UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Localized post-transcriptional gene silencing of the very long chain fatty acid condensing enzyme, CER6,… Houlahan, Nora Kathleen 2004

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_2004-0491.pdf [ 15.19MB ]
JSON: 831-1.0091737.json
JSON-LD: 831-1.0091737-ld.json
RDF/XML (Pretty): 831-1.0091737-rdf.xml
RDF/JSON: 831-1.0091737-rdf.json
Turtle: 831-1.0091737-turtle.txt
N-Triples: 831-1.0091737-rdf-ntriples.txt
Original Record: 831-1.0091737-source.json
Full Text

Full Text

Localized post-transcriptional gene silencing of the very long chain fatty acid condensing enzyme, CER6, and analysis of transcript accumulation in the anthers of developing flower buds of Arabidopsis thaliana  by  Nora Kathleen Houlahan B.Sc. University of Guelph, 2000  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF ,  MASTER OF SCDSNCE In THE FACULTY OF GRADUATE STUDIES (Department of Botany)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA August 2004 © Nora Kathleen Houlahan, 2004  FACULTY OF GRADUATE STUDIES  THE UNIVERSITY OF BRITISH COLUMBIA  Library Authorization  In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  Nora Kathleen Houlahan Date (dcr/mm/yyyy)  Name of Author (please print)  Title of Thesis:  Localized post-transcriptional gene silencing of the very long chain fatty acid condensing enzyme, CER6, and analysis of transcript accumulation in the anthers of developing flower buds of Arabidopsis thaliana  Degree:  Master of Science  Department of  Botany  Year:  2004  The University of British Columbia Vancouver, BC Canada  grad.ubc.ca/forms/?formlD=THS  page 1 of 1  last updated: 4-Aug-04  11  Abstract  In Arabidopsis thaliana, both the stem wax and the pollen lipids are generated from very long chain fatty acid (VLCFA) precursors. A mutation in the gene encoding CER6, a condensing enzyme involved in VLCFA synthesis, results in plants with alterations in both the stem wax and the pollen coat lipid profiles. Phenotypically these plants lack epicuticular wax crystals on their stems and are conditionally male sterile. When grown under conditions of low humidity, the pollen of cer6 plants does not hydrate on a receptive stigma, although hydration will occur under high humidity growth conditions. This conditional male sterility can be exploited for use in breeding systems. Pollination of cer6 plants as the female parent will ensure no self-pollinated contaminants and the line can be easily propagated for future use through growth in high humidity. Alterations in stem wax profile are, however, not desirable as they may negatively influence the fitness of the plant. To address this problem, an RNAi construct was designed to induce localized silencing of the CER6 gene in the tissues where the pollen coat is produced. Transgenic plants expressing this silencing construct displayed a wildtype stem wax load and exhibited conditional male sterility. Analysis of RNA extracted from flower buds of the TI generation showed a decrease in CER6 expression at stage 12 of floral development in plants that were phenotypically conditionally male sterile. Analysis of the subsequent T2 and T3 generations revealed a reversion to a wildtype fertile phenotype in all of the initially silenced lines. Examination of CER6 transcript levels confirmed that this fertility was correlated to wildtype levels of expression at stage 12 of floral development. This loss of silencing was not the result of an aberrant  transgene transcript that was unable to initiate silencing, as no such transcript could be detected in any of the plants examined. Rather, methylation of the transgene promoter was observed, indicating the transcriptional silencing of transgene expression.  iv  Table of Contents Abstract Table of Contents List of Tables List of Figures List of Abbreviations  ii iv vi vii ix  ;  Chapter 1: Introduction 1.1 Floral organs and their reproductive structures 1.1.1 Overview of floral organs 1.1.2 The female reproductive structures 1.1.3 The male reproductive structures 1.1.4 The stages of anther development 1.1.5 Anther dehiscence 1.2 The tapetal cells of the anther 1.3 Pollen development 1.3.1 Stages of pollen development 1.3.2 The lipids of pollen grains 1.3.2a The lipids of pollen grains: the exine 1.3.2b The lipids of pollen grains: the pollen coat 1.4 Pollen adhesion, hydration, and germination 1.5 Wax biosynthesis 1.5.1 Fatty acid synthesis 1.5.2 Fatty acid elongation 1.5.3 Wax biosynthesis 1.5.4 eceriferium mutants 1.6 Male sterility 1.7 Post-transcriptional gene silencing \ 1.8 Thesis objectives  1 1 1 5 7 11 14 19 19 22 22 30 32 36 36 37 42 43 4  ?i  Chapter 2: Materials and Methods 2.1 Plasmid DNA preparation , DNA gel purification, DNA ligation, and DNA sequencing 2.2 Bacterial growth conditions 2.3 Nucleic acid analysis 2.3.1 Isolation of genomic DNA 2.3.2 Isolation of total RNA 2.3.3 RNA quantification and Reverse Transcriptase (RT) reaction 2.3.4 Optimization of PCR from cDNA template 2.4 Polymerase chain reaction (PCR) 2.4.1 Amplification of the CER6 3 'UTR 2.4.2 Amplification of the MS2 promoter 2.4.3 PCR from cDNA template 2.4.4 Amplification of aberrant transcripts in T2 and T3 generations  6  53 : 59  60 60 61 61 '. 61 62 62 63 63 63 64 65  2.5  2.6  2.4.5 Amplification to test for methylation in T2 and T3 generations Construct design 2.5.1 Generation of RNAi construct ; 2.5.2 Generation of MS2 promoter::GUS construct 2.5.3 Generation of vectors for plant transformation Plant growth conditions, transformation, and selection 2.6.1 Plant growth conditions 2.6.2 Floral dip 2.6.3 Screening for transgenic Arabidopsis in the TI and further generations 2.6.4 Segregation of antibiotic resistance in the T2 and T3 generations GUS histochemical assays Pollen microscopy :  2.7 2.8  65 66 66 68 68 69 69 70 70 71 71 72  Chapter 3: Results and Discussion - Analysis of TI Generation 3.1 Promoter::GUS confirms anther specificity of the MS2 promoter 73 3.2 Transgenic plants expressing the RNAi construct are conditionally male sterile 77 3.3 RT-PCR of CER6 transcript in wildtype buds 82 3.4 CER6 expression is reduced in conditionally male sterile plants 84 3.5 Summary , 89 Chapter 4: Results and Discussion - Analysis of T2 and T3 4.1 Segregation ratios fit 3:1 as expectedfroma single insertion of the transgene _-_ 4.2 Phenotypes of T2 plants - Fertility is restored 4.3 Phenotypes of T2 and T3 plants - Fertility is restored 4.3.1 Phenotypic analysis of transgenic plants 4.3.2 RT-PCR analysis reveals no aberrant transcripts in T2 and T3 plants 4.3.3 PCR analysis reveals methylation of transgene pdk intron ... 4.3.4 PCR analysis reveals methylation of transgene MS2 promoter sequence 4.4 Summary  90 92 93 93 100 101 103 105  Chapter 5: Conclusions and Future Directions 5.1 Analysis of silencing using GFP fluorescence 5.1.1 Generation of GFP:C£i?r5+3'UTR construct 5.1.2 Plant transformation, selection, and phenotypic analysis .... 5.1.3 Future strategies 5.2 Conclusion  107 107 109 110 Ill  References  113  vi  List of Tables Chapter 1: Introduction Table 1.1: Summary of the stages of flower development Table 1.2: Correlation of the stages of floral, anther, and pollen development Table 1.3: Male sterile mutants of Arabidopsis thaliana Chapter 4: Results and Discussion - Analysis of T2 and T3 Table 4.1: Segregation analysis of T2 seedlings Table 4.2: Segregation analysis of T3 seedlings  2 18 49  91 95  Vll  List of Figures Chapter 1: Introduction Figure 1.1: Floral diagrams of the Arabidopsis thaliana flower Figure 1.2: Stages of anther development Figure 1.3: Correlation of tapetal and pollen development . Figure 1.4: The structure of the pollen exine wall Figure 1.5: Summary oide novo fatty acid synthesis Figure 1.6: Summary of fatty acid elongation Figure 1.7: Current model for PTGS  2 6 17 26 , 38 41 56  Chapter 2: Material and Methods Figure 2.1: Cloning strategy for generation of the CER6 3 'UTR RNAi construct  67  Chapter 3: Results and Discussion - Analysis of TI Figure 3.1: MS2 promoter is not active in stems or cauline leaves 73 Figure 3.2: lOx brightfield magnification of proM&2::GUS 75 Figure 3.3: 20x brightfield magnification of proM52::GUS 76 Figure 3.4: RNAi plant #7 as representative of stem waxiness and male sterility 78 Figure 3.5: Siliques of RNAi#72-80 when grown at low humidity 80 Figure 3.6: Examples of TI plants after growth in high humidity 82 Figure 3.7: RT-PCR analysis of CER6 expression in wildtype buds . 83 Figure 3.8: RT-PCR analysis of CER6 expression in selected TI plants 85 Figure 3.9: RT-PCR analysis of CER6 expression in selected TI plants 87 Figure 3.10: Microscopic analysis of wildtype Columbia and transgenic TI plants expressing the CER6 silencing transgene 88 Chapter 4: Results and Discussion - Analysis of T2 and T3 Figure 4.1: Fertile pheontypes of selected plants in the T2 generation .... Figure 4.2: RT-PCR analysis of wildtype, TI, T2, and T3 representative plants from self-pollination of RNAi#30 Figure 4.3: Phenotypes of selected plants from the T3 generation Figure 4.4: RT-PCR analysis of CER6 transcript level in wildtype, RNAi#16-3-4 AND #16-3-5 (T3 representatives) at stages 11 and 12 of floral development Figure 4.5: RT-PCR analysis of CER6 transcript levels of wildtype, RNAi#19-3-l and #19-3-3 (T3 representatives) at stages 11 and 12 of floral development Figure 4.6: PCR analysis of methylation of the pdk intron present in the transgene insertion Figure 4.7: PCR analysis of methylation of the MS2 promoter in the transgene insertion Figure 4.8: PCR analysis of methylation of the MS2 promoter in the transgene insertion  92 94 96  98  99 102 104 105  viii Chapter 5: Conclusions and Future Directions Figure 5.1: Cloning strategy used for generation of the proCER6::GFP:CERG±3'UTR construct Figure 5.2: Fertility is restored in plants containing the cassette with the 3'UTR in the forward orientation when grown under normal humidity  108  110  IX  List of Abbreviations  ACCase ACP bp/kbp C4H CAD CaMV CCR cDNA Cer CHS CoA DNA EDTA EST FAE FAR FAS GFP GUS JA KAS KCS LB MCAT mL MMC MSUD nt  ocs PAL PCR pdk PTGS RdRp RNA RNAi RT-PCR SDS  acetyl-CoA carboxylase acyl carrier protein base pairs/kilobase pairs Cinnamate-4-hydroxylase Cinnamyl alcohol dehydrogenase cauliflower mosaic virus 35S promoter Cinnamoyl CoA reductase complementary DNA Eceriferum Chalcone Synthase CoenzymeA Deoxyribonucleic acid Ethylene diamine tetra acetic acid Expressed sequence tag Fatty acid elongation fatty-acyl CoA reductase Fatty acid synthesis Greenfluorescentprotein P-glucuronidase Jasmonic acid P-ketoacyl synthase P-ketoacyl CoA synthase Luria broth Malonyl-CoA-ACP transacylase Millilitres Microspore mother cells Meiotic silencing of unpaired DNA nucleotide Octopine synthase terminator derivedfromAgrobacterium tumefaciens Phenylalanine ammonia-lyase Polymerase chain reaction pyruvate orthophosphate dikinasefromFlaveria trinervia Post-transcriptional gene silencing RNA-dependent RNA polymerase Ribonucleic acid RNA interference Reverse transcriptase- polymerase chain reaction Sodium dodecyl sulphate  X  SEM siRNA TI T2 T3 TAG TE TEM TGS Tm Tris UTR UV VLCFA w/v WS wt X-gluc  Scanning electron microscope Small interfering RNA First transgenic generation Second transgenic generation Third transgenic generation Triacylglycerol Tris-EDTA Transmission electron microscope Transcriptional gene silencing Melting temperature Tri s(hydrozymethyl)- aminomethane Untranslated region Ultraviolet Very long chain fatty acid weight per volume Wax synthase Wildtype 5-Bromo-4-chloro-3-indolyl b-D-glucuronide cyclohexylamine salt  1  Chapter 1: Introduction 1.1 Floral organs and their reproductive structures 1.1.1 Overview of floral organs The development and structure of the flowers of Arabidopsis thaliana are characteristic of plants in the Brassicaceae. In 1990, Smyth et al. published a comprehensive analysis of the morphological characteristics of the developing Arabidopsis inflorescence. As the control of floral organ identity is a complex issue that will not be discussed in this document, a summary of the key events and timelines that occur after the determination of floral organ identity is summarized in Table 1.1 (Smyth etal, 1990). Arabidopsis flowers consist of four whorls of organs. From the outermost inward, these whorls consist of sepals, petals, stamens and carpels (Figure 1.1). Arabidopsis flowers are radially symmetrical (Irish and Yamamoto, 1995). In wildtype Arabidopsis, there are four of each of the sepals and petals, six stamens, and two fused carpels. The stamens are the male reproductive organs, whereas the carpels are the female reproductive organs. 