETHE1(+/-) Siliques 95 4.4 Endosperm Development is Abnormal in ethe1(-/-) Seeds 98 4.5 ETHE1 Expression Patterns 100

viii 4.6 Immunolocalization of ETHE1 in Wild-type Buds 103 5.1 Generation of Inducible RNAi 125 5.2 Time Course Studies of Induced ETHE1-RNAi in Arabidopsis Cells 128 5.3 Effects of ETHE1-RNAi on Cell Culture Growth and Survival 131 5.4 Morphology of Arabidopsis Cell Cultures Expressing ETHE1-RNAi 133 5.5 EMA Toxicity in Arabidopsis Cell Cultures 134 5.6 Inducible ETHE1-RNAi Studies on Seed Germination 137 5.7 Effects of Inducible ETHE1-RNAi on Plant Development 138 6.1 Molecular Analysis of Arabidopsis ETHE1 154

6.2 Localization and Purification of ETHE1256 and ETHE1294 From Transgenic Arabidopsis Cell Cultures 158

6.3 Peptide Mapping of ETHE1294 using MALDI-TOF MS Spectroscopy 161 6.4 Expression Analysis of Transgenic Plants Over-expressing Either

ETHE1256 or ETHE1294 166 6.5 Metabolic Routes of EMA Production 168 6.6 Effect of Exogenous Valine on Seedling Growth 171

6.7 The Effect of Over-expressing Either ETHE1256 or ETHE1294 on Plant Growth in Arabidopsis 173

6.8 The Effect of ETHE1256-OE in Arabidopsis on Senescence, Seed Yield, Dry Mass, and Inflorescence Stem Thickness 176

6.9 Effects of Over-expressing Arabidopsis ETHE1256 in Nicotiana Tobacum 178 .

ix Chapter 1

Introduction

1.1 Ethylmalonic Encephalopathy

1.1.1 Symptoms of Ethylmalonic Encephalopathy. Ethylmalonic encephalopathy (EE) is a rare, autosomal recessive disorder mainly found in families of Mediterranean (1-5) or Arabic (6) decent. Only 35 cases worldwide have been documented since the disease was first described by Burlina et al. in 1991, demonstrating the rarity of this disorder (7). Common symptoms seen in EE patients include chronic diarrhea, a delay in neural development, symmetric brain lesions, relapsing petechiae, muscle hypotonia, and acrocyanosis of the extremities, which ultimately lead to death within the first decade of life (1, 2). The onset and degree of severity of these symptoms vary from patient to patient; however, the majority of these symptoms occur early in development (7). All patients with EE display some level of muscle weakness and spasms as well as abnormal cranial MRI scans with varying levels of brain atrophy depending on disease progression (3, 7). The majority of these patients showed severe delays in both cognitive and psychomotor functions (1-4, 7, 8). 1.1.2 Biochemical Markers of EE. In addition to the common symptoms associated with EE, patients also exhibit several biochemical traits including lactic acidemia, reduction of mitochondrial respiratory complex IV activity in the skeletal muscle, increase in carnitines, C4-6 acylglycines, and elevated levels of ethylmalonic acid (EMA) in their urine (3, 9, 10). Lactic acidemia is a very common marker in a number of metabolic disorders and can result from a cascade of events in cellular respiration that depletes the cellular central energy pathways. build up results from a blockage in glycolysis, which in turn, limits acetyl-CoA thereby inactivating the citric acid cycle (CAC). This inactivation of the CAC further leads to the depletion of NADH and reduces the activity of the electron transport system (11). Additionally, the excretion

1 of carnitines and acylglycines suggests a buildup of acyl-CoAs found typically when there is a block in fatty acid oxidation, further leading to a depletion of energy metabolism (11). Perhaps the most unique biochemical marker associated with EE is the presence of elevated concentrations of EMA. 1.1.3 Ethylmalonic Acid. EMA is a toxic metabolite that is capable of inhibiting mitochondrial creatine kinase activity (Mi-CK), which also plays a central role in energy metabolism (12). Creatine kinase catalyzes the conversion of creatine to phosphocreatine through the consumption of adenosine triphosphate (ATP), thereby generating adenosine diphosphate (ADP). In high energy-consuming tissues, such as skeletal muscle, brain, and smooth muscle, phosphocreatine serves as a reservoir for the rapid regeneration of ATP and is therefore necessary for energy metabolism in these tissues. EMA has been shown to have the greatest inhibitory effects on the brain-specific isoform of Mi-CK, which has been suggested to be a cause of the severe neurological features present in patients with EE and other diseases exhibiting high accumulations of EMA (12). Studies have shown that the incubation of EMA with the antioxidant , as well as with ascorbic acid, can block the inhibitory effect of EMA on brain tissue. This has led to the suggestion that EMA possibly functions through oxidation by free radicals of a critical or other groups present on the creatine kinase essential for its activity (12). 1.1.4 Role of EMA in Metabolic Pathways. The production of EMA is elevated when there are high levels of butyryl-CoA that can be carboxylated through propionyl- CoA carboxylase to ethylmalonyl-CoA allowing for the hydrolysis of ethylmalonyl-CoA to free EMA (13). The abnormal accumulation of butyryl-CoA can occur through at least two possible pathways: disorders of the short-chain fatty acid cycle resulting in a build- up of butyryl-CoA or the R-isoleucine catabolism pathway from 2-ethylmalonic- semialdehyde (Figure 1.1) (14, 15). Evidence for the production of EMA as a result of defects in β-oxidation is seen in disorders involving mutations in the short-chain acyl- CoA dehydrogenase (SCAD) (16), as well as glutaric academia type 2 (17), and Jamaican vomiting sickness (18), although none of these disorders can account for all the symptoms associated with EE, including diarrhea and the recurrent petechiae seen in EE patients (7). There is also evidence that symptoms of EE may be associated with a defect in the R-isoleucine pathway (Figure 1.1) (15). Patients with EE that were given excess

2

Figure 1.1: Metabolic Routes of EMA Production. The production of EMA is elevated by the accumulation of butyryl-CoA, where it can be carboxylated through propionyl-CoA carboxylase to ethylmalonyl-CoA. This can occur through disorders of short-chain fatty acid β-oxidation pathways as well as through R- isoleucine catabolism.

3 isoleucine produced higher levels of EMA compared with normal individuals (15). Because of this finding, it was suggested that EE could result from a blockage at or after 2-methyl-3-hydroxybutyrate (2M-BCAD) in L-isoleucine degradation, which would cause a shift toward the R-pathway of isoleucine catabolism (15).

1.1.5 Current Treatments for EE. All of the biochemical markers as well as the common symptoms associated with EE suggest that this disease is a result of a metabolic disorder; however, the specific biochemical pathway responsible for EE is still unknown (1-3, 7, 10). Some recent success has been seen in studies in which EE patients received treatment with riboflavin, coenzyme Q, and carnitine to promote the activity of fatty acid

β-oxidation (8). Some symptoms of EE were alleviated in these case studies, most notably the chronic diarrhea. However, this treatment was not able to eliminate all of the symptoms, and therefore, more studies are necessary to monitor its effectiveness (8).

1.2 Ethylmalonic Encephalopathy Protein 1 (ETHE1).

In 2004, homozygosity mapping revealed the molecular defect associated with EE (9). In this technique, mapped restriction length polymorphisms (RFLPs) in the DNA of affected EE children from inbred marriages were analyzed, and the disease detected using the assumption that the adjacent region of the locus was homozygous by descent in inbred children (9). Using this technique, it was shown that mutations in the gene Ethylmalonic Encephalopathy protein 1 (ETHE1) are directly associated with the disease EE (9, 10). Preliminary characterization of ETHE1 revealed that it contains orthologs in a wide range of organisms, including archaebacteria, fungi, plants, and animals, suggesting that ETHE1 serves a fundamental biochemical role. However, this role has yet to be fully characterized in any organism. A pairwise alignment of ETHE1-like proteins from a variety of organisms (Fig 1.2) identified a number of highly-conserved residues including the amino acids Y38, T136, C161, R163, and L185, of which mutations were found in EE patients (9, 10). Therefore, these residues are likely critical for ETHE1 function.

4

Figure 1.2: Pairwise alignment of ETHE1 Homologs. Identical residues are highlighted in black and residues associated with mutations observed with EE are labeled with an *.

5 1.3 Metallo-β-lactamase Fold Family of Proteins

A BLAST P (Basic Local Alignment Search Tool Protein) search revealed that ETHE1 is most similar to the metallo-β-lactamase fold family of proteins, which are defined by a common αβ/βα fold and a conserved metal binding motif, T-H-X-H-X-D (19). β-lactamase proteins are grouped into four separate classes; A, B, C, and D (20). Classes A, C, and D all have an essential Ser residue in the active site and were grouped together based on amino acid sequence similarity. Class B, also known as the metallo-β- lactamases, require 1 or 2 Zn (II) ions bound to their active sites for activity (20). The metallo-β-lactamases belong to a large superfamily of proteins that utilize a variety of substrates. The majority of the enzymes cleave an ester linkage (21). In addition to the class B β-lactamases, which hydrolyze lactams, this superfamily includes: glyoxalase II, aryl sulfatases, rubredoxin: (ROO), and ZiPD families (19, 21, 22). The β-lactamase fold typically binds Zn(II), however, the metal binding domain of the β-lactamase fold has also been shown to bind Fe and Mn (23, 24), demonstrating that the β-lactamase fold can accommodate several different , depending on the enzyme and the nature of the reaction .

1.4 The Glyoxalase System

1.4.1 The Glyoxalase System. Of the metallo-β-lactamase family of proteins, ETHE1 is most similar to the glyoxalase II family. A pairwise alignment between human ETHE1 and glyoxalase II showed that ETHE1 had low levels of identity (13 %) and similarity with glyoxalase II (Figure 1.3). The majority of the conserved residues are in the metal binding ligand domain. The glyoxalase system is ubiquitous in nature and has been studied in a number of organisms (25, 26). It consists of two enzymes: glyoxalase I (lactoylglutathione ) and glyoxalase II (hydroxacylglutathione ). The exact role of this system is unknown. However, it has been proposed that the detoxification of 2-oxoaldehydes, toxic by-products of carbohydrate and lipid metabolism, is its primary function, with the primary physiological substrate thought to be (2-oxopropanal), a reactive α-oxoaldehyde (25). Methylglyoxal is a cytotoxic and mutagenic by-product of respiration that can inactivate proteins, modify

6

Figure 1.3: Sequence Alignment of Human ETHE1 and Glyoxalase II. A pairwise alignment was generated using Clustal W v. 1.82 between ETHE1 (NP_055112) and GLX2 (CAA62483). Identical residues are highlighted in black while residues of the conserved dinuclear metal binding site are highlighted red. Residues highly conserved for GLX2 substrate binding are indicated by ▼.

7 guanylate residues, and create interstrand DNA crosslinks (27-29). Cellular detoxification of methylglyoxal begins with its non-enzymatic reaction with glutathione to produce a thiohemiacetal, which is then converted to S-D-lactoylglutathione (SLG) by Glyoxalase I (30). Glyoxalase II then hydrolyzes SLG to form D-lactic acid and free glutathione (31) (Figure 1.4).

Glyoxalase I, which is present as a single isozyme, has been well characterized in several systems (32-38). However, much less is known about the structure and function of the glyoxalase II enzymes, which exist as multiple isozymes in several organisms including , plants, and animals (32, 39). Glyoxalase II enzymes have been isolated and characterized from both the cytosol and mitochondria (33, 40, 41). In mammals, a single gene encodes both the cytosolic and mitochondrial GLX2 proteins. In contrast, multiple GLX2 genes exist in yeast and higher plants (40). 1.4.2. Arabidopsis Glyoxalase II Enzymes. So far, the best studied plant glyoxalase II enzymes are those from Arabidopsis thaliana. Five putative glyoxalase II isozymes have been identified in the Arabidopsis genome based on : GLX2-1 (At2g43430), GLX2-2 (At3g10850), GLX2-3 (At1g53580), GLX2-4 (At1g06130), and GLX2-5 (At2g31350) (41). Of the five isozymes, GLX2-2 is located in the cytosol and is most similar to human glyoxalase II, while GLX2-1, GLX2-4, and GLX2-5 are all predicted to be localized in the . Prior to the work presented here, the location of GLX2-3 was unknown. Interestingly, of the Arabidopsis GLX2 family, GLX2-3 shows the greatest similarity to human ETHE1. A pairwise alignment of GLX2-3 and ETHE1 shows that GLX2-3 is 54% identical to ETHE1 and shares all of the conserved residues of ETHE1 (Figure 1.5). A pairwise alignment of GLX2-3 with the other Arabidopsis GLX2 isozymes demonstrates that GLX2-3 is only 13% identical to the other glyoxalase II enzymes and the majority of the identity is accounted for in the predicted metal binding ligands (Figure 1.6). In addition, GLX2-3 contains only 2 out of the 5 residues predicted to be critical for binding of SLG (42, 43). The low level of similarity and the absence of the glyoxalase II substrate binding residues suggests that GLX2-3 is not a glyoxalase II enzyme after all, and is in fact, the Arabidopsis homolog of ETHE1. Because of this, we propose changing the name of Arabidopsis GLX2-3 to ETHE1.

8 COO- O O O Non-enzymatic N H SG N COO- NH3 + O O OH SH Glutathione (GSH) Methylglyoxal Hemithioacetal

Glyoxalase I

OH OH O Glyoxalase II SG GSH + O O

D-lactic Acid S-D-Lactoylglutathione

Figure 1.4: The Glyoxalase System. Methyglyoxal spontaneously reacts with glutathione to produce a thiohemiacetal adduct, which is then catabolized by glyoxalase I into S-D-lactoylglutathione. Glyoxalase II further hydrolyzes SLG into D-lactic acid and free glutathione.

9

Figure 1.5: Sequence Alignment of GLX2-3(ETHE1) and ETHE1. A pairwise alignment was generated between GLX2-3 (ETHE1) (NP_974018) and ETHE1 (NP_055112). Identical residues are highlighted black, while residues conserved in ETHE1-like proteins are indicated by ♦. Residues implicated in the EE disease when mutated are indicated by ∞.

10 1.5 Hepatoma Subtracted-cDNA Library Clone One

Prior to the demonstration that mutations in ETHE1 are responsible for the disease EE, ETHE1 was formally known as HSCO (hepatoma subtracted-cDNA library clone one) because its transcripts are elevated in hepatoma cells (44). Subsequently it was suggested that ETHE1 functions as a nuclear-cytoplasmic shuttle protein that regulates the nuclear localization of NF-κB, a factor associated with the regulation of apoptotic responses (44, 45). Most recently, human ETHE1 was shown to associate with 1 (HDAc1) and facilitate the HDAc1-dependent deacetylation, and therefore degradation of the . It is well established that p53 is an important cellular regulator that when activated is known to mediate cell-cycle arrest, apoptosis, senescence, differentiation, DNA repair, and the inhibition of angiogenesis and metastasis (46-48). Furthermore, cell lines over-expressing ETHE1 have shown resistance to p53-controlled apoptosis induced by DNA damage (47). Given the wide phylogenetic distribution of ETHE1, this role in inhibiting p53-dependent apoptosis is not expected to be a common feature of the protein.

1.6 Arabidopsis thaliana

1.6.1 Arabidopsis thaliana as a Model System. Because of its relatively small genome and ease of study, Arabidopsis thaliana, a small plant of the mustard family, has become a model system for basic research (49, 50). Laboratory experiments can be performed quickly and in small spaces considering that the entire life cycle of the plant from seed germination to the maturation of the first seed is less than 6 weeks and that the maximum height of growth is only 20 cm (49). Additionally, one plant can produce up to 5000 progeny making it useful for large scale genetic analyses (49). Another main reason why Arabidopsis has become a model system is that a broad knowledge base has been established, including a fully-sequenced genome (49, 50). One large resource of data that sets Arabidopsis apart from other model organisms is the publicly available populations of Arabidopsis lines containing transformed DNA (T-DNA) mutations. These T-DNA lines were established through insertional mutagenesis and allow researchers to analyze particular genes of interest through reverse genetics (49, 50). Because many fundamental life processes at the molecular and cellular levels are

11

Figure 1.6: Sequence Alignment of the Arabidopsis Glyoxalase II isozymes. A pairwise alignment between ETHE1 (GLX2-3) and the other Arabidopsis Glyoxalase II isozymes. Identical residues are highlighted in black while residues highlighted in red indicated the highly conserved metallo-β-lactamase fold family metal binding residues. Residues highly conserved for GLX2 substrate binding are indicated by ▼.

12 common to all higher organisms, information gained from studying Arabidopsis genes should provide insight into the role the gene plays in higher (50). 1.6.2 Loss of Function Mutants of Arabidopsis thaliana. One way to study the function of a protein is through reverse genetics by manipulating the gene of interest and then analyzing its effects. This cause and effect analysis is typically performed through knockout mutations, which, in mammals, is done through homologous recombination where a modified gene is substituted for a wild-type copy in vivo. In plant systems, knockout analyses are generally performed utilizing large scale Agrobacterium-mediated transformation in which a T-DNA is randomly inserted into the gene disrupting its function (51). Using this system, large populations of knockout mutations can be easily screened using the polymerase chain reaction (PCR) to identify a mutation in the gene of interest. The homozygous mutant can then be analyzed for phenotypic effects of the mutation (52).

1.6.3 Embryo Lethal Mutants of Arabidopsis thaliana. To date, there are about 350 non-redundant, essential genes for Arabidopsis thaliana development in the database (www.seedgene.org). Many of these genes encode proteins for basal cellular functions, transcription factors, or signaling components; however, a wide range of functions have been reported (53). The various embryo defective mutants differ in many aspects of their development including among other things, stage of abortion, extent of abnormal development, size and color of aborted seeds and embryos, efficiency of transmission through male and female gametes, and the ability to produce mutant seedlings (54). Analysis of these mutants prior to their arrest can provide insight into the function of the protein. Preliminary analysis of Arabidopsis plants containing a knockout mutation in ETHE1 showed that it resulted in the inability to produce viable homozygous plants, suggesting that the mutation was lethal (54). Given the debilitating and fatal consequences of mutations in the ETHE1 gene in humans, the essential nature of ETHE1 is not unexpected. However, it appears that plants may be more sensitive to the absenceof ETHE1 activity. Further investigation of seed development is necessary to better understand ETHE1’s role in plant development.

1.6.4 Seed Development in Arabidopsis thaliana. (Figure 1.7) The seeds of Arabidopsis develop in the fruits of the plant known as siliques. Each individual silique

13

Figure 1.7: Life Cycle of Arabidopsis. Seed development consists of two stages of development, gametogenesis and embryogenesis. Gametogenesis is initiated through the asexual generation of two spores, the microspore (male) and the megaspore (female) in the developing bud. The microspore undergoes two mitotic divisions to produce the mature pollen consisting of a vegetative nucleus (vn) and two sperm nuclei (sn). The megaspore undergoes three rounds of nuclear division followed by cellularization resulting in a seven cell female gametophyte containing three antipodal cells (ap), a diploid central cell (cc), two synergid cells (syn), and an egg cell (ec). A double fertilization event of the two sperm cells with the egg and central cells allow for the formation of the embryo (e) and the endosperm, tissue, respectively. Nuclear division in the endosperm results in the formation of three regions: the micropylar endosperm (MCE), the peripheral endosperm (PEN), and the chalazal region (CZE). As the embryo develops it expands to fit the entirety of the seed consuming the non-persistent

14 endosperm. Once fully matured, the seed undergoes dessication allowing for germination and plant growth.

15 contains about 40-60 seeds separated into two long rows that elongate as the seeds develop (55). The full cycle of reproductive development from fertilization to seed desiccation is completed in approximately two weeks (56). Prior to fertilization of the seed, the ovule undergoes a series of developmental events involving the development of the female gametophyte, also called the embryo sac. The male gametophyte, which develops in the anther, at full maturity contains a large vegetative cell and two sperm cells that are involved in a double fertilization event in the embryo sac (56). Conversly, the female gametophyte consists of seven cells: three antipodals, one large diploid central cell, two synergids, and an egg cell (57, 58). In the first stage of female gametogenesis, FG1, three of the four megaspores forming a tetrad undergo cell death allowing for the enlargement of the remaining megaspore. FG2 is defined through the mitotic event of the surviving megaspore resulting in a two cell female gametophyte. Stages FG3-FG4 are characterized by the migration of the two nuclei to the micropylar and chalazal poles of the embryo sac followed by a second round of mitosis forming a four nucleate gametophyte. These nuclei undergo a third round of mitosis producing an eight nuclei gametophyte (FG5). At this stage, one nucleus from the chalazal pole migrates towards the micropylar region forming polar nuclei and subsequently, cellularization occurs. This event is followed by the fusion of the two polar nuclei generating a diploid central cell (FG6) and the degeneration of the three antipodal cells (FG7). The degeneration of one of the synergid cells from the entry of the pollen tube marks the onset of FG8. A double fertilization event occurs after the pollen tube delivers two sperm nuclei to the embryo sac. One sperm nucleus fuses with the egg to form the zygote, and the other sperm nucleus fuses with the central cell to form the endosperm marking the beginning of embryogenesis (57, 58). In embryogenesis (Figure 1.7), the fertilized central cell divides into two nuclei, which migrate to opposite poles of the developing seed, the micropylar and the chalazal regions. This initial division is followed by several synchronized syncytial nuclear divisions until cellularization occurs in a wave-like pattern beginning once the embryo has reached late globular stage. Cellularization begins in the micropylar region and progresses through to the chalazal endosperm by the time heart stage is reached in the embryo (57).

16 Once the onset of nuclear division occurs in the endosperm, the zygote begins to divide forming the embryo proper and suspensor. Cellular division continues throughout the embryo proper forming the two-cell, quadrant, octant, dermatogen, early globular, globular, heart, torpedo, bent cotyledon, and lastly, the mature embryo (57). The growing seed remains relatively small during the first few divisions, however once heart stage is reached, the seed undergoes organ expansion to accommodate the growing embryo. As the embryo continues to develop, the non-persistent cellularized endosperm tissue is consumed (57). Once the embryo reaches maturation, the seed undergoes desiccation, characterized by the brown pigmentation of the seed resulting from the loss of water and the subsequent hardening of the seed (57).

1.7 Sections of the Dissertation

To date, a detailed characterization of ETHE1 has not been performed. Therefore, neither the structure nor the function has been established for the enzyme in any organism. In this dissertation, a full structural characterization of the Arabidopsis homolog of ETHE1 as well as information on its functional role in plants will be presented. We have obtained the first crystal structure of an ETHE1-like protein as well as performed metal and preliminary substrate analyses providing new insights into the possible role and substrate of ETHE1. In addition, I demonstrate that ETHE1 is essential for both plant growth and development. Chapter 2 describes the crystallization conditions and the crystal structure of Arabidopsis ETHE1. This work has been published in Acta Crystal (59). Chapter 3 describes our efforts to characterize the metal binding center and active site of ETHE1, both structurally and functionally utilizing NMR, EPR, gel filtration, and substrate analyses. This chapter is to be submitted to Journal of Bioinorganic Chemistry. Chapter 4 describes our functional studies of ETHE1 in seed and endosperm development. This work has been submitted for publication in the Plant Journal. Chapter 5 describes the characterization of inducible ETHE1 knock-down lines to determine the effects of reduced ETHE1 levels in cell cultures and at various stages of plant growth. Chapter 6 describes the growth enhancing effects seen with the constitutive over-expression of ETHE1 in Arabidopsis thaliana. This work has been submitted for a US patent (60) and will be submitted in the journal of Plant Molecular

17 Biology. Lastly, Chapter 7 summarizes all of the research described in this dissertation and potential roles of ETHE1 in plants and humans. Finally, this chapter also describes potential future directions for this project.

18 1.8 References

1. Burlina, A., Zacchello, F., Dionisi-Vici, C., Bertini, E., Sabetta, G., Bennet, M. J., Hale, D. E., Schmidt-Sommerfeld, E., and Rinaldo, P. (1991). New Clinical of Branched-chain acyl-CoA Oxidation Defect. Lancet 338, 1522-1523

2. Burlina, A. B., Dionisi-Vici, C., Bennett, M. J., Gibson, K. M., Servidei, S., Bertini, E., Hale, D. E., Schmidt-Sommerfeld, E., Sabetta, G., and Zacchello, F. (1994). A New Syndrome with Ethylmalonic Aciduria and Normal Fatty Aci Oxidation in Fibroblasts. J. Pediatr. 124, 79-86

3. Zafeiriou, D. I., Augoustides-Savvopoulou, P., Haas, D. et al. (2007). Ethylmalonic Encephalopathy: Clinical and Biochemical Observations. Neuropediatrics 38, 78-82

4. Grosso, S., Mostardini, R., Farnetani, M. A., Molinelli, M., Berardi, R., Dionisi-Vici, C., Rizzo, C., Morgese, G., and Balestri, P. (2002). Ethylmalonic Encephalopathy: Further Clinical and Neuroradiological Characterization. J. Neurol. 249, 1446-1450

5. Garavaglia, B., Colamaria, V., Carrara, F., Tonin, P., Rimoldi, M., and Uziel, G. (1994). Muscle Deficiency in two Italian Patients with Ethylmalonic Acidurai and Peculiar Clinical Phenotype. J. Inherit. Metab. Dis. 17, 301-303

6. Ozand, P. T., Rashed, M., Millington, D. S., Sakati, N., Hazzaa, S., Rahbeeni, Z., al Odaib, A., Youssef, N., Mazrou, A., and Gascon, G. G. (1994). Ethylmalonic Aciduria: an Organic Acidemia with CNS Involvement and Vasculopathy. Brain Dev. 16 Suppl, 12-22

7. Plecko, B. (2007). Lessons to Learn from Rare Invorn Errors of Metabolism. Neuropediatrics 38, 59-60

8. Yoon, H. R., Hahn, S. H., Ahn, Y. M., Jang, S. H., Shin, Y. J., Lee, E. H., Ryu, K. H., Eun, B. L., Rinaldo, P., and Yamaguchi, S. (2001). Therapeutic Trial in the First

19 Three Asian Cases of Ethylmalonic Encephalopathy: Response to Riboflavin. J. Inherit. Metab. Dis. 24, 870-873

9. Tiranti, V., D'Adamo, P., Briem, E. et al. (2004). Ethylmalonic Encephalopathy is Caused by Mutations in ETHE1, a Gene Encoding a Protein. Am. J. Hum. Genet. 74, 239-252

10. Tiranti, V., Briem, E., Lamantea, E. et al. (2006). ETHE1 Mutations are Specific to Ethylmalonic Encephalopathy. J. Med. Genet. 43, 340-346

11. Bralley, J. A. (2001) Organic acids. In Laboratory evaluations in molecular medicine pp. 173-222. Institute for Advances in Molecular Medicine.

12. Leipnitz, G., Schuck, P. F., Ribeiro, C. A., Dalcin, K. B., Assis, D. R., Barschak, A. G., Pulrolnik, V., Wannmacher, C. M., Wyse, A. T., and Wajner, M. (2003). Ethylmalonic Acid Inhibits Mitochondrial Creatine Kinase Activity from Cerebral Cortex of Young Rats in Vitro. Neurochem. Res. 28, 771-777

13. Hegre, C. S., Halenz, D. R., and Lane, M. D. (1959). The Enzymatic Carboxylation of Butyryl Coenzyme A. J. Am. Chem. Soc. 81, 6526-6527

14. Mamer, O. A., Tjoa, S. S., Scriver, C. R., and Klassen, G. A. (1976). Demonstration of a New Mammalian Isoleucine Catabolic Pathway Yielding an Rseries of Metabolites. Biochem. J. 160, 417-426

15. Nowaczyk, M. J., Lehotay, D. C., Platt, B. A., Fisher, L., Tan, R., Phillips, H., and Clarke, J. T. (1998). Ethylmalonic and Methylsuccinic Aciduria in Ethylmalonic Encephalopathy Arise from Abnormal Isoleucine Metabolism. Metabolism 47, 836- 839

16. Corydon, M. J., Gregersen, N., Lehnert, W. et al. (1996). Ethylmalonic Aciduria is Associated with an Amino Acid Variant of Short Chain Acyl-coenzyme A Dehydrogenase. Pediatr. Res. 39, 1059-1066

20 17. Mantagos, S., Genel, M., and Tanaka, K. (1979). Ethylmalonic-adipic Aciduria. In Vivo and In Vitro Studies Indicating Deficiency of Activities of Multiple Acyl-CoA Dehydrogenases. J. Clin. Invest. 64, 1580-1589

18. Tanaka, K., Ramsdell, H. S., Baretz, B. H., Keefe, M. B., Kean, E. A., and Johnson, B. (1976). Identification of Ethylmalonic Acid in Urine of Two Patients with the Vomitting Sickness of Jamaica. Clin. Chim. Acta 69, 105-112

19. Melino, S., Capo, C., Dragani, B., Aceto, A., and Petruzzelli, R. (1998). A - binding Motif Conserved in Glyoxalase II, Beta-lactamase and Arylsulfatases. Trends Biochem. Sci. 23, 381-382

20. Fisher, J. F., Meroueh, S. O., and Mobashery, S. (2005). Bacterial Resistance to beta- lactam Antibiotics: Compelling Opportunism, Compelling Opportunity. Chem. Rev. 105, 395-424

21. Aravind, L. (1999). An Evolutionary Classification of the Metallo-beta-lactamase Fold Proteins. In Silico Biol. 1, 69-91

22. Daiyasu, H., Osaka, K., Ishino, Y., and Toh, H. (2001). Expansion of the Zinc Metallo-hydrolase Family of the Beta-lactamase Fold. FEBS Lett. 503, 1-6

23. Schilling, O., Wenzel, N., Naylor, M., Vogel, A., Crowder, M., Makaroff, C., and Meyer-Klaucke, W. (2003). Flexible Metal Binding of the Metallo-beta-lactamase Domain: Glyoxalase II Incorporates Iron, , and Zinc in Vivo. Biochemistry 42, 11777-11786

24. Frazao, C., Silva, G., Gomes, C. M. et al. (2000). Structure of a Dioxygen Reduction Enzyme from Desulfovibrio gigas. Nat. Struct. Biol. 7, 1041-1045

25. Thornalley, P. J. (1993) The Glyoxalase System in Health and Disease. Mol. Aspects Med. 14, 287-371

21 26. Thornalley, P. J. (1990). The Glyoxalase System: New Developments Towards Functional Characterization of a Metabolic Pathway Fundamental to Biological Life. Biochem. J. 269, 1-11

27. Lo, T. W., Westwood, M. E., McLellan, A. C., Selwood, T., and Thornalley, P. J. (1994). Binding and Modification of Proteins by Methylglyoxal Under Physiological Conditions. A Kinetic and Mechanistic Study with N Alpha- acetylarginine, N Alpha-acetylcysteine, and N Alpha-acetyllysine, and Bovine Serum Albumin. J. Biol. Chem. 269, 32299-32305

28. Papoulis, A., al-Abed, Y., and Bucala, R. (1995). Identification of N2-(1- carboxyethyl) Guanine (CEG) as a Guanine Advanced Glycosylation End Product Biochemistry 34, 648-655

29. Rahman, A., Shahabuddin, and Hadi, S. M. (1990). Formation of Strand Breaks and Interstrand Cross-links in DNA by Methylglyoxal. J. Biochem. Toxicol. 5, 161-166

30. Mannervik, B., and Ridderstrom, M. (1993). Catalytic and Molecular Properties of Glyoxalase I. Biochem. Soc. Trans. 21, 515-517

31. Allen, R. E., Lo, T. W., and Thornalley, P. J. (1993). Purification and Characterization of Glyoxalase II from Human Red Blood Cells. Eur. J. Biochem. 213, 1261-1267

32. Creighton, D. J., and Hamilton, D. S. (2001). Brief History of Glyoxalase I and What We Have Learned about Metal Ion-dependent, Enzyme-catalyzed Isomerizations. Arch. Biochem. Biophys. 387, 1-10

33. Bito, A., Haider, M., Hadler, I., and Breitenbach, M. (1997). Identification and Phenotypic Analysis of Two Glyoxalase II Encoding Genes from , GLO2 and GLO4, and Intracellular Localization of the Corresponding Proteins. J. Biol. Chem. 272, 21509-21519

34. Clugston, S. L., Daub, E., Kinach, R., Miedema, D., Barnard, J. F., and Honek, J. F. (1997). Isolation and Sequencing of a Gene Coding for Glyoxalase I Activity from

22 Salmonella typhimurium and Comparison with other Glyoxalase I Sequences. Gene 186, 103-111

35. Inoue, Y., and Kimura, A. (1996). Identification of the Structural Gene for Glyoxalase I from Saccharomyces Cerevisiae. J. Biol. Chem. 271, 25958-25965

36. Lu, T., Creighton, D. J., Antoine, M., Fenselau, C., and Lovett, P. S. (1994). The Gene Encoding Glyoxalase I from Pseudomonas putida: Cloning, Overexpression, and Sequence Comparisons with Human Glyoxalase I. Gene 150, 93-96

37. Rhee, H., and Kimura, A. (1988). Glyoxalase I and its Application to Production of S-lactoylglutathione. Tanpakushitsu Kakusan Koso 33, 1610-1614

38. Ridderstrom, M., and Mannervik, B. (1996). The Primary Structure of Monomeric Yeast Glyoxalase I Indicates a Resulting in Two Similar Segments Homogous with the Subunit of Dimeric Human Glyoxalase I. Biochem. J. 316 ( Pt 3), 1005-1006

39. Richter, U., and Krauss, M. (2001). Active Site Structure and Mechanism of Human Glyoxalase I-an ab initio Theoretical Study. J. Am. Chem. Soc. 123, 6973-6982

40. Cordell, P. A., Futers, T. S., Grant, P. J., and Pease, R. J. (2004). The Human Hydroxyacylglutathione Hydrolase (HAGH) Gene Encodes both Cytosolic and Mitochondrial Forms of Glyoxalase II. J. Biol. Chem. 279, 28653-28661

41. Maiti, M. K., Krishnasamy, S., Owen, H. A., and Makaroff, C. A. (1997). Molecular Characterization of Glyoxalase II from Arabidopsis thaliana. Plant Mol. Biol. 35, 471-481

42. Cameron, A. D., Ridderstrom, M., Olin, B., and Mannervik, B. (1999). Crystal Structure of Human Glyoxalase II and its Complex with a Glutathione Thiolester Substrate Analogue. Structure 7, 1067-1078

23 43. Zang, T. M., Hollman, D. A., Crawford, P. A., Crowder, M. W., and Makaroff, C. A. (2001). Arabidopsis Glyoxalase II Contains a Zinc/iron Binuclear Metal Center that is Essential for Substrate Binding and Catalyisis. J. Biol. Chem. 276, 4788-4795

44. Higashitsuji, H., Higashitsuji, H., Nagao, T., Nonoguchi, K., Fujii, S., Itoh, K., and Fujita, J. (2002). A Novel Protein Overexpressed in Hepatoma Accelerates Export of NR-kappa B from the Nucleus and Inhibits p53-dependent Apoptosis. Cell 2, 335-346

45. Karin, M., and Lin, A. (2002). NF-κB at the Crossroads of Life and Death. Nat. Immunol. 3, 221-227

46. Higashitsuji, H., Higashitsuji, H., Masuda, T., Liu, Y., Itoh, K., and Fujita, J. (2007). Enhanced Deacerylation of p53 by the Anti-apoptotic Protein HSCO in Association with Histone Deacetylase 1. J. Biol. Chem. 282, 13716-13725

47. Ito, A., Kawaguchi, Y., Lai, C.H., Kovacs, J.J., Higashimoto, Y., Appella, E., and Yao, T.P. (2002). Processing of Mitochondrial Protein Precursors. EMBO J. 21, 6236-6245

48. Liu, G., and Chen, X. (2006). Regulation of the p53 Transcriptional Activity. J. Cell. Biochem. 15, 448-458

49. Meinke, D. W., Cherry, J. M., Dean, C., Rounsley, S. D., and Koornneef, M. (1998). Arabidopsis thaliana: a Model Plant for Genome Analysis. Science 282, 662, 679-82

50. Leonelli, S. (2007). Arabidopsis, the Botancial Drosophila: From Mouse Cress to . Endeavour 31, 34-38

51. Koncz, C., and Schell, J. (1992) T-DNA Transformation and Insertion Mutagenesis. In Methods in Arabidopsis research pp. 224-273. Edited by C. Koncz, N. Chua & J. Schell, Singapore, World Scientific Publishing Co.

