|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 FUNCTIONAL CHARACTERIZATION OF 3 ISOPROPYLMALATE DEHYDROGENASES AND ISOPROPYLMALATE ISOMERASES IN ARABIDOPSIS By YAN HE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT O F THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010
2 2010 Yan He
3 To those who have love and faith in me
4 ACKNOWLEDGMENTS I would like to begin by giving my exceptional thanks to Dr. Sixue Chen, my superv isory committee chair. His trust, endless support and guidance throughout my graduate study have been invaluable and greatly appreciated. I would also like to express my appreciation to the other members of the supervisory committee, Dr. Harry Klee, Dr. Ba la Rathinasabapathi and Dr. Nancy Denslow for their willingness to serve on the committee, their input and guidance during my research, and for reviewing the papers and dissertations. Special thanks go to my cooperators, Dr. Joseph Jez at Washington Univ ersity Dr. Thomas P. Mawhinney at University of Missouri, Dr. ByungHo Kang and Dr. Bernard Hauser at University of Florida, for their assistance regarding the experiments. I thank Dr. Liqun Chen, Qiuying Pang, Mengmeng Zhu, Ning Zhu, Johanna Strul and other member s in my lab for their assistance. My special thanks go to my lovely wife and adorable baby, my father, brother and my parent s in law and my family for their constant faith and love in me.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 8 LIST OF FIGURES .......................................................................................................... 9 ABST RACT ................................................................................................................... 11 CHAPTER 1 LITERATURE REVIEW .......................................................................................... 14 Glucosinolate Metabolism ....................................................................................... 14 Chemical Structure and Systemic Distribution .................................................. 14 Biological Function: GlucosinolateMyrosinase System ................................... 16 Aliphatic Glucosinola te Biosynthesis ................................................................ 17 The evolutionary link between methionine chain elongation and leucine biosynthesis ............................................................................................ 17 Aliphatic glucosinolate biosynthesis: core structure ................................... 19 Aliphatic glucosinolate biosynthesis: secondary transformations ............... 24 Regulation of Glucosinolate Biosynthesis: Signaling Networks ........................ 26 Transcriptional factors controlling aliphatic glucosinolate biosynthesis ...... 27 Transcriptional factors controlling indole glucosinolate biosynthesis ......... 28 Other known regulators of glucosinolate biosynthesis ............................... 2 9 Signaling networks of glucosinolate biosynthesis ...................................... 30 Conclusions and Future Prospects ................................................................... 31 2 A REDOX ACTIVE ISOPROPYLMALATE DEHYDROGENASE FUNCTIONS IN THE BIOSYNTHESIS OF GLUCOSINOLATE AND LEUCINE IN ARABIDO PSIS ....................................................................................................... 36 Introduction ............................................................................................................. 36 Materials and Methods ............................................................................................ 38 Plant Materials and Chemicals ......................................................................... 38 DNA Extraction and Genotyping ....................................................................... 39 RNA Extraction, RT PCR and Arabidopsis Transformation .............................. 39 Promoter GUS Fusion and GUS Assay ............................................................ 40 Immunocytochemistry and Chloroplast Fractionation ....................................... 40 Glucosinolate and Amino Acid Analysis ........................................................... 41 Protein Expression, Western Blotting and Visualization of Protein Sulfhydryl Groups .......................................................................................................... 41 Redox and Enzymatic Activity Assays .............................................................. 42 Results .................................................................................................................... 43 IPMDH1 Is I nvolved in Glucosinolate and Leucine Biosynthesis ...................... 43
6 IPMDH1 Exhibits 3 Isopropylmalate Dehydrogenase Activity .......................... 44 Glucosinolate P rofiles and Leucine Biosynthesis Are Substantially Affected by Mutation of IPMDH1 ................................................................................. 45 IPMDH1 Expression Displays Temporal and Spatial Regulation and Is Triggered by Wounding ................................................................................. 47 IPMDH1 Protein and Activity Are Localized to Chloroplast Stroma .................. 48 IPMDH1 Is Regulated by ThioredoxinMediated Redox Regulation ................. 49 Thiol Based Switch Is Essential for Regulating IPMDH1 Activity ..................... 50 Discussion .............................................................................................................. 52 The Expression and Subcellular Localization of IPMDH1 Highlight Its Potential Functional Significance .................................................................. 52 The Dual Functions of IPMDH1 in Leucine Biosynthesis and the Methionine ChainElonga tion Cycle of Aliphatic Glucosinolate Biosynthesis ................... 53 Thiol Based Redox Regulation of Glucosinolate Biosynthesis and Leucine Biosynthesis .................................................................................................. 55 3 FUNCTIONAL CHARACT ERIZATION OF ISOPROPYLMALATE DEHYDROGENASES REVEALS THEIR ROLES IN GAMETOPHYTE DEVELOPMENT ..................................................................................................... 73 Introduction ............................................................................................................. 73 Materials and Methods ............................................................................................ 75 Plant Materials and Growth Conditions ............................................................ 75 DNA Extraction and Genotyping ....................................................................... 75 Genetic Analysis ............................................................................................... 76 Complementation of IPMDH mutants ............................................................... 77 Recombinant Protein Ex pression and Purification ............................................ 77 Light Microscopy and Phenotype Analysis ....................................................... 78 Transmission and Scanning Electron Microscopy ............................................ 79 Gene Expression Analysis ................................................................................ 80 Chlorophyll, Glucosinolate and Amino Acid Analysis ....................................... 81 Chloroplast Fractionation and In Vivo IPMDH Activity Assay ........................... 81 Results .................................................................................................................... 82 IPMDH2 and IPMDH3 Are Key Enzymes in Leucine Biosynthesis ................... 82 Double Mutant ipmdh2 ipmdh3 Is Lethal .......................................................... 82 The ipmdh2 ipmdh3 Is Not Transmitted through the Male ................................ 83 Wild Type IPMDH2 and IPMDH3 Can Rescue the Distorted Transmission ..... 84 ipmdh2 ipmdh3 Pollen Grains Show Abnormal Morphology ............................. 85 ipmdh2 ipmdh3 Pollen Grains Are Defective in Pollen Germination ................. 86 Ultrastructure of ipmdh2 ipmdh3 pollen ............................................................ 87 Genetic Confirmation of ipmdh2 ipmdh3 Male Gametophytic Lethality ............ 88 Female Gametophyte Development Slowed in ipmdh2 ipmdh3 Mutants ......... 89 HaploInsufficiency of IPMDH2 Results in Sporophytic Phenotype .................. 90 Enzyme Activities and Leucine Levels Were Affected in the Mutants .............. 91 IPMDH2 and IPMDH3 Function in Glucosinolate Biosynthesis ........................ 93 TissueSpecific Expression of the IPMDH2 and IPMDH3 ................................. 94 Discussion .............................................................................................................. 95
7 IPMDH2 and IPMDH3 Are Functionally Redundant in Leucine Metabolism ..... 95 IPMDH2 and IPM DH3 Are Essential for Male Gametophyte Development ...... 96 Involvement of IPMDH2 and IPMDH3 in Embryo Sac Development ................ 98 IPMDH2 and I PMDH3 Affect Sporophytic Development ................................ 100 Involvement of IPMDH2 and IPMDH3 in Glucosinolate Biosynthesis ............. 101 4 FUNCTIONAL SPECIFI CATION OF THE SMALL SUBUNITS OF ISOPROPYLMALATE ISOMERASE IN GLUCOSINOLATE AND LEUCINE BIOSYHTHESIS ................................................................................................... 125 Introduction ........................................................................................................... 125 Materia ls and Methods .......................................................................................... 126 Plants and Chemicals ..................................................................................... 126 DNA Extraction and Genotyping ..................................................................... 126 RNA Interference of AtLeuD2 and Plant Transformation ................................ 127 RNA Extraction and RTPCR ......................................................................... 127 Promoter GUS Fusion and GUS Assay.......................................................... 127 Complementation of Arabidopsis atleuc Mutant ............................................. 128 Protein Expression and Antibody Production ................................................. 128 Chloroplast Fractionation ................................................................................ 128 Coimmunoprecipitation and Western Blotting ................................................. 128 Glucosin olate and Amino Acid Analysis ......................................................... 129 Complementation of E. coli leuC and leuD Mutants ....................................... 129 Results .................................................................................................................. 131 AtLeuD1 and AtLeuD2 Are Redundantly Functional in Glucosinolate Biosynthesis ................................................................................................ 131 Temporal and Spatial Regulation of AtLeuC and AtLeuDs ............................. 132 AtLeuDs Associate Physically with AtLeuC in vivo ......................................... 133 IPMI Complex Is Localized to Chloroplast Stroma ......................................... 134 Complementation of E. coli Leucine Auxotrophs by Coexpression of AtLeuC and AtLeuDs ............................................................................................... 134 atleud3 Mutants Exhibit Lethal Phenotype ..................................................... 135 Discussion ............................................................................................................ 136 AtLeuC Is Functional in Both Leucine and Glucosinolate Biosynthesis .......... 136 AtLeuD1 and A tLeuD2 Are Involved in Glucosinolate Biosynthesis ............... 137 AtLeuD3 Is an Essential Gene ....................................................................... 138 LIST OF REFERENCES ............................................................................................. 149 BIOGRAPHICAL SKETCH .......................................................................................... 164
8 LIST OF TABLES Table page 2 1 Primers used in this study ................................................................................. 58 2 2 Substrate and kinetics of recombinant IPMDH1 (red, reduced form; ox, oxidized form) ..................................................................................................... 60 3 1 Primers used in this study. ............................................................................... 103 3 2 Enzyme affinity and activity toward 3isopropylmalate ..................................... 104 3 3 Transmission efficiency of ipmdh2 ipmdh3 allele in reciprocal crosses. ........... 105 3 4 Genetic analysis of the ipmdh2 ipmdh3 allele in transgenic lines harboring the complementation construct ......................................................................... 106 3 5 The transmission efficiency of ipmdh2 ipmdh3 gamete using the transgenic plants as pollen donor ...................................................................................... 107 3 6 Transmission efficiency of ipmdh2 ipmdh3 gamete through female from delayed pollination test. .................................................................................... 108 3 7 Amino acid profiles of wild type and different mutants ...................................... 109 4 1 P rimers used in this study. ............................................................................... 139 4 2 Amino acid profiles of wild type and different mutants ...................................... 141
9 LIST OF FIGURES Figure page 1 1 Chainelongation process of aliphatic glucosinolate biosynthesis in Arabidopsis and leucine biosynthes is in microbes .............................................. 33 1 2 Biosynthesis of t he core glucosinolate structure ................................................. 34 1 3 Side chain modification of methioninederive d glucosinolates in Arabidopsis .... 35 2 1 Multiple sequence alignment of 3isopropylmalate dehydrogenases from different organisms ............................................................................................. 61 2 2 Purification and identification of recombinant IPMDH1 ....................................... 62 2 3 Temperature and pH dependence of IPMDH1 activity ....................................... 63 2 4 Characterization of TDNA knockout mutant of IPMDH1 .................................... 64 2 5 Glucosinolate profiles of wild type, ipmdh1 mutant and ipmdh1 complementation plants ...................................................................................... 65 2 6 Free amino acid profiles of wild type and ipmdh1 mutant ................................... 66 2 7 Spatial and temporal expression patterns of IPMDH1 analyzed using IPMDH1 promoter GUS plants ........................................................................... 67 2 8 Subcellular localization of IPMDH1 and in vivo activity analysis ......................... 68 2 9 Redox regulation of recombi nant IPMDH1 ......................................................... 69 2 1 0 Redox sensitivity of IPMDH1 cysteine mutants .................................................. 7 0 2 1 1 Accurate molecular weight analysis of reduced and oxidi zed IPMDH1 .............. 71 2 1 2 In vivo redox sensitivity of IPMDH ...................................................................... 72 3 1 Heterologous expression of three IPMDHs in E. coli and purification using Nikel affinity chromatography e ......................................................................... 110 3 2 Schematic diagram of the genomic structure of IPMDHs with T DNA insertion sites (triangles) ................................................................................................. 111 3 3 Generation and confirmation of single, double and triple mutants of IPMDH genes ................................................................................................................ 112 3 4 Microscopic phenotypes of pollens from wildtype (left), IPMDH2/ipmdh2 ipmdh 3/ipmdh3 ( middle) and ipmdh2/ipmdh2 IPMDH 3/ipmdh3 plants (right). .. 113
10 3 5 Germination of pollens from wildtype, IPMDH2/ipmdh2 ipmdh 3/ipmdh3 and ipmdh2/ipmdh2 IPMDH 3/ipmdh3 mutant plants ............................................... 114 3 6 Transmission Electron Microscopy (TEM) of IPMDH2/ipmdh2 ipmdh 3/ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipmdh3 mutant pollen grains .............................. 115 3 7 Poll en development and germination of ipmdh2 ipmdh3 in qrt background. .... 116 3 8 Seeds are developed normally in wildtype siliques ......................................... 117 3 9 Loss of function of IPMDH2 and IPMDH3 causes delayed fusion of two polar nuclei in embryo sacs ....................................................................................... 118 3 1 0 Sporophytic phenotypes of IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh1/ipmdh1 IPMDH2/ipm dh2 ipmdh3/ipmdh3 plants .................................. 119 3 1 1 In vivo IPMDH activity in wild type, the single, double and triple IPMDH mutants ............................................................................................................. 120 3 1 2 Gluc osinolate profiles in wildtype and different IPMDH mutants ..................... 121 3 1 3 E xpression pattern of IPMDH2 and IPMDH3 analyzed using promoter GUS plants ................................................................................................................ 122 3 1 4 Quantitative real time RT PCR analysis of IPMDH transcript levels at various tissues .............................................................................................................. 123 3 1 5 Ultra structure of chloroplasts in first true leaves .............................................. 124 4 1 Characterization of TDNA knockout mutant of AtLeuD1 and RNA interference mutants of AtLeuD2 ...................................................................... 142 4 2 Glucosinolate profiles of wild type, at leud1 mutant and LeuD2RNAi plants .... 143 4 3 Spatial and temporal expression patterns of AtLeuC analyzed using promoter GUS plants ....................................................................................................... 144 4 4 Glucosinolate profiles of wild type, atleuc mutant and the complemented plants ................................................................................................................ 145 4 5 Direct Interaction between AtLeuC and AtLeuDs in vivo .................................. 146 4 6 Subcellular localization of AtLeuC and AtLeuDs within Arabidopsis ................. 147 4 7 Complementation of E.coli leuC and leuD auxotrophic mutants. ...................... 148
11 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy F UNCTIONAL CHARACTERIZATION OF 3 ISOPROPYLMALATE DEHYDROGENASES AND ISOPRO PYLMALATE ISOMERASES IN ARABIDO P S IS By Yan He May 2010 Chair: Sixue Chen Major: Plant Molecular and Cellular Biology Isopropylmalate dehydrogenases ( IPMDHs ) and Isopropylmalate isomerase s (IPMIs ) catalyze the oxidative decarboxylation of 3 isopropylmalate (3IPM) and the isomerization of 2isopropylmalate (2IPM) in leucine biosynthesis respectively Here the functions of the genes encoding IPMDHs and IPMIs in Arabidospis were characterized using multidisciplinary approaches Arabidops is IPMDH1 shares high homology with enzymes from bacteria and yeast known to f unction in leucine biosynthesis In plants, IPMDH1 co expressed with nearly all known genes in aliphatic glucosinolate biosynthesis. Mutation of IPMDH1 led to a significant reduc tion in the levels of leucine and glucosinolates with sidechains of four carbons or longer. Complementation of the mutant phenotype by ectopic expression of IPMDH1 together with the enzymes s ubstrate specificity, implicates IPMDH1 in both glucosinolate and leucine biosynthesis. This functional assignment is substantiated by the subcellular localization of the protein in the chloroplast stroma and the gene expression patterns in different tissues. Interestingly, IPMDH1 activity was regulated by a thiol based redox modification. Collectively, this study has characterized the first enzyme in plants that catalyzes the oxidative decarboxylation step in both leucine
12 biosynthesis (primary metabolism) and methionine chain elongation of glucosinolates (specialized metabolism). It provides evidence for the hypothesis that the two pathways share a common origin and suggests a role for redox regulation of glucosinolate and leucine synthesis in plants. Compared to IPMDH1, Arabidopsis IPMDH2 and IPMDH3 exhibit significa ntly higher affinity toward 3 IPM which is indicative of pivotal role in leucine biosynthesis. Single mutants of IPMDH2 or IPMDH3 lacked a discernible phenotype. Genetic analysis showed that the ipmdh2 ipmdh3 double mutant was lethal in male gametophytes and had reduced transmission through the female gametophytes. Microscopic analysis indicated that the aborted pollen grains were small, abnormal in cellular structure, and arrested in germination. In addition, half of the double mutant embryo sacs exhibited slowed development. The IPMDH2/ipmdh2 ipmdh3/ipmdh3 genotype exhibited abnormal vegetative phenotypes, suggesting haploinsufficiency of IPMDH2 in the ipmdh3 background. The mutant and a triple mutant containing one allele of IPMDH2 or IPMDH3 had decreas ed leucine biosynthetic enzyme activities and lower free leucine levels. The latter showed changes in glucosinolate profiles that were significantly different from those in the ipmdh1 mutant. These results demonstrate that IPMDH2 and IPMDH3 primarily funct ion in leucine biosynthesis, are essential for pollen development and needed for embryo sac development. Arabidopsis encodes heterodimeric IPMIs consisting of a large subunit encoded by one gene ( AtLeuC ), and a small subunit encoded by one of three homologous genes ( AtLeuD1, AtLeuD2 and AtLeuD 3 ). The involvement of AtLeuC in both glucosinolate biosynthesis and leucine metabolism has been reported. H owever, the specific function
13 of each At LeuD is not known. Compared to wildtype, there was no glucosinolate change in atleud1 single mutant. When expression of AtLeuD2 was reduced in atleud1 mutant background, substantial changes in glucosinolate profiles were observed, but the levels of leucine and other amino acid levels were not affected, indicating that AtLe uD1 and AtLeuD2 are functionally redundant in glucosinolate biosynthesis. The redundancy between AtLeuD1 and AtLeuD2 was also reflected by their nearly identical gene expression patterns. In addition, biochemical analysis revealed in vivo interaction between AtLeuC and AtLeuD proteins. Furthermore, coexpression of AtLeuC with one of the AtLeuDs in leu C and leuD bacterial mutants restored the autotrophic growth of the bacteria. This result suggests that all three AtLeuDs have the ability to form a functional IPMI with AtLeuC. Inte res tingly, loss of function of AtLeuD3 result ed in lethality and the underlying mechanism remains to be elucidated.
14 CHAPTER 1 LITERATURE REVIEW Glucosinolate Metabolism Glucosin olates, once collectively referred to as mustard oil glucosides, have long been part of human life for their distinctive flavor and aroma in B rassica vegetables, the unique benefit to human nutrition and characteristic involvement in plant defense and growth (Kliebenstein et al., 2005; Grubb and Abel 2006). In the past decades, the importance of these sulfur containing secondary metabolites has drawn great attention because of their roles in cancer prevention (Hecht, 2000) and nonhost r esistance to bacteria and fungi (Bednarek et al., 2009; Clay et al., 2009). Glucosinolates occur almost exclusively in members of the Brassicales an order that includes cabbage, broccoli, mustard and rape (Bak et al, 1999). The presence of glucosinolates in the model plant, Arabidopsis thaliana has inspired intensive researc h into these unusual amino acid derived compounds, leading to the complete elucidation of the biosynthetic pathway, identification of many transcriptional regulators of the pathway and the evolutionary links to other pathways (Halkier and Gershenzon, 2006; Yan and Chen, 2007) Chemical Structure and Systemic Distribution The core chemical structure of all glucosinola thioglucose linked via a sulfur atom to a (Z) N hydroximinosulfate es ter, plus a variable side chain (R group) derived from one of eight amino acids (Fahey et al., 2001). Different precursor amino acids and R group variations caused by sidechain elongation and modification account for the chemical diversity of the over 120 glucosinolate structures (Fahey et al., 2001). Glucos inolate s can be classified according to their precursor amino ac ids, i.e.
15 those derived from methionine, ala nine, l eucine, iso l eucine, or v al ine are named aliphatic glucos inolates, those derived from tr yptophan are named indole glucosinolates and those syn thesized from phenylalanine or t yr osine are called aromatic glucosino late s. Aliphatic glucosinolates represent the most abundant group in Arabidopsis and many other species of the Brassicaceae. Plant g lucosin olate concentration is highly variable and is normally about 1% of dry weight in some tissues of the Brassica vegetables (Kushad et al., 1999). Most species contain a limite d number of glucosinolates (usually less than one dozen), but in Arabidopsis over 30 different glucosinolates have been identified (Haughn et al., 1991; Reichelt et al., 2002). Glucosinolate composition and concentration vary significantly in different tissues at different developmental stages (Peterson et al., 2002). In Arabidopsis, high glucosinolate concentrations are found in reproductive organs such as seeds, siliques, flowers and developing inflorescences, followed by roots, young leaves, stem, and fully expanded leaves, whereas senescing rosette leaves have the lowest amounts (Brown et al., 2003). The sulfur rich cells (Scell) situated in the floral stalk between the phloem of the vas cular bundle and endodermis, have extraordinarily high concentrations of glucosinolates based on a sulfur assay (Koroleva et al., 2000). The extremely low activity of glucosinolate biosynthesis in seeds suggests t he presence of an import system The phloem transport of glucosinolate s from leaves to inflorescence and developing seeds has been characterized (Chen et al., 2001), and is further supported by the evidence that many glucosinolate biosynthetic genes showed strong expression in vascular tissues (Mikkelsen et al., 2000; Reintanz et al., 2001; Chen et al., 2003; Grubb et al., 2004; Schuster et al., 2006; He et al., 2009).
16 Biological Function: GlucosinolateMyrosinase System Even though intact glucosinolates may confer resistance to certain insects (Kim and Jander, 2007), the concept that glucosinolates are hydrolyzed by myrosinases into defense compounds has long been recognized (Halkier and Gershenzon, 2006, Hopkins et al., 2009). The glucosinolatemyrosinase system represents a twocomponent defense system, often referred to as the mustard oil bomb (Ratzka et al., 2002). In this system, intact gluco si nolates are stored separately from their degradation thioglucosidases) Upon tissue damage (e.g. by cutting or chewing), the contact of glucosinolates with myrosinases leads to the rapid generation of unstable thiohydroximateO sulfate intermediates, which undergo elimination of the sulfate group and spontaneous degradation, resulting in the production of a suite of biologically active compounds, including nitriles, epithionitriles, thiocyanates, oxazolidine 2 thiones and isothiocyanates (Wittstock and Halkier, 2002) The final chemical products produced by the reaction depend mainly on the structure of the side chains and reaction conditions such as pH, the availability of ferrous ions and the presence of epithiospecifier proteins (ESP) or ESP modifier protein (Lambrix et al., 2001; Burow et al., 2006) Until 2009, the mechani sms of glucosinolates in plant defense against bacteria and fungi were not as well defined as compared to their widely demonstrated anti insect function (Bednarek et al., 2009; Clay et al., 2009). Unlike the passive property of damagetriggered activation by myrosinases, glucosinolates are now identified to be active in living plant cells and recruited for broadspectrum antifungal defense responses (Bednarek et al., 2009). Once plants are attack ed by a fungus, a P450 monooxygenase, encoded by Arabidopsis C YP81F2 catalyzes the accumulation of 4-
17 methoxy indole 3 ylmethylglucosinolate (4MI3G), which is hydrolyzed by the atypical PEN2 thioglucoside glucohydrolase). The hydrolyzed products are then actively transported to the cell periphery at the fungal entry sites by a PEN3 encoded plasma membraneresident ABC transporter for antifungal defense (Bednarek et al. 2009). In addition, the study by Clay et al., (2009) showed that PEN2 and PEN3 dependent 4MI3G metabolism was required for the microbeassociated molecule pattern (MAMP) triggered callose defense response, highlighting the link of the glucosinolate hydr olytic products as potential signaling molecules to a classical innate immune response to both adapted and nonadapted microbial pathogens. Aliphatic Glucosinolate Biosynthesis Biosynthesis of methioninederived glucosinolates proceeds in three separate phases. First, methionine undergoes a sequential addition of one to six additional methylene groups to produce chainelongated methionine derivatives (Figure 1 1, He et al., 2009). Second, the derived chainelongated methionines enter the pathway for synthes is of the core glucosinolate structure. (Figure 1 2, Halkier and Gershenzon, 2006). Last, the initially formed glucosinolates are modified by further secondary sidechain transformations (Figure 1 3, Grubb and Abel, 2006). T he evolutionary link between met hionine chain elongation and leucine b iosynthesis Based largely on in vivo studies with isotopically labeled trace r s, the initial methionine is deaminated to form the corresponding 2o xo acid, i.e. 4methythio2 oxobutanoic acid (Schuster et al., 2006; Kni ll et al., 2008). Next is a threestep cycle in which the 2 oxo acid condense s with acetylCoA to produce a substituted 2malate derivative (2(2 methylthioethyl) malate) (Textor et al., 2004, 2007). The 2malate
18 derivative is then isomerized to a 3malat e derivative via a 1,2hydroxyl shift (Knill et al., 2009). The 3malate derivative undergoes oxidative decarboxylation to yield a 2oxo acid extended by one m ethylene group (He et al., 2009). After each turn of the cycle, the extended 2oxo acid can be tr ans a minated to produce the corresponding elongated methionine (e.g. homomethionine), and enters into the glucosinolate core structure pathway to yield C3 glucosinolate (Knill et al., 2008). Alternatively, the extended 2oxo acid can undergo iterative acety l CoA condensation, isomerization, and oxidationcarboxylation, leading to further chainelongation. Consequently, the newly extended 2oxo acid can enter the core glucosinolate synthesis to yield C4 C8 glucosinolate (He et al., 2009). Similar 2 oxo acid based chainelongation sequences involving condensation, isomerization and oxidative decarboxylation also occur in leucine biosynthesis (Figure 2 1, 2oxoisovalerate, 2isopropylmalate, 3isoporpylmalate, 2oxoisocaproate) and in the TCA cycle (oxaloacetat e, citrate i socitrate, 2 oxoglutarate) (Textor et al., 2004). R ecent literature indicated in the order of Capparales at least in Arabidopsis, the enzymes involved in leucine biosynthesis are related to or participate in the methionine chainelongation process of aliphatic glucosinolate biosynthesis (Field et al., 2004; Schuster et al., 2006; Binder et al. 2007; Textor et al., 2007; He et al., 2009). For example, isopropylmalate synthase is encoded by four genes in Arabidopsis (Col 0), two of which ( IPMS1 and IPMS2 ) are functional in leucine biosynthesis and the other two genes encode methylthioalkylmalate ( MAM ) synthases ( MAM1 and MAM3 ) involved in methionine chainelongation pathway of aliphatic glucosinolate biosynthesis (Field et al., 2004; Textor et al ., 2004 and 2007). In addition, as in bacteria but not in yeast,
19 Arabidopsis isopropylmalate isomerase ( IPMI) appears to be a heterodimeric enzyme, consisting of a large subunit encoded by a single gene and a smal l subunit encoded by one of three genes (Bi nder et al., 2007; Knill et al., 2009). Metabolic profiling of the large subunit mutant revealed significant accumulation of intermediates in both the leucine pathway and the methionine chain elongation pathway, demonstrating the function of the IPMI large subunit in both leucine and glucosinolate biosynthesis (Knill et al., 2009). In contrast, the small subunits seem to be specialized to either leucine biosynthesis or methionine chain elongation, although the precise function of each gene is still elusive (Knill et al., 2009; Sawada et al., 2009). Furthermore, among the six branch chain aminotransferases (BCATs) in Arabidopsis, BCAT4 was demonstrated to be specifically involved in glucosinolate biosynthesis whereas BCAT3 functions in both amino acid and glu cosinolate biosynthesis (Schuster et al., 2006; Knill et al., 2008) Taken together, the close relationsh ip between methionine chain elongation and leucine biosynthesis suggests that both pathways are derived from a common ancestral reaction or the methionine chainelongation reaction is evolutionally recruited from primary metabolism, i.e., leucine biosynthesis. Aliphatic glucosinolate biosynthesis: core structure Synthesis of the core glucosinolat es is composed of six steps involving intermediates common to all glucosinolates including N hyd r oxy l amino acids, aldoximes, aciNitro or nitrile oxides compound, GSH conjugated S alkyl thiohydroximate, Cys Gly conjugated S alkyl thiohydroximate, S alkyl thiohydroximate and desulfoglucosinolate (Figure 1 2, Gr ubb and Abel, 2006; Halkier and Gershenzon, 2006; Geu Flores et al., 2009). The genes responsible for all steps have been characterized in the model plant Arabidopsis since 2000 (Figure 1 2).