1.1.2 The female reproductive structures The carpel whorl consists of two fused carpels, and forms the stigma, style and ovary (Figure 1.1). It is in the ovary that megagametogenesis occurs to produce the ovules, which are later fertilized by the pollen. The location of the ovary can be described as either superior or inferior depending on its location relative to the sepal whorl. In flowers such as Arabidopsis thaliana, the ovary is located atop the sepal whorl and is therefore classified as superior. When the ovary is located below the sepal whorl it  2  is considered to be inferior. The style is the region above the ovary, through which the pollen tubes must grow to reach the ovules. Finally the stigma is the region atop the style that provides the initial contact with the pollen grain. This initial interaction site provides the opportunity for the plant to recognize and eliminate any incompatible pollen. In plants that are not able to self-fertilize, this recognition is known as the selfincompatibility (SI) response. This initial contact, and the interaction between the pollen and the stigma that follows, are of utmost importance in the successful production of a viable zygote. There are two important stigma characteristics that will influence this response: morphological structure of the stigmatic surface and the nature of the stigmatic exudate (Heslop-Harrison and Shivanna, 1977). These two characteristics are often used in the taxonomic classification of plants (Heslop-Harrison and Shivanna, 1977).  Stigma Style Ovary  Figure 1.1: Floral diagrams of the Arabidopsis thaliana flower showing the four whorls in aerial and lateral view. T h e diagram on t h e left (aerial view) shows t h e number of organs in each whorl w h e r e a s the diagram o n the right (lateral view) show the organs that make up the carpel whorl.  3 Table 1.1: Summary of the stages of flower development from Smyth et al:, 1990 focusing on the stages where organ differentiation is complete. Duration and age of flower are estimated to the nearest 6 hours. At stage 13, duration and age of flower are reset to zero and all subsequent events are measured with reference to bud opening.  6 7 8 9 10 11 12  Sepals enclose bud Long stamen primordia stalked at base Locules appear in long stamens Petal primordia stalked at base Petals level with short stamens Stigmatic papillae appear Petals level with long stamens  30 24 24 60 12 30 42  Age of Flower at End of Stage (days) 5.25 6.25 7.25 9.75 10.25 11.5 13.25  13 14 15 16 17 18 19 20  Bud opens, petals visible, anthesis Long anthers extend above stigma Stigma extends above long anthers Petals and sepals withering All organs fall from green siliques Siliques turn yellow Valves separatefromdry siliques Seeds fall  6 18 24 12 192 36 Up to 24  0.5 1 2 2.5 10.5 12 13  Stage  Landmark Event at Beginning of Stage  Duration (hrs)  Stigmatic cells are modified epidermal cells that serve as the point of contact between the male and female parent. Stigmas can be classified morphologically as having papillate or non-papillate surface cells (Heslop-Harrison and Shivanna, 1977). Stigmatic papillae can be unicellular or multicellular, but generally project above the surface of the stigma (Heslop-Harrison and Shivanna, 1977). In papillate stigmas such as Arabidopsis, the cells at the tip of the stigma are elongate, creating finger like projections. This morphological characteristic should increase the surface area available to trap pollen grains. It also allows for a physical separation of pollen grains, preventing  4  cross signalling between compatible and incompatible pollen when the signalling molecules are carried in the pollen coating (Heslop-Harrison and Shivanna, 1977). The second character to consider is the nature of the stigmatic exudate. There are two general types of stigmas: wet and dry. Wet stigmas, when receptive to pollen, are covered by a fluid (sometimes sticky) secretion (Heslop-Harrison and Shivanna, 1977; Zinkl and Preuss, 2000) that is an extracellular matrix composed of sugars, amino acids, proteins, phenolics, fatty acids, lipids and glycolipids (Sanders and Lord, 1992). Some of these components are believed to be signalling molecules necessary for pollen compatibility and germination. In the lily flower, for example, the wet stigmatic exudate is aqueous and the pollen grain is surrounded by a lipidic coating (Heslop-Harrison, 1968b), whereas in tobacco the wet stigmatic exudate is lipidic and there is no associated pollen coat (de Graaf et al, 2001). Typically, species with a wet stigma will also have bicellular pollen and a gametophytic SI response (Heslop-Harrison and Shivanna, 1977). The nature of the SI response is determined as a result of these exudates allowing the non-specific binding and germination of pollen (Heslop-Harrison and Shivanna, 1977). The self-incompatibility response will typically occur after pollen germination or during pollen tube growth. SI is therefore mediated by interactions between signals produced in the pollen tube under the control of the pollen (gametophytic) genome arid the stigma (Heslop-Harrison and Shivanna, 1977). In contrast, species such as Arabidopsis thaliana have dry stigmas, which are not covered with an exudate. They are instead coated with a thin proteinaceous layer known as the pellicle (Heslop-Harrison and Shivanna, 1977; Elleman et al, 1992; Zinkl and Preuss, 2000). The pellicle is highly hydrophobic and may contain (glyco)proteins,  5 carbohydrates, lipids and enzymes involved in pollen germination and pollen tube growth, such as cutinases, esterases, and phosphatases (de Graaf et al, 2001). Typically, species with dry stigmas are fertilized by tricellular pollen surrounded by a lipidic pollen coat. They exhibit a sporophytic SI response that occurs at the stigmatic surface (HeslopHarrison and Shivanna, 1977). In such cases, many of the signal molecules necessary for hydration, germination, and the self-incompatibility response are carried in the pollen coating (Zinkl and Preuss, 2000). The ubiquitous presence of lipids, either in the stigmatic exudate, or in the pollen coating suggests an essential nature in pollination. 1.1.3 The male reproductive structures Each stamen is composed of a long slender filament that supports a four-lobed anther. It is inside the anther that the pollen develops. The lobes of the anther consist of five tissue layers (Figure 1.2, stage 5) (Sanders et al, 1999; Owen and Makaroff, 1995): the microspore mother cells (MMC), the tapetum, the middle layer, the endothecium and the epidermis. Maturation of the anther is associated with the degeneration of several of these cell layers (Goldberg et al, 1993). The innermost layer is the reproductive tissue whereas the remaining four layers are non-reproductive (Goldberg et al, 1993). In addition, the anther also contains vascular tissue at the point of attachment to the filament, connective tissue, septum, and stomium (Figure 1.2) (Sanders et al, 1999). Both the septum and the stomium are highly specialized cells. The septum cells separate two lobes in a theca (Zhang et al, 2001), whereas the stomium are differentiated epidermal cells that are necessary for anther dehiscence (Goldberg et al, 1993).  6  Stage 5  Stage 6  Stage 12  Stage 11  Stage 13  Figure 1.2: Diagram representing the stages of anther d e v e l o p m e n t in Arabidopsis thaliana as classified by Goldberg et al. (1993) and corresponding bright field micrographs. Micrographs are f r o m Sanders et al. (1999) and are stained with toluidine blue. Diagram representations of tissues are indicated at stage five. Since the epidermal, endothecial, vascular, and connective tissues persist throughout anther development, they are labeled only once. Green colouring represents epidermal tissue and connective tissue, yellow represents the endothecium, light blue represents the middle layer, red represents the t a p e t u m , and dark blue represents the microspore mother cells ( M M C ) and free pollen. Tissue degeneration is represented by a disappearance and/or lightening of the corresponding section of the diagram. Diagram not to scale.  7 1.1.4 The stages of anther development Goldberg et al. (1993) divided anther development into two distinct phases: a histospecification phase where morphology is established and a cell degeneration/dehiscence phase where pollen grains differentiate and are released. Sanders et al. (1999) further subdivided these phases into a series of fourteen stages based on observed cytological changes within the anther. Selected stages of anther development are shown in Figure 1.2 and the key events in each stage are summarized in Table 1.2. The first phase of anther development encompasses stages 1-8 and spans cell differentiation from the stamen primordia to the release of free microspores into the anther locule (Sanders et al, 1999). Cell fates are determined by stage 5 (Sanders et al, 1999) when the microspore mother cell undergoes meiosis. Prior to this event, the MMCs are connected to each other, as well as to the tapetal cells, by plasmodesmata (Owen and Makaroff, 1995). These connections are lost as a callose (P-l,3-glucan) wall is synthesized by the MMCs and deposited between the pectic-cellulosic primary cell wall and the plasma membrane (Owen and Makaroff, 1995; Rhee and Somerville, 1998). While the MMCs are no longer connected to the tapetal cells, they retain communication with each other via cytoplasmic channels (Owen and Makaroff, 1995; McCormick, 1993). Because of these connections, meiotic events are amazingly synchronized within each lobe of the anther (McCormick, 1993). These connections are lost at the end of meiosis and each tetrad of microspores is completely enclosed in a callose wall (McCormick, 1993; Owen and Makaroff, 1995). As the anther develops, the tapetal cells expand and become vacuolated, essentially crushing the cells of the middle layer  8 (Sanders et al, 1999) although remnants may persist until the degeneration of the tapetum (Owen and Makaroff, 1995). This degeneration of the middle layer marks stage 6 of anther development (Sanders et al, 1999). In addition, a fluid filled space called the locule develops in each lobe; this space contains the microspore tetrads and is bordered by the tapetum. The single layer of tapetal cells provides many of the nutrients and building blocks for the developing pollen grains. Another essential function of the tapetum is the synthesis and secretion of cell wall degrading enzymes such as polygalacturonase (Rhee et al, 2003; Jenkins et al, 1999) and callase (P-l,3-glucanase) (Hird et al, 199'4; Bedinger, 1992) at the appropriate time of development. Enzymes of both types are required for wildtype pollen development and mutations in genes encoding such enzymes may lead to aberrant pollen development (Rhee et al, 2003) and potentially male sterility (Fei and Sawhney, 1999). The release of the free microspores denotes the end of both stage 8 and the histospecialization phase of anther development (Sanders et al, 1999; Goldberg et al, 1993). The second phase of anther development, the cell degeneration/dehiscence phase, involves radical changes in the remaining anther tissues (Sanders et al, 1999). In Arabidopsis these changes include the mitotic divisions of the microspores to generate tricellular pollen. As pollen development will be described in detail in a later section, the present discussion will focus on the non-reproductive tissues of the anther. In these nonreproductive tissues, degeneration of the tapetum, thickening of the cell walls of the endothecium, expansion of both the epidermis and the endothecium, as well as rupture of both the septum and the stomium occur (Sanders et al, 1999; Goldberg et al, 1993; Owen and Makaroff, 1995).  9 The tapetum is a highly specialized and fascinating layer of cells that is essential for pollen development (Mariani et al, 1992). There are two general classifications of the tapetum found in higher plants: the amoebic tapetum and the secretory tapetum (Fumess and Rudall, 2001; Shivanna, 2003). In both cases the tapetum synthesizes and secretes components that are required for the development of a mature pollen grain (Shivanna, 2003). In the amoeboid tapetum, the tapetal cells fragment and intersperse around the developing microspores. In this type, the locule isfilledwith microspores and tapetal cells, and there is direct contact between the two cell types (Shivanna, 2003). This tapetal morphology is more common in monocotyledonous species, and not often seen in dicotyledonous species (Furness and Rudall, 2001). In contrast, the tapetal cells of the secretory tapetum remain in a cell layer surrounding the locule (Heslop-Harrison, 1968a; Owen and Makaroff, 1995). Tapetal components are secreted into the locule to be taken up by the microspores. The tapetal cells are not necessarily in direct contact with the microspores and consequently the amount of compounds any given pollen grain receives may be influenced by its location within the locule and its distance from the tapetum (Scott, 1994). Arabidopsis has a secretory tapetum, as is characteristic of the Brassicaceae (Furness and Rudall, 2001). At the time of pollen meiosis, the tapetal cells have undergone a mitotic division without cytokinesis, resulting in binucleate cells (Owen and Makaroff, 1995). In addition, the primary cell wall begins to degenerate although the tapetal cells remain connected via plasmodesmata (Bedinger, 1992; Owen and Makaroff, 1995). At this stage, the tapetum is highly metabolically active, as evidenced by the stacks of rough endoplasmic reticulum at the site of callase secretion (Fei and Sawhney, 1999).  