24 52. Krysan, P. J., Young, J. C., Tax, F., and Sussman, M. R. (1996). Identification of Transferred DNA Insertions within Arabidopsis Genes Involved in Signal Transduction and Ion Transport. Proc. Natl. Acad. Sci. U. S. A. 93, 8145-8150

53. Tzafrir, I., Pena-Muralla, R., Dickerman, A. et al. (2004). Identification of Genes Required for Embryo Development in Arabidopsis. Plant Physiol. 135, 1206-1220

54. Rhee, J. S. (2001). Characterization of Glyoxalase II Isozymes in Arabidopsis thaliana. Dissertation Thesis

55. Meinke, D. W. (1994) Seed development in Arabidopsis thaliana. In Arabidopsis thaliana pp. 253-295. Edited by E. M. Meyerowitz, & C. R. Somerville, New York, Cold Spring Harbor Laboratory Press.

56. McCormick, S. (2004). Control of Male Gametophyte Development. Plant Cell 16 Suppl, S142-53

57. Jurgens, G., and Mayer, U. (1994) Arabidopsis. In Color Atlas of Development pp. 7- 21. Edited by J. Bard, London, Elsevier Science.

58. Berger, F., Grini, P. E., and Schnittger, A. (2006). Endosperm: an Integrator of Seed Growth and Development. Curr. Opin. Plant Biol. 9, 664-670

59. McCoy, J. G., Bingman, C. A., Bitto, E., Holdorf, M. M., Makaroff, C. A., and Phillips, G. N.,Jr. (2006). Structure of an ETHE1-like Protein from Arabidopsis thaliana. Acta Crystallogr. D Biol. Crystallogr. 62, 964-970

60. Holdorf, M., and Makaroff, C. (2007). Transgenic Plants with Enhanced Characteristics. US Patent application #60 909 649

25

Chapter 2

Structure of an ETHE1-like Protein from Arabidopsis thaliana

Jason G. McCoy,a Craig A. Bingman,a Eduard Bitto,a Meghan M. Holdorf,b Christopher A. Makaroffb and George N. Phillips Jr.a*

aDepartment of Biochemistry and Center for Eukaryotic Structural Genomics at the University of Wisconsin Madison, USA, and bDepartment of Chemistry and Biochemistry at Miami University, USA.

* Corresponding author:

George N. Phillips Jr. Department of Biochemistry and Center for Eukayotic Structural Genomics University of Wisconsin Madison

E-mail: [email protected]

This paper has been published in Acta Cryst.(2006). D62, 964-970

Authors contribution to work: In this paper, the crystal structure was solved by J.G. McCoy, C.A. Bingman, E.Bitto, and G. N. Phillips, Jr. All other work in this paper including the cloning, over-expression, purification, and analysis of the ETHE1 protein was contributed by M.M.Holdorf.

26

List of abbreviations used:

EE-Ethylmalonic Encephalopathy; ETHE1-Ethylmalonic Encephalopathy Protein 1; FPLC-Fast Protein Liquid Chromatography; GLX1-Glyoxalase I; GLX2-Glyoxalase II; GLX2-5-Arabidopsis glyoxalase 2-5; ICP-AES-Inductively Coupled Plasma Spectrometer with Atomic Emission Spectroscopy Detection; MG-methylglyoxal.

27 2.1 Summary

The protein product of gene At1g53580 from Arabidopsis thaliana possesses 54 % sequence identity to a human enzyme that has been implicated in the rare disorder ethylmalonic encephalopathy. The structure of the At1g53580 protein has been solved to a nominal resolution of 1.48 Å. This structure reveals tertiary structure differences between the ETHE1-like enzyme and glyoxalase II enzymes that are likely to account for differences in reaction chemistry and multimeric state between the two types of enzymes. In addition, the Arabidopsis ETHE1 protein is used as a model to explain the significance of several mutations in the human enzyme that have been observed in patients with ethylmalonic encephalopathy.

28 2.2 Introduction

The Arabidopsis thaliana gene, At1g53580, encodes a 255 residue protein whose sequence places it in the metallo-β-lactamase superfamily (SUPFAM E-value = 3x10-13) {{255 Gough,J. 2001; }}. This protein was originally identified as one of five glyoxalase II isozymes in Arabidopsis {{9 Maiti,M.K. 1997; }}. Structures of two glyoxalase II enzymes are currently known. One corresponds to a cytoplasmic isozyme from Homo sapiens {{38 Cameron,A.D. 1999; }}, and the other is a mitochondrial isozyme from Arabidopsis thaliana (Marasinghe et al., 2005){{65 Marasinghe,G.P. 2005; }}, which has been designated AtGLX2-5 (At2g31350). Glyoxalase II, also known as hydroxyacylglutathione hydrolase, along with glyoxalase I make up the glyoxalase system that acts to convert a variety of α-keto into hydroxyacids in the presence of glutathione {{106 Thompson, Guy A. Jr 1993; }}. Aromatic and aliphatic α-keto aldehydes react spontaneously with glutathione to form thiohemiacetals, which are converted to S-(2-hydroxyacyl) glutathione derivatives by GLX1. GLX2 hydrolyzes these derivatives to regenerate glutathione and produce hydroxyacids. Glyoxalase I utilizes a number of α-ketoaldehydes {{256 Davidson, G. 2000; }}. However the primary physiological substrate of the enzyme is thought to be methylglyoxal (MG), a cytotoxic and mutagenic compound formed primarily as a by-product of carbohydrate and lipid metabolism {{258 Thornalley,P.J. 1998;257 Thornalley,P.J. 1995; }}. Therefore, the glyoxalase system is thought to play an important role in chemical detoxification. Glyoxalase II enzymes, like other members of the metallo-β-lactamase superfamily, have been shown to contain dinuclear metal centers {{26 Crowder,M.W. 1997; }}. Interestingly, different glyoxalase II enzymes have differing specificities for iron, zinc, and manganese {{143 Wenzel,N.F. 2004;65 Marasinghe,G.P. 2005; 26 Crowder,M.W. 1997; 38 Cameron,A.D. 1999; }}. Recently it has been shown that the predicted At1g53580 enzyme shows greater sequence identity (54%) to a human enzyme from a gene named ETHE1 than to glyoxalase II enzymes. Therefore, the Arabidopsis gene locus At1g53580 has been named AtETHE1. We will subsequently refer to the protein product of this gene as

29 AtETHE1 as well. While the ETHE1 protein (referred to hereafter as ETHE1) shows significant sequence similarity to glyoxalase II, it does not possess glyoxalase II activity {{10 Tiranti,V. 2004; }}. No function has been determined for the enzyme; however, it has been implicated in a rare autosomal recessive disorder known as ethylmalonic encephalopathy, and a number of mutations in the ETHE1 protein of affected individuals have been identified {{10 Tiranti,V. 2004; }}. The first description of this disorder included the following symptoms: mental retardation, orthostatic acrocyanosis, relapsing petechiae, progressive pyramidal signs, chronic diarrhea, and symmetric brain lesions {{12 Burlina,A. 1991;13 Burlina,A.B. 1994; }}. There were also high levels of ethylmalonic acid, C4-C5 acylglycines, and acylcarnitines detected in patients' urine as well as elevated levels of lactic acid and short and branched chain acylcarnitine levels in the blood {{13 Burlina,A.B. 1994;12 Burlina,A. 1991; }}. In this paper, we describe the structure of AtETHE1 and demonstrate the structural differences between AtETHE1 and glyoxalase II enzymes. We further illustrate the structural significance of several mutations within the ETHE1 enzyme found in sufferers of ethylmalonic encephalopathy.

2.3 Material and Methods

2.3.1 Cloning, Expression, and Purification

The AtETHE1 gene was cloned into pET24b as a Nde1 and XhoI fragment following PCR amplification using the primers TCTTCTCATATGAAGCTTCTCTTTCGTCAAC

and 5’-GAGTCGACTCGAGCTCTAGATC(T)16. For high-level expression in E. coli, the N-terminal eleven amino acids were removed, and the amino terminal methionine was placed amino acid 12 of the predicted protein sequence. After verification by DNA sequencing, pET24b-At ETHE1 was transformed into E.coli BL21-Codon Plus (DE3)- RIL cells and used for protein over-expression in ZY media containing 50μg/ml

kanamycin and 50μM Fe(NH4)2(SO4)2 as previously described {{3 Zang,T.M. 2001; }}. AtETHE1 was purified from cleared lysates by FPLC using a Q-Sepharose column as described previously {{26 Crowder,M.W. 1997; }}. Protein purity was determined by

30 SDS-polyacrylamide gel electrophoresis, and protein concentrations were determined using the extinction coefficient of 10,240 M-1cm-1, which is based on amino acid composition of AtETHE1 {{27 Gill,S.C. 1989; }}. Metal analyses were performed on the purified enzyme using a Varian-Liberty 150 inductively coupled plasma spectrometer with atomic emission spectroscopy detection (ICP-AES) as described {{26 Crowder,M.W. 1997; }}.

2.3.2 Crystallization:

AtETHE1 crystals were grown by the hanging-drop vapor-diffusion method at 25 °C. The reservoir solution contained 24% (w/v) polyethylene glycol methyl ether 5K, 0.05 M sulfate, and 0.10 M N-(2-hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid) (HEPES), pH 8.5. The crystals were grown at room temperature, and diffraction-quality crystals appeared after several months. Additional crystals used for phasing were grown within a week following microseeding and then soaked in a solution of mother liquor containing 2 mM thimerosal (C9H9HgNaO2S) for two days. Crystals were then cryoprotected by soaking in solutions of mother liquor with increasing amounts of ethylene glycol up to 20% (v/v).

2.3.3 Data Collection:

Diffraction data from the native AtETHE1 crystal were collected at liquid temperatures on beamline 22-ID at Argonne National Laboratories at a wavelength of 1.23984 Å to a maximum resolution of 1.48 Å. Diffraction data from the mercury derivatized crystal were collected on beamline 23-ID-B at Argonne National Laboratories at a wavelength of 0.98244 Å to a maximum resolution of 2.04 Å. The diffraction images were integrated and scaled using HKL2000 {{259 Otwinowski, Z. 1997; }}. The unit cell belonged to space group and had parameters a = 66.6 Å, b = 64.5 Å, c = 127.9 Å, α = γ = 90.0 °, and β = 97.8 °.

2.3.4 Structure Determination and Refinement:

31 The mercury substructure of the derivatized crystal was determined using HySS from PHENIX {{262 Weeks, C.M. 2003;261 Adams, P.D. 2002; }}, and the mercury positions were input into autoSHARP to calculate phases using single isomorphous replacement with anomalous scattering phasing techniques {{263 Bricogne, G. 2003; }}. Auxiliary programs used by autoSHARP were from the CCP4 suite {{264 Collaborative Computational Project, N. 1994; }}. Density modification was carried out with Solomon {{265 Abrahams, J.P. 1996; }}. ARP/wARP was used to build the initial model {{266 Blanc, E. 2004; }}. The model was completed with alternate rounds of model building with COOT {{267 Emsley, P. 2004; }} and restrained refinement via REFMAC {{268 Murshudov, G.N. 1997; }}. The final model contained four protein molecules, four iron (II) ions, 1037 water molecules, 14 ethylene glycol molecules, and a sulfate molecule. Eight of the C-terminal residues for two of the protein chains and the final C-terminal residue for the other two chains were left unmodeled. The discrepancy in the observable length of the four chains was due to additional, non-biological contacts that the C-terminal regions of chains A and C were able to make with chains D and B. The Ramachandran plot showed that 92% of the residues were in the most favorable region. The remainder were in the generously allowed region of the plot. The data collection and refinement statistics are summarized in Table 1.

2.4 Results and discussion

2.4.1 Overall Fold:

The overall fold of AtETHE1 is typical of the β-lactamase superfamily. It contains two central mixed β-sheets, each containing six strands, surrounded on both sides by helices (see Fig. 2.1). The β-sheet topology is of the order A, B, C, D, E, F and G, H, I, J, K, L. β-strands A, B, and C are aligned antiparallel, whereas C, D, E, and F are parallel. In the second β-sheet, strands G, H, I, and J are aligned antiparallel, J and K are aligned parallel, and K and L are aligned antiparallel.

32

Table 2.1

Summary of crystallographic data-collection and refinement statistics Values in parentheses refer to the highest resolution shell.

33 A VAST search indicated that the most structurally similar known enzymes are the human and Arabidopsis (AtGLX2-5) glyoxalase II enzymes with scores of 27.7 and 27.2 respectively. The overall folds of AtETHE1 and AtGLX2-5 are highly similar, but differ in three regions, as shown in Fig. 2.1. The first two regions are outside the active site but make contacts with one another. This structure includes a two helix bundle that extends from residues 172 to 206 and an extended loop consisting of residues 223 to 240 in AtGLX2-5. Both of these features are missing in AtETHE1. Additionally, the extended C-terminus of AtETHE1 reaches across the opening of the active site, greatly limiting the possible size of potential substrates. These gaps are further illustrated in Fig. 2.2.

2.4.2 Dimer:

While previously described glyoxalase II enzymes are monomers, the crystal structure of AtETHE1 reveals a dimeric organization for the protein (see Fig. 2.3). The human ETHE1 protein was shown to be a dimer via gel filtration chromatography. Interestingly, the dimerization interface for AtETHE1 appears to be in a region that was blocked by the two-helix bundle of the AtGLX2-5 enzyme. This may represent a distinguishing feature between ETHE1-like and glyoxalase II enzymes. The interface between the AtETHE1 dimers is not extensive, with only 830 Å2 of buried surface area (25). The interface contains 58% nonpolar area and involves 10 residue to residue bonds. The interactions are identical between the two subunits with Arg-17 forming hydrogen bonds with Glu-206 from the other subunit, and likewise Gln-18 with Glu-200, Phe-20 with Gly-198, Arg-53 with Lys-197, Glu-60 with Arg-158, and vice versa.

2.4.3 Metal Binding Site:

Only one metal was located within the electron density for AtETHE1. ICP-AES results indicated that the purified enzyme contained two molar equivalents of iron; however, ICP-MS metal analysis of the protein after being subjected to freezing and

34

Figure 2.1 The At1g53580 monomer. Helices are labeled 1 - 8 and β-strands are labeled A - L. Below. Overlay of the AtETHE1 (magenta) and the GLX2-5 (cyan) monomers. The metal ions from the GLX2-5 structure are colored orange, and the iron ion from the AtETHE1 structure is colored grey. Arrows point to changes between the folds of the two enzymes.

35

Figure 2.2 Sequence alignment of AtETHE1, ETHE1, GLX2-5, and human glyoxalase II (HumanGLX2). The alignment was created by running a structure- based alignment with VAST between AtETHE1 and GLX2-5, and then performing a pairwise alignment between AtETHE1 and ETHE1, and between GLX2-5 and human glyoxalase II. Residues involved in dimer formation in AtETHE1 are colored green; residues involved in substrate binding in human glyoxalase II are colored orange; residues which belong to metal site 1 are colored blue; residues which belong to metal site 2 are colored red; and residues that have been implicated in ethylmalonic encephalopathy are marked with an asterix.

36

Figure 2.3 The AtETHE1 dimer. The two subunits are colored cyan and magenta.

37 storage gave a metal/protein ratio of 0.56. This closely matches what was observed in the crystal structure, where an occupancy of 0.5 for the iron ion gave the best refinement results. Despite this partial occupancy, the electron density for the coordinating ligands is well-defined without any indication of heterogeneity or multiple conformers. The iron ion location is identical to one of the two metal ions in the AtGLX2-5 protein structure. Fig. 2.4 depicts an overlay of the residues involved in metal binding in the AtGLX2-5 and AtETHE1 enzymes. In the AtGLX2-5 protein structure, one metal ion was tetrahedrally-coordinated to three (His-54, His-56, and His-112) and a bridging water molecule. The coordination of the equivalent iron ion in the AtETHE1 structure was octahedral. In addition to three waters, the iron ion was bound to His-72 and His-128, homologs of His-54 and His-112 in AtGLX2-5. The homolog of Asp-131, Asp-153, which binds the second metal atom in AtGLX2-5, is slightly shifted such that it also directly coordinates the iron ion in AtETHE1. The homolog of His-56, His-74, does not coordinate the metal in AtETHE1. In the AtGLX2-5 structure there is a single turn helix containing His-56 near its N-terminal end. In the AtETHE1structure, this helix has been pulled apart such that the sidechain of His-74 is no longer directed towards the metal atom. This change also displaces the sidechains of Asp-76 and His-77, whose structural equivalents in AtGLX2- 5 coordinate the second metal. It is unclear if the unwinding of this helix is simply due to a missing metal atom or if it accurately represents an active conformation of the protein. There are two sequence features of AtETHE1 that indicate this unraveling may be more likely to occur in this enzyme than in the glyoxylase II isozymes. The unwinding of this helix places the sidechain of Ala-75 directly into the active site of the enzyme. In glyoxylase II enzymes, this alanine is replaced by a residue with a bulky sidechain, for instance a tyrosine in AtGLX2-5, which may cause additional steric problems upon unwinding of the helix. In addition, there is a modification to a conserved glyoxylase II CGK(L/F)(F/Y)EG motif (Cys-138, Gly-139, Lys-140, Leu-141, Phe-142, Glu-143, and Gly-144 in AtETHE1) which alters the sequence to CGRTDFQEG. The sidechain of the inserted glutamine, Gln-166, directly occupies the space where the side chain of His-74 would need to be to coordinate the Fe ion and forms a hydrogen bond with the carbonyl of Val-73. This displacement is further stabilized by hydrogen bonds formed between the

38

Figure 2.4 Overlay of metal-binding residues in the AtETHE1 enzyme (magenta) and the GLX2-5 enzyme (cyan). The metal ion from the AtETHE1 structure is labeled purple. The metal ions from the GLX2-5 structure are labeled orange. The backbone of both enzymes is also depicted in order to illustrate the unwinding of the helix near the metal-binding site. Labels correspond to AtETHE1.

39 sidechains of Asp-164 and Arg-162. These residues appear to be strongly conserved in ETHE1 proteins. The second metal atom in AtGLX2-5 was coordinated by His-59, His-169, Asp- 58, Asp-131 and a bridging water molecule. As mentioned previously, the homologs of His-59 and Asp-58 in AtETHE1 are displaced due to the unraveling of a single turn helix; however, the homolog of Asp-58 (Asp-76) is still within coordination distance of a second metal atom. The homolog of His-59 (His-77) is pulled away from the putative second metal atom site and in its current position is unlikely to coordinate a metal atom. The sidechain of His-74 in the AtETHE1 enzyme is near the location His-77 would need to occupy if it were to coordinate a second metal atom, and may serve as a replacement for binding a second metal atom. In the AtETHE1 structure a carboxyl oxygen of the homolog of Asp-131 is 2.1 Å from the iron ion, and also near the expected location of the second metal atom. The homolog of His-169, His-194, is positioned identically to His- 169 and could conceivably coordinate a second metal atom in AtETHE1. Ultimately, all of the residues necessary for binding a second metal atom similarly to GLX2-5 are present in AtETHE1, however, some structural rearrangements would have to occur to obtain the necessary orientation of the metal binding sidechains. The metal and potential substrate binding residues in AtETHE1 are for the most part involved in the same hydrogen bonding networks observed in AtGLX2-5. The primary exception is for His-194, which does not coordinate a metal in the AtETHE1 structure. The equivalent in AtGLX2-5 was stabilized by interactions with the carboxyl oxygen of an aspartate sidechain. The remaining carboxyl oxygen was hydrogen bonded to a lysine sidechain nitrogen. In AtETHE1, this aspartate is replaced with a serine, limiting further electron delocalization. Also, the imidazole sidechain of His-194 is flipped in the AtETHE1 structure, and the ND1 nitrogen interacts instead with the sidechain of Asp-153.

2.4.4 Active Site:

Residues involved in substrate binding in the human glyoxalase II are almost entirely conserved in AtGLX2-5. This is not the case with AtETHE1. An overlay of the human

40 glyoxalase II substrate binding site with the equivalent residues in AtETHE1is depicted in Fig. 2.5. The side chains of residues Arg-249 and Lys-252 make hydrogen bonds with the glycine portion of glutathione in the human enzyme. These residues are replaced with Met-225 and Leu-228 in AtETHE1. Leu-228 is also pulled away from the substrate binding region. The backbone amino group of Lys-143, and the sidechain of Tyr-175 in human glyoxalase II were shown to hydrogen bond to the cysteine portion of glutathione. In AtETHE1, Lys-143 is replaced with Arg-162; however, the backbone amino groups of these two residues overlap. Replacement of this lysine with arginine appears to be common among ETHE1 proteins, and loss of this arginine has been linked with ethylmalonic encephalopathy in humans (11). The AtETHE1 homolog of Tyr-175, Tyr- 196, is also positioned similarly to that of the human enzyme. In human glyoxalase II the side chains of Tyr-145 and Lys-143 form hydrogen bonds with the glutamate portion of the glutathione. AtETHE1 has a phenylalanine in place of Tyr-145; however, Arabidopsis glyoxalase II, AtGLX2-5 has the same substitution. Lys-143 is replaced with Arg-162 in AtETHE1. The arginine side chain could conceivably interact with the substrate as well. The sidechain of Arg-162 appears to be firmly held in place by two hydrogen bonds with the sidechain carboxyl of Asp-164 and is not likely to be as flexible as Lys-143. The active site of AtETHE1 has significantly less room for substrate binding than that of the human and Arabidopsis glyoxalase II enzymes due to the unwound helix described above and the extended C-terminal region that covers up much of the active site. While there appears to be enough room to fit a glutathione group, there is substantially less space available for the other portion of the thioester.

2.4.5 Structural Basis for Encephalopathy:

A number of single residue mutations have been identified in the ETHE1 protein of patients with ethylmalonic encephalopathy (11, 26) and a model of human ETHE1 was created based on the human glyoxalase II crystal structure (26). Our results confirm many of the observations of this study and provide some additional details. Mutations in human ETHE1 include R163Q, C161Y, Y38C, L185R, and T136A. The Arabidopsis equivalent of Arg-163, Arg-162, is located within the expected active site of the enzyme, and forms

41

Figure 2.5 Overlay of the substrate binding residues from the human glyoxalase II (cyan) with the equivalent residues in the AtETHE1 enzyme (magenta). The S-hydroxybromophenylcarbamoyl glutathione bound to glyoxalase II is labeled yellow. Labels correspond to AtETHE1.

42 two hydrogen bonds with the sidechain of Asp-164. Given the location of this residue and the substantial electron delocalization provided by the aspartate, it seems likely that this residue may be directly involved in the catalytic mechanism of the enzyme. The Arabidopsis equivalent of Cys-161, Cys-160, is also near the active site. Mutation of this residue into a bulky tyrosine would clearly reposition Arg-162 and possibly other amino acids involved in substrate binding. The Arabidopsis equivalent of Thr-136, Thr-129, forms a hydrogen bond with a backbone amine nitrogen that may help stabilize His-128 which coordinates the iron ion in AtETHE1. The Arabidopsis equivalent of Tyr-38, Tyr- 29, is part of the hydrophobic interface between the two internal β-sheets and sits in a pocket of cyclic, aromatic sidechains which include Phe-27, Phe-16, Phe-156, and Tyr- 191. In addition the Tyr-29 hydroxyl group forms a hydrogen bond with the sidechain of Gln-18, which is part of the dimer interface. Mutation of this residue could subsequently have repercussions on the stability of both the tertiary and quaternary structures. The Arabidopsis equivalent of the final relevant human residue, Leu-184, is far from the active site and sits in the region between the C-terminal β-sheet and the C-terminal α- helical region. The residue is near the sidechains of Arg-123 and Arg-147. Mutation of this residue into another arginine would likely be highly destabilizing.

2.4.6 Sequence Analysis:

A series of BLAST searches were conducted to determine the prevalence of ETHE1-like enzymes. The assumption was made that the presence of Arg-162, the absence of residues involved in binding glutathione, and the absence of the two-helix bundle observed in glyoxalase II are features that distinguish between ETHE1 and glyoxalase II enzymes. ETHE1-like enzymes were observed in almost all forms of life including animals, plants, fungi, eubacteria, and archaebacteria. Unlike glyoxalase II enzymes, multiple isozymes of the ETHE1 were not detected within a species. It was also noted that in some archaebacteria, such as halobacteria, the ETHE1-like fold was coupled to a rhodanese -like domain.

43 2.5 Conclusions

The crystal structure of AtETHE1 has been solved and refined to 1.48 Å. The structure reveals a fold that varies from the closely-related enzyme glyoxalase II. The removal of a two-helix bundle in AtETHE1 results in the formation of a dimer interface missing from the glyoxalase II enzymes. The extended C-terminus which aligns the active site, as well as several changes in the substrate binding residues of glyoxalase II enzymes allow for a different, unknown reaction chemistry. Also, the structure revealed a metal binding site as well as a possible second metal site given some structural rearrangement. In addition, the structure of AtETHE1 is the closest model available for human ETHE1 and provides a structural explanation for the deleterious effects of several mutations corresponding to the onset of ethylmalonic encephalopathy as well as reveals the active site architecture involved in binding an as yet unknown substrate.

44 2.6 Acknowledgements

This work was supported by the National Institutes of Health, Protein Structure Initiative P50 GM 64598, U54 GM 074901 (John L. Markley, P.I., George N. Phillips Jr. and Brian G. Fox, Co-Investigators), NLM training grant T15 LM007359 (J.G.M), and the National Science Foundation MCB-9817083 (CAM). The Advanced Photon Source is supported by the U.S. Department of Energy, BasicEnergy Sciences, Office of Science, under contract No. W-31-109-ENG-38. Data were collected at Southeast Regional Collaborative Access Team (SER-CAT) 22-ID beamline at the Advanced Photon Source, Argonne National Laboratory. Supporting institutions may be foundat www.ser- cat.org/members.html. GM/CA CAT is supported by the National Cancer Institute (Y1- CO-1020) and the National Institute of General Medical Science (Y1-GM-1104). We thank Ward Smith for assistance at GM/CA-CAT. We specially thank Gary Wesenberg, Dave Aceti, Janelle Warick, and other members of the CESG staff. Data were collected at Southeast Regional Collaborative Access Team (SER-CAT) 22-ID (or 22-BM) beamline at the Advanced Photon Source, Argonne National Laboratory.

45

2.7 References

1. Gough, J., Karplus, K., Hughey, R., and Chothia, C. (2001). Assignment of Homology to Genome Sequences using a Library of Hidden Markov Models that Represent all Proteins of Known Structure. J. Mol. Biol. 313, 903-919

2. Maiti, M. K., Krishnasamy, S., Owen, H. A., and Makaroff, C. A. (1997). Molecular Characterization of Glyoxalase II from Arabidopsis thaliana. Plant Mol. Biol. 35, 471-481

3. Cameron, A. D., Ridderstrom, M., Olin, B., and Mannervik, B. (1999). Crystal Structure of Human Glyoxalase II and its Complex with a Glutathione Thiolester Substrate Analogue. Structure 7, 1067-1078

4. Marasinghe, G. P., Sander, I. M., Bennett, B., Periyannan, G., Yang, K. W., Makaroff, C. A., and Crowder, M. W. (2005). Structural Studies on a Mitochondrial Glyoxalase II. J. Biol. Chem. 280, 40668-40675

5. Thompson, Guy A. Jr. (1993) Responses of Lipid Metabolism to Developmental Change and Environmental Perturbation. In Lipid Metabolism in Plants pp. 591-619. Edited by Moore, Thomas S. Jr., Boca Raton, Florida, CRC Press.

6. Davidson, G., Clugston, S. L., Honek, J. F., and Maroney, M. J. (2000). XAS Investigation of Actie Sit Structure in Glyoxalase I. Inorg. Chem. 39, 2962-2963

7. Thornalley, P. J. (1998). Glutathione-dependent Detoxification of Alpha- oxoaldehydes by the Glyoxalase System: Inolvement in Disease Mechanisms and Antiproliferative Activity of Glyoxalase I Inhibitors. Chem. Biol. Interact. 111-112, 137-151

46 8. Thornalley, P. J. (1995). Advances in Glyoxalase Research. Glyoxalase Expression in Malignancy, Anti-proliferative Effects of Methylglyoxal, Glyoxalase I Inhibitor Diester and S-lactoylglutathione, and Methylglyoxal-modified Protein Binding and Endocytosis by Advanced Glycation Endproduct Receptor. Crit Rev Oncol Hematol 20, 99-128

9. Crowder, M. W., Maiti, M. K., Banovic, L., and Makaroff, C. A. (1997). Glyoxalase II from A. thaliana Requires Zn(II) for Catalytic Activity. FEBS Lett. 418, 351-354

10. Wenzel, N. F., Carenbauer, A. L., Pfiester, M. P., Schilling, O., Meyer-Klaucke, W., Makaroff, C. A., and Crowder, M. W. (2004). The Binding of Iron and Zinc to Glyoxalase II Occurs Exclusively as Di-metal Centers and is Unique within the Metallo-beta-lactamase Family. J. Biol. Inorg. Chem. 9, 429-438

11. Tiranti, V., D'Adamo, P., Briem, E. et al. (2004). Ethylmalonic Encephalopathy is Caused by Mutations in ETHE1, a Gene Encoding a Mitochondrial Matrix Protein. Am. J. Hum. Genet. 74, 239-252

12. Burlina, A., Zacchello, F., Dionisi-Vici, C., Bertini, E., Sabetta, G., Bennet, M. J., Hale, D. E., Schmidt-Sommerfeld, E., and Rinaldo, P. (1991). New Clinical Phenotype of Branched-chain Acyl-CoA Oxidation Defect. Lancet. 338, 1522-1523

13. Burlina, A. B., Dionisi-Vici, C., Bennett, M. J., Gibson, K. M., Servidei, S., Bertini, E., Hale, D. E., Schmidt-Sommerfeld, E., Sabetta, G., and Zacchello, F. (1994). A New Syndrome with Ethylmalonic Aciduria and Normal Fatty Acid Oxidation in Fibroblasts. J. Pediatr. 124, 79-86

14. Zang, T. M., Hollman, D. A., Crawford, P. A., Crowder, M. W., and Makaroff, C. A. (2001). Arabidopsis Glyoxalase II Contains a Zinc/iron Binuclear Metal Center that is Essential for Substrate Binding and . J. Biol. Chem. 276, 4788-4795

15. Gill, S. C., and von Hippel, P. H. (1989). Calculation of Protein Extinction Coefficients from Amino Acid Sequence Data. Anal. Biochem. 182, 319-326

47 16. Otwinowski, Z., and Minor, W. (1997). Processing of X-ray Diffraction Data Collected in Oscillation Modes. Methods Enzym , 307-326

17. Weeks, C. M., Adams, P. D., Berendzen, J. et al. (2003). Automatic Solution of Heavy-Atom Substructures. Methods Enzym 374, 37-83

18. Adams, P. D., Grosse-Kunstleve, R. W., Hung, L. W., Ioerger, T., McCoy, A. J., Moriarty, N. W., Read, R. J., Sacchettini, J. C., Sauter, N. K., and Terwilliger, T. C. (2002). PHENIX: Building New Software for Automated Crystallographic Structure Determination. Acta Crystallogr. D Biol. Crystallogr. 58, 1948-1954

19. Bricogne, G., Vonrhein, C., Flensburg, C., Schiltz, M., and Paciorek, W. (2003). Generation, Representation, and Flow of Phase Information in Structure Determination: Recent Development in and around SHARP 2.0. Acta. Crystallogr. D Biol. Crystallogr. 59, 2023-2030

20. Collaborative Computational Project, N. (1994). The CCP4 Suite: Programs for Protein Crystallography. Acta Crystallogr. D Biol. Crystallogr. 50, 760-763

21. Abrahams, J. P., and Leslie, A. G. (1996). Methods Used in the Structure Determination of Bovine Mitochondrial F1 ATPase. Acta Crystallogr. D Biol. Crystallogr. 52, 30-42

22. Blanc, E., Roversi, P., Vonrhein, C., Flensburg, C., Lea, S. M., and Bricogne, G. (2004). Refinement of Severely Incomplete Structures with Maximum Likekihood in BUSTER-TNT. Acta Crystallogr. D Biol. Crystallogr. 60, 2210-2221

23. Emsley, P., and Cowtan, K. (2004). Coot: Model-Building Tools for Molecular Graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126-2132

24. 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

48 25. Jones, S., and Thornton, J. M. (1995). Protein-protein Interactions: A Review of Protein Dimer Structures. Prog Biophys Mol Biol 63, 31-65

26. Tiranti, V., Briem, E., Lamantea, E. et al. (2006). ETHE1 Mutations are Specific to Ethylmalonic Encephalopathy. J. Med. Genet. 43, 340-346

49

Chapter 3

Spectroscopic Studies of Arabidopsis ETHE1, a Glyoxalase II-like Protein

Meghan M. Holdorf1, Brian Bennett2, Michael Crowder1; and Christopher Makaroff1

1 Department of Chemistry and Biochemistry, Miami University, Oxford OH 45056 2 National Biomedical EPR center, Department Of Biophysics, Medical College of Wisconsin, Milwaukee, Wisconsin 53226

This paper to be submitted to: Journal of Inorganic Biochemistry

Authors contribution to work: All work presented in this chapter was contributed by M. M. Holdorf with the exception of the EPR studies which were performed by Brian Bennett.

50 3.1 Summary: ETHE1 is a β-lactamase fold containing protein that is essential for the survival of a range of organisms. In spite of the apparent importance of this enzyme, very little is known about its function or biochemical properties. In this study we demonstrate that Arabidopsis ETHE1 contains a mononuclear Fe center. ETHE1 is most similar to the GLX2 family of proteins which have been shown to contain a dinuclear metal center. While ETHE1 contains all of the conserved GLX2 metal binding ligands, ETHE1 appears to have evolved to accept a single metal atom in an octahedral configuration. Metal analyses demonstrate that ETHE1 binds tightly to a single Fe. Spectroscopic studies are consistent with the presence of a single Fe (II) atom that is bound in a modified zinc 1 site of the metallo-β-lactamase metal binding motif. In addition, we present a new experimentally determined extinction coefficient for ETHE1 that is more consistent with our experimental results than the previously calculated extinction coefficient. These results represent the first detailed structural characterization of an ETHE1-like enzyme from any organism.

51 3.2 Introduction

Ethylmalonic Encephalopathy protein 1 (ETHE1) is a recently discovered gene, which when mutated is the cause the complex human disease Ethylmalonic Encephalopathy (EE) (1, 2). The symptoms of EE are well characterized, yet the underlying physiological cause of the disease remains unknown (2). EE is an autosomal recessive disease with symptoms that include: chronic diarrhea, a delay in neural development, symmetric brain lesions, relapsing petechiae, lactic academia, and acrocyanosis of the hands and feet, all of which lead to death within the first few years of life (3). Additional biological markers of EE are elevated levels of C4 and C5 plasma acylcarnitines and elevated levels of ethylmalonic acid (EMA). Based on these biochemical alterations it has been suggested that ETHE1 may be involved in the β- oxidation of short-chain acyl-CoAs (2). ETHE1-like enzymes are found in most organisms including archaebacteria, fungi, plants, and animals, suggesting that it serves a fundamental biochemical role in nature; however the exact biochemical role of ETHE1 is currently unknown. ETHE1 shows the greatest sequence similarity to the glyoxalase II family of proteins. Glyoxalase II catalyzes the second step of the glyoxalase pathway, which consists of glyoxalase I () (GLX1) and glyoxalase II (hydroxacylglutathione hydrolase) (GLX2), and is thought to be involved in the detoxification of 2-oxaldehydes, toxic byproducts of carbohydrate and lipid metabolism (4). The primary physiological substrate appears to be methylglyoxal (2-oxopropanal), a reactive α-oxoaldehyde (4). Cellular detoxification of methylglyoxal begins with its non-enzymatic reaction with glutathione to produce a glutathione hemiacetal, which is then converted to S-D- lactoylglutathione (SLG) by GLX1 (5). GLX2 then hydrolyzes SLG to form D-lactic acid and free glutathione (6). GLX2 enzymes belong to the metallo-β-lactamase fold family of proteins, which are typically dinuclear metallohydrolases. The metallo-β-lactamases typically bind two equivalents of Zn(II) and hydrolyze β-lactams (10, 11). GLX2 enzymes, which hydrolyze SLG, contain dinuclear metal centers that can bind Zn, Fe, and Mn (12).