20 Amino acids and chain elongated amino acids enter the core pathway beginning with oxidation to aldoximes catalyzed by cytochrome P450 monooxygenases in the CYP79 family ( CYP79s Hansen et al., 2001; Chen et al., 2003; Tantikanjana et al., 2004). The Arabidopsis genome encodes seven CYP79s and five of them are characterized with respect to substrate specificity (Grubb and Abel, 2006). CYP79F1 and CYP79F2 are responsive for aldoxime production in the biosynthesis of the aliphatic glucosinolates derived from chainelongated methionine derivatives (Hansen et al., 2001; Chen et al., 2003). CYP79B2/CYP79B3 and CYP79A2 catalyze the production of aldoximes from tryptophan and phenylalanine, respectively (Hull and Celenza, 2000; Wittstock and Halkier, 2002; Zhao et al., 2002). However, the role of CYP79C1 and CYP79C2 remains to be identified (Halkier and Gershenzon, 2006). In t wo independent genetic screens, two Arabidopsis mutants called supershoot ( sps ) or bushy ( bus ), with severe morphological alteration including massive proliferation of shoots and increased number of meristems in leaf axils, were recognized to be caused by the disruption in the gene encoding CYP79F1 (Reintanz et al., 2001; Tantikanjana et al., 2001). The sps/bus/cyp79F1 plants completely lack short chain aliphatic glucosinolate, suggesting that CYP79F1 metabolizes the short chain methionine derivatives. In contrast, in cyp79F2 mutant plant, the levels of longchain aliphatic glucosinolates is significantly reduced and the level of short chain aliphatic glucosinolate is not affected, suggesting that CYP79F2 met abolized the longchain methionine derivatives (Hansen et al., 2001; Chen et al., 2003). In addition, the substrate specificities of CYP79F1 and CYP79F2 are further supported by the biochemical characterization that CYP79F1 metabolizes monoto hexa h omomet hioni n e
21 whereas CYP79F2 exclusively metabolizes pentaand hexahomomethionines (Hansen et al., 2001; Chen et al., 2003). Despite the fact that sps/bus/cyp79F1 plants resemble some hormone mutants, the role of CYP79F1 in hormone homeostasis in planta is elu sive (Tantikanjana et al., 2001). The regiospecific post aldoxime enzymes are less specific for the sidechains considering the fact that plants can transform nonendogenous or artificial aldoximes into glucosinolates (Petersen et al., 2001). In the second step of glucosinolate core structure biosynthesis, aldoximes are oxidized by cytochromes P450 in the CYP83 family to yield the reactive acinitro compounds or nitrile oxides intermediates (Barlier et al., 2000; Bak et al., 2001). It has been determind that CYP83A1 primarily metabolizes aliphatic aldoximes, and the aromatic and indolic aldoximes derived from phenylalanine, tyrosine and tryptophan are primarily metabolized by CYP83B1 (Bak and Feyereisen, 2001; Naur et al., 2003). Interestingly, the cyp83B1 m utant plants displayed a characteristic highauxin phenotype, hereafter it is referred as sur2 highlighting the existence of a common metabolic link IAOx, which act as the joint intermediate between indole glucosinolates and the plant hormone indole3 ac etic acid (IAA). Consequently, the cyp83B1 mutants gain their phenotype by channeling excess IAOx into IAA because of the blockage of the post aldoxime reactions. This conclusion is further support ed by the finding that overexpression of CYP79B2 leads to t he increase of both indole glucosinolate levels and IAA, resulting in plants with phenotypes identical to knockout mutant (Mikkelson et al., 2000; Zhao et al., 2002). The products from the CYP83s are activated, oxidized form of the aldoxime, e g. an acinitro compound or their dehydrated analogs, nitrile oxides. Due to its instability,
22 the product has escaped isolation (Grubb and Abel, 2006). The enzyme catalyzing the incorporation of reduced sulfur is the only enzyme that remains to be discovered, and is proposed to be a glutathioneS transferase (GST) or GSTlike enzyme (Halkier and Gershenzon, 2006). Traditionally, cysteine was suggested as the li kely sulfur donating molecule, which is evidenced by in vivo feeding study (Wetter and Chisholm, 1968) as we ll as the finding that active acinitro compound can react efficiently with nucleophilic S donors to form S alkyl thiohydroximates (Bak et al., 2001; Handen et al., 2001). However, a recent study demonstrates that instead of cysteine, glutathione (GSH) is the bone fide sulfur donating molecule in glucosinolate biosynthesis (Schlaeppi et al., 2008; GeuFlores et al., 2009). Transformation of CYP79A2 and CYP83B1 into N. benthamiana led to a substantial accumulation of a GSH conjugate, the identity of which is confirmed by LC MS analysis using chemically synthesized S [(Z)phenylacetohydroximoyl] L glutathione as internal standard (GeuFlores et al., 2009). However, as the author s pointed out a metabolic bottleneck occur red when GSH was considered as the sulf ur donating molecule because the resulting GSH conjugate is not an acceptable substrate for the downstream C S lyase (GeuFlores et al., 2009). The problem has been recently resolved by the identification of a glutamine aminotransferase enzyme, which is e ncoded by GGP1 (At4g30530). This enzyme is glutamyl peptide bond in a GSH conjugate to create a free amino group which would be accepted by the C S lyase (Geu Flores et al., 2009) S alkylthiohydroximate conjugates are cleaved by a C S lyase into thiohydroximic acid, pyruvate and ammonia. The identification of a C S lyase involved in glucosinolate biosynthesis was unsuccessful until the similarity between animal C S lyases and the
23 protein encoded by Superroot ( SUR1 ) from Arabidopsis was recognized through the bioinformatic approaches (Mikkelsen et al., 2004). This is further evidenced by the complete absence of aliphatic and indole glucosinolates in the sur1 knockout mutant, indicating that the SUR1 encodes a single gene and is responsible for gluco sinolate synthesis wit hout side chain specificity. This suggests that resembling the sur2 mutant, the highauxin phenotype of sur1 is caused by the accumulation endogenous IAOx which is subs equently channeled into the biosynthesis of IAA (Mikkelsen et al., 2004). A key direct ion for future studies of this gene is to characterize its potential to form a metabolon with CYP83 and an S donating enzyme in order to conduct this specific sulfur metabolism without losing reactive su lfur intermediates to the other route I n view of the fact that thiohydroximates are unstable and reactive, the remaining enzymes in the core pathway are considered to be enzymes recruited from the detoxification processes. A UDP glucose dependent glucosyltransferase, UGT74B1 (At1g24100), catalyzing the spec ific thiohydroximates to the corresponding desulfoglucosinolates was c haracterized (Grubb et al., 2004). Aliphatic and indolyl gluco si nolates are significantly decreased in ugt74B1 mutants but not completely abolished, suggestive of the presence of addit ional enzymes active ly toward thiohydroximates (Grubb et al., 2004). The final step in synthesis of the core structure is the 3phosphoadenosine 5 phosphosulfate (PAPS) dependent sulfation of desulfoglucosinolates. Three PAPS:desulfoglucosinolate sulfotr ansferase, AtST5a AtST5b and AtST5c were identified in Arabidopsis (Piotrowski et al., 2004). AtST5a catalyzes the sulfation reaction with the preference of tryptophanand phenylalaninederived
24 desulfoglucosinolates, whereas longchain aliphatic desulfoglucosinolates are the preferred substrates of AtST5b and AtST5c (Piotrowski et al., 2004). Aliphat ic glucosinolate biosynthesis: secondary t ransformations The initially formed parent glucosinolates can be subject to a wide range of sidechain modifications (Grubb and Abel, 2006). These reactions are of particular importance because they not only influence the route of glucosinolate hydrolysis but also the activity of the hydrolysis products (Halkier and Gershenzon, 2006). The side chains of methioninederi ved aliphatic glucosinolates are especially entailed by oxidations, eliminations, alkylations or esterifications to produce a substantial structural variation such as methylthioalkyl, methylsulfinylalkyl, alkenyl, hydroxyalkenyl and benzoylxyl glucosinolat es (Figure 13, Grubb and Abel, 2006). To date, in Arabidopsis, the genes regulating the modifications have been functionally characterized, including GSL OX GSL ALK and GSL OH The conversion of methythioalkyl into methysulfinylalkyl glucosinolates is under the control of the glucosinolate S oxygenation ( GS OX ) QTLs in both Arabidopsis and Brassica napus (Kliebenstein et al., 2001a, 2001b), as suggested by quantitative traits locus analysis (QTL) but the genetic basis of this locus is not determined unt il recently. FMOGS OX1, encoding a flavinmonooxygenase enzyme, was identified to catalyze the conversion of methythioalkyl into methysulfiny lalkyl GSLs In comparison to the almost complete conversion of methythioal kyl into methylsulfinylalkyl glucosinolate s in the FMOGS OX1 overexpression lines, the ratio of methylthioalkyl: methysulfinylalkyl was significantly increased (Hansen et al. 2007). This is further support ed by the biochemical result that the heter o logously expressed FMOGS OX1 was able to catal yze the S ox ygenation of methylthioalkyl glucosinolates (Hansen et al., 2007). In
25 Arabidopsis, there are 29 putative FMO genes, which could be grouped into three major clades as illustrated after constructing a phylogenetic tree of plant FMOs FMOGS OX1 w a s grouped w ithin a cluster of seven genes four of which (FMOGS OX2 to FMOGS OX5) were capabl e of S oxygenating aliphatic glucosinolate s with distinct substrate specificity (Li et al., 2008). The conversion of methylsulfinylalkyl to alkenyl and hydroxylal kyl glucosinolate s in Arabidospis is controlled by the GSL AOP locus, which encodes two tandem ly linked and duplicated 2oxoaciddependent dioxygenases ( 2 ODDs ). They are members of a large family with at least 100 members in Arabidopsis, catalyzing a range of reactions such as hydroxylations, epoxidations and desaturations (Figure 1 3, Kliebenstein et al., 2001b). AOP2 catalyzes the conversion of methylsulfinylalkyl to alkenyl glucosinolates and is only expressed in ecotypes accumulating alkenyl glucosinol ate s. In contrast AOP3 directs t he formation of hydroxyalkyl glucosinolates from methylsulfinylalkyl glucosinolates and is expressed only in ecotypes accumulating hydroxyalkyl glucosinolates (Figure 13 Kliebenstein et al., 2001b). The natural variations of AOP activity among ecotypes is due to changes in the expression of the AOP2 and AOP3 genes, and a deletion in the open reading frame of AOP2 which cause a truncated protein (Kliebenstein et al., 2001b). GSL OH locus (At2g25450), which is responsible for hydroxylating but 3 enyl glucosinolates to 2hydroxybut 3 enyl glucosinolates in planta, was recently cloned and characterized (Hansen et al., 2008). TDNA insertional mutant in this gene completely abolish the presence of 2hydroxybut 3 enyl glucosinolate. The other important feather of this enzyme is the stereoselectivity of its 2 hydroxylase which varies among different
26 plants species within the Brassicales, leading to the produc tion of both the 2R and 2S enantiomers in a constant ration (2R:2S) of 1:3 in Arabidopsis but only the production 2R enantiomer in Brassica napus (Hansen et al., 2008). The gene which encodes the enzyme catalyzing the production of benzoyloxyglucosinolate from hydroxyglucosinolate was identified to be BZO1 (At1g65880) (Kliebenstein et al., 2007). The functional involvement of this enzyme in glucosinolate biosynthesis was supported by genetic evidence that the benz o yloxy glucosinolate levels was significantly decreased in bzo1 mutant and by biochemical evidence that it encoded a functional benzo yl CoA ligase, catalyzing the esterification of the free hydroxyl groups of benzoic acids (Kliebenstein et al., 2007) Interestingly, the longchain aliphatic gluco si nolates are elevated in bzo1 mutant, suggesting metabolic competition f or the common short chain aliphatic glucosinolate precursors between both longchain aliphatic and benzoyloxyglucosinolate biosynthesis (Kliebenstein et al., 2007). Regulation of Glucosinolate Biosynthesis: Signaling Networks Unlike the well elucidated biosynthetic pathway, where most of the genes have been characterized in Arabidopsis, the regulators of glucosinolate biosynthesis have just started to be discovered, including the subgroup 12 R2R3MYB transcription factors and components acting upstream such as IQD1 SLIM1 or AtDof1.1 (Gigolashvili et al., 2008). To unravel the signal transduction cascade from incoming signals to early response genes and further downstream regulators, as well as the new interacting partners and targeted genes of those regulat ors is our future goal to fully understand the complex molecular network s controlling glucosinolate biosynthesis (Yan and Chen, 2007)
27 T ranscriptional factors controlling aliphatic glucosinolate biosynthesis MYB28 MYB76 and MYB29 also referred to as HAG1 (high aliphatic glucosinolate1/ MYB28 ) HAG2 and HAG3, respectively, have been characterized as transcriptional activator s of aliphatic glucosinolate biosynthesis through three different approaches independently In 2007, Hirai et al studied the function of MYB28 and MYB29 using an integrated omics approach assuming that a set of coexpressed genes are very likely to involve in the same or related metabolic pathway s. MYB28 and MYB29 were found to be closel y co regulated with all the known genes in glucosinolate metabolism. In addition, QTL analysis revealed MYB28 as a regulator of aliphatic glucosinolates MYB28 was shown to be located within a genomic region determining the aliphatic glucosinolate concentration and the transcript levels of genes involved in t he aliphatic glucosinolate biosynthesis (Sonderby et al., 2007). Furthermore MYB28 MYB29 and MYB76 were identified in a screen for their trans activation potential toward aliphatic glucosinolate biosynthetic genes (Gigolashvili et al., 2007b 2008). Over expression of MYB28 MYB29 and MYB76 resulted in an increase in total aliphatic glucosinolates but not indole glucosinolates in leaves and suspension cells. The expression levels of aliphatic glucosinolate biosynthetic genes were increas ed, while those of indole gluco s i nolate genes were repressed, suggesting reciprocal negative control of aliphatic and indole glucosinolate pathways (Hirai et al., 2007; Sonderby et al., 2007; Gigolashvili et al., 2007b, 2008). Knockout mutants defective in MYB28 showed signi ficantly decreased levels of both short and longchain aliphatic glucosinolates, whereas only the short chain products were reduced in myb29 mutant, suggesting that MYB29 regulates the production of short chain aliphatic glucosino lates, while MYB28 control both short and longchain glucosinolates. Double knockout mutant
28 myb28 myb39 almost completely abolished the production of aliphatic glucosinolates, suggesting that MYB28 and MYB29 were the master factors and MYB76 has an accessory role in regulating the methioninederived aliphatic glucosinolate biosynthesis (Sonderby et al., 2007; Gigolashvili et al., 2007b, 2008). Transcripti onal factors controlling indole glucosinolate biosynthesis The first positive regulator of indole gluco si nolate biosynthesis is Altered T r yptophan R egulation 1 ( ATR1/MYB34 ). A dominant overexpression allele, atr1D confers constitutively activated expression of tryptophan synthesis ( ASA1 and TSB1 ) and indole glucosinolate genes ( CYP79B2 CYP79B3 and CYP83B1 ) genes, while the expres sion of CYP79F1 is not altered. Compared to wild type, there were a ten fold increase in the accumulation of indole glucosinolates and a two fold increase in IAA levels, but the levels of aliphatic glucosinolates did not change (Celenza et al., 2005). Loss of function mutant of atr1 1 shows a reduced expression of core genes ( CYP79B2 CYP79B3 and CYP83B1 but not ASA1 and TSB1 ) and decreased levels of indole glucosinolates (Celenza et al., 2005). Phylogenetic analysis showed that two other MYB factors, MYB5 1 and MYB122 were closely related to MYB34 and all of them belonged to the subgroup 12 of the R2R3MYB transcription factor family. Subsequent studies demonstrated that MYB51 and MYB122 were also the positiv e regulators controlling indole glucosinolate biosynthesis. The role of MYB51 was confirmed by the elevated accumulation of I3M in MYB51 overexpression lines and decreased levels of I3M in a myb51 knockout mutant (Gigolashvili et al., 2007a). Likewise, overexpression of MYB122 caused an increased accum ulation of I3M and IAA, but only in the presence of a functional MYB51 suggesting that the coordination between MYB51 and MYB122 in
29 the regulation of indole glucosinolate biosynthesis and IAA homeostasis (Gigolashvili et al., 2007a) Other known regulator s of glucosinolate biosynthesis IQD1 encodes a basic nuclear localized calmodulinbinding protein, which modulates the expression of both aliphatic and indole glucosinolate biosynthetic pathway genes (Levy et al., 2005). Gainand loss of function iqd1 alleles correlated with significant but mild increase and decrease in glucosinolate accumulation, respectively. Overexpression of IQD1 induced the expression of the genes encoding for key enzymes o f the indole glucosinolate biosynthesis, while genes encoding enzymes related to aliphatic glucosinolate biosynthesis and glucosinolate degradation were reduced in expression. Although expression of IQD1 did not seem to be substantially regulated by the class ical plant defense hormone signaling pathways, it was moder ately stimulated by mechanical stimuli. This suggests that IQD1 may integrate early woundor pathogen/elicitor induced changes of cytoplasmic Ca2+ signatures to coordinate an array of defense responses, inc luding glucosinolate production (Levy et al., 2005). SLIM1 an ethyleneinsensitive3 like transcriptional factor, functions as a central transcriptional regulator controlling sulfate uptake, repressing the biosynthesis of glucosinolates and activating the degradation of glucosinolates under sulfur deficiency conditions (MaruyamaNakashita et al., 2006). Another known transcription factor positively controlling glucosinolate biosynthesis is AtDof1.1 (DNA binding with one finger). It was shown to regulate the transcription of at least CYP83B1 Overexpression of AtDof1 result ed in the moderate increase in the levels of aliphatic and indole glucosinolates. The expression of AtDof1.1 was activated by mechanical wounding and herbivore attack (Skirycz et al., 2006)
30 Signaling networks of glucosinolate biosynthesi s Glucosinolate metabolism has been shown to be responsive for many different environmental or endogenous stimuli such as pathogen challenge, herbivore damage, mechanical wounding, altered mineral nutrit ion, jasmonate, salicylic acid and ethylene levels (W ittstock and Halkier,2002; MaruyamaNakashita et al., 2003; Mewis et al., 2005; Grubb and Abel, 2006; Yan and Chen, 2007; Gigolashvili et al., 2009). An integrated transcriptomics and metabolomics analysis discovered coordinated expression of many genes in glucosinolate biosynthesis under sulfate deficiency conditions (MaruyamaNakashita et al., 2003). A bioinformatic approach based on transcriptional coexpression analysis revealed that almost all the genes involved in methionine chainelongation cycle and the core biosynthetic pathways were coordinately regulated (Hirai et al., 2007). It has been mentioned in several studi es that pathogenand herbivoreinduced pathway gene expression and glucosinolate production are mediated by major plant hormones associated with specific and broadspectrum defense, such as jasmonic acid (JA), salicylic acid (SA) and ethylene (Kliebenstein et al., 2002; Reymond et al., 2004; Mewis et al., 2005; Grubb and Abel, 2006). E xogenous application of methyl jasmonate (MeJA) led to an increase in both aliphatic and indole glucosinolates via multiple signaling pathways (Brader et al., 2001; Mikkelsen et al., 2003; Mewis et al., 2005). The expression of MYB29 but not that of MYB28 was induced by MeJA treatment, suggesting the involvem ent of MYB29 in MeJA signaling (Mikkelsen et al., 2003). Although MYB76 seemed to play a minor role in the regulation of aliphatic glucosinolate biosynthesis under normal condition, the expression of MYB76 could be dramatically stimulated by mechanical sti muli such as wounding, highlighting its key function in
31 glucosinolate metabolism under mechanical stress (Gigolashvilie et al., 2008). Interestingly, in myc/jin1 mutant defective in a gene known as jasmonate signaling component, the MeJAinduced expression of MYB51 was enhanced compared to the wild type. H owever, the MeJAmediated stimulation MYB34 was reduced, indicating that MYC2/JIN1 acted as a positive regulator of MeJAmediated MYB34 expression, but as a negative regulator of the MeJAmediated MYB51 in duction (Dombrecht et al., 2007). In addition to its association with MeJAdependent signaling, MYB51 was identified to play an important role in early response to other biotic stresses through mediating the accumulation of indole glucosinolates. For example, MYB51 expression was induced by pathogen attack or the application of pathogen elicitor s (Thilmony et al., 2006), and it was significantly stimulated by wounding in a rapid but transient mode (Gigolashvili et al., 2007). In addition, overexpression of MYB51 resulted in an increased resistance to the generalist herbivore (Gigolashvili et al 2007). Conclusions and Future Prospects Using multidisciplinary approaches including molecular biology genetic s and functional genomics, there has been remarkable pr ogress in glucosinolate research in recent years The core biosynthetic pathway has been elucidated, a suite of regulators of aliphatic and indole glucosinolates have been characterized, and metabolic and evolutionary links to closely related pathway s have been identified (Yan and Chen 2007) Although much has already been learned, much more await s exploration so that we can completely understand why and how plant s produce glucosinolates. Current research increasingly focuses on regulatory mechanism s of glucosinolate synthesis, distribution and degradation, defense against bacterial and fungal pathogens, as well as met abolic engineering of custom glucosinolate profiles. An ultimate understanding of the
32 regulation of glucosinolate biosynthesis will not only r equire the clarification of the signaling cascade from incoming signals to the downstream regulators, but also require knowledge about the glucosinolate biochemistry including subcellular localization of enzymes, intracellular channeling and trafficking of intermediates, major flux control ling regulators and metabolon organi zation as well as enzyme catalysis regulated by covalent modifications and effectors. Characterization of the regulatory and metabolic networks controlling glucosinolate metabolism would enable rational engineering of glucosinolate metabolism for human health and plant defense.
33 Figure 11. Chainelongation process of aliphatic glucosinolate biosynthesis in Arabidopsis (left panel) and leucine biosynthesis in microbes (right panel). The dotted line indicates further chainelongation of 2oxo acids. 2 -Isopropylmalate IPMS 2,3 -Dihydroxy -3 -isovalerate 3 -Methyl -2 oxobutyrate 3 -Isopropylmalate 4 -Methyl -2 oxovalerate LEU1 LEU2 BCAT Leucine Methionine BCATs MAM1, MAM3 4 -Methylthio2 -oxobutyrate 2 -(2 -Methylthio)ethymalate 3 -(2 -Methylthio)ethymalate AtIPMDH1 5 -Methylthio2 -oxopentanoate BCATs Chain-elongated methionine?AtLeuC/AtLeuDs
34 Figure 12. Biosynthesis of the core glucosinolate structure. Abbreviations: R, variable side chain; CYP79, cytochrome P450 monooxygenases of the CYP79 family; CYP8 3, cytochrome P450 monooxygenases of the CYP83 family; GST, glutathione S transferase; GATase, glutamine ami n otransferase; UGT, UDP dependent glucosyl transferase; ST, sulfotransferase. R NH2COOH Amino Acid CYP79 Oxidation R N Aldoxime CYP83 Oxidation R N Aci Nitro compound OH-O Conjugation GST? R N GSH conjugate OH S (GSH) R N OH S (Cys -Gly ) Cleavage GATase Cys -Gly conjugate C -S lyase C -S cleavage R N OH SH Thiohydroximic acid UGT Glucosylation R N OH S-Glc Desulfo glucosinolate S T Sulfation R N OSO3 S-Glc Glucosinolate
35 Figure 13. Side chain modification of methionine derived glucosinolates in Arabidopsis. FMO indicate one small family of the flavin monooxygenase; AOP2, AOP3 and GSL OH represent the corresponding the 2oxoaciddependent dioxygenase catalyzing these reaction types in Arabidopsis; and BZO1 encode a benzoylCoA ligase. (CH2)nN OSO3 S-Glc Methylthioalkylglucosinolate S (CH2)nN OSO3 S-Glc Methlsulfinylalkylglucosinolate S O FMO (CH2)nN OSO3 S-Glc Hydroxyalkyl glucosinolate HO (CH2)nN OSO3 S-Glc Alkenyl -glucosinolate AOP2 AOP3 (CH2)nN OSO3 S-Glc Benzoyloxyl-glucosinolate BZO1 OBz (CH2)nN OSO3 S-Glc 2 -Hydroxy -3 butenyl-glucosinolate OH GSL OH Hydroxylation Esterification Alkylation Hydroxylation Oxidation
36 CHAPTER 2 A REDOX ACTIVE ISOPROPYLMALATE DEHYDROGENASE FUNCTIONS IN THE BIOSYNTHESIS OF GLUCOSINOLATE AND LEUCIN E IN ARABIDO PSIS Introduction Glucosinolates are a diverse group of sulfur containing plant specialized metabolites found in the order of Brassicales which includes many important vegetable and oi lseed crops and the model plant Arabidopsis (Grubb and Abel, 2006; Halkier and Gershenzon, 2006). Aliphatic glucosinolates are derived from methionine through a threest ep chainelongation cycle that sequentially adds one to six methylene groups to the methionine sidechain (Field et al., 2004; Schuster et al., 2006) (Figure 12). Although the initial and final reactions of this pathway are understood, little is known about the middle steps of this biosynthetic transformation. In the initial step of aliphatic glucosinolate biosynthesis, methionine is deaminated to the corresponding 2 oxo acid, i.e., 4methylthio2 oxobutanoic acid, by branchedchain aminotransferases ( BCAT 3 and BCAT4 ) (Schuster et al., 2006; Knill et al., 2008) (Figure 12). Methylthioalkylmalate synthases ( MAM1 and MAM3 ) catalyze the condensation of the 2oxo acid with acetylCoA to produce a substituted 2malate derivative (2(2 methylthioethyl) malate) (Textor et al., 2004; Textor et al., 2007). The 2malate derivative then isomerizes to a 3malate derivative that undergoes oxidative decarboxylation to yield a 2oxo acid extended by one methylene group. The enzymes that catalyze the isomerization and oxidation reactions remain to be characterized (Field et al., 2004; Textor et al., 2007; Knill et al., 2008). If the extended 2 oxo acid is transaminated by BCAT3 to form the corresponding elongated methionine (e.g., homomethionine), it enters into the glucos inolate pathway to yield C3glucosinolates (Knill et al., 2008). Alternatively, the extended 2oxo acid undergoes iterative
37 condensations with acetylCoA, the newly extended 2oxo acids can enter the pathway to yield C4 to C8 glucosinolates. A 2oxo acid based chainelongation system is also used in leucine biosynthesis (Chisholm and Wetter 1964; Massey et al., 1976; Halkier and Gershenzon, 2006). The chemical similarity and the proposed common origins of methionine chainelongation and leucine biosynthesis suggest that homologous genes encode the enzymes in these pathways (Schuster et al., 2006; Hirai et al., 2007) (Figure 1 1 ). The close relationship between the two pathways is also evident at the genetic level. The families of MAM and BCAT proteins invol ved in leucine biosynthesis are related to those found in the methionine elongation steps (Schuster et al., 2006; Binder et al., 2007). Moreover, recent evidence shows that BCAT3 potentially functions in both pathways (Knill et al., 2008). In the leucine biosynthesis pathway found in bacteria and yeast, isopropylmalate dehydrogenase ( IPMDH) catalyzes the NAD+dependent oxidation and decarboxylation of 3isopropylL malate to produce 4methyl2 oxovalerate that is further transaminated to le ucine (Drevland et al., 2007). In plants, the IPMDHs cloned from rape (Ellerstrom et al., 1992), potato (Jackson et al., 1993) and Arabidopsis (Nozawa et al., 2005) complement yeast with a mutation in the Leu2 gene, which encodes a functional IPMDH. The experiments suggest that a potential IPMDH functions in plant leucine biosynthesis, but functional analysis of such a plant enzyme remains to be performed. In this study, we describe the detailed functional characterization of an Arabidopsis IPMDH, IPMDH1 (At5g14200) that displays high amino acid sequence homology with other bacterial and yeast IPMDHs In vitro activity tests showed that the Arabidopsis enzyme exhibits isopropylmalate dehydrogenase activity. Disruption of IPMDH1 in
38 Arabidopsis caused a dramatic decrease in the concentrations of glucosinolates with side chains of four carbons or longer and a decrease in free leucine levels. The mutant chemical phenotypes were nearly completely complemented by the reintroduction of a functional IPMDH1 gene into the mutant. IPM DH1 activity is regulated by thiol based redox modification. Together with the subcellular localization of the protein in the chloroplast stroma and the gene expression pattern in different tissues, these results indicate that IPMDH1 plays an important rol e in the biosynthesis of aliphatic glucosinolates and participates in leucine biosynthesis. Materials and Methods Plant Materials and Chemicals Seeds of Arabidopsis ecotype Columbia (Col 0, CS3879) and Salk mutant lines ipmdh1 (Salk_063423C), ipmdh2 (Sa lk_152647C) and ipmdh3 (Salk_142322C) were obtained from the Arabidopsis Biological Resource Center. The seeds were sterilized using 50% bleach for 10 min, and thereafter washing four times with sterilized water. Seeds were germinated on a half strength Murashige Skoog agar medium containing 1% sucrose and transferred to a growth chamber under a photosynthetic flux of 140 mol photons m2s1 with a photoperiod of 16 hours at 22C and 18C at night, and 70% relative humidity for 10 days. The seedlings were t ransferred to soil and grown under the same conditions. For oxidative stress experiments, Arabidopsis seedlings were grown and treated as previously described (Hicks et al., 2007). Unless otherwise specified, all the chemicals and reagents were obtained fr om Sigma Aldrich, Merck or Fisher Scientific except that 3isopropylmalate was purchased from Wako Pure Chemical (Osaka, Japan).