10 There have been a variety of genes isolated that show a tapetum-specific expression pattern in Nicotania tobaccum (Koltunow et al, 1990), Brassica napus (Hong et al, 1997), and in Arabidopsis (Hird et al, 1994; Paul et al, 1992; Aarts et al, 1997; Rubinelli et al, 1998; Ariizumi et al, 2002). Of these, ATA7 (Rubinelli et al, 1998) and LTP12 (Ariizumi et al, 2002) encode putative lipid transfer proteins which may have a role in transporting lipids from the tapetal cytoplasm to the locule. PGA4 (Ariizumi et al, 2002) and QRT3 (Rhee et al, 2003) encode putative polygalacturonases, enzymes required for the degradation of the primary cell wall of the MMC (Rhee and Somerville, 1998; Rhee et al, 2003). MS2 encodes a putative fatty acyl reductase involved in the wax biosynthetic pathway, possibly in the generation of sporopollenin precursors (Aarts etal, 1997). Tapetal cells remain metabolically active even as they begin to degrade (Owen and Makaroff, 1995). This programmed cell death commences at stage 11 of anther development, and is synchronized with pollen mitosis (Sanders et al, 1999). Degeneration is characterized by cell shrinkage, vacuolation (Wu and Cheung, 2000) as well as by the disappearance of the nuclei and the disruption of the ER and tapetal plastids (Owen and Makaroff, 1995). In Arabidopsis, the lipophillic pollen coat is derived entirelyfromtapetal constituents. The lipid bodies present in tapetal cells will be examined in the context of pollen development in a following section. Many cellular changes occur in the remaining tissues at stage 11 of anther development (Sanders et al, 1999). In addition to the degeneration of the tapetum, the cells of the endothecium begin to expand and develop characteristic lignified fibrous thickenings of their cell walls (Sanders et al, 1999; Owen and Makaroff, 1995; Keijzer,  11 1987a). These wall thickenings cover all sides of the cell that are not adjacent to the epidermis. They therefore have a U-shaped appearance when viewed in section (Keijzer, 1987a). The Arabidopsis Ms35 mutant plants lack these thickenings and consequently fail to dehisce. The integrity of the endothecial cells is not maintained, and although all other aspects of anther development appear similar to wildtype plants, the anthers fail to release their fertile pollen (Dawson et al, 1999). The septum is a group of highly specialized endothecial cells that lack the characteristic wall thickenings (Keijzer, 1987a; Sanders et al, 1999). They separate the two locules present in a theca. At stage 12 of anther development, an enzyme-mediated breakdown of the cell wall material causes the septum to separate. This separation joins the two locules and creates a bi-lobed anther (Keijzer, 1987a). Concurrently, the stomium differentiate from the epidermal cells (Sanders  al, 1999). This cell type is  essential for the proper release of pollen, as plants whose stomia have been destroyed by targeted cell ablasion do not undergo dehiscence. Instead of releasing their pollen, these plants retain viable pollen inside of the closed anther until flower senescence (Beals and Goldberg, 1997). An expansion of the epidermal and endothecial cells will cause an inward bending of the locule walls, causing a mechanical rupture of the septum (Keijzer, 1987a) and dehiscence of the anther (Sanders et al, 1999). 1.1.5 A n t h e r dehiscence  Jasmonic acid plays an essential role in the appropriate timing of anther dehiscence. A delayed dehiscence is observed in Arabidopsis plants lacking a functional copy of DED1. This gene encodes a 12-oxyphytodienoate reductase involved in the JA biosynthetic pathway (Sanders et al, 2000). This same effect is observed in Arabidopsis  12 mutants lacking linolenic acid (18:3), the precursor for JA biosynthesis. DAD1 encodes a phospholipase A l required for the release of linolenic acidfrommembrane phospholipids (Ishiguro et al, 2001) and FAD3, FAD7, and FAD8 are fatty-acyl desaturases involved in the conversion of linoleic acid (18:2) to linolenic acid (18:3) (McConn and Browse, 1996). The mutant plants dadl, dedl, and the fad3fad7fad8 triple mutant all exhibit shortened anther filaments, and delayed anther dehiscence as a result of a delay in stomium rupture. In wildtype anthers, dehiscence is preceded by a loss of waterfromthe endothececium, connective tissue and the locule. This water loss is correlated with a sudden increase infilamentheight. Indeed, expression of both DAD1 and DED1 genes could be localized to the filaments at this stage of development, indicating that JA may have a role in regulating water transport in the anther, drawing it out of the tissues of the anther and into thefilament(Ishiguro et al, 2001; Sanders et al, 2000). In addition, mutants lacking COI1, an enzyme involved in the JA signalling pathway, show a phenotype similar to the fad3fad7fad8 triple mutant, suggesting a requirement for JA in pollen development (Xie et al, 1998; Feys et al, 1994). Recently, evidencefromZea mays (Schreiber et al, 2004) and Nicotiana tobaccum (Cecchetti et al, 2004) point to a role for auxin in anther dehiscence. Cecchetti et al. (2004) used chiameric expression of the RolB (Agrobacterium oncogene that confers auxin sensitivity) transgene to generate tobacco with an increased sensitivity to auxin. These transgenic plants displayed shortened filaments and a late dehiscence due to a delay in the rupture of the stomium. Schreiber et al. (2004) describe a MADS box transcription factor^ ZmMADS2, whose expression is localized to the endothecium, connective tissue, and pollen in the late stages of floral development. Plants lacking a  13 functional copy of ZmMADS2 displayed an identical phenotype to that already described. Interestingly, the ZmMADS2 promoter contains an auxin response element and a water stress response element (Schreiber et al, 2004). It has been suggested that auxin delays dehiscence by repressing the enzymatic activity of the cellulases and pectinases that are responsible for the degradation of the degenerating cells (Cecchetti et al, 2004). Perhaps then auxin is produced in the earlier stages of anther development to repress the activity of cellulases and pectinases, but also activates MADS box transcription factors which, in turn, activate these cell wall degrading enzymes essential for anther dehiscence. Finally, Jenkins et al (1999) found that the promoter of a polygalacturonase enzyme fused to the GUS reporter gene showed GUS staining in the stomium cells of Brassica napus anthers. Transgenic plants containing this promoter fused to the cytotoxic barnase gene were shown to be sterile because of a lack of anther dehiscence. These results indicate that cell wall degrading enzymes such as polygalacturonase may indeed play a vital role in the dehiscence of anthers (Jenkins et al, 1999). The regulation of the second phase of anther development, the cell degeneration phase, is a complex process that is not fully understood. Its regulation may involve JA, auxin, and many enzymes expressed at precise times during development. The stages of Arabidopsis anther development are summarized in Table 1.2 and correlated to both the stages of floral development and the stages of pollen development. 1.2 T h e tapetal cells of the anther  In Arabidopsis, the lipophillic pollen coat is derived entirely from tapetal constituents (Evans et al, 1992). It is therefore important to examine the lipid bodies  present within the tapetum. Tapetal cells have three main lipid components that are involved in the development of pollen grains (Murphy, 2001). Elaioplasts are specialized plastids present only in tapetal cells. Orbicules (or Ubish bodies) are bodies that appear heterogeneous under examination with a transmission electron microscope (TEM). Finally, tapetosomes, or lipid bodies, are aggregates that are also released into the locule to contribute to the pollen coating. The accumulation of these organelles in the tapetum is temporally regulated to correspond to different stages of pollen development (Evans et al, 1992; Murphy, 2001; Wu et al, 1997). Tapetal cells contain a unique form of plastid, the elaioplast, that differentiates early during cell development (Piffanelli et al, 1998). Elaioplasts are non-pigmented plastids devoid of grana. They may, however, have a small number of membranous thylakoid-like inclusions (Owen and Makaroff, 1995). After pollen meiosis, these plastids begin to loose their starch granules and any thylakoid-like membranes disappear (Keijzer, 1987b). At the bicellular pollen stage, they can be seen under the TEM to be accumulating small lipid bodies (Hernandez-Pinzon et al, 1999). The number of these lipid-filled plastids increases immediately prior to the degeneration of the tapetum (Piffanelli and Murphy, 1998). In Brassica napus, the main lipidic components of these plastids are sterol esters and triacylglycerols (Costa et al, 2003; Hernandez-Pinzon et al, 1999; Murphy, 2001). Of these, the sterol esters have been identified in the Brassica napus pollen coat. Plastidial derived sterol esters are also found in the pollen coat of Arabidopsis, whereas TAGs are absent in the pollen coatings of both species (HernandezPinzon et al, 1999; Wu et al, 1997). The contents of the elaioplasts are released when  15 the tapetal cells degrade, and their contents contribute to the formation of the pollen coating. Orbicules are enigmatic cytoplasmic membrane bound organelles that are possibly derivedfromthe endoplasmic reticulum (Heslop-Harrison, 1968b). These organelles begin to accumulate as pro-orbicules during tetrad formation (Piffanelli and Murphy, 1998). In many species the orbicular membranes then become lined with a heterogeneously staining electron dense coating of what is believed to be sporopollenin precursors (Keijzer, 1987b; Piffanelli et al, 1998; Wang et al, 2003). Because of this pre-sporopollenin composition, one proposed function of orbicules is in the formation of the pollen exine, which* is composed mainly of sporopollenin (Keijzer, 1987b; Murphy, 2001). This hypothesis is supported by Wang et al (2003) who isolated and characterized a proteinfromrice and wheat, RAFTIN, that localized to both the orbicules in the tapetum and the exine of the microspore. Moreover, silencing the expression of the RAFTIN gene resulted in mutant plants that displayed collapsed pollen and a tapetum that failed to degenerate (Wang et al, 2003). Interestingly, the orbicules of some species, such as those of the lily, have been observed to localize to the tapetal membrane adjacent to the middle layer. The orbicules remain in this location even after the tapetum has degraded (Keijzer, 1987b). This localization is proposed to hold pollen in the locule after anther dehiscence until removed by a pollination vector (Keijzer, 1987b). Contrary to the role of orbicules in exine formation, these organelles lack the sporopollenin-like coating in members of the Brassicaceae and are conspicuously absent in Arabidopsis, which has a well-developed exine (Wang et al, 2003). Correspondingly, Wang et al. (2003) failed to identify a gene homologous to RAFTIN in Arabidopsis. These observations suggest that  16 orbicules have no role in exine formation in Arabidopsis, and possibly in other members of Brassicaceae. Despite much conjecture, the role of orbicules in pollen and anther development remains unclear. Lastly, tapetosomes are patches of electron-dense TAGs that appear heterogeneous with no clearly defined boundaries (Wu et al, 1997). They are large organelles, ranging in size from 1-5 um (Hesse, 1986; Murphy and Ross, 1998). While the TAGs that make up the major lipid component of the tapetum are not incorporated into the pollen coat, they may be taken up and metabolized by the developing pollen grain (Hernandez-Pinzon et al, 1999). Evidence for this hypothesis includes the observance of channels across the pollen secondary cell wall (the exine), connecting the locule and the pollen grain (Rowley et al, 2003). Such channels provide an excellent opportunity for the transport of TAGs from the tapetosomes to the pollen grain. Tapetosomes also appear to contain a fibrous, possibly proteinaceous, network (Owen and Makaroff, 1995; Murphy and Ross, 1998; Altman, 1999; Wu et al, 1997) associated with oleosin-like proteins similar to those found in seed lipid bodies (Murphy and Ross, 1998; Hernandez-Pinzon et al, 1999). In the locule, the N-terminal region is cleaved from the C-terminal oleosin region and the C-terminal peptide is found in the pollen coating (Murphy and Ross, 1998). Whereas it was previously believed that the contents of both the elaioplasts and the tapetosomes were deposited directly onto the pollen grain, evidence from the oleosin-like proteins and from lipid profiles indicates that modification of these constituents occurs in the locule (Murphy and Ross, 1998; Hernandez-Pinzon et al, 1999). Figure 1.3 outlines the contribution of the tapetal lipid constituents to the development of the pollen grain.  