52 Additional protein families containing the β-lactamase fold include the rubredoxin:oxygen oxidoredutase (ROO) and ZiPD families, which have been shown to bind divalent Fe and divalent Zn respectively (13, 14). Therefore, the β-lactamase fold can accommodate several different metals and is present in enzymes that can catalyze a wide range of reactions. The GLX2 dinuclear metal binding center typically consists of one metal site that is tetrahedrally coordinated by three histidines , while the second metal site, is coordinated by two histindines, two aspartates, and two waters (15-17). Arabidopsis ETHE1 is 54% identical to human ETHE1. Crystal structure comparisons between Arabidopsis ETHE1, which was previously chacterized as a glyoxalase enzyme (7) and Arabidopsis GLX2-5, a mitochondrial GLX2 enzyme, revealed that while the proteins show only 13% sequence identity, ETHE1 is structurally very similar to GLX2 (18). ETHE1 enzymes share several structural features with GLX2 enzymes, including sequence and structural characteristics of the metal binding domain and the β-lactamase fold consisting of two central mixed β-sheets surrounded on both sides by helices (15, 18). Interestingly, even though ETHE1 and GLX2 enzymes exhibit extensive similarity in their metal binding regions, the ETHE1 crystal structure showed it only bound a single metal atom (18). In fact, the crystal structure was best refined to a metal occupancy of 0.5 (18). However, metal analysis of purified ETHE1 prior to crystallization had indicated a metal/protein ratio of 2, raising the question of the actual metal content of this enzyme. Several other features were identified that set ETHE1 apart from the GLX2 family. An important difference between the two proteins involves the substrate binding pocket. Sequence alignments, as well as the crystal structure comparison, identified several amino acid substitutions in ETHE1 of residues involved in SLG binding in GLX2 (15, 18). These changes in addition to the presence of an extended C-terminal tail are predicted to render the substrate binding pocket of ETHE1 too small for the GLX2 substrate SLG. Additionally, unlike GLX2, ETHE1 was predicted to function as a dimer due to the absence of a two helix bundle in ETHE1 that is present in the GLX2 enzymes (15, 18). We have over-expressed, purified, and conducted a biochemical and spectroscopic characterization of Arabidopsis ETHE1 to investigate these differences

53 further and to better understand the functional role of ETHE1. The results of these studies show that Arabidopsis ETHE1 is homodimeric in solution, exhibits low-level esterase activity, and specifically binds a single Fe (II) atom in the active site.

3.3 MATERIALS AND METHODS

3.3.1 Over-expression and Purification. The Arabidopsis ETHE1 cDNA was obtained from Arabidopsis bud RNA and cloned into pET24b (Novagen) as a Nde1 and XhoI fragment following reverse transcription and PCR amplification using the primers GLX2-3 (TCTTCTCATATGAAGCTTCTCTTTCGTCAAC) and a 3’poly (A) anchor primer. During the cloning the N-terminal leader peptides was removed for high-level expression of ETHE1 in E. coli, resulting in the amino terminal methionine, which corresponds to amino acid 51 of the predicted protein sequence. This is the same form of the protein that was used for crystal structure determination (18). After verification by DNA sequencing, pET24b- ETHE1 was transformed into BL21-Codon Plus (DE3)-RIL cells and used for protein over-expression in ZY media as previously described (12). ETHE1 was purified from cleared lysates by Fast Performance Liquid Chromatography (FPLC) using a Q-Sepharose column as described previously (19). Protein purity was determined by SDS-polyacrylamide gel electrophoresis, and protein concentrations were initially determined using the extinction coefficient 10,240 M-1cm-1, which is based on the ETHE1 amino acid composition (20).

3.3.2 Extinction Coefficient Determination. The concentration of purified ETHE1 was also determined using a BCA protein assay kit (Pierce) and a BSA standard curve to evaluate the calculated extinction coefficient of ETHE1. Using the protein

standard curve method and an Abs280 reading, an ETHE1 extinction coefficient was determined (21). The validity of the two extinction coefficients was further evaluated through SDS PAGE protein loading comparisons between known amounts of GLX2-5 and calculated amounts of ETHE1. A series of known amounts of GLX2-5 (0.5 μg, 1 μg, 3 μg, and 5 μg) and predicted amounts (based on the new extinction coefficient) of

54 ETHE1 were run side by side on an SDS-PAGE gel and the amount of Coomassie Blue- stained protein measured using densitometry. Integrated density values were determined for each protein at the various concentrations and differences in the amount of the two proteins were determined.

3.3.3 Metal Analysis. Metal analyses were performed on the purified enzyme using a Varian-Liberty 150 inductively coupled plasma spectrometer with atomic emission spectroscopy detection (ICP-AES) as described (15). The purified protein was diluted to 10 μM in 50 mM TRIS, pH 7.2, and analyzed for the presence of zinc, manganese, iron, and . The metal data presented here represent the average of at least three preparations for each metal addition experiment.

3.3.4 Native Molecular Weight Determination. The native molecular weight of ETHE1 was determined utilizing a Sephadex S200 column in 10mM MOPS, pH 7.2, containing 0.15 M NaCl. ETHE1 (1 mg), purified as described above, was mixed with 1 mg of each of the protein standards: Blue Dextran, BSA, Ovalbumin, Aldolase, and Ribonuclease A. One milliliter fractions were collected with a flow rate of 0.5 ml/min,

and samples containing protein were identified by monitoring A280 and by SDS-PAGE gel analysis.

3.3.5 Substrate Analysis. Enzymatic assays were conducted at 25 ºC in 10 mM MOPS buffer, pH 7.2, on a Cary IE UV-Vis Spectrophotometer. A series of thioesters of glutathione were synthesized as previously described (22), and all other substrates were purchased commercially. The hydrolysis of S-D-lactoylglutathione (Sigma), S-D-acetylglutathione, S-D-acetoacetylglutathione, S-D-formylglutathione, S- D-glyocosylgutathione or S-D-pyruvylglutathione was monitored at 240 nm. S-D- mandeloylglutathione hydrolysis was monitored at 263 nm (22). Hydrolysis of p- nitrophenyl phosphate (Sigma), p-nitrophenyl sulfate (Sigma), and p-nitrophenyl acetate (Sigma) was monitored at 400 nm (23-25). The hydrolysis of L-alanine-p-nitroanilide (Sigma) was measured at 404 nm (24). Ala-ala-ala-p-nitroanilide (Sigma) and y-L-Glu- p-nitroanalide (Sigma) hydrolysis was monitored at 410 nm. Hydrolysis of

55 benzoylglycyl-L-phenylalanine was monitored at 254 nm (26). Nitrocefin (Becton- Dickinson) hydrolysis was monitored at 485 nm (27). Methylglyoxal was assayed colorimetrically by using the 2,3-dinitrophenylhydrazine-alkali reaction (28). Assays were performed for 5 min using 80 μM substrate and varying concentrations of pure enzyme both as isolated and after loading with excess iron. Additional substrate screening was performed using the Micronaut-Taxa Profile E (Merlin GmbH) microtiter plate (13). Each well was filled with 25 μl of 40 μM ETHE1, and reactions prepared according to the plate’s directions. A negative control was performed using MOPS buffer pH 7.2 to rule out non-specific reactions. The reactions were monitored visually for 24 hours at 37 ºC, and positive reactions were recorded. Each reaction was performed in duplicate using enzyme loaded with iron.

3.3.6 EPR Spectroscopy. EPR was carried out at 9.63 GHz and 10 K using a Bruker EleXsys E580 spectrometer equipped with an ER4116DM cavity, and an Oxford ESR900 liquid helium flow cryostat and ITC503 temperature controller. 12 G (1.2 mT) field modulation at 100 kHz was employed. Samples contained 4.1 mM ETHE1 protein in 50 mM TRIS, pH 7.2.

3.3.7 1H NMR Spectroscopy. NMR spectra were collected on a Bruker Avance 500 spectrometer operating at 500.13 MHz, 298 K, and a magnetic field of 11.7 tesla, recycle delay (AQ), 41 ms; sweep width, 400 ppm. Concentrated samples of ETHE1 (3.5

mM) contained 10% D2O for locking or 90% D2O for monitoring of solvent-

exchangeable peaks. Protein chemical shifts were calibrated by assigning the H2O signal a value of 4.70 ppm, and a modified presaturation pulse sequence (zgpr) was used to suppress the signals that originated from the water molecules.

3.4 RESULTS

3.4.1 Over-expression, Purification, and Characterization of Arabidopsis ETHE1

56 Based on publicly-available localization prediction programs (pSORT II, Mitoprot), recombinant ETHE1 was cloned into the pET24b expression vector after removing the predicted 51 amino acid N-terminal leader sequence, which generated an N-terminus of MKLLFRQ (Fig. 3.1). This plasmid was transformed into BL21 (RIL) Rosetta cells and over-expressed as described in the Material and Methods. ETHE1 was purified using FPLC Q-Sepharose chromatography, eluting from the column at ~125 mM NaCl. Purified ETHE1 protein was obtained in high yield (~400-600 mg of protein/L) and was >95% pure (data not shown).

3.4.2 ETHE1 Functions as a Dimer

In contrast to the GLX2 enzymes, which exist as monomers in solution (35, 36), the crystal structure of Arabidopsis ETHE1 suggested that it has a dimeric quaternary structure (18). Gel filtration studies were performed on the recombinant ETHE1 enzyme to test this hypothesis. ETHE1 eluted from an S200 column between standard proteins adolase (158 kDa) and ovalbumin (44 kDa), resulting in a calculated molecular weight of 58.3 kDa, which is roughly twice the recombinant monomeric weight of 26.8 kDa (Fig. 3.2). Therefore, Arabidopsis ETHE1 exists as a dimer in solution.

3.4.3 Calculation of the ETHE1 Extinction Coefficient

Given that the GLX2 metal binding residues are conserved in ETHE1, the metal coordination chemistry was expected to be similar between GLX2 and ETHE1 (15). It was therefore predicted that ETHE1, like the GLX2 enzymes, would bind two equivalents of metal. While metal analyses, using a predicted extinction coefficient of 10,240 M-1cm-1, which is based on the amino acid composition of the protein using a method described by Gill and Von Hippel (20), showed that ETHE1 contained two equivalents of iron, the ETHE1 crystal structure only showed a single metal atom in the metal binding pocket (18). Furthermore, the ETHE1 crystal structure was best refined at

57

Figure 3.1: Sequence Alignment of ETHE1 and select metallo-β-lactamase fold proteins. Protein sequences were aligned using CLUSTAL W (1.83). Identical residues are highlighted with black and similar residues with gray highlight. Conserved metal binding ligands are indicated by closed triangles. Residues labeled with an asterisk are mutated in patients with EE. N-terminal processing site is indicated by an arrow.

58

Figure 3.2: Gel filtration elution profile of internal protein standards and ETHE1. Approximately 1 mg of each of the following internal protein standards was separated on a Sephadex S200 column: Blue Dextran, BSA, Ovulbumin, Aldolase, and Ribonuclease A. Protein containing fractions were monitored by

A280 as well as SDS-PAGE.

59 a metal occupancy of 0.5 suggesting that not all of the protein actually contained bound metal (18). The discrepancy in the two results could be due to either (1) an inaccurate protein concentration or that (2), the metal disassociating from the protein during the crystallization process. Given that the previously determined protein concentration was based on a predicted extinction coefficient, we determined the extinction coefficient using a second method. An ETHE1 extinction coefficient of 1,386 M-1cm-1 was obtained using the standard protein curve method of molar absorptivity determination (21). This number is 86.5% lower than the calculated value using the Gill and Von Hippel method (20). We next evaluated which of the two extinction coefficients seemed to best estimate the true concentration of ETHE1 by comparing predicted amounts of ETHE1 with known amounts of GLX2-5 (Fig 3.3). Arabidopsis GLX2-5 has been shown to bind two equivalents of metal and is similar to ETHE1 in both molecular weight and amino acid composition (29). When predicted amounts of ETHE1, which were based on the standard curve extinction coefficient, were compared to known amounts of GLX2-5 using SDS-PAGE, the loadings were similar, but not identical. This suggests that the true extinction coefficient for ETHE1 is likely between the two determined values, but closer to the standard curve molar absorptivity of 1,386.7 M-1cm-1 (Fig 3.3). Integrated density values for ETHE1 and GLX2-5 from the PAGE experiments were used to calculate a new temporary extinction coefficient (2,500 M-1cm-1) for ETHE1, which is used in the remainder of these studies (Table 3.1). Amino acid analysis is currently being performed to obtain a more accurate extinction coefficient, which can be used in subsequent studies.

3.4.4 Metal Analysis

Further studies on the metal content of the purified recombinant ETHE1 protein were then performed using the new extinction coefficient. GLX2 enzymes have been shown to bind iron, zinc, and manganese, and therefore ETHE1 was analyzed for all three of these metals and also copper (29, 30). When grown and over-expressed in rich media

containing 250 μM Fe(NH4)2(SO4)2 and Zn(SO4)2, as isolated ETHE1 was found to

60

Figure 3.3: Protein quantification of ETHE1 based on known amounts of GLX2-5. SDS-PAGE gel of side by side comparisons of known amounts (0.5 μg- 5 μg) of GLX2-5 and predicted equivalent amounts of ETHE1 based on an -1 extinction coefficient of 1386.7 M cm-1.

61

Measurement of ETHE1 extinction coefficient IDV μg ETHE1 loaded μg ETHE1 loaded μg GLX2-5 IDVb % based on ETHE1 based on ETHE1 ε loadeda GLX2-5 ETHE1 Difference ε =1386.7 M-1cm-1 =10240 M-1cm-1 0.5 μg 212891 273464 0.78 0.89 6.68 1 μg 436608 538590 0.81 1.8 13.4 3 μg 646506 942278 0.69 5.1 40.1 5 μg 756051 966857 0.78 8.9 66.8

a Amounts loaded were based on Abs280 reading and GLX2-5's calculated ε coefficient. Similar amounts of ETHE1 were loaded based on concentration determined by BCA. b Integrated Density Values

Table 3.1. Determination of ETHE1’s extinction coefficient. All measurements are from the electrophoresis experiment of GLX2-5 and ETHE1 shown in Figure 6. IDV readings were calculated using Alpha Ease Fc Software version 4.0.0. Extinction coefficient for ETHE1 calculated to be 2,496 M-1 cm-1 based on these experiments.

62 contain 0.13 ± 0.04 equivalents of iron, and less than 0.001 eq. of zinc, manganese, and copper. While low, this suggested that ETHE1 may preferentially bind iron. Interestingly, when the purified enzyme was incubated with excess iron, ETHE1 was found to bind 0.46 ± 0.06 eq. of iron, and no detectable traces of zinc, manganese, or copper. This result is consistent with the crystal structure, which showed that when prepared this way, ETHE1 only binds 0.5 equivalents of iron. Unlike the GLX2 enzymes, the addition of excess zinc to the isolated enzyme resulted in no additional metal binding, suggesting that ETHE1 may be specific for iron.

3.4.5 ETHE1 Does Not Hydrolyze SLG

ETHE1 had previously been predicted to be a GLX2-like enzyme (7). However, a careful sequence comparison revealed that ETHE1 is lacking several highly-conserved amino acids known to participate in the hydrogen bonding of SLG in GLX2 (15, 18). Furthermore, crystallographic analysis of Arabidopsis ETHE1 showned that the substrate binding pocket is too small to accommodate SLG (18). These results suggested that Arabidopsis ETHE1 does not utilize SLG as a substrate. ETHE1 both as-isolated and after incubation with excess iron, was assayed for thioesterase activity with SLG and various other thioester derivatives of glutathione to test this hypothesis. Consistent with the small substrate binding pocket and absence of critical SLG binding residues, purified ETHE1 did not hydrolyze any of the glutathione thioesters. As isolated ETHE1 and Fe-enriched ETHE1 were also assayed against 188 different substrates, including 95 substrates for peptidases, 17 substrates for diverse reactions, and 76 substrates for glycolytic enzymes, phosphatases, and esterases using a commercially-available substrate screening plate (13) to investigate potential substrates. Three potential substrates were identified through this screening process: Ala-Ala-Ala- βNA, Glu-pNA, and p-nitrophenyl acetate. Steady-state kinetic studies were then conducted using the three compounds. Upon further characterization, Ala-Ala-Ala-βNA and Glu-pNA were found not to be substrates for ETHE1. However, ETHE1 did exhibit a low level of activity against p-NPA (0.79 ± 0.46 nmols/min/mg of enzyme). This low level of activity is similar to the esterase activity (10.1 nmols/min/mg) observed in

63 recombinant rat carbonic anhydrase III when reacted with p-nitrophenyl acetate (31). The activity of ETHE1 towards p-nitrophenyl acetate was inhibited by the presence of a metal chelator, 1,10-o-phenanthroline, indicating that p-nitrophenyl acetate hydrolysis by ETHE1 requires bound metals. Therefore, based on these results and those of the crystal structure, we predict that the ETHE1 substrate may be a relatively small ester.

3.4.6 Spectroscopic Studies on ETHE1 The metal binding site of the iron bound ETHE1 was investigated using 1H NMR spectroscopy (Fig 3.4). The spectra revealed the presence of at least 3 paramagnetically shifted resonances between 110 and -30 ppm. In this spectrum peaks a and c integrate to 1 proton each, while peak b integrates to 3 (Fig 3.4). To further investigate the oxidation state of the bound Fe, solvent exchangeable peaks were monitored in the

presence of D2O (Fig. 3.4). Two of the three protons were found to be solvent- exchangeable (Peak b). Based on the resonance positions and line widths, the exchangable peaks can be assigned to the protons bound to the N-H protons on histidines bound to Fe(II) or possibly to an antiferromagnetically-coupled Fe(III)Fe(II) center (32, 33). The ETHE1 crystal structure shows that the both of the histidine ligands are bound through the ε nitrogen and therefore, peak c and the non-exchangable proton from peak b

are likely due to the β-CH2 protons on the metal bound histidines (Fig 3.4) (18). This is surprising because the β-methionine protons don’t normally shift out to these positions. We cannot rule out the possibility that these peaks are due to ortho protons on metal bound histidines, but they are usually too broad to detect (34). Peak a is likely due to the

β-CH2 protons on the bound Asp ligand (Fig 3.4) (18). This result was surprising, since we initially predicted that ETHE1 would have a metal binding site similar to those in the GLX2 family. NMR specra from GLX2-5, which contains a dinuclear iron center, shows at least eight paramagnetically shifted resonances in-between 110 and -30 ppm that correspond to the to the ligands bound to a Fe(III)Fe(II) antiferromagnetically coupled center (29). The recent crystal structure of ETHE1 demonstrated that His232 may not be in a position to coordinate a bound metal ion (18), which is consistent with NMR spectrum of ETHE1.

64

Figure 3.4: 1H NMR spectrum of 3.5 mM iron bound ETHE1 at pH 7.2. The spectra were collected on a Bruker Avance 500 spectrometer operating at 500.13 MHz, 298 K, and a magnetic field of 11.7 tesla, recycle delay (AQ), 41 ms; sweep width, 400 ppm. The solvent exchangeable peak is labeled with an asterisk.center (20, 24, 26). However, the recent crystal structure of ETHE1 demonstrated that His232 may not be in a position to coordinate a bound metal ion (23).

65 EPR spectroscopy on several different forms of ETHE1 was conducted to further investigate the ETHE1 metal center. The spectrum of as-isolated ETHE1 containing 0.13

eq Fe (‘Feiso-ETHE1’) indicated the presence of rhombic Fe(III) (Fig 3.5A). The signal

was dominated by a broad derivative feature at geff = 4.27; no features that could be considered diagnostic for protein-bound Fe(III) were observed and the origin of the signal is, therefore, unclear. Although precise quantitation of high-spin systems for which more than one Kramers’ doublet is populated is not trivial, double integration and comparison with a similar EPR signal due to Fe(III) from a well-characterized (4-hydroxyphenyl) pyruvate (35) indicated an Fe(III) content of about 0.025 mM. Therefore about 95 % of the iron in Feiso-ETHE1 is EPR-silent. An Fe(III) signal with a resonance at g ~ 4.3 and additional absorption features at lower field were also observed from the analysis of the ETHE1 enzyme containing 0.5 eq

Fe ( ‘Fe0.5-ETHE1) (Fig 3.5B). In this spectrum, the resonance positions indicated a dominant zero-field splitting term (i.e. D,E >> gβBS; the lower field resonances centered at g ~ 5.5 and g ~ 9 terminate abruptly at g = 6 and g = 10, respectively); the lack of resolved structure and the broad absorption from g ~ 4.3 to g ~ 10 indicated a fairly broad distribution of the rhombic zero-field splitting parameter (i.e. strains in E/D) and thermal population of at least two Kramers’ doublets (36). Again, the signal was not definitive for site-specific binding of Fe(III) by ETHE1, although the form of the signal, particularly the inflection on the g ~ 4.3 crossover due to incomplete rhombicity (i.e. E/D 1 is slightly less than /3), is highly reminiscent of the spectrum due to Fe(III) bound to the active site of the metallo-β-lactamase GOB (33). The intensity of the signal was somewhat lower than that of the as isolated ETHE1 and accounted for < 1 % of the total iron. Interestingly, there was no evidence for a ‘g ~ 1.7, 1.8, 1.9’ signal due to an Fe(II)- Fe(III) center (33).

We attempted to disrupt the protein structure of Fe0.5-ETHE1 by applying freeze- thaw cycles, but the EPR signal remained relatively constant (Fig. 3.5C; fewer scans were averaged). The intensity of the signal was found to increase by a factor of 4 upon aerobic incubation for 100 h at 4 °C (Fig. 3.5D). The form of this spectrum is clearly distinct

66

Figure 3.5: EPR spectram of 4.1 mM ETHE1 under different conditions. A, ETHE1 as isolated containing 0.13 equivalents of Fe; B, iron enriched ETHE1 containing 0.46 equivalents of Fe; C, iron enriched ETHE1 after 2 cycles of freeze/thaw; D, iron enriched ETHE1 after 4 days at 4º C. Spectra were collected on a Bruker ESP-300E spectrometer containing an ESR-900 helium flow cryostate operating at 4.7 K with 2 milliwatts of microwave power at 9.48 GHz and 10 G-field modulation at 100 kHz.

67 from that of Feiso-ETHE1, but still accounts for < 5 % of the total Fe. Because of a weak background signal due to trace Mn(II) and Cu(II), the possibility of the presence of a signal due to an Fe(II)-Fe(III) center cannot be ruled out completely, but the population of such a center, if not zero, must be very low.

3.5 DISCUSSION

Results presented in this paper represent the first detailed structural characterization of an ETHE1 protein from any organism. ETHE1 is most similar to the GLX2-family of proteins, which are members of the metallo-β-lactamase superfamily. This superfamily consists of proteins that catalyze a wide range of reactions, but share the common metal binding motif H-x-[EH]-x-D-[CRSH]-X50-70-[CSD]-X , which is part of the common β-lactamase fold (37). This motif typically coordinates two metal ions that are essential for the activity of the majority of the enzymes (37). In most β-lactamase fold containing proteins the coordination of the first metal (zinc 1 site) is tetrahedral and consists of three histidines, while the second metal binding site (zinc 2 site) is trigonal pyramidal. The site 2 metal binding ligands are more variable but always contain a histidine and aspartic acid (Fig 3.6A) (37). Because of the similarity of ETHE1 to the GLX2 family of enzymes, it was predicted that ETHE1 would also bind two equivalents of metal. All metal analyses of ETHE1 done prior to this paper were based on a calculated extinction coefficient of 10,240 M-1cm-1 (18). Metal analyses using this extinction coefficient indicated that ETHE1 bound two equivalents of iron; however the ETHE1 crystal structure reported only a 0.5 iron metal occupancy (18). It was originally thought that this discrepancy resulted from the loss of metal during the crystallization processes (18). However, results presented here suggest that the extinction coefficient used for these experiments may be inaccurate. When additional methods, including a BCA assay and integrated density values of standardized SDS- PAGE protein loading were used to determine the ETHE1 extinction coefficient, a much lower value, 2,500 M-1cm-1 was obtained. This is about 76% lower than what was previously calculated. Using 2.500 M-1cm-1 as the extinction coefficient, we found that

68

Figure 3.6: Dinuclear metal binding center models for A.thaliana GLX2 and A. thaliana ETHE1. (A) Dinuclear metal binding center model for A.thaliana GLX2-2. This model is based from the metal binding center from the crystal structure of human GLX2. (B) Metal binding center model of A. thaliana ETHE1. This model is based on the metal binding center from the crystal structure of A. thaliana ETHE1. These models imply neither site-specific nor distributed binding. Me = metal.

69 after the addition of excess iron ETHE1 showed an iron to protein ratio of 0.5 equivalents. This number is likely low, but it is consistent with the ETHE1 crystal structure (18). Based on the uncertainity in the extinction coefficient we predict that ETHE1 most-likely binds one iron per protein monomer. Consistent with the crystal structure of ETHE1, our results suggest that ETHE1 contains a single Fe (II) that is coordinated by two histidines (Fig 3.5, 3.6B) (18). GLX2 enzymes enriched in iron typically contain an antiferromagnically coupled Fe(II)-Fe(III) center in their dinuclear metal binding site (29, 30). Results from our EPR studies argue against the possibility that ETHE1 contains an Fe(II)-Fe(III) site. Two distinct EPR signals were observed for ETHE1, both due to magnetically isolated Fe(III) and both accounting for only a very small proportion (1 – 5 %) of the total iron. Even extensive exposure to air was unsuccessful in increasing the Fe (III) signal. Therefore, the population of ETHE1 molecules containing an Fe(II)-Fe(III) center must be very small, perhaps zero. The very low intensities of the Fe(III) signals can either be due to most of the iron being in the Fe(II) state, or most of the iron residing in an anti-ferromagnetically- coupled S’ = 0 dinuclear site. The lack of signals due to Fe(III)-Fe(II) where Fe(III) is not observed and the crystallographic identification of a monometallated active site (18) both argue against a predominant Fe(III)-Fe(III) S’ = 0 species and support Fe(II) as the major form of metal in ETHE1. While neither of the two Fe(III) EPR signals was diagnostic for Fe(III) bound tightly and specifically by ETHE1, the similarity of the signals from oxidized, Fe(II)-treated ETHE1 and GOB suggest that this form, rather than as-isolated ETHE1, is more likely to contain Fe(III) bound at the active site (33). ETHE1 protein samples have been sent for amino acid analysis to more accurately determine the ETHE1 extinction coefficient value, which will allow us to accurately determine the metal content of the protein. Even though ETHE1 contains all of the highly conserved metal binding ligands of the metallo β-lactamase family of proteins (27), ETHE1 apparently does not bind two equivalents of metal. In agreement with this result, the crystal structure identified changes in the tertiary structure of the metal binding domain that do not allow for the coordination of a second metal atom (Fig 3.6B) (18). A single-turn helix in ETHE1 containing His112 is pulled away from the active site, displacing His112 away from the

70 metal atom in ETHE1 relative to GLX2-5. This unwinding of the helix also displaces the side chains of Asp 115 and His 116, which have been shown to coordinate the second metal in GLX2-5 (Fig 3.6A-B) (18). Therefore, subtle changes in protein conformation have displaced several of the metal binding ligands, ultimately limiting the ability of ETHE1 to bind two metals. There are other examples of metallo-β-lactamase proteins that only bind one metal. CPhA from Aeromonas hydrophilia and GOB-1 from Elizabethkinga meningosptica both only bind a single Zn(II) atom in the zinc 2 position (Asp, Cys, His) (38, 39). However, in these enzymes, the zinc 1 site is altered either by the replacement of a histidine with an asparagine in CPhA or a histidine with glutamine in GOB-1 (38- 40). To our knowledge ETHE1 is the first example of a β-lactamase family protein that contains all of the conserved metal-binding ligands, yet only binds one metal. In addition, it has been shown that the presence of soft or hard ligands in the metal binding site can effect the specificity of the metal binding (37). All of the metallo- β-lactamases exclusively bind zinc (II) (37). The incorporation of an aspartic acid and an additional histidine in the zinc 2 site likely allows for the variable binding of Fe, Zn, and Mn seen in the GLX2 enzymes (Fig 3.6A) (12, 15, 29, 41). Likewise, the replacement of two soft ligands by an aspartic acid and a glutamate as observed in ROO allows the formation of a di-iron center (14, 37). Interestingly, in the ETHE1 crystal structure, the iron (II) atom is bound in the zinc 1 site which has been modified by the removal of a histidine and the shifting the GLX2 bridging aspartic acid to specially coordinate the iron (Fig 3.6B) (18). The Fe(II) atom is further coordinated by three water molecules resulting in an octahedrally bound metal (18, 37), unlike the tetrahedral coordination of metal normally seen in the zinc 1 site of metallo-β-lactamases. Therefore, ETHE1 proteins appear to represent a new class of ETHE1 in the metallo-β-lactamase fold family of proteins. Although structurally very similar to the GLX2 enzymes, ETHE1 appears to have evolved to accept a single iron atom in an octahedral configuration. Metal analyses and spectroscopic data suggest that unlike GLX2 enzymes ETHE1 tightly binds to a single Fe (II) atom in a modified zinc 1 site of the metallo-β-lactamase metal binding motif. However questions still remain concerning the correct extinction coefficient and therefore concentration of ETHE1. More

71 experiments are necessary to confirm if native ETHE1 does in fact bind a single Fe(II) atom. Finally we show that ETHE1 is homodimeric in solution and exhibits low levels of esterase activity suggesting that ETHE1 might hydrolyze a short-chain ester. Further experiments are required to further classify the ETHE1 substrate and ultimately understand its biochemical role.

72 3.6 References

1. Tiranti, V., Briem, E., Lamantea, E. et al. (2006). ETHE1 Mutations are Specific to Ethylmalonic Encephalopathy. J. Med. Genet. 43, 340-346

2. Tiranti, V., D'Adamo, P., Briem, E. et al. (2004). Ethylmalonic Encephalopathy is Caused by Mutations in ETHE1, a Gene Encoding a Mitochondrial Matrix Protein. Am. J. Hum. Genet. 74, 239-252

3. Burlina, A., Zacchello, F., Dionisi-Vici, C., Bertini, E., Sabetta, G., Bennet, M. J., Hale, D. E., Schmidt-Sommerfeld, E., and Rinaldo, P. (1991). New Clinical Phenotype of Branched-chain Acyl-CoA Oxidation Defect. Lancet 338, 1522-1523

4. Thornalley, P. J. (1993). The Glyoxalase System in Health and Disease. Mol. Aspects Med. 14, 287-371

5. Mannervik, B., and Ridderstrom, M. (1993). Catalytic and Molecular Properties of Glyoxalase I. Biochem. Soc. Trans. 21, 515-517

6. Allen, R. E., Lo, T. W., and Thornalley, P. J. (1993). Purification and Characterization of Glyoxalase II from Human Red Blood Cells. Eur. J. Biochem. 213, 1261-1267

7. Maiti, M. K., Krishnasamy, S., Owen, H. A., and Makaroff, C. A. (1997). Molecular Characterization of Glyoxalase II from Arabidopsis thaliana. Plant Mol. Biol. 35, 471-481

8. Bito, A., Haider, M., Hadler, I., and Breitenbach, M. (1997). Identification and Phenotypic Analysis of Two Glyoxalase II Encoding Genes from Saccharomyces cerevisiae, GLO2 and GLO4, and Intracelluluar Localization of the Corresponding Proteins. J. Biol. Chem. 272, 21509-21519

9. Cordell, P. A., Futers, T. S., Grant, P. J., and Pease, R. J. (2004). The Human hydroxyacylglutathione Hydrolase (HAGH) Gene Encodes both Cytosolic and Mitochondrial Forms of Glyoxalase II. J. Biol. Chem. 279, 28653-28661

73 10. Aravind, L. (1999). An Evolutionary Classification of the Metallo-beta-lactamase Fold Proteins. In. Silico Biol. 1, 69-91

11. Crowder, M. W., Spencer, J., and Vila, A. J. (2006). Metallo-beta-lactamases: Novel Weaponry for Antibiotic Resistance in Bacteria. Acc. Chem. Res. 39, 721-728

12. Zang, T. M., Hollman, D. A., Crawford, P. A., Crowder, M. W., and Makaroff, C. A. (2001). Arabidopsis Glyoxalase II Contains a Zinc/iron Binuclear Metal Center that is Essential for Substrate Binding and Catalyisis. J. Biol. Chem. 276, 4788-4795

13. Vogel, A., Borchers, T., Marcus, K., Meyer, H. E., Krebs, B., and Spener, F. (2002). Heterologous Expression and Characterization of Recombinant Purple Acid Phosphatase from Red Kidney Bean. Arch. Biochem. Biophys. 401, 164-172

14. Frazao, C., Silva, G., Gomes, C. M. et al. (2000). Structure of a Dioxygen Reduction Enzyme from Desulfovibrio gigas. Nat. Struct. Biol. 7, 1041-1045

15. Cameron, A. D., Ridderstrom, M., Olin, B., and Mannervik, B. (1999). Crystal Structure of Human Glyoxalase II and its Complex with a Glutathione Thiolester Substrate Analogue. Structure 7, 1067-1078

16. Concha, N. O., Rasmussen, B. A., Bush, K., and Herzberg, O. (1996). Crystal Structure of the Wild-spectrum Binuclear Zinc Beta-lactamase from Bacteroides fragilis. Structure 4, 823-836

17. Wang, Z., Fast, W., Valentine, A. M., and Benkovic, S. J. (1999). Metallo-beta- lactamase: Structure and Mechanism. Curr. Opin. Chem. Biol. 3, 614-622

18. McCoy, J. G., Bingman, C. A., Bitto, E., Holdorf, M. M., Makaroff, C. A., and Phillips, G. N.,Jr. (2006). Structure of an ETHE1-like Protein from Arabidopsis thaliana. Acta Crystallogr. D Biol. Crystallogr. 62, 964-970

19. Crowder, M. W., Maiti, M. K., Banovic, L., and Makaroff, C. A. (1997). Glyoxalase II from A. thaliana Requires Zn(II) for Catalytic Acitivity. FEBS Lett. 418, 351-354

74 20. Gill, S. C., and von Hippel, P. H. (1989). Calculation of Protein Extinction Coefficients from Amino Acid Sequence Data. Anal. Biochem. 182, 319-326

21. Bollag, D. M., Rozycki, M. D., and Edelstein, S. J. (1996) Protein Methods, Wiley.

22. Uotila, L. (1981). Thioesters of Glutathione. Methods Enzymol. 77, 424-430

23. Anderson, R. A., Bosron, W. F., Kennedy, F. S., and Vallee, B. L. (1975). Role of Mangnesium in Escherichia coli Alkaline Phosphatase. Proc. Natl. Acad. Sci. U. S. A. 72, 2989-2993

24. Brandt, J. J., Chatwood, L. L., Yang, K. W., and Crowder, M. W. (1999). Continuous Assay for VanX, the D-alanyl-D-alanine Dipeptidase Requried for High-level Vancomycin Resistance. Anal. Biochem. 272, 94-99

25. Armstrong, J. M., Myers, D. V., Verpoorte, J. A., and Edsall, J. T. (1966). Purification and Properties of Human Erythrocyte Carbonic Anhydrases. J. Biol. Chem. 241, 5137-5149

26. Sabastian, J. F., and Lo, W. Y. (1978). Mechanisms for Activation and Inhibition of Carboxypeptidase A Catalyzes Hydrolyses of Pepties and Esters. Can. J. Biochem 56, 329-333

27. Crowder, M. W., Wang, Z., Franklin, S. L., Zovinka, E. P., and Benkovic, S. J. (1996). Characterization of the Metal-binding Sites of the Beta-lactamase from Bacteroides fragilis. Biochem. J. 35, 12126-12132

28. Cooper, R. A. (1975) Methylglyoxal synthase. In Methods in Enzymology. pp. 502- 508. Edited by J. N. Abelson, & M. I. Simon, New York, Academic Press.