39 DNA Extraction and Genotyping Genomic DNA was extracted from 3week old rosette leaves by grinding with sand in 100 mM Tris H Cl, pH 9.5, 1 M KCl and 10 mM EDTA. After incubation at 70 C for 30 min and centrifugation for 10 min, the supernatant was transferred into a new tube and dilute d two fold with MilliQ water (Millipore, USA). PCR reactions were performed in a total volume 2 forward and reverse primers, and 1 unit of Taq DNA polymerase (Invitrogen, USA). The PCR program was as follows: 30 s at 94 C, 30 s at 50 C 55 C (annealing temperature dependent on the sp ecific primer pairs, Table 11), and 1 min/kb at 72 C for a total of 37 cycles. The TDNA specific primers were LBa1 and LBb1 and the IPMDH1 specific primers were IMD1 LP and IMD1RP. RNA Extraction, RT PCR and Arabidopsis Transformation RNA was extracted from different tissues using an RNeasy plant mini kit (Qiagen, USA). cDNA was synthe 20 according to the kit instructions. PCR conditions were the same as those described for genomic PCR in the previous section, except the cycles for tissue specific expression and wounding were 32 and 25, respectively, using the IPMDH1 specific primers IMD1 2L and IMD12R. Actin1 F and actin1R primers were use to amplify Actin1 for control purposes. IPMDH1 was amplified using primers IMD13L with IM D1 3R, and cloned into the pGEM T vector. Restriction sites of NcoI and PmlI were introduced by PCR using primers IMD14L and IMD1 4R. After verification by sequencing, the gene was subcloned into the pCAMBIA1305 to construct 35S:: IPMDH1. Agorbacterium tum efaciens strain GV3101
40 was used to transform the construct through floral dipping (Clough and Bent, 1998). Promoter GUS Fusion and GUS Assay The IPMDH1 promoter (from 1834 to 53bp) was amplified from genomic DNA using primers IMD 1L and IMD 1R, and cloned into a pGEM T vector (Promega, USA). The promoter sequence was subcloned into pCAMBIA1305 using Sac I and Nco I restriction sites. Histochemical GUS assays were performed on T2 transgenic plants as described previously (Jefferson et al., 1987). Immunocytochemistry and Chloroplast Fractionation Arabidopsis leaf discs were loaded in hexadecane, frozen in a highpressure freezer (Baltec HPM 010; Technotrade, USA), and transferred to liquid nitrogen. Substitution was performed in 0.1% uranyl acetate in acetone at 80C for 172 h and warmed to 50C. After several acetone rinses, samples were infiltrated with Lowicryl HM20 (Electron Microscopy Sciences, USA) for 48 h and polymerized at 50C under UV light for 24 h. Sections were mounted on formvar coated nickel grids and blocked for 30 min with 20% (w/v) fetal bovine serum albumin in TBST (30 mM Tris, 37 mM NaCl, pH 7.5, 0.05% Tween20). Sections were incubated with IPMDH1 antibody for 1 h. After rinsin g with TBST, they were transferred to the secondary antibody conjugated to 15 nm gold particles for 1 h. Controls were performed by omitting the primary antibody. The sections were observed using a Leo 912 transmission electron microscope (Zeiss, Germany). Chloroplasts were isolated from 4week old rosette leaves and purified as previously described (Aronsson et al., 2002). Purified chloroplasts were resuspended in a lysis buffer (62.5 mM Tris HCl, pH 7.5, 2 mM MgCl2 and 1% (v/v) protease inhibitor
41 cocktail from Sigma Aldrich, USA) and incubated for 15 min on ice with occasional stirring. The lysate was centrifuged at 14,000 g at 4 C for 5 min. The supernatant contained the stroma fraction and the pellet contained chloroplast envelope and thylakoids (Smith et al., 2003). Glucosinolate and Amino Acid Analysis Seeds (20 mg) and 4week old rosette leaves (100 mg) were used for glucosinolate and free amino acid analysis. Glucosinolates were analyzed using HPLC MS as previously described (Chen et al., 2003; Alva rez et al., 2008). Free amino acids were extracted using a methanol and chloroform method (Colebatch et al., 2004; Schauer et al., 2005) and profiled using an HPLC based precolumn derivatization protocol (Schuster et al., 2006). Protein Expression, Wester n Blotting and Visualization of Protein Sulfhydryl Groups The cDNA of IPMDH1 without the N terminal plastidtargeting peptide was cloned into pET28a expression vector (Novagen, USA) using primers IMD15L and IMD15R. To generate the cysteine mutants of IPM DH1, the two cysteine residues were substituted by serine residues using a sitedirected mutagenesis kit (Stratagene, USA) using primers IMD16L and IMD16R for C232S, and IMD17L and IMD17R for C390S. The fidelity of the mutation in the constructs of IPM DH1C232S, IPMDH1C390S and IPMDH1C232S/C390S was confirmed by DNA sequencing. The constructs were expressed in E. coli strain BL21(DE3) respectively by growing at 37 C in Terrific Broth medium to an OD600 of 0.6 and then inducing wit h 1mM IPTG at 20 C for 15 h. IPMDH1 and mutant proteins were purified as His tagged proteins using Mi di PrepEase kit (EMD, USA). Arabidopsis Trx f1 (At3g02730) and Trx m1 (At1g03680) were cloned into the
42 pET28a vector using the primers TRXF11L and TRXF1 1R, and TRXM11L and TR XM1 1R, respectively. The expression and purification of the recombinant proteins were the same as described above. Purified proteins were dialyzed against 20 mM Tris HCl, pH 8.0 and 1 mM phenylmethylsulfonyl fluoride at 4 C overnight, and then transferre d to fresh buffer for another 5 h. The protein preparations were concentrated by ultrafiltration with a 10kDa cutoff membrane (Amicon, USA) at 4 C. Protein concentration was determined by the Bradford protein assay with bovine serum albumin as a standard. The homogeneity of the purified proteins was tested by SDS PAGE analysis and the identity was conf irmed by HPLC MS. The purified IPMDH1 (1 mg) was sent to Cocalico Biologicals Inc. (Pennsylvania, USA) for antibody produc tion. Immunoblotting with the IPMD H1 antibody (1:2000), anti his tag antibody (1:10,000) (Invitrogen, USA), RbcL antibody (1:500) and PsbO antibody (1:500) was done as previously described (Chen and Halkier 1999). To visualize protein sulfhydryl groups, proteins were incubated with 90 mM monobromobimane (mBBr) in the dark for 30 min. Proteins were then separated on a 15% (w/v) nonreducing polyacrylamide gel. Following electrophoresis, the gels were fixed in 12% trichloroacetic acid for 1 h and then incubated overnight in 40% methanol and 10% acetic acid. Fluorescently labeled protein bands were visualized under UV light and imaged using an ImageQuant CCD camera (GE Healthcare, USA). Redox and Enzymatic Activity Assays Freshly prepared protein (5 ) uCl2 for 1 h at 30 C. Oxidized protein was dialyzed against 25 mM Tris HCl (pH 7.5) to remove CuCl2 incubated with 5 mM DTT or with different concentrations of reduced thioredoxin
43 proteins at 30 C for 1 h. The redox state of the IPMDH1 protein was analyzed using nonreducing SDS gels. The enzyme assay mixture consisted of 100 mM Tris HCl buffer, pH 7.5, 1 mM MgCl2, 100 mM KCl, 5 mM NAD+ and 1 mM 3isopropylmalate (different concentr ations for kinetics analysis). The reaction was conducted in a quartz cuvette and initiated by mixing with 10 g enzyme. Enzyme activity was measured by monitoring the production of NADH at 340 nm using a Beckman DU 65 spectrophotometer. Results IPMDH1 Is Involved in Glucosinolate and Leucine Biosynthesis Transcriptional co expression analysis is a powerful tool for revealing genes involved in linked processes and pathways (Persson et al., 2005; Hirai et al., 2007). To explore the potential pathways where I PMDH1 (At5g14200) may function, IPMDH1 was used to search a public co expression database (http://www.arabidopsis.leeds.ac.uk/act/) (Manfield et al., 2006). As suspected, the IPMDH1 gene exhibited strong coexpression with many genes known to function in g lucosinolate biosynthesis (Table 22, Gigolashvili et al., 2009). In particular, BCAT3 BCAT4 and MAM1 the three key players in the methionine chainelongation cycle (Binder et al., 2007), exhibited strong positive correlation with IPDMH1 In addition, a large subunit and two small subunits of isopropylmalate isomerases, which might catalyze the isomerization step of methionine chainelongation, also exhibited high coexpression with IPMDH1 When any of the known genes in glucosinolate biosynthesis was used as query, IPMDH1 did show up with a high score (data not shown), indicating that IPMDH1 is regulated in a similar manner as glucosinolate biosynthetic genes. Furthermore, IPMDH1 exhibits a high similarity to known and potential isopropylmalate
44 dehydrogenases in other organisms (Figure 21) (Hsu and Kohlhaw 1980; Ellerstrom et al., 1992; Wallon et al., 1997; Nozawa et al., 2005). It is therefore conceivable that IPMDH1 encodes a plant isopropylmalate dehydrogenase likely to function in leucine biosynthesis and in the methionine chainelongation of glucosinolate biosynthesis. IPMDH1 Exhibits 3Isopropylmalate Dehydrogenase Activity To i nvestigate the possibility of IPMDH1 being an isopropylmalate dehydrogenase as suggested by the bioinformatic analysis, IPMDH1 gene was overexpressed in E. coli without its plastid targeting sequence (Figure 21) and the protein was purified for detailed biochemical studies. A typical preparation with 150 ml bacterial culture produced about 4 mg IPMDH1 protein, which appeared on a reducing SDS gel as one band of approximately 45 kDa, comparable to its theoretical mass of 43.4 kDa. The intactness and identity were characterized by Western blotting and mass spectrometry (MS) (Figure 2 2). When tandem MS spectra were searched agains t total NCBI database, IPMDH1 was unambiguously identified as the top hit. A protein from Brassica napus was also identified because it shared high homology to IPMDH1 No other proteins were identified in the pr otein sample, indicating that IPMDH1 of high purity and homogeneity was obtained (Figure 22). The activity of IPMDH1 was tested with different potential substrates (Kawaguchi et al., 2000; Drevland et al., 2007). The protein exhibited Michaelis Menten kinetics and showed a high activity towards 3i sopropylmalate (0.93 0.01 mol/minmg protein) and NAD+ (1.01 0.01 mol/minmg protein) and a low activity towards D malate (0.16 0.03 mol/minmg protein). The Km values for 3isopropylmalate, NAD+ and D malate were 25.2 M, 187.1 M and 1.1 mM, res pectively (Table 23). No activity could be detected with 2isopropylmalate, NADP+, L malate, citrate and isocitrate. Oxidization of
45 IPMDH1 by 50 M CuCl2 led to decreased activity towards 3isopropylmalate, NAD+ and D malate (Table 23). S imilar to other known IPMDHs, IPMDH1 had a pH optimum of 7.6 and a temperature optimum of 60C (Figure 23). These experiments demonstrated that IPMDH1 is an isopropylmalate dehydrogenase. Because of the lack of 3methylthioalkylmalate derivatives, the intermediates of the methionine chainelongation cycle, no biochemical reactions could be conducted to test oxidative decarboxylation of 3 methylthioalkylmalates to 2oxo acids in the methionine chainelongation cycle. Glucosinolate Profiles and Leucine Biosynthesis Are Substantially Affected by Mutation of IPMDH1 A T DNA mutant of the IPMDH1 gene was identified from the SALK collection (Alonso et al., 2003) and homozygotes were selected. The position of the T DNA insertion was located by genomic PCR with genespecific and TDNA specific primers followed by sequencing of the resulting PCR product (Figure 24A). Disruption of gene transcription by TDNA insertion was confirmed by RTPCR. Using gene specific primer pairs spanning the TDNA insertion site, the transcript of IPMD H1 was detectable in wild type plants but not in the mutant (Figure 24B), further confirming the complete knockout of the gene expression and the homozygous state of the mutant. To examine the role of IPMDH1 in glucosinolate biosynthesis, glucosinolates in the mutant were analyzed. Compared to wild type plants, ipmdh1 mutant showed dramatic changes in glucosinolate profiles. In both seeds and leaves, C4 glucosinolates were reduced by approximately 50% in the mutant and longer aliphatic glucosinolates were reduced to extremely low levels. In mutant leaves, C6 to C8 glucosinolates were not detectable (Figure 25). Meanwhile, there was a commensurate accumulation of C3 glucosinolates as well as a marked increase in indole glucosinolates (Figure 25). In
46 spi te of the increase of C3 glucosinolates, the total level of aliphatic glucosinolates in the seeds of the ipmdh1 mutant was about 70% of the wild type plants. In leaves, the total levels of aliphatic glucosinolates were not significantly different from the wild type plants To investigate whether the changes in glucosinolate profiles observed in ipmdh1 is d irectly caused by the loss of IPMDH1 function, we generated ipmdh1 complementation plants expressing full length IPMDH1 driven by the cauliflower mosaic v irus 35S promoter. Out of 20 independent transformants, 15 showed identical complementation profiles in glucosinolates. In both seeds and leaves, the levels of C4 and longer aliphatic glucosinolates increased to levels comparable to those in wild type plants. Meanwhile, C3 and indole glucosinolates decreased to approximate wild type levels (Figure 25). The rescue of the altered glucosinolate profiles in ipmdh1 mu tant by ectopic expression of IPMDH1 strongly suggests that IPMDH1 plays an important role in t he chainelongation process of aliphatic glucosinolate biosynthesis. As predicted by bioinformatic analysis, if IPMDH1 functions in leucine biosynthesis, disruption of IPMDH1 gene might cause reduction of cellular free leucine levels. We therefore determi ned the free amino acid levels in the ipmdh1 plants. In the seeds of ipmdh1, free leucine levels were reduced to 54% of the levels in wild type. Isoleucine levels showed a slight decrease, phenylalanine showed 18% increase and tryptphan showed about 10% reduction compared to wild type (Figure 26). These results suggest that IPMDH1 contributes to leucine biosynthesis. In leaves, however, no significant difference in the levels of leucine and other amino acids was observed between wild type and the mutant (data not shown). This result is consistent with a
47 recent report (Sawada et al., 2009) and may not be surprising, considering the existence of possibly redundant IPMDH1 homologs and the nature of leaves as source tissues. IPMDH1 Expression Displays Temporal and Spatial Regulation and Is Triggered by Wounding To investigate the tissuespecific expression pattern of IPMDH1 we generated transgenic Arabidopsis containing an IPMDH1 glucuronidase (GUS) construct. Of 12 independent transgenic lines, 11 displayed similar patterns of GUS expression. Three days after germination, a strong GUS staining was observed in all the tissues except root tip (Figure 27A). At six days, intense staining was found in cotyledons, newly emerging leaves and the lower regi on of roots (Figure 27B). At rosette stage, IPMDH1 promoter activity was predominantly present in the main veins of leaves (Figure 27C). At flowering stage, blue staining was found in petal and pistil, as well as in the two ends of young siliques (Figure 2 7D, E). Many glucosinolate biosynthetic genes are induced by wounding and methyl jasmonate (MeJA) (Mikkelsen et al., 2003; Schuster et al., 2006; Alvarez et al., 2009). To test whether IPMDH1 is responsive to wounding and MeJA, leaves were clipped with scissors, squeezed with forceps, pierced with a pin as previously reported (Schuster et al., 2006), and incubated different ways of wounding (Schuster et al., 2006) (Figure 27F) and by MeJA treatment (Figure 27G). The levels of IPMDH1 expression was also analyzed by RTPCR. IPMDH1 was expressed in all the tissues examined with high expression in leaves, stems and roots, and low expression in flowers and siliques (Figure 2 7H). The experiments have also
48 confirmed the wounding induction of IPMDH1 (Figure 27I). Collectively, the profile of IPMDH1 expression exhibited the typical pattern observed for genes involved in the formation of aliphatic glucosinolates (Chen et al., 2003; Mikkelsen et al., 2003; Schuster et al., 2006). IPMDH1 Protein and Activity Are Localized to Chloroplast Stroma All potential plant IPMDHs possess putative chloroplast targeting peptides based on bioinformatic analysis using ChloroP and TargetP ( http ://www.cbs.dtu.dk/services/ ). Consistent with the prediction, IPMDH1 was actually identified in chloroplasts using proteomics (Zybailov et al., 2008). To test the subcellular localization of IPMDH1 immunocytochemical labeli ng of leaf sections with anti IP MDH1 antibody was conducted. Positive IPMDH1 signal was obser ved in chloroplasts with anti IPMDH1 antibody followed by a goldlabeled secondary antibody (Figure 28A, C). The signal was absent in other cellular compartments and in chloroplasts treated with the secondary antibody alone (Figure 28B, D). To consolidate this localization, intact chloroplasts were isolated and fractionated into stroma and thylakoid/envelope membrane. Proteins from different fractions were subjected t o Western analysis using ant i IPMDH1 antibody, an antibody against a thylakoid luminal extrinsic subunit of PSII PsbO (Haussuhl et al., 2001), and an antibody against Rubsico large subunit RbcL (Weigel et al., 2003), respectively. As shown in Figure 28E, a weak signal of IPMDH1 was detected in the intact chloroplasts and a very strong signal was found in the stroma fraction. No signal was observed in the thylakoid/envelope fraction. This pattern is similar to that of RbcL and different from PsbO In addition, IPMDH activity in differ ent fractions was measured using 3isopropylmalate. In stroma, the activity was nearly ten times as high as in intact chloroplasts. No detectable activity was obtained in
49 thylakoid/envelope (Figure 28F). These results demonstrate that IPMDH1 is localized in chloroplast stroma, where the methionine chainelongation enzymes such as MAM (Falk et al., 2004; Textor et al., 2007) and BCAT3 (Knill et al., 2008) were found. IPMDH1 Is Regulated by ThioredoxinMediated Redox Regulation In Chlamydomonas reinhardtii a unicellular photosynthetic eukaryote, IPMDH1 was isolated by thioredoxin affinity chromatography (Lemaire et al., 2004). In addition, our kinetics data showed that oxidation of IPMDH1 reduced catalytic efficiency (Table 23). These findings led to the hypothes is that IPMDH1 may be subject to thioredoxinmediated redox regulation. To test the hypothesis, IPMDH1 was heterologously expressed in E. coli and purified to homogeneity (Figure 22). Addition of 5 mM dithiothreitol (DTT) led to reduction of the pr otein, while incubation with an oxidant (50 M CuCl2) (Ikegami et al., 2007; Balsera et al., 2009) caused oxidation, which could be reversed by DTT (Figure 29A). The activities of the enzyme were followed. Reduced IPMDH1 was more active and CuCl2 oxidatio n led to a significant decrease of IPMDH1 activity (Table 23, Figure 29A). After incubating the oxidized enzyme with DTT, the enzyme activity could be restored (Figure 2 9A). These data showed that IPMDH1 can be reversibly activated and inactivated by the redox state. Because IPMDH1 contains two cysteines which are conserved in higher plants and C. reinhardtii (Figure 21) and is localized to chloroplasts, we then investigated whether the redox regulation is mediat ed by the thioredoxin system. IPMDH1 was firstly oxidized by CuCl2, and then incubated with reduced bacterial thioredoxin, Arabidopsis chloroplast thioredoxin f and m, respectively. All three types of thioredoxin increas ed the activities of oxidized IPMDH1, with thioredoxin m being the most effec tive (Figure 29A). As the concentration of reduced thioredoxin m increased, the oxidized enzyme was
50 gradually shifted to the reduced form, accompanied by the increase of enzyme activity (Figure 29B). Taken together, these results clearly showed that IPMD H1 was a target of thioredoxin. Thiol Based Switch Is Essential for Regulating IPMDH1 Activity To examine the potential role of the two c onserved cysteine residues in IPMDH1, each cysteine residue was mutated to a serine residue to yield the C232S and C390S mutants, respectively. A double mutant C232S/C390S was also generated in which both cysteines are replaced with a serine. Under native conditions, all the mutants showed similar activities as the native protein. Under oxidized conditions, none of the mutants displayed a significant decrease in activities, while native enzyme activity decreased nearly twofold (Figure 210A). These data showed that mutation of either cysteine or both rendered the enzyme insensitive to oxidation, indicating that both cysteines are esse ntial for redox regulation of IPMDH1. When the purified IPMDH1 was analyzed under reducing SDS gel conditions (Figure 22), the oxidized form observed in Figure 29A was not observed. When the cysteine mutants and the fully oxidized IPMDH1 we re separated on reducing SDS gels, all the proteins migrated to the reduced protein position (data not shown). On the nonreducing SDS gels, only oxidized IPMDH1 protein migrated to the oxidized protein position (Figure 29). Nearly all C232S, C390S and C232S/C390S (reduced or oxidized) proteins migrated to the reduced position (data not shown). This experiment suggests that when disulfide bond was formed, the shape of protein changed significantly so that it exhibited differential migration in nonreducing SDS gels. To confirm the redox stat e of the sulfhydryl groups of IPMDH1 and the mutants under reducing and oxidized conditions, free sulfhydryl groups were labeled with a
51 specific fluorescent tag monobromobimane (mBBr) and visualized under UV excitation. As shown in Figure 210B, no mBBr signal w as present in oxidized native IPMDH1 and the double mutant C232S/C390S, indicating complete oxidation and loss of cysteine sulfhydryl groups, respectively. In addition, the free sulfhydryl groups of C232S and C390S wer e remarkably less than native IPMDH1, and they became even less after oxidation. Furthermor e, alkylation of the original IPMDH1 protein with iodoacetamide turned the enzyme insensitive toward reducing and oxidation reagents (data not shown). Moreover, accurate molecular weight analysis (Jez et al., 2004) showed that reduced IPMDH1 was two mas s units heavier than oxidized IPMDH1 (Figure 211). Collectively, these results support a mechanism that the two cysteines (Cys232 and Cys390) serve as a redox swi tch to regulate IPMDH1 activity. To investigate redox regulation of IPMDH1 activity in vivo stroma fraction of Arabidopsis chloroplasts was freshly prepared, oxidized with CuCl2 and reduced with thioredoxin, respectively. Enzyme activity assay and Western analysis showed that the increase of activity was associated with more reduced enzyme form and vice visa (Figure 212A, B). In addition to the treatment of the stroma, Arabidopsis seedlings were grown in liquid culture medium and then treated with water, hydrogen peroxide (5 mM and 10 mM for 2 h) and heavy metal (CdCl2) (50 M for 2 h and 4 h), which induces changes in cellular redox state (Hicks et al., 2007). Plant extracts were prepared, separated by nonreducing gel and examined by Western blotting. Treatment of seedlings with CdCl2 shifted the distribution of IPMDH1 to the oxidized form compared with control treatments. Treatment with hydrogen peroxide led to similar c hanges in the distribution of IPMDH1 in favor of the oxidized form (Figure 212C).
52 D iscussion The Expression and Subcellular Localization of IPMDH1 Highlight Its Potential Functional Significance Genes that function in the same or related metabolic pathways are often under the control of the same regulatory system, and genes involved in glucosinolate biosynthesis are coexpressed (Hirai et al., 2004, 2005, 2007; MaruyamaNakashita et al., 2006; Yan and Chen 2007). With the available Arabidopsis microarray data sets, we conducted coexpression analysis to identify candidate genes involved in the uncharacterized methionine chainelongation steps of aliphatic glucosinolate biosynthesis. In this work, we have identified the IPMDH1 gene as displaying a strong coexpression with nearly all known glucosinolate biosynthetic genes. Its high homology to known isopropylmalate dehydrogenases further suggests potential functions in methionine chainelongation and/or leucine biosynthesis (Table 21). In accordance with the above bioinformatic analysis, the spatial and temporal expression pattern of IPMDH1 d isplayed high similarity with the typical pattern observed for genes involved in the formation of aliphatic glucosinolates (Chen et al., 2003; Mikkelsen et al., 2003; Schuster et al., 2006). In addition, wounding and MeJA induction of IPMDH1 is consistent with the response of many genes involved in glucosinolate biosynthesis (Figure 27) (Mikkelsen et al., 2003; Schuster et al., 2006; Alvarez et al., 2008). Furthermore, localization of IPMDH1 in the chloroplast stroma (Figure 28) is another line of evidenc e supporting the function of IPMDH1 in methionine chainelongation and leucine biosynthesis. It is known that branchedchain amino acids, such as leucine are synthesized in chloroplasts (Binder et al., 2007; Knill et al., 2008). It is also known that BCAT2 3 and 5 are located to chloroplasts (Schuster et al., 2006;
53 Binder et al., 2007). MAM activity (Falk et al., 2004; Binder et al., 2007) and protein (Textor et al., 2007) were only found in chloroplasts. Based on the N terminal sequences, all the potential isopropylmalate isomerases and dehydrogenases are predicted to target to chloroplasts (data not shown). It is therefore reasonable to predict that the enzyme machinery for leucine biosynthesis and methionine chainelongation is mainly localized in chloroplasts. The Dual Functions of IPMDH1 in Leucine Biosynthesis and the Methionine ChainElongation Cycle of Aliphatic Glucosinolate Biosynthesis This study presented several lines of evidence supporting the dual function of IPMDH1 in leucine biosynthesis and in methionine chain elongation of aliphatic glucosinolate bi osynthesis. First, recombinant IPMDH1 showed a clear substrate preference for 3isopropylmalate, a known intermediate in leucine biosynthesis. IPMDH1 displayed similar biochemical characteristic s as those characterized IPMDHs in bacteria and y east, including high affinity for 3 isopropylmal ate, high activity under reducing conditions, using NAD+ as cosubstrate, and the same order of magnitude of Km and optimal pH values (Hsu and Kohlhaw 1980; Wal lon et al., 1997; Kaw aguchi et al., 2000). Second, IPMDH1 was able to complement E. coli and yeast Leu2 mutants lacking leucine autotrophy (data not shown; Nozawa et al., 2005), confirming the functional capability of IPMDH1 in leucine biosynthesis. Third, a knockout mutant of IPMDH1 showed significant reduction of leucine levels in seeds. Fourth, the regulation of IPMDH1 by redox and thioredoxin is consistent with the findings of potential redox regulation of IPMDHs in other organisms (Hsu and Kohlhaw 1980; Lemaire et al., 2004). Although leucine is one of the essential amino acids for humans and animals and has to be obtained primarily from plants, it is surprising that genes involved in plant leucine
54 biosynthesis have been barely characterized before. Our results have provided strong evidence that IPMDH1 is a functional isopropylmalate dehydrogenase catalyzing the oxidative decarboxylation step of plant leucine biosynthesis. In addition to tissuespecific expression and chlor oplast localization results, strong evidence for the role of IPMDH1 in methionine chainelongation cycle came from the reverse genetics experiments. Homozygous IPMDH1 mutant showed remarkable glucosinolate changes, i.e., reduction of the contents of C4 and longer chain aliphatic glucos inolates and increase of C3 glucosinolates and indole glucosinolates (Figure 25). These changes can be complemented by reintroduction of the functional IPMDH1 gene, confirming the direct involvement of IPMDH1 As the manuscript related to the study in this chaper was under final revision, an accepted paper appeared on line, in which several methionine derived metabolites were found accumulated in the IPMDH1 mutant (Sawada et al., 2009), corroborating the function of IPMDH1 in methionine chainelongation. The Arabidopsis genome contains two more potential IPMDH genes, IPMDH2 (At1g31180) and IPMDH3 (At1g80560). No significant changes of glucosinolate profiles were observed in homozygous IPMDH2 mutant and IPMDH3 mutant (Chapter 3) These results indicate that IPMDH1 is the major player in catalyzing the oxidative decarboxylation step of the methionine chainelongation cycle of aliphatic glucosinolate biosynthesis. IPMDH2 and/or IPMDH3 might play a minor role, which can not be fully excluded because in IPMDH1 n ull mutants low amounts of C4 (in seeds and leaves), C5 (in leaves) and C8 glucosinolates (in seeds) could still be observed. Our future studies using double and triple mutants will help resolve the individual functions of IPMDHs in glucosinolate biosynthesis.
55 It is interesting to note that there was a commensurate accumulation of C3 glucosinolates as well as a marked increase in indole glucosinolates in the IPMDH1 mutants. The increase of C3 glucosinolates is an interesting phenomenon, which was observed in mutants of genes involved in the early steps of methionine chain elongation (Kroymann et al., 2001; Textor et al., 2007; Sawada et al., 2009). It implies existence of another set of enzymes capable of performing the initial C2 to C3 chainelongation step. The enzymes and underlying mechanisms await further investigation. The negative correlation between aliphatic glucosinolates and indole glucosinolates has been observed several times, indicating that there might be crosstalk between the biosynthesis of aliphatic and indole glucosinolates (Reintanz et al., 2001; Chen et al., 2003; Textor et al., 2007; Gigolashvili et al., 2009). Our results have provided strong evidence that leucine biosynthesis and aliphatic glucosinolate biosynthesis share an evolutionary origin. The possible crosstalk between indole glucosinolate and aliphatic glucosinolate pathways is intriguing. Thiol Based Redox Regulation of Glucosinolate Biosynthesis and Leucine Biosynthesis Glucosinolate metabolism constitutes an indis pensable part of Brassica plants ability to defend against pathogens and pests (Grubb and Abel 2006; Halkier and Gershenzon 2006; Yan and Chen 2007). Since the core pathway for glucosinolate biosynthesis has been elucidated in Arabidopsis, the molecular regulation of glucosinolate metabolism has become an interesting and pressing question. Here we conducted a detailed exami nation of redox regulation of IPMDH1 using a combination of biochemical assays, sitedirected mutagenesis and mass spectrometry. As demonstrated i n the previous sections, IPMDH1 catalyzed the third step, i.e., oxidative
56 decarboxylation of the leucine biosynthetic pathway and the methionine chainelongation cycle of aliphati c glucosinolate biosynthesis. IPMDH1 contains two cysteines that are conserve d in photosynthetic organisms (Figure 21). Considering oxidative stress commonly associated with plant defense pr ocesses, we hypothesized that IPMDH1 was subjected to thiol based redox regulation that might directly regulate glucosinolate metabolism (Yan and Chen 2007). Multiple lines of evidence support the above hypothesis. As shown in Figures 29 to 212, the activity and conformation of IPMDH1 was highly sensit ive to redox changes. Reduced IPMDH1 was more active than the oxidized protein. Addition of DTT or reduced thioredox in was necessary for oxidized IPMDH1 to recover its a ctivity, thus confirming that IPMDH1 is regulated by redox environment and is a direct substrate of thioredoxin. Mutagenesis of Cys232 and Cys390 demonstrates these sulfhydryls ar e important for regulating activity in a thioredoxinmediated process. Mutation of either cysteine or both residues abrogated the redox responsiveness of the protein (Figure 210). Our results suggest that Cys232 and Cys390 may form a regulatory disulfide bond and, as shown in Figure 210, mediate a conformational change between oxidized and reduced forms of the protein. Further studies that compare the threedimensional structure of the plant IPMDH, which are redox regulated, with the bacterial homologs, w hich are not redox regulated, will provide insight on this regulatory mechanism. The localization of IPMDH1 in chloroplast stroma is a key factor for the redox regulation that is physiologically meaningful. Chloroplasts contain different isoforms of thioredoxin and many chloroplast enzymes are regulated by redox (Dai et al., 2004; Buchanan and Balmer 2005; Buchanan and Luan 2005; Hicks et al., 2007). For
57 example, Calvin cycle enzymes, which are localized in stroma, are regulated by redox and their activities are enhanced in the reduced state (Lemaire et al., 2007). In a similar way, IPMDH1 can sense cellular redox changes through thioredoxin. Interestingly, evidence has accumulated indicating that the other enzymes in the methionine chainelongation cycle may be subject to redox regulation (Textor et al., 2004; Alkhalfioui et al., 2007). Recently, a mutant with reduced glutathione levels exhibited decreased glucosinolate levels (Schlaeppi et al., 2008). Given the fact that many enzymes in glucosinolate biosyn thesis contain at least two cysteines and are predicted to form disulfide bonds, this research has opened a door for studying redox regulation of glucosinolate metabolism. The thiol based regulation of IPMDH1 characterized in this study provides an important post translational switch that regulates leucine and glucosinolate homeostasis in Arabidopsis.
58 Table 21. Primers used in this study Primer Sequence Temperature LBa1 TGGTTC ACG TAG TGG GCC ATC G 55 C LBb1 GCG TGG ACC GCT TGC TGC AAC T IMD1 LP TAA CTC TCA CGT GGT GTG GTG 50 C IMD1 -RP CAG AGC CAT CTC AGG TCT CAG IMD1 1L CGG ACG GAA CCA TAA CCG AAC C 54 C IMD1 1R GAG TTT GGT ACT TAG ATA GGT GTG CAC G IMD1 2L GAA CAT CAG TCT GAA TGC GAT CAA 52 C IMD1 2R GGT ACA ATG TTG CAT GGA TTA GCA I MD1 3L ATG GCG GCG TTT TTG CAA AC 55 C IMD1 3R TGT TGC ATG GAT TAG CAT ATA AGT GAC C IMD1 -4L CGG ACG GAA CCA TAA CCG AAC C 54 C IMD1 4R GAG TTT GGT ACT TAG ATA GGT GTG CAC G IMD1 5L GAA TTC GGG AAA AAA CGG TAT AAC AT 53 C IMD1 5R GTC GAC TTA AA C AGT AGC TGG AAC TT T G IMD1 -6L GGC GTG GCA AAC TTT CCT CTG TTG ACA AAG CCA ATG TGT TGG 55 C IMD1 -6R CCA ACA CAT TGG CTT TGT CAA CAG AGG AAA GTT TGC CAC GCC IMD1 -7L CCT GGA AAT AAA CTG GTG GGA TCC AAG GAA ATG GGT GAG GAG G 55 C IMD1 -7R CCT CCT CA C CCA TTT CCT TGG AT C CCA CCA GTT TAT TTC CAG G Actin1 F TGG AAC TGG AATGGT T AA GGC TGG 57 C Actin1-R TCT CCA GAG TCG AGC ACA ATA CCG TRXF1 1L GAA TTC GTT GGT CAG GTG ACG GA 55 C TRXF1 1R GTC GAC TCA TCC GGA AGC AGC AG TRXM1 1L GGA TCC GAC ACT GC T ACA GGA ATT 55 C TRXM1 1R GAG CTC TTA CAA GAA TTT GTT GAT GC
59 Table 22. List of 30 topranked Arabidopsis genes that coexpress with IPMDH1 Gene locus Annotation R value AT2G43100 a Small subunit of isopropylmalate isomerase 0.90 AT3G 58990 a Small subunit of isopropylmalate isomerase 0.89 AT4G13430 a Large subunit of isopropylmalate isomerase 0.87 AT1G78370 a Glutathione S transferase (ATGSTU20) 0.84 AT3G03190 a Glutathione S transferase (ATGSTU11) 0.83 AT1G74090 b Sulfotransferas e family protein (AtST5b) 0.82 AT4G12030 b Bile acid:sodium symporter family protein 0.81 AT1G16410 b Cytochrome P450 (CYP79F1) 0.80 AT1G16400 b Cytochrome P450 (CYP79F2) 0.80 AT1G31230 Aspartate kinase 0.79 AT4G13770 b Cytochrome P450 (CYP83A1) 0.78 AT3G49680 b Branched chain amino acid aminotransferase (BCAT3) 0.78 AT3G19710 b Branched-chain amino acid aminotransferase (BCAT4) 0.78 AT2G20610 b Alkylthiohydroximate C S lyase 0.77 AT2G31790 a UDP glucosyl transferase 0.77 AT5G61420 b Myb transcrip tion factor (MYB28) 0.75 AT3G02020 Aspartate kinase 0.74 AT1G65860 b Flavin containing monooxygenase (FMO) 0.74 AT1G73600 Phosphoethanolamine N methyltransferase 3 (NMT3) 0.72 AT1G18590 b Sulfotransferase (AtST5c) 0.71 AT5G23010 b 2 isopropylmalate sy nthase (MAM1) 0.69 AT2G46650 Cytochrome b5 0.67 AT1G21440 Mutase family protein 0.65 AT1G62560 b Flavin containing monooxygenase (FMO) 0.65 AT4G14680 Sulfate adenylyltransferase (APS3) 0.65 AT1G48600 Phosphoethanolamine N methyltransferase (NMT2) 0.63 AT5G04590 Sulfite reductase 0.62 AT5G67150 Anthranilate N hydroxycinnamoyl/ benzoyltransferase 0.60 AT5G23020 b 2 isopropylmalate synthase (MAM3) 0.58 AT5G07690 b Myb family transcription factor (MYB29) 0.58 a genes potentially involved in glucosinol ate biosynthesis. b genes known to be involved in glucosinolate biosynthesis.