17  S $ ro E E - a> o o. Q> E 3 c o ro o ro  ^_  lat  ©  -i—'  CD o 0 Q. o CO o ro o O > co c o c —-  V)  CD CD  1 — 1 —  licr  J  £ o £ o c  "Z.  w  CO  C CD "O CD Q. C 3  $3  _  =3? 4 3 CD  CO -6 -  PS i  O _  =3 CO  18 Table 1.2: Correlation of the stages of floral development (Smyth et al, 1990), anther development (Goldberg et al, 1993), and pollen development (Regan and Moffatt, 1990) listing several key events at eac l stage.  Flower Anther Pollen stage stage stage 7  3  8  4  9  5  3  6  10  11-12  7  4  8  5  9  6-7  10  11  8  12  9 10  13-14  13  15-16  14  17  15  Key developmental events Long stamen primordia stalked at base Mitotic divisions of archesporial cells differentiate tissue types in the anther Locules appear in long stamens Vascular region initiated in anther Petal primordia stalked at base Locules established and microspore mother cells appear in the anther Callose wall deposited around MMCs MMCs enter meiosis, middle layer elongated, tapetum vacuolated Middle layer degenerating Meiosis completed Primexine deposited Petals level with short stamens Callose wall degenerates and microspores enter free microspore stage Anther growth and septum differentiation Large vacuole forms in cytoplasm of pollen (vacuolate microspore stage), nucleus displaced to periphery of cell Exine deposited and intine synthesis begins Stigmatic papillae appear Petals level with long stamens Tapetum degeneration begins Endothecial layer expands and lignified fibrous bands appear in endothecium and connective cells. Septum degeneration and stomium differentiation begin Vacuole reabsorption and first pollen mitosis occurs Second pollen mitosis occurs, storage bodies visible Septum breakage generates bilocular anthers Stomium fully differentiated Mature pollen is dessicated immediately preceding anther dehiscence Bud opens, petals visible Long anthers extend above stigma Rupture of stomium cells Dehiscence and pollen release Stigma extends above long anthers Petals and sepals withering Senescence and shrinkage of stamen tissues All organs fall from green siliques  19 1.3 Pollen development 1.3.1 Stages of pollen development Pollen development begins with tissue differentiation in the anther. Regan and Moffatt (1990) divided this process into ten reference stages. Table 1.2 correlates the stages of pollen development with those of anther and of floral development, listing several key events at each stage, and Figure 1.3 relates the stages of pollen development to tapetal development. The microspore mother cells (MMCs) differentiate from the L2 layer of archeosporial cells to form the reproductive tissue of the anther (Goldberg et al, 1993; Yang and Sundaresan, 2000). At this initial stage, the MMC has a cellulosic primary cell wall surrounding the plasma membrane. Callose is synthesized and secreted by the MMC to form a secondary wall that lies between the primary wall and the plasma membrane (Keijzer, 1987b). This callose layer both isolates the microspores from tapetal influences and fixes tetrad configuration (Barnes and Blackmore, 1986). The MMCs undergo meiosis to form a tetrad of haploid microspores; in Arabidopsis the gametophytes are held in a tetrahedral configuration (Barnes and Blackmore, 1986). Callose continues to be synthesized by the gametophytes and is deposited on the cell plates during meiotic cytokinesis to enclose each microspore in the tetrad (Hird et al, 1994). During this time, the microspores also synthesize and deposit a cellulosic primexine wall outside their plasma membrane (Owen and Makaroff, 1995; Scott, 1994; McCormick, 1993). The primexine serves as a scaffolding for the later deposition of the sporopollenin precursors that form the exine wall. The microspores are released from the tetrad when callase is secreted from the tapetum to digest the callose wall. Many reports suggest that callase is the key enzyme  required for this release. A thorough analysis of pollen mutants, however, indicated that callase alone is insufficient for this release. Preuss et al. (1994) identified two Arabidopsis mutants, qrtl and qrt2, whose pollen failed to enter thefreemicrospore stage. Digestion of the callose wall was unaffected and the pollen grains were fully fertile in their tetrahedral configuration (Rhee and Somerville, 1998). An identical phenotype was observed in the qrt3 mutant (Rhee and Somerville, 2003). Cloning revealed the QRT3 locus encoded an endopolygalacturonase essential for the degradation of the primary cell wall to the MMC. Therefore, synthesis and secretion of both callase and polygalacturonase are required for the release of microsporesfromthe tetrad. At the free microspore stage, small vacuoles begin to appear in the cytoplasm of the developing pollen grains. During this time, sporopollenin precursors are secreted from the tapetum and deposited on the primexine, where they are polymerized to form the exine pollen wall. This process is characterized by an increase in the thickness and the density of the microspore wall (Owen and Makaroff, 1995). The small vacuoles coalesce into one large vacuole that occupies much of the pollen grain, and the nucleus migrates to the periphery of the cell (Bedinger, 1992).  This vacuolate stage of pollen  development is associated with a large increase in size. The first mitotic division accompanies the disappearance of the large vacuole and the reappearance of many small vacuoles (Yamamoto et al, 2003). This division is an asymmetric division that results in the vegetative and the generative cells (McCormick, 1993). The vegetative cell is large and metabolically active. It is responsible for the synthesis of many of the compounds required for development of viable pollen as well as for the synthesis and deposition of the inner pollen wall, the intine (Twell, 1994). In contrast, the generative cell is small  21 and completely enclosed by the vegetative cell. It is also relatively metabolically inactive. At the bicellular pollen stage, exine deposition is essentially complete and intine synthesis has begun (Owen and Makaroff, 1995). The intine is a pecto-cellulosic layer that is deposited during the free microspore stage, at the time of the second pollen mitosis. The initial deposition occurs atop the pores destined to become apertures and then spreads to form a continuous layer between the plasma membrane and the exine (McCormick, 1993). The intine may also contain enzymes, such as cutinase, that are required for interactions with the stigma and may help to regulate water loss and absorption in the mature pollen grain (Dickinson et al, 2000). The generative cell will undergo one further mitotic division to produce the two sperm cells. One of these sperm cells will fertilize the two female polar nuclei to form the triploid endosperm while the other will fertilize the egg to form the diploid zygote (McCormick, 1993). In bicellular pollen, the microspores are released from the anther before this second mitotic division has occurred. Successful interaction with a female parent will result in pollen germination, tube growth, and the second mitotic division. In tricellular pollen, such as that of Arabidopsis, the second mitosis occurs inside the locule of the anther. During this time, the degeneration of the tapetum has released the lipidic and proteinaceous components of the pollen coating to form the pollen coat (Evans et al, 1992). Immediately preceding anther dehiscence is a period of significant water loss from the microspores and their surrounding tissues. This water loss is associated with an increase in filament and petal length, but also results in the desiccated state of mature Arabidopsis pollen. Because of this desiccation, it is imperative that the Arabidopsis <• pollen is able to rehydrate extremely quickly once it contacts a receptive stigma (Heslop-  22 Harrison and Heslop-Harrison, 1994; Shivanna, 2003). Among the structures integral to this rehydration are the various lipidic components of the pollen grain. 1.3.2 The lipids of pollen grains There are four main lipidic components in a mature pollen grain of Arabidopsis thaliana: the intracellular membrane lipids, lipidic storage bodies, the exine, and the pollen coat. Both the membrane lipids and the storage bodies are synthesized by the gametophyte, whereas the exine and the pollen coat are synthesized by the sporophytic tapetum (Piffanelli et al, 1997; Piffanelli et al, 1998). The pollen lipid storage bodies accumulate in the cytoplasm during the later stages of pollen development (Owen and Makaroff, 1995; Piffanelli et al, 1997). They are analogous to the storage oil bodies of the Arabidopsis seed (Piffanelli et al, 1998). These lipidic components contain mostly TAGs and are synthesized solely by the microspore itself (Piffanelli et al, 1997). These lipid storage bodies are believed to provide energy to the pollen grain during the metabolically demanding period of germination and pollen tube growth (Piffanelli et al, 1997). 1.3.2a The lipids of pollen grains: the exine The exine, or outer pollen wall, is a complex lipidic layer surrounding the outside of the pollen grain, that is characteristic of all pollen grains. It is made of sporopollenin, a remarkably stable, inert molecule that is resistant to acetolysis. Because of this lack of reactivity, it has proved difficult to elucidate the exact composition of sporopollenin (Scott, 1994). Although previous theories suggested that either long chain lipids or polymerized and esterified carotenoids were the main component of sporopollenin, the current belief is that sporopollenin is composed of a heterogeneous mixture of saturated  23 very long chain fatty acids and phenylpropanoids (Scott, 1994; Piffanelli et al, 1998). While the stability of sporopollenin proves an enigma to biological chemists, for this same reason it is indispensable to pollen grains. The sequencing of the Arabidopsis genome and the use of forward and reverse genetic techniques provide an alternate route by which to approach the enigma of sporopollenin composition. Recently, Costa et al. (2003) undertook an extensive in silico study of hypothetical phenylpropanoid biosynthetic enzymes in Arabidopsis thaliana expressed sequence tag (EST) databases. Through examination of the available EST data, eight genes were found to be expressed in the flower buds. The enzymes predicted to be encoded by these genes included a cinnamate-4-hydroxylase (C4H), several 4coumarate CoA ligases, a cinnamoyl CoA reductase (CCR) and a cinnamyl alcohol dehydrogenase (CAD) (Costa et al, 2003). Phenylalanine ammonia-lyase (PAL) is a key enzyme in the initial steps of phenylpropanoid biosynthesis. Matsuda et al. (1996) introduced a gene encoding PAL2 from sweet potato under the control of a tapetum specific promoter into tobacco. Plants showing antisense-suppression also exhibited reduced pollen fertility. Although PAL enzymes have been shown to be active in the anthers of Arabidopsis (Rookes and Cahill, 2003), Brassica oleracea (Kishitani et al, 1993) and Tulipa (Rittscher Wiermann, 1983), there were no PAL ESTs detected in cDNA from flower buds by Costa et al. (2003). This discrepancy could reflect the stage of floral development at which RNA was extracted. If PAL enzymes were indeed involved in the biosynthesis of sporopollenin precursors, expression of PAL genes would be expected at the tetrad stage until the bicellular pollen stage. Consequently, if RNA was extracted from buds at later stages of development, the tapetum may have already  24 degenerated and no PAL ESTs would be detectable. While this presents an obvious limitation, when combined with what is known about the timing of sporopollenin biosynthesis, the use of genetics may aid in the isolation of genes responsible for the phenylpropanoid constituents of sporopollenin. In a similar manner, the fatty acyl precursor of sporopollenin may be isolated using genetic approaches. The Arabidopsis MS2 gene was identified and cloned because of the male sterile phenotype exhibited by ms2 mutant plants (Aarts et al, 1993). In these mutants, both the tapetum and the microspores are aborted at the time of microspore releasefromthe tetrad. An