29. Marasinghe, G. P., Sander, I. M., Bennett, B., Periyannan, G., Yang, K. W., Makaroff, C. A., and Crowder, M. W. (2005). Structural Studies on a Mitochondrial Glyoxalase II. J. Biol. Chem. 280, 40668-40675

75 30. Schilling, O., Wenzel, N., Naylor, M., Vogel, A., Crowder, M., Makaroff, C., and Meyer-Klaucke, W. (2003). Flexible Metal Binding of the Metallo-beta-lactamase Domain: Glyoxalase II Incorporates Iron, Manganese, and Zince in Vivo. Biochemistry 42, 11777-11786

31. Kim, G., Selengut, J., and Levine, R. L. (2000). Carbonic Anhydrase III: the Phosphatase Activity is Extinsic. Arch. Biochem. Biophys. 377, 334-340

32. Bertini, I., Turano, P., and Vila, A. J. (1993). Nuclear Magnetic Resonance of Paramagnetic . Chem. Rev. 93, 2833-2932

33. Moran-Barrio, J., Gonzalez, J. M., Lisa, M. N. et al. (2007). The Metallo-beta- lactamase GOB is a Mono-Zn(II) Enzyme with a Novel Active Site. J. Biol. Chem. 282, 18286-18293

34. Wang, Z., Ming, L. J., Que, L.,Jr, Vincent, J. B., Crowder, M. W., and Averill, B. A. (1992). 1H NMR and NOE Studies of the Purple Acid Phosphatases from Procine Uterus and Bovine Spleen. Biochemistry 31, 5263-5268

35. Purpero, V. M., and Moran, G. R. (2006). Catalytic, Noncatalytic, and Inhibitory Phenomena: Kinetic Analysis of (4-hydroxyphenyl)pyruvate Dioxygenase from Arabidopsis thaliana. Biochemistry 45, 6044-6055

36. Copik, A. J., Waterson, S., Swierczek, S. I., Bennett, B., and Holz, R. C. (2005). Both Nucleophile and Substrate Bind to the Catalytic Fe(II)-center in the Type-II Methionyl Aminopeptidase from Pyrococcus Furiosus. Inorg. Chem. 44, 1160-1162

37. Gomes, C. M., Frazao, C., Xavier, A. V., Legall, J., and Teixeira, M. (2002). Functional Control of the Binuclear Metal Site in the Metallo-beta-lactamase-like Fold by Subtle Amino Acid Replacements. Protein Sci. 11, 707-712

38. Garau, G., Bebrone, C., Anne, C., Galleni, M., Frere, J. M., and Dideberg, O. (2005). A Metallo-beta-lactamase Enzyme in Action: Crystal Structures of the Monozince Carbapenemase CphA and its Complex with Biapenem. J. Mol. Biol. 345, 785-795

76 39. Bellais, S., Aubert, D., Naas, T., and Nordmann, P. (2000). Molecular and Biochemical Heterogeneity of Class B Carbapenem-hydrolyzing Beta-lactamases in Chryseobacterium meningosepticum. Antimicrob. Agents Chemother. 44, 1878-1886

40. Vanhove, M., Zakhem, M., Devreese, B. et al. (2003). Role of Cys221 and Asn116 in the Zinc-binding Sites of the Aeromonas hydrophila Metallo-beta-lactamase. Cell Mol. Life Sci. 60, 2501-2509

41. Wenzel, N. F., Carenbauer, A. L., Pfiester, M. P., Schilling, O., Meyer-Klaucke, W., Makaroff, C. A., and Crowder, M. W. (2004). The Binding of Iron and Zinc to Glyoxalase II Occurs Exclusively as Di-metal Centers and is Unique Within the Metallo-beta-lactamase Family. J. Biol. Inorg. Chem. 9, 429-438

77 Chapter 4

ETHE1, a Gene Associated with Human Ethylmalonic Encephalopathy, is Essential for Endosperm Development in A. thaliana

Meghan M. Holdorf1, Jeongmi S. Rhee1, Heather A. Owen2, Christopher A. Makaroff1*

1Department of Chemistry/Biochemistry, Miami University, Oxford OH 2Department of Biological Sciences, University of Wisconsin-Milwaukee, Milwaukee, WI *Corresponding author: Chris Makaroff Dept of Chemistry and Biochemistry Miami University Oxford OH 45056 Email: [email protected]

This work submitted to Plant Journal. Authors contribution to this work: Initial studies presented in this paper specifically, the isolation of the T-DNA ETHE1 knockout and the initial observation of seed lethality was performed by J.S.Rhee. SEM studies were performed by H.A.Owen. All other work and analyses presented in this paper was contributed by M.M.Holdorf.

78

List of Abbreviations BCAA-branched chain amino acid; CZE-chalazal endosperm; EE-ethylmalonic encephalopathy; EMA-ethylmalonic acid; ETHE1-ethylmalonic encephalopathy protein 1; GLX2-glyoxlase II; LSCM-Laser Scanning Confocal Microscopy; MCE-micropylar endosperm; PEN-peripheral endosperm; SLG-S-D-lactoylglutathione; T-DNA- transformed DNA

79 4.1 Summary

Mutations in the ETHE1 gene result in the complex metabolic disease ethylmalonic encephalopathy, which is characterized by symmetric brain lesions, lactic academia, elevated excretion of ethylmalonic acid, and death in the first decade of life. ETHE1-like genes are found in a wide range of organisms; however, the biochemical and physiological roles of this protein are still unknown. To investigate the role(s) of ETHE1, we characterized the Arabidopsis thaliana ETHE1 homolog and investigated the effect of an ETHE1 loss-of-function mutation. Seeds homozygous for a T-DNA insertion in ETHE1 exhibit an early arrest in endosperm development that then affects embryo development beginning at heart stage resulting in seed lethality. To our knowledge, ethe1(-/-) is one of the earliest characterized mutants displaying defects primarily in the endosperm. Strong ETHE1 labeling was observed in the peripheral and chalazal endosperm of wild-type seeds prior to cellularization. Taken together, these results demonstrate that ETHE1 is essential for endosperm development and provide evidence for a direct role of the endosperm in embryo development.

80 4.2 Introduction

Seed development in angiosperms begins with a double fertilization event in which sperm cells unite with an egg and a diploid central cell resulting in the embryo and triploid endosperm, respectively. Embryo and endosperm development has been extensively studied in a number of plants including Arabidopsis, yet the role of the endosperm during seed development is still largely unknown (1-6). In Arabidopsis, endosperm development can be divided into two separate stages: early endosperm development or the syncytial phase, characterized by successive nuclear divisions without cytokinesis, and late endosperm development, where the endosperm nuclei undergo cellularization and then the cellularized endosperm is consumed by the developing embryo. Early endosperm development begins with the division of the fertilized central cell before the first division of the embryo. Several synchronized syncytial nuclear divisions, accompanied by cell growth, result in a large multinucleate cell that is divided into three distinct domains: the micropylar endosperm (MCE) surrounding the embryo, the peripheral endosperm (PEN) in the central chamber, and the chalazal endosperm (CZE) opposite to the developing embryo (5, 7). By the time the embryo has reached globular stage, the endosperm has undergone seven rounds of free nuclear division and consists of roughly 100 free nuclei. At this stage in development, the endosperm nuclei begin to cellularize signaling the beginning of late endosperm development (7). Cellularization of the endosperm takes place in a wave beginning first with the MCE and progressing through the PEN region. The CZE is the last to cellularize and does so independently of the first wave of cellularization (3, 8). By the time the embryo reaches late heart stage, the endosperm is fully cellularized. Subsequently, it is gradually depleted as the embryo continues to grow and develop (3). Several roles have been proposed for the endosperm in embryo development. One generally accepted role is that the endosperm serves as a physical support for the growing embryo (3). The endosperm is also predicted to be active in nutrient flux, protecting the embryo from physical and osmotic stress (2, 3). A third role of the

81 endosperm is thought to be in nutrient storage. In particular, this is the case in cereal crops where nutrients are stored in the form of starch (9). Unlike monocots, the endosperm in the dicot Arabidopsis does not undergo endoreduplication; however, the positioning and activity of the chalazal region suggests that it might serve a role in nutrient transfer (3, 10, 11). Consistent with this hypothesis, ultrastructural studies on the CZE showed that endoplasmic reticulum, golgi bodies, and small vacuoles are abundant throughout the chalazal cyst with mitochondria concentrated on the top (3). The overall structure and organization of the chalazal endosperm cyst therefore suggests that it could function in the uptake, reprocessing, and release of metabolites (3). Several Arabidopsis mutants with defects in endosperm development have been isolated and characterized (reviewed in (12)). The majority exhibit defects in late endosperm development and are the result of mutations in genes involved in cytokinesis such as the KNOLLE, KEULE, and HINKEL mutants, which also cause defects in embryo development (5, 13-15). Plants containing the KNOLLE, KEULE, and HINKEL mutations all contain embryo and endosperm nuclei of irregular shapes and sizes (13-16). Mutations in KNOLLE and HINKEL have also been shown to disrupt cellularization in the endosperm (5). However, another mutant defective in cytokinesis, spatzle (spz), demonstrated that cellularization of the endosperm is not essential for seed development and germination (5). In spz seeds, cellularization of the endosperm is inhibited; however, the embryo still develops normally until the torpedo stage when growth is stalled. This growth arrest was suggested to be due to the inability of the endosperm to nurture the embryo (5). Surprisingly, even with the cytokinesis defects, seeds homozygous for the knolle and hinkel mutations are able to germinate but die during seedling development (13-15). Seeds homozygous for spz produce plants indistinguishable from wild-type siblings (5). Very few mutants have been isolated that exhibit defects in early endosperm development and therefore, little is known about what role(s) the endosperm plays early in embryo and seed development. One mutant that exhibits early endosperm defects is the haiku2 mutant. In seeds homozygous for the haiku2 mutation, which are ~25% the size of the wild-type seed, the embryo develops normally until late heart stage at which point it shows a decrease in cell proliferation that is coupled with an early onset of

82 endosperm cellularization (2, 17)). This decrease in cell proliferation and precocious cellularization was suggested to be a direct result of the smaller seed size (17). The reduction in embryo growth was in turn predicted to result from the reduced endosperm size, leading to the suggestion that HAIKU2 might be involved in providing nutrient storage and/or delivery from the endosperm to the developing embryo (2). Recently, HAIKU2 was shown to encode a leucine-rich repeat receptor kinase suggesting that it has a role in signal transduction (18). If the endosperm does in fact act as a sink for the storage, processing, and delivery of nutrients to the embryo, then mutations that disrupt metabolic pathways essential for this activity should disrupt this nutrient sink and result in compromised endosperm and ultimately embryo growth. Consistent with this theory, removal of short- chain acyl-CoA oxidase activity, which catalyzes the conversion of acyl-CoA to 2-trans- enoyl-CoA in the first step of the β-oxidation pathway, results in embryo lethality (19). It has been suggested that (a) this lack of short-chain acyl-CoA oxidase activity causes the accumulation of short-chain acyl-CoAs, resulting in toxicity to the developing embryo, or (b) that the activity is directly essential for embryo development (19). β−Oxidation is necessary for fatty acid degradation and also for the degradation of the branched chain amino acids (BCAAs) valine, leucine, and isoleucine. A recently discovered gene in humans, Ethylmalonic Encephalopathy Protein 1 (ETHE1), which results in the severe infantile metabolic disease Ethylmalonic Encephalopathy (EE), has been suggested to have a role in the β-oxidation of short-chain acyl-CoAs (20). Common symptoms of EE patients include chronic diarrhea, a delay in neural development, symmetric brain lesions, relapsing petechiae, lactic acidemia, and acrocyanosis of the extremities, which lead to death within the first decade of life (21,

22). Biochemical alterations associated with the disease include elevated levels of C4 and

C5 plasma acylcarnitines, and elevated excretion of ethylmalonic acid (EMA) in the patient’s urine (20). Possible sources of elevated levels of EMA are mutations in the short-chain fatty acid cycle resulting in the carboxylation of butyryl-CoA or a mutation in isoleucine catabolism that causes a shift to the R-isoleucine catabolic pathway, which produces EMA from 2-ethylmalonic-semialdehyde (23, 24). However, the exact biochemical role of ETHE1 in any organism is currently unknown.

83 ETHE1 orthologs are present in a wide range of organisms, including archaebacteria, fungi, plants, and animals, suggesting that it serves a fundamental biochemical role. In order to gain insight into the biochemical role of ETHE1 in general and specifically what role the protein plays in plant growth and development, we analyzed the effect of a knockout mutation in the Arabidopsis homolog of ETHE1. In this paper we report the characterization of Arabidopsis ETHE1 and demonstrate that ETHE1 is essential for early endosperm and ultimately embryo development in plants. Seeds lacking ETHE1 exhibit endosperm defects as early as the 4 cell octant zygote stage followed by the arrest in embryo development by early heart stage. Our results provide evidence for an important role of the early endosperm in embryo development.

4.3 Materials and Methods:

4.3.1 Plant Material

The Arabidopsis thaliana ETHE1 loss of function mutant was obtained from the DuPont p2800 T-DNA tagged seed pool. Seeds were vernalized at 4˚C for 48 hours prior to being placed in a growth chamber for growth on commercial potting soil at 23˚C with a 16:8 hour light/dark cycle. Wild-type Wassilewskja (WS) was used as the control in the development and immunolocalization studies. Stress studies were performed on wild-type Arabidopsis Columbia plants grown hydroponically. Seeds were germinated on Rockwool (GrodanHP, Agro Dynamics Inc., East Brunswick, NJ, USA) suspended above Hoagland’s solution (25) and grown in a growth chamber maintained at 23˚C with 16 hour light, 8 hour dark cycles. Plants stressed with abscisic acid (ABA), were grown hydroponically for 7 days and then stressed for 3 hours with 10 μM ABA. Plants were grown hydroponically for 18 days prior to inducing stress for 24 hours with either 150 mM NaCl or 300 mM Mannitol.

4.3.2 Phylogenetic Analysis of β-lactamase Proteins

84 A phylogenetic tree was derived from multiple alignments of β-lactamase-fold containing proteins using Clustal W version 1.82. A neighbor-joining phylogenetic analysis was conducted with MEGA version 3.1 using the Poisson correction amino acid substitution model and the complete deletion gaps option (26). Bootstrap values from 500 replicates were calculated and are indicated at branch points on the neighbor-joining tree. The phylogenetic analysis included H. sapiens ETHE1 (NP_055112), M. musculus ETHE1 (NP_075643), Xenopus laevis ETHE1 (NP_001079404), Xenopus tropicalis ETHE1 (NP_001005706), Arabidopsis ETHE1 (NP_974018), O. sativa Os01g066200 (NP_001043807), Burkholderia phytofirmans (ZP_01508561), Stigmatella auranticaca ETHE1 (ZP_01459266), Myxococcus xanthus (YP_633997), Arabidopsis GLX2-1 (NP_973679), Arabidopsis GLX2-2 (NP_187696), Arabidopsis GLX2-4 (NP_849599), Arabidopsis GLX2-5 (NP_850166), H. sapiens GLX2 (CAA62483), M. musculus GLX2 (NP_077246), O. sativa GLX2 (AAL14249), Brassica juncea GLX2 (AA026580), S. cerevesia GLO2 (CAA71335), S. cerevesia GLO4 (CAA99230), O. sativa Os03g0332400 (BAF11930), O. sativa OsJ_010290 (EAZ26807), and Stentrophomonas maltophilia L1 β-lactamase (CAB63489).

4.3.3 Molecular Analysis of ETHE1

ETHE1 cDNA was reverse transcribed from Arabidopsis total bud RNA and PCR amplified using primers GLX2-3NdeI (Fig. 4.1) and a 3’poly (A) anchor primer, generating a NdeI site at the predicted amino terminus of the ETHE1 mature protein and a XhoI restriction site, respectively. The amplified DNA was digested with Nde1 and Xho1, cloned into the expression vector pET15b (Novagen), sequenced, and introduced into E. coli BL21-RIL cells for protein over-expression. Cell cultures of E. coli BL21- RIL cells containing the ETHE1 plasmid were grown and induced, and ETHE1 was purified as described previously (27). Protein purity was assessed by SDS-PAGE gels -1 -1 and quantified at Abs280 using a molar extinction coefficient of 10,240 M cm . This value was determined using the method of Gill and von Hippel (28). The enzymatic activity of ETHE1 with S-D-lactoylglutathione (Sigma) was measured in 5 minute assays using 80 μM substrate and varying concentrations of pure enzyme as-isolated or after

85 addition of two equivalents of supplemented iron as previously described (29). For antibody production purified ETHE1 (280 mg) was resuspended in 1 ml PBS and mixed with either Freund’s complete or incomplete adjuvant (Sigma) and used to raise polyclonal antibody in rabbits using standard procedures. ETHE1 RNA levels were determined by RT-PCR on total RNA (1 μg) isolated from cotyledon, root, stem, leaf, silique, and bud, and seedlings subjected to abiotic stresses. ETHE1 cDNA was synthesized using a Thermoscript RT-PCR kit, followed by 25 cycles of PCR using the primer pair 2-3-5’ and 2-3-2. ACT8 was used as a control (23). A T-DNA insertion in ETHE1 was identified by performing PCR on pooled genomic DNA isolated from a population of Arabidopsis T-DNA insertion lines as previously described (30, 31). Two rounds of PCR were conducted with combinations of left border (LB) and gene specific primer pairs. Amplification products were verified through DNA sequencing. T1 seeds from pool p2800 containing the ETHE1 T-DNA line were obtained from the ARBC. Genomic DNA was isolated and screened by PCR to identify plants containing the ETHE1 T-DNA insertion. Genotyping was performed by PCR using a combination of T-DNA insertion and wild-type ETHE1 primer pairs. Complementation studies were conducted with a 3400 bp BamH1 and EcoR1 genomic DNA fragment containing ETHE1 that was amplified from BAC clone T3F20 using primers 448 and 383. The fragment was cloned into pCAMBIA-1390 (www.cambia.org), sequenced, and then transformed into GV3101/PMP90 Agrobacterium cells. A segregating population of ETHE1(+/-) plants was transformed (32). Transformants were selected using hygromycin (25mg/L) and confirmed by PCR screening using the pCAMBIA-1390 vector specific primer 557 (CGCAAGACCCTTCCTCTA) and gene-specific primer 2-3-2. Genotyping of hygromycin-resistant plants was performed by PCR using primers specific to the pCAMBIA vector, the T-DNA insertion, and the wild-type ETHE1 locus. Primer pair LB2 and 2-3-2 was used to detect the presence of the T-DNA insert. The complementation construct was identified with primer pair 557/2-3-2. The presence of the wild-type genomic locus was identified with primer pair 609 and 618 (AGTAAGTTGTGTTGTGTCACAC), which does not amplify the complementation

86 construct. The ability of the complementation construct to rescue the ethe1(-/-) phenotype was determined through the analysis of 40 mature siliques of ethe1(-/-) plants.

4.3.4 Microscopy

Seed development in siliques from ETHE1(+/-) plants was initially examined by whole silique analysis. Mature siliques (84) from ETHE1+/- plants and wild-type plants (20) were treated with a 4% (w/v) sodium hydroxide solution for 16 hours at room temperature, and the numbers of aborted and healthy seeds were determined by examination under a dissecting microscope. Embryo and endosperm development were analyzed in siliques from 40 ETHE1(+/-) plants at various stages of development using Laser Scanning Confocal Microscopy (LSCM) essentially as described previously (33). The samples were cleared, and individual seeds were dissected in methyl salicylate and viewed using an Olympus IX-81 fluorescence deconvolution microscope system. Data were analyzed with Image Pro Plus (Media Cybernetics) and organized with Photoshop. SEM was performed on staged flowers/siliques. Plant material was fixed overnight at room temperature in 2.8% (v/v) glutaraldehyde in 0.1 M HEPES buffer ,pH 7.2, and 0.02% (v/v) Triton X-100). Material was rinsed twice (15 minutes each) in 0.1

M HEPES, pH 7.2, then post-fixed in 1% (w/v) aqueous OsO4 overnight. Following post-fixation, the material was rinsed in 0.1 M HEPES, pH 7.2. Individual gynoecia were dissected open in drops of buffer to expose the ovules and placed into a critical point drier specimen basket submerged in buffer. Material in the basket was dehydrated through a graded ethanol series (10% increments, one hour each) with 3 changes of 100%

ethanol and critical point dried with a Balzers CPD 020 using CO2 as the transitional fluid. Dried specimens were mounted, sputter-coated with gold and examined with a Hitachi S-570 SEM operating at 15 kV. Digital images were captured using an Oxford Link ISIS microanalysis system. Immunolocalization studies to examine the distribution of ETHE1 were conducted on flowers and siliques 0.4 cm-0.8 cm in length. Samples were fixed in FAA

87 under vacuum overnight, rinsed in ddH2O, and dehydrated in a graded ethanol series (1 hour in each 10% increment). At 95% ethanol, samples were infiltrated with tertiary butanol and embedded in paraffin. Blocks were sectioned at 12 microns on a microtome and adhered to poly-lysine coated slides. Slides were rinsed in two changes of 1X PBS for 10 minutes each, followed by an hour of equilibration buffer (1X PBS, 1% Triton x- 100, 1 mM EDTA). Slides were blocked in blocking buffer (1X PBS, 5% BSA, 1 mM EDTA, 0.1% Tween-20) with shaking at room temperature for 1 hour, followed by incubation with anti-ETHE1 (1:1000 dilution) in blocking buffer overnight at 4˚C. Slides were washed eight times for 20 minutes each in wash buffer (1X PBS, 0.1% Tween-20, 1 mM EDTA), and the anti-ETHE1 antibody was detected with alkaline-phosphatase goat- anti-rabbit antibody (1:1000). Samples were observed and analyzed as described above.

4.4 Results

4.4.1 Molecular Analysis of ETHE1 The Arabidopsis gene At1g53580 (Fig 4.1A) had previously been predicted to encode a glyoxalase II isozyme (34). However, after further analysis we show here that it represents Arabidopsis ETHE1. Analysis of a cDNA (NM_202289) for At1g53580 revealed that it has the potential to encode a 294 amino acid protein that belongs to the β- lactamase family of proteins. A pairwise protein alignment between At1g53580 and the Arabidopsis cytoplasmic glyoxalase II (GLX2) isozyme (AtGLX2-2) revealed only 13% identity, while a pairwise alignment between At1g53580 and human ETHE1 revealed 54% identity, indicating that At1g53580 is likely not a GLX2 enzyme, but rather is the Arabidopsis ETHE1 protein (Fig 4.1B). GLX2 and ETHE1 both belong to the metallo-β-lactamase family of proteins, which are defined by a common αβ/βα fold and a conserved metal binding motif, T-H- X-H-X-D. A-type flavoproteins, and zinc phosphodiesterases also belong to the

88

Figure 4.1: Molecular Characterization of ETHE1. (A). Map of the ETHE1 locus. Exons are labeled and shown in white boxes. The positions and directions of primers used in this study are shown as horizontal arrows. A T-DNA insert is shown as an inverted triangle. (B) Alignment of ETHE1 with select β-lactamase fold proteins. Protein sequences were aligned using CLUSTAL W (1.83). Identical residues are highlighted with black and similar residues with gray highlight. Residues labeled with an asterisk are conserved in ETHE1-like enzymes. Residues labeled with an inverted triangle are mutated in patients with EE. Lower case residues are metal binding ligands. (C) Phylogenic Analysis of

89 ETHE1. The tree was derived from multiple β-lactamase fold proteins, which are aligned with Clustal W and followed by a neighbor joining phylogenetic analysis conducted with MEGA version 3.1. Bootstrap values are shown at branch points. See methods section for the accession numbers of sequences used for this analysis.

90 metallo-β-lactamase family. ETHE1 enzymes are most similar to the GLX2 family (35). There are, however, several biologically and structurally-distinguishing features of ETHE1 that set it apart from the glyoxalase II family. VAST alignments between glyoxalase II enzymes and ETHE1 show that ETHE1 is missing five out of the six highly conserved GLX2 substrate hydrogen-bonding residues (35, 36). In GLX2, Arg-249 and Lys-252 hydrogen bond to the glycine portion of glutathione, and in ETHE1 the residues are replaced by Met275 and Leu278. Residues Tyr145 and Lys143 of GLX2 hydrogen bond with the glutamate portion of glutathione. The corresponding amino acid residues in ETHE1 are replaced with Phe215 and Arg212 (35,36). Lastly, GLX2 residues Lys143 and Tyr175 interact with the cysteine portion of glutathione (35,36). ETHE1 retains the conserved tyrosine; however, Lys143 is replaced by Arg162. Finally, comparison of the ETHE1 and GLX2-5 crystal structures suggested that the substrate binding pocket of ETHE1 is too small to accommodate SLG, the glyoxalase II substrate (29, 35). These results suggest that ETHE1 does not utilize SLG as a substrate. In order to test this prediction, Arabidopsis ETHE1 was over-expressed, purified, and its activity with SLG was tested. As predicted, ETHE1 exhibited no enzymatic activity towards SLG (data not shown), confirming that it is not a glyoxalase II enzyme. In addition, glyoxalase II enzymes function as monomers; however, gel filtration studies on ETHE1 showed that it exists as a dimer (data not shown). Taken together, these results confirm that At1g53580 is not a glyoxalase II enzyme but rather is the Arabidopsis ETHE1 homolog. A Blast-P search revealed that ETHE1-like proteins are present in most organisms including animals, plants, fungi, and bacteria. Interestingly, an ETHE1-like protein does not appear to be present in yeast. A neighbor-joining tree was generated based on a Clustal W alignment of select metallo-β-lactamase proteins including ETHE1-like and GLX2 proteins. Analysis of the phylogenic tree, rooted in the common ancestor β- lactamase family protein, demonstrates that β-lactamases diverged through ancient duplication events into two separate lineages: ETHE-1 like and GLX2 proteins (Fig 4.1C). As expected, ETHE1 grouped together with human ETHE1 and predicted ETHE1 proteins from mouse, frog, fish, rice, broccoli, and bacteria. Likewise, the Arabidopsis GLX2 enzymes grouped together with GLX2 proteins from human, mouse, rice, and

91 yeast (Fig 4.1C). In addition to the conserved metal-binding residues, all ETHE1 proteins share a number of conserved residues, including R163, C161, Y38, L185, and T136 (Fig 4.1B). Mutations (R163Q, C161Y, Y38C, L185R and T136A) in these residues have been associated with EE in humans, suggesting that they may be required for the catalytic activity of ETHE1 (20). We have therefore renamed GLX2-3 as ETHE1.

4.4.2 ETHE1 is Required for Early Seed Development

In order to gain insights into the role of ETHE1 in general and plants in particular, an Arabidopsis line containing a T-DNA insertion in the gene was isolated and characterized. PCR screening of a population of T-DNA insertion mutants resulted in the identification of a line that contained a T-DNA insertion in exon four of ETHE1 (Fig 4.1A). In a segregation analysis on the progeny of ETHE1(+/-) plants, 312 (32%) were homozygous for the genomic wild-type locus and 673 (68%) were heterozygous for the ETHE1 T-DNA insertion. No plants homozygous for the T-DNA insertion were identified. The 1:2:0 segregation ratio suggested that ETHE1 is an essential gene and that the mutation may result in embryo lethality. Seed formation was then investigated in mature siliques to determine if in fact the ETHE1 T-DNA insertion is associated with alterations in seed development. A total of 107 siliques from 51 self-pollinated ETHE1(+/-) plants were examined and compared to those of wild-type plants. Approximately 4% (27/638) of the seeds contained in wild- type siliques appeared aborted, defined as either small masses or empty spots (Fig 4.2A). In contrast, siliques of ETHE1(+/-) plants contained 24.8% (752/3353) aborted seeds, which appeared as shrunken, shriveled masses, suggesting that the mutation does in fact affect seed development.

Analysis of seed development in immature siliques of ETHE1(+/-) plants revealed several phenotypic differences between mutant and wild-type seeds. The most obvious alteration is that ethe1(-/-) seeds are dramatically smaller in size than their wild-type counterparts (Fig 4.2B). Differences in seed size were first detected in siliques 7-8 mm in length, which is equivalent to the zygote/globular stages in wild-type embryo

92

Figure 4.2: Inactivation of ETHE1 disrupts seed development. (A) Mature siliques of wild-type, ETHE1(+/-), and ethe1(-/-) complementation plants observed by light microscopy. Aborted seeds are indicated by arrows. Size bar =1.5mm. (B) ETHE1(+/-) developing siliques observed by light microscopy. Seeds homozygous for the ethe1(-/-) mutation are indicated with an *. (C-E) SEM of developing seeds in ETHE1(+/-) siliques. Panels (C and D) show a comparison between the outer integument in post globular-staged seeds of wild-type and the corresponding ethe1(-/-) seeds. Higher magnification image of the outer integument of an earlier staged wild-type seed is shown in panel (E). Size bar = 10 μm.

93 development. The reduced size of ethe1(-/-) seeds relatively early in development suggested that the mutation may affect early endosperm development, which has been shown to be coupled to cell elongation and development of the seed coat (37). Consistent with this hypothesis, analysis of developing seeds using Scanning Electron Microscopy (SEM) revealed that no differences between seeds in the siliques of ETHE1(+/-) plants at early stages in seed development (data not shown). However at later stages in seed development, the cells of the outer integument in ethe1(-/-) seeds appeared smoother and less well defined than those of wild-type and heterozygous seeds in the same silique (Fig 4.2C and 4.2D). This is similar to the appearance of the integument during early stages of development (Fig 4.2C-E), suggesting that seed coat development arrests early in development.

Embryo and endosperm development was then investigated in siliques of ETHE1(+/-) plants using laser scanning confocal microscopy to more specifically determine the nature of the defect. Although differences in seed size were already apparent, no alterations in embryo development were observed in the seeds of ETHE1(+/-) siliques early in development, including at the zygote and early globular stages (compare Fig 4.3A and 4.3D). The first clear developmental difference in embryo development was observed when wild-type embryos were at early heart stage. When most seeds were at heart stage, 23% (129 of 560) of the seeds in ETHE1(+/-) siliques, all of which were small in size, remained at the 32 cell globular stage (compare Fig 4.3B and 4.3E). By the time most of the seeds had progressed to late torpedo and cotyledon stages, ethe1(-/-) seeds in the same silique had only developed to early heart stage (compare Fig 4.3C and 4.3F). With the exception of the delayed/arrested development, no dramatic cellular abnormalities were detected in ethe1(-/-) embryos during early and late globular stages (Fig 4.3G-4.3H). However, at heart stage many ethe1(-/-) embryos contained elongated cells (Fig 4.3I). As discussed above, the overall size of ethe1(-/-) seeds at the globular stage of development is much smaller than wild-type seeds at the same developmental stage (compare Fig 4.4A-D,B-E). Several studies have shown that the endosperm plays a direct role in early seed size suggesting that the ethe1 mutation may cause defects in endosperm development (2, 6). A detailed analysis of endosperm development revealed

94

Figure 4.3: Embryo Development in ETHE1(+/-) siliques. Confocal Light Microscopy images of Feulgan stained segregating ETHE1(+/-) seeds. Wild-type seeds are shown at embryonic stages globular (A), heart (B) and walking stick (C). ethe1(-/-) seeds from siliques with wild-type embryos shown in (A-C). The ethe1(-/-) embryos developed to early globular (D), globular (E), and early heart (F) stages. High magnification images ethe1(-/-) embryos at early globular (G), globular (H), and early heart (I) stages. Arrows denote elongated cells. Size bar = 20μm.

95 that this is in fact the case. Specifically, ethe1(-/-) seeds have fewer endosperm nuclei than wild-type seeds at most stages of development. These differences are most pronounced in the PEN region (compare Fig 4.4B and 4.4H, and 4.4E and 4.4K), which accounts for the majority of the seed’s size. In a developing wild-type seed at the 8-cell octant stage, endosperm nuclei in the PEN region have undergone 7 rounds of division resulting in approximately 100 free endosperm nuclei (Fig 4.4A-B). Cellularization of the endosperm nuclei begins shortly after this stage progressing in a wave from the micropylar region through the CZE by the end of heart stage (38). This, however, is not the case in ethe1(-/-) seeds. In contrast to the ca. 100 free endosperm nuclei in the PEN region of wild-type seeds at the globular stage, only 10-15 endosperm nuclei were observed in globular ethe1(-/-) seeds (Fig 4.4D-E). By the time ethe1(-/-) seeds approached heart stage, only 40-50 endosperm nuclei were observed in comparison to the 200 nuclei seen in the wild-type PEN region (compare Fig 4.4G-H and Fig 4.4J-K). In addition, unlike wild-type seeds, the endosperm of ethe1(-/-) seeds did not undergo cellularization (compare Fig 4.4H and Fig 4.4K).

Alterations in the CZE were also noted in the mutant beginning at the 4-cell zygote stage (compare Fig 4.4C and Fig 4.4I with Fig 4.4F and Fig 4.4L). The CZE in ethe1(-/-) seeds (Fig 4.4F and 4.4L) appeared underdeveloped, containing very little cytoplasm surrounding the chalazal cyst and few, if any, chalazal nodules compared to wild-type seeds at the same developmental stage. Therefore, defects are observed in the endosperm as early as the 4-cell zygote stage. This suggests that the primary effect of the ethe1 mutation is disruption of endosperm development, which in turn, slows and ultimately causes the arrest of embryo development.

The tight genetic linkage between the ETHE1 T-DNA insertion and the observed embryo lethality indicated that ETHE1 is an essential gene that is required for early embryo and endosperm development. A complementation study was performed to verify this conclusion. A 3.4 Kbp genomic DNA fragment containing ETHE1 and 1248 and 141 bp of 5’ and 3’ flanking DNA, respectively, was transformed into a segregating population of ETHE1(-/+) plants. Three independent lines that were homozygous for the ethe1 T-DNA mutation and contained the ETHE1 complementation construct were identified. To further verify that complete complementation had occurred, individual

96

97

Figure 4.4: Endosperm development is abnormal in ethe1(-/-) seeds. Confocal microscopy images of Feulgan stained segregating ETHE1(+/-) seeds. Panels (A- C) show endosperm development in globular-staged wild-type seed (A) in the PEN (B) and CHZ (C) regions. Similar regions of a globular-staged ethe1(-/-) seed (D) are shown in (E-F). Panels (G-I) show endosperm development of a wild-type heart-staged seed (G) in the PEN (H) and CHZ (I) regions. The same regions are shown in a heart-staged ethe1(-/-) seed in panels (J-L). Size bar = 20μm. e = embryo; chz = chalazal region.

98 siliques from ethe1(-/-) plants, which contained the complementation construct, were analyzed for the presence of abnormal/aborted seeds. A total of 59 (3.7%) out of 1341 seeds examined in siliques from the complementation lines were aborted, which is similar to the number of aborted seed typically observed in wild-type siliques (Fig 4.2A). These results confirm that the ethe1 mutation is responsible for the seed defects we observed and suggests that the complementation construct is able to restore normal ETHE1 function.