60 Table 23. Substrat e and kinetics of recombinant IPMDH1 (red, reduced form; ox, oxidized form) Substrate V max ( mol/minmg) SE K m ( M) SE IPMDH1 red IPMDH1 ox IPM DH1 red IPMDH1 ox 3 isopropylmalate 0.93 0.01 0.45 0.02 25.21 2.30 68.62 8.31 NAD + 1.01 0.01 0.82 0.05 187.11 11.79 242.10 22.50 D malate 0.16 0.03 0.11 0.03 1100 100 1400 200
61 Figure 21. Multiple sequence al ignme nt of 3isopropylmalate dehydrogenases from different organisms. The sequences were aligned using ClustalW and MultAlign ( http://bioinfo.genotoul.fr/multalin/ ). The black arrow indicates putative chloroplast signal peptide, and the red background indi cates conserved residues. The two conserved cysteine residues are highlighted with red stars. At, Arabidopsis thaliana; Bn, Brassica napus (rape); Ca, Capsicum annuum (pepper); Cr, Chlamydomonas reinhardtii (alga); Ec, Escherichia coli (bacteria); Os, Ory za sativa (rice); Sc, Saccharomyces cerevisiae (yeast); Tf, Thiobacillus Ferrooxidans (bacteria); Vv, Vitis vinifera (grape). cTP** cTP**
62 Figure 22. Purification and identification of recombinant IPMDH1. A) SDS gel showin g vector control and purified IPMDH1. B) Western blot detection of IPMDH1 us ing antibody produced against IPMDH1. C) Mass spectrometry identification of purified IPMDH1 shown in (A). Molecular mass Vector, crude Vector, purified IPMDH1, crude IPMDH1, purified Molecular mass Vector, crude Vector, purified IPMDH1, crude IPMDH1, purified
63 Figure 23. Tem perature and pH dependence of IPMDH1 activity. A) Temperature. B) pH.
64 Figure 24. Characterization of TDNA knockout mutant of IPMDH1 A) Schematic diagram of the genomic structure of IPMDH1 with T DNA insertion site indicated by an inverted triangle. Bars represent exons and lines represent introns. ATG, start codon; STOP, stop codon. B) RTPCR analysis of homozygous mutant plants with specific primers for IPMDH1 Actin1 was used as an equal loading control. IPMDH1 IPMDH1 Actin1Wild type ipmdh1
65 Figure 25. Glucosinolate prof iles of wild type, ipmdh1 mutant and ipmdh1 complementation plants. Aliphatic glucosinolates are grouped according to their chain length (the number of methylene carbons in the sidechain, C3C8), and indole glucosinolates were summed as one group. Data sh own are the means and standard deviations (SD) of at least three replicates. A) seeds; B) leaves. Wild type ipmdh1 35S::ipmdh1 Wild type ipmdh135S::ipmdh1
66 Figure 26. Free amino acid profiles of wild type and ipmdh1 mutant. Data shown are the means and standard deviations (SD) of at least three replicates. Insert shows the levels of amino acids with significant changes. Wild type ipmdh1
67 Figure 27. Spatial and temporal expression patterns of IPMDH1 analyzed using IPMDH1 promoter GUS plants. A) 3 day old seedling; B) 6 day old seedling; C) 3 week old leaf; D) flower; E) young silique; F) leaf cut with scissors (left), squeezed with forceps (middle), and pierced with a needle (right); G) control quant itative RT PCR analysis of IPDMH1 expression in different tissues of wild type (Col0) plants (top row). Actin1 was used as control for equal loading (bottom row); I) Semi quantitative RTPCR analysis of samples shown in F. Controls are leaves not treated with mechanical wounding (top row ). IPMDH1 Actin1
68 Figure 28. Subcellular localization of IPMDH1 and in vivo activity analysis. A) and C) Immunocy tochemical labeling with anti IPMDH1 antibody and a goldlabeled secondary antibody. Label is ass ociated with chloroplast (cp, arrows) rather than the cytoplasm (c), vacuole (v) and cell wall (cw). B) and D) Immunocytochemical labeling with the goldlabeled secondary antibody only. E) Western blot of diff erent fractions of chloroplasts. F) In vivo IPM DH activity determined for the above fractions (n > 3, SD). IPMDH1 PsbO RbcL
69 Figure 29. Re dox regulation of recombinant IPMDH1. A) Effect of oxidation and reduction on IPMDH1 redox state (top) and activity (bottom). The positions of the reduced (re d) and oxidized (ox) forms of IPMDH1 are indicated by arrows. B) Effect of Trx m concentration on the reduction and activity of oxidized IPMDH1. The redox states of IPMDH1 were indicated by arrows (top) and the activities (bottom) were determined from at least three replicate experiments (mean SD).
70 Figur e 210. Redox sensitivity of IPMDH1 cysteine mutants. A) Activity of the reduced form (red, open bar) and oxidized form (ox, closed bar) of wild type, single a nd double cysteine mutants of IPMDH1. Data shown are the means and SDs of at least three replicates. B) Reduced and oxidized wild type and mutant IPMDH1 labeled with mBBr and separated by nonreducing SDS gel electrophoresis. Proteins labeled with mBBr were visualize d under UV illumination. IPMDH1 C232S C390S C232S/ C390S IPMDH1 C232S C390S C232S/ C390S
71 Figure 211. Accurate molecular weight analysis of reduced and oxidized IPMDH1. A) Electr ospray MS spectrum of reduced IPMDH1 (top) and deconvoluted spectrum indicating the molecular weight of 43250 dalton (bottom). B) Electro spray MS spectrum of oxidized IPMDH1 (top) and deconvoluted spectrum indicating the molecular weight of 43248 dalton (bottom).
72 Figure 212. In vivo redox sensitivity of IPMDH. A) IPMDH a ctivity in stroma fractions that were freshly isolated (untreated), treated with 50 M CuCl2 (oxidized) and 3 M Trx m (reduced), respectively. B) Redox state of IPMDH1 in stroma that ME, control), 50 M CuCl2 (oxidized) and 3 M Trx m (reduced), respectively. C) Effect of treatment of Arabidopsis seedlings with 50 M CdCl2 for 2 h and 4h, and H2O2 5 mM and 10 mM for 2 h, respectively, on the in vivo redox state of IPMDH1.
73 CHAPTER 3 FUNCTIONAL CHARACTERIZATION OF I S O PROPYLMALATE DEHYDRO GENASES REVEALS THEIR IMPORT ANT ROLES IN GAMETOPHYTE DEVELOPMENT Introduction Valine, isoleucine and leucine, which have branched aliphatic side chains, can not be synthesized by animals. So they are essential amino acids (Singh 1999; Binder et al., 2007). In plants, valine and isoleucine are synthesized in two parallel pathways using a set of ident ical enzymes, resulting in the generation of either 2oxo 3 methylvalerate or 2oxoisovalerate, which is transaminated to isoleucine or valine, respectively. The valine precursor, 2oxoisovalerate, can also serve as substrate for the biosynthesis of leucine via an additional four step conversion process. Briefly, the intermediate 2oxoisovelarate i s condensed with acetylisopropylmalate, isopropylmalate through an isomerization ketoisocaproate, which subsequently undergoes reductive amination to produce leucine (Binder et al., 2007). In Arabidopsis, the enzymes involved in leucine biosynthesis are related to or participate in the methionine chain elongation process during the biosynthesis of aliphatic glucosinolates (Field et al., 2004; Schust er et al., 2006; Binder et al., 2007; Textor et al., 2007; He et al., 2009). Isopropylmalate synthase is encoded by four genes in Arabidopsis, two of which ( IPMS1 and IPMS2 ) are functional in leucine biosynthesis and the other two genes encode methylthioal kylmalate (MAM) synthases ( MAM1 and MAM3 ) involved in methionine chainelongation pathway (Field et al., 2004; Textor et al., 2004; de Kraker et al., 2007; Textor et al., 2007). In addition, as in bacteria but not in yeast, Arabidopsis isopropylmalate isom erases (IPMI) appear to function as heterodimers, consist of a large subunit encoded by a single gene and a small subunit
74 encoded by one of the three genes (Binder et al., 2007; Knill et al., 2009). Mutation of the large subunit caused intermediates in bot h leucine biosynthesis and methionine chainelongation pathways to accumulate, demonstrating that IPMI large subunit functions in both leucine and glucosinolate biosynthesis (Knill et al., 2009). In contrast, the small subunits seem to be specialized to ei ther leucine biosynthesis or methionine chain elongation, although the precise function of each gene is unknown (Knill et al., 2009; Sawada et al., 2009). Furthermore, among the six branch ed chain aminotransferases (BCATs) in Arabidopsis, BCAT4 was demonst rated to be specifically involved in glucosinolate biosynthesis whereas BCAT3 functions in both amino acid and glucosinolate biosynthesis (Schuster et al., 2006; Knill et al., 2008). Recently, we have shown that an isopropylmalate dehydrogenase gene IPMDH1 (At5g14200) is active in both leucine biosynthesis and methionine chainelongation of aliphatic glucosinolate biosynthesis (He et al., 2009). The IPMDH1 knockout mutant exhibited wildtype levels of free leucine in leaves, but had reduced levels in seeds. Other IPMDH proteins may be responsible for the remaining enzyme activities. Based on bioinformatic analyses, IPMDH2 (At1g31180) and IPMDH3 (At1g80560) appear to be the most likely candidates to carry out this function (Binder et al., 2007; He et al., 2009). In this chaper we describe detailed functional characterization of IPMDH2 and IPMDH3 using multiple approaches. Substrate specificity analysis of recombinant IPMDHs, phenotypic analysis of IPMDH mutants, in vivo enzyme activity analysis, and amino acid and glucosinolate profiling led us to conclude that IPMDH2 and IPMDH3 are functionally redundant in leucine metabolism, and their involvement in aliphatic glucosinolate biosynthesis is dominated by IPMDH1 In addition, the distorted
75 segregation ratios in male gametes and female gametes, and the developmental defects of IPMDH2 and IPMDH3 double mutant demonstrate that IPMDH2 and IPMDH3 are essential for development of functional gametophytes. These defects in pollen and embryo sac development suggest that active leucine biosynthesis is essential for gametophyte formation, which is an indispensable process in plant life cycle. Materials and Methods Plant Materials and Growth Conditions Seeds of Arabidopsis thaliana ecotype Columbia (Col 0) (CS3879), qrt mut ant (CS25041) and Salk mutant lines ipmdh1 (Salk_063423), ipmdh2 (Salk_152647) and ipmdh 3 (Salk_013237) were obtained from the Arabidopsis Biological Resource Center (ABRC, Ohio State University, Columbus, OH). Double or triple mutants were created by cros sing as depicted by Figure 3 1 For all crosses, immature flower buds were emasculated and manually cross pollinated with pollen from a parental flower. Seed germination and plant growth conditions were the same as previously described (He et al., 2009). DNA Extraction and Genotyping Genomic DNA was extracted from young leaves and ground with sand in 200 mM Tris HCl, pH 7.5, 250 mM KCl, 10 mM EDTA and 0.5% SDS (w/v). After centrifugation at 14,000 rpm, 4oC for 5 min, the supernatant was transferred into a new microtube, precipitated with 1 volume of 2propanol, rinsed with cold 70% ethanol and subsequently resuspended in TE buffer (10 mM Tris HCl, pH 8.0, 1 mM EDTA. PCR USA) 1.5 mM MgCl2 of Taq DNA polymerase (Invitrogen, USA). The PCR program was as follows: 30 s at 94
76 C, 30 s at 50 C 55 C (Table 31), and 1 min/kb at 72 C for 37 cycles. All PCR primers used in this study are listed in Table 31 Primers used for identifying SALK TDNA insertion mutants were designed by the SIGnAL TDNA Express Arabidopsis Gene Mapping Tool ( http://signal.salk.edu/tdnaprimers.2.html ) (Alonso et al., 2003). ipmdh1 mutant was isolated using left gene primers IPMDH1LP, IPMDH1 RP and TDNA left border primer LBb1. ipmdh2 was isolated using gene primers IPMDH2LP, IPMDH2 RP and LBb1. ipmdh3 was isolated using gene primers IPMDH3LP, IPMDH3 RP and LBb1. Double or triple mutants were identified by two or three relevant pairs of the above primers. Genetic Analysis To determine gametophytic transmission of ipmdh2 and ipmdh3 alleles, reciprocal crosses were performed between wildtype and mutant lines (Table 33 ). Approximately 20 independent reciprocal crosses were done. Seeds were harvested and planted. Their progenies were genotyped for the presence of IPMDH2 and/or IPMDH3 T DNA insertions by PCR. Tetrad analysis was performed by generating two triple mutants. The pollens from qrt/qrt mutant were used to pollinate emasculated flowers of the IPMDH2/ipmdh2 ipmdh3/ipmdh3 plant. We selected the progenies on kanamycin plates for IPMDH2/ipmdh2 IPMDH3/ipmdh3 QRT/qrt and then allow positive plants to self pollinate. Seeds from the selfing were grown on kanamycin plates, and plants that showed qrt phenotypes ( IPMDH2/ipmdh2 ipmdh3/ipmdh3 qrt/qrt and ipmdh2/ipmdh2 IPMDH3/ipmdh3 qrt/qrt) were selected.
77 Complementation of IPMDH mutants A 3245bp IPMDH2 genomic fragment was amplified using a hi gh fidelity Plantinum Taq DNA polymerase (Invitrogen, USA) with the primer pair IPMDH24L/IPMDH2 4R (Table 31). The PCR products were cloned into a pGEM T easy vector (Promega, USA) and sequenced to confirm fidelity. The correct sequence was cut out from the vector with Xba I and Pml I and inserted into the Xba I and Pml I sites of the pCAMBIA1305 vector (http://www.cambia.org). For ipmdh3 complementation, a full length genomic copy of IPMDH3 was excised from a BAC clone (F28K20) obtained from ABRC using Spe I and Pvu II and inserted into the Xba I and Pml I sites of the pCAMBIA1305 vector. Subsequently, the resulting vectors containing genomic IPMDH2 or IPMDH3 were introduced into Agrobacterium tumefaciens strain GV3101 and transformed into ipmdh2/ipmdh2 IPMDH3/ip mdh3 or IPMDH2/ipmdh2 ipmdh3/ipmdh3 plants, respectively, through floral dipping (Clough and Bent, 1998). Transgenic plants Recombinant Protein Expression and Purification The cDNAs of IPMDH1 IPMDH2 and IPMDH3 without the N terminal plastidtargeti ng peptide were cloned into a pET28a expression vector (Novagen, USA) using the primer pairs IPMDH1 2L/IPMDH1 2R, IPMDH2 2L/IPMDH2 2R and IPMDH32L/IPMDH3 2R, respectively. The constructs were expressed in E. coli strain BL21(DE3) grown at 37C in Terrific Broth medium to an OD600 of 0.6 and then induced with 1 mM IPTG at 20C for 15 h. Soluble protein was purified using a Midi PrepEase kit according to the manufacturers instructions (EMD, USA). The purified proteins were then dialyzed against 20 mM Tris H Cl (pH 8.0) at 4C overnight. Proteins were concentrated by ultrafiltration with a 10kDa cutoff membrane (Amicon, USA) at
78 4C. Enzyme purity was examined using 12% denaturing gel electrophoresis and protein concentrations were determined using the Bradfor d method (Bradford, 1976) with BSA as standard. The enzyme activities were determined using a standard assay (He et al., 2009). Light Microscopy and Phenotype Analysis Unless otherwise stated, all the microscopy analyses were performed by using a Leica C TR6000 microscope equipped for differential interference contrast (DIC) and fluorescence sources. Images were captured using a Retiga EXi Digital CCD Camera and analyzed using an Openlab software package ( http://www.cellularimaging.com/products/openlab /). Alexander staining for pollen viability was assessed by incubating fresh pollen grains in Alexanders solution for 10 min at room temperature, followed by observation under bright field condition (Alexander, 1969; JohnsonBrousseau and McCormick, 2004). The pollen nuclear constitution was analyzed by staining the pollens with 1 mg/ml 46 diamidino 2 phenylindole (DAPI) as previously described (Park et al., 1998). Then the pollen grains were visualized under UV light. To examine the terminal phenotype of mut ant ovules, the pistils at floral stage 12c (Smyth et al., 1990) were emasculated and the ovule phenotype was analyzed 12h, 24h, and 48 h after emasculation. For ovule clearing, pistils were dissected longitudinally, fixed overnight at room temperature in an FAA solution (50% ethanol (v/v), 10% formalin (v/v), 5% glacial acetic acid (v/v) and 35% water (v/v)), and then dehydrated by 70% ethanol for 30 min before clearing in chloral hydrate: glycerol: water (8 g: 2 mL: 1 mL) for 16 h. The DIC optics was use d for observation. In vitro pollen tube germination was performed at 22C for 4 6 h on solid medium as previously described (Boavida and McCormick, 2007).
79 Assays were repeated at least five times and the germination rate was estimated. Microscopy imaging w as conducted using the bright field optics. To analyze seed setting, siliques at 8 to 10 day after fertilization were placed on doublesided tape and longitudinally dissected under a dissecting microscope, and then undeveloped and developed ovules were co unted. Images of the dissected siliques were captured using a Nikon SMZ800 stereoscope. Transmission and Scanning Electron Microscopy Pollen sac samples from wildtype and mutant Arabidopsis flowers at stage 10~13 (Smyth et al., 1990) were dissected out and fixed in 4% paraformadehyde, 2.5% glutaraldehyde in 0.1M sodium cacodylate buffer, pH 7.24. For efficient fixation, dehydration and resin embedding, pollen sac samples were slit open with a scalpel in the fixative. Fixed tissues were processed with the aid of a Pelco BioWave laboratory microwave (Ted Pella, CA, USA). The samples were washed in 0.1M sodium cacodylate pH 7.24, post fixed with 2% OsO4, water washed and dehydrated in a graded ethanol series 25%, 50%, 75%, 95%, 100% followed by 100% acetone. D ehydrated samples were infiltrated in graded acetone/Spurrs epoxy resin (Ellis, 2006) 30%, 50%, 70%, 100% and cured at 60oC. Cured resin blocks were trimmed, thin sectioned and collected on formvar copper slot grids, post stained with 2% aqueous Uranyl ace tate and Reynolds lead citrate as described in Kang et al. (2003). Sections were examined with a Hitachi H 7000 TEM (Hitachi High Technologies America Inc., Schaumburg, IL) and digital images acquired with a Veleta 2k x 2k camera and iTEM software (Olympu s Soft Imaging Solutions Corp, Lakewood, CO, USA). For scanning electron microscopy, pollen grain samples were fixed and dehydrated as mentioned above in a graded ethanol series of 25%, 50%, 75%, 95%
80 and 100%, followed by critical point drying (Bal Tec CPD 030, Leica Microsystems, IL, USA). Samples were mounted on carbon adhesive tabs on aluminum specimen mount, Au/Pd sputter coated (DeskII, Denton Vacuum, Moorestown, NJ, USA) and high resolution digital micrographs acquired with a fieldemission scanning electron microscope (S 4000, Hitachi High Technologies America Inc., Schaumburg, IL USA). Gene Expression Analysis RNA extraction and cDNA synthesis were conducted as previously described (He et al., 2009). Regular RTPCR was conducted as the genomic PCR described before, except the following genespecific primers were used: IPMDH11L/IPMDH1 1R for IPMDH1 IPMDH2 1L/IPMDH2 1R for IPMDH2 and IPMDH3 1L/IPMDH3 1R for IPMDH3 ACTIN1 was used as control (Table 31). Quantitative real time PCR was performed on a MyIQ singlecolor detection system (BioRad, USA) using iQ SYBR Green Supermix from BioRad with the following thermal cycling program: 95C for 4 min, followed by 40 cycles of 95C for 10 s, 55C for 30 s. Amplification was followed by a melt curve determ ination as follows: 180 cycles, each cycle persisting for 10 s and the first cycle at 60C, with an increase of 0.2C each cycle after cycle 2. Primer pairs specific for each gene were IPMDH13L/IPMDH1 3R for IPMDH1 IPMDH2 3L/IPMDH2 3R for IPMDH2 and IPMD H3 3L/IPMDH3 3R for IPMDH3 (Table 31). The specificity of each primer pair was tested by sequencing the PCR product and by the above melting curve analysis. Expression levels were quantified relative to ACTIN2/8 expression (Actin2/8L/Actin2/8R) and aver aged from three independent experiments. For promoter GUS experiment, the IPMDH3 promoter region (from 252 to +33bp) was amplified from genomic DNA using primers IPMDH34L/IPMDH3 4R, inserted into the pGEM T easy vector and sequence verified by sequencing. The promoter region
81 was subsequently subcloned into the pCAMBIA1305 vector using Nco I and Sal I sites. For constructing IPMDH2 promoter GUS, GUS sequence was firstly amplified from pCAMBIA1305 plasmid using the primer pair GUS1L/GUS 1R ( Table 3 1) and subsequently cloned into the pCAMBIA1305 vector containing the full length IPMDH2 genomic sequence, which was used in the mutant complementation experiment. Plant transformation and selection were conducted as in the complementation experiment. Histochemica l GUS assays were performed on T2 transgenic plants as described previously (Jefferson et al., 1987). Chlorophyll, Glucosinolate and Amino Acid Analysis Chlorophyll was extracted from 4week old rosette leaves and analyzed using a spectrophotometer as prev iously described (Aronsson and Jarvis, 2002). Chlorophyll concentration was calculated using the formula: Chlorophyll (g/g fresh weight) = (17.76 x A645 + 7.34 x A663)/leaf fresh weight (Porra 2002). Seeds of 20 mg and rosette leaves of 100 mg were used f or glucosinolate and free amino acid analysis. Glucosinolates were analyzed using HPLC MS (Chen et al., 2003; Alvarez et al., 2008). Free amino acids were extracted using a methanol and chloroform method (Colebatch et al., 2004; Schauer et al., 2005) and profiled using an HPLC based precolumn derivatization protocol (Schuster et al., 2006). Chloroplast Fractionation and In Vivo IPMDH Activity Assay Chloroplast fraction and IPMDH activity assay were conducted as described in He et al. (2009).
82 Results IPMDH2 and IPMDH3 Are Key Enzymes in Leucine Biosynthesis The Arabidopsis genome has three IPMDH genes (He et al., 2009). To elucidate the biochemical characteristics of each IPMDH, cDNAs coding for the mature forms of all three IPMDHs without the putative trans it signal peptides were cloned into the bacterial expression vector pET28a with an N terminal 6XHis affinity tag, and transformed into BL21 (DE3) cells. As shown in Figure 3 1, all three proteins could be expressed i n bacteria and purified to homogeneity. The activity of each enzyme on 3 isopropylmalate (3IPM), the intermediate of leucine biosynthesis, was investigated using established enzyme assays (He et a l., 2009). As shown in Table 32 IPMDH2 and IPMDH3 exhibited significantly higher affinity and higher catalytic efficiency than IPMDH1, suggesting that IPMDH2 and IPMDH3 genes encode key enzymes catalyzing the oxidative decarboxylation of 3IPM in leucine biosynthesis. Double Mutant ipmdh2 ipmdh3 Is Lethal To identify A. thaliana lines that carry TDNA insertions in IPMDH loci, the Salk T DNA Insertion collections (Alonso et al., 2003) were screened. One line (Salk_063423, named ipmdh1) contains a TDNA insertion in the first exon of IPMDH1 Another line (Salk_152647, named ipmdh2 ) h arbors a TDNA inser tion in the fifth exon of IPMDH2 and a third line (Salk_013237, named ipmdh3 ) contains a TDNA in the eighth exon of IPMDH3 (Figure 32). Different genespecific primers were combined with the respective TDNA border primers for genotyping (Table 31). RT PCR amplification of each of the gene products using genespecific primers revealed that all of the mutants are null alleles. (Table31; Figure 33).
83 No discernible vegetative or reproductive phenotypes were observed for ipmdh1, ipmdh2 or ipmdh3 single m utants (data not shown), which indicates that loss of a single IPMDH gene function has no obvious impact on Arabidopsis growth and development. To address whether IPMDH2 and IPMDH3 were functionally redundant in leucine biosynthesis, we attempted to generate the ipmdh2 ipmdh3 double mutants. Reciprocal crosses of ipmdh2 and ipmdh3 mutants were done and the subsequent F1 pl ants self fertilized (Figure 33 ). The F2 seedlings from individual F1 plants were genotyped using PCR. IPMDH2 and IPMDH3 genes are locat ed at opposite ends of chromosome 1 and behave as unlinked loci. According to the law of Mendelian genetics, the expected frequency of double homozygous mutant plants in the F2 population should be 1:16. However, a double mutant was never found in the F2 population (n=378), indicating a segregation abnormality. To increase the probability of finding the double mutant, we screened the F3 progeny of self pollinated plants where TDNA insertions were homozygous at one locus and heterozygous at the other locus ( ipmdh2/ipmdh2 IPMDH3/ipmdh3 or IPMDH2/ipmdh2 ipmdh3/ipmdh3). Homozygous double mutants were not found in the populations, which is consistent with the ipmdh2 ipmdh3 double mutant being an embryo or gametophyte lethal mutant. Therefore, plants with ipmdh2/ ipmdh2 IPMDH3/ipmdh3 or IPMDH2/ipmdh2 ipmdh3/ipmdh3 genotype were used for further experiments. In contrast, homozygous ipmdh1 ipmdh2 and ipmdh1 ipmdh3 doubl e mutants were found (Figure 33 ). The ipmdh2 ipmdh3 Is Not Transmitted through the Male To determi ne the cause of the anomalous segregation, we assessed the transmission efficiencies (TE) of double mutant alleles through male and female gametes and compared them to the TE of each single mutant Assuming random
84 segregation during meiosis and the absence of post meiotic selection, the genotypes of the gametes should be equally distributed (Howden et al., 1998). The heterozygous mutant lines were reciprocally backcrossed with wild type and the number of plants with each possible genotype was scored. For single mutant allele, the TEs of ipmdh2 (male: 94.4% and female: 95.8%) were nearly normal, whereas the TE of ipmdh3 was slightly reduced in both the male (81.7%) and female (91.1%), suggesting that a small part of ipmdh3 gametes were aborted (Table 3 3 ). Wh en double heterozygous mutant plants were used as the pollen donors, however, the TE of ipmdh2 ipmdh3 gametes was zero, indicating that both mutant alleles could not be transmitted th rough male gametophytes (Table 33 ). This was further supported by recipr ocal cross between IPMDH2/ipmdh2 ipmdh3/ipmdh3 or ipmdh2/ipmdh2 IPMDH3/ipmdh3 and wild type, where no ipmdh2 ipmdh3 gametes were inherited from the male parent (Table 2). On the other hand, transmission of both mutant alleles could occur in female gametophytes, but it did so at a reduced frequency (Table 3 3 ). Our results demonstrated that IPMDH2 and IPMDH3 are genetically redundant in gametophyte development. Wild Type IPMDH2 and IPMDH3 Can Rescue the Distorted Transmission To unequivocally establish that the disruption of IPMDH2 and IPMDH3 caused nonfunctional male gametes, the full length genomic DNA of IPMDH2 including promoter and terminator regions was introduced into ipmdh2/ipmdh2 IPMDH3/ipmdh3 plants. In addition, the full length genomic copy of IPM DH3 was introduced into IPMDH2/ipmdh2 ipmdh3/ipmdh3 plants. The subsequent homozygous ipmdh2 ipmdh3 double mutant plants were identified in the offspri ng of transgenic plants via successive hygromycin resistant selection and PCR based genotyping (Table 34 ). In all cases, the transgene restored normal transmission frequencies of ipmdh2 ipmdh3 mutant alleles
85 ( Table 35 ). We therefore conclude that male gametophytes were lethal and female gametophytes exhibited reduced viability. ipmdh2 ipmdh3 Pollen Grains S how Abnormal Morphology Pollen grains from IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipmdh3 were analyzed for viability, morphology and nuclear anomalies. Half of the pollens grains should have an ipmdh2 ipmdh3 genotype. Pollen viability was ass essed by Alexander staining (Alexander, 1969). Viable pollen grains are stained red with a green outline, and aborted pollen grains are entirely green. As shown in Figure 34A, pollen grains from wildtype plants were plump, round and redstained with a gr een halo. However, pollen grains from IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipmdh3 plants were a mixture of sizes (Figure 34A). Approximately onehalf of pollen grains (54.1%, n=1566 and 48.1%, n=1327) were smaller than wildtype. Some of t he small pollen grains were shrunken, presumably due to cytoplasmic degeneration. Given that the percentage of pollen grains showing morphological alterations correlated well with the expected frequency of double mutant male gametophytes, we inferred that the genotype of these small pollen grains was ipmdh2 ipmdh3. Scanning electron microscopy (SEM) was used to examine wildtype and mutant pollen with only one functional IPMDH2 or IPMDH3 allele. Pollen grains from wild type anthers were uniform, but pollen grains from the IPMDH2/ipmdh2 ipmdh3/ipmdh3 or ipmdh2/ipmdh2 IPMDH3/ipmdh3 anthers were clearly a mixture of sizes (Figure 34B). Roughly half of the pollen appeared of normal size and half appeared small. Although the mutant pollen grains were smaller in size and appeared shrunken, the basic structure of the exine layer appeared normal (Figure 34B), showing that the heterozygous sporophyte tissue developed normally.