ultrastructural study indicated that, at the time of MMC meiosis, the tapetum is highly vacuolated. Similarly, upon their release from the tetrad, microspores are vacuolated and show no sign of a pollen wall (Aarts et al, 1997). Occasionally a pollen grain developed to maturity in a later flower. These pollen grains completely lack an exine wall and are instead coated with a thin unknown substance that confers no protection against acetolysis (Aarts et al, 1997). Fusion of the GUS reporter gene to the MS2 promoter supported in situ data showing expression of MS2 in the tapetal cells of Arabidopsis from tetrad release until completion of exine formation at the bicellular pollen stage. MS2 sequence analysis revealed homology to a fatty acyl reductase (JJFAR)fromjojoba (Simmondsia chinensis) that is involved in the reduction of fatty acids to alcohols in the formation of jojoba seed wax esters (Aarts et al, 1997). The phenotype ofms2 mutants, expression data and sequence homology to the JJFAR suggest a role for MS2 in the formation of the fatty acyl component of sporopollenin. Additionally, several male sterile mutants of Arabidopsis thaliana have been isolated that lack a structured exine wall. Analysis of the phenotypes of two such  25  mutants, ms9 and msl2, has been limited to ultrastractural characterization (Taylor et al, 1998). Both of these mutants exhibit an abnormal early tapetal degeneration. Although microspores oims9 mutants show no exine sculpturing, they do display an intact bottom layer, or nexine (Figure 1.4). In addition, globular material was observed to aggregate on the pollen at thefreemicrospore stage, but does not polymerize into sporopollenin. This globular material is presumably the contents of the prematurely degraded tapetum (Taylor et al, 1998). In contrast to ms9, msl2 mutants completely lack an exine wall. They also lack apertures. As both the patterning of the exine and the location of the apertures are influenced by the deposition of the cellulosic primexine, the msl2 phenotype suggests a defect in this process (Taylor et al, 1998). Unfortunately, the gene products and their function in sporopollenin synthesis and deposition remain to be determined. The male sterile phenotypes of mutants showing defective exines illustrate the essential nature of the exine wall in pollen viability. The exine forms an elaborately sculptured layer around the pollen grain. The patterning of the exine and the exceptional stability of sporopollenin form a cornerstone of the disciplines of palynology and paleopalynology (Scott, 1994). Palynology, the study of pollen and spores, is used in crop analysis, sedimentation studies, and forensics. In paleopalynology, the examination of the micro fossils of pollen and spores can lead to the identification of extinct species (Heslop-Harrison, 1968a; Taylor and Zavada, 1986). Certain characteristics of the exine are correlated with specific pollination methods, allowing inferences to be drawn about the plantsfromwhich these microfossils came (Crane, 1986). Because of the multi-disciplinary interest in the pollen exine, the terminology can be quite complex. Figure 1.4 shows a summary of terminology  26 commonly encountered in the literature. For the sake of simplicity, I will adopt the terminology used by Heslop-Harrison (1968a), as shown on the left side of the figure.  Tectu m-  i—Tectu  LEktexine 1 -  r~ S e x i n e n Columella  Baculae  Exine" r  L N e x i n &  L  -  Nexine 1 Nexine 2 Intine'  Foot layer—  Exine 1  Endexine ^ -  —  Intine  Figure 1.4: T h e structure of the pollen exine wall. A d a p t e d from Shivanna (2003)  Typically an exine is subdivided into two categories: the nexine and the sexine. Whereas the nexine lies directly in contact with the pollen grain and forms an almost continuous layer surrounding the grain, the sexine is a discontinuous highly sculptured layer (Heslop-Harrison, 1968a). The mature sexine consists of the baculae (or columellae), which arise perpendicular to the plasma membrane, and further projections of sporopollenin that run approximately parallel to the PM and are called the tectae (Scott, 1994; Dickinson and Sheldon, 1986; Dahl, 1986). The tectae can fuse to completely cover the pollen grain, as in the pollen of tobacco, or they can remain open, leaving spaces in between the baculae, as in Arabidopsis . The patterns formed by the fusion of the tectae are typically hexagonal or pentagonal and are referred to as the muri (Heslop-Harrison, 1968a; Taylor and Zavada, 1986). The nexine can be further subdivided into two categories: the nexine 1 (foot layer) and the nexine 2 (Heslop-  27 Harrison, 1968b; Heslop-Harrison, 1968a). The nexine 1 is deposited as an unpatterned layer that is uniformly distributed over the non-apertural regions of the microspore. Under examination with a transmission electron microscope, the nexine 1 and the nexine 2 appear to be separated by an electron lucent band (Heslop-Harrison, 1968a). The nexine 2 is believed to consist of sporopollenin-like material although is has different staining patterns than the sexine and the nexine 1. Like the intine, the nexine 1 is thought to be synthesized by the microspore (Heslop-Harrison, 1968a), however, unlike the intine, it is not deposited over apertural areas (Barnes and Blackmore, 1986). Formation of the exine begins with the deposition of the primexine during the tetrad stage of pollen development when the microspore remains surrounded with a callose wall (Barnes and Blackmore, 1986). Primexine synthesis is believed to be carried out by the microspore itself. Analysis of pollen grainsfromFi hybrids of two lily varieties showed no segregation of exine patterning, implicating the sporophytic genome control of patterning (Shivanna, 2003). As there is already evidence of an exine layer when the microspores are releasedfromthe callose wall, deposition of the primexine by the tapetum cannot entirely account for the sporophytic determination of exine patterning. It is hypothesized that long-lived RNA molecules or proteins synthesized in the MMC may be partitioned during meiosis and may play a role in primexine synthesis and deposition (Shivanna, 2003). Further evidence for the synthesis and deposition of the primexine by the microspores lies in examination of the apertures. Apertures are the germination zones of pollen grains. They provide a placefromwhich the pollen tube can emerge, modulate volume changes due to hydration and desiccation, as well as facilitate ion exchange  28 (Thanikaimoni, 1986). Mature Arabidopsis pollen is classified as tricolpate as it has three apertures that form elongated invaginations along the length of the grain (Furness and Rudall, 2004; Thanikaimoni, 1986). Although there is some evidence that the position of these apertures is influenced by the spatial relationship with the other microspores in the tetrad (Furness and Rudall, 2004), the most conspicuous feature of the aperture is the lack of an exine wall (Heslop-Harrison, 1968a). This lack of exine is thought to be controlled twofold by the microspore. First, there is no primexine deposited above the region destined to become an aperture (Scott, 1994). This lack of primexine deposition effectively blocks the polymerization of a patterned sexine layer. Secondly, immediately adjacent to the plasma membrane in regions destined to become apertures, lie stacks of endoplasmic reticulum. These stacks are occasionally referred to as onci in the literature (Barnes and Blackmore, 1986). These stacks of ER are thought to block the deposition of both the primexine and of the nexine 2 layer (Barnes and Blackmore, 1986). Concurrent with this hypothesis, Dickinson and Sheldon observed an 'aperture stencil', at the telophase II stage of MMC meiosis; the same developmental stage at which the primexine is deposited (Dickinson and Sheldon, 1986). The patterning of the primexine determines the future patterning of the exine, forming distinct probaculae that radiate perpendicular to the plasma membrane (HeslopHarrison, 1968a). The importance of the primexine in determining exine patterning in Arabidopsis is illustrated by the dexl mutant. DEX1 encodes a membrane associated protein with Ca binding domains that is expressed throughout the plant (Paxson2+  Sowders et al, 2001). Plants lacking DEX1 activity are male sterile, showing a delayed arid greatly reduced primexine. Sporopollenin appears to be unable to anchor to the  29 microspore and thus forms abnormal aggregations distributed randomly over the microspore plasma membrane. The identity and function of this gene product, as well as its role in other tissues, is currently unknown (Paxson-Sowders et al, 2001). In wildtype Arabidopsis the primexine acts as a scaffolding for the adherence and polymerization of sporopollenin precursors. Exine deposition starts before the free microspores are released from the tetrad, while they are still enclosed in a callose wall (Scott, 1994). Immediately upon release, the sporopollenin components are less resistant to acetolysis and are thus termed protosporopollenin (Heslop-Harrison, 1968b). This susceptibility could indicate that enzymes from the tapetum are required for the polymerization of precursors into mature sporopollenin (Scott, 1994). In addition, sporopollenin precursors are synthesized and secreted by the tapetal cells after the microspores are released from the tetrad. In Arabidopsis, which lacks the sporopollenincontaining orbicules, the exact nature of these precursors is unknown. Metz et al. (2000) propose that some of these precursors may be derivedfromthe TAG components of the tapetosomes. This suggestion is intriguing because, indeed, the TAGsfromtapetosomes are not present in the pollen coat, and it has been speculated that they degrade upon release into the locule or are taken up by the pollen grain (Piffanelli et al, 1997). Because the tapetosomes are released in the late stages of anther development, however, they cannot entirely account for the tapetally derived sporopollenin precursors in Arabidopsis. Where the sculptured sexine is not deposited there remains a cellulosic matrix onto which the pollen coat is deposited. In species where the tectae of the exine are fused,fragmentsof this cellulosic matrix can be seen trapped in between the baculae,  30 indicating that this matrix may have a role in deposition of the primexine, as well as of the pollen coat (Heslop-Harrison, 1968a). 1.3.2b The lipids of pollen grains - the pollen coat The pollen coat, also referred to as pollenkitt or tryphine, is a lipophillic layer deposited in the structured exine. These three terms are commonly used interchangeably in the literature. Despite this trend, there has been a distinction drawn between a pollenkitt and a tryphine. According to Shivanna (2003) and the references therein, historically a pollenkitt is the hydrophobic lipidic component that contains pigmentation whereas a tryphine is the hydrophilic substance released and deposited on the pollen grain following tapetal cell degradation. Because of this discrepancy, the term pollen coat will be used for the purpose of this document to refer to all tapetally derived components that are deposited on the pollen exine, including but not limited to, lipids and proteins. The constituents of the pollen coat then are derived from lipidic components, phenolic signalling molecules, and proteins that are synthesized in the tapetal cells of the anther. The pollen coat serves many functions, including attracting and adhering to pollination vectors, maintaining the desiccated state of mature pollen, and facilitating the hydration of pollen upon contact with a receptive stigma (Piffanelli et al, 1997). The pollen coat is typical of species with dry stigmas and is rarely seen in species with wet stigmas (Heslop-Harrison and Shivanna, 1977). The components of the pollen coat essentially mimic the constituents found in the exudate of a wet stigma (Heslop-Harrison and Shivanna, 1977). In Brassica napus, the pollen coat has been shown to contain medium chain sterol esters as its main lipidic component (Hernandez-Pinzon et al, 1999). Detailed analysis  31 of the lipid composition of elaioplasts, tapetosomes, pollen coatings, and coatless pollen grains revealed that the sterol esters in the pollen coat are derived from the tapetal elaioplasts (Hernandez-Pinzon et al, 1999). Additionally, the pollen coat contains fatty aldehydes and very long chain acyl esters. Analysis of the pollen coats isolated from Arabidopsis thaliana yielded similar results (Hernandez-Pinzon et al, 1999). The major very long-chain lipid constituents of the Arabidopsis pollen coat are the C29 alkanes (nnonacosane) and ketones (15-nonocosanone), as well as the C30 primary alcohols (triacontanol) (Preuss et al, 1993; Mayfield and Preuss, 2000). The main class of proteins associated with the Brassica napus pollen coat are the oleosin-like proteins (Hong et al, 1997; Murphy and Ross, 1998). Similarly, the major classes of proteins found in the Arabidopsis pollen coat are lipases and oleosin-like proteins (Mayfield et al, 2001). Oleosin proteins are found associated with seed lipid bodies of most seed-oil plants. They are thought to be involved with the stabilization of storage lipid bodies (Murphy, 1996). Before deposition in the pollen coat, oleosin-like proteins identified in B. napus (Murphy and Ross, 1998) and Arabidopsis (Mayfield et al, 2001; Mayfield and Preuss, 2000) are present exclusively in the tapetosomes in the tapetal tissues (Ferreira et al, 1997; Murphy and Ross, 1998). Unlike seed oleosins, the oleosin-like proteins are cleaved at their mature state. The C-terminal oleosin-like domain is found in the pollen coat, while the N-terminal structural domain appears to be degraded (Murphy and Ross, 1998). It is hypothesized that the mature oleosin-like proteins may play a role in stabilizing the lipid components of the pollen coat (Murphy and Ross, 1998).  