4.4.3 ETHE1 Expression and Localization

Results from our T-DNA knockout studies indicate that ETHE1 is an essential gene involved in early endosperm development. In order to determine if ETHE1 plays a role throughout plant development, we investigated its expression patterns. ETHE1 is present on both the Affymetrix AG: 8k array and the ATH1: 22k array Genechip®. Therefore there is a considerable amount of information concerning ETHE1 expression patterns in the public databanks (39). Analysis of the available data indicated that ETHE1 RNA is present in every tissue examined with the lowest levels detected in reproductive tissues. The distribution of ETHE1 transcripts in various tissues was determined using RT-PCR in order to confirm these results. The ACTIN8 (ACT8) gene was used to standardize the reactions due to its relatively constant expression (40). ETHE1 transcript levels were detected in all tissues examined. Consistent with the microarray data, lower levels were seen in siliques and flower buds (Fig 4.5A). Further analysis of the microarray data revealed that ETHE1 transcript levels are elevated during senescence, osmotic stresses (e.g. mannitol and NaCl), and biotic stresses (e.g. infection with B. cinerea and P. syringae). Significant changes (greater than 2 log) were not observed under other conditions. To verify these results, wild-type Arabidopsis Columbia plants were grown hydroponically, subjected to several different stresses, and ETHE1 RNA levels were determined. Three different abiotic stresses (mannitol, NaCl, and abscisic acid) resulted in elevated ETHE1 transcript levels (Fig 4.5B). Treatment with ABA and NaCl resulted in approximately two-fold increases in ETHE1 transcript

99

Figure 4.5: ETHE1 Expression Patterns. (A) RT-PCR analysis of ETHE1 transcript patterns in different tissues. (B) RT-PCR analysis of ETHE1 transcript levels of hydroponically grown Wt/Ws plants in response to environmental stresses NaCl, Mannitol, and ABA. ACT 8 was used as an internal control.

100 levels, while treatment with mannitol resulted in an increase of approximately 2.5 fold. Therefore, ETHE1 transcripts are present throughout the plant and their levels increase during stress. We next investigated the distribution of ETHE1 during flower and seed development in order to investigate why the ETHE1 mutation appears to have a more pronounced effect on endosperm development. Immunolocalization studies on flower buds, flowers, and developing siliques revealed ETHE1 signals above background levels in all tissues examined (Fig 4.6). Consistent with the results from the microarray data and our RT-PCR experiments, reproductive organs in general do not contain high levels of ETHE1. However, strong ETHE1 signal is present in the tapetal cells, the nutrient cells of the anther (Fig 4.6A-D). In early stages of seed growth, including from the zygote to the globular embryo stages, ETHE1 signal is present in the developing embryo and throughout the endosperm (Fig 4.6E-I). Closer examination revealed that early in development ETHE1 signal is observed primarily in the nuclei of the embryo proper and the PEN region of young, zygote-staged, seeds (Fig 4.6F and 4.6G). As development progresses into the globular stage, ETHE1 signal becomes more pronounced and is the strongest in the PEN and CZE tissues (Fig 4.6M-O). At this stage in embryo development, the MCE has begun to cellularize, while the PEN and CZE are still rapidly dividing. The strongest ETHE1 signal is observed in PEN endosperm nuclei and at the base of the chalazal cyst, consistent with the theory that ETHE1 plays an important role in early endosperm development (Fig 4.6N-O).

4.5 DISCUSSION

In humans, mutations in ETHE1 are responsible for ethylmalonic encephalopathy (20). The symptoms of this disease are well characterized, but a role for ETHE1 has not yet been determined (20-22, 41). Arabidopsis ETHE1 has been characterized to provide further insight into the function(s) of ETHE1. The Arabidopsis protein is 54% identical to human ETHE1 and contains all the ETHE1 family-conserved residues. ETHE1 is ubiquitously expressed and is an essential gene in plants. Analysis of an ETHE1 loss of

101

102 Figure 4.6: Immunolocalization of ETHE1 in Wild-type Buds. Sections of wild-type buds (A,B), anthers (C, D) or seeds (E-O) were prepared and treated with pre-immune serum (A, C, E, H, J-L) or the ETHE1 antibody (B, D, F-G, I, M- O). ETHE1 cross reactive material is purple, non specific signals from the detection system are brown. ETHE1 is present in the tapetal cells of the anther (B, D), young 2 cell embryo seeds, in particular, surrounding the nuclei of the young embryo (F) and PEN nuclei (G). ETHE1 is also present in globular-staged seeds (I), in particular, in the globular embryo (M), the PEN nuclei (N) and the chalazal cyst (O). The embryo and chalazal cyst are labeled e and chz respectively. Size bar = 50μm.

103 function mutant demonstrated that ETHE1 is essential for early endosperm development and seed viability in plants.

4.5.1 ETHE1 is Essential for Early Endosperm Development

Analysis of a T-DNA knockout mutation of Arabidopsis demonstrated that ETHE1 is essential for endosperm and embryo development. Approximately 25% of the seeds in siliques of plants heterozygous for the ETHE1 T-DNA mutation exhibited defects early in development. CLSM studies of seed development in ETHE1(+/-) siliques demonstrated that ethe1(-/-) embryos exhibit developmental delays beginning at the globular stage and arrest once heart stage is reached (Fig 4.3D-F). With the exception of the developmental delay and the presence of small numbers of elongated cells in some heart-stage embryos, alterations were not observed in ethe1(-/-) embryos (Fig 4.3G-I). Developmental delays were also noted in both the PEN and CZE of ethe1(-/-) seeds beginning soon after fertilization with most endosperm nuclei arresting after only a few divisions (Fig 4.4E-F, 4.4K-L). Typically fewer than 50 endosperm nuclei were observed in the PEN region of ethe1(-/-) seeds along with only a few chalazal nodes (Fig 4.4F and 4.4L). In addition, cellularization of the endosperm never occurred in ethe1(-/-) seeds. The observation that the earliest and most dramatic effects of the mutation are on the endosperm suggest that the mutation primarily affects endosperm development. Consistent with this is our observation that while ETHE1 is present in both embryo and endosperm tissue during very early stages of seed development (Fig 4.6G-H), the strongest ETHE1 signals are observed in endosperm nuclei of the PEN region and the chalazal cyst as development progresses towards the globular stage (Fig 4.6M-O). Therefore, we believe that the primary defect of ethe1(-/-) seeds is in the endosperm and that defects in early endosperm development ultimately result in embryo arrest by heart stage. However, it is also possible that ETHE1 is also critical for embryo development and embryo arrest results from the lack of ETHE1 activity and not from the underdeveloped endosperm. Several mutants have been isolated showing defects in late endosperm development, including the absence of endosperm cellularization. The majority of these

104 mutations are in genes required for nuclear or cellular division and affect structural elements, such as the pilz, titan, knolle and hinkel mutants. All of these mutations result in defects in both the embryo and endosperm (14, 15, 42, 43). Several mutants, including spatzle, fis1, and fis2 have been identified where no apparent cytokinesis defects are apparent in the developing embryo, however the endosperm tissues fails to cellularize (5, 18, 44, 45). Interestingly, even though endosperm cellularization is blocked in these mutants, the resulting seed is comparable in size to the wild-type and is able to develop into a fully functional plant suggesting that late endosperm development is not required for embryo viability (5). Therefore, the absence of endosperm cellularization is likely not the cause of the embryo arrest in ethe1(-/-) seeds. Analysis on the haiku2 mutant recently showed that defects in endosperm development can affect integument elongation and seed size (2, 17, 18). The reduced size and delayed/restricted development of the integument associated with ethe1(-/-) seeds is consistent with the theory that early endosperm development plays an important role in seed size and integument development. Interestingly, haiku2(-/-) seeds while smaller than wild-type are still viable (2). Results from our studies show that the endosperm of ethe1(- /-) seeds arrests at a much earlier stage of development than do haiku2(-/-) seeds, which exhibit a precocious arrest of endosperm cellularization, reduced proliferation of the endosperm and a reduction of embryo growth at the torpedo stage (2). Taken together, these results suggest that a critical number of syncytial mitoses are required to support the development of a viable embryo. Consistent with this hypothesis are the results of an endosperm ablation study that expressed an endosperm specific diphtheria toxin (6). In this study, degradation of the endosperm shortly after the second round of nuclear divisions resulted in small seeds in which the embryo arrested at heart stage (6). Prior to arrest, the embryos in the KS221>>DTA seeds exhibited a number of abnormalities. In particular, preglobular embryos displayed swollen protodermal cells and/or altered division planes; however, they did not show signs of cell death (6). Defects in ethe1(-/-) seeds are not as dramatic as those observed in the KS221>>DTA seeds. In particular, the endosperm in ethe1(-/-) seeds undergoes several rounds of nuclear division, and while embryos arrest at heart stage, they do not exhibit distorted division patterns. Taken together these results suggest

105 that the endosperm may play several roles in embryo development. Very early embryo development is closely tied with endosperm development with the first several syncytial divisions being critical to establish the proper division patterns in the embryo. Later, a threshold level of endosperm nuclei appears to be necessary to support the growth and development of the endosperm beyond the heart stage. Our results are consistent with the predicted role of the CZE in functioning in the uptake, processing, and transfer of metabolites to the endosperm and suggest that the chalazal cyst is necessary for embryo development beyond heart stage. Once the embryo has developed beyond heart stage the endosperm appears to be dispensable.

4.5.2 Potential Role(s) of ETHE1

ETHE1 is most closely related to the GLX2 family of enzymes. GLX2, along with GLXI, make up the glyoxalase system, which is ubiquitous in nature and has been studied in a number of organisms. The exact role of this system is unknown; however, it has been proposed that it is responsible for the glutathione-dependent detoxification of 2- oxoaldehydes, toxic by-products of carbohydrate and lipid metabolism (46). GLX2 catalyzes the hydrolysis of the thioester bond of S-D-lactoglutathione to produce glutathione and D-lactic acid (47). ETHE1 lacks several highly conserved glutathione binding residues found in GLX2 and does not utilize SLG or other glutathione-based thioesters. Therefore, it is not involved in a glutathione-based detoxification event. Recombinant ETHE1 does however exhibit weak esterase activity (Chapter 3). Based on these results and the nature of the ETHE1 active site, we predict that ETHE1 may hydrolyze a short chain ester. In humans, mutations in ETHE1 have been shown to cause EE; one symptom of this disease is the excretion of high levels of EMA (20). Known pathways that can lead to the production of EMA are limited and include the carboxylation of butyryl-CoA, which could result from a defect in the short-chain fatty acid β-oxidation pathway, and the R-pathway of isoleucine catabolism (20, 23, 24). Several diseases, including glutaric academia type 2 and Jamaican vomiting sickness are caused by defects in short-chain

106 acyl-CoA dehydrogenase activity that result in an accumulation of butyryl-CoA, which is then metabolized to EMA (48). Therefore, it is possible that ETHE1 participates in the short chain β-oxidation pathways. Consistent with this possibility, it was recently shown that the accumulation of fatty acids is necessary for the growing embryo (49). There are six acyl-CoA oxidase (ACX) genes in Arabidopsis, which catalyze the conversion of acyl-CoA to 2-trans- enoyl-CoA in the first step of the β-oxidation pathway (19, 49-52). It was recently shown that, unlike medium and long-chain fatty acid acyl-CoA oxidation, short-chain fatty acid acyl-CoA oxidase activity is essential for early stages of embryo growth (49). Through our loss of function studies, we have shown that ETHE1 is essential for early endosperm development in Arabidopsis. It has been speculated that the endosperm in some dicots, such as Arabidopsis, serves to protect the embryo from physical and osmotic stress, as well as being involved in the transfer of nutrients to the developing embryo (2, 3). Unlike monocot species, the dicot Arabidopsis endosperm does not accumulate starch for nutrient storage; however, it has been shown to accumulate lipids, in particular fatty acids (53). In addition, it has been suggested that the chalazal cyst serves in nutrient transport and/or as a temporary nutrient storage site (3, 3, 54). We have shown that the loss of ETHE1 disrupts CZE development (Fig 4.5F, 4.5L). An early arrest of the CZE could block nutrient storage and transfer to the developing embryo, depleting the necessary energy reserves for continued embryo development. The possibility that ETHE1 plays a role in short-chain fatty acid β-oxidation also is consistent with the increased expression during senescence as seen in the microarray data. However, the increased expression of ETHE1 during stress is not as clear in this regard. A second proposed role of human ETHE1, also called HSCO, is more consistent with the increased expression of ETHE1 observed upon abiotic and biotic stresses. It has been suggested that human ETHE1 acts as a nuclear-cytoplasmic shuttle protein that regulates the nuclear localization of NF-κB (55). Human ETHE1 was further shown to associate with histone deacetylase 1 (HDAc1) and facilitate the HDAc1-dependent deacetylation of p53 (56). Given the wide phylogenetic distribution of ETHE1, this role in inhibiting p53-dependent apoptosis is not expected to be a common feature of the protein. However, it is interesting to note that while our immunolocalization experiments

107 showed that ETHE1 is typically found in the cytoplasm, strong ETHE1 signal was found in the endosperm nuclei of the PEN region (Fig 4.6N). In summary, we have shown that ETHE1 is essential for early aspects of endosperm development. To our knowledge, ethe1(-/-) is one of the earliest characterized mutants that affect endosperm development. Defects in endosperm development in the ethe1(-/-) seed block embryo development beginning at heart stage, clearly demonstrating the link between early embryo and endosperm development. Our results suggest that ETHE1 may participate directly or indirectly in the β-oxidation of fatty acids. However, further experiments are necessary to more specifically define the role of ETHE1 in general and in plants in particular. Studies to investigate the physiological role(s) of ETHE1 in plants are complicated by the fact that the knockout mutation is lethal. Therefore, we are currently in the process of generating ETHE1-inducible RNAi and over-expression lines to further investigate the role of this essential protein.

108 4.6 Acknowledgements Thanks go to Dr. Richard Moore, Dr. Lara Strittmatter, Ling Jiang, and Xiaohui Yang for helpful discussions and technical advice, Dr. Richard E. Edelmann and Matthew Duley for support with the microscopes, Sue Gibson for the pooled DNA’s, and the ABRC for the seed stocks. This work was supported in part by a grant from the National Science Foundation (MCB9817083) to C.A.M.

109 4.7 References

1. Costa, L. M., Gutierrez-Marcos, J. F., and Dickinson, H. G. (2004). More than a Yolk: the Short Life and Complex Times of the Plant Endosperm. Trends Plant Sci. 9, 507- 514

2. Garcia, D., Saingery, V., Chambrier, P., Mayer, U., Jurgens, G., and Berger, F. (2003). Arabidopsis haiku Mutants Reveal New Controls of Seed Size by Endosperm. Plant Physiol. 131, 1661-1670

3. Nguyen, H., Brown, R. C., and Lemmon, B. E. (2000). The Specialized Chalazal Endosperm in Arabidopsis thaliana and Lepidium virginicum (Brassicaceae). Protoplasma 212, 99-110

4. Olsen, O. A. (2004). Nuclear Endosperm Development in Cereals and Arabidopsis thaliana. Plant Cell 16 Suppl, S214-27

5. Sorensen, M. B., Mayer, U., Lukowitz, W., Robert, H., Chambrier, P., Jurgens, G., Somerville, C., Lepiniec, L., and Berger, F. (2002). Cellularisation in the Endosperm of Arabidopsis thaliana is Coupled to Mitosis and Shares Multiple Components with Cytokinesis. Development 129, 5567-5576

6. Weijers, D., Van Hamburg, J. P., Van Rijn, E., Hooykaas, P. J., and Offringa, R. (2003). Diphtheria Toxin-mediated Cell Ablation Reveals Interregional Communication During Arabidopsis Seed Development. Plant Physiol. 133, 1882- 1892

7. Brown, R. C., Lemmon, B. E., Nguyen, H., and Olsen, O. (1999). Development of Endosperm in Arabidopsis thaliana. Sex. Plant Reprod. 12, 32-42

8. Mansfield, S. G. (1994) Endosperm development. In Arabidopsis: An Atlas of Morphology and Development pp. 385-397. Edited by J. L. Bowman, Berlin Heidelberg New York Tokyo, Springer.

110 9. Lopes, M. A., and Larkins, B. A. (1993). Endosperm Origin, Development, and Function. Plant Cell 5, 1383-1399

10. Larkins, B. A., Dilkes, B. P., Dante, R. A., Coelho, C. M., Woo, Y. M., and Liu, Y. (2001). Inverstigating the Hows and Whys of DNA Endoreduplication. J. Exp. Bot. 52, 183-192

11. Pacini, E., Simoncioli, C., and Cresti, M. (1975). Ultrastructure of Necellus and Endosperm of Diplotaxis erucoides During Embryogenesis. Caryologia 28, 525-538

12. Berger, F. (2003). Endosperm: The Crossroad of Seed Development. Curr. Opin. Plant Biol. 6, 42-50

13. Assaad, F. F., Mayer, U., Wanner, G., and Jurgens, G. (1996). The KEULE Gene is Involved in Cytokinesis in Arabidopsis. Mol. Gen. Genet. 253, 267-277

14. Lukowitz, W., Mayer, U., and Jurgens, G. (1996). Cytokinesis in the Arabidopsis Embryo Involves the Syntaxin-rlated KNOLLE Gene Product. Cell 84, 61-71

15. Strompen, G., El Kasmi, F., Richter, S., Lukowitz, W., Assaad, F. F., Jurgens, G., and Mayer, U. (2002). The Arabidopsis HINKEL Gene Encodes a Kinesin-related Protein Involved in Cytokinesis and is Expressed in a Cell Cycle-dependent Manner. Curr. Biol. 12, 153-158

16. Waizenegger, I., Lukowitz, W., Assaad, F., Schwarz, H., Jurgens, G., and Mayer, U. (2000). The Arabidopsis KNOLLE and KEULE Genes Interact to Promote Vesicle Fusion During Cytokinesis. Curr. Biol. 10, 1371-1374

17. Garcia, D., Fitz Gerald, J. N., and Berger, F. (2005). Maternal Control of Integument Cell Elongation and Zygotic Control of Endosperm Growth are Coordinated to Determine Seed Size in Arabidopsis. Plant Cell 17, 52-60

18. Luo, M., Dennis, E. S., Berger, F., Peacock, W. J., and Chaudhury, A. (2005). MINISEED3 (MINI3), a WRKY Family Gene, and HAIKU2 (IKU2), a Leucine-rich

111 Repeat (LRR) KINASE Gene, are Regulators of Seed Size in Arabidopsis. Proc. Natl. Acad. Sci. U. S. A. 102, 17531-17536

19. Rylott, E. L., Rogers, C. A., Gilday, A. D., Edgell, T., Larson, T. R., and Graham, I. A. (2003). Arabidopsis Mutants in Short- and Medium-chain Acyl-CoA Oxidase Activities Accumulate Acyl-CoAs and Reveal that Fatty Acid Beta-oxidation is Essential for Embryo Development. J. Biol. Chem. 278, 21370-21377

20. Tiranti, V., D'Adamo, P., Briem, E. et al. (2004). Ethylmalonic Encephalopathy is Caused by Mutations in ETHE1, a Gene Encoding a Mitochondrial Matrix Protein. Am. J. Hum. Genet. 74, 239-252

21. Burlina, A., Zacchello, F., Dionisi-Vici, C., Bertini, E., Sabetta, G., Bennet, M. J., Hale, D. E., Schmidt-Sommerfeld, E., and Rinaldo, P. (1991). New Clinical Phenotype of Branched-chain Acyl-CoA Oxidation Defect. Lancet. 338, 1522-1523

22. Burlina, A. B., Dionisi-Vici, C., Bennett, M. J., Gibson, K. M., Servidei, S., Bertini, E., Hale, D. E., Schmidt-Sommerfeld, E., Sabetta, G., and Zacchello, F. (1994). A New Syndrome with Ethylmalonic Aciduria and Normal Fatty Acid Oxidation in Fibroblasts. J. Pediatr. 124, 79-86

23. Mamer, O. A., Tjoa, S. S., Scriver, C. R., and Klassen, G. A. (1976). Demonstration of a New Mammalian Isoleucine Catabolic Pathway Yielding an Rseries of Metabolites. Biochem. J. 160, 417-426

24. Nowaczyk, M. J., Lehotay, D. C., Platt, B. A., Fisher, L., Tan, R., Phillips, H., and Clarke, J. T. (1998). Ethylmalonic and Methylsuccinic Aciduria in Ethylmalonic Encephalopathy Arise from Abnormal Isoleucing Metabolism. Metabolism 47, 836- 839

25. Cowgill, U. M., and Milazzo, D. P. (1989) The Culturing and Testing of Two Species of Duckweed. In Aquatic Toxicology and Hazard Assessment pp. 379-391. Edited by W. J. Adams, & G. A. Chapman, Philadelphia, PA, American Society for Testing and Materials.

112 26. Kumar, S., Tamura, K., and Nei, M. (2004). MEGA3: Integrated Software for Molecular Evolutionary Genetics Analysis and Sequence Alignment. Brief. Bioinform. 5, 150-163

27. Crowder, M. W., Maiti, M. K., Banovic, L., and Makaroff, C. A. (1997). Glyoxalase II from A. thaliana Requires Zn(II) for Catalytic Activity. FEBS Lett. 418, 351-354

28. Gill, S. C., and von Hippel, P. H. (1989). Calculation of Protein Extinction Coefficients from Amino Acid Sequence Data. Anal. Biochem. 182, 319-326

29. Marasinghe, G. P., Sander, I. M., Bennett, B., Periyannan, G., Yang, K. W., Makaroff, C. A., and Crowder, M. W. (2005). Structural Studies on a Mitochondrial Glyoxalase II. J. Biol. Chem. 280, 40668-40675

30. Krysan, P. J., Young, J. C., Tax, F., and Sussman, M. R. (1996). Identification of Transferred DNA Insertions within Arabidopsis Genes Involved in Signal Transduction and Ion Transport. Proc. Natl. Acad. Sci. U. S. A. 93, 8145-8150

31. Todd, J., Post-Beittenmiller, D., and Jaworski, J. G. (1999). KCS1 Encodes a Fatty Acid Elongase 3-ketoacyl-CoA Synthase Affecting Wax Biosynthesis in Arabidopsis thaliana. Plant J. 17, 119-130

32. Chung, M. H., Chen, M. K., and Pan, S. M. (2000). Floral Spray Transformation can Effeciently Generate Arabidopsis Transgenic Plants. Transgenic Res. 9, 471-476

33. Braselton, J. P., Wilkinson, M. J., and Clulow, S. A. (1996). Feulgen Staining of Intact Plant Tissues for Confocal Microscopy. Biotech. Histochem. 71, 84-87

34. Maiti, M. K., Krishnasamy, S., Owen, H. A., and Makaroff, C. A. (1997). Molecular Characterization of Glyoxalase II from Arabidopsis thaliana. Plant Mol. Biol. 35, 471-481

35. McCoy, J. G., Bingman, C. A., Bitto, E., Holdorf, M. M., Makaroff, C. A., and Phillips, G. N.,Jr. (2006). Structure of an ETHE1-like Protein from Arabidopsis thaliana. Acta Crystallogr. D Biol. Crystallogr. 62, 964-970

113 36. Cameron, A. D., Ridderstrom, M., Olin, B., and Mannervik, B. (1999). Crystal Structure of Human Glyoxalase II and its Complex with a Glutathione Thiolester Substrate Analogue. Structure 7, 1067-1078

37. Sundaresan, V. (2005). Control of Seed Size in Plants. Proc. Natl. Acad. Sci. U. S. A. 102, 17887-17888

38. Bowman, J. L. (1994) Embryogenesis. In Arabidopsis: An atlas of morphology pp. 351-361. Edited by J. L. Bowman, & S. G. Mansfield, New York, Springer-Verlag.

39. Zimmerman, P., Hirscho-Hoffman, M., Hennig, L., and Gruissem, W. (2005). Genevestigator. Arabidopsis Microarray Database and Analysis Toolbox. Trends Plant Sci.

40. An, Y. Q., McDowell, J. M., Huang, S., McKinney, E. C., Chambliss, S., and Meagher, R. B. (1996). Strong, Constitutive Expression of the Arabidopsis ACT2/ACT8 Actin Subclass in Vegetative Tissues. Plant J. 10, 107-121

41. Tiranti, V., Briem, E., Lamantea, E. et al. (2006). ETHE1 Mutations are Specific to Ethylmalonic Encephalopathy. J. Med. Genet. 43, 340-346

42. Liu, C. M., and Meinke, D. W. (1998). The Titan Mutants of Arabidopsis are Disrupted in Mitosis and Cell Cycle Control During Seed Development. Plant J. 16, 21-31

43. Mayer, U., Herzog, U., Berger, F., Inze, D., and Jurgens, G. (1999). Mutations in the PILZ Group Genes Disrupt the Cytoskeleton and Uncouple Cell Cycle Progression from Cell Division in Arabidopsis Embryo and Endosperm. Eur. J. Cell Biol. 78, 100-108

44. Luo, M., Bilodeau, P., Koltunow, A., Dennis, E. S., Peacock, W. J., and Chaudhury, A. M. (1999). Genes Controlling Fertilization-independent Seed Development in Arabidopsis thaliana. Proc. Natl. Acad. Sci. U. S. A. 96, 296-301

114 45. Ohad, N., Yadegari, R., Margossian, L., Hannon, M., Michaeli, D., Harada, J. J., Goldberg, R. B., and Fischer, R. L. (1999). Mutations in FIE, a WD Polycomb Group Gene, Allow Endosperm Development without Fertilization. Plant Cell 11, 407-416

46. Phillips, S. A., and Thornalley, P. J. (1993). The Formation of Methylglyoxal from Triose Phosphates. Investigation Using a Specific Assay for Methylglyoxal. Eur. J. Biochem. 212, 101-105

47. Thornalley, P. J. (1990). The Glyoxalase System: New Developments Towards Funcional Characterization of a Metabolic Pathway Fundamental to Biological Life. Biochem. J. 269, 1-11

48. Corydon, M. J., Gregersen, N., Lehnert, W. et al. (1996). Ethylmalonic Aciduria is Associated with an Amino Acid Variant of Short Chain Acyl-Coenzyme A Dehydrogenase. Pediatr. Res. 39, 1059-1066

49. Penfield, S., Helen, M., Pinfield-Wells et al. (2006) Storage Reserve Mobilisation and Seedling Establishment in Arabidopsis. In The Arabidopsis Book Rockville, MD, American Society of Plant Biologists.

50. Pinfield-Wells, H., Rylott, E. L., Gilday, A. D., Graham, S., Job, K., Larson, T. R., and Graham, I. A. (2005). Sucrose Rescues Seedling Establishment but not Germination of Arabidopsis Mutants Disrupted in Peroxisomal Fatty Acid Catabolism. Plant J. 43, 861-872

51. Eastmond, P. J., Hooks, M. A., Williams, D., Lange, P., Bechtold, N., Sarrobert, C., Nussaume, L., and Graham, I. A. (2000). Trapping of a Novel Medium- chain Acyl-CoA Oxidase which is Induced Transcriptionally During Arabidopsis Seed Germination. J. Biol. Chem. 275, 34375-34381

52. Adham, A. R., Zolman, B. K., Millius, A., and Bartel, B. (2005). Mutations in Arabidopsis Acyl-CoA Oxidase Genes Reveal Distinct and Overlapping Roles in Beta-oxidation. Plant J. 41, 859-874

115 53. Penfield, S., Rylott, E. L., Gilday, A. D., Graham, S., Larson, T. R., and Graham, I. A. (2004). Reserve Mobilization in the Arabidopsis Endosperm Fuels Hypocotyl Elongation in the Dark, is Independent of Abscisic Acid, and Requires Phosphoenolpyruvate Carboxykinase 1. Plant Cell 16, 2705-2718

54. Otegui, M. S., Capp, R., and Staehelin, L. A. (2002). Developing Seeds of Arabidopsis Store Diferent Minerals in Two Types of Vacuoles and in the Endoplasmic Reticulum. Plant Cell 14, 1311-1327

55. Higashitsuji, H., Higashitsuji, H., Nagao, T., Nonoguchi, K., Fujii, S., Itoh, K., and Fujita, J. (2002). A Novel Protein Overexpressed in Hepatoma Accelerates Export of NF-kappa B From the Nucleus and Inhibits p53-dependent Apoptosis. Cancer Cell 2, 335-346

56. Higashitsuji, H., Higashitsuji, H., Masuda, T., Liu, Y., Itoh, K., and Fujita, J. (2007). Enhanced Deacetylation of p53 by the Anti-apoptotic Protein HSCO in Association with Histone Deacetylase I. J. Biol. Chem. 282, 13716-13725

116 Chapter 5

Functional Studies on Arabidopsis Plants and Cell Cultures Expressing an Inducible ETHE1 RNAi

117

5.1 Summary

Ethylmalonic Encephalopathy Protein 1 (ETHE1) is essential for development in both humans and plants, yet the biochemical and physiological role(s) that ETHE1 plays is still unknown. Mutations in human ETHE1 have been directly linked Ethylmalonic Encephalopathy (EE), a disease that causes among other symptoms chronic diarrhea, delays in neural development, symmetric brain lesions, relapsing petechiae, lactic academia, and acrocyanosis of the hands and feet. Likewise, inactivation of Arabidopsis ETHE1 disrupts early endosperm development resulting in embryo lethality and seed abortion. In order to study the role of ETHE1 at other time points in plant development, we generated a chemically-inducible ETHE1-RNAi construct. In these studies we demonstrate that in addition to its critical role in seed development, ETHE1 also plays an essential role in germination and in suspension cell survival. These results provide further support for a potential role of ETHE1 in fatty acid β-oxidation

118 5.2 Introduction

Ethylmalonic encephalopathy (EE) is a rare autosomal recessive disorder found mainly in patients of Arabic and Medditerranean descent (1-3). Studies have shown that EE patients contain loss of function mutations in the gene Ethylmalonic Encephalopathy protein 1 (ETHE1) (3). Patients with EE are generally diagnosed shortly after birth and display a number of common symptoms, including chronic diarrhea, delays in neural development, symmetric brain lesions, relapsing petechiae, muscle hypotonia, and acrocyanosis of the extremities, which lead to death within the first decade of life (1, 2).

EE patients also exhibit a marked increase in lactic acid, carnitines, C4-6 acylglycines, and elevated urinary excretion of ethylmalonic acid (EMA) as well as a reduction of mitochondrial respiratory complex IV (Mi-CK) activity in the skeletal muscle (3-5). Elevated levels of EMA are typically observed in metabolic disorders that result in the accumulation of butyryl-CoA (6). Abnormal accumulation of butyryl-CoA can result from disorders affecting the short-chain fatty acid cycle, or the R-isoleucine catabolism pathway from 2-ethylmalonic-semialdehyde (7, 8). Accumulation of butyryl- CoA results in its carboxylation to ethylmalonyl-CoA that can be further hydrolyzed to EMA (6), which has been shown to be very inhibitory to energy metabolism (9, 10). At this time there are no known treatment options for EE patients because the function and substrate of ETHE1 are still unknown in any organism. More research is clearly needed on this important enzyme to determine its biological role. We have previously shown through the analysis of plants containing a mutation in Arabidopsis ETHE1 homolog that the protein also performs a critical role in early plant development (Chapter 4, this dissertation). These studies demonstrated that inactivation of ETHE1 disrupts early endosperm development resulting in embryo lethality and seed abortion (Chapter 4, this dissertation). The endosperm is thought to serve a protective role against physical and osmotic stresses, as well as serve a role in nutrient transfer to the developing embryo in dicots (11, 12). Of particular importance in the endosperm tissue is the chalazal cyst (CHZ), which has been speculated to serve in the transport of nutrients and/or the temporary storage of nutrients for the developing embryo (12)

119 Interestingly the CHZ is severely underdeveloped and is the first to arrest in ethe1(-/-) mutant seeds. We hypothesize that disruption of the chalazal cyst might block nutrient storage and transport to the embryo, which would in turn deplete necessary energy reserves for the continued development of the embryo (Chapter 4, this dissertation). This would be consistent with the predicted role of ETHE1 in short-chain fatty acid β- oxidation. Therefore, similar to the situation in humans, our data show that ETHE1 plays an essential role in plants, illustrating that Arabidopsis provides a model system to study ETHE1 function. However, because ethe1(-/-) loss of function mutants arrest at such an early stage in development, biochemical studies to determine ETHE1’s potential role(s) in plant development are very difficult. A chemically-inducible ETHE1-RNAi construct was generated, and the effects of ETHE1 removal at various time points of plant and suspension cell culture growth was studied in order to overcome this problem and further elucidate the biochemical role(s) that ETHE1 serves. In this chapter, we show that the reduction of ETHE1 levels in cell cultures results in morphological abnormalities of the suspension cells that are consistent with toxicity and nutrient starvation. In addition, we show that while ETHE1 is essential for seed germination it does not appear to have a critical role later in the life cycle of plants. These results provide further support for a role of ETHE1 in fatty acid β-oxidation.

5.3 Materials and Methods

5.3.1 Plant Material: Studies in Arabidopsis thaliana Heyhn were performed on the Wassilewskija (WS) ecotype. Seeds were vernalized at 4˚C for 48 hrs prior to being placed in a growth chamber to ensure equal germination. For germination studies, wild-type and transgenic seeds were surfaced sterilized with 70% (v/v) ethanol for 30 seconds followed by 10% (v/v) bleach for 30 min. Seeds were then rinsed 3 times with sterile deionized water. Sterile seeds were plated on half strength MS plates solidified with 0.6% (w/v) Gelrite and supplemented with 1.5% sucrose, pH 6.0. All other plant studies were performed on commercial potting soil in a controlled growth chamber at 23˚C on a 16:8hr light/dark cycle.

120 5.3.2 Cell Culture: Arabidopsis thaliana (ecotype Landsberg Erecta) cell suspensions were maintained in 50ml liquid growth medium (1X Murashige and Skoog Basal Salts, 1X Gamborg’s B5 vitamins, 3% (w/v) sucrose, 0.59 g/L MES, 0.5 mg/L NAA, and 0.05 mg/L BAP, pH 5.7) at 25ºC by gentle agitation (130 rpm) in 16:8 hour light:dark cycles. An aliquot of 6ml was transferred to 50 ml of fresh medium weekly.

5.3.3 Generation of Inducible ETHE1-RNAi Arabidopsis Plants:

A pX7-AtPDSi construct containing a 17 β-estradiol-inducible Cre/loxP (CLX) recombination system was obtained from Dr. Nam-Hai Chua’s lab (13). The additional restriction sites PacI, AscI, and NcoI were introduced into pX7-AtPDSi at the XhoI and SpeI sites by ligation of two 5’ sticky end complementary oligos 805 (5’- TCGAGATCCATGGAAGGCGCGCCATTTAATTAAAA-3’) and 806 (5’- CTAGTTTTAATTAAATGGCGCGCCTTCCATGGATC-3’). A 621 bp fragment of ETHE1 cDNA was amplified using primers 807 (5’- ACCTTTAATTAAGGCGCGCCCGACGTTTCTCATCCT-3) and 808 (5’- TCGCCTAGGATTTAAATAGATTTGACATAATGGTTT-3’) and cloned in both the sense and antisense directions into the modified pX7 construct from above. Constructs were introduced into Agrobacterium tumefaciens GV3101 and transformed into wild-type plants (14). Transgenic plants were identified through hygromycin selection (25 mg/L) and confirmed by PCR analysis. The ETHE1-RNAi construct was induced by the application of 2μM 17β-estradiol (Sigma) to plants as previously described (13).

5.3.4 Generation of Inducible ETHE1-RNAi Arabidopsis Suspension Cell Cultures Wild-type Arabidopsis suspension cell cultures were transformed as follows: seven day old suspension cells were subdivided into fresh medium at a high density (1:5 dilution) and allowed to rotate under normal conditions for 36 hours. After 36 hours, the suspension cell culture was diluted (1:2) with fresh medium and allowed to rotate for an additional 12 hours. The cells were again diluted (1:10) and 10 ml aloquots placed into a 250 mL Erlenmeyer flasks. Agrobacterium cultures containing the pX7-ETHE1-RNAi construct described above were grown to OD600 ~0.8, diluted (1:10) in fresh medium and

121 washed three times in the Arabidopsis medium. The resulting Agrobacterium pellet was resuspended in 1 mL of Arabidopsis media and 100 μl added to the diluted Arabidopsis culture cells. The cells were placed under 16:8 hour light:dark conditions for 48 hours with no shaking. After 48 hours, the cells were washed three times in fresh medium and resuspended in 10 ml fresh medium containing 200 mg/L cefotaxime. The cells were allowed to rotate under normal conditions for the next three days and then densely plated on solid medium containing hygromycin (25 mg/L) and cefotaxime (200 mg/L). Positive transformants were confirmed by PCR screening and reintroduced back into culture medium containing cefotaxime and hygromycin. The ETHE1-RNAi construct was induced by the addition of 2μM 17β-estradiol (Sigma) to the cell culture medium.