86 To determine whether the morphological change of mutant pollen was accompanied with a nuc lear abnormality, pollen mitosis was examined by 46diamidino 2 phenylindole (DAPI) staining and fluorescence microscopy. Overall, no differences between the nuclei in wildtype and the mutant pollen were found (Figure 34C). The vegetative nucleus and the two generative nuclei were distinguishable in the small pollen grains. The substantial size difference could only be found in pollen grains from IPMDH2/ipmdh2 ipmdh3/ipmdh3 or ipmdh2/ipmdh2 IPMDH3/ipmdh3 plants (Figure 34). ipmdh2 ipmdh3 Pollen Grains A re Defective in Pollen Germination While the ipmdh2 ipmdh3 pollen grains were noticeably smaller, they had not completely degenerated. To define the terminal abortion caused by the double mutation, in vitro germination rates of pollen from the mutants har boring only one functional copy of IPMDH2 or IPMDH3 were investigated and compared with wildtype pollen. As shown in Figure 35 pollen tubes were observed to emerge from full, round pollen grains, but not from the smaller, misshapen grains. The germinati on rate of wild type pollen was 82.4% (n=2314), and the germination rate of pollen in IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipmdh3 plants was reduced to 35.6% (n=1989) and 38.3% (n=2508), respectively (Figure 35 ). This is almost half of wil d type, which is consistent with a loss of pollen growth in ipmdh2 ipmdh3 double mutants. As a control, the germination rate of pollen from single mutant lines was assessed and compared to the wild type. Slight decreases in germination rates of ipmdh2 poll en (76%, n=702) and ipmdh3 pollen (72%, n=533) were found, but were not significantly different from wildtype, indicating that the male gametes from the single mutants were functional. These
87 data indicate that the germination defect of the double mutant pollen account for the male sterility. Ultrastructure of ipmdh2 ipmdh3 pollen Thin section transmission electron microscopy (TEM) analysis was done to observe the ultrastructure of male ipmdh2 ipmdh3 gametophytes. Prior to the first mitotic division, no abnormalities were found in any of the pollen grains (data not shown). After the first mitotic division, some of pollen grains from the ipmdh2/ipmdh2 IPMDH3/ipmdh3 and IPMDH2/ipmdh2 ipmdh3/ipmdh3 mutant plants produced large numbers of starch granules in the plastids (Figures 3 6A and 36 B). It has been shown that normal Arabidopsis pollen grain plastids accumulate starch granules of varying sizes at bicellular and tricellular stages (Kuang and Musgrave, 1996; Zhang et al., 2002). Calculating segregation ratios of the pollen grains based on starch deposition was not successful because two size classes could not easily be determined. Among the maturing pollen grains collected from dehiscing anthers, however, it was evident that approximately half of the pollen grains (29 out of 55 pollen grains from the two double mutant plants) deviate from the normal pollen grain development (Figure 36 C). The abnormal pollen grains were distinguished by their large elec tron dense aggregates (Figures 36 D F), u ndulated cell sur face (Figures 36D and 36 E), and lack of small round particles in the vegetative cytoplasm (Figures 36 F H). The darkly stained aggregates are the most striking defect in the abnormal pollen grains. Their heavy staining might be related to lipid bodies. M ost of the aggregates were spatially associa ted with ER cisternae (Figures 36E and 36F). In the pollen grains containing dark aggregates, the intine layer was irregular, often growing into the cytoplasm. In normal pollen grains, there appeared numerous c ytoplasmic particles with sizes and
88 shapes consistent with the previously described particles (Owen and Makaroff, 1995; Kua ng and Musgrave, 1996) (Figure 36 G). In contrast, the affected pollen grains were devoid of such particles (Figure 36 F). Plastids with several starch granules were found in abnormal pollen grains, but were not seen in the normal pollen grains (Figure 36 E). We did not see differences in pollen phenotypes between the two mutants, ipmdh2/ipmdh2 IPMDH3/ipmdh3 and IPMDH2/ipmdh2 ipmdh3/ipm dh3, which might be explained if there is functional redundancy between the IPMDH2 and IPMDH3 genes Genetic Confirmation of ipmdh2 ipmdh3 Male Gametophytic Lethality In the qrt mutant, the products of meiosis, a tetrad of four haploid microspores, fails t o separate during male gametophyte development, resulting in quartets of pollen grains (Preuss et al. 1994). It is easy to differentiate sporophytic and gametophytic mutants in the qrt background (Robertson et al., 2004; Gusti et al., 2009; Suzuki et al. 2009). If the mutation has sporophytic defects, the number of normal and defective pollen can differ in a given quartet. In contrast, if the mutation is gametophytic, each quartet always will have two normal and two mutant gametophytes. The mutant ipmdh2 /IPMDH2 ipmdh3/ipmdh3 was crossed into the qrt background. IPMDH2/ipmdh2 ipmdh3/ipmdh3 qrt/qrt and ipmdh2/ipmdh2 IPMDH3/ipmdh3 qrt/qrt plants were identified in the F2 generation, and their pollen grains were inspected. As shown in Figure 37 A, quartets of mature pollen grains from qrt single mutants are perfectly ordered with four uniformly shaped and sized microspores. Tetrads of I PMDH2/ipmdh2 ipmdh3/ipmdh3 qrt/qrt and ipmdh2/ipmdh2 IPMDH3/ipmdh3 qrt/qrt always produced two normal and two small, misshapen microspores. Occasionally some of the small pollen grains were shrunken, suggesting possible cytoplasmic degradation in these smaller pollen grains (Figure 37 B). Germination of these aborted pollen grains was examined and the
89 aberrant pollen grai ns faile d to germinate (Figure 37 C). These results provide strong evidence that the smaller gametophytes were ipmdh2 ipmdh3 mutants, which are lethal. Female Gametophyte Development Slowed in ipmdh2 ipmdh3 Mutants According to the TE data, the female gametophyte development was impaired ipmdh2 ipmdh3 double mutants (Table 33). Evidence for this was seen in the developing siliques of IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipmdh3 mutant plants, in which small, white, unfertilized ovules were often obs erved that rarely were found in wildtype siliques (Figure 38). The actual frequencies of aborted ovules were 23.7% (n=1245) and 21.9% (n=1545) for IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipmdh3 plants, respectively. This is consistent with t he expected values (23.6% and 21.6%, respectively) on the basis of the TE data (Table 33). Consequently, seeds set in IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipmdh3 siliques showed 25% and 24% reduction, respectively, compared to the wildtyp e plants. To analyze the nature of the abortion in the female gametophytes, the pistils at floral stage 12c (Smyth et., 1990) were emasculated and cleared. The embryo sacs were examined under Nomarski optics 12h after emasculation (stage 13). In ovaries of wild type plants, 90% (n=145) of the gametophytes developed correctly then had one central cell, one egg cell, and two synergid cells (Figure 39A). While in ovaries of IPMDH2/ipmdh2 ipmdh3/ipmdh3, 72% (n=234) of embryo sacs di splayed this phenotype, whi le 26% of embryo sacs were at a younger stage of gametophyte development with polar nuclei that had not fused and antipodal cells that were intact (Figure 39B). Similarly, 75% (n=212) of the analyzed embryo sacs in ovaries of
90 ipmdh2/ipmdh2 IPMDH3/ipmdh3 a ppeared to be the wildtype phenotype and 22% were at a younger stage of development (Figure 39C). To assess whether these embryo sacs could mature, IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipmdh3 plants were emasculated and the embryo sac all owed to grow for 48h and then examined under microscope. Surprisingly, 87% (n=192) and 89% (n=202) of the embryo sacs from IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipmdh3 plants, respectively, formed mature female gametophytes. These results indicate that the ipmdh2 ipmdh3 embryo sacs developed more slowly (Figures 39D F). To determine whether the ipmdh2 ipmdh3 embryo sacs were functional, pistils at floral stage 12c were emasculated and fertilized with wild type pollen 12h (stage 13), 24h (stage 14) and 48h after emasculation (stage15). Interestingly, for IPMDH2/ipmdh2 ipmdh3/ipmdh3 plants, a 84% (n=175) TE was identified at 48h, which is apparently higher than 79% (302) at 24h and 71% (317) at 12h (Table 36), suggesting that this delayed poll ination rescued the observed defect of these embryo sacs. Similarly, for ipmdh2/ipmdh2 IPMDH/ipmdh3 plants, the female TE of the ipmdh2 ipmdh3 gamete is enhanced to 91% (n=223) at 48h from the 73% (n=288) at 12h and 79% (n=198) at 24h. These results demons trate that ipmdh2 ipmdh3 embryo sacs develop more slowly than wildtype, causing them to miss the time of fertilization. HaploInsufficiency of IPMDH2 Results in Sporophytic Phenotype No obvious morphology phenotypes were observed in the following mutants: ipmdh1 ipmdh2, ipmdh3 single mutant, ipmdh1 ipmdh2 and ipmdh1 ipmdh3 double mutants, and ipmdh2/ipmdh2 IPMDH3/ipmdh3 mutant. Strikingly, plants with IPMDH2/ipmdh2 ipmdh3/ipmdh3 genotype showed a set of sporophytic abnormality.
91 The most profound phenotype was that the rosette leaves of IPMDH2/ipmdh2 ipmdh3/ipmdh3 plants exhibited interveinal chlorosis (Figure 310A). This occurred in cotyledons and young expanding rosette leaves. Chlorophyll contents were analyzed in these leaves. The IPMDH2/ipmdh2 ipmdh3/ ipmdh3 plants contained significantly less chlorophyll levels than wildtype plants (Figure 310B). In addition, the IPMDH2/ipmdh2 ipmdh3/ipmdh3 plants displayed a reduced growth rate (Figure 310B). In contrast, ipmdh2/ipmdh2 IPMDH3/ipmdh3 mutant leaves did not show these phenotypic defects, suggesting that IPMDH3 is responsible for a majority of isopropylmalate dehydrogenase activity in leaves (Figure 311). Interestingly, the ipmdh1/ipmdh1 IPMDH2/ipmdh2 ipmdh3/ipmdh3 mutant had a severe loss of total ch lorophyll and decrease in total fresh weight (Figure 310). This indicates that IPMDH1 IPMDH2 and IPMDH3 were active in leaves in a genedosage dependent manner. These leaf phenotypes could not be rescued by exogenously applying leucine, branchedchain am ino acids valine or isoleucine. Similarly, adding these amino acids to the nutrient solution of hydroponically grown plants had no effect (Data not shown). Enzyme Activities and Leucine Levels Were Affected in the Mutants To determine whether changes at t he transcription level of IPMDHs lead to any changes in enzyme activity in planta, isopropylmalate dehydrogenase activity was measured in 4weeks old leaves in a variety of different genotypes (Figure 33). As shown in Figure 311, enzyme activities were r educed in ipmdh1, ipmdh2 and ipmdh3 mutants by 14%, 13% and 27%, respectively. In double mutant ipmdh1 ipmdh2 and ipmdh1 ipmdh3, the enzyme activity was very similar to that of ipmdh2 and ipmdh3, indicating IPMDH1 contributed little to the overall activity IPMDH2/ipmdh2
9 2 ipmdh3/ipmdh3 plants showed a 64% reduction in activity, while ipmdh2/ipmdh2 IPMDH3/ipmdh3 plants had a 14% reduction. This result indicates that IPMDH3 provides a large portion of enzyme activity in leaves. The loss of this catalytic activ ity is consistent with the observed vegetative phenotype of the IPMDH2/ipmdh2 ipmdh3/ipmd h3 mutants (Figure 311), Likewise, the decline of enzyme activity could be further enhanced by the loss of ipmdh1, as shown by 73% and 39% reduction for ipmdh1/ipmdh1 IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh1/ipmdh1 ipmdh2/ipmdh2 IPMDH3/ipmdh3 respectively (Fi gure 311). To test whether the changes of enzyme activity in the mutants cause any changes in free leucine levels, free amino acid levels were analyzed by HPLC in the 4week old leaves. In ipmdh1, ipmdh2 and ipmdh3 single mutant, ipmdh1 ipmdh2 and ipmdh1 ipmdh3 double mutants, no significant variations in the levels of leucine or any other amino acids were found when compared to those in wild type leaves (Table 37 ). However, the leucine contents in IPMDH2/ipmdh2 ipmdh3/ipmdh3 plants were 59% of those present in wildtype. This result indicates that IPMDH2 and IPMDH3 work redundantly to catalyze leucine biosynthesis in vivo Additionally, histidine, isoleucine, lysi ne and valine levels were pleiotropically affected in IPMDH2/ipmdh2 ipmdh3/ipmdh3 plants ( Table 37 ). In ipmdh1/ipmdh1 IPMDH2/ipmdh2 ipmdh3/ipmdh3 leaves, leucine levels were reduced by 50% and the levels of aspartic acid, asparagine, histidine and methionine were significantly changed ( Table 37). No significant changes in leucine levels were observed in ipmdh2/ipmdh2 IPMDH3/ipmdh3 and ipmdh1/ipmdh1 ipmdh2/ipmdh2 IPMDH3/ipmdh3 plants. This result revealed that IPMDH3 plays a larger role than IPMDH2 in leuc ine biosynthesis
93 IPMDH2 and IPMDH3 Function in Glucosinolate Biosynthesis To test the hypothesis that IPMDH2 and IPMDH3 have overlapping functions with IPMDH1 in glucosinolate biosynthesis, the glucosinolate profiles were analyzed in seeds and leaves of the ipmdh2 and ipmdh3 single mutant, ipmdh1 ipmdh2 and ipmdh1 ipmdh3 double mutants as well as IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipmdh3 mutants. As shown in Figure 312, there were no glucosinolate changes in ipmdh2 or ipmdh3 single mutant plants, suggesting that the loss of single IPMDH2 or IPMDH3 has no impact on glucosinolate biosynthesis. In addition, the glucosinolate profiles in ipmdh1 ipmdh2 or ipmdh1 ipmdh3 double mutant plants appear to be identical to those in the single ipmdh1 mutant, suggesting that the loss of either IPMDH2 or IPMDH3 does not exert any additive effects to the glucosinolate changes attributed to the ipmdh1 mutation. Furthermore, no glucosinolate changes in IPMDH2/ipmdh2 ipmdh3ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipm dh3 plants were observed, suggesting that IPMDH2 and IPMDH3 are not involved in glucosinolate biosynthesis despite the obvious phenotypic and biochemical impacts (Figure 310 and Figure 311 ). Investigation of the ipmdh1/ipmdh1 IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh1/ipmdh1 ipmdh2/ipmdh2 IPMDH3/ipmdh3 genotypes revealed unexpected findings. The glucosinolate profiles in these genotypes were significantly different from those in the ipmdh1 single mutant (He et al., 2009). Glucosinolates with chainlength longer than 4C were undetectable, and 4C glucosinolate levels were significantly lower than in the ipmdh1 single mutant. Surprisingly, 3C glucosinolate levels were still present at high levels, as observed in ipmdh1 mutant plants (Figure 9). These results show ed that IPMDH2 and IPMDH3 work additively in glucosinolate biosynthesis when IPMDH1 was disabled. The functional allele of IPMDH2 and IPMDH3 in the mutants may
94 contribute to the presence of 3C and 4C glucosinolates in the mutants. However, the existence of other genes specialized in the first cycle of methionine chainelongation leading to 3C glucosinolates can not be excluded. TissueSpecific Expression of the IPMDH2 and IPMDH3 To understand tissuespecific expression of IPMDH2 and IPMDH3 the coding seque glucu ronidase gene was cloned inframe into the pCAMBIA1305 vectors containing the full length genomic IPMDH2 or IPMDH3 promoter sequence, respectively. The generated IPMDH2GUS and IPMDH3 GUS reporter constructs were transformed into wildtype pl ants. For each construct, fifteen independent T2 transgenic lines were tested for GUS activity, and all exhibited similar expression patterns during vegetative and reproductive growth. The IPMDH2GUS was active at significantly lower levels (Figure 313). IPMDH2 GUS and IPMDH3 GUS signals were present in cotyledons, leaves, hypocotyls and roots. During reproductive growth, higher IPMDH3GUS expression is consistent with a function in gametophyte development (Figures 313). In open flowers, expression of GUS was evidently detected in carpels, sepals, anther, mature pollen and stigma, but not in ovules. High GUS activity could be observed in embryo sacs of mature pistils that had been cut to facilitate the diffusion of 5 bromo4 chloro3 indolyl glucu ronide (X Gluc), indicating that the carpel tissue surrounding embryo sacs might inhibit GUS substrate penetration. The expression patterns of IPMDH2 and IPMDH3 were further analyzed by real time PCR and IPMDH1 was included for comparison. The transcripts of both I PMDH2 and IPMDH3 were present in seedlings, roots, stems, leaves, green buds, open flowers, and pollen. IPMDH2 transcripts were found at much lower levels in the above tissues (Figure 314). In contrast, IPMDH1 could barely be detected in the flowers, desp ite its
95 high expression in the vegetative tissues. This result is consistent with previous findings (He et al., 2009). IPMDH2 and IPMDH3 have overlapping expression profiles in agreement with the genetic evidence that IPMDH2 and IPMDH3 are functionally redundant. In addition, the expression patterns of IPMDH2 and IPMDH3 are clearly distinct from those of IPMDH1 suggesting that the differential expression of IPMDHs contributes to their functional divergence in plants. Discussion IPMDH2 and IPMDH3 Are Functi onally Redundant in Leucine Metabolism IPMDH2 and IPMDH3 share 89% amino acid identity suggesting that they might have redundant function. Biochemical analysis showed that IPMDH2 and IPMDH3 have higher affinity and activity t oward 3IPM than IPMDH1 (Table 3 2 ), implying that IPMDH2 and IPMDH3 are involved in the leucine metabolism. Neither ipmdh2 nor ipmdh3 single mutant showed any reduction in leucine levels or obvious phenotypic abnormalities, indicating that IPMDH2 and IPMDH3 are functionally redundant. The double knockout mutant ipmdh2 ipmdh3 could not be obtained in the progeny of self fertilized double heterozygous mutant, which indicates one of the two genes is required for viability. The doublemutant was likely a gametophyte or embryo lethal. Further molecular and genetic characterization confirmed that IPMDH2 and IPMDH3 were redundant and essential for male gametophyte development and affected female gametophyte development. Noticeably, the importance of IPMDH in gametophyte development has escaped forward genetic screening so far, most likely due to the genetic redundancy of IPMDH2 and IPMDH3 IPMDH2 and IPMDH3 transcript expression overlaps extensively (Figure 314). The substantial reduction of IPMDH activity and free leucine abundance was observ ed as copies of IPMDH2 and IPMDH3
96 were lost (Figure 311 ). Moreover, the bifunctional feature of IPMDH1 which was recognized previously (He et al., 2009) is further confirmed in this study of plants with the ipmdh1/ipmdh1 IPMDH2/ipmdh2 ipmdh3/ipmdh3 genoty pe, where the simultaneous loss of function of IPMDH1 added to the resultant morphological and chemical phenotypes of the mutant plants with IPMDH2/ipmdh2 ipmdh3/ipmdh3 genotype (Figures 310 and 311, Table 47 ). IPMDH2 and IPMDH3 Are Essential for Male G ametophyte Development Our g enetic and phenotypic analyses demonstrated that the presence of a functional IPMDH2 or IPMDH3 is essential for pollen development. The doublemutant pollen was smaller than wildtype and the cellular contents were altered. The electrondense aggregates might be caused by alterations in lipid accumulation. The nuclear division occurred normally in these double mutants (Figures 34). The double mutant pollen had an irregular intine layer, and lacked small round particles in the ve getative cytoplasm, consistent with the cellular defects of these mutant microspores occurring late in development Finally, pollen development was arrested during germination (Figures 35). I t is tempting to speculate that the null ipmdh2 ipmdh3 allele re sulted in pollen abortion at the maturation stage. The pollen defects were fully rescued with a full length genomic copy of IPMDH2 or IPMDH3 (Table 34 and 35), which demonstrated that the pollen phenotype was caused by TDNA insertions in these two genes and not by positional effects of the TDNA insertion or other gene mutations. During development of haploid microspores, pollen grains show high metabolic activity. Starch, protein and other nutrients accumulate in the pollen, and there are demands for energy and metabolic resources during pollen tube growth (McCormick, 2004, Leon et al., 2007). A mutation in the plastidic phosphate translocator GPT1
97 causes a disruption in the oxidative pentose phosphate cycle and leads to malegametophyte lethality (Niewi adomski et al., 2005). Serine palmitoyltransferase, an enzyme involved in the de novo synthesis of sphingolipids, was shown to be essential for pollen development (Teng et al., 2008). More recently, it has shown that the pur4 mutant, defective in de novo p urin e biosynthesis, compromised male gametophyte development and to a smaller extent female gametophyte development (Berthom et al., 2008). Given that IPMDH2 and IPMDH3 are unambiguously involved in leucine metabolism, the defects observed in double mutants could be due to the lack of leucine, causing the loss of essential metabolites for normal pollen development Analysis of genetic lines with fewer functional IPMDH alleles led us to propose that autonomous leucine biosynthesis impaired in the double ipm dh2 ipmdh3 mutant is required for pollen development. In addition, strong activities of IPMDH2 and IPMDH3 promoters in pollen supported their functional relevance and indicated a direct mode of action as opposed to an indirect effect by activities in surrounding tissues. Furthermore, IPMDH2GUS and IPMDH3 GUS activities were very low prior to mitosis, but became very high after mitosis (Figure 313). This result suggests that IPMDH activity may not be required for the early meiosis. A fraction of leucine pool could be carried over from the diploid microspore mother cell, which could sufficiently allow the beginning steps of male gametophyte development to occur, including tetrad formation and mitotic divisions. Once the original leucine pool was depleted, autonomous leucine biosynthesis becomes indispensable for pollen development. Mutation in IPMDH2 and IPMDH3 clearly affected leucine biosynthesis and metabolism. Our interpretation is consistent with the
98 idea that active metabolism activities are required for gametogenesis (McCormick 2004). Involvement of IPMDH2 and IPMDH3 in Embryo Sac Development The reduced TE of ipmdh2 ipmdh3 genotype in female gametophytes indicates IPMDH2 and IPMDH3 are also important for normal embryo sac development. Indeed, a substantial number of aborted ovules were found in developing siliques of double heterozygotes and mutants containing only one functional copy of IPMDH2 or IPMDH3 (Figure 38). Both of these genes are expressed in embryo sac (Figure 313) and this phenotype supports their role during female gametophyte development. The nuclear division of cells in the gametophyte was slowed, but seemed to be less affected than in pollen. This is likely due to continued uptake of leucine from maternal tissue. At cellularization st age, however, fusion of the two polar nuclei phenotype was consistently postponed. The frequency of this occurrence correlated well with the fraction of the aborted ovules genetically that were predicted to harbor the double mutant genotype. Importantly, the abortion rate in mutants was markedly reduced by manually fertilizating pistils later in gametophyte formation (Table 3 6). This result indicates that the embryo sacs of the double mutants were able to reach the terminal development stage (containing one secondary nucleus, one egg cell nucleus and two synergid cell nuclei) if the time for their development was extended. This phenotype is different from those of gfa2, gpt1 and sdh1 mutants (Christensen et al. (2002), Niewiadomski et al. (2005) and Leon et al. (2007), respectively), in which embryo sac development arrested at a stage before the fusion of the polar nuclei. Thus, the phenotype in ipmdh2 ipmdh3 female gametophytes does not result from a developmental defect, but from a reduction in the pool of a required metabolite.
99 It was noticed that a fraction of the ipmdh2 ipmdh3 female gametophytes (approximately 50%) still remains functional. Therefore, in contrast to its fully penetrant male gametophytic character, the double mutant was only partially p enetrant in female gametophyte development, suggesting that some double mutant gametophytes had sufficient leucine to develop. It is plausible that 1) the female gametophytes benefit from a larger quantity of inputs from surrounding sporophytic cells than male gametophyte, then a stochastic inheritance and persistence determines the variable distribution of nutrient among embryo sacs (Stadler et al., 2005). In this context, it makes sense that the male mutant gametophytes have a more severe phenotype. 2) IP MDH2 and IPMDH3 are more highly expressed in male than in female, which could be related to each gametophytic phase having specific requirements at the levels of enzyme and/or leucine. As a result the double mutant was fully impaired in male gametogenesis, whereas those mutants exhibited a weaker effect on female gametophyte development. 3) given that IPMDH1 is highly homologous to IPMDH2 and IPMDH3 and IPMDH1 transcripts could be detected in pistils, it could be speculated that IPMDH1 might be able to part ly substitute for IPMDH2 and IPMDH3 in female specificity (He et al., 2009). However, this possibility could be ruled out by the fact that still about half of the ipmdh1 ipmdh2 ipmdh3 female gametophytes are viable based on that the female transmission ef ficiency of ipmdh1 ipmdh2 ipmdh3 gametes is 49.2% (n=109) and 48.3% (n=113) for ipmdh1/ipmdh1 ipmdh2/ipmdh2 IPMDH3/ipmdh3 or ipmdh1/ipmdh1 IPMDH2/ipmdh2 ipmdh3/ipmdh3 mutants, respectively. IPMDHs are essential for pollen development and important for embr yo sac development.
100 IPMDH2 and IPMDH3 Affect Sporophytic Development As pointed out by Feldmann et al. (1997), gametophytelethal mutations can be separated into two types. The first type refers to mutation that affects the gametophytic phase of the life c ycle specifically. The second type refers to mutation that affects both gametophytic and sporophytic phase of the life cycle. Because leucine is a basic precursor for protein synthesis and is also related to hormone regulation (Staswick and Tiryaki, 2004), defects in leucine biosynthesis can impose severe impact on plant sporophytic growth. Such an essential role is reflected by the mRNA expression profiles of IPMDH2 and IPMDH3 which extends to nearly all the tissues analyzed. Because double ipmdh2 ipmdh3 mutation causes gametophytelethal, the study of sporophyte development was not possible for the double mutants. However, the plants with partial loss of IPMDH2 and complete loss of IPMDH3 showed a number of sporophytic phenotype. The most striking phenoty pe was the appearing reticulate leaf with palegreen color. Such a phenotype has previously been described for several mutations in genes encoding chloroplast proteins (Kinsman and Pyke et al., 1998, Streatfield et al., 1999, Voll et al., 2003). Because the chloroplast ultrastructure did not show any obvious differences between mutants and wildtype plants (Figure 315), we reason that chloroplast functionality might be affected in the mutants. Interestingly, the iil mutant with decreased expression of LeuC show the similar chlorotic leaf phenotype, suggesting that this phenotype was possibly pathway related, rather than single stepspecific (Sureshkumar et al., 2009). In contrast, the plants with partial loss of IPMDH3 and complete loss of IPMDH2 did not s how obvious sporophytic phenotype. This can be explained by the hypothesis that IPMDH3 might be responsible for the majority of IPMDH activity in leaf tissues. This hypothesis is supported by the evidence that
101 IPMDH3 is consistently expressed at higher lev els than IPMDH2 ( Figures 313 and 315). Additionally, the haploinsufficiency of IPMDH2 under IPMDH3 knockout background also suggests the gene dosage effect of IPMDH activity. The detailed cellular and molecular bases of the aforementioned phenotypes re main unknown. Exogenous application of leucine and other branchedchain amino acids did not rescue the phenotypes. It is possible that leucine was not readily transported to the sites of action when applied exogenously. Previous studies with mutants in met hionine and glycine biosynthesis showed that exogenous application of the needed amino acids did not rescue the mutant phenotypes (Joshi et al., 2006, Joshi and Jander et al., 2009). The relationship between gametophytic phenotype and sporophytic phenotype is not clear. Generation of transgenic lines, in which the gametophytic defects of the double mutants are removed (e.g., by complementing the double mutant with IPMDH2 or IPMDH3 driven by pollenspecific promoters), would be beneficial to address the spor ophytic phenotype caused by the double mutants. Involvement of IPMDH2 and IPMDH3 in Glucosinolate Biosynthesis Recently, IPMDH1 was characterized as a key enzyme involved in the chainelongation cycle of methioninederived aliphatic glucosinolate biosynthesis (He et al., 2009). Because the ipmdh1 knockout mutant still has limited capability to complete the methionine chainelongation reaction in the absence of a functional IPMDH1 protein (e.g., longchain aliphatic glucosinolates were not completely abolis hed and 3C glucosinolate levels dramatically increased), it is reasonable to postulate that there are other gene (s) responsible for the remaining isopropylmalate dehydrogenase activity (He et al., 2009). The Arabidopsis genome encodes two other IPMDH homologs, IPMDH2 and IPMDH3 whose products share over 80% sequence identity with
102 IPMDH1 This posed the possibility that IPMDH2 and IPMDH3 could participate in the methionine chainelongation cycle. However, no glucosinolate alterations were observed in singl e ipmdh2 or ipmdh3 mutant. In addition, the glucosinolate composition in ipmdh1 ipmdh2 and ipmdh1 ipmdh3 double mutants did not show any changes in glucosinolate profiles that were not in the single ipmdh1 mutant. This indicates that IPMDH2 and IPMDH3 do not compensate for IPMDH1 during glucosinolate biosynthesis. Even when the enzyme activities were significantly reduced and the morphological/growth phenotypes were obvious in IPMDH2/ipmdh2 ipmdh3/ipmdh3 plants, there were no significant changes in glucosinolate profiles (Figure 312 ). However, by investigating glucosinolate profiles in the following genotypes, ipmdh1/ipmdh1 IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh1/ipmdh1 ipmdh2/ipmdh2 IPMDH3/ipmdh3 we discovered that the glucosinolate variations observed in the ipmdh1 mutant was exacerbated. We therefore conclude that IPMDH2 and IPMDH3 are involved in glucosinolate biosynthesis, but their contribution was undetectable if IPMDH1 is functional. Since 3C and 4C aliphatic glucosinolates were present in these genotypes, other enzymes catalyzing the oxidative decarboxylation reaction in methionine chainelongation might be found in Arabidopsis. The activation/replacement of IPMDH activity by different alleles offers plants metabolic flexibility and may be important for plant adaptation during evolution and under environmental stress conditions.