In addition to the oleosin-like family of proteins, aquaporins have  also been detected in the pollen coat. This is a particularly alluring discovery, as these  32  proteins may be directly involved in the rehydration of the pollen grain upon contact with a stigma (Ikeda et al, 1997).  1.4 Pollen adhesion, hydration and germination The mechanisms of pollen adhesion, hydration, and germination are very complex, involving many different steps and very fast signalling between the pollen grain and the stigma. In order for the pollen grain to germinate on plants with dry stigmas, such as Arabidopsis, the stigma must release water, allowing the pollen grain to hydrate (Zinkl et al, 1999). The pollen coat is involved in both pollen-stigma signalling and pollen hydration (Preuss et al, 1993). Its adhesive properties allow the pollen to bind to the relatively dry stigmatal surface (Lolle and Pruitt, 1999). The pollen coat is also essential in maintaining the semi-hydrated state of the pollen grain. This layer prevents excessive moisture loss and/or premature hydration, ensuring that the pollen grain will remain viable and germinate at the correct time (Shivanna, 2003). Typically tricellular pollen grains with pollen coats do not have a long life outside of the anther and do not undergo periods of dormancy (Heslop-Harrison and Heslop-Harrison, 1994). Thus it is essential that the pollen remains in the optimal state until it comes into contact with the stigma. Once contact is established, it is essential that rapid hydration take place (Heslop-Harrison and Heslop-Harrison, 1994). Following initial contact between the dry stigma and a compatible pollen grain, a strong adhesion occurs very rapidly in members of Brassicaecae. This adhesion seems to have two phases: an initial strong adhesion that occurs within seconds of contact, and a second adhesion that progressively strengthens until germination (Preuss et al, 1993;  33 Zinkl et al, 1999). In Arabidopsis, the initial adhesion is thought to be mediated by the lipidic components of the exine (Zinkl et al, 1999; Zinkl and Preuss, 2000). This hypothesis was formed after analysis of pollen lacking a pollen coat but having an intact exine. Such pollen was obtained bothfromthe coatless mutant cer6 and chemical removal of the coatfromwildtype pollen (Zinkl et al, 1999). In both cases, the initial adhesion of pollen grains to the stigma was comparable to that of wildtype pollen. The pollen grain could not be removed by treatment with proteolytic enzymes, indicating that this adhesion was not mediated by protein-protein interactions (Zinkl et al, 1999). Furthermore, analysis of less adherent pollen (lapl) mutants showed that the decrease in adhesion was accompanied by a defective exine (Zinkl and Preuss, 2000). Luu et al (1997) demonstrated that treatment of Brassica oleracea stigmas with proteinase K resulted in a reduction in initial adhesion force, but not elimination of adhesion. This suggests that the adhesion is not entirely mediated by lipidic molecules; proteins on the surface of the stigma likely also have a role in the initial adhesion. Although the exine has been implicated in the initial binding of the pollen grain to the stigma, the pollen coat has a role in the later phase of this binding. Unlike wildtype Arabidopsis pollen, the initial adhesion did not strengthen in coatless pollen grains. Such grains also failed to hydrate or germinate (Zinkl et al, 1999). Moreover, Heizmann et al. (1994) reported a decrease in adhesion forces in the cer6 and cerl mutants, both of which have an aberrant pollen coat. These observations implicate factors carried in the pollen coat in the strengthening of adhesion forces between the pollen grain and the stigma, maintaining contact until germination.  34 Once the lipophilic molecules begin adhering to the stigma, the pollen coat mobilizes and spreads along the surface of the stigma, forming a 'foot layer' and thereby maximizing the contact surface between pollen and stigma (Elleman et al, 1992; Elleman and Dickinson, 1996; Zinkl et al, 1999; Lolle and Pruitt, 1999). This movement of the pollen coat is accompanied by a change in electron density when observed under a TEM (Elleman et al, 1992; Elleman and Dickinson, 1990). The initially homogeneous electron lucent coat becomes heterogeneous with visible membranous inclusions and vesicles (Doughty et al, 1992). This change in pollen coat triggers an expansion of the stigma cell wall at the point of contact with the pollen grain (Elleman and Dickinson, 1996). The pellicle of the stigma expands and the underlying cytoplasm is rich in ER, Golgi, and vesicles, implying an increase in metabolic activity (Elleman and Dickinson, 1996). At this time, the stigma wall can be observed invading regions between baculae of the exine (Zinkl et al, 1999). Furthermore, isolated pollen coats can elicit this expansion of the stigma, underscoring the importance of the pollen coat in pollen-stigma interactions (Wheeler et al, 2001). Signalling from the pollen coat triggers hydration of the pollen grain, a step that is essential for germination. Mutants with altered pollen coats are unable to properly hydrate (Preuss et al, 1993; Hulskamp et al, 1995). This failure could be remedied by copollination with wildtype grains or by treatment with isolated pollen coat (Hulskamp et al, 1995; Wheeler et al, 2001). Moreover, lapl pollen has an aberrant exine but an intact pollen coat, and consequently retains the ability to hydrate and germinate (Zinkl and Preuss, 2000). Because hydration of the pollen grain can be initiated by treatment with isolated pollen coat, the signalling molecules necessary for the release of water from  35 the stigma must be present in the coat and not in the pollen grain itself (Hulskamp et al, 1995). Indeed, Mayfield and Preuss (2000) identified an oleosin-like protein, GRP17, present in the pollen coat of Arabidopsis thaliana. Nonfunctional copies of this gene result in a delay in pollen hydration and a consequent reduction in pollen fitness (Mayfield and Preuss, 2000). As oleosin-like proteins constitute the major family of Arabidopsis coat proteins, it is not surprising that grpl 7 plants retained some ability to hydrate and germinate (Mayfield and Preuss, 2000; Mayfield et al, 2001). This retention indicates a possible functional redundancy between oleosin-like proteins and further emphasizes the importance of the coat constituents in interactions with the stigma. Many of the mutants with altered pollen coats are cer mutants, which have altered stem wax phenotypes (Koornneef et al, 1989; Dellaert et al, 1979; Hannoufa et al, 1993). Male sterility, observed in some cer mutants, may result as a pleotropic effect caused by a deficiency in the lipids forming the pollen coat (Preuss et al, 1993). Thus waxes, and their fatty acids, are important not only in the actual physical formation and viability of the pollen grain, but also in the regulation of fertility by contributing to the key components of the pollen coat. In the pollen coat, the lipids may maintain the integrity of the signalling molecules or may act as signalling components themselves. From this discussion, it is evident that lipids play an essential role in pollen development and viability. The following is a brief overview of the synthesis and study of lipids and waxes in Arabidopsis.  36 1.5 Wax Biosynthesis 1.5.1 Fatty acid synthesis Fatty acids are synthesized de novo in the plastids of all plant cells (Harwood, 1996). The process of fatty acid synthesis is summarized in Figure 1.5. The first committed step in fatty acid synthesis is the conversion of acetyl-Coenzyme A (CoA) to malonyl-CoA by an acetyl-CoA carboxylase (ACCase) (Rawsthorne, 2002; Harwood, 1996; Baud et al, 2003; Ohlrogge and Jaworski, 1997). Malonyl-CoA-ACP transacylase (MCAT) then transfers the malonyl moietyfromthe CoA to an acyl carrier protein (ACP) (Harwood, 1996). The subsequent steps involve a series of four reactions, each of which is catalyzed by a separate enzyme. The four reactions form one cycle of the fatty acid syntheses process resulting in an addition of 2 carbons (C) to the growing fatty acid chain, p-ketoacyl synthase III (KASIII) catalyzes the initial condensation of acetyl-CoA with malonyl-ACP to form a P-ketoacyl-ACP. This initial condensation is followed by a reduction to p-hydroxyacyl-ACP by P-ketoacyl-ACP reductase. The subsequent dehydration by P-hydroxyacyl-ACP dehydrase releases a water molecule to form a transA -enoyl-ACP moiety. Thefirstcycle of fatty acid synthesis is complete with the action 2  of an enoyl-ACP reductase to generate a four carbon backbone acyl-ACP fatty acid. Each subsequent cycle of fatty acid synthesis includes the same four enzymatic reactions resulting in the elongation of the growing acyl chain by two carbons. In each cycle of reactions, malonyl-ACP acts as the two carbon donor. Three separate P-ketoacyl-ACP synthase (von Aderkas and Dawkins, 1993) enzymes, each specific for certain acyl chain lengths, catalyze the condensation reactions. KASIII is specific for acyl chain lengths of C2-C4, and thus mediates the condensation reaction in the initial round of fatty acid  37 synthesis (Clough et al, 1992), while KASI is specific for C4-C14 and KASII for C16C18 (Shimakata and Stumpf, 1982). The three other enzymes involved in fatty acid synthesis have no acyl chain length specificities and participate in fatty acid synthesis of C2-C18 acyl groups. When the carbon chain reaches 16 or 18C in length, the formation of the fatty acid is terminated. Termination can occur via one of two reactions. Plastidial acyltransferases can transfer the acyl group from acyl-ACP to glycerol-3-phosphate to produce glycerolipids via the prokaryotic (plastidial) pathway, or acyl-ACP thioesterases (FatA and FatB classes) can hydrolyze the acyl-ACP moiety and release free fatty acids (Ohlrogge and Jaworski, 1997; Liu and Post-Beittenmiller 1995). Thesefreefatty acids are exported from the plastid and re-esterified to CoenzymeA (CoA) (Post-Beittenmiller, 1996). Before exportation, C18:0 can be processed by stearoyl-ACP desaturase to produce C18:l. C16:0, C18:0 and C18:l can be used in the formation of glycerolipids in the plastid via the prokaryotic pathway, or exportedfromthe plastid and used for production of glycerolipids via the eukaryotic pathway, sphingolipids, waxes and/or cutin and suberin. C16:0 and C18:l are the major precursors of glycerolipids. Because the epicuticular waxes do not contain unsaturated fatty acids, either Cl6:0, C 18:0, or both are the major precursors of these molecules. 1.5.2 Fatty acid elongation  The synthesis of very long chain fatty acids (VLCFAs) occurs by an elongation of the C16 and C18 exported from the plastid (Cassagne et al, 1994). The enzymes that catalyze this elongation are found in the microsomalfractionand are likely localized in the endoplasmic reticulum (Bessoule et al, 1989; Fehling et al, 1992). Elongation  38 MCAT  ACCase  (malonyl-CoA-ACP  (acetyl-CoA carboxylase)  O  transacylase)  O  H3C-C-S-C0A  OOC-CH -C-S-Co/r  Acetyl-  Malonyl-CoA  co  2  O  CoA  H C-<y-CH -C-S-ACP  KAS I or KASII 0  0  H C-CH -CH -(!!-CH2-C-S-ACP 3  2  2  2  B-ketoacyl-ACP  Bj -ik\ ve t o a c y l - A C P IVI  NADPH NADP H  B-ketoacyl-ACP reductase O  +  HaC-C-CrVc'-S-ACP I OH  NADPH-K. 1  frans-A -enoyl-ACP  E  S  O  A  C  Y  L  .  A  C  P  O  OH B-hyd r o x y a c y I - A C P B-hydroxyacylACP H 0  y  2  I H  K  H3C-CH2-CH2-C-CH2-C-S-ACP  O  .i=c-d! S-ACP  .  