5.3.5 Molecular Analyses ETHE1 RNA levels were determined in plants and cell cultures using RT-PCR on total RNA isolated from either leaves or cell cultures. ETHE1 cDNA was synthesized from total RNA (1 μg) using a Thermoscript RT-PCR kit and mRNA levels were measured by 25 cycles of PCR using the primer pair 2-3-5’ (5’- TGGACAAGACTGTGGATAGAGA) and 2-3-2. ACTIN 8 was used as a control (15). ETHE1 protein levels were determined through western blot analysis. Total protein was isolated by grinding cell cultures that had been harvested in liquid nitrogen. This was followed by incubation with Laemeli isolation buffer (0.15 M TRIS-HCL, pH 6.8, 4.8% w/v SDS, 24% v/v glycerol, 1% v/v 2-mercaptoethanol, 1% v/v protease inhibitor cocktail (Sigma)) at 60° C for 20 min followed by centrifugation at 20,000 g for 5 min. Equal amounts (10 μg) of total protein or protein from the cell fractionation steps were separated by SDS-PAGE and subjected to western blot analysis with anti-ETHE1 antibody (1:1000) followed by detection with alkaline phosphatase-conjugated goat anti- rabbit secondary antibody (1:3000) using standard procedures (16).

5.3.6 EMA Toxicity studies in Arabidopsis Cell Cultures Wild-type cell cultures that had been freshly-subdivided were incubated with varying concentrations of EMA (0 mM, 0.5 mM, 1 mM, 2 mM, and 2.5 mM) and allowed

122 to grow normally. On day four of incubation with EMA, total protein was isolated and subjected to western blot analysis as described above.

5.3.7 Microscopy The morphology of Arabidopsis suspension cells was observed by Nomarski light microscopy. Cell vitality experiments were performed on Arabidopsis suspension cells at various time points, by measuring cell wet weight as well as by staining with a 0.4% solution of Typan blue (Sigma). All images were collected on an Olympus AX70 light microscope and arranged using Adobe Photoshop.

5.4 Results

Because the ETHE1 loss of function mutants are lethal, the role of the protein on other aspects of plant development can not be determined. In order to allow for a much broader functional study and gain better insight into the biochemical and physiological role(s) of ETHE1 in plants and potentially in humans, we generated a chemically- inducible ETHE1-RNAi construct that allowed us to monitor the effects of depleting ETHE1 from Arabidopsis suspension cell cultures and plants at various time points in plant development.

5.4.1 Generation of transgenic plant and suspension cell cultures expressing the inducible pX7-ETHE1-RNAi construct An inducible ETHE1-RNAi construct was generated by cloning a 607 bp fragment (nucleotides 194-801) in the sense and anti-sense orientations behind the 17-β- estradiol promoter (Fig 5.1A-B). This construct was introduced into wild-type Landsburg Erecta Arabidopsis suspension cell cultures and wild-type (ws) plants. Three independent transgenic lines from both the cell cultures and transgenic plants selected positive for hygromycin resistance and PCR screening.

5.4.2 Time Course Expression Analysis of pX7-ETHE1-RNAi in Cell Cultures Previous research using the pX7 inducible vector system with various genes has shown high levels of expression from approximately 12 hours to 2 weeks after induction,

123 (A)

124

Figure 5.1: Generation of inducible RNAi. (A) Gene map of Arabidopsis ETHE1. Arrows show the direction and location of primers used in this study. (B) Schematic diagram of inducible pX7-ETHE1-RNAi construct. DNA fragments

125 encoding sense an antisense ETHE1 RNA were cloned into the pX7 vector. XVE, a transactivator containing regulator domain of an estrogen receptor; HYG, a hygromycin-resistance marker; Cre, the bacteriophage P1 Cre recombinase; loxP, specific recognition sites of Cre;G10-90, strong constitutive promoter. (C) Upon the induction by 17-β-estradiol, Cre/loxP-mediated recombination leads to the activation of ETHE1-RNAi transcription by bringing it immediately downstream of the constitutive G10-90 promoter.

126 depending on the age of the plant and the gene being expressed (13, 17). Because there is no published data on the expression patterns of this system in suspension cell cultures, a time course study of ETHE1 RNA levels was conducted for 12 days after the addition of the 17-β-estradiol to the cell culture media. RNA was isolated and ETHE1 transcript levels were measured at 2 day intervals from the first day of induction to the 12th day of growth in the pX7-ETHE1-RNAi transgenic cells with wild-type cell cultures as a control (Fig 5.2). As expected, there was no effect of the inducer on wild-type cells; however, by day 8, all three transgenic RNAi lines showed either drastically reduced or no ETHE1 RNA (Fig 5.2). Of the three transgenic lines, line pX7-ETHE1-RNAi 19 showed the most drastic reduction in ETHE1 transcript levels with the complete absence of the ETHE1 transcript by day 2 (Fig 5.2).

5.4.3 ETHE1 is Critical for Arabidopsis Cell Suspension Growth Induction of the ETHE1-RNAi resulted in major changes in the growth properties of the cultures compared with wild-type cells grown under the same conditions. Transgenic cells expressing the ETHE1 RNAi stopped dividing by day 2 and became yellow in color by day 6 after induction. They also began to show signs of cell lysis (data not shown). To investigate whether the absence of ETHE1 results in cell death, total cell weight was measured at various times after ETHE1 RNAi induction. As expected, wild- type cells continued to increase in cell weight over time indicating a normal pattern of growth and division (Fig 5.3A). In contrast, the wet cell weight of the transgenic cell cultures increased very little beginning two days after treatment with the inducer suggesting a cease in cell division and ultimately cell death. Samples were stained with Typan blue at various times after ETHE1 RNAi induction to detect cell death and the percentage of live cells was calculated using a hemocytometer in order to investigate the possibility that the absence of ETHE1 results in cell death. Although the viability of cell cultures normally decreases over time after division of the culture, the ETHE1-RNAi cultures began to exhibit decreased cell viability as early as two days after induction with 17-β-estradiol inducer (Fig 5.3B-C). By the end of the 12 day study, the average viability of wild-type cells had decreased by 40% while the viability of the ETHE1-RNAi lines had decreased an average of 85% with lines 10 and

127

Fig 5.2: Time course studies of induced ETHE1-RNAi in Arabidopsis cell cultures with RT-PCR. ETHE1 levels were analyzed on RNA isolated from three transgenic cell culture lines (#6, 10, 19) at various days after the induction of the ETHE1-RNAi and compared to wild-type cultures. ACT8 was used as an internal loading control.

128 19 showing almost complete cellular death (Fig 5.3B). This is consistent with our observations that lines 10 and 19 exhibited the fastest and most complete reduction of ETHE1 RNA after RNAi induction.

5.4.4 Inducible Expression of ETHE1-RNAi Results in Filamentous Growth and Arrest of Normal Division in Arabidopsis Cell Culture Cell cultures expressing the ETHE1 RNAi constructs were analyzed by light microscopy to determine if loss of ETHE1 results in morphological changes prior to cellular arrest and lysis. Interestingly, just two days after RNAi induction, all three lines were found to contain cells of irregular shape and morphology. In particular, these cells were found to contain enlarged central vacuoles, granulated cytoplasm, and a thick cell wall in contrast to wild-type cells under the same conditions, which were spherical and contained a dense cytoplasm (Fig 5.4). The morphological features of the RNAi cultures resembled those of cells undergoing vacuolar autophagy (18), which is a process of nutrient recycling that is initiated in times of nutrient starvation in fungi, plants, and animals (18). This process involves the enveloping of large portions of cytoplasm, including organelles into membrane-bound vesicles to form autophagosomes that are transported to the vacuole for degradation in order to recycle spare nutrients for other pathways (19). This process naturally occurs in wild-type cells grown in culture for 21 days at which point they lose the ability to divide (20). It has been proposed that this is due to an accumulation of undesirable products as a result of metabolism along with a depletion of nutrients (20). Consistent with this proposal is the induction of autophagy in both Nicotiana and Arabidopsis cell cultures under sucrose starvation (19, 21). These results suggest therefore, that the removal of ETHE1 might result in the depletion of critical nutrient(s) or an accumulation of a toxin. Given the known toxicity of EMA, it is possible that EMA accumulation might play a role in the induction of autophagy and cell death after RNAi induction.

5.4.5 Incubation of Arabidopsis Cell Culture with EMA Undergoes Autophagy EMA is a known toxic metabolite, and studies have shown its inhibitory effect on the activity of the metabolic regulator Mi-CK in humans (9), however to our knowledge, there have been no studies on the effects of EMA on plants to date. To address whether

129

130 Fig 5.3: Effects of ETHE1-RNAi on cell culture growth and survival. (A) Measurement of wet cell weight of wild-type Landsberg erecta Arabidopsis suspension cell cultures and pX7-ETHE1-RNAi cell culture lines over 8 days of induction. RNAi represents the average of the three lines +/- SD. (B) Percentage of total live cells from wild-type and the pX7-ETHE1-RNAi lines were counted on a hemocytometer and recorded through the 12 day time course study. Four readings were taken and averaged per day for each plant cell culture line (mean +/- SD). (C) Light microscopy of wild-type and pX7-ETHE1-RNAi cells in the presence of the inducer at day 8 of growth. Dead cells, indicated with arrowheads, were identified by typan blue staining.

131 EMA can cause morphological effects similar to those observed after depletion of ETHE1, wild-type cell cultures were incubated with concentrations of EMA ranging from 0 to 2.5 mM for 4 days. Minimal effects were observed for EMA concentrations lower than 1 mM. However, for cell cultures incubated with 2 and 2.5 mM EMA, an increase in cell death was observed compared to normal conditions (Fig 5.5A). After four days the viability of cultures treated with 2.5 mM EMA was 46.4 % of the four day wild-type control (Fig 5.5A). Likewise, Nomarski imaging revealed numerous cells undergoing vacuole autophagy (Fig 5.5B-C). Cells displaying autophagy were not observed in the wild-type control culture. This experiment suggests that the detrimental effects of ETHE1 depletion observed in the pX7-ETHE1-RNAi cell culture studies could be the result of a build up of the toxic metabolite EMA. However, the morphology of the EMA treated cells was not identical to the ETHE1-RNAi cells. It has been shown that the toxic effects of EMA to the Mi-CK activity can be reduced upon incubation of EMA with the anti-oxidants glutathione (GSH) and ascorbic acid (9). Given the similarity of ETHE1 to the glyoxalase II enzymes, which function in the detoxification of methylglyoxal (22), ETHE1 might perform a similar detoxification reaction with EMA. To test this hypothesis, ETHE1 protein levels were monitored using western blot analysis on samples prepared from wild-type cells after four days of growth in the presence of various concentrations of EMA. If ETHE1 plays a specific role in the detoxification of EMA, an increase in ETHE1 levels might be observed. No difference in ETHE1 levels between wild-type cultures incubated with EMA and the control cells was observed (data not shown). This suggests that the induction of ETHE1 is not necessary for the detoxification of EMA. However, more experiments are necessary to test this hypothesis.

5.4.6 ETHE1 is Essential for Plant Germination

Previously we showed that a ETHE1 loss of function mutation blocks seed development with the arrest occurring at a very early stage in endosperm development (chapter 4). However, because the arrest occurs so early in seed development, the amount of information that can be deduced from the previous studies is limited. Given that ETHE1 is critical for seed development, it seemed likely that ETHE1 plays critical

132

Fig 5.4: Morphology of Arabidopsis cell cultures expressing ETHE1-RNAi. Nomarski imaging of cells from both wild-type and the transgenic cells of pX7- ETHE1-RNAi line 19 at day 4 of induction. Size bars = 10 μM.

133

Fig 5.5: EMA toxicity in Arabidopsis cell cultures. (A) Effect of EMA on cell viability of wild-type cells after 4 days of growth. (B) Light microscope images of

134 cells after 4 days of growth in the presence and absence of 2.5 mM EMA. (C) High magnification of cells after 4 days of growth in the presence and absence of 2.5 mM EMA. Size bars = 10 μM.

135 roles at other stages of plant development. pX7-ETHE1-RNAi seeds were germinated in the presence of the 17-β-estradiol to monitor the effect of blocking ETHE1 expression and test this hypothesis (Fig 5.6). Seeds containing the pX7-ETHE1-RNAi construct were unable to germinate on MS plates containing 17-β-estradiol, while both wild-type seeds grown in the presence of the inducer and pX7-ETHE1-RNAi seeds grown on MS plates without the inducer were able to germinate and grow normally (Fig 5.6). This study demonstrates that in addition to its critical role in seed development, ETHE1 is also essential for seed germination. Given ETHE1’s critical role in seed development and germination, we hypothesized that ETHE1 might also play an important role later in plant development. To test this hypothesis we monitored the effects of removing ETHE1 after seedling germination and establishment. After seedling germination and establishment, plants were continuously root drenched with 17-β-estradiol every four days to ensure continued RNAi expression. Total RNA was collected, and ETHE1 RNA levels were monitored through RT-PCR. We were able to drastically reduce/knockout ETHE1 levels during plant growth on soil (Fig 5.7A). Surprisingly, no obvious phenotypic effects were observed. Both the control and pX7-ETHE1-RNAi plants appeared to exhibit normal vegetative growth (Fig 7B). No differences in the size or morphology of the plants were observed. They also bolted and flowered normally (Fig 5.7C). Interestingly, the pX7- ETHE1-RNAi plants also produced siliques. However, because ETHE1 appears to be critical to endosperm development, more studies are necessary to monitor the viability of these seeds.

5.5 Discussion

Mutations in ETHE1 have shown that the protein is essential in both humans and plants. In humans ETHE1 mutations are responsible for the autosomal recessive disorder EE (3, 5). In plants ETHE1 mutations block very early stages in seed development (chapter 4). However, in spite of the critical role(s) of ETHE1 in development, the biochemical role(s) of ETHE1 are still unknown. To date, the effect of ETHE1 mutations

136

Fig 5.6: ETHE1 is required for seed germination. pX7-ETHE1-RNAi seeds were germinated in the presence or absence of 17-β-estradiol. Wild-type seeds germinated on plates containing 17-β-estradiol were used as a control. Images are taken after 7 days of growth.

137

Fig 5.7: Effects of inducible ETHE1-RNAi on plant development. (A) RT- PCR to measure ETHE1 RNA levels in wild-type and plant lines #1-3 expressing pX7-ETHE1-RNAi. (B) Comparison of vegetative growth at day 13 between

138 wild-type plants and plants expressing pX7-ETHE1-RNAi in the presence of the inducer. (C) Comparison of growth at day 25 of wild-type plants and pX7- ETHE1-RNAi plants in the presence of the inducer.

139 has only been studied under complete knockout conditions. However, the early seedling lethality caused by the mutation has limited the amount of knowledge gained in these studies. Using an inducible ETHE1-RNAi construct, we have generated a system where the physiological role of the protein at specific stages of growth can be studied. Results obtained from studies in this chapter provide further support for the theory that ETHE1 may play a role in the β-oxidation of fatty acids.

5.5.1 ETHE1 is Critical for Cell Survival in Arabidosis Suspension Cell Culture Analysis of Arabidopsis cell cultures expressing the inducible ETHE1-RNAi construct demonstrated that ETHE1 is essential for cell growth in culture. Removal of ETHE1 resulted in an average 73% loss in viability compared with wild-type cells under the same conditions (Fig 5.3). Additionally, Nomarski imaging of the inducible ETHE1- RNAi cell lines two days after RNAi induction showed numerous cells undergoing filamentous, irregular growth (Fig 5.4). These results are typical of cells undergoing autophagy, which is generally initiated under periods of nutrient starvation to recycle nutrients for other uses (18). Studies in both tobacco and Arabidopsis cell cultures after four days of sucrose starvation identified cells exhibiting similar to those observed in the ETHE1-RNAi cultures (19, 21). It was shown that these starving cells undergo a loss of oxidation potential, as well a drastic decrease in the rate of respiration after 24 hours of nutrient starvation, which appears to stem from a breakdown in the mitochondria (23). Wild-type cells typically undergo autophagy after 21 days of growth when sucrose levels drop in unrefreshed culture medium (20). In addition to these observations, several genes are up-regulated during nutrient starvation with over 15% of these genes linked to metabolic processes including carbohydrate metabolism, tyrosine, isoleucine, and valine metabolism. These pathways probably participate in nutrient recycling and have been suggested to be linked primarily to stress responses (19). Given the results from our studies and the results from studies on nutrient starvation and autophagy, it is possible that the depletion of ETHE1 causes nutrient depletion in cultured cells. Because EMA is a known toxic metabolite and high levels of EMA are found in the urine of EE patients, we incubated wild-type cell cultures with EMA. Cells

140 exhibiting properties of autophagy were seen when wild-type cells were treated with EMA (Fig 5A-C). Given the similarity of ETHE1 to the detoxifying enzyme glyoxalase II (GLX2), it was also possible that ETHE1, like GLX2, could play a role in the removal of EMA using a -dependpent pathway. However, our studies showed that cell cultures incubated with EMA did not show an increase in ETHE1 protein levels. Therefore, ETHE1 may not be involved in the direct detoxification of EMA, but may play a role in a metabolic pathway preventing the accumulation of EMA. A role in the detoxification of EMA cannot be dismissed at this time because we only measured ETHE1 levels at a single time point. It is possible that ETHE1 may be recruited for expression within the first few hours EMA incubation, returning to wild-type levels shortly after. More studies are necessary to investigate this possibility.

5.5.2 ETHE1 is Essential for Seed Germination Previously we had shown that ETHE1 plays an essential role in seed development and in particular, endosperm development (chapter 4, this dissertation). The ETHE1 loss of function mutant exhibited an arrest in endosperm development beginning soon after fertilization, ultimately resulting in embryo arrest at early heart stage (chapter 4, this dissertation). Studies have shown that the breakdown of fatty acids is necessary for both embryo development and seed germination (24, 27). The Acyl-CoA oxidase (ACX) genes in Arabidopsis, which catalyze the conversion of acyl-CoA to 2-trans-enoyl-CoA in the first step of the β-Oxidation pathway are elevated during germination and early post-germination growth, which is consistent with a rapid period of fatty acid degradation (24-29). Arabidopsis contains six ACX members, each specific for catalyzing a different length of fatty acid. Substrate specificities have been determined for four of the acyl-CoA oxidases: ACX1 (long chain), ACX2 (very-long chain), ACX3 (medium chain), and ACX4 (short chain) (26, 29-32). It was found that long chain ACXs are essential for seedling germination, while short-chain ACXs are required for embryo development (24, 27). Interestingly, our studies showed that the removal of ETHE1 from wild-type seeds inhibited the ability of the seeds to germinate (Fig 5.6). This suggests that ETHE1 may participate either directly or indirectly in the β-oxidation of short chain fatty acids (chapter 4, this

141 dissertation). In particular, ETHE1 may not only be involved directly or indirectly in short-chain fatty acid β-oxidation but may also have a role in long-chain fatty acid β- oxidation. Because ETHE1 plays such a critical role in early plant development, it was surprising that we did not identify observable phenotypes when ETHE1 was depleted later in development (Fig 5.7). Plants expressing the ETHE1-RNAi after germination and early seedling development appeared to develop normally in their vegetative state, as well as bolt, flower, and produce siliques like wild-type plants. However, given that the ETHE1 knockout mutant is lethal in seed development, it is unlikely that seed produced from these plants are viable (Chapter 4, this dissertation). More studies are necessary to determine the viability of these ETHE1-RNAi seeds. Consistent with our observations, very few known β-oxidation mutants show obvious phenotypes after seedling establishment (33). This suggests that β-oxidation may not play a critical role in plant development under normal conditions after seedling establishment, which could explain the absence of a phenotype in mature ETHE1-RNAi induced plants. It is known that β-oxidation is necessary for the synthesis of the plant hormone jasmonic acid (JA), which accumulates under periods of stress, namely wounding and insect resistance, to help in the regulation of growth inhibition, senescence, and leaf absicission (34). If ETHE1 is closely linked to β-oxidation, removal or over-expression of ETHE1 under certain stress conditions might yield more information on its biochemical role(s). While further experiments are required to more clearly define the biochemical role of ETHE1, the results of studies presented here have shown that ETHE1 is critical for both cell culture survival as well as seed germination, suggesting a strong link to the β-oxidation of fatty acids. Furthermore, the inducible ETHE1-RNAi lines developed here should provide a system to address the important questions that still remain concerning the biochemical role of ETHE1.

142 5.6 References

1. Burlina, A. B., Dionisi-Vici, C., Bennett, M. J., Gibson, K. M., Servidei, S., Bertini, E., Hale, D. E., Schmidt-Sommerfeld, E., Sabetta, G., and Zacchello, F. (1994). A New Syndrome with Ethylmalonic Aciduria and Normal Fatty Acid Oxidation in Fibroblasts. J. Pediatr. 124, 79-86

2. Burlina, A., Zacchello, F., Dionisi-Vici, C., Bertini, E., Sabetta, G., Bennet, M. J., Hale, D. E., Schmidt-Sommerfeld, E., and Rinaldo, P. (1991). New Clinical Phenotype of Branched-chain Acyl-CoA Oxidation Defect. Lancet. 338, 1522-1523

3. Tiranti, V., D'Adamo, P., Briem, E. et al. (2004). Ethylmalonic Encephalopathy is caused by Mutations in ETHE1, a Gene Encoding a Mitochondrial Matrix Protein. Am. J. Hum. Genet. 74, 239-252

4. Zafeiriou, D. I., Augoustides-Savvopoulou, P., Haas, D. et al. (2007). Ethylmalonic Encephalopathy: Clinich and Biochemical Observations. Neuropediatrics 38, 78-82

5. Tiranti, V., Briem, E., Lamantea, E. et al. (2006). ETHE1 Mutations are Specific to Ethylmalonic Encephalopathy. J. Med. Genet. 43, 340-346

6. Hegre, C. S., Halenz, D. R., and Lane, M. D. (1959). The Enzymatic Carboxylation of Butyryl Coenzyme A. J. Am. Chem. Soc. 81, 6526-6527

7. Mamer, O. A., Tjoa, S. S., Scriver, C. R., and Klassen, G. A. (1976). Demonstration of a New Mammalian Isoleucine Catabolic Pathway Yielding an Rseries of Metabolites. Biochem. J. 160, 417-426

8. Nowaczyk, M. J., Lehotay, D. C., Platt, B. A., Fisher, L., Tan, R., Phillips, H., and Clarke, J. T. (1998). Ethylmalonic and Methylsuccinic Aciduria in Ethylmalonic Encephalopathy Arise from Abnormal Isoleucine Metabolism. Metabolism 47, 836- 839

143 9. Leipnitz, G., Schuck, P. F., Ribeiro, C. A., Dalcin, K. B., Assis, D. R., Barschak, A. G., Pulrolnik, V., Wannmacher, C. M., Wyse, A. T., and Wajner, M. (2003). Ethylmalonic acid Inhibits Mitochondrial Creatine Kinase Activity from Cerebral Cortex of Young Rats in Vitro. Neurochem. Res. 28, 771-777

10. Brustovetsky, N., Brustovetsky, T., and Dubinsky, J. M. (2001). On the Mechanisms of Neuroprotection by Creatine and Phosphocreatine. J. Neurochem. 76, 425-434

11. Garcia, D., Saingery, V., Chambrier, P., Mayer, U., Jurgens, G., and Berger, F. (2003). Arabidopsis haiku Mutants Reveal New Controls of Seed Size by Endosperm. Plant Physiol. 131, 1661-1670

12. Nguyen, H., Brown, R. C., and Lemmon, B. E. (2000). The Specialized Chalazal Endosperm in Arabidopsis thaliana and Lepidum virginicum (Brassicaceae). Protoplasma 212, 99-110

13. Guo, H. S., Fei, J. F., Xie, Q., and Chua, N. H. (2003). A Chemical-regulated RNAi System in Plants. Plant J. 34, 383-392

14. Chung, M. H., Chen, M. K., and Pan, S. M. (2000). Floral Spray Transformation can Efficiently Generate Arabidopsis Transgenic Plants. Transgenic Res. 9, 471-476

15. An, Y. Q., McDowell, J. M., Huang, S., McKinney, E. C., Chambliss, S., and Meagher, R. B. (1996). Strong, Constitutive Expression of the Arabidopsis ACT2/ACT8 Actin Subclass in Vegetative Tissues. Plant J. 10, 107-121

16. Harlow, E., and Lane, D. (1999) Using Antibodies: A Laboratory Manuel, Cold Spring Harbor Laboratory, N.Y., Cold Spring Harbor.

17. Zuo, J., Niu, Q. W., and Chua, N. H. (2000). Technical Advance: An Estrogen Receptor-based Transactivator XVE Mediats Highly Inducible in Transgenic Plants. Plant J. 24, 265-273

18. Huang, W. P., and Klionsky, D. J. (2002). Autophagy in Yeast: a Review of the Molecular Machinery. Cell Struct. Funct. 27, 409-420

144 19. Contento, A. L., Kim, S. J., and Bassham, D. C. (2004). Transcriptome Profiling of the Response of Arabidopsis Suspension Culture Cells to Suc Starvation. Plant Physiol. 135, 2330-2347

20. Kobae, Y., Mizutani, M., Segami, S., and Maeshima, M. (2006). Immunochemical Analysis of Aquaporin Isoforms in Arabidopsis Suspension-cultured Cells. Biosci. Biotechnol. Biochem. 70, 980-987

21. Moriyasu, Y., and Ohsumi, Y. (1996). Autophagy in Tobacco Suspension-cultured Cells in Response to Sucrose Starvation. Plant Physiol. 111, 1233-1241

22. Thornalley, P. J. (1993). The Glyoxalase System in Health and Disease. Mol. Aspects Med. 14, 287-371

23. Journet, E. P., Bligny, R., and Douce, R. (1986). Biochemical Changes During Sucrose Deprivation in Higher Plant Cells. J. Biol. Chem. 261, 3193-3199

24. Penfield, S., Helen, M., Pinfield-Wells et al. (2006) Storage Reserve Mobilization and Seedling Establishment in Arabidopsis. In The Arabidopsis book Rockville, MD, American Society of Plant Biologists.

25. Adham, A. R., Zolman, B. K., Millius, A., and Bartel, B. (2005). Mutations in Arabidopsis Acyl-CoA Oxidase Genes Reveal Distinct and Overlapping Roles in Beta-oxidation. Plant J. 41, 859-874

26. Eastmond, P. J., Hooks, M. A., Williams, D., Lange, P., Bechtold, N., Sarrobert, C., Nussaume, L., and Graham, I. A. (2000). Promoter Trapping of a Novel Medium- chain acyl-CoA oxidase, which is Induced Transcriptionally during Arabidopsis Seed Germination. J. Biol. Chem. 275, 34375-34381

27. Pinfield-Wells, H., Rylott, E. L., Gilday, A. D., Graham, S., Job, K., Larson, T. R., and Graham, I. A. (2005). Sucrose Rescues Seedling Establishment but not Germination of Arabidopsis Mutants Disrupted in Peroxisomal Fatty Acid Catabolism. Plant J. 43, 861-872

145 28. Germain, V., Rylott, E. L., Larson, T. R., Sherson, S. M., Bechtold, N., Carde, J. P., Bryce, J. H., Graham, I. A., and Smith, S. M. (2001). Requirement for 3-ketoacyl- CoA Thiolase-2 in Peroxisome Development, Fatty Acid Beta-oxidation and Breakdown of Triacylglycerol in Lipid Bodies of Arabidopsis Seedlings. Plant J. 28, 1-12

29. Rylott, E. L., Rogers, C. A., Gilday, A. D., Edgell, T., Larson, T. R., and Graham, I. A. (2003). Arabidopsis Mutants in Short- and Medium-chain Acyl-CoA Oxidase Activities Accumulate Acyl-CoAs and Reveal that Fatty Acid Beta-oxidation is Essential for Embryo Development. J. Biol. Chem. 278, 21370-21377

30. Hooks, M. A., Kellas, F., and Graham, I. A. (1999). Long-chain Acyl-CoA Oxidases of Arabidopsis. Plant J. 20, 1-13

31. Hayashi, H., De Bellis, L., Ciurli, A., Kondo, M., Hayashi, M., and Nishimura, M. (1999). A Novel Acyl-CoA Oxidase that can Oxidize Short-chain Acyl-CoA in Plant Peroxisomes. J. Biol. Chem. 274, 12715-12721

32. Froman, B. E., Edwards, P. C., Bursch, A. G., and Dehesh, K. (2000). ACX3, a Novel Medium-chain Acyl-coenzyme A Oxidase from Arabidopsis. Plant Physiol. 123, 733-742

33. Goepfert, S., and Poirier, Y. (2007). Beta-oxidation in Fatty Acid Degradation and Beyond. Curr. Opin. Plant Biol. 10, 245-251

34. Hedden, P., and Thomas, S. (2006) Plant hormone signaling. APR, Oxford, UK, Blackwell Publishing.

146

Chapter 6

Over-expression of Arabidopsis ETHE1 Results in Enhanced Growth Properties of Plants

Meghan M. Holdorf, Lara Strittmatter, and Christopher A. Makaroff

Department of Chemistry and Biochemistry at Miami University

* Corresponding author:

Christopher A. Makaroff

Department of Chemistry and Biochemistry

Miami University

E-mail: [email protected]

To be submitted to Journal of Plant Molecular Biology

Authors contribution of work: The work on vascular cross-sections of Arabidopsis inflorescences was performed by L. Strittmatter. All other analysis and work presented in this paper was performed by M.M. Holdorf.

147

6.1 Summary

ETHE1 is an essential gene in both humans and plants yet the biochemical and physiological roles of this protein are still unknown. We demonstrate that Arabidopsis ETHE1 is normally localized in the mitochondrion and using MALDI-TOF peptide mapping identify the mature N-terminus of the protein. Because ETHE1 mutants in Arabidopsis result in the early arrest of seed development, we investigated the role(s) of ETHE1 by constitutively over-expressing the protein in plants. Over-expression of the native form of the protein had no observable effect on plant growth or development. However, over-expression of ETHE1 in the cytosol results in plants that bolt faster, have a more upright growth habit, and set more seed. Based on the phenotype of the transgenic plants we predict that the mis-localized ETHE1 may indirectly alter cytokinin levels.

148 6.2 Introduction

Genetic engineering has become an important and effective tool for enhancing agricultural traits in plants, including disease resistance, drought tolerance, and the protection of plants against abiotic and biotic stresses (1, 2). Genetic engineering allows for the manipulation of traits by moving genes that code for advantageous traits from one species to another. Commercially successful examples of genetically engineered crops include the biofortification of vitamin A in golden rice, herbicide resistance crops for weed control, and pest control seen in bt corn. Of particular economic interest is seed yield enhancement, which can be defined in terms of quantity and/or quality and is dependent on several factors including the number and size of the organs, plant architecture, seed production, root development, nutrient uptake, stress tolerance, and plant population (3-7). Optimizing one or more of the above mentioned factors can contribute to an overall increase in crop yield. One area that has received considerable recent attention is the enhancement of stress tolerance (2, 8). Recently it was shown that plants that over-express glyoxalase system enzymes (glyoxalase I and glyoxalase II) exhibit enhanced tolerance to salinity and zinc stresses (9-12). The glyoxalase system is ubiquitous in nature and has been studied in a number of organisms (13, 14). It is thought that the primary role of the glyoxalase system is the detoxification of 2-oxaldehydes, toxic byproducts of carbohydrate and lipid metabolism, with the primary physiological substrate being methylglyoxal (2-oxopropanal) (13). Methylglyoxal is a cytotoxic and mutagenic byproduct of respiration that can inactivate proteins, modify guanylate residues, and create interstrand DNA crosslinks (15-17). Cellular detoxification of methylglyoxal begins with its non-enzymatic reaction with glutathione to produce a thiohemiacetal which is then converted to S-D-lactoylglutathione (SLG) by Glyoxalase I (GLX1) (18). Glyoxalase II (GLX2) then hydrolyzes SLG to form D-lactic acid and free glutathione (19). It is thought that the enhanced stress tolerance observed with glyoxalase over- expression is due to the rapid removal of toxic byproducts that are produced during times of respiratory stress (9).

149 GLX1, which is present as a single isozyme, has been well characterized in several systems (20-25). However, much less is known about the structure and function of the GLX2 enzymes, which have been isolated from several organisms including yeast, plants, and animals (26-28). GLX2 enzymes have been isolated and characterized from both the cytosol and mitochondria (26-28). In mammals, a single gene encodes both the cytosolic and mitochondrial GLX2 proteins; however multiple GLX2 genes exist in yeast and higher plants (26-28). We recently showed that Arabidopsis Ethylmalonic Encephalopathy Protein 1 (ETHE1), a GLX2-like protein with unknown function, is essential for seed development and shows enhanced expression in plants exposed to the abiotic stresses mannitol, NaCl, and abscisic acid (Holdorf, unpublished). In addition, publicly available microarray data revealed that ETHE1 transcript levels are also elevated during biotic stresses (e.g. infection with B. cinerea and P. syringae) as well as during senescence (29). These findings and the similarity of ETHE1 to the GLX2 enzymes raised the possibility that over-expression of ETHE1 could confer enhanced growth properties to plants. In order to test this hypothesis we over-expressed ETHE1 in transgenic Arabidopsis and tobacco plants. In this paper we demonstrate that constitutive over- expression of ETHE1 results in plants that bolt sooner, have a more upright growth habit, and set more seed. Interestingly, the enhanced growth properties are only observed when the protein is targeted to the cytosol, but not the mitochondrion, the normal location of ETHE1. These results suggest that the cytoplasmically localized protein catalyzes a reaction that may not normally occur in plant cells. Based on the phenotype of the transgenic plants we predict that the mis-localized ETHE1 may indirectly alter cytokinin levels.

6.3 Materials and Methods

6.3.1 Plant Material and Growth Conditions Studies in Arabidopsis thaliana Heyhn were performed on the Wassilewskija (WS) ecotype. Seeds were vernalized at 4˚C for 48 hours prior to being placed in a

150 growth chamber. For measurement of toxicity and germination rates, seeds were surfaced sterilized with 70% ethanol for 30 seconds followed by 10% bleach for 30 minutes, rinsed three times with sterile deionized water and plated on half strength Murashige and Skoog (MS) plates solidified with 0.6% Gelrite and supplemented with 1.5% sucrose, pH 6.0. Plates were vernalized for four days at 4º C and then placed in a controlled growth chamber at 23º C on a 16:8 light/dark cycle.

Wild-type Arabidopsis (Ws), ETHE1256-OE, and ETHE1294-OE seeds were sown onto half-strength MS plates with 0 mM to 2 mM valine, propionate, ethylmalonic acid, , or crotonic acid (all purchased from Sigma). Neutralized, filter-sterilized metabolites were added after the medium was autoclaved. Once toxicity levels for wild- type seeds were obtained, wild-type, ETHE1256-OE, and ETHE1294-OE seeds were plated together on the same plates. Root lengths and cotyledon growth from 100 plants per stress were measured and compared after 7 days of growth. Unless otherwise noted, plants were grown on commercial potting soil in a controlled growth chamber at 23˚C on a 16:8hr light/dark cycle. The effect of over- expression of ETHE1 on plant height, bolting rate, flowering time, and days to senescence was examined using 10 plants from each of four independent T2 lines of

ETHE1256-OE plants and from six independent T2 lines of ETHE1294-OE plants. Plant height was measured from the base of the primary inflorescence. Bolting rate was measured as the days between germination and the first presence of the inflorescence. Flowering time was measured from germination to the day of the first open flower. Time to senescence was measured from germination to the first stage of senescence as defined

by the yellowing of the rosette leaves. Four Arabidopsis ETHE1256-OE lines were analyzed for dry weight, number of siliques per plant, and number of seeds per silique. Dry weight was determined by cutting the base of the inflorescence at the soil line after the last flower had opened and drying the material allowing the material to dry fully between two papers for ten days in a 37˚C incubator. Weights of ten plants from each of the four transgenic lines were determined and compared with WT. Siliques were counted on ten individual plants of each transgenic line after the completion of flowering. The number of seeds per silique was determined under a dissecting microscope on five siliques from each of the ten plants per transgenic line.

151 Studies in tobacco were performed on the Petite Havana ecotype. Seeds were vernalized at 4˚ C for 48 hours prior to being placed in a green house. Plants were grown on commercial potting soil with constant liquid feed and watered as needed. Seeds from

six T2 ETHE1256-OE transgenic lines were grown on soil for analysis of plant bolting rate, flowering time, seed yield, and days to senescence. Seed yield was determined by collecting seed pods over the lifetime of the plant and seed number determined by weight. Arabidopsis (ecotype Landsberg Erecta) cell suspensions were grown in 50ml liquid growth medium (1X Murashige and Skoog Basal Salts, 1X Gamborg’s B5 vitamins, 3% (w/v) Sucrose, 0.59 g/L MES, 0.5 mg/L NAA, and 0.05 mg/L BAP, pH 5.7) at 25ºC with gentle agitation (130 rpm) in 16:8 hour light:dark cycles. A six ml aliquot was transferred to 50 ml fresh medium each week.

6.3.2 Generation of ETHE1 Over-expression Plants and Cell Cultures Based on the cDNA sequence of Arabidopsis ETHE1 (genebank id: 79606538) available at the time these experiments were initiated, a 35S over-expression construct of a 256 amino acid Arabidopsis ETHE1 protein (ETHE1256) was generated. A second larger ETHE1 EST (gi:145362330) was subsequently identified and used to generate an

over-expression construct that produced a 294 amino acid ETHE1 protein (ETHE1294).

Fragments containing ETHE1256 and ETHE1294 were generated by PCR with primers Nco1 and 2-3-3’ and 1013 and 2-3-3’, respectively, and cloned into pFGC 5941 (Genbank assession # AY310901) (Fig 6.1A-B). After sequence confirmation, the constructs were transformed into Agrobacterium GV3101/PMP90 cells and used to transform wild-type plants (Chung et al., 2000). Transgenic plants were identified by BASTA screening (1:4000 dilution) and the presence of the over-expression construct was confirmed by PCR with a vector specific primer and primer 2-3-2.

Over-expression constructs that produce ETHE1256 or ETHE1294 with C-terminal FAST tags (30) were generated to facilitate subcellular localization and N-terminus determination studies. ETHE1 fragments were amplified from their respective pFGC clones using a pFGC vector specific promoter primer, and 774, a primer that eliminated

152

153

Fig. 6.1: Molecular analysis of Arabidopsis ETHE1. (A) Gene map of ETHE1. Arrows show the direction and location of primers used in this study. Bent arrows show the start and Genebank accession numbers of ESTs used for cloning. (B) Sequence alignment of ETHE1 with GLX2-2 using CLUSTAL W. Amino acids in gray lettering correspond to the extended protein sequence in

ETHE1294.which is not present in ETHE1256. Identical residues are highlighted in black.

154 the ETHE1 stop codon, and cloned into pFAST (Fig 6.1A). The constructs were sequenced and then transformed into GV3101/PMP90 Agrobacterium cells. Wild-type Arabidopsis suspension cell cultures were transformed as follows: seven day old suspension cells were subdivided into fresh liquid growth medium at a high density (1:5 dilution) and allowed to rotate under normal conditions for 36 hours. The culture was further diluted (1:2) with fresh medium and allowed to rotate for an

additional twelve hours. Agrobacterium cultures were grown to OD600 ~0.8, diluted (1:10) in fresh Arabidopsis liquid growth medium and washed three times in that same medium. The suspension cells were diluted (1:10) into a final volume of ten ml fresh medium in a 250 ml Erlenmeyer flask and 100 μl of the Agrobacterium dilution added. The cells were placed under normal lighting conditions with no shaking. After 48 hours, the cells were washed three times in fresh medium and resuspended in ten ml fresh medium containing 200 mg/L cefotaxime. The cells were allowed to rotate under normal conditions for three days and then plated on solid medium containing kanamycin (50 mg/L) and cefotaxime (200 mg/L). Positive transformants were confirmed by PCR screening using primers 557 and 255. Positive calli were reintroduced into liquid growth medium containing kanamycin (50 mg/L) and cefotaxime (200 mg/L) for localization and N-terminal determination experiments. In order to test the effect of over-expression of ETHE1 in tobacco, an EcoRI and

XbaI fragment from ETHE1256 PFGC5941 was transferred into pPZP111 (31). After sequence confirmation, the construct was transformed into Agrobacterium GV3101/PMP90 cells and used to transform wild-type Petite Havana tobacco plants

through leaf disk infection (32). ETHE1256-OE/pPZP111 transformants were identified through kanamycin selection and confirmed by PCR using a vector specific primer and gene specific primer 2-3-2. The pPZP111 vector was transformed into separate plants as a control. ETHE1 RNA levels in wild-type and ETHE1 over-expression plants were determined using RT-PCR on total RNA (1 μg) isolated from leaves using a Thermoscript RT-PCR kit. PCR was conducted for 25 cycles using the primer pair 2-3- 5’ and 2-3-2 at 56˚ C. ACT8 was used as a control (33).

155 6.3.3 Microscopy

Morphology of the vascular tissue in ETHE1256-OE and wild-type plants was analyzed by light microscopy. Sections (1 cm) of the primary inflorescence from Arabidopsis plants at stage three (as defined by (34)), three cm from the base, were cut

for analysis on four plants from each of four transgenic ETHE1256-OE lines and 23 wild type plants. Sections were fixed in FAA (formalin:acetic acid:acetone 10:5:50) for 16

hours under vacuum, rinsed twice with ddH2O, one hour each, and dehydrated under a graded ethanol series: 20%, 30%, 50%, 70%, 95%, 100% for at least an hour each. Samples were left in 100% ethanol for further analysis. Cross sections of 50 μM relative thickness were generated on a Vibratome Series 1000 sectioning system and stored in distilled water. Cross sections were stained in 0.1% Toluidine blue and mounted in a semi-permanent gelatin-glycerine mount (10 g non- flavored gelatin, 150 ml glycerine, 0.1g sodium salycilate, 170 ml ddH20). Images were observed using light microscopy on an Olympus IX-81 fluorescence deconvolution microscope system. Data were analyzed with Image Pro Plus (Media Cybernetics) and organized with Photoshop. Statistical analysis of variance was performed using the two sample t-test assuming unequal variances (Excel).

6.3.4 N-terminal Analysis and Protein Localization

The locations of ETHE1294 and ETHE1256 were determined by subcellar fractionation of pFAST transgenic cell cultures using differential centrifugation. Harvested cells were lysed in isolation buffer (0.35 M Sorbitol, 25 mM MOPS, 0.1% BSA, 2 mM EDTA, 0.1% BME, 1% PVP-40) using a blender (2 g/5 ml buffer) and centrifuged at 2500xg for 5 minutes to remove the chloroplasts and cellular debris. The supernatant was then centrifuged at 15000 x g for 15 minutes to pellet the mitochondria. The mitochondrial pellet was washed three times in wash buffer (0.4 M Sucrose, 2 mM EDTA, 10 mM MOPS, 0.1% BSA) using alternating low and high speed spins. The mitochondrial pellet was resuspended in lysis buffer (50 mM TRIS pH 8.0, 0.5% Triton X-100) and placed on ice for 30 minutes. The lysed mitochondrial fraction was centrifuged for 30 minutes at 12000 x g. Protein was quantitated using a BCA assay (PIERCE), and ten μg was loaded onto an SDS-PAGE gel for western detection using

156 anti-FLAG antibody (Sigma). Mitochondrial cytochrome c oxidase was used as a control to monitor the purity of the fractions. Cytosolic or mitochondrial fractions from cultures over-expressing

ETHE1256FLAG or ETHE1294FLAG, respectively, were subjected to affinity-purification using anti-FLAG chromatography (Sigma) for N-terminal determination. Purified ETHE1 was eluted using FLAG peptide (Sigma F3290) and analyzed by SDS-PAGE gel and western blotting. Purified ETHE1 proteins were further resolved by SDS-PAGE. Gel slices containing the proteins were digested using trypsin (Sigma) and GluC (pH 7, Pierce), and the resulting peptides were analyzed on a MALDI-TOF mass spectrometer (Bruker Reflex III) in positive ion mode essentially as described (35). Mass spectral data were analyzed using the PAWS program at http://prowl.rockefeller.edu.

6.4 Results and Discussion:

6.4.1 Localization and N-terminal Determination Studies Arabidopsis ETHE1 was previously reported as a 256 amino acid (27.7 kDa) predicted mitochondrial protein showing similarity to GLX2 (27). Consistent with this prediction, a sequence alignment of Arabidopsis ETHE1 with the cytosolic isozyme of Arabidopsis GLX2 showed that ETHE1 contains an N-terminal extension, which resembles a mitochondrial leader peptide (Fig 6.1A-B). To determine if ETHE1 is in fact localized in the mitochondrion and to determine the ETHE1 N-terminal processing site, a construct

based on the reported ETHE1 cDNA (ETHE1256) was generated in the pFAST expression

vector. The construct, which expresses ETHE1256 with a C-terminal FLAG tag, was transformed into Arabidopsis suspension cell cultures. Fractionation studies were performed on the transgenic cell culture followed by western blot analysis using anti-

FLAG antibody to determine the subcellular location of ETHE1256. In contrast to what was expected, the fractionation study showed that ETHE1256 is clearly cytosolic (Fig 6.2A). Duplicate blots were probed with an antibody to the mitochondrial cytochrome c oxidase as a control. As expected, cytochrome c oxidase was found in the total protein and mitochondrial fractions (Fig 6.2A). It was not detected in the cytosolic fraction. Our

157

Fig. 6.2: Localization and purification of ETHE1256 and ETHE1294 from

transgenic Arabidopsis cell cultures. (A) Subcellular localization of ETHE1256 or ETHE1294 FLAG proteins in Arabidopsis cell cultures was determined through western blot analysis using cytochrome c oxidase (COX) as a mitochondrial

fraction control (B) ETHE1256 FLAG and ETHE1294 FLAG proteins purified from transgenic Arabidopsis cell cultures was detected by anti-FLAG western blot analysis

158 finding that Arabidopsis ETHE1256 is present in the cytosol is inconsistent with a previous observation that human ETHE1 is a mitochondrial protein (36), suggesting that

ETHE1256 may not represent the full-length protein. A second Arabidopsis ETHE1 cDNA was subsequently identified that encodes a protein with an additional 38 N-terminal amino acids. This cDNA, which we confirmed as full-length through RT-PCR, has the potential to encode a protein of 294 amino acids with a molecular weight of 32.3 kDa (Fig 6.1B). Therefore, transgenic cell cultures

expressing the full-length, ETHE1294 protein in the pFAST expression vector were generated and subjected to the same fractionation studies as described above. As expected, ETHE1294 was found in the total protein and mitochondrial fractions (Fig 6.2A). It was not detected in the cytosol. Interestingly, FAST-tagged proteins of

approximately 28 kDa were identified in both ETHE1256 FLAG and ETHE1294 FLAG cells suggesting that the N-terminal processing site of ETHE1294 is near the N-terminus

of ETHE1256 (Fig 6.2B). Although it was previously shown that human ETHE1 is localized to the mitochondrion, the mature N-terminus of an ETHE1 protein is not known in any

organism (36). Therefore, we purified the ETHE1256 FLAG and ETHE1294 FLAG proteins using FLAG affinity chromatography and subjected them to N-terminal analysis. Several attempts of N-terminal sequencing proved unsuccessful, likely due to an N- terminal block. Therefore, we mapped the N-terminal sequences of the proteins using peptide mapping with the endopeptidases trypsin and GluC. Peptide fragments were analyzed using MALDI-TOF mass spectrometry and the data compiled from the peak lists of both endopeptidase cleavages were mapped to the ETHE1 sequence using PAWS (Fig 6.3A-C, Table 6.1). The peptides produced good coverage and showed several overlapping peptides from the various cleavages allowing an accurate prediction of the

processed form of ETHE1 (Fig 6.3C). Peptides obtained from ETHE1256 were observed beginning at the methionine located at amino acid 39 of the full-length form of the protein (data not shown). As expected, this corresponds to the unprocessed form of the protein. Localization prediction programs (Mitoprot-2, TargetP) predicted two possible N-

terminal processing sites for ETHE1294, the first at proline 31 and the second at lysine 52

159

160

Fig. 6.3: Peptide mapping of ETHE1294 using MALDI-TOF spectroscopy. MALDI-TOF spectra collected in positive ion mode of (A) trypsin digest and (B)

Glu C digests of affinity purified ETHE1294 FLAG protein. (C) ETHE1 protein map showing peptide coverage obtained from the MALDI-TOF spectra. Amino acids in italics correspond to the C-terminal tag attached to the protein for purification.

161 The blue and green underlines represent peptide matches from trypsin and Glu C digests of ETHE1294, respectively.

162 of the full length protein. Both of these sites would produce endopeptidase cleavage

fragments with masses within range of the MALDI-TOF detection. ETHE1294 processed

at P31 would produce trypsin and GluC fragments of 826.47 Da. and 3305.67 Da.,

respectively. A processing site at K52 of ETHE1294 would not produce a fragment detectable by trypsin digest; however, it would produce a fragment with a mass of 1192.69 after GluC digestion. The majority of the endopeptidase digestion peaks found in the MALDI-TOF spectra were identified and produced good sequence coverage, yet, surprisingly none of the masses could be detected in MALDI-TOF spectra associated with either of the two predicted processing sites (Fig 6.3A-C, Table 6.1). However, two peak masses that correspond to overlapping fragments in the two endopeptidase digestions: a trypsin digest mass of 1796.23 Da. and a GluC digest mass of 2310.32 Da.,

corresponding to the peptides M39-R56 and M39-E60, respectively were identified (Fig

6.3A-B, Table 6.1). Therefore, mapping analysis of ETHE1294 indicated a leader sequence processing site at the methionine located at position 39 of the full length protein (Table 6.1).

It should be noted that while M39 was not initially predicted as a cleavage site, it is still consistent will all the sequence features necessary for the Mitochondrial Processing Peptidase (MPP) recognition and activity (37). Specifically, ETHE1 contains an Arg at the -3 position and an additional distal basic residue at the -11 position (38-40). MPP has also been shown to exhibit a preference for aromatic amino acids and, to a lesser extent, hydrophobic amino acids at position 1 (39) which is also consistent with the processing site at M39 of ETHE1. In addition, the M39 processing site contains several hydrophillic/hydroxyl amino acids at positions -2 and -3 which should enhance the activity of MPP (40). These results demonstrate that Arabidopsis ETHE1 is actually larger than previously predicted (27) and contains a 38 amino acid leader peptide that is essential for the protein’s localization within the mitochondrion. While sequencing experiments are required to confirm this prediction, there are several observations that strengthen our confidence in the result. (1) We are able to identify peptides in both the trypsin and GluC

digests corresponding to an N-terminal peptide beginning with M39. (2) Similar peptide

maps were obtained for the purified ETHE1256 and ETHE1294 proteins. (3) This

163

ETHE1294 Peptide Masses Position Experimental of peptide # mass Error peptide Sequence 1 1796.23 -1.37 39-56 MGSSSSFSSSSSKLLFRa 2 2310.32 1.8 39-60 MGSSSSFSSSSSKLLFRQLFE 3 1116.53 -0.04 63-72 SSTFTYLLAD 4 1793.93 1.06 78-93 KPALLIDPVDKTDVDRDb 5 2778.58 0.86 136-161 ASGSKALDFLEPGDKRSIGDIVLER 6 1104.22 0.33 142-151 ADLFLEPGDK 7 2348.65 0.58 142-162 ADLFLEPGDKRSIGDIVLEVR 8 1137.08 -0.47 151-161 KVSIGDIYLEc 9 1263.19 0.5 152-162 VSIGDIVLEVR 10 1193.15 0.51 187-197 MAFTGDAVLIR 10 1193.12 0.48 187-197 MAFTGDAVLIRb 11 1165.99 0.49 187-201 MAFTGDAVLIRGCGRb 12 2755.6 0.7 202-225 TDFQEGSSDQLESVHSQIFTLPK 13 1335.14 0.51 226-236 DTLIVPAHDYK 14 1110.08 -0.47 253-261 LTKDKETFK 15 1110.03 -0.42 253-261 LTKDKETFKc 16 1887.42 -1.45 258-273 ETFKTIMSNLNLSYPKa 17 1380.24 0.61 262-273 TIMSNLNLSYPK 18 1380.1 0.61 262-273 TIMSNLNLSYPKb 19 1266.35 0.34 289-299 VPSQANmdykd a=aceylated b=oxidated c=sodiated

Table 6.1: Peptide matches of ETHE1294 purified protein. Masses obtained from MALDI-TOF analysis were matched to ETHE1 using PAWS. Peptide sequences unbolded refer to peptides matched to the protein sequence of ETHE1 after digestion by trypsin. Peptide sequences bolded refer to matched peptides from Glu C digestion. The position of the peptides correspond to the amino acid numbering of ETHE1294.

164 processing site is also consistent with the observed molecular weight seen in the western blot (Fig 6.2B). Knowledge of the mature N-terminus of Arabidopsis ETHE1 may be useful in the prediction of the mature N-termini of other ETHE1-like proteins and will be important for the over-expression of recombinant protein.

6.4.2 Generation of ETHE1 Over-expressing Arabidopsis Plants

Plants that over-express either ETHE1256 or ETHE1294 were generated in order to investigate possible functional roles of ETHE1 in plant growth and development and to determine if over-expression of ETHE1, like GLX1 and GLX2 (9, 10), can enhance the growth properties of plants by providing stress resistance. Over-expression constructs of

both ETHE1256 and ETHE1294 driven by the constitutive 35S promoter were generated and transformed into wild-type Arabidopsis (Ws) plants. Four independent Basta-

resistant ETHE1256 transgenic lines and six independent Basta-resistent ETHE1294 transgenic lines were selected and confirmed by PCR. RT-PCR was performed on the individual lines and all of the transgenic plants showed significantly increased levels of ETHE1 transcript in comparison to wild type plants (Fig 6.4 A-B).

6.4.3 ETHE1256-OE Plants Show Resistance to High Concentrations of Valine While the biochemical role of ETHE1 is not known in any organism, physiological changes observed in patients with EE have been used to predict potential biochemical roles for the protein. Mutations in human ETHE1 result in elevated levels of

C4 and C5 plasma acylcarnitines and markedly elevated urinary excretion of ethylmalonic

acid, and C4-6 acylglycines (36). Ethylmalonic acid is primarily derived from the carboxylation of butyryl-CoA, which is derived from the β-oxidation of short chain fatty acids, or from 2-ethylmalonic-semialdehyde, the final product of the R-pathway for the catabolism of isoleucine (Fig 6.5) (41, 42). This raised the possibility that human ETHE1 and by analogy Arabidopsis ETHE1 may be involved in the removal of a hydroxyacid or CoA ester that is formed as part of amino acid and/or lipid metabolism. We hypothesized that if ETHE1 is actually involved in the detoxification of byproducts from amino acid metabolism, then over-expression of the enzyme may eliminate the toxic effects observed when plants are grown in the presence of high levels of certain amino acids or their toxic

165

Fig. 6.4: Expression analysis of transgenic plants over-expressing either

ETHE1256 or ETHE1294. RT-PCR analysis of ETHE1 transcript levels in

Arabidopsis plants over-expressing either (A) ETHE1256 or (B) ETHE1294. ACT 8 was used as an internal control.

166 intermediates. In order to test this hypothesis we examined the sensitivity of ETHE1-OE lines to valine, propionate, ethylmalonic acid, butyric acid, and crotonic acid (Fig 6.5).

No differences were observed between ETHE1294-OE and wild-type plants at the various concentrations tested for the different metabolites (data not shown). These results suggest that ETHE1 may not normally be involved in the removal of toxic metabolites of branched chain amino acid metabolism or the β-oxidation of fatty acids. However, these results also do not rule out this possibility. Even though no differences were observed in plants over-expressing the full-

length, ETHE1294 protein, ETHE1256-OE plants did exhibit growth differences on plates containing exogenous valine (Fig 6.6 A-B). Valine concentrations of 1.5 mM and greater are toxic to wild-type plants, inhibiting root growth by approximately 80% over 7 days of

growth. The same level of valine only inhibited the growth of ETHE1256-OE plants by

about 40% (Fig 6.6A-B). However, in contrast to our predictions ETHE256-OE plants displayed the same sensitivity to propionate, ethylmalonic acid, butyric acid, and crotonic acid as wild-type plants (data not shown). Mutations in acetolactate synthase (ALS), an enzyme that catalyses the first common step of branched chain amino acid biosynthesis can confer resistance to high levels of valine (43, 44). ALS is regulated through feedback inhibition by the end products of the pathway (45). Complete inhibition of ALS leads to plant death primarily through starvation for essential amino acids (46); however single point mutations that reduce the binding affinity of ALS to its inhibitors have been reported that have little effect on its native enzymatic function, while providing plants with resistance to exogenous sources of valine (43, 47). Exogenous valine toxicity can also arise from methylacrylyl-CoA, a toxic degradation product, which can accumulate in the presence of high concentrations of valine (48-50). There is also evidence for the production of methylmalonate in patients compromised in methylmalonic semialdehyde dehydrogenase which catalyses the conversion of methylmalonate semialdehyde to propionyl-CoA (51, 52). Methylmalonate is a known toxin whose toxic effects have been studied extensively in patients displaying methylmalonic aciduria (reviewed in (53). Preliminary studies have shown that methylmalonate is also toxic in plants (54). Given the similarity between

167

Fig. 6.5: Metabolic Routes of EMA Production. EMA levels can be elevated by the accumulation of butyryl-CoA, which can be carboxylated through propionyl-CoA carboxylase to ethylmalonyl-CoA. This is known to occur in

168 disorders of short-chain fatty acid β-oxidation pathways as well as through alterations in R-isoleucine catabolism. Metabolites labeled with asterisks were used in the metabolic stress studies.

169 methylmalonate and ethylmalonate and the excretion of ethylmalonic acid observed in EE patients, it is possible that ETHE1 be directly/indirectly involved in the hydrolyzing methylmalonate or one of its precursors, which in turn could promote the increased resistance of exogenous valine. However, why high levels of cytoplasmic ETHE1 can confer resistance to inhibitory levels of valine, while the mitochondrial protein can not, is not clear at this time. Further studies are required to investigate this question.

6.4.4 Over-expression of ETHE1256 Leads to Earlier Bolting and Flowering in Arabidopsis

Plants that over-express cytosolic, ETHE1256, and mitochondrial, ETHE1294, display drastically different phenotypes. ETHE1294-OE plants appear normal in every respect. At this time we are unable to identify any specific effects of over-expressing

ETHE1294. In contrast, plants that over-express ETHE1256 display several changes in their growth properties. The phenotypes described below were observed over three generations and in several independent transgenic lines, indicating the changes are due to the over-expression of ETHE1256 and are relatively stable. Wild-type plants and

ETHE1256-OE transgenic seedlings grown on solid MS growth medium for fourteen days did not exhibit any noticeable differences. Likewise no differences were observed between the two lines during the first eighteen days of growth in soil (data not shown).

However ETHE1256-OE plants bolted significantly earlier than wild-type plants grown

under the same conditions (Fig 6.7A-C). ETHE1256-OE plants typically bolted 19 days

(+/- 1) after germination while wild-type and ETHE1294-OE plants did not bolt until day

24 (+/- 1) (Fig 6.7A-D). Consistent with the earlier bolting time, ETHE1256-OE plants also flowered significantly (P<0.05) earlier when compared to wild-type plants (Fig

6.7E). In contrast to wild-type and ETHE1294-OE plants that typically flowered 29 days

(+/- 2) after germination, ETHE1256-OE plants flowered on average 23 (+/- 1) days after germination (Fig 6.7E-F). The timing of flowering is an important developmental event that contributes to crop productivity, especially in regions with short growing seasons. If a plant moves from vegetative growth into its reproductive stage too early, seed yield can be limited due to an inadequate supply of energy from the reduced number of roots and leaves.

170

Fig. 6.6: Effect of exogenous valine on seedling growth. (A) Light microscopy images of 7 day old ETHE1256-OE and wild-type seedlings grown on MS plates containing exogenous valine in concentrations from 0 mM to 1.5 mM. (B) Measurement of root lengths of plants obtained from the valine study in (A). These studies were repeated in triplicate and represent the average +/- SD of 50 plants per concentration and plant line.

171 Alternately, if the plant moves too slowly from its vegetative state to reproductive growth, it may not have enough time to produce mature seeds (55). There are several cues that guide the transition from vegetative to reproductive growth, including environmental cues such as photoperiod and vernalization, as well as phytohormonal cues, such as abscisic acid (ABA), cytokinin (CK) and gibberellin (GA) levels (reviewed in (56). ABA is generally described as an inhibitor of flowering (57), which is consistent with the early flowering phenotype of Arabidopsis mutants that are deficient in or insensitive to ABA (58). CKs influence cell division and shoot formation, and are classified as promoters of flowering (56). Transgenic plants deficient in CKs flower late (59), whereas plants that are enriched in CKs have been shown to flower early (60, 61). Exogenous treatment with GAs has also been shown to cause flower formation (62) and GA-deficient and insensitive mutants of Arabidopsis provide evidence that GAs are required for flowering under short days (63). The early bolting and flowering observed

in ETHE1256–OE plants therefore could result from elevated levels of either CKs or GA.

The observation that ETHE1256 but not ETHE1294 can promote the earlier flowering suggests that the cytoplasmic form of ETHE1 may catalyze a reaction that is different from that of the native, mitochondrially localized enzyme, or that it acts on a substrate that is not accessible to the mitochondrial enzyme.

6.4.5 ETHE1256 Expression Enhances Seed Yield in Arabidopsis

In addition to flowering earlier than wild-type plants, ETHE1256-OE plants also exhibited delayed senescence, flowering 6 days longer than wild-type plants (Fig 6.8A).

ETHE1256-OE plants typically stopped flowering 42 (+/- 3) days after germination. In contrast, wild-type plants stopped flowering 36 (+/- 3) days after germination. In the absence of environmental cues such as osmotic stress or pathogenic attack, the onset of senescence is generally determined by the age of the plant; however, this process, similar to the onset of flowering, can be influenced by the levels of certain phytohormones (64). The most widely known hormones involved in controlling senescence are CKs (64). Studies have shown delayed senescence upon the application of CKs (65) as well in transgenic plants overproducing a key enzyme of cytokinin biosynthesis, isopentenyl

transferase (ipt) (66-68). The delayed senescence phenotype of ETHE1256-OE plants

172

Fig. 6.7: The effect of over-expressing either ETHE1256 or ETHE1294 on plant

growth in Arabidopsis. Plants on day 19 of growth are shown for ETHE1256-OE

(A) and ETHE1294-OE (B) respectively. The average number of days to bolting

173 for wild-type ETHE1256-OE (C), and ETHE1294-OE plants (D). The average

number of days to flowering are shown for wild-type, ETHE1256-OE (E) and

ETHE1294-OE (F). The values are represented as the mean +/- SD.

174 along with the early bolting and flowering suggests that ETHE1256-OE plants may contain increased endogenous CK levels.

Because ETHE1256–OE plants display both earlier and longer flowering times compared with wild-type plants, we predicted that they would have a greater seed yield as well. The number of seeds per individual silique, and the number of siliques each plant produced were measured on plants from the individual transgenic lines and compared with wild-type plants (Fig 6.8B-D). No statistically significant differences were observed in the number of seeds per silique produced between wild-type and

ETHE1256-OE plants (Fig 6.8B). It should be noted that while ETHE1256-OE line 1 appears to show an increase in the number of seeds produced per silique (42 +/- 12) in comparison to wild-type (36 +/- 6) a statistical t-test did not show a significant difference (P<0.05). A significant difference was however observed in the number of siliques

produced in ETHE1256-OE plants compared with wild-type plants (Fig 6.8C). Consistent

with an increase in the duration of flowering, ETHE1256-OE plants produced on average

31% more siliques than wild-type plants (ETHE1256-OE 59 siliques: wild-type 45

siliques) (Fig 6.8C). These results demonstrate that high level expression of ETHE1256 in the cytoplasm increases the reproductive life time of the plant and overall seed yield.

6.4.6 Arabidopsis ETHE1256-OE Plants Have Thicker Primary Inflorescence Stems

Arabidopsis plants expressing ETHE1256 are taller and also display thicker, more upright primary inflorescence stems than wild-type plants at the same stage of growth. Consistent with this observation we found an overall increase in dry weight of 24% in the

ETHE1256-OE plants (ETHE1256-OE 60 mg per plant: wild-type 49 mg) (Fig 6.8D). The

growth habit of ETHE1256-OE plants was investigated further by examining cross-

sections of the basal region of the primary inflorescence of both wild-type and ETHE1256- OE transgenic plants at the same stage of growth (Fig 6.8E). Analysis of cross-sections demonstrated a significant difference in the area of the primary inflorescence between plants over-expressing ETHE1256 and wild-type plants (Fig 6.8E). ETHE1256-OE plants were found to have on average a 28% increase in stem area compared with wild-type 2 2 plants (ETHE1256-OE: 920.8 mm ; wild-type: 666.1 mm ) (Fig 6.8F). Upon closer examination, no significant differences in the interfasicular fibers or the number of

175

Fig. 6.8: The effect of ETHE1256-OE in Arabidopsis on senescence, seed yield, dry mass, and inflorescence stem thickness. The effect of over-

expression of ETHE1256 on time to senescence (A), Number of seeds per silique (B), the mean number of siliques per plant (C), and the mean dry mass of the full grown plant (D) was determined. Cross-sectional analysis of the primary

inflorescence of ETHE1256-OE (E) shows an increased area compared with wild- type plants (F). Measurements represent the mean area +/- SD.

176 vascular bundles were observed between the ETHE1256-OE and wild-type plants (data not shown). The increase in the area of the primary inflorescence observed in plants expressing ETHE1256-OE appears to be primarily a result of cell expansion in the central parenchyma (Fig 6.8E). A similar increase in primary inflorescence area through parenchyma expansion has previously been observed in transgenic tobacco plants over- expressing ipt (69). The parenchyma cells are involved in photosynthesis, storage, secretion, movement of water, and transport of food depending on their localization within the plant body and therefore, expansion and enlargement of these cells could potentially provide enhanced growth properties.

6.4.7 ETHE1256 Over-expression in Nicotiana tabacum Results in Enhanced Growth Properties To determine if the growth enhancing properties observed from the over-

expression of ETHE1256 in Arabidopsis plants can be replicated in other species, an

ETHE1256 over-expression construct was introduced into tobacco plants. Transgenic plants were identified through kanamycin resistance and confirmed through PCR. Six lines were identified as positive for both. RT-PCR was performed on the individual lines

and all ETHE1256-OE transgenic plants showed increased levels of the ETHE1 transcript

in comparison to wild type plants, with tobacco lines ETHE1256-OE 8, 9, 10, and 36 showing the greatest increase in mRNA levels (Fig 6.9A). The phenotypes described below represent data from two generations of the six lines and include data from 15 plants from each line.

Tobacco plants over-expressing ETHE1256 displayed phenotypes similar to those

observed in Arabidopsis ETHE1256-OE lines. In general, plants that had higher levels of ETHE1 mRNA differed most from wild-type tobacco plants. Consistent with the phenotype observed in Arabidopsis ETHE1256-OE plants, tobacco transgenic lines that

over-express ETHE1256 bolted on average 16 +/- 3 days after germination on soil compared to 21 +/- 1 days in tobacco plants containing an empty vector control (Fig 6.9

B-C). Likewise, ETHE1256-OE tobacco plants flowered earlier, on average of 61 +/- 10 days after germination, in contrast to the 77 +/- 12 days after germination observed in the control plants (Fig 6.9D). Interestingly, in contrast to the increased time to senescence

177

Fig 6.9: Effects of over-expressing Arabidopsis ETHE1256 in Nicatiana

tabacum. (A) RT-PCR of ETHE1 in ETHE1256 tobacco plants transcript levels.

(B) Comparison of tobacco ETHE1256-OE and control plants at day 17 of growth. The times to bolting, flowering, and senescence measurements are given in

178 graphs (C), (D), and (E) respectively of ETHE1256-OE tobacco and the empty vector control plants. (F) Comparison of tobacco ETHE1256 and control plants at day 90 of growth. Mean (+/- SD) measurement of seed yield is given in (G). Statistically significant differences (p< 0.05) are indicated by an asterisk.

179 observed in ETHE1256-OE Arabidopsis plants, no significant difference was observed in

the time to senescence between ETHE1256-OE tobacco plants and the empty vector control (Fig 6.9E). Perhaps the most dramatic phenotype of ETHE1256-OE tobacco plants

was the increase in overall seed yield (Fig 6.9F-G). ETHE1256-OE tobacco plants had on

average a 22% increase in seed yield (ETHE1256-OE: 12,335 seeds: Control: 9,682 seeds) (Fig 6.9F-G). The most dramatic increase in seed yield was observed in line 8, which exhibited a 34% average overall increase in seed yield. The increase in seed yield appears to be due to both a shorter time to flowering

and a loss of apical in tobacco ETHE1256-OE plants (Fig 6.9F). Consistent

with our prediction that over-expression of the ETHE1256 protein may be altering cytokinin levels, it is well established that cytokinins along with auxin are capable of controlling apical dominance. High cytokinin to auxin ratios have been shown to result in a loss of apical dominance resulting in the production of more lateral branching (70).

Our studies on transgenic tobacco plants that over-express ETHE1256 demonstrate that the enhanced growth properties observed in Arabidopsis can be conferred to other plants, and further suggest that cytoplasmic ETHE1 may increase cytokinin levels.

6.4.8 Concluding remarks We recently showed using an ETHE1 is loss of function mutant that ETHE1 essential for early seed development and may have a potential role in β-oxidation of short chain fatty acids (Holdorf, unpublished). However, biochemical studies on ethe1 loss of function mutants are difficult due to seed lethality. In the results presented here we demonstrated that Arabidopsis ETHE1 contains a 38 amino acid leader sequence that is essential for its proper processing and native localization in the mitochondrion. Over- expression of native ETHE1 in Arabidopsis resulted in no obvious phenotypic changes under normal growth conditions. However, given the elevated levels of ETHE1 mRNA observed under abiotic stresses (Chapter 4), further experiments are necessary to

determine if over-expressing ETHE1294 might confer increased stress tolerance. Surprisingly, we found that high levels of cytosolic ETHE1 leads to changes in the growth properties of both Arabidopsis and tobacco. Specifically, plants that over-

express ETHE1256 exhibit resistance to exogenous valine, flower earlier and longer, have

180 a more upright growth habit, and produce more seed. These affects are not observed in

plants that over-express ETHE1294 suggesting that cytosolic ETHE1 may catalyze a reaction not normally catalyzed by the native mitochondrial enzyme or that is has access to a substrate not found in the mitochondrion. The phenotypes exhibited by ETHE1256- OE plants resemble those of plants containing mutations that increase CK levels, suggesting that cytoplasmic ETHE1 may indirectly alter CK levels within the plant. The natural ETHE1 substrate is not yet known. However, further experiments using the ETHE1 over-expression plants and cell cultures should allow us to directly test this possibility. Finally, we have shown that cytoplasmic over-expression of ETHE1 can enhance the growth properties of both Arabidopsis and tobacco plants and increase seed yield. This raises the possibility that ETHE1 could ultimately be used to increase seed yield in agriculturally important plants.

181 6.5 References

1. Altman, A. (1999). Plant Biotechnology in the 21st Century: the Challenges Ahead. Electronic Journal of Biotechnology 2, 51-55

2. Miflin, B. (2000). Crop Improvement in the 21st Century. J. Exp. Bot. 51, 1-8

3. Harper, J. L. (1977) Population biology of plants, London England, Academic Press.

4. Primack, R. B. (1978). Regulation of Seed Yield in Plantago. Journal of Ecology 66, 835-847

5. Primack, R. B., and Antonoics, J. (1981) Evolution 111, 451-461

6. Pritts, M. P., and Hancock, J. F. (1984). Independence of Life History Parameters in Populations of Vaccinium angustifolium (Ericaceae). Bullentin of Torrey Botanical Club 111, 451-461

7. Pritts, M. P., and Honcock, J. F. (1985). Lifetime Biomass Partitioning and Yield Component Relationships in the High-bush Blueberry, Vaccinium corymbosum L. (Ericaceae). American Journal of Botony 72, 446-452

8. Singh, R. P., USHA, Rizvi, S. M. H. et al. (2003) Genetic engineering for enhancing abiotic stress tolerance. In Improvement Strategies of Leguminosae Biotechnology pp. 223-245. Edited by P. Jaiwal, & R. P. Singh, Springer.

9. Singla-Pareek, S. L., Reddy, M. K., and Sopory, S. K. (2003). Genetic Engineering of the Glyoxalase Pathway Enzymes Grow and Set Viable Seeds in Zinc-spiked Soils. Proc. Natl. Acad. Sci. U. S. A. 100, 14672-14677

10. Singla-Pareek, S. L., Yadav, S. K., Pareek, A., Reddy, M. K., and Sopory, S. K. (2006). Transgenic Tobacco Overexpressing Glyoxalase Pathway Enzymes Grow and Set Viable Seeds in Zinc-spiked Soils. Plant Physiol. 140, 613-623

182 11. Yadav, S. K., Singla-Pareek, S. L., Ray, M., Reddy, M. K., and Sopory, S. K. (2005). Methylglyoxal Levels in Plants under Salinity Stress are Dependent on Glyoxalse I and Glutathione. Biochem. Biophys. Res. Commun. 337, 61-67

12. Yadav, S. K., Singla-Pareek, S. L., Reddy, M. K., and Sopory, S. K. (2005). Transgenic Tobacco Plants Overexpressing Glyoxalase Enzymes Resist an Increase in Methylglyoxal and Maintain Higher Reduced Glutathione Levels under Salinity Stress. FEBS Lett. 579, 6265-6271

13. Thornalley, P. J. (1993). The Glyoxalse System in Health and Disease. Mol. Aspects Med. 14, 287-371

14. Thornalley, P. J. (1990). The Glyoxalase System: New Developments Towards Functional Characterization of a Metabolic Pathway Fundamental to Biological Life. Biochem. J. 269, 1-11

15. Papoulis, A., al-Abed, Y., and Bucala, R. (1995). Identification of N-(1- carboxyethyl)guanine (CEG) as a Guanine Advanced Glycosylation End Product. Biochemistry 34, 648-655

16. Lo, T. W., Westwood, M. E., McLellan, A. C., Selwood, T., and Thornalley, P. J. (1994). Binding and Modification of Proteins by Methylglyoxal under Physiological Conditions. A kinetic and Mechanistic Study with N Alpha-acetylarginine, N Alpha- acetylcysteine, and N Alpha-acetyllysine, and Bovine Serum Albumin. J. Biol. Chem. 269, 32299-32305

17. Rahman, A., Shahabuddin, and Hadi, S. M. (1990). Formation of Strand Breaks and Interstrand Cross-links in DNA by Methylglyoxal. J. Biochem. Toxicol. 5, 161-166

18. Mannervik, B., and Ridderstrom, M. (1993). Catalytic and Molecular Properties of Glyoxalase I. Biochem. Soc. Trans. 21, 515-517

19. Allen, R. E., Lo, T. W., and Thornalley, P. J. (1993). Purification and Characterization of Glyoxalase II from Human Red Blood Cells. Eur. J. Biochem. 213, 1261-1267

183 20. Clugston, S. L., Daub, E., Kinach, R., Miedema, D., Barnard, J. F., and Honek, J. F. (1997). Isolation and Sequencing of a Gene Coding for Glyoxalase I Activity from Salmonenlla typhimurium and Comparison with other Glyoxalase I Sequences. Gene 186, 103-111

21. Creighton, D. J., and Hamilton, D. S. (2001). Brief History of Glyoxalase I and What We Have Learned about Metal Ion-dependent, Enzyme-catalyzed Isomerizations. Arch. Biochem. Biophys. 387, 1-10

22. Inoue, Y., and Kimura, A. (1996). Identification of the Structural Gene for Glyoxalase I from Saccharomyces cerevisiae. J. Biol. Chem. 271, 25958-25965

23. Lu, T., Creighton, D. J., Antoine, M., Fenselau, C., and Lovett, P. S. (1994). The Gene Encoding Glyoxalase I form Pseudomonas putida: Cloning, Overexpression, and Sequence Comparisons with Human Glyoxalase I. Gene 150, 93-96

24. Rhee, H., and Kimura, A. (1988). Glyoxalase I and its Application to Production of S-lactoylglutathione. Tanpakushitsu Kakusan Koso 33, 1610-1614

25. Ridderstrom, M., and Mannervik, B. (1996). The Primary Structure of Monomeric Yeast Glyoxalase I Indicates a Gene Duplication Resulting in Two Similar Segments Homologous with the Subunit of Dimeric Human Glyoxalase I. Biochem. J. 316 ( Pt 3), 1005-1006

26. Bito, A., Haider, M., Hadler, I., and Breitenbach, M. (1997). Identification and Phenotypic Analysis of Two Glyoxalase II Encoding Genes from Saccharomyces cerevisiae, GLO2 and GLO4, and Intracellular Localization of the Corresponding Proteins. J. Biol. Chem. 272, 21509-21519

27. Maiti, M. K., Krishnasamy, S., Owen, H. A., and Makaroff, C. A. (1997). Molecular Characterization of Glyoxalase II from Arabidopsis Thaliana. Plant Mol. Biol. 35, 471-481

184 28. Cordell, P. A., Futers, T. S., Grant, P. J., and Pease, R. J. (2004). The Human Hydroxyacylglutathione Hydrolase (HAGH) Gene Encodes Both Cytosolic and Mitochondrial Forms of Glyoxalase I. J. Biol. Chem. 279, 28653-28661

29. Zimmerman, P., Hirscho-Hoffman, M., Hennig, L., and Gruissem, W. (2005). Genevestigator. Arabidopsis Microarray Database and Analysis Toolbox. Trends Plant Sci.

30. Ge, X., Dietrich, C., Matsuno, M., Li, G., Berg, H., and Xia, Y. (2005). An Arabidopsis Aspartic Protease Functions as an Anti-cell-death Component in Reproduction and Embryogenesis. EMBO Rep. 6, 282-288

31. Hajdukiewicz, P., Svab, Z., and Maliga, P. (1994). The Small, Versatile pPZP family of Agrobacterium Binary Vectors for Plant Transformation. Plant Mol. Biol. 25, 989- 994

32. Dandekar, A. M., and Fisk, H. J. (2005). Plant Transformation: Agrobacterium- mediated Gene Transfer. Methods Mol. Biol. 286, 35-46

33. An, Y. Q., McDowell, J. M., Huang, S., McKinney, E. C., Chambliss, S., and Meagher, R. B. (1996). Strong, Consitutive Expression of the Arabidopsis ACT2/ACT8 Actin Subclass in Vegetative Tissues. Plant J. 10, 107-121

34. Altamura, M. M., Possenti, M., Matteucci, A., Baima, S., Ruberti, I., and Morelli, G. (2001). Development of the Vascular System in the Inflorescence Stem of Arabidopsis. New Phytologist 151, 381-389

35. Nilsson, C. L., Cooper, H. J., Hakansson, K., Marshall, A. G., Ostberg, Y., Lavrinovicha, M., and Bergstrom, S. (2002). Characterization of the P13 Membrane Protein of Borrelia burgdorferi by Mass Spectrometry. J. Am. Soc. Mass Spectrom. 13, 295-299

36. Tiranti, V., D'Adamo, P., Briem, E. et al. (2004). Ethylmalonic Encephalopathy is Caused by Mutations in ETHE1, a Gene Encoding a Mitochondrial Matrix Protein. Am. J. Hum. Genet. 74, 239-252

185 37. Ito, A. (1998). Processing of Mitochondrial Protein Precursors. Seikagaku 70, 1401- 1412

38. Shimizu, H., Ito, M., Miyahara, M., Ichikawa, K., Okubo, S., Konishi, T., Naka, M., Tanaka, T., Hirano, K., and Hartshorne, D. J. (1994). Characterization of the Myosin-binding Subunit of Smooth Muscle Myosin Phosphatase. J. Biol. Chem. 269, 30407-30411

39. Ogishima, T., Niidome, T., Shimokata, K., Kitada, S., and Ito, A. (1995). Analysis of Elements in the Substrate Required for Processing by Mitochondrial Processing Peptidase. J. Biol. Chem. 270, 30322-30326

40. Song, M. C., Ogishima, T., and Ito, A. (1998). Importance of Residues Carboxyl Terminal Relative to the Cleavage Site in Substrates of Mitochondrial Processing Peptidase for their Specific Recognition and Cleavage. J. Biochem. 124, 1045-1049

41. Mamer, O. A., Tjoa, S. S., Scriver, C. R., and Klassen, G. A. (1976). Demonstration of a New Mammalian Isoleucine Catabolic Pathway Yielding an Rseries of Metabolites. Biochem. J. 160, 417-426

42. Nowaczyk, M. J., Lehotay, D. C., Platt, B. A., Fisher, L., Tan, R., Phillips, H., and Clarke, J. T. (1998). Ethylmalonic and Methylsuccinic Aciduria in Ethylmalonic Encephalopathy Arise from Abnormal Isoleucine Metabolism. Metabolism 47, 836- 839

43. Hervieu, F., and Vaucheret, H. (1996). A Single Amino Acid Change in Acetolactate Synthase Confers Resistance to Valine in Tobacco. Mol. Gen. Genet. 251, 220-224

44. Relton, J. M., Wallsgrove, R. M., Bourgin, J. P., and Bright, S. W. J. (1986). Altered Feedback Sensitivity of Acetohydroxyacid Synthase from Valine-resistant Mutants of Tobacco. Planta 169, 46-50

45. Miflin, B. J. (1971). Cooperative Feedback Control of Barley Acetohydroxyacid Synthetase by Leucine, Isoleucine, and Valine. Arch. Biochem. Biophys. 146, 542- 550

186 46. Shaner, D. L., Anderson, P. C., and Stidham, M. A. (1984). Imidazolinones: Potent Inhibitors of Acetohydroxyacid Synthase. Plant Physiol. 76, 545-546

47. Tranel, P. J., and Wright, T. R. (2002). Resistance of Weeds to ALS-inhibiting Herbicides: What have we learned? Weed Science 50, 700-712

48. Lucas, K. A., Filley, J. R., Erb, J. M., Graybill, E. R., and Hawes, J. W. (2007). Peroxisomal Metabolism of Propionic Acid and Isobutyric Acid in Plants. J. Biol. Chem. 282, 24980-24989

49. Lange, P. R., Eastmond, P. J., Madagan, K., and Graham, I. A. (2004). An Arabidopsis Mutant Disrupted in Valine Catabolism is also Compromised in Peroxisomal Fatty Acid Beta-oxidation. FEBS Lett. 571, 147-153

50. Zolman, B. K., Monroe-Augustus, M., Thompson, B., Hawes, J. W., Krukenberg, K. A., Matsuda, S. P., and Bartel, B. (2001). chy1, an Arabidopsis Mutant with Impaired Beta-oxidation, is Defective in a Peroxisomal Beta-hydroxyisobutyryl-CoA Hydrolase. J. Biol. Chem. 276, 31037-31046

51. Roe, C. R., Struys, E., Kok, R. M., Roe, D. S., Harris, R. A., and Jakobs, C. (1998). Methylmalonic Semialdehyde Dehydrogenase Deficiency: Psychomotor Delay and Methylmalonic Aciduria Without Metabolic Decompensation. Mol. Genet. Metab. 65, 35-43

52. Chambliss, K. L., Gray, R. G., Rylance, G., Pollitt, R. J., and Gibson, K. M. (2000). Molecular Characterization of Methylmalonate Semialdehyde Dehydrogenase Deficiency. J. Inherit. Metab. Dis. 23, 497-504

53. Deodato, F., Boenzi, S., Santorelli, F. M., and Dionisi-Vici, C. (2006). Methylmalonic and Propionic Aciduria. Am. J. Med. Genet. C. Semin. Med. Genet. 142, 104-112

54. Chaves, N., Sosa, T., Alias, J. C., and Escudero, J. C. (2001). Identification and Effects of Interaction Phytotoxic Compounds from Exudate of Cistus Ladanifer Leaves. J. Chem. Ecol. 27, 611-621

187 55. Franke, D. F., Ellis, A. G., Dharjwa, M., Freshwater, M., Fujikawa, M., Padron, A., and Weis, A. E. (2006). A Steep Cline in Flowering Time for Brassica rapa in Southern California: Population-level Variation in the Field and the Greenhouse. Int. J. Plant Sci. 167, 83-92

56. Bernier, G., and Perilleux, C. (2005). A Physiological Overview of the Genetics of Flowering Time Control. Plant. Biotechnol. J. 3, 3-16

57. Bernier, G. (1988). The control of Floral Evocation and Morphogenesis. Annu. Rev. Plant Physiol. Plant Mol. Biol. 39, 175-219

58. Martinez-Zapater, J. M., Coupland, G., Dean, C. et al. (1994) The Transition to Flowering in Arabidopsis. In Arabidopsis pp. 403-433. Edited by E. M. Meyerowitz, & C. R. Somerville, Cold Spring Harbor, Cold Spring Harbor Laboratory Press.

59. Werner, T., Motyka, V., Laucou, V., Smets, R., Van Onckelen, H., and Schmulling, T. (2003). Cytokinin-deficient Transgeneic Arabidopsis Plants show Multiple Developmental Alterations Indicating Opposite Functions of Cytokinins in the Regulation of Shoot and Root Meristem Acitivity. Plant Cell 15, 2532-2550

60. Chaudhury, A. M., Letham, S., Craig, S., and Dennis, E. S. (1993). Amp1-A Mutant with High Cytokinin Levels and Altered Embryonic Patter, Faster Vegetative Growth, Constitutive Photomorphogenesis and Precocious Flowering. Plant J. 4, 907-916

61. Srivastava, S., Emery, R. J. Neil, Rahman, M. H., and Kav, Nat N. V. (2007). A Crucial Role for Cytokinins in Pea ABR-17 Mediated Enhanced Germination and Early Seedling Growth of Arabidopsis thaliana Under Saline and Low Temperature Stresses. J. Plant Growth Regul. 26, 26-37

62. Zeevaart, J. A. D. (1983) Gibberellians and flowering. In The Biochemistry and Physiology of Gibberellians pp. 333-374. Edited by A. Crozier, New York, Praeger.

188 63. Wilson, R. N., Heckman, J. W., and Somerville, C. R. (1992). Gibberellin is Required for Flowering in Arabidopsis thaliana under Short Days. Plant Physiol. 100, 403-408

64. Buchanan-Wollaston, V., Earl, S., Harrison, E., Mathas, E., Navabpour, S., Page, T., and Pink, D. (2003). The Molecular Analysis of Leaf Senescence—a Genomics Approach. Plant. Biotechnol. J. 1, 3-22

65. Van Staden, J., Cook, E. L., and Nooden, L. D. (1988) Cytokinins and Senescence. In Senescence and Aging in Plants pp. 281-328. Edited by L. D. Nooden, & A. C. Leopold, San Diego, CA, Academic Press.

66. Estruch, J. J., Prinsen, E., VAN Onckelen, H., Schell, J., and Spena, A. (1991). Viviparous Leaves Produced by Somatic Activation of an Inactive Cytokinin- Synthesizing Gene. Science 254, 1364-1367

67. Li, Y., Hagen, G., and Guilfoyle, T. J. (1992). Altered Morphology in Transgenic Tobacco Plants that Overproduce Cytokinins in Specific Tissues and Organs. Dev. Biol. 153, 386-395

68. Smart, C. M., Scofield, S. R., Bevan, M. W., and Dyer, T. A. (1991). Delayed Leaf Senescence in Tobacco Plants Transformed with tmr, a Gene for Cytokinin Production in Agrobacterium. Plant Cell 3, 647-656

69. Van Loven, K., Beinsberger, Susy E. I., Valcke, R. L., Van Onckelen, Henri A., and Clijsters, Herman M. M. (1993). Morphometric Analysis of the Growth of Phsp70- ipt Transgenic Tobacco Plants. Journal of Experimental Botany 44, 1671-1678

70. Nijhoff, M. (1987) Hormonal regulation of apical dominance. In Plant Hormones and Their Role in Plant Growth and Development Edited by P. J. Davies, Boston, MA,

189

Chapter 7

Conclusions

7.1 Scientific problems addressed in this dissertation

It has been shown that mutations in the gene ETHE1 are directly linked to the autosomal recessive disorder Ethylmalonic Encephalopathy in which patients display symptoms of symmetric brain lesions, lactic academia, elevated excretion of ethylmalonic acid, and death in the first decade of life (1-4). ETHE1-like genes are found in a wide range of organisms; however, the biochemical and physiological roles of this protein are still unknown. In order to investigate this question and determine a functional role(s) for this protein, a structural study was conducted on Arabidopsis ETHE1 homolog in which we (1) obtained the first crystal structure of an ETHE1-like protein (2) performed a full metal analysis (3) tested potential substrates. In addition, we also initiated studies on ETHE1’s functional role(s) in Arabidopsis that involved the characterization of an ethe1 knock out mutant including a full knock out study as well determining the effect of knocking down ETHE1 mRNA levels at different time points in wild-type plant and cell culture growth. Lastly, we analyzed the effects of over- expressing ETHE1 and discovered a novel role for ETHE1 when it is targeted to the cytoplasm. This dissertation represents the first full characterization of an ETHE1-like protein in any organism.

7.2 Structural analysis of ETHE1 in Arabidopsis

Results presented here represent the first crystal structure of an ETHE1-like protein (5). ETHE1 had previously been predicted to be a glyoxalase II-like enzyme (GLX2); however, using recombinant enzyme that contained a predicted N-terminus, we

190 were able to obtain a crystal structure that was structurally similar to that of the GLX2 enzymes (5-7), but exhibited several key differences that set the two families apart (5, 6). Notable differences included ETHE1’s existence as a dimer as the presence of a unique active site (5-7). We further showed that ETHE1 does not use SLG as a substrate but does exhibit weak esterase activity. In contrast to GLX2 enzymes which tightly bind two metal for full activity, ETHE1 only binds a single Fe(II) atom (Chapter 3, (6, 8)). Because ETHE1’s substrate is unknown, perhaps the most important finding in the dissertation is the determination of the processed N-terminus (Chapter 6). In this dissertation, it was shown that ETHE1 is localized in the mitochondrion and through MALDI-TOF peptide mass finger printing we determined the proper processing site for the mature ETHE1 protein (Chapter 6). This provides the first known mature N-terminus processing site for ETHE1 in any organism. The actual ETHE1 processing site was found to be 11 amino acids upstream from the predicted cleavage site used in our recombinant studies. Given the high level of structural similarity to the GLX2 enzymes and ETHE1’s solvent exposed N-terminus, it is not expected that the protein structure will fluctuate greatly from the solved crystal. However, experiments are currently underway to express the correct mature protein in a recombinant system for use in substrate analysis and metal binding studies.

7.3 Functional analysis of ETHE1 in Arabidopsis

Prior to this dissertation, no role for ETHE1 in any organism was known nor had a functional study of ETHE1 been initiated. All predicted pathways and functions of ETHE1 were based on the symptoms of patients diagnosed with EE (1, 2, 9, 10). The most noted and unique symptom of EE is the presence of elevated levels of ethylmalonic acid (EMA) excreted in the patient’s urine, suggesting a metabolic disorder (3, 4). Production of EMA can come from at least two possible pathways; disorders of the short- chain fatty acid cycle resulting in a build-up of butyryl-CoA, or the R-isoleucine catabolism pathway from 2-ethylmalonic-semialdehyde (9, 10). Therefore, it has been predicted that ETHE1 plays a role in either isoleucine catabolism or the β-oxidation of short-chain fatty acids (9, 10). To address these two possibilities, we characterized

191 Arabidopsis ETHE1 through loss of function mutant as well as in transgenic plants that over-express ETHE1 (Chapter 4-6). Inactivation of ETHE1 is lethal to the seed and results in an arrest in early endosperm and ultimately, early embryo development (Chapter 4). These results suggested that ETHE1 could play a critical role in either the transport and/or storage of an essential nutrient(s) which is consistent with a role in fatty acid β-oxidation. Short- chain fatty acid acyl-CoA oxidase activity is essential for early stages of embryo growth (11). However, the limitations imposed by this lethality made it difficult to speculate further into ETHE1’s role in plant development and potential substrate. To overcome this problem, we generated and analyzed inducible ETHE1-RNAi constructs, which allowed us to study the effects associated with the removal of ETHE1 at various time points throughout wild-type plant development (Chapter 5). These studies demonstrated that in addition to ETHE1 being critical to seed development, it is also essential for seed germination as well as necessary for the viability of suspension cell cultures (Chapter 5). These results are also consistent with those observed in mutants of β-oxidation; however, it was found that long-chain fatty acids, unlike short-chain fatty acids, are not required for embryo development, but are critical for seedling germination and establishment (12). This suggests that ETHE1 may not only be involved directly or indirectly in short-chain fatty acid β-oxidation, but also has a direct or indirect role in long-chain fatty acid β- oxidation. To further elucidate the role of ETHE1 in plant development, we over-expressed ETHE1 and monitored its phenotypic effects (Chapter 6). It had previously been predicted that ETHE1 was localized in mitochondria and that the mature protein

contained 256 amino acids (ETHE1256) (13). However more recently a full-length cDNA was published in Gene Bank that suggested the full length protein contained an additional N-terminal 38 amino acids resulting in a full length protein of 294 amino acids

(ETHE1294). We determined that ETHE1294 is localized in the mitochondrion and is

processed at amino acid 39 while transgenic cells expressing ETHE1256 accumulate the protein in the cytosol (Chapter 6). These results are consistent with the observation that

human ETHE1 is found in the mitochondria (1), and indicate that ETHE1294 is the correct

form of the full-length protein. Transgenic plants over-expressing ETHE1294 developed

192 identically to wild-type plants showing no obvious phenotypes under normal growth conditions. However the elevated levels of ETHE1 normally observed under environmental and abiotic stresses suggest that additional experiments are necessary under these conditions to further elucidate the functional role of ETHE1. Various metabolites of the branched-chain amino acid degradation pathways were

tested on plants over-expressing ETHE1294 in order to test one of the predicted roles of ETHE1 in isoleucine metabolism. However no differences in phenotype were apparent

(Chapter 6). Surprisingly, though, ETHE1254-OE plants showed a strong resistance to exogenous valine suggesting that the cytosolic version of ETHE1 performs a separate reaction chemistry than that of the native enzyme (Chapter 6). Further analysis of the

growth and development of plants over-expressing ETHE1256 revealed distinct growth enhancing properties; specifically plants over-expressing ETHE1256 bolted earlier,

flowered earlier and longer, and senesced later than both the wild-type and ETHE1294-OE plants expressing the native ETHE1 protein, thus increasing the plants’ overall seed yield. These plants also had thicker stems. While the growth enhancement properties displayed by ETHE1256-OE plants may not help in the determination of a role for ETHE1, plants with an ability to flower quicker and longer, produce more seed, and have thicker more upright inflorescence stems than wild-type plants could benefit the agricultural community and therefore may be useful in transgenic crop studies

7.4 Potential roles of ETHE1

ETHE1 is most closely related to the GLX2 family of enzymes which has been proposed, along with GLX1, to be responsible for the glutathione-dependent detoxification of 2-oxaldehydes, toxic byproducts of carbohydrate and lipid metabolism (14). Even though the primary sequences of these two families of proteins are quite diverse, the crystal structure obtained in our studies shows that ETHE1 is structurally very similar to GLX2 (5). ETHE1 does however lack several highly conserved glutathione binding residues found in GLX2 and does not utilize SLG, the GLX2 substrate, or other glutathione-based thioesters. Therefore, it is not involved in a glutathione based detoxification event. Recombinant ETHE1 does however exhibit weak

193 esterase activity. Based on these results and the nature of the ETHE1 active site, we predict that ETHE1 may hydrolyze a short chain ester. In humans, mutations in ETHE1 have been shown to cause EE which leads to the elevated levels of EMA excreted in the patient’s urine (1). Production of EMA can result from the carboxylation of butyryl-CoA found to occur in defects of the short-chain fatty acid β-oxidation pathway, as well as in the R-pathway of isoleucine catabolism (1, 9, 10). The results presented in this dissertation provide evidence that ETHE1 could be involved either directly or indirectly in the β-oxidation of fatty acids. Consistent with this possibility and our observations, it has been shown that β- oxidation plays an essential role in both seed development and seed germination (11, 15). There are six Acyl-CoA oxidase (ACX) genes in Arabidopsis, which catalyze the conversion of acyl-CoA‘s to 2-trans-enoyl-CoA’s in the first step of the β-oxidation pathway (12, 15-18). It was recently shown that, unlike medium and long-chain fatty acid acyl-CoA oxidation, short-chain fatty acid acyl-CoA oxidase activity is essential for early stages of embryo growth (11). However, long-chain fatty acid acyl-CoA oxidation, unlike medium and short-chain fatty acid acyl-CoA oxidation, was essential for seed germination (15). The possibility that ETHE1 plays a role in β-oxidation also is consistent with its increased expression during senescence as seen in the microarray data. However, the increased expression of ETHE1 during stress is not as clear in this regard. Studies have shown, however, that β-oxidation is necessary for the synthesis of the plant hormone Jasmonic acid (JA) which accumulates under periods of stress, namely wounding and insect resistance, to help in the regulation of growth inhibition, senescence, and leaf absicission (19).

7.5 Future aspects and directions of ETHE1 analysis

It has been shown in humans that mutations in ETHE1 are directly linked to the autosomal recessive disorder EE; however, ETHE1’s biochemical role and substrate are unknown (1). This thesis represents the first report of a full characterization of an ETHE1-like protein in any organism. These studies strongly suggest a role for ETHE1 in β-oxidation of fatty acids; however, more studies are necessary to confirm this hypothesis and to determine a substrate for ETHE1.

194 7.5.1 Microarray The use of microarrays allows for quantitative gene expression pattern analysis and has been widely used to monitor expression patterns of genes during studies of developmental processes as well as responses to environmental stresses in higher plants (20). In addition to monitoring these effects, microarrays can be used to analyze expression patterns between wild-type and transgenic plants. The limiting amount of ethe1 material available from our loss of function mutants limits its usefulness. However microarray analysis could be performed on the inducible pX7-ETHE1-RNAi lines at various stages of development and compared to wild-type. This would allow us to monitor and compare genes that are up regulated as well as down regulated in response to the removal of ETHE1. If ETHE1 plays a role in a metabolic pathway, monitoring changes in gene expression could potentially lead to a grouping of genes expressed within or connected to a particular pathway potentially providing evidence linking

ETHE1 to a biochemical role. Microarray studies on ETHE1256-OE plants should also provide insight intot he physiological changes that are connected with the growth enhanced traits. There are several limitations, however, to using this type of global approach (20). The amount of information obtained in a microarray study is extremely large and interpretation is time consuming and challenging. Additionally, because of the sensitivity of the array, experiments must be run a minimal of three times, therefore rendering cost as a factor in this technique (20). However, if the correct controls are taken and the analysis is consistent, a tremendous amount of information can be gained using microarray analysis.

7.5.2 NMR Metabolomics and NMR Metabolite Tracing Much like microarray analysis, NMR metabolomics uses a global approach to study gene function. However instead of measuring gene expression patterns, metabolomics utilizes NMR spectroscopy to measure the unique metabolites produced from cellular processes. Unlike the microarray analysis, this technique would allow us to monitor differences in metabolic profiles that result from alterations in metabolism. Given that EE patients show an increased excretion of EMA, it is likely that there are metabolite differences that can be used to assign a particular metabolic pathway of

195 ETHE1. Due to limitations of sample size in the Arabidopsis ETHE1 loss of function mutants, a metabolomic study on the inducible pX7-ETHE1-RNAi plants at various time points in plant development would prove useful for monitoring changes in metabolic profiling compared to wild-type. Similar to microarray analysis, a lot of data can be obtained from the metabolomic studies; therefore, samples need to be run a multitude of times, and data must be carefully interpretated. Data obtained from this study could be correlated with the results obtained from the microarray analysis at equivalent time points, providing a much fuller picture of the biochemical role of ETHE1. Once a predicted pathway for ETHE1 is determined, a NMR metabolic tracing study can be employed. In this technique, NMR spectroscopy can be used to study and compare the metabolic pathway of interest in both wild-type and the inducible pX7- ETHE1-RNAi plants by utilizing 13C-labeled metabolites of the predicted pathway. Through the use of the 13C-label, it is possible to trace the metabolic pathway and compare the two separate plant lines potentially locating a particular site and/or substrate of ETHE1 that accumulates when ETHE1 is removed.

7.5.3 Inducible pX7-ETHE1-OE Constructs Monitoring the gene expression patterns and metabolite profiles of plant lines over-expressing ETHE1 might also be beneficial in determining the biological role of ETHE1. However, it is predicted that the results from these studies on a constitutively over-expressing ETHE1 line might yield confusing results given the possibility that the plant will adapt to the over-expressed gene. Therefore plants and cell cultures expressing the inducible pX7-ETHE1-OE construct have been generated. Preliminary results obtained from these studies showed no difference in growth patterns and development from wild-type plants under normal conditions (data not shown); however, differences may appear in gene expression as well as metabolic profiles of the inducible pX7- ETHE1-OE plant versus wild-type. Likewise, in the presence of various abiotic and environmental stresses, microarray and metabolomic studies of the inducible pX7- ETHE1-OE plants in comparison to wild-type might provide insight into the biological role of ETHE1.

196 7.5.4 Characterization of Human ETHE1 Even though ETHE1 is critical for development in both plants and humans, it is possible that ETHE1 serves different biochemical roles in humans than that which we have predicted in plants. To address this possibility, a full characterization of the human ETHE1 needs to be conducted. Through a collaboration with Valeria Tiranti’s lab we have obtained various N-terminal cDNA clones and have over-expressed and begun preliminary metal and substrate analyses on the proteins. We have also begun assessing conditions for crystal formation to obtain the crystal structure. Based on the conserved sequence, it is likely that the plant and human enzymes will share many of the same biochemical and structural properties. However, it is possible that the two enzymes evolved different specific roles based on the metabolism of the different organisms. Even though the exact substrate for ETHE1 is still unknown, the results presented in this dissertation have demonstrated that ETHE1 plays a critical role in endosperm development, seed germination, and the viability of suspension cell cultures, which suggests that ETHE1 may function either directly or indirectly in the β-oxidation of fatty acids in plants. Ultimately, information obtained from these studies may be able to guide future studies in humans and provide insight as to the reason such large metabolic changes take place in patients with EE.

197 7.6 References

1. Tiranti, V., D'Adamo, P., Briem, E. et al. (2004). Ethylmalonic Encephalopathy is Caused by Mutations in ETHE1, a Gene Encoding a Mitochondrial Matrix Protein. Am. J. Hum. Genet. 74, 239-252

2. Tiranti, V., Briem, E., Lamantea, E. et al. (2006). ETHE1 Mutations are Specific to Ethylmalonic Encephalopathy. J. Med. Genet. 43, 340-346

3. Burlina, A., Zacchello, F., Dionisi-Vici, C., Bertini, E., Sabetta, G., Bennet, M. J., Hale, D. E., Schmidt-Sommerfeld, E., and Rinaldo, P. (1991). New Clinical Phenotype of Branched-chain Acyl-CoA Oxidation Defect. Lancet 338, 1522-1523

4. Burlina, A. B., Dionisi-Vici, C., Bennett, M. J., Gibson, K. M., Servidei, S., Bertini, E., Hale, D. E., Schmidt-Sommerfeld, E., Sabetta, G., and Zacchello, F. (1994). A New Syndrome with Ethylmalonic Aciduria and Normal Fatty Acid Oxidation in Fibroblasts. J. Pediatr. 124, 79-86

5. McCoy, J. G., Bingman, C. A., Bitto, E., Holdorf, M. M., Makaroff, C. A., and Phillips, G. N.,Jr. (2006). Structure of an ETHE1-like Protein from Arabidopsis thaliana. Acta Crystallogr. D Biol. Crystallogr. 62, 964-970

6. Cameron, A. D., Ridderstrom, M., Olin, B., and Mannervik, B. (1999). Crystal Structure of Human Glyoxalse II and its Complex with a Glutathione Thiolester Substrate Analogue. Structure 7, 1067-1078

7. Marasinghe, G. P., Sander, I. M., Bennett, B., Periyannan, G., Yang, K. W., Makaroff, C. A., and Crowder, M. W. (2005). Structural Studies on a Mitochondrial Glyoxalase II. J. Biol. Chem. 280, 40668-40675

8. Schilling, O., Wenzel, N., Naylor, M., Vogel, A., Crowder, M., Makaroff, C., and Meyer-Klaucke, W. (2003). Flexible Metal Binding of the Metallo-beta-lactamase

198 Domain: Glyoxalse II Incorporates Iron, Manganese, and Zince in Vivo. Biochemistry 42, 11777-11786

9. Mamer, O. A., Tjoa, S. S., Scriver, C. R., and Klassen, G. A. (1976). Demonstration of a New Mammalian Isoleucine Catabolic Pathway Yielding an Rseries of Metabolites. Biochem. J. 160, 417-426

10. Nowaczyk, M. J., Lehotay, D. C., Platt, B. A., Fisher, L., Tan, R., Phillips, H., and Clarke, J. T. (1998). Ethylmalonic and Methylsuccinic Aciduria in Ethylmalonic Encephalopathy Arise from Abnormal Isoleucine Metabolism. Metabolism 47, 836- 839

11. Penfield, S., Helen, M., Pinfield-Wells et al. (2006) Storage Reserve Mobilisation and Seedling Establishment in Arabidopsis. In The Arabidopsis Book Rockville, MD, American Society of Plant Biologists.

12. Adham, A. R., Zolman, B. K., Millius, A., and Bartel, B. (2005). Mutations in Arabidopsis Acyl-CoA Oxidase Genes Reeal Distinct and Overlapping Roles in Beta-Oxidation. Plant J. 41, 859-874

13. Maiti, M. K., Krishnasamy, S., Owen, H. A., and Makaroff, C. A. (1997). Molecular Characterization of Glyoxalase II from Arabidopsis thaliana. Plant Mol. Biol. 35, 471-481

14. Phillips, S. A., and Thornalley, P. J. (1993). The Formation of Methylglyoxal from Triose Phosphates. Investigation Using a Specific Assay for Methylglyoxal. Eur. J. Biochem. 212, 101-105

15. Pinfield-Wells, H., Rylott, E. L., Gilday, A. D., Graham, S., Job, K., Larson, T. R., and Graham, I. A. (2005). Sucrose Rescues Seedling Establishment but not Germination of Arabidopsis Mutants Disrupted in Peroxisomal Fatty Acid Catabolism. Plant J. 43, 861-872

16. Eastmond, P. J., Hooks, M. A., Williams, D., Lange, P., Bechtold, N., Sarrobert, C., Nussaume, L., and Graham, I. A. (2000). Promoter Trapping of a Novel Medium-

199 chain Acyl-CoA Oxidase, Which is Induced Transcriptionally During Arabidopsis Seed Germination. J. Biol. Chem. 275, 34375-34381

17. Penfield, S., Rylott, E. L., Gilday, A. D., Graham, S., Larson, T. R., and Graham, I. A. (2004). Reserve Mobilization in the Arabidopsis Endosperm Fuels Hypocotyl Elongation in the Dark, is Independent of Abscisic Acis, and Requires Phosphoenolpyruvate Carboxykinase 1. Plant Cell 16, 2705-2718

18. Rylott, E. L., Rogers, C. A., Gilday, A. D., Edgell, T., Larson, T. R., and Graham, I. A. (2003). Arabidopsis Mutants in Short- and Medium-chain Acyl-CoA Oxidase Activities Accumulate Acyl-CoAs and Reveal that Fatty Acid Beta-Oxidation is Essential for Embryo Development. J. Biol. Chem. 278, 21370-21377

19. Hedden, P., and Thomas, S. (2006) Plant Hormone Signaling. A.P.R., Oxford, UK, Blackwell Publishing.

20. Kehoe, D. M., Villand, P., and Somerville, S. (1999). DNA Microarrays for Studies of Higher Plants and Other Photosynthetic Organisms. Trends Plant Sci. 4, 38-41

200