103 Table 31. Primers used in this study. Primers Primer Sequence (5' 3') Tm IMD2g L GCTCTAGACGTTCATCTCAAGTTTGGATTCATCTA 55 IMD2g R GCCACGTGACCAAGATACGCAATCCCAATGAACAT 55 IMD1 1L GCGAATTCGGGAAAAAACGGTATAACAT 55 IMD1 1R TGTCGACTTAAACAGTAGCTGGAACTTTGGAT 55 IMD2 1L GCGAATTCGGTAAAAAACGATACACCA 55 IMD2 1R GCGTCGACTTAAACAGAAGCTGGAACT 55 IMD3 1L GCGAATTCGTGAAAAAACGGTATAACA 55 IMD3 1R GCGTCGACTTAAACAGGAACTTTGGAG 55 IMD1 2L GA ACATCAGTCTGAATGCGATCAA 51 IMD1 2R GGTACAATGTTGCATGGATTAGCA 51 IMD2 2L CTACGTTCAGAGCTGTAAG 51 IMD2 2R AGTCCCTCTAATAATCTCAC 51 IMD3 2L CTGGAGATCATACCGGGAA 53 IMD3 2R AACCAAGGGCTACACACTT 51 IMD1 3L TGAAGCACTTGAAAAAGTTTGTG 55 IMD1 3R TTGCAGAATCGATCCCAAA T 55 IMD2 3L TGAAGAGTTCATTTCCAACAACA 55 IMD2 3R TCAACGTTTCAGTTCGATCA 55 IMD3 3L GAAAAGTGTGTAGCCCTTGGTT 55 IMD3 3R TTGCAGAATCGATCCCAAAT 55 Actin2/8 L GGTAACATTGTGCTCAGTGGTGG 55 Actin2/8 R AACGACCTTAATCTTCATGCTGC 55 IMD3 4L CCTTCCCATTGTTGGTGTCGC 54 I MD3 4R CAGACGGATGTTAGTTTGC 51 GUS L CCGCGGCGATGGTAGATCTGAGGAAC 55 GUS R CCGCGGTCACACGTGATGGTGATG 55
104 Table 32 Enzyme affinity and activity toward 3isopropylmalate K m SE (M) V max SE (mol min -1 mg -1 ) K cat (min -1 ) K cat /K m (M -1 min -1 ) IPMDH1 25.21 2.30 0.93 0.10 0.37 X 102 0.15 X 107 IPMDH2 10.85 1.32 4.05 0.36 3.73 X 102 3.44 X 107 IPMDH3 9.17 1.43 4.98 0.33 5.43 X 102 5.92 X 107 The activities were measured according to the standard assay as described by He et al., 2009. Data are presented as mean values of three independent experiments. The numbers after represent standard errors.
105 Table 33 Transmission efficiency of ipmdh2 ipmdh3 allele in reciprocal crosses. Cross performed Resultant progeny % TE b (b) % TE b (d) % TE b (bd) Female parent a Male parent a BBDD BBDd BbDD BbDd Wt BbDD 90 85 94.4 c BbDD Wt 96 92 95.8 d Wt BBDd 93 76 81.7 c BBDd Wt 112 102 91.1 d Wt BbDd 76 56 72 0 54.5 c 37.8 c 0 c BbDd Wt 72 68 70 39 77.9 d 75.4 d 54.2 d Cross performed Resultant progeny % TE b (bd) Female parent Male parent BBDd BbDd Wt Bbdd 127 0 0 c Bbdd Wt 212 112 52.8 d Cross performed Resultant progeny % TE b (bd) Female parent Male parent BbDD BbDd Wt bbDd 155 0 0 c bbDd Wt 260 148 56.9 d a B, IPMHD2 ; b, ipmdh2; D, IPMDH3 ; d, ipmdh3. The progeny genotype was determined by PCR and used to infer the genotype of the gamete contributed by the mutant parent in each cross. b TE was calculated as follows: TE (%) = (observed no.of mutant allele/observed no.of wild-type allele) X 100 (Ebel et al, 2004) c refer to male TE. d refer to female TE.
106 Table 34. Genetic analysis of the ipmdh2 ipmdh3 allele in transgenic lines harboring the complementation construct T 1 genoty pes Construct a bbDD bbDd IPMDH2 19 11 T 2 genotypes b bbDD bbDd c bbdd c 7 16 (7) 6 (2) T 1 genotypes IPMDH3 BBdd Bbdd 20 10 T 2 genotypes b BBdd Bbdd c bbdd c 9 17 (6) 9 (3) a The full -length genomic copy of IPMDH2 including promoter and terminor regions were reintroduced into IPMDH2/ipmdh2 ipmdh3/ipmdh3 plants and the full -length genomic copy of IPMDH3 were reintroduced into IPMDH2/ipmdh2 ipmdh3/ipmdh3 plants, respectively. b For each construct, one of T1 transgenic lines was randomly selected for further genetic analysis based on the subsequently observed 3:1 ratio (hygromycin-resistant versus hygromycin-sensitive) within T2 generation, which is symptomatic of the single T-DNA insertion. c Numbers in parentheses indicate the number of plants harboring the homozygous transgenic construct
107 Table 35 The transmission efficiency of ipmdh2 ipmdh3 gamete using the transgenic plants as pollen donor Cross performed Resultant progeny a % TE c (bd) Female parent Male parent a b BBDd BbDd Wt Bbdd 43 33 87 Cross performed Resultant progeny a % TE c (bd) Female parent Male parent a, b BbDD BbDd Wt bbDd 39 32 90 a The full -length genomic copy of IPMDH2 including promoter and terminor regions were reintroduced into IPMDH 2/ipmdh2 ipmdh3/ipmdh3 plants and the full -length genomic copy of IPMDH3 were reintroduced into IPMDH2/ipmdh2 ipmdh3/ipmdh3 plants, respectively. b For each construct, one of T1 transgenic lines was randomly selected for further genetic analysis based on the subsequently observed 3:1 ratio (hygromycin-resistant versus hygromycin-sensitive) within T2 generation, which is symptomatic of the single T-DNA insertion.
108 Table 36 Transmission efficiency of ipmdh2 ipmdh3 gamete through female from delay ed pollination test. Parent Genotype a HAE b Resultant progeny TE BBDd BbDd Bbdd 12 317 225 71% 24 302 239 79% 48 1 75 154 88% Resultant progeny BbDD BbDd bbDd 12 288 210 73% 24 198 156 79% 48 245 223 91% Pistils of IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipmdh3 plants were emasculated at floral 12c (Shi et al., 2005; Li et al., 2009) and pollinated with pollens from wildtype plants at 12h (stage 13), 24h (stage 14), or 48h (stage 15) after emasculation (Smyth et al., 1990) a B, IPMDH2 ; b, ipmdh2; D, IPMDH3 ; d, ipmdh3. The progeny genotype was determined by PCR and used to infer the genotype of the gamete contributed by the mutant parent in each cross. b Hours after emasculation
109 Table 37 Amino acid profiles of wild type and different m utants indicate a significant difference when compared to wild -type with p<0.05. a A, IPMDH1 ; a, ipmdh1 ; B, IPMDH2 ; b, ipmdh2; D, IPMDH3 ; d, ipmdh3. Amino Acids Wild type aaBBDD AAbbDD AABBdd aabbDD aaBBdd AABbdd AAbbDd aaBbdd aabbDd Ala 0.269 0.018 0.265 0.010 0.259 0.014 0.263 0.016 0.268 0.024 0.258 0.019 0.263 0.015 0.268 0.012 0.288 0.012 0.275 0.020 Arg 0.009 0.001 0.011 0.002 0.011 0.001 0.010 0.001 0.011 0.001 0.012 0.002 0.012 0.002 0.011 0.002 0.012 0.003 0.009 0.003 Asn 0.092 0.008 0.093 0.008 0.093 0.006 0.094 0.008 0.092 0.006 0.095 0.009 0.094 0.007 0 .095 0.005 0.121 0.010* 0.092 0.008 Asp 0.443 0.022 0.439 0.016 0.451 0.010 0.436 0.016 0.444 0.017 0.445 0.015 0.441 0.020 0.445 0.025 0.385 0.024* 0.392 0.007* Gln 0.274 0.021 0.268 0.013 0.267 0.017 0.257 0.014 0.285 0.016 0.281 0.022 0.283 0.019 0.271 0.021 0.241 0.018 0.226 0.011* Glu 1.473 0.036 1.464 0.022 1.460 0.043 1.453 0.025 1.482 0.022 1.458 0.027 1.465 0.030 1.472 0.033 1.458 0.037 1.453 0.028 Gly 0.015 0.003 0.011 0.002 0.011 0.002 0.014 0.002 0.012 0.002 0.015 0.001 0.015 0.002 0.014 0.002 0.013 0.003 0.018 0.002 His 0.011 0.002 0.013 0.002 0.012 0.002 0.013 0.001 0.013 0.003 0.014 0.003 0.017 0.002* 0.012 0.002 0.018 0.002* 0.012 0.002 Ile 0.012 0.001 0.015 0.002 0.015 0.002 0.016 0.002 0.015 0.003 0.015 0.003 0.017 0.002* 0.014 0.001 0.013 0.003 0.013 0.002 Leu 0.012 0.001 0.013 0.002 0.014 0.002 0.014 0.002 0.014 0.002 0.015 0.003 0.007 0.002* 0.010 0.002 0.006 0.003* 0.009 0.002 Lys 0.020 0.002 0.019 0.002 0.021 0.002 0.022 0.002 0.023 0.002 0.024 0.004 0.027 0.002* 0.023 0.003 0.019 0.003 0.021 0.003 Met 0.029 0.002 0.033 0.002 0.028 0.003 0.027 0.003 0. 032 0.002 0.034 0.003 0.028 0.003 0.027 0.002 0.044 0.004* 0.038 0.004* Phe 0.019 0.002 0.023 0.002 0.020 0.002 0.021 0.003 0.023 0.002 0.022 0.003 0.022 0.002 0.022 0.002 0.023 0.004 0.018 0.003 Pro 0.050 0.004 0.058 0.002 0.051 0.005 0.053 0.003 0.045 0.003 0.050 0.006 0.051 0.005 0.052 0.004 0.052 0.007 0.055 0.003 Ser 0.096 0.006 0.094 0.010 0.091 0.011 0.094 0.008 0.102 0.013 0.106 0.010 0.087 0.002 0.086 0.006 0.098 0.003 0.099 0.004 Thr 0.173 0. 010 0.176 0.012 0.179 0.012 0.172 0.010 0.169 0. 011 0.182 0.012 0.184 0.016 0.170 0.010 0.183 0.015 0.167 0.010 Tyr 0.005 0.001 0.006 0.001 0.006 0.001 0.007 0.002 0.006 0.001 0.007 0.002 0.007 0.00 1 0.006 0.002 0.005 0.001 0.005 0.001 Val 0.046 0.004 0.054 0.002 0.051 0.003 0.054 0.004 0.056 0.005 0.051 0.005 0.056 0.004* 0.049 0.003 0.043 0.006 0.041 0.004
110 Figure 31. Heterologous expression of three IPMDHs in E. coli and purification using Nickel affinity chromatography. A total of 10 g crude lysate and purified protein were loaded onto a 12% SDS PAGE gel, respectively. The purified enzymes appeared homogeneous as judged on the gel stained with BioSafe Coomassie.
111 Figure 32. Schematic diagram of the genomic structure of IPMDHs wit h TDNA insertion sites (triangles). Bars indicate exons and lines represent introns. ATG, start codon; STOP, stop codon; Scale bar, 100 bp.
112 Figure 33. Generation and confirmation of single, double and triple mutants of IPMDH genes. A) Diagram of the crossing experiments and the progeny genotypes (A, IPMDH1 ; a, ipmdh1; B, IPMDH2 ; b, ipmdh2; D, IPMDH3 ; d, ipmdh3). The mutants were numbered 210. B) Transcript analysis of the selected mutants (numbered the same as in the top panel) using gene specific primers Actin1 gene was used as a loading control.
113 Figure 34. Microscopic phenotypes of pollens from wildtype (left), IPMDH2/ipmdh2 ipmdh3/ipmdh3 (middle) and ipmdh2/ipmdh2 IPMDH3/ipmdh3 plants (ri ght). A) Bright field images of wholemount Alexander stained pollens. B) Scanning electron micrographs of mature pollen grains. C) Fluorescent images of wholemount DAPI stained pollens, showing one centrally located vegetative nucleus and two sperm nuclei. Bar = 20 m.
114 Figure 35. Germination of pollens from wildtype, IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipmdh3 mutant plants. Light field images of in vitro germination of pollen grains. The germination rates listed below the images were calculated according to this formula: (number of germinating pollens / number of observed pollens) X 100. The mean and standard error from five independent experiments were presented. Approximately 300 pollen grains were counted in each replicate. Bar = 20 m.
115 Figure 36. Transmission Electron Microscopy (TEM) of ipmdh2/ipmdh2 IPMDH3/ipmdh3 and IPMDH2/ipmdh2 ipmdh3/ipmdh3 mutant pollen grains. A) and B) Pollen grains at bicellular stage. Plastids containing large numbers of starch granules are enclosed in a dashed oval in A) and marked with an arrow in B). VN, vegetative nucleus; SN, sperm nucleus. C) Low magnification image showing two normal pollen grains (black arrowheads) and abnormal pollen grains (white arrowheads) in a dehiscing pollen sac Electron dense aggregates in the abnormal pollen grains are marked with dashed circles. D) to F) High magnification images of abnormal pollen grains. Irregular intine wall is indicated wit h arrowheads in D) and E). Note the ER cisternae associated with dark aggregates (arrows in E) and F)). G, Golgi stack; M, mitochondrium; P, plastid (starch particles are marked with *); IN, intine cell wall. G) and H) High magnification image of a mature normal pollen grain from the mutant plant G) and from a wildtype plant H). In both pollen grains, intine cell wall is flat, lipid bodies (L) are regular in size (also see C)), and the cytoplasm has round particles (arrow). Scale bars in A) and C): 10
116 Figure 37. Pollen development and germination of ipmdh2 ipmdh3 in qrt background. Left, middle and right panels represent qrt parental control, IPMDH2/ipmdh2 ipmdh3/ipmdh3 qrt/qrt and ipmdh2/ipmdh2 IPMDH3/ipmdh3 qrt/qrt plants, respectively. A) Light field images of tetrads. B) Alexander staining of tetrads. C) Light microscopy of in vitro germination of a tetrad from qrt showing four normal pollen tubes (left), and tetrads from IPMDH2 /ipmdh2 ipmdh3/ipmdh3 qrt/qrt (middle) and ipmdh2/ipmdh2 IPMDH3/ipmdh3 qrt/qrt (right) showing segregation of healthy and aberrant pollen tubes. Bar = 20 m.
117 Figure 38. Seeds are developed normally in wildtype siliques A ), while 25% (n=1245) and 24% (n=1545) aborted ovules (arrows) are found in B) IPMDH2/ipmdh2 ipmdh3/ipmdh3 and C ) ipmdh2/ipmdh2 IPMDH3/ipmdh3 siliques, respectively.
118 Figure 39. Loss of function of IPMDH2 and IPMDH3 caus es delayed fusion of two polar nuclei in embryo sacs. The pistils at floral stage 12c of wildtype, IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh2/ipmdh2 IPMDH3/ipmdh3 mutant plants were emasculated, and wholemount preparations of ovules were analyzed by differential interference contrast (DIC) Nomarskioptics after 12h or 48h emasculation. Embryo sac nuclei are indicated by arrows. The inserted numbers represent the percentages of the ovules shown in the image to the total number of ovules (n) examined. CC, central cell nucleus; EC, egg cell nucleus; PN, polar nucleus; SYN, synergidcell nucleus. A) and D) DIC images of wild type embryo sacs after 12h and 48h emasculation, respectively. B) and E) DIC images of ovules harboring mutant embryo sacs in the ovaries of IP MDH2/ipmdh2 ipmdh3/ipmdh3 plants after 12h and 48h emasculation, respectively. C) and F) DIC images of ovules harboring mutant embryo sacs in the ovaries of ipmdh2/ipmdh2 IPMDH3/ipmdh3 plants after 12h and 48h
119 Figure 310. Sporophytic phenotypes of IPMDH2/ipmdh2 ipmdh3/ipmdh3 and ipmdh1/ipmdh1 IPMDH2/ipmdh2 ipmdh3/ipmdh3 plants. Plants were grown in MS medium containing 1% (w/v) sucrose for one week before being transferred into soil. A) P h otographs of four week old wild type (left), IPMDH2/ipmdh2 ipmdh3/ipmdh3 (middle) and ipmdh1/ipmdh1 IPMDH2/ipmdh2 ipmdh3/ipmdh3 plants. B) Chlorophyll and fresh weight measurements of the plants in A). Chlorophll was extracted from rosette leaves according to Strain et al.,1971. Fresh weight of the aerial part was determined from seven individual plants. The error bars represent standard errors.
120 Figure 311. In vivo IPMDH activi ty in wild type, s ingle, double and triple IPMDH mutants. The numbers are the same as in Figure 32. 1 wild type, 2 ipmdh1, 3 ipmdh2 4 impdh3, 5 ipmdh1 ipmdh2, 6 ipmdh1 ipmdh3, 7 IPMDH2/ipmdh2 ipmdh3/ipmdh3, 8 ipmdh2/ipmdh2 IPMDH3/ipmdh3 9 ipmdh1/ipmdh1 IPMDH2/ipmdh2 ipmdh3/ ipmdh3 10. ipmdh1/ipmdh1 ipmdh2/ipmdh2 IPMDH3/ipmdh3 S tandard enzyme assay was performed using chloroplast fractions prepared from 4week old rosette leaves. Plotted are the mean standard errors from three independent experiments.
121 Figure 312. Glucosinolate prof iles in wild type and different IPMDH mutants. The numbers are the same as in Figure 32. 1 wild type, 2 ipmdh1, 3 ipmdh2, 4 impdh3, 5 ipmdh1 ipmdh2, 6 ipmdh1 ipmdh3, 7 IPMDH2/ipmdh2 ipmdh3/ipmdh3, 8 ipmdh2/ipmdh2 IPMDH3/ipmdh3 9 ipmdh1/ipmdh1 IPMDH2/ipmdh 2 ipmdh3/ ipmdh3 10. ipmdh1/ipmdh1 ipmdh2/ipmdh2 IPMDH3/ipmdh3 A) seeds, B) leaves.
122 Figure 313. E xpression patterns of IPMDH2 and IPMDH3 analyzed using promoter GUS plants. A) to E) transgenic plants harboring IPMDH2GUS construct. F) to J) transgenic plants harboring IPMDH2GUS construct. A) and F) 3day old seedling. A) and G) 3week old leaf. C) and H) flower stalk with a buds and flowers at different stages. D) and ( I) mature flower. E) and J) embryo sac.
123 Figure 314. Quantitative real time RT PCR analysis of IPMDH transcript levels in various tissues Relative levels to ACTIN2/8 gene are presented with standard deviation. Data were from three independent experiments.
124 Figure 3 15. Ultra structure of chloroplasts in first true leaves. A) wildtype, B) IPMDP2/ipmdh2 ipmdh3/ipmdh3, C) ipmdh2/ipmdh2 IPMDH3/ipmdh3.
125 CHAPTER 4 FUNCTIONAL SPECIFICATION OF THE SMALL SUBUNITS OF ISOPROPYLMALATE ISOMERASE IN GLUCOSINOLATE AND LEUCINE BIOSYNTHESIS Introduction IPMI is an enzyme that catalyzes the stereospecific isomerization of 2 isopropylL malate isopropylmalate) to give 3isopropylL isopropylmalate) via the formation of cis dimethylcitraconate (Drevland et al., 2007) IPMIs h ave been identified in several prokaryotic and eukaryotic organisms, and have been classified into two groups based on the forms of protein assembly ( Yasutake et al., 2004). Fungal IPMI is a monomeric protein with two distinct domains. In bacteria and archaea, the two domains are expressed as separate protein subunits to produce a heterodimeric enzyme, i.e., a sm all polypeptide interacts with a large Fe4S4containing polypeptide to form an active enzyme ( Drevland et al., 2007; Yasu take et al., 2004) All the IPMI enzymes are relatively conserved and belong to the aconitase superfamily. In Arabidopsis one gene se quence ( AtLeuC / At4g13430) was found with high similarity to sequences encoding the large subunit of IPMI and three gene sequences ( AtLeuD1/At2g43100, AtLeuD2/At3g58990 and AtLeuD3/ At2g43090) with high similarity to sequence s encoding the small subunit of I PMI Recently, the large subunit has been identified to be involved in both chainelongation cycle of glucosinolate biosynthesis and leucine metabolism (Knill et al., 2009; S awada et al., 2009). However, the specific function of each small subunit has not been characterized. In this chapter w e demonstrate that AtLeuD1 and AtLeuD2 are functionally redundant in the methionine chainelongation cycle of aliphatic glucosinolate biosynthesis This function is support ed by several lines of evidence: 1) S imultaneous disruption of AtLeuD1 and AtLeuD2 caused significant change s in gluco sinolate profile. 2) AtLeuD1 and AtLeuD2 showed
126 an overlapping tissuespecific expression pattern, which was distinct from AtLeuD3. 3 ) L ike AtLeuD3 AtLeuD1 and AtLeuD2 encode smal l su bunit s of IPMI that were able to rescue the growth of the auxotropic E.coli leu C and leu D mutant s after forming a heter odimeric complex with the large subunit In addition, in vivo interaction between AtLeuC and the A tLeuDs was revealed by coimmunoprecipit ation analysis Futhermore, no homozygous knockout mutant s of AtLeuD3 could be obtained, suggesting that AtLeuD3 plays an essential role in Arabidopsis. However, the exact molecular basis underlying atleud3 mutant lethality remains to be identified. Materi als and Methods Plants and Chemicals The following Arabidopsis ecotype Columbia (Col 0) seeds and Salk mutant lines were obtained from the Arabidopsis Biological Resource Center: Col 0 (CS3879), atleuc (Salk_029510), atleud1 (Salk_048320), atleud31 (Salk_ 111666) and atleud32 (Salk_115899). Plants were surface sterilized and germinated on agar plates containing Murashige and Skoog salts plus vitamins, 2% sucrose, 2.5 mM MES and 0.8% agar, pH 5.7. The seeds were stratified for 3 day s at 4C in the dark before transferring to a controlled growth chamber for germination at 22C, 16h light, and 70% relative humidity. Small seedlings of 12day s old were transferred to soil and grown in the same growth chamber DNA Extraction and Genotyping Genomic DNA extrac tion and P CR reac tions were performed as previously described (He et al., 2009) All PCR primers used in this study are listed in Table 41. Homozygous atleud1 mutant was isolated using gene specific primers LeuD1LP, LeuD1RP and TDNA left border primer LBa1. Mutant atleud31 was is olated using
127 gene primers LeuD3LP 1, LeudD3RP1 and LBa1 and atleud3 2 was i solated using gene primers LeuD 3 LP 2, LeuD3RP2 and LBa1. RNA Interference of AtLeuD2 and Plant Transformation A fragment spanning nucleotides 6 244 was PCR amplified with the primer pairs LeuD2L 1/LeuD2R 1 and LeuD2 2/LeuD2R 2 The PCR products were ligated with the uidA fragment in both antisense and sense orientations. The resulting construct was then cloned into an overexpression vector pCAMBIA1305. Agro bacterium tumefaciens strain C58C1 was used to transform the construct into an Arabidopsis plant through hygromycin. RNA Extraction and RT PCR RNA extraction, cDNA synthesis and RTPCR were conducted as previously described (He et al., 2009), except the following PCR primers were used: LeuD11L and LeuD11R for AtLeuD1 LeuD21L and LeuD21R for AtLeuD2, and LeuD31L and LeuD31R for AtLeuD3 Actin1 F and Actin1R p rimers were use to amplify Actin1 as equal loading control Promoter GUS Fusion and GUS Assay The AtLeuC promoter (from 2402 to 46 bp) was amplified fr om genomic DNA using primers LeuC 1L and LeuC 1R, and cloned into a pGEM T vector (Promega, USA). The p romoter sequence was verified by sequencing and subcloned into pCAMBIA1305 us ing Sa l I and Nco I restriction sites. The construct was transformed into Arabidopsis as described above using the floral dipping method. Histochemical GUS assays were performed on T2 transgenic plants as described (Jefferson et al., 1987).
128 Complementation of Arabidopsis atleuc M utant Flagtagged AtLeuC was generated using PCR with primers LeuC 2L and LeuC 2R, where 3 end of LeuC 2R primer containing the Flag sequence (DYKDDDDK). I t was then cloned into the pGEM T vector. After verification of the sequence by sequencing, AtLeuC was subcloned into the pCAMBIA1305 vector to create 35S:: AtLeuC using Nco I and Pml I sites. To generate native promoter:: AtLeuC construct, the 35S promoter se quence was replaced with the AtLeuC promoter sequence using SalI and Nco I sites. These constructs were transformed into the homozygous atleuc mutant. Protein Expression and Antibody Production The cDNA sequences of At LeuD1, AtLeuD2 and AtLeuD3 without the N termi nal plastid targeting peptide were cloned into pET28a expression vector (Novagen, USA) using primer pairs LeuD12L/LeuD12R, LeuD22L/LeuD22R and LeuD32L/LeuD32, respectively E xpression and purification of the recombinant proteins were the same as described (He et al., 2009). P urified AtLeuD3 protein (1 mg) was sent to Cocalico Biologicals Inc. (Pennsylvania, USA) for antibody production. Chloroplast Fractionation Chloroplast isolation and stroma fraction purification were conducted as previous ly described (He et al., 2009). Coimmunoprecipitation and Western Blotting T otal protein was extracted from the purified stroma fraction using immunoprecipitation lysis/binding buffer (50 mM Tris/HCl, pH7.5, 150 mM NaCl, 10 mM MgCl2, 1 mM EDTA, 1 mM EGTA, 0.5% Triton, 0.5% NP 40, 1 mM DTT, 10% glycerol, 1mM PMSF and 1X protease inhibitor cocktail ). The extract was centrifuged at 14,000
129 rpm for 10 min at 4 C, and the protein concentration in the supernatant was determined using the Bradford assay (BioRad). A total of 2 mg protein was incubated with 3 mg of affinity purified anti AtLeuD2 antibody for 3 h at 4 C followed by precipitation with protein A/G Sepharose beads for 2 h. Alternatively aliquots of 50 l of Anti Flag antibody conjugated agarose beads (Sigma Aldrich) were mix ed with 1 mg of the total protein extracts for 3 h at 4 C. A fter washing the beads four times with lysis/binding buffer, the affinity bound proteins were eluted by boiling for 5 min in 2X SDS Laemmli sample buffer before loading onto a 12% SDS polyacrylamide gel Western blot was conducted as described (Chen and Halkier, 1999). Proteins immunoprecipitated with anti AtLeuD2 were probed with anti FLAG antibody (1:10,000) and vice versa (anti AtLeuD2, 1:2,000). Glucosinolate and Amino Acid Analysis Seeds and 4week old ro sette leaves were used for glucosinolate and free amino acid analysis. Glucosinolates were analyzed using HPLC MS (Chen et al., 2003; Alvarez et al., 2008). Free amino acids were extracted using a methanol and chlorofor m method (Colebatch et al., 2004; Schauer et al., 2005) and profiled using an HPLC based precolumn derivatization protocol (Schuster et al., 2006). Complementation of E. coli l euC and leuD Mutants AtLeuC cDNA without the N termi nal plastidtargeting pepti de sequence was cloned into CDFDuet 1 vector (Novagen, USA) using primer pairs LeuC 3L/LeuC 3R to create CDFDuet AtLeuC Then t he cDNA sequences of At LeuD1, AtLeuD2 and AtLeuD3 were subcloned into CDFDuet AtLeuC to construct CDFDuet AtLeuC/AtLeuD1, CDFDuet AtLeuC/AtLeuD2 and CDFDuet AtLeuC/AtLeuD3 using the primer pairs LeuD13L/3R, LeuD23L/3R and LeuD23L/3R, respectively. It should be
130 noted that 6X His tags were included in each LeuD sequence at the N terminus by in cluding the 6X His tags sequence in the p r imers. E. coli strain CV522 (deficient in LeuC ) and CV524 (deficient in LeuD ) were used for complementation studies. The strains were able to grow on M9minimal medium (containing 1 mM MgSO4, 0.1 mM CaCl2, 2% glucose and 1 mM thiamine) when supplemented with casaminoacids (Difco) or transformed with the constructs expressing functional IPMI To ensure expression of T7 polymerase driven constructs, the IPTG inducible T7polymerase gene was intergrated into strain CV522 ( leuC222/F+) and CV524 ( leuD211 /F+) lysogeni zation kit (Novagen, USA) producing CV5 2 2 (DE3) and CV524 (DE3) (Textor et al., 2007) The plasmids CDFDuet (empty vector), CDFDuet LeuC, CDFDuet LeuD1, CDFDuet LeuD2, CDFDuet LeuD3, CDFDuet LeuC/LeuD1, CDFDuet LeuC/LeuD2 and CDFDuet LeuC/LeuD3 were chemically transformed into CV522 (DE3) and CV524 (DE3), respectively For the complementation test, one colony containing each of the constructs grown on LuriaBertani (LB) media was streaked on solid (1.5% agar) M9 minimal medium, supplem ented with 50 g/ml spectrom ycin 1 mM IPTG, and 0.3 mM leucine. The plates were incubated overnight at 37C. Alternatively, each colony was grown in liquid LB medium and 5 ml of the saturated suspension was centrifuged at 6,000 rpm for 3 min. The bacterial pellet was suspended and diluted in LB media back to an OD600=0.05. It was then added to M9 m inimal media supplemented with 1 mM IPTG and incubated at 37C, 250 rpm.
131 Results At LeuD1 and AtLeuD2 Are Redundantly Functional in Glucosinolate Biosynthesis There are three genes in the Arabidopsis genome encoding the putative small subunit of IPMI. Two of them ( AtLeuD1 and AtLeuD2) showed close coexpression with glucosinolate biosynthetic genes (Hirai et al., 2007; Sawada et al., 2009). It could be expected that one or both was likely to interact with AtLeuC to function in the isomerization step of methionine chainelongat ion cycle. Because no glucosinolate changes were observed in the homozygous atleud1 mutant (Figure 41A, B) and no TDNA mutants of AtLeuD2 gene were available, RNA interference (RNAi) was used to suppress AtLeuD2 in wild type and the atleud1 mutant. In particular, the AtLeuD2RNAi construct was introduced into wild type plants to generate atleud2 and into atleud1 plant to generate AtLeuD2RNAi/ a tleud1. To this end, a 239 bp coding fragment of AtLeuD2 was used to generate the AtLeuD2RNAi construct (Figure 41C). Twenty transgenic lines were chosen in each background wildtype or atleud1 mutant As shown by RTPCR analysis (Figure 4 1D), AtLeuD2 e xpression was remarkably decreased in the RNAi plants. In most of the transgenic lines, AtLeuD3 still showed identical expression levels as in wild type plants, suggesting high specificity of AtLeuD2 silencing (Figure 41D). To test the function of AtLeuD 1 and AtLeuD2 in glucosinolate biosynthesis, glucosinolate profiles were determined in seeds and 4week old leaves of atleud1 mutant and AtLeuD2RNAi lines. As shown in both leaves and seeds, the glucosinolate profiles in atleud1 and AtLeuD2RNAi plants we re indistinguishable from wildtype plants (Figure 42). However, in AtLeuD2RNAi/ atleud1 plants, the glucosinolate profiles changed substantially. There was a remarkable accumulation of 3C aliphatic
132 glucosinolates and a significant decrease of longer chai n aliphatic glucosinolates (Figure 42). This indicates that AtLeuD1 and AtLeuD2 are functionally redundant in glucosinolate biosynthesis. In addition, AtLeuD3 did not seem to be involve d in this reaction since cosuppression of AtLeuD3 with AtLeuD1 or AtLe uD2 did not show any additive effect on glucosinolate levels (Figure 42). It should be noted that the plants used in this study were from the T1 transgenic generation due to the instability of LeuD2RNAi in the T2 generation ( Wang et al., 2005; Small, 200 7 ) To test whether AtLeuD1 and AtLeuD2 are also redundant in leucine biosynthesis, the levels of free amino acids in seeds were analyzed. No significant changes in leucine levels were observed, although several other amino acids displayed significant changes in AtLeuD2 RNAi/ atleud1 plants as compared to wildtype (Table 4 2). Temporal and Spatial Regulation of AtLeuC and AtLeuDs To investigate tissue specific patterns of AtLeuC expression, we generated transgenic Arabidopsis plants containing an AtLeuC p romoter glucuronidase (GUS) construct. Of 12 independent ProAtLeuC::GUS transgenic lines selected, all displayed similar patterns of GUS expression. S trong GUS expression was observed at three days after germination in hypocotyls and roots (Figure 43A). Twelve days after germination, intense GUS stain was detected in roots and newly emerging leaves and weak stain ing in hyp ocotyls and cotyledons (Figure 43 B). At rosette stage (3 week old), GUS activity was predominantly present in th e main veins of leave s (Figure 43 C). At the flowering stage (9week old), GUS activity was high in all floral tissues including sepal, stamen, pistil as well as pollen grains (Figure 4 3D), while it was hardly present in developing siliques (Figure 43F)
133 The expression pattern of AtLeuC was further confirmed by semi quantitative RTPCR with total RNA extracted from various wild type tissues. AtLeuC was constitutively expressed in all the tissues examined with high expression in roots, leave s, stems and flowers, but low in sil iques (Figure 43G) Collectively, the profile of AtLeuC expression exhibited the typical pattern observed for genes involved in the biosynthesis of aliphatic glucosinolates (Chen et al., 2003; Schuster et al., 2006; He et al., 2009). The expression of th e three AtLeuDs was also determined using the RTPCR analysis. AtLeuD1 and AtLeuD2 showed nearly identical expression patterns with the highest level in stems (Figure 43G ). AtLeuD3 exhibited high expression in roots, leaves, flowers and stems. In general, the above gene expression profile is consistent with the data from the publicly available Arabidopsis Microarray Database (GENVESTIGATOR; https://www.genevestigator.ethz.ch/). AtLeuDs Associate Physically with AtLeuC in vivo To investigate potential inte raction between AtLeuC and AtLeuD in planta, we constructed a chimeric gene using the AtLeuC native promoter to drive expression of the full length AtLeuC protein with a Flagtag at the C terminus. The chimeric gene was transformed into the atleuc mutant, and T2 progeny carrying a singlelocus transgene were selected. As shown in Figure 44, the glucosinolate phenotype of the atleuc mutant was rescued completely in the LeuCFlag lines, suggesting that the introduced LeuCFLAG protein was functional in planta. It also confirmed the role of AtLeuC in glucosinolate biosynthesis. To examine whether the AtLeuCFLAG fusion protein interacts with AtLeuDs in vivo monoclonal anti Flag antibody was used to precipitate the AtLeuCFLAG fusion protein from the LeuCFlag tran sgenic plants. As expected, immunoprecipitation of AtLeuCFLAG
134 readily pulled down endogenous AtLeuDs (Figure 45A). In a reciprocal experiment, anti LeuD antibody was used for an immunoprecipitation experiment, which not only immunoprecipitated AtLeuD2, but also pull ed down AtLeuCFLAG (Figure 45B), demonstrating the existence of the IPMI complex in vivo IPMI Complex Is Localized to Chloroplast Stroma It is known that both leucine biosynthesis and the methioni n e chainelongation cycle occur in chloroplasts (Falk et al. 2004; Binder et al. 2007; Textor et al. 2007; Knill et al. 2008; He et al., 2009), implying that the IPMI should be localized in this organelle. Consistent with this expectation, AtLeuD1 was shown to be localized in chloroplasts using GFP fus ion protein analysis (Knill et al., 2009). To determine the specific subcellular localization of AtLeuC and AtLeuDs intact chloroplasts wer e isolated from LeuCFLAG transgenic plants and fractionated into di fferent suborganelle fractions. Proteins extract ed from different chloroplastic fractions were subjected to immunoblot analysis using anti Flag antibody, anti LeuD antibody, an antibody against a thylakoidmembrane PsbO, and an antibody against the stromalocalized RbcL (Haussuhl et al., 2001; Weigel et al., 2003; He et al., 2009) As shown in Figure 46, the signal s of AtLeuC and AtLeuDs were readily detected in the intact chloroplasts and the stroma fraction. N o positive signal could be observed in the thylakoid/envelope fraction. These results demonst rate the Arabidopsis IPMI complex is localized in chloroplast stroma Complementation of E. coli Leucine A uxotrophs by Coexpressi on of AtL euC and AtL euDs To test the function of IPMI (formed by AtLeuC and one of the AtLeuDs) in vivo the auxotropic E.coli strain s bacterial expression vector constructs containing the empty vector as control (CDFDuet -
135 empty), AtLeuC (CDFDuet AtLeuC), AtLeuD s (C DFDuet AtLeuD1, CDFDuet AtLeuD2 and CDFDuet AtLeuD3) as well a s the combinations of AtLeuC with each of AtLeuDs (CDFDuet AtLeuC/AtLeuD1, CDFDuet AtLeuC/AtLeuD2 and CDFDuet AtLeuC/AtLeuD3). The constructs producing the functional IPMI were expected to complement the E.coli mutat ion by restoring the ability of mutants to grow on minimal medium without supplemental leucine. C omplementation of the leu C mutant or leu D mutant on minimal medium was only observed with the constructs containing both AtLeuC and one of AtLeuDs (Figure 47). Neither AtLeuC itself nor any of the A tLeuD genes alone was able to complement the E.coli mutant (Figure 47), suggesting that the AtLeuC or the AtLeuD subunit can t form a functional enzyme with the E.coli LeuD or LeuC subunit, respectively (Xu et al., 2004). atleud3 Mu t ants Exhibit Lethal P henotype Two mutant alleles were identified to carry T DNA insertions inside the AtLeuD3 gene ( atleud3 1 : Salk 111666 and atleud3 1 : Salk_115589) In contrast to the homozygous knockout atleud1 mutant, no homozygous plants could be found for the T DNA alle le in either atleud3 line, suggesting AtLeuD3 is an essential gene in Arabidopsis and an AtLeuD3 null mutant could cause em bryo and/or gamete lethality This result is consistent with the report by Knill et al., 2009. Based on the evidence that the number of mature seeds was substantially reduced in two independent lines and no homozygous mutants could be obtained, they concluded that AtLeuD3 is crucial for the development of mature viable seeds (Knill et al., 2009). However, the precise mechanisms underlying the lethality are not known because the reduced seed number could be also caused by a gametophyte defect. Therefore, further investigation is
136 needed to elucidate AtLeuD3 function in leucine biosynthesis and its potential involvement in gametophyte devel opment. Discussion AtLeuC Is Functional in Both Leucine and Glucosinolate Biosynthesis The presence of one gene encoding the large subunit and three genes encoding small subunit s of IPMI suggest a bacterial type heterodimeric structure of this enzyme in Ar abidopsis (Binder et al., 2007; Knill et al., 2009). Compared to wildtype plant s, 2 IPM and 2(3 methylsulfinyl)propylmalate, the intermediate of leucine and glucosinolate biosynthesis respectively, were substantially accumulated in atleuc knockdown mut ant s, indic a ting that AtLeuC functions in both pathways (Knill et al., 2009). In addition, the involvement of AtLeuC in glucosinolate biosynthesis is also demonstrated by the results that glucosinolate composition was significantly altered in the mutant (K nill et al., 2009; Sawada et al., 2009) and this alteration could be fully restored by the reintroduction of a functional copy of AtLeuC (Figure 44). Interestingly, the glucosinolate phenotype in the atleuc mutant appeared to be identical to that in the ipmdh1 mutant, which is exemplified by the increase in shorter chain glucosinolates and the reduction in longchain glucosinolates (He et al., 2009). In contrast, the defects in the other genes in the chainelongation pathway, i.e. MAM1 MAM3 BCAT3 and BC AT4 led to different changes in glucosinolate profiles (Field et al., 2004; Schuster et al., 2006; Knill et al., 2007; Textor et al., 2007) This suggest s that IPMI and IPMDH function in a similar manner It is possible that IPMI and IPMDH might interact to channel substrates and intermediates in the isomerization and decarboxylation reactions of aliphatic glucosinolate biosynthesis.
137 The presence of short chain glucosinolates and the unaffected leucine abundance implies the existence of residual isomeriza tion activity in the atleuc mutant. This could be simply explained by the fact that the mutant, atleuc is a knock down mutant, and the remaining decreased AtLeuC transcript may be sufficient for sustaining leucine biosynthesis and at least the beginning cy cles of the methionine chainelongation in aliphatic glucosinolate biosynthesis (Knill et al., 2009). However, the possibility of the occurrence of a bypass pathway involved in these reactions cannot be ruled out. AtLeuD1 and AtLeuD2 Are Involved in Glucosinolate Biosynthesis The specificity of the large subunit of IPMI toward the substrates of leucine biosynthesis and the chainelongation cycle of aliphatic glucosinolate biosynthesis was suggested to be determined by its partner, the small subunit of IPMI (Gruer et al., 1997; Yasutake et al., 2004). Indeed, three genes encoding the small subunits in Arabidopsis appear to be specialized to either leucine biosynthesis or the methionine chain elongation cycle (Knill et al., 2009) Several lines of evidence dem onstrate that AtLeuD1 and AtLeuD2 are redundantly involved in glucosinolate biosynthesis. First the gene coexpression network analysis showed that AtLeuD1 and AtLeuD2 were closely co regulated with all the known genes in aliphatic glucosinolate biosynthes is (Hirai et al., 2007; He et al., 2009; Sawada et al., 2009). Second, a single mutation of AtLeuD1 or AtLeuD2 did not cause any glucosinolate alterations. In contrast simultaneous disruption of both led to large change s in glucosinolate profiles (Figure 4 2). Last AtLeuD1 and AtLeuD2 exhibited overlapping tissuespecific expression pattern, which is distinct from AtLeuD3. It should be noted that the glucosinolate composition in AtLeuD2RNAi/ atleud1 plant s is very similar to what was observed in the atle uc mutant (Figures 42, 44) This
138 indicates that the shift towards shorter chain C3 glucosinolates is a general phenomenon when the IPMI step is impaired. Genetic c omplementation studies showed that like AtLeuD3, AtLeuD1 and AtLeuD2 can form a functional enzyme with AtLeuC to rescue the growth of auxotropic E.coli leuC and leuD mutants. This result suggests that these genes function in leucine biosynthesis. However, we did not observe any changes of leucine level s in plants when AtLeuD1 and AtLeuD2 functions were simultaneous ly disabled. Therefore, we propose that the functional complex, AtLeuC/AtLeuD1 and AtLeuC/AtLeuD2, are primarily involved in glucosinolate biosynthesis in planta. AtLeuD3 Is an Essential Gene Consistent with the result described by Knil l et al., 2009, no homozygous plants could be obtained for atleud3 mutants in this study. This indicates that AtLeuD3 is an essential gene in Arabidopsis, most likely due to its participation in leucine biosynthesis since it has been shown by cyp79f1/cyp79f2 and myb28/myb29 mutants that the complete loss of aliphatic glucosinolates did not result in lethality in plant development (Tantikanjana et al., 2004 Beekwilder et al., 2008). C ompar ed with wild type, the leucine level s were not changed in the mutants with the decreased transcript ion of AtLeuD3, suggesting that the residual transcript left may be sufficient for leucine biosynthesis The cellular mechanism underlying the lethality caused by the mutation of AtLeuD3 is unknown. S eed germination results show ed that both mutant alleles had the comparable germination rates as wild type plants (data not shown) suggesting that the defect in gametophytes rather than in embryogenesis may be responsible for the lethality. Intensive investigation using genetic appr oaches is under way to elucidate the function of AtLeuD3 in gametophyte development.
139 Table 41. P rimers used in this study Name Sequence (5' 3') Tm Purpose LeuC LP GCTTGATACAATGAAACCTCAATG 63 Genotyping Salk_029510 LeuC RP ACCAGGACCACAAACATCATG 65 LeuD1 LP TAAGCCCAAAAAGGGAAATTG 64 Genotyping Salk_048320 LeuD1 RP TACATTTCCCCAAGCCTTACC 64 LeuD3 LP1 CTTCTCTGCAATCAGCAAACC 64 Genotyping Salk_111666 LeuD3 RP1 TGGAAAAGATCAGCCAACATC 64 LeuD3 LP2 GTGGTGGCAAAATCATGAAAG 64 Genotyping Salk_115899 LeuD3 RP2 TAAACAAACCGGAGCATGTTC 64 LeuD2 L1 TCTAGAGACTTCTCAGCAATTTTTAAAC 62 Clone the sense strand to construct AtLeuD2RNAi LeuD2 R1 GGATCCCTTTCAAGACGAAGCAGAGG 70 LeuD2 L2 AGATCTGACTTCTCAGCAATTTTTAAAC 64 Clone the antisense strand to construct AtLeuD2 RNAi LeuD2 R2 GATATCCTTTCAAGACGAAGCAGAGG 68 LeuD1 1L CAAGCCTTACCTTGCTCGTC 64 AtLeuD1 specific primer for semiquantitative RT PCR LeuD1 1R TCACTCAGCTCGATCGTCACCGT 70 LeuD2 1L CTCCTCCGCCACGATCATCACA 72 AtLeuD2 specific primer for semiquantitative RT PCR L euD2 1R GTCTTCCTTGTGTTCGATTGTCACCA 71 LeuD3 1L ATTCAACATCCGTCGCTTCC 66 AtLeuD3 specific primer for semiquantitative RT PCR LeuD3 1R TCCCTCAACTCAACAGTCGCAACA 72 LeuC 1L GTCGACGCTAGGAAGGGAATCTAT 66 Clone AtLeuC promoter LeuC 1R CCATGGCGGAGATTAGAGGCGTGT GT 76 LeuC 2L CCATGGCTTCTGTTATCTCTTC 61 Clone Flag tagged AtLeuC LeuC 2R CACGTGCTACTTATCATCATCATCTTTATAATCCTGC AAGAACTCCCTTGG 80 LeuD1 2L GCCATATGCTAGCCAACACAACCTTCC 73 Clone AtLeuD1 without signal peptide into pET28a LeuD1 2R GCGGATCCTTAAGCTAATGATGG AATC 69 LeuD2 2L GCCATATGATAACCAGAGAGACTTTCCA 68 Clone AtLeuD2 without signal peptide into pET28a LeuD2 2R GCGGATCCTCAAGCAGAAGGAATCAT 73 LeuD3 2L GCCATATGCAAGAAAGAAAAACCTTCCA 71 Clone AtLeuD3 without signal peptide into pET28a LeuD3 2R TAGGATCCTCAAGC AGCAGCAGATGGA 75
140 Table 4 1. Continued LeuC 3L GCGAATTCATGACAATGACGGAGAAG 72 Clone AtLeuC without sig n al peptide into CDFDuet LeuC 3R GCGTCGACCTACTGCAAGAACTCCCTT 74 LeuD1 3L CATATGCATCACCATCATCACCACCTAGCCAACACAACC TTCC 84 Clone AtLeuD1 without signal peptide into CDFDuet LeuD13R GATATCTTAAGCTAATGATGGAATC 58 LeuD2 3L CATATGCATCACCATCATCACCACATAACCAGAGAGACTTTC 80 Clone AtLeuD2 without signal peptide into CDFDuet LeuD2 3R GATATCTCAAGCAGAAGGAATCAT 61 LeuD3 3L GCCATATGCATCACCATCATCACCACCAAGAAAGAAAAA CCTTC 82 Clone AtLeuD3 without signal peptide into CDFDuet LeuD3 3R CTCGAGTCAAGCAGCAGCAGATGG 73
141 Table 42. Amino acid profiles of wild type and different mutants Amino Acid Wt atleud1 AtLeuD2 RNAi::Wt 1 AtLeuD2 RNAi::Wt 5 AtLeuD2 RNAi::atleud1 5 AtLeuD2 RNAi::atleu d1 2 Ala 0.223 0.011 0.229 0.017 0.241 0.211 0.297 0.033* 0.246 0.042 0.333 0.027* Arg 0.071 0.007 0.075 0.004 0.071 0.004 0.081 0.009 0.079 0.006 0.125 0.008* Asn 0.489 0.019 0.495 0.023 0.521 0.039 0.519 0.037 0.5 07 0.036 0.665 0.036* Asp 0.402 0.010 0.419 0.027 0.372 0.037 0.432 0.046 0.409 0.038 0.432 0.037 Gln 0.095 0.006 0.098 0.008 0.111 0.023 0.145 0.013* 0.094 0.012 0.166 0.026* Glu 1.871 0.021 1.898 0.026 1.951 0.089 1.961 0.037* 1.895 0.053 2.129 0.052* Gly 0.236 0.018 0.231 0.018 0.225 0.023 0.264 0.036 0.253 0.029 0.253 0.026 His 0.041 0.003 0.043 0.003 0.049 0.007 0.062 0.004* 0.064 0.004* 0.065 0.004* Ile 0.085 0.006 0.089 0.006 0.078 0.007 0.082 0.006 0.087 0.007 0.121 0.010* Leu 0.071 0.004 0.071 0.003 0.073 0.004 0.071 0.003 0.075 0.006 0.073 0.006 Lys 0.037 0.002 0.032 0.003 0.042 0.005 0.044 0.005 0.039 0.004 0.040 0.005 Met 0.055 0.004 0.057 0.004 0.053 0.004 0.076 0.006* 0.087 0.007* 0.089 0.005* Phe 0.155 0.011 0.158 0.009 0.169 0.018 0.163 0.010 0.140 0.018 0.212 0.029* Pro 0.149 0.007 0.142 0.010 0.139 0.015 0.155 0.011 0.135 0.025 0.217 0.023* Ser 0.371 0.016 0.392 0.028 0.369 0.029 0.369 0.033 0.392 0.035 0.412 0.047 Thr 0.155 0.011 0.166 0.012 0.174 0.028 0.196 0.019* 0.168 0.022 0.224 0.024* Trp 0.536 0.025 0.521 0.015 0.519 0.042 0.551 0.037 0. 531 0.046 0.577 0.047 Tyr 0.064 0.003 0.068 0.003 0.057 0.007 0.065 0.006 0.070 0.007 0.071 0.006 Val 0.215 0.011 0.227 0.014 0.241 0.037 0.222 0.011 0.212 0.021 0.299 0.025* indicate a significant difference when compared to wild -type with p<0.01
142 Figure 41. Characterization of TDNA knockout mutant of AtLeuD1 and RNA interference mutants of AtLeuD2 A) Schematic representation of the genomic structure of AtLeuD1 (top) with T DNA insertion sites as well as AtLeuD 2 wit h the RNAi targeted site (bottom ). Solid bars indicate exons. Solid lines represent introns. ATG, start codon; STOP, stop codon. B) Genotyping of the At LeuD1 mutant plants. A set of genespecific and TDNA specific primers used for genomic PCR are indicated. W, wild type; T, T DNA insertion. C) Schematic presentation of the AtLeuD2 RNAi construct. D) RT PCR analysis of wild type (line 1), homozygous atleud1 (line 2), and representative AtLeuD2RNAi suppression lines under wildtype (lane 3, AtLeuD2 RNAi::Wt 1 and line 4, AtLeuD2 RNAi::Wt 5) and under atleud1 background (line 5, AtLeuD2RNAi:: atleud15 and line6, AtLeuD2RNAi::atleud12 ) with specific primers for AtLeuD1, AtLeuD2, and AtLeuD3. The A ctin 1 gene was used as an internal control for analysis of the amounts of template
143 Figure 42 Glucosinolate profiles of wild type, atleud1 mutant and LeuD2RNAi plants. Aliphatic glucosinolates are grouped according to their chain length, and indole glucosinolates were summed as one group. D ata shown are the means and standard deviations (SD) of at least three replicates. A) Leaves; B) Seeds.
144 Figure 43 Spatial and temporal expression patterns of AtLeuC analyzed using promoter GUS plants. A) 3 day old seedling; B) 12day old seedling; C) 3week old leaf; D) flower; E) Pollen Grains; F) young silique; G) Semi quantitative RTPCR confirmation of AtLeuDs and AtLeuC expression. Actin1 was used as control for equal loading (bottom row);
145 Fi gure 44 Glucosinolat e profiles of wild type, atleuc mutant and the comple me nted plants. Aliphatic glucosinolates are grouped according to their chain length, and indole glucosinolates were summed as one group. Data shown are the means and standard deviations (SD) of at least three replicates. A) Leaves; B) Seeds
146 Figure 45 Direct Interaction between AtLeuC and AtLeuDs in vivo A) Chloroplastic protein prepared from wildtype (Wt) and AtLeuC flag transgenic (LeuCFLAG) plants were inc ubated with anti Flag antibody coupled agarose. The immunoprecipitates (IP) and the total extracts (T) were run on SDS PAGE, subjected to immunoblot analysis with antibodies against Flag and AtLeuDs B) Chloroplastic protein prepared from Wt and LeuCFLAG p lants were incubated with anti AtLeuDs antibody conjugated beads. The IP and the T were run on SDS PAGE, subjected to immunoblot analysis with antibodies against Flag and AtLeuDs. E, empty bead immunoprecipitation control
147 Figure 46 Subcellular localization of AtLeuC and AtLeuDs within Arabidopsis chloroplast compartments Protein extract e d from the purified intact chloroplast and twenty microgram proteins per lane. SDS PAGE (12% [w/v] ) and Western blots performed with the anti FLAG and antibodies raised against the recombinant AtLeuDs. Antibodies Rbcl and PsbO were specific for stroma or thylakoid correspondingly.
148 Figure 47 Complementation of E.coli leu C and leu D auxotrophic mutant s. A) and B) Growth of CV522 and CV524 containing the indicated constructs on M9 minimal medium with or without leucine. The appropriate antibiotics and inducer were included in the medium. Line 1, CDFDuet empty vector; Line 2, CDFDuet AtLeuD1; Line 3, CDF Duet AtLeuD2; Line 4, CDFDuet AtLeuD3; Line 5, CDFDuet AtLeuC; Line 6, CDFDuet AtLeuC/AtLeuD1; Line 7, CDFDuet AtLeuC/AtLeuD2; Line 8, CDFDuet AtLeuC/AtLeuD3. C) and D) Growth curves of CV522 and CV524 containing the indicated constructs were grown overni ght LB media, washed three times in sterilized H2O, diluted to OD600=0.005, then added to M9 minimal media supplemented with 0.1mM IPTG and incubated for a timecourse growth analysis at 37C. Each point represents the average of three independent experime nts
149 LIST OF REFERENCES Alexander, M.P. (1969). Differential staining of aborted and nonaborted pollen. Stain technology 44, 117122. Alkhalfioui, F., Renard, M., Vensel, W.H., Wong, J., Tanaka, C.K., Hurkman, W.J., Buchanan, B.B., and Montrichard, F. (2007). Thioredoxinlinked proteins are reduced during germination of Medicago truncatula seeds. Plant physiology 144, 15591579. Alonso, J.M., et al (2003). Genomewide insertional mutagenesis of Arabidopsis thaliana. Science 301, 653657. Alvarez, S. He, Y., and Chen, S. (2008). Comparative investigations of the glucosinolatemyrosinase system in Arabidopsis suspension cells and hypocotyls. Plant & cell physiology 49, 324333. Aronsson, H., and Jarvis, P. (2002). A simple method for isolating import competent Arabidopsis chloroplasts. FEBS letters 529, 215220. Bak, S., Olsen, C.E., Petersen, B.L., Moller, B.L., and Halkier, B.A. (1999). Metabolic engineering of phydroxybenzylglucosinolate in Arabidopsis by expression of the cyanogenic CYP79A1 from Sorghum bicolor. Plant J 20, 663671. Bak, S., and Feyereisen, R. (2001). The involvement of two p450 enzymes, CYP83B1 and CYP83A1 in auxin homeostasis and glucosinolate biosynthesis. Plant physiology 127, 108118. Bak, S., Tax, F.E., Feldmann, K.A., G albraith, D.W., and Feyereisen, R. (2001). CYP83B1 a cytochrome P450 at the metabolic branch point in auxin and indole glucosinolate biosynthesis in Arabidopsis. The Plant cell 13, 101111. Balsera, M., Goetze, T.A., KovacsBogdan, E., Schurmann, P., Wagner, R., Buchanan, B.B., Soll, J., and Bolter, B. (2009). Characterization of Tic110, a channel forming protein at the inner envelope membrane of chloroplasts, unveils a response to Ca2+ and a stromal regulatory disulfide bridge. The Journal of biological chemistry 284, 26032616. Barlier, I., Kowalczyk, M., Marchant, A., Ljung, K., Bhalerao, R., Bennett, M., Sandberg, G., and Bellini, C. (2000). The SUR2 gene of Arabidopsis thaliana encodes the cytochrome P450 CYP83B1 a modulator of auxin homeostasis. Pr oc. Natl. Acad. Sci. USA 97, 1481914824. Bednarek, P., PislewskaBednarek, M., Svatos, A., Schneider, B., Doubsky, J., Mansurova, M., Humphry, M., Consonni, C., Panstruga, R., Sanchez Vallet, A., Molina, A., and Schulze Lefert, P. (2009). A glucosinolate metabolism pathway in living plant cells mediates broadspectrum antifungal defense. Science 323, 101106.
150 Beekwilder, J., (2008). The impact of the absence of aliphatic glucosinolates on insect herbivory in Arabidopsis. PLoS One 3, e2068. Berthome, R., Thomasset, M., Maene, M., Bourgeois, N., Froger, N., and Budar, F. (2008). pur4 mutations are lethal to the male, but not the female, gametophyte and affect sporophyte development in Arabidopsis. Plant physiology 147, 650660. Binder, S., Knill, T. and Schuster, J. (2007) Branchedchain amino acid metabolism in higher plants. Physiol. Plant. 129, 68 78 Boavida, L.C., and McCormick, S. (2007). Temperature as a determinant factor for increased and reproducible in vitro pollen germination in Arabidopsis thali ana Plant J 52, 570582. Brader, G., Tas, E., and Palva, E.T. (2001). Jasmonatedependent induction of indole glucosinolates in Arabidopsis by culture filtrates of the nonspecific pathogen Erwinia carotovora. Plant physiology 126, 849860. Bradford, M.M (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of proteindye binding. Analytical biochemistry 72, 248254. Brown, P.D., Tokuhisa, J.G., Reichelt, M., and Gershenzon, J. (2003). Variation of glucosinolate accumulation among different organs and developmental stages of Arabidopsis thaliana. Phytochemistry 62, 471 481. Buchanan, B.B., and Balmer, Y. (2005). Redox regulation: a broadening horizon. Annual review of plant biology 56, 1872 20. Buchanan, B.B., and Luan, S. (2005). Redox regulation in the chloroplast thylakoid lumen: a new frontier in photosynthesis research. Journal of experimental botany 56, 14391447. Burow, M., Markert, J., Gershenzon, J., and Wittstock, U. (2006). Comparative biochemical characterization of nitrileforming proteins from plants and insects that alter myrosinase catalysed hydrolysis of glucosinolates. The FEBS journal 273, 24322446. Celenza, J.L., Quiel, J.A., Smolen, G.A., Merrikh, H., Silvestro, A.R., Normanly, J., and Bender, J. (2005). The Arabidopsis ATR1 Myb transcription factor controls indolic glucosinolate homeostasis. Plant physiology 137, 253262. Chen, S., and Halkier, B.A. (1999). Functional expression and characterization of the myrosinase MYR1 from Brassica napus in Saccharomyces cerevisiae. Protein expression and purification 17, 414420.
151 Chen, S., Petersen, B.L., Olsen, C.E., Schulz, A., and Halkier, B.A. (2001). Longdistance phloem transport of glucosinolates in Arabidopsis. Plant phys iology 127, 194201. Chen, S., Glawischnig, E., Jorgensen, K., Naur, P., Jorgensen, B., Olsen, C.E., Hansen, C.H., Rasmussen, H., Pickett, J.A., and Halkier, B.A. (2003). CYP79F1 and CYP79F2 have distinct functions in the biosynthesis of aliphatic glucosinolates in Arabidopsis. Plant J 33, 923937. Chisholm, M.D., and Wetter, L.R. (1964). Biosynthesis of Mustard Oil Glucosides. Iv. the Administration of Methionine C14 and Related Compounds to Horseradish. Canadian journal of biochemistry and physiology 42, 10331040. Christensen, C.A., Gorsich, S.W., Brown, R.H., Jones, L.G., Brown, J., Shaw, J.M., and Drews, G.N. (2002). Mitochondrial GFA2 is required for synergid cell death in Arabidopsis. The Plant cell 14, 22152232. Clay, N.K., Adio, A.M., Denoux, C ., Jander, G., and Ausubel, F.M. (2009). Glucosinolate metabolites required for an Arabidopsis innate immune response. Science 323, 95101. Clough, S.J., and Bent, A.F. (1998). Floral dip: a simplified method for Agrobacterium mediated transformation of A rabidopsis thaliana. Plant J 16, 735743. Colebatch, G., Desbrosses, G., Ott, T., Krusell, L., Montanari, O., Kloska, S., Kopka, J., and Udvardi, M.K. (2004). Global changes in transcription orchestrate metabolic differentiation during symbiotic nitrogen fixation in Lotus japonicus Plant J 39, 487512. Dai, S., Johansson, K., MiginiacMaslow, M., Schurmann, P., and Eklund, H. (2004). Structural Basis of Redox Signaling in Photosynthesis: Structure and Function of Ferredoxin:thioredoxin Reductase and Target Enzymes. Photosynthesis research 79, 233248. de Kraker, J.W., Luck, K., Textor, S., Tokuhisa, J.G., and Gershenzon, J. (2007). Two Arabidopsis genes ( IPMS1 and IPMS2 ) encode isopropylmalate synthase, the branchpoint step in the biosynthesis of leucine. Plant physiology 143, 970986. Dombrecht, B., Xue, G.P., Sprague, S.J., Kirkegaard, J.A., Ross, J.J., Reid, J.B., Fitt, G.P., Sewelam, N., Schenk, P.M., Manners, J.M., and Kazan, K. (2007). MYC2 differentially modulates diverse jasmonatedependent funct ions in Arabidopsis. The Plant cell 19, 22252245. Drevland, R.M., Waheed, A., and Graham, D.E. (2007). Enzymology and evolution of the pyruvate pathway to 2oxobutyrate in Methanocaldococcus jannaschii Journal of bacteriology 189, 43914400.
152 Ebel, C., Mariconti, L., and Gruissem, W. (2004). Plant retinoblastoma homologues control nuclear proliferation in the female gametophyte. Nature 429, 776780. Ellerstrom, M., Josefsson, L.G., Rask, L., and Ronne, H. (1992). Cloning of a cDNA for rape chloroplast 3isopropylmalate dehydrogenase by genetic complementation in yeast. Plant molecular biology 18, 557566. Ellis EA. (2006). Solutions to the problem of substitution of ERL 4221 for vinyl cyclohexene dioxide in Spurr low viscosity embedding formulations. Mi crosc Today 14: 32. Fahey, J.W., Zalcmann, A.T., and Talalay, P. (2001). The chemical diversity and distribution of glucosinolates and isothiocyanates among plants. Phytochemistry 56, 5 51. Falk, K.L., Vogel, C., Textor, S., Bartram, S., Hick, A., Pickett, J.A., and Gershenzon, J. (2004). Glucosinolate biosynthesis: demonstration and characterization of the condensing enzyme of the chain elongation cycle in Eruca sativa Phytochemistry 65, 10731084. Feldmann, K.A., Coury, D.A., and Christianson, M.L. (1997). Exceptional segregation of a selectable marker (KanR) in Arabidopsis identifies genes important for gametophytic growth and development. Genetics 147, 14111422. Field, B., Cardon, G., Traka, M., Botterman, J., Vancanneyt, G., and Mithen, R. (2004). Glucosinolate and amino acid biosynthesis in Arabidopsis. Plant physiology 135, 828839. GeuFlores, F., Nielsen, M.T., Nafisi, M., Moldrup, M.E., Olsen, C.E., Motawia, M.S., and Halkier, B.A. (2009). Glucosinolate engineering identifies a gammaglutamyl peptidase. Nature chemical biology 5, 575577. Gigolashvili, T., Yatusevich, R., Berger, B., Muller, C., and Flugge, U.I. (2007). The R2R3 MYB transcription factor HAG1/MYB28 is a regulator of methioninederived glucosinolate biosynthesis in Arabidopsis t haliana. Plant J 51, 247261. Gigolashvili, T., Berger, B., Mock, H.P., Muller, C., Weisshaar, B., and Flugge, U.I. (2007). The transcription factor HIG1/MYB51 regulates indolic glucosinolate biosynthesis in Arabidopsis thaliana. Plant J 50, 886901. Gig olashvili, T., Engqvist, M., Yatusevich, R., Muller, C., and Flugge, U.I. (2008). HAG2/MYB76 and HAG3/MYB29 exert a specific and coordinated control on the regulation of aliphatic glucosinolate biosynthesis in Arabidopsis thaliana. The New phytologist 177, 627642.
153 Gigolashvili, T., Yatusevich, R., Rollwitz, I., Humphry, M., Gershenzon, J., and Flugge, U.I. (2009). The plastidic bile acid transporter 5 is required for the biosynthesis of methioninederived glucosinolates in Arabidopsis thaliana. The Plant cell 21, 18131829. Grubb, C.D., Zipp, B.J., LudwigMuller, J., Masuno, M.N., Molinski, T.F., and Abel, S. (2004). Arabidopsis glucosyltransferase UGT74B1 functions in glucosinolate biosynthesis and auxin homeostasis. Plant J 40, 893908. Grubb, C.D., and Abel, S. (2006). Glucosinolate metabolism and its control. Trends in plant science 11, 89100. Gusti, A., Baumberger, N., Nowack, M., Pusch, S., Eisler, H., Potuschak, T., De Veylder, L., Schnittger, A., and Genschik, P. (2009). The Arabidopsis thaliana F box protein FBL17 is essential for progression through the second mitosis during pollen development. PloS one 4, e4780. Halkier, B.A., and Gershenzon, J. (2006). Biology and biochemistry of glucosinolates. Annual review of plant biology 57, 303333. H ansen, C.H., Wittstock, U., Olsen, C.E., Hick, A.J., Pickett, J.A., and Halkier, B.A. (2001). Cytochrome p450 CYP79F1 from arabidopsis catalyzes the conversion of dihomomethionine and trihomomethionine to the corresponding aldoximes in the biosynthesis of aliphatic glucosinolates. The Journal of biological chemistry 276, 1107811085. Hansen, B.G., Kliebenstein, D.J., and Halkier, B.A. (2007). Identification of a flavinmonooxygenase as the S oxygenating enzyme in aliphatic glucosinolate biosynthesis in Arabidopsis. Plant J 50, 902910. Hansen, B.G., Kerwin, R.E., Ober, J.A., Lambrix, V.M., MitchellOlds, T., Gershenzon, J., Halkier, B.A., and Kliebenstein, D.J. (2008). A novel 2oxoaciddependent dioxygenase involved in the formation of the goiterogenic 2hydroxybut 3 enyl glucosinolate and generalist insect resistance in Arabidopsis. Plant physiology 148, 20962108. Haughn, G.W., Davin, L., Giblin, M., and Underhill, E.W. (1991). Biochemical Genetics of Plant Secondary Metabolites in Arabidopsis thaliana: The Glucosinolates. Plant physiology 97, 217226. Haussuhl, K., Andersson, B., and Adamska, I. (2001). A chloroplast DegP2 protease performs the primary cleavage of the photodamaged D1 protein in plant photosystem II. The EMBO journal 20, 713722. He, Y., Mawhinney, T.P., Preuss, M.L., Schroeder, A.C., Chen, B., Abraham, L., Jez, J.M., and Chen, S. (2009). A redox active isopropylmalate dehydrogenase functions in the biosynthesis of glucosinolates and leucine in Arabidopsis. Plant J 60, 679690.
154 Hecht, S.S. (2000). Inhibition of carcinogenesis by isothiocyanates. Drug metabolism reviews 32, 395411. Hicks, L.M., Cahoon, R.E., Bonner, E.R., Rivard, R.S., Sheffield, J., and Jez, J.M. (2007). Thiol based regulation of redox active glutamatecysteine ligas e from Arabidopsis thaliana. The Plant cell 19, 26532661. Hirai, M.Y., Yano, M., Goodenowe, D.B., Kanaya, S., Kimura, T., Awazuhara, M., Arita, M., Fujiwara, T., and Saito, K. (2004). Integration of transcriptomics and metabolomics for understanding of global responses to nutritional stresses in Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA 101, 1020510210. Hirai, M.Y., Klein, M., Fujikawa, Y., Yano, M., Goodenowe, D.B., Yamazaki, Y., Kanaya, S., Nakamura, Y., Kitayama, M., Suzuki, H., Sakurai, N., S hibata, D., Tokuhisa, J., Reichelt, M., Gershenzon, J., Papenbrock, J., and Saito, K. (2005). Elucidation of geneto gene and metaboliteto gene networks in arabidopsis by integration of metabolomics and transcriptomics. The Journal of biological chemistry 280, 25590 25595. Hirai, M.Y., Sugiyama, K., Sawada, Y., Tohge, T., Obayashi, T., Suzuki, A., Araki, R., Sakurai, N., Suzuki, H., Aoki, K., Goda, H., Nishizawa, O.I., Shibata, D., and Saito, K. (2007). Omics based identification of Arabidopsis Myb tra nscription factors regulating aliphatic glucosinolate biosynthesis. Proc. Natl. Acad. Sci. USA 104, 64786483. Hopkins, R.J., van Dam, N.M., and van Loon, J.J. (2009). Role of glucosinolates in insect plant relationships and multitrophic interactions. Annual review of entomology 54, 5783. Howden, R., Park, S.K., Moore, J.M., Orme, J., Grossniklaus, U., and Twell, D. (1998). Selection of TDNAtagged male and female gametophytic mutants by segregation distortion in Arabidopsis. Genetics 149, 621631. Hsu Y.P., and Kohlhaw, G.B. (1980). Leucine biosynthesis in Saccharomyces cerevisiae. Purification and characterization of betaisopropylmalate dehydrogenase. The Journal of biological chemistry 255, 72557260. Hull, A.K., Vij, R., and Celenza, J.L. (2000). Arabidopsis cytochrome P450s that catalyze the first step of tryptophandependent indole3 acetic acid biosynthesis. Proc. Natl. Acad. Sci. USA 97, 23792384. Ikegami, A., Yoshimura, N., Motohashi, K., Takahashi, S., Romano, P.G., Hisabori, T., Takamiya, K., and Masuda, T. (2007). The CHLI1 subunit of Arabidopsis thaliana magnesium chelatase is a target protein of the chloroplast thioredoxin. The Journal of biological chemistry 282, 1928219291.
155 Jackson, S.D., Sonnewald, U., and Willmitzer, L. (1993). Cl oning and expression analysis of betaisopropylmalate dehydrogenase from potato. Mol Gen Genet 236, 309314. Jefferson, R.A., Kavanagh, T.A., and Bevan, M.W. (1987). GUS fusions: betaglucuronidase as a sensitive and versatile gene fusion marker in higher plants. The EMBO journal 6, 3901 3907. Jez, J.M., Cahoon, R.E., and Chen, S. (2004). Arabidopsis thaliana glutamatecysteine ligase: functional properties, kinetic mechanism, and regulation of activity. The Journal of biological chemistry 279, 3346333470. JohnsonBrousseau, S.A., and McCormick, S. (2004). A compendium of methods useful for characterizing Arabidopsis pollen mutants and gametophytically expressed genes. Plant J 39, 761775. Joshi, V., Laubengayer, K.M., Schauer, N., Fernie, A.R., and Jander, G. (2006). Two Arabidopsis threonine aldolases are nonredundant and compete with threonine deaminase for a common substrate pool. The Plant cell 18, 35643575. Joshi, V., and Jander, G. (2009). Arabidopsis methionine gammalyase is regulated according to isoleucine biosynthesis needs but plays a subordinate role to threonine deaminase. Plant physiology 151, 367378. Kang, B.H., Rancour, D.M., and Bednarek, S.Y. (2003). The dynaminlike protein ADL1C is essential for plasma membrane maintenance during pollen maturation. Plant J 35, 1 15. Kawaguchi, H., Inagaki, K., Matsunami, H., Nakayama, Y., Tano, T., and Tanaka, H. (2000). Purification and characterization of 3isopropylmalate dehydrogenase from Thiobacillus thiooxidans Journal of bioscience and bioengineering 90, 459461. Kim, J.H., and Jander, G. (2007). Myzus persicae (green peach aphid) feeding on Arabidopsis induces the formation of a deterrent indole glucosinolate. Plant J 49, 10081019. Kinsman, E.A., and Pyke, K.A. (1998). Bundle sheath c ells and cell specific plastid development in Arabidopsis leaves. Development 125, 18151822. Kliebenstein, D.J., Kroymann, J., Brown, P., Figuth, A., Pedersen, D., Gershenzon, J., and Mitchell Olds, T. (2001). Genetic control of natural variation in Arabidopsis glucosinolate accumulation. Plant physiology 126, 811825. Kliebenstein, D.J., Lambrix, V.M., Reichelt, M., Gershenzon, J., and Mitchell Olds, T. (2001). Gene duplication in the diversification of secondary metabolism: tandem 2 oxoglutaratedepend ent dioxygenases control glucosinolate biosynthesis in Arabidopsis. The Plant cell 13, 681693.
156 Kliebenstein, D.J., Figuth, A., and Mitchell Olds, T. (2002). Genetic architecture of plastic methyl jasmonate responses in Arabidopsis thaliana. Genetics 161, 16851696. Kliebenstein, D.J., Kroymann, J., and Mitchell Olds, T. (2005). The glucosinolatemyrosinase system in an ecological and evolutionary context. Current opinion in plant biology 8, 264271. Kliebenstein, D.J., D'Auria, J.C., Behere, A.S., Kim, J.H., Gunderson, K.L., Breen, J.N., Lee, G., Gershenzon, J., Last, R.L., and Jander, G. (2007). Characterization of seedspecific benzoyloxyglucosinolate mutations in Arabidopsis thaliana. Plant J 51, 10621076. Knill, T., Schuster, J., Reichelt, M., Ger shenzon, J., and Binder, S. (2008). Arabidopsis branched chain aminotransferase 3 functions in both amino acid and glucosinolate biosynthesis. Plant physiology 146, 10281039. Knill, T., Reichelt, M., Paetz, C., Gershenzon, J., and Binder, S. (2009). Arab idopsis thaliana encodes a bacterial type heterodimeric isopropylmalate isomerase involved in both Leu biosynthesis and the Met chain elongation pathway of glucosinolate formation. Plant molecular biology 71, 227239. Koroleva, O.A., Davies, A., Deeken, R ., Thorpe, M.R., Tomos, A.D., and Hedrich, R. (2000). Identification of a new glucosinolaterich cell type in Arabidopsis flower stalk. Plant physiology 124, 599608. Kroymann, J., Textor, S., Tokuhisa, J.G., Falk, K.L., Bartram, S., Gershenzon, J., and M itchellOlds, T. (2001). A gene controlling variation in Arabidopsis glucosinolate composition is part of the methionine chain elongation pathway. Plant physiology 127, 10771088. Kuang, A., and Musgrave, M.E. (1996). Dynamics of vegetative cytoplasm duri ng generative cell formation and pollen maturation in Arabidopsis thaliana. Protoplasma 194, 8190. Kushad, M.M., Brown, A.F., Kurilich, A.C., Juvik, J.A., Klein, B.P., Wallig, M.A., and Jeffery, E.H. (1999). Variation of glucosinolates in vegetable crops of Brassica oleracea. Journal of agricultural and food chemistry 47, 15411548. Lambrix, V., Reichelt, M., MitchellOlds, T., Kliebenstein, D.J., and Gershenzon, J. (2001). The Arabidopsis epithiospecifier protein promotes the hydrolysis of glucosinolates to nitriles and influences Trichoplusia ni herbivory. The Plant cell 13, 27932807. Lemaire, S.D., Michelet, L., Zaffagnini, M., Massot, V., and IssakidisBourguet, E. (2007). Thioredoxins in chloroplasts. Current genetics 51, 343365.
157 Leon, G., Holuigue, L., and Jordana, X. (2007). Mitochondrial complex II Is essential for gametophyte development in Arabidopsis. Plant physiology 143, 15341546. Levy, M., Wang, Q., Kaspi, R., Parrella, M.P., and Abel, S. (2005). Arabidopsis IQD1 a novel calmodulinbin ding nuclear protein, stimulates glucosinolate accumulation and plant defense. Plant J 43, 7996. Li, J., Hansen, B.G., Ober, J.A., Kliebenstein, D.J., and Halkier, B.A. (2008). Subclade of flavinmonooxygenases involved in aliphatic glucosinolate biosynt hesis. Plant physiology 148, 17211733. Li, N., Yuan, L., Liu, N., Shi, D., Li, X., Tang, Z., Liu, J., Sundaresan, V., and Yang, W.C. (2009). SLOW WALKER2 a NOC1/MAK21 homologue, is essential for coordinated cell cycle progression during female gametophy te development in Arabidopsis. Plant physiology 151, 14861497. Manfield, I.W., Jen, C.H., Pinney, J.W., Michalopoulos, I., Bradford, J.R., Gilmartin, P.M., and Westhead, D.R. (2006). Arabidopsis Coexpression Tool (ACT): web server tools for microarray based gene expression analysis. Nucleic acids research 34, W504509. Maruyama Nakashita, A., Inoue, E., WatanabeTakahashi, A., Yamaya, T., and Takahashi, H. (2003). Transcriptome profiling of sulfur responsive genes in Arabidopsis reveals global effects of sulfur nutrition on multiple metabolic pathways. Plant physiology 132, 597605. Maruyama Nakashita, A., Nakamura, Y., Tohge, T., Saito, K., and Takahashi, H. (2006). Arabidopsis SLIM1 is a central transcriptional regulator of plant sulfur response and metabolism. The Plant cell 18, 32353251. Massey, L.K., Sokatch, J.R., and Conrad, R.S. (1976). Branchedchain amino acid catabolism in bacteria. Bacteriological reviews 40, 4254. McCormick, S. (2004). Control of male gametophyte development. The Plant cell 16 Suppl, S142153. Mewis, I., Appel, H.M., Hom, A., Raina, R., and Schultz, J.C. (2005). Major signaling pathways modulate Arabidopsis glucosinolate accumulation and response to both phloem feeding and chewing insects. Plant physiology 138, 11491162. Mikkelsen, M.D., Hansen, C.H., Wittstock, U., and Halkier, B.A. (2000). Cytochrome P450 CYP79B2 from Arabidopsis catalyzes the conversion of tryptophan to indole3 acetaldoxime, a precursor of indole glucosinolates and indole3 acetic acid. The Journal of biological chemistry 275, 3371233717. Mikkelsen, M.D., Petersen, B.L., Glawischnig, E., Jensen, A.B., Andreasson, E., and Halkier, B.A. (2003). Modulation of CYP79 genes and glucosinolate profiles in Arabidopsis by defense signaling pathways. Plant physiology 131, 298308.
158 Mikkelsen, M.D., Naur, P., and Halkier, B.A. (2004). Arabidopsis mutants in the C S lyase of glucosinolate biosynthesis establish a critical role for indole 3 acetaldoxime in auxin homeostasis. Plant J 37, 770777. Naur, P., Peter sen, B.L., Mikkelsen, M.D., Bak, S., Rasmussen, H., Olsen, C.E., and Halkier, B.A. (2003). CYP83A1 and CYP83B1 two nonredundant cytochrome P450 enzymes metabolizing oximes in the biosynthesis of glucosinolates in Arabidopsis. Plant physiology 133, 6372. Niewiadomski, P., Knappe, S., Geimer, S., Fischer, K., Schulz, B., Unte, U.S., Rosso, M.G., Ache, P., Flugge, U.I., and Schneider, A. (2005). The Arabidopsis plastidic glucose 6phosphate/phosphate translocator GPT1 is essential for pollen maturation and embryo sac development. The Plant cell 17, 760775. Nozawa, A., Takano, J., Miwa, K., Nakagawa, Y., and Fujiwara, T. (2005). Cloning of cDNAs encoding isopropylmalate dehydrogenase from Arabidopsis thaliana and accumulation patterns of their transcripts. Bioscience, biotechnology, and biochemistry 69, 806810. Owen HA, Makaroff CA (1995). Ultrastructure of micro sporogenesis and microgametogenesis in Arabidopsis thaliana (L.) Heynh. ecotype Wassilewskija (Brassicaceae). Protoplasma 185: 7 21. Park, S.K., Howden, R., and Twell, D. (1998). The Arabidopsis thaliana gametophytic mutation gemini pollen1 disrupts microspore polarity, division asymmetry and pollen cell fate. Development 125, 37893799. Persson, S., Wei, H., Milne, J., Page, G.P., and Somerville, C.R. (2005). Identification of genes required for cellulose synthesis by regression analysis of public microarray data sets. Proc. Natl. Acad. Sci. USA 102, 86338638. Petersen, B.L., Andreasson, E., Bak, S., Agerbirk, N., and Halkier, B.A. (2001). Char acterization of transgenic Arabidopsis thaliana with metabolically engineered high levels of phydroxybenzylglucosinolate. Planta 212, 612618. Petersen, B.L., Chen, S., Hansen, C.H., Olsen, C.E., and Halkier, B.A. (2002). Composition and content of gluco sinolates in developing Arabidopsis thaliana Planta 214, 562571. Piotrowski, M., Schemenewitz, A., Lopukhina, A., Muller, A., Janowitz, T., Weiler, E.W., and Oecking, C. (2004). Desulfoglucosinolate sulfotransferases from Arabidopsis thaliana catalyze t he final step in the biosynthesis of the glucosinolate core structure. The Journal of biological chemistry 279, 5071750725. Porra, R.J. (2002). The chequered history of the development and use of simultaneous equations for the accurate determination of c hlorophylls a and b. Photosynthesis research 73, 149156.
159 Preuss, D., Rhee, S.Y., and Davis, R.W. (1994). Tetrad analysis possible in Arabidopsis with mutation of the QUARTET (QRT) genes. Science 264, 14581460. Ratzka, A., Vogel, H., Kliebenstein, D.J., Mitchell Olds, T., and Kroymann, J. (2002). Disarming the mustard oil bomb. Proc. Natl. Acad. Sci. USA 99, 1122311228. Reintanz, B., Lehnen, M., Reichelt, M., Gershenzon, J., Kowalczyk, M., Sandberg, G., Godde, M., Uhl, R., and Palme, K. (2001). Bus a bushy Arabidopsis CYP79F1 knockout mutant with abolished synthesis of short chain aliphatic glucosinolates. The Plant cell 13, 351367. Reichelt M, Brown PD, Schneider B, Oldham NJ, Stauber EJ, Tokuhisa J, Kliebenstein DJ, Mitchell Olds T, Gershenzon J (2 002) Benzoic acid glucosinolate esters and other glucosinolates from Arabidopsis thaliana. Phytochemistry 59, 663 671 Reymond, P., Bodenhausen, N., Van Poecke, R.M., Krishnamurthy, V., Dicke, M., and Farmer, E.E. (2004). A conserved transcript pattern in response to a specialist and a generalist herbivore. The Plant cell 16, 31323147. Robertson, W.R., Clark, K., Young, J.C., and Sussman, M.R. (2004). An Arabidopsis thaliana plasma membrane proton pump is essential for pollen development. Genetics 168, 16771687. Roth, M. (1971). Fluorescence reaction for amino acids. Analytical chemistry 43, 880882. Sarwar, G., and Botting, H.G. (1993). Evaluation of liquid chromatographic analysis of nutritionally important amino acids in food and physiological samples. Journal of chromatography 615, 1 22. Sawada, Y., Kuwahara, A., Nagano, M., Narisawa, T., Sakata, A., Saito, K., and Hirai, M.Y. (2009). Omics based approaches to methionine side chain elongation in Arabidopsis: characterization of the genes encoding methylthioalkylmalate isomerase and methylthioalkylmalate dehydrogenase. Plant & cell physiology 50, 11811190. Schauer, N., Zamir, D., and Fernie, A.R. (2005). Metabolic profiling of leaves and fruit of wild species tomato: a survey of the Solanum lycopers icum complex. Journal of experimental botany 56, 297307. Schlaeppi, K., Bodenhausen, N., Buchala, A., Mauch, F., and Reymond, P. (2008). The glutathionedeficient mutant pad21 accumulates lower amounts of glucosinolates and is more susceptible to the insect herbivore Spodoptera littoralis Plant J 55, 774786.
160 Schuster, J., Knill, T., Reichelt, M., Gershenzon, J., and Binder, S. (2006). Branchedchain aminotransferase4 is part of the chain elongation pathway in the biosynthesis of methioninederived glucosinolates in Arabidopsis. The Plant cell 18, 26642679. Shi, D.Q., Liu, J., Xiang, Y.H., Ye, D., Sundaresan, V., and Yang, W.C. (2005). SLOW WALKER1 essential for gametogenesis in Arabidopsis, encodes a WD40 protein involved in 18S ribosomal RNA biogenesis. The Plant cell 17, 23402354. Singh BK. (1999). Biosynthesis of valine, leucine, and isoleucine. In BK Singh, ed, Plant Amino Acids: Biochemistry and Biotechnology. Marcel Dekker, New York, pp 227247 Skirycz, A., Reichelt, M., Burow, M., Birkemeyer C., Rolcik, J., Kopka, J., Zanor, M.I., Gershenzon, J., Strnad, M., Szopa, J., Mueller Roeber, B., and Witt, I. (2006). DOF transcription factor AtDof1.1 (OBP2) is part of a regulatory network controlling glucosinolate biosynthesis in Arabidopsis. Plant J 47, 10 24. Small I. 2007. RNAi for revealing and engineering plant gene functions. Current Opinion in Biotechnology 18, 148 153 Smith, M.D., Schnell, D.J., Fitzpatrick, L., and Keegstra, K. (2003). In vitro analysis of chloroplast protein import. Current protocols in cell biology Chapter 11, Unit11 16. Smyth, D.R., Bowman, J.L., and Meyerowitz, E.M. (1990). Early flower development in Arabidopsis. The Plant cell 2, 755767. Sonderby, I.E., Hansen, B.G., Bjarnholt, N., Ticconi, C., Halkier, B.A., and K liebenstein, D.J. (2007). A systems biology approach identifies a R2R3 MYB gene subfamily with distinct and overlapping functions in regulation of aliphatic glucosinolates. PloS one 2, e1322. Stadler, R., Lauterbach, C., and Sauer, N. (2005). Cellto cell movement of green fluorescent protein reveals post phloem transport in the outer integument and identifies symplastic domains in Arabidopsis seeds and embryos. Plant physiology 139, 701712. Staswick, P.E., and Tiryaki, I. (2004). The oxylipin signal jas monic acid is activated by an enzyme that conjugates it to isoleucine in Arabidopsis. The Plant cell 16, 21172127. Strain, H.H., Cope, B.T., and Svec, W.A. (1971). Analytical procedures for the isolation, identification, estimation and investigation of t he chlorophylls. Methods Enzymol 23: 452476.
161 Streatfield, S.J., Weber, A., Kinsman, E.A., Hausler, R.E., Li, J., PostBeittenmiller, D., Kaiser, W.M., Pyke, K.A., Flugge, U.I., and Chory, J. (1999). The phosphoenolpyruvate/phosphate translocator is requir ed for phenolic metabolism, palisade cell development, and plastiddependent nuclear gene expression. The Plant cell 11, 16091622. Sureshkumar, S., Todesco, M., Schneeberger, K., Harilal, R., Balasubramanian, S., and Weigel, D. (2009). A genetic defect c aused by a triplet repeat expansion in Arabidopsis thaliana. Science 323, 10601063. Suzuki, G. (2009). Recent progress in plant reproduction research: the story of the male gametophyte through to successful fertilization. Plant & cell physiology 50, 18571864. Tantikanjana, T., Yong, J.W., Letham, D.S., Griffith, M., Hussain, M., Ljung, K., Sandberg, G., and Sundaresan, V. (2001). Control of axillary bud initiation and shoot architecture in Arabidopsis through the SUPERSHOOT gene. Genes & development 15, 15771588. Tantikanjana, T., Mikkelsen, M.D., Hussain, M., Halkier, B.A., and Sundaresan, V. (2004). Functional analysis of the tandem duplicated P450 genes SPS/BUS/CYP79F1 and CYP79F2 in glucosinolate biosynthesis and plant development by Ds transpositi on generated double mutants. Plant physiology 135, 840848. Teng, C., Dong, H., Shi, L., Deng, Y., Mu, J., Zhang, J., Yang, X., and Zuo, J. (2008). Serine palmitoyltransferase, a key enzyme for de novo synthesis of sphingolipids, is essential for male gam etophyte development in Arabidopsis. Plant physiology 146, 13221332. Textor, S., Bartram, S., Kroymann, J., Falk, K.L., Hick, A., Pickett, J.A., and Gershenzon, J. (2004). Biosynthesis of methioninederived glucosinolates in Arabidopsis thaliana: recombi nant expression and characterization of methylthioalkylmalate synthase, the condensing enzyme of the chainelongation cycle. Planta 218, 10261035. Textor, S., de Kraker, J.W., Hause, B., Gershenzon, J., and Tokuhisa, J.G. (2007). MAM3 catalyzes the forma tion of all aliphatic glucosinolate chain lengths in Arabidopsis. Plant physiology 144, 6071. Thilmony, R., Underwood, W., and He, S.Y. (2006). Genomewide transcriptional analysis of the Arabidopsis thaliana interaction with the plant pathogen Pseudomonas syringae pv. tomato DC3000 and the human pathogen Escherichia coli O157:H7. Plant J 46, 3453.
162 Underhill, E.W., Wetter, L.R. and Chisholm, M.D. ( 1973) Biosynthesis of glucosinolates. Bioch emical Society Symposium 38, 303 326. Voll, L., Hausler, R.E., Hecker, R., Weber, A., Weissenbock, G., Fiene, G., Waffenschmidt, S., and Flugge, U.I. (2003). The phenotype of the Arabidopsis cue1 mutant is not simply caused by a general restriction of the shikimate pathway. Plant J 36, 301317. Wallon, G., Kryger, G., Lovett, S.T., Oshima, T., Ringe, D., and Petsko, G.A. (1997). Crystal structures of Escherichia coli and Salmonella typhimurium 3 isopropylmalate dehydrogenase and comparison with their thermophilic counterpart from Thermus thermophilus Journal of molecular biology 266, 10161031. Wang, T., Iyer, L.M., Pancholy, R., Shi, X. and Hall, T.C. (2005) Assessment of penetrance and expressivity of RNAi mediated silencing of the Arabidopsis phytoene desaturase gene. New Phytol. 167, 751 760. Weigel, M., Varotto C., Pesaresi, P., Finazzi, G., Rappaport, F., Salamini, F., and Leister, D. (2003). Plastocyanin is indispensable for photosynthetic electron flow in Arabidopsis thaliana. The Journal of biological chemistry 278, 3128631289. Wetter, L.R., and Chisholm, M.D. (1968). Sources of sulfur in the thioglucosides of various higher plants. Canadian journal of biochemistry 46, 931935. Wittstock, U., and Halkier, B.A. (2002). Glucosinolate research in the Arabidopsis era. Trends in plant science 7, 263270. Xu, H., Zhang, Y., Guo, X., Ren, S., Staempfli, A.A., Chiao, J., Jiang, W., and Zhao, G. (2004). Isoleucine biosynthesis in Leptospira interrogans serotype lai strain 56601 proceeds via a threonineindependent pathway. Journal of bacteriology 186, 54005409. Yan, X., and Chen, S. (2007). Regulation of plant glucosinolate metabolism. Planta 226, 13431352. Yasutake, Y., Yao, M., Sakai, N., Kirita, T., and Tanaka, I. (2004). Crystal structure of the Pyrococcus horikoshii isopropylmalate isomerase small subuni t provides insight into the dual substrate specificity of the enzyme. Journal of molecular biology 344, 325333. Zhang, C., Guinel, F.C., and Moffatt, B.A. (2002). A comparative ultrastructural study of pollen development in Arabidopsis thaliana ecotype C olumbia and malesterile mutant apt13 Protoplasma 219, 5971.
163 Zhao, Y., Hull, A.K., Gupta, N.R., Goss, K.A., Alonso, J., Ecker, J.R., Normanly, J., Chory, J., and Celenza, J.L. (2002). Trpdependent auxin biosynthesis in Arabidopsis: involvement of cytochrome P450s CYP79B2 and CYP79B3 Genes & development 16, 31003112. Zybailov, B., Rutschow, H., Friso, G., Rudella, A., Emanuelsson, O., Sun, Q., and van Wijk, K.J. (2008). Sorting signals, N terminal modifications and abundance of the chloroplast proteome. PloS one 3, e1994.
164 BIOGRAPHICAL SKETCH Yan He was born in 1978, in Tieling, Laioning Province, China. He received a B.S. in a gronomy (2001) from Shenyang Agricultural University. After c ompleting his undergraduate studies he continued to purs ue his first master s degree at the National Maize Improvement Center of China Agricultural University (CAU), one o f the best agricultural universities in China After he obtained his first masters degree from CAU (2004), he joined a cooperative program between CAU and Missouri State University (MSU) where he got his second master s degree at the Division of Plant Science of MSU (2006). After completion of master studies he was enrolled in the Plant Molecular and Cellular Biology Program at the Universit y of Florida to pursue his Ph.D. degree under the supervision of Dr. Sixue Chen.