H  4  H,C  R  recuctase  C-hydroxyacyl-ACP 13-hydroxyacyl _ ACP H2O ^ dehydrase H  dehydrase  H  0  H C-CH -C=C-CH,-C-S-ACP 3  2  2  NADPH NADP  enoyl-ACP reductase  +  trans-A -e n o y I -A C P 2  NADPH " ^ N l NADP < ^ +  enoyl-ACP reductase  O H3C-CH2-CH2-C-S-ACP  acyl-ACP  Malonyl-ACP  )  O  3  ACP  A D P + Pi  KASIII (ketoacyl ACP synthase  CoA  'OOC-CH2-C-S-ACP  2  ATP  0  O H3C-CH2-CH2-CH2-CH2-C-S-ACP acyl-ACP  Figure 1.5: S u m m a r y of de novo fatty acid synthesis  involves a condensation, a P-keto reduction, a dehydration and an enoyl-reduction reaction, and proceeds in a manner analogous to de novo fatty acid synthesis (Cassagne et al, 1994). This process is summarized in Figure 1.6. In this four enzyme model for fatty acid elongation (FAE), the carbon backbone of the fatty acid is elongated by 2 carbons, and malonyl-CoA serves as the carbon donor. Condensation of the acyl-CoA and malonyl-CoA are carried out by a P-ketoacyl-CoA synthase (Todd et al, 1999) to generate p-ketoacyl CoA. This is then converted to a P-hydroxy-acyl-CoA by a P-keto reductase. The following steps involve a dehydration to a trans-2-acyl-CoA and a final reduction by an enoyl reductase to an acyl-CoA moiety that has been extended by two carbon atoms. Collectively, the four enzymes involved in fatty acid elongation are referred to as an elongase (von Wettstein-Knowles, 1987; Post-Beittenmiller, 1996). As in fatty acid synthesis, the initial condensation reaction is the rate limiting and chain length specific step, with the remaining three enzymes being used with a variety of condensing enzymes in different elongases (Millar and Kunst, 1997). In Arabidopsis, four condensing enzymes, FAE1, FDH, KCS1, and CER6, have been cloned and studied (Kunst et al, 1992; James et al, 1995; Millar et al, 1998; Todd et al, 1999; Yephremov et al, 1999; Preuss et al, 1994; Millar et al, 1999). FAE1 has been shown to be a condensing enzyme (Kunst et al, 1992; Millar et al, 1998), and by homology, the others are considered putative condensing enzymes. FAE1 is expressed solely in the seeds of Arabidopsis and is specific for acyl chain lengths of 18 and 20 carbons (Millar and Kunst, 1997). FDH is expressed solely in the epidermal cells of the aerial plant tissues (Yephremov et al, 1999; Pruitt et al, 2000). KCS1 encodes a 3ketoacyl-CoA synthase (Todd et al, 1999) found in both the vegetative tissues and the  40 roots of the plant. Mutants deficient in KCS1 activity show a reduction in the level of C26 and C30 wax aldehydes, although none showed a complete waxless phenotype (Todd et al, 1999). Finally, the CER6 protein is an epidermis specific condensing enzyme expressed in the aerial organs of the plant. In addition to its epidermal expression, CER6 transcript is also present in the tapetal cells of the anther (Millar et al, 1999; Hooker etal, 2002). Putative P-ketoacyl reductases have been identified in maize and yeast (Saccharomyces cerevisiae) and a homologue identified in Arabidopsis. In maize, Glossy8 (GL8) mutants show reduced wax on the surfaces of seedling leaves with a leaf wax load one third that of wildtype (Xu et al, 1997). Subcellular localization studies placed GL8 in the endoplasmic reticulum of maize leaf cells (Xu et al, 2002) where fatty acid elongation is thought to occur. Reaction of the GL8 protein with leek acyl-CoA elongase antibodies and inhibition of reductase activity by GL8 antibodies suggest that GL8 is the P-ketoacyl-CoA reductase component of the maize elongase complex (Xu et al, 2002). A P-keto reductase was also isolated from Saccharomyces cerevisiae (Beaudoin et al, 2002). YBR159 encodes an integral membrane protein that is believed to be localized to the endoplasmic reticulum. Although disruption of this gene severely reduced reductase activity (Han et al, 2002), the activity was not completely eliminated, nor was the loss of function lethal, suggesting functional redundancy. Reductase activity could be rescued in Aybrl59 by an identified Arabidopsis homologue (Beaudoin et al, 2002). Interestingly, the Arabidopsis gene Atybrl59 (Atlg67730) is the same gene annotated as encoding the Arabidopsis homologue to maize GL8 (Xu et al, 1997).  o  o  H C(CH ) -C-S-CoA 3  2  "OOC-CH -(B-S-CoA 2  n  Acyl-CoA  Malonyl-CoA  .  _  KCS (ketoacyl-CoA synthase) CoA + ^ | C0 ^ ' FAEI, F D H , • CERfi. 2  o  o  H C(CH ) -C-CH -C-S-ACP 3  2  n  2  C-ketoacyl-CoA NADPH NADP  B-ketoacyl-CoA reductase H  O  H C(CH ) -c!-CH -(!!-S-ACP 3  2  2  n  OH O-hydroxyacyl-CoA ]^  B-hydroxyacyl-CoA dehydratase  H  O  A  11  H C(CH ) -C=C-C-S-ACP 3  2  n  H frans-A -enoyl-CoA 2  NADP*  ^  "°"- ° reductase O  e  C  A  HqCfCH9)n-CH?-CHp-i C(CH2)n-CH2-CH2-C-S-ACP 3  acyl-CoA  Figure 1.6: S u m m a r y of fatty acid elongation  J  42 Finally, an Arabidopsis homologue to the yeast enoyl-reductase has been identified and named AtTSC13 (Gable et al, 2004). Expression of the Arabidopsis gene (At3g55360) rescued the lethality of a temperature sensitive enoyl-reductase/condensing enzyme double mutant. Furthermore, the yeast enoyl-reductase TSC13 was replaced with the Arabidopsis AtTSC13 coding region and expressed. Immunoprecipitation of the yeast elongase resulted in the co-precipitation of AtTSC13, indicating a physical interaction between the Arabidopsis protein and the components of the yeast elongase. The homology to the yeast enoyl-reductase, rescue of the yeast mutant phenotype and the recombinant interactions all indicate that this gene may indeed encode the Arabidopsis enoyl reductase (Gable et al., 2004). 1.5.3 Wax biosynthesis In Arabidopsis, wax biosynthesis is believed to occur via two parallel pathways: the decarbonylation pathway and the acyl-reduction pathway (Post-Beittenmiller, 1996). The decarbonylation pathway produces even carbon chain length aldehydes and odd carbon chain length alkanes, secondary alcohols and ketones whereas the acyl-reduction pathway produces even carbon chain length primary alcohols and wax esters. The products of the decarbonylation pathway account for approximately 90% of the stem epicuticular wax of Arabidopsis (Millar et al, 1999). In this pathway, a fattyacyl CoA reductase (FAR) converts VLCFA precursors to aldehydes (Vioque and Kolattukudy, 1997). A fatty aldehyde decarbonylase then catalyzes the conversion of the aldehydes to alkanes, with the release of CO (Schneider-Belhaddad and Kolattukudy, 2000). These alkanes are then hydroxlyated into secondary alcohols, which can be subsequently oxidized into ketones (Kolattukudy et al, 1973).  43 The acyl-reduction pathway produces primary alcohols and wax esters of even chain length and accounts for approximately 10% of Arabidopsis stem wax (Millar et al, 1999). The production of primary alcohols is catalyzed by an alcohol generating fatty acyl-CoA reductase (FAR) and is thought to occur via an aldehyde intermediate that is not released. FAR enzymes have been identified in jojoba (Metz et al, 2000) and assayed in pea (Vioque and Kolattukudy, 1997). In Arabidopsis, the tapetum-specific MS2 protein has been characterized as a putative FAR essential for exine deposition and pollen fertility (Aarts et al, 1993; Aarts et al, 1997). Wax esters are formed by a fatty acyl-CoA: fatty alcohol acyltransferase, or wax synthase (WS), that transfers a fatty acyl-CoA to a fatty alcohol. A WS enzyme has been purifiedfromjojoba and the WS gene isolated (Lardizabal et al, 2000). Furthermore, database searches reveal homologues in Arabidopsis. Two different acyl-CoA reductases have also been isolatedfromPisum sativum (Vioque and Kolattukudy, 1997). One of these reductases acts on fatty acyl CoA to form fatty aldehydes while the other acts on fatty acyl-CoA to form alcohols withoutfreefatty aldehyde intermediates (Schneider-Belhaddad and Kolattukudy, 2000). This finding indicates that aldehydes are formed exclusively by the decarbonylation pathway and not by the acyl-reduction pathway of wax biosynthesis and supports the two pathway model for wax biosynthesis. 1.5.4 eceriferium mutants  In Arabidopsis twenty-two mutants have been isolated with altered stem wax phenotypes and designated eceriferium (cer) mutants (Koornneef et al, 1989; Hannoufa et al, 1993). An altered wax profile may alter stem wax crystal composition, causing a  44 green glossy phenotype instead of the white glaucous wildtype phenotype. The cer mutants have thus been classified according to the level of visual glossiness and fertility under conditions of low humidity (McNevin et al, 1993). Only four of the genes altered in the cer mutants have been cloned. These genes are CER1 (Aarts et al, 1995), CER2 (Negruk et al, 1996; Xia et al, 1996), CER3 (Hannoufa et al, 1996), and CER6 (Millar et al, 1999; Preuss et al, 1993; Fiebig et al, 2000). The cerl mutant has a glossy stem phenotype and is infertile when grown under low humidity (McNevin et al, 1993). This sterility is due to a lack of pollen germination and can be rescued by transfer to high humidity conditions; it is therefore termed conditional male sterility. Interestingly, the cerl mutant sterility is not caused by a lack of a pollen coat. Although the pollen coat is present in the cerl mutant, the lipid droplets within the coat are much smaller and more abundant than those of wildtype (Aarts et al., 1995). The cerl mutant has a highly glossy stem phenotype due to the presence of very few wax crystals on the stems (McNevin et al, 1993). Unlike cerl however, cer2 is fully fertile (McNevin et al, 1993). The wildtype CER2 transcript is present at high levels in the stems and the flowers and at low levels in all other tissues (Negruk et al, 1996; Xia et al, 1997). This expression profile is particularly interesting because despite high levels of wildtype CER2 expression in the flowers, pollen fertility is not affected. Analysis of the expression of the GUS reporter gene fused to the CER2 promoter revealed expression of CER2 in the anthers of developing flower buds (Negruk et al, 1996; Xia et al, 1997) although the specific tissue within the anther was not investigated. Potentially CER2 is  45 expressed in neither the tapetum nor the developing pollen grain and thus has no effect on pollen fertility. cer3 mutants are also highly glossy and conditionally male sterile (McNevin et al, 1993). CER3 is expressed at high level in all tissues: leaves, stems, roots, flowers and apical meristems (Hannoufa et al, 1996). Although it was shown that there is a reduction of free fatty acids and a corresponding accumulation of longer chain alkanes on the stems of cer3 mutants (Jenks et al, 1995), the effect on cer3 pollen is unknown. Finally, the stems of cer6 mutants are wax deficient and their pollen is conditionally sterile (McNevin et al, 1993). In these mutants, the pollen attaches to the stigma but does not hydrate (Preuss et al, 1993). Examination of the severe cer6-2 allele revealed a complete absence of the pollen coat (Preuss et al, 1993). Partial restoration of the activity of CER6 is sufficient to restore fertility, but not to produce a normal layer of stem epicuticular wax (Fiebig et al, 2000). These observations indicate that the conditional sterility may be due to an absence of signalling molecules that are carried within the pollen coat and necessary for the initiation of pollen hydration. Based on sequence similarity to FAE1, the CER6 protein is thought to be a VLCFA condensing enzyme necessary for epicuticular wax biosynthesis (Millar et al, 1999). Expression of the CER6 mRNA is restricted to the epidermal cells of the aerial tissues and the tapetal cells of the anther. CER6 transcripts accumulate in the tapetal cells and remain until the breakdown of the tapetal cell layer (Hooker et al, 2002). It is interesting to note that, despite the lack of pollen coat, the exine is intact in cer6 pollen. Since CER6 is thought to generate the VLCFA precursors for the wax biosynthetic pathway (and thus enzymes such as MS2) the lack of exine phenotype in  46 these mutants is curious. Perhaps CER6 activity is not required for the formation of sporopollenin, and a separate condensing enzyme carries out this function in the tapetal cells. Despite this question, the lack of a pollen coat in cer6 mutants and the homology of CER6 to VLCFA condensing enzyme FAE1 indicate a potential role for CER6 in the formation of pollen coat lipids (Millar et al, 1999; Hooker et al, 2002).  1.6 Male sterility Male sterility is the inability of the pollen grain (male gametophyte) to fertilize the ovule (Nonomura et al, 2003; Chaudhury, 1993). This sterility can be caused by defects in sporogenous tissues, such as the tapetum, pollen mother cells, or tissues of the anther, and thus be classified as sporophytic male sterility. Additionally, deficiencies in pollen maturation, germination, or in pollen tube growth can be classified as gametophytic male sterility because they are under the control of the haploid genome. Deficiencies in the pollen coat and the exine are deemed sporophytic male sterility because their deposition is under the control of the sporophytic genome (Chaudhury, 1993). Further classification of plants exhibiting male sterility relies on observations of the phenotypic manifestation and genetic inheritance of the sterility (Horner and Palmer, 1995). Phenotypically, male sterility can be divided into sporogenous, structural, and functional sterility. Sporogenous male sterility is used to describe plants where stamen development proceeds normally but a defect manifests during the time of early microsporogenesis to late microgametogenesis that leads to nonfunctional pollen. In instances where stamen formation is either not initiated or abnormal resulting in a failure  47 to initiate microsporogenesis or an aborted microsporogenesis, the term structural male sterility is utilized. When pollen is produced but is not released or cannot reach the stigma, the defect is termed functional male sterility. The genotypic classification of male sterile plants is based on the mode of inheritance of the sterility phenotype and can be genie (nuclear), cytoplasmic (controlled by maternally inherited mitochondrial genes), or genic-cytoplasmic (cytoplasmic male sterility with nuclear fertility restorer genes). A combination of these phenotypic and genotypic descriptions can be used in the examination of both naturally occurring and artificially induced male sterility (Horner and Palmer, 1995). The complexity of anther and pollen development allows for many possible causes of male sterility. It can result from adverse growing conditions such as low temperatures (Kuranouchi et al, 2000) and affliction of disease (Budar and Pelletier, 2001), or can arisefrommutations in one of the many genes essential for pollen development and viability (Chaudhury, 1993). Because of the essential nature of pollen lipids, and because of their originfromeither the developing pollen grain or the tapetal cells, many mutants in Arabidopsis (and in many other plant species) have been isolated and found to have deficiencies in their pollen lipids. As seen in the previous sections, these deficiencies can impair the development, release, hydration, and/or germination of the pollen grain. Table 1.3 presents a selection of male sterile mutants of Arabidopsis thaliana whose gene products have been examined. Those with a predicted role in pollen lipid formation are indicated in red. Naturally occurring male sterility has been exploited in such agricultural breeding systems as carrot, onion, maize, sugar beet, sunflower, rice, canola, and soybean (Budar  48 and Pelletier, 2001; Horner and Palmer, 1995). Heterosis refers to the increased vigour and competitive advantage seen in hybrid plants. Because inbreeding does not allow for new favourable allele combinations, many breeding systems force hybridization to exploit this increased hybrid vigour (Budar and Pelletier, 2001; Marais et al, 2000; Goetz et al, 2001). In these systems, the use of a male sterile plant as a female parent eliminates the possibility of contamination of valuable hybrid seed (Horner and Palmer, 1995). When naturally occurring male sterile variants are not available, the effect can be generated by a variety of methods such as manual emasculation (Budar and Pelletier, 2001), varying environmental conditions (Kuranouchi et al, 2000), treatment with gametocides (Loussaert, 2004), crossing in sterility, or by genetic manipulation. Many of these procedures are costly, labour intensive, or time consuming, requiring physical removal of stamens, extended exposure to certain conditions and/or many generations of backcrossing sterility into desired plants (Horner and Palmer, 1995). Additionally, treatment with chemical gametocides can impair development of other organs and the effect is not specific to the plants of interest (Loussaert, 2004). Many of these costs can be circumvented by genetically manipulating desirable plants to be male sterile. Consequently, much effort has been focussed on developing such systems in barley, wheat, tobacco, cabbage, sugar beet, and petunia (Marais et al, 2000; Zhang et al, 2001; Deblock et al, 1997; Lee et al, 2003; Kuranouchi et al, 2000). Typically these strategies target the cells of the tapetum by using tapetal specific promoters such as the A9 promotersfromtobacco and Arabidopsis (Paul et al, 1992; Fuerstenberg et al, 2000). One popular strategy has been to induce targeted cell ablasion by transformation  49 CN O CN o O CN O (N  cu  u  e  Os  o\  <N O  ON ON  a  «  53 CD  CO  pi  N U  fl  <3  cu  cci  60  CO  fl CD  .fl N  >  fN  —)  CD  cu 60 fl CO a M cd cd  cu CD O Id  o  O  60 fl  fl  CO  t! t!  CO1 O M  cd cd  <<  60 ON cd i> ON - A  al S N  -a co h  CD  ^ CO  •*-»  co CO CD  I.  cu  i2 o  13  i_> 4-*  CD  o  C  In  CD CU —)  CO  3  —)  c u CD 60 ~j o a OH  CD  CD"  fl JU  60  & —J  S  CO  CD „ cO  o  u  _  co  g  a  CD  CL M  co  60 o  60  cu  Id co  CO  CD o 3co CL  2  0 co O X) rj cd 13 _  CD fl  CD  B  CO  O  CD  2 °  CD  £o  O  CO CO  o o c  S  O  —  U i  c2  -6 *  CO CD  O  a  OH  a  o  CL  . f l CO CO O  l-i UA 60 O O ' §  u  "ed «l §  g  cS co Ji cd 3 CD  CD  CO  co  CO  u  p , _cd  CD < CD u  —  CD  O _ CM cfa  OH , 0 CD CD —;  _  cu s-  .23 co  11  Q.  CL  O CL co _ O  O cO  o  ,2  O  x fl  | £ CU  3  CO CD CD _  a  CD  *J o  ^  o  t/l  22  - B  CO  o  o  CD  OH  •3  CO CD CD  CU  co  c_ CD •fl fl  c  CD  CO CD  CD  60  C C  « U <u  T3  « o  3 Ch „. CO 60 o  *3 —1 S  £s  C o  c o  c CU  >>  o s  >1  CD  "5^  cu  e  g X)  CD  CO  O  5 13  « ti  c at  5 -y  3  IS  CD  fl  CO CO  <U  C (U  CD  X) X)  co  _c0 "fl <u  & s « eg ed •!-(  5 £  R S3 m a,  CO  CL  2 So"  s .a s  o c o  o  60  CO 1  u — O  •o  CD fl CD  CO  CD  _c0  CO  CD  X)  CD  CD  O  e C X) 'OH C CO fl  W Q  O P.  duce  O  60  —J c  xine  O  —J co „ >  B  0  .3 c  o a o  u O  -a a CO  cu  > O bO  •8  1 cu  fl  O ,c0 60 o fl  60 S ,  C  "  ^ Q  !N  IS »  - < OH  ~C3  f  cy CO  60  g .3 CO XI  CO  CO  O c  » p  K  -  o ^  I-  23  s5 _§B i *2^ 4H CD  43 O  ^  43  fl f  C  . 2 S "S 3 •  Q S U xi  — rO  _ ON O  ~~  i  > «  -~ "2 ^ » »  <^ ON O » O O .  ^  <N ©  O N  b DC  c  ^  >  ° C  oo  ro O O q O  as  CN  ON  m  3  23  3 O  i2  -H  S  <  <u co  >N  ON  «  (3  © cu 13  S < ~«" 2 •S U-< o * J 1-1 ^ o 3 o a o ^ Q 3 <w> CO  cu 3  ^  3.  o ©  a)  K S3  •  CN  fl CO  'EL  -o u  -t—»  uc "o Cu  cd  5  u  S O  •a >> ,fl O  -o u  Cu r f l v-c cd  o b  g  co  ca  CD  ro  g;  C O <u ca  ca  a  ^  (L)  o  PH - U pq  00  <D  3 ID  I1 -a  ca  _ o C3  ca C  3  QJ CO  S* s  § > +H  O  C «J  u  co cd to-  (5 5 < .S  o  CO  Id  <L>  "cd o  C  T3  "  o  hJ  o  >> g ft  <  CN-)  uJ  o ^00  UH  «  cd U  o  o  CO  o  C  0 >  Cu 00  c  .fl  O  o  •S J2 c « '« t:  11'll W  3  CO HH  O o >» U  Cu co O Cu  M  O jo ca  -s -s cu cu  o CN  o  •a 3 o  „  r a  «  a, o  '•8 ~ i  ON ON  O so ^  I a .s>,8  |£  ON ON  „ ~ —^  «N  CN — v  1  O  51 with a gene encoding a cytotoxic protein. This strategy has been successful using the barnase ribonuclease in tobacco (Beals and Goldberg, 1997) and wheat (Deblock et al, 1997), targeting destruction of the stomium cells and tapetum respectively. Similarly, the diphtheria toxin A-chain destroys tapetal cells and aborts tetrad formation when expressed in the tapetum of cabbage (Lee et al., 2003), and a ribosome inactivating protein has the same cytotoxic effect in tobacco (Cho et al, 2001). While the use of cytotoxic genes to generate male sterility has been quite effective, there has also been interest in generating sterility by silencing the expression of endogenous genes necessary for pollen viability. Such strategies include antisense suppression of the SnRkl protein kinase in barley, resulting in inviable pollen arrested at the binucleate stage of development (Zhang et al, 2001). As mentioned earlier, suppression of key enzymes involved in phenylpropanoid synthesis (PAL) or flavanoid synthesis (Rowley et al, 2003) also result in male sterility in tobacco, broccoli, and petunia (Matsuda et al, 1996; Kishitani et al, 1993; van der Meer et al, 1992). Finally, antisense suppression of a mitochondria-targeted pyruvate dehydrogenase Ela subunit has been localized to the tapetum in sugar beet (Yui et al. 2003). This localized suppression resulted in malformed tapetosomes, elaioplasts, and orbicules in the tapetum, and an aberrant exine on the microspores. This result is particularly intriguing as silencing of this enzyme inhibits the TCA cycle, which consequently reduces the levels of acetyl-CoA. Since acetyl-CoA is a substrate for fatty acid synthesis, it is tempting to speculate that a reduction in available acetyl-CoA limits fatty acid synthesis and further modification into the sporopollenin components of the exine, thus resulting in the aberrant exine observed in transgenic plants (Yui et al, 2003).  52 Although genetic induction of male sterility is proving to be a useful tool in many crop species, there remains a need to maintain the male sterile line for ongoing breeding programs. This can be accomplished by continual transformation into desired plants, which is time and resource consuming, or by maintaining the sterile line in a heterozygous, fertile state. The latter strategy requires much growing space and active selection, as only one quarter of the progeny (from a self-pollination by the heterozygote) will be sterile. In contrast, the male sterility seen in the cer6 mutants is reversible by growth in high humidity conditions (Hulskamp et al, 1995). Conditional sterility would allow the use of the male sterile plants in breeding systems and their propagation under permissive conditions. This self-propagation would allow for savings in space, maintenance, and time. While the conditional sterility of the cer6 plants is optimal for breeding systems, their altered wax phenotype is not desirable. The cuticular wax on the aerial tissues is involved in protecting the plantfromUV radiation, non-stomatal water loss, and in plant-insect interactions (Eglinton and Hamilton, 1967; Eigenbrode, 1996). The importance of cuticular wax to the plant eliminates the use of cer6 mutants directly in breeding systems. However, as with other transgenic male sterile strategies, expression of CER6 could be silenced by antisense suppression exclusively in the tapetal cells of the developing anther. This localized silencing would mimic the conditional sterility of the cer6 mutants while maintaining a wildtype wax load. Silencing expression of a given protein by antisense inhibition however, occurs only at a lowfrequencyin transformed populations (Ratcliff et al, 1997). The mechanism of antisense inhibition has been shown to be a double stranded RNA mediated post-transcriptional gene silencing.  53 1.7 Post-transcriptional gene silencing  Post-transcriptional gene silencing (PTGS), also known as RNA interference (RNAi) in animals and quelling in fungi, was originally identified in Petunia plants engineered to overexpress the gene encoding chalcone synthase and generate deep purple flowers . Unexpectedly, some transgenic flowers showed regions of white colouring or were entirely white (Napoli et al, 1990). It was observed that this white colouring was accompanied by a reduction in both transgenic and endogenous CHS mRNA. Subsequent revertants showed purple colouring, correlated with an increase in CHS mRNA levels (Napoli et al, 1990). The phenomenon of RNAi has since been extensively studied in Drosophila and C. elegans, and plant PTGS is only now being extensively investigated. Neibel et al (1995) observed that silencing a B-l,3-glucanase transgene also resulted in silencing of the endogenous gene expression. Subsequent analysis revealed no change in the nuclear transgene mRNA, but a dramatic decrease in levels of mRNA in the cytoplasm (Niebel et al, 1995). Analysis of Petunia lines showing silenced CHS activity indicated a region of homology between the 3' region of the CHS gene and its 3'UTR. These regions were predicted to form an RNA duplex and trigger a cycle of RNA endonucleolytic cleavage and degradation of mRNAs sharing sequence identity with this aberrant RNA molecule (Metzlaff et al, 1997). Indeed, PTGS (or RNAi) has been shown to be triggered by the formation of dsRNA molecules in plants (English et al, 1996; Waterhouse et al, 1998; Wang and Waterhouse, 2000), Drosophila (Hammond et al, 2000), and C. elegans (Fire et al, 1998). The mechanism of RNAi silencing is believed to act as a defence against viral replication and against transposable elements. Approximately 90% of plant viruses have  54 ssRNA genomes that replicate via dsRNA and class I transposable elements amplify via a reverse transcription of an RNA intermediate (Waterhouse et al, 2001). The role of the RNAi degradation pathway in viral defence is supported by the observations of Mourrain et al. (2000) that sgs2 and sgs3 mutants of Arabidopsis are both impaired in PTGS and more susceptible to viral infection. The current model (Figure 1.7) for the PTGS/RNAi pathway involves the recognition of dsRNA in the cytoplasm by a multi-domain enzyme Dicer. The Dicer family of enzymes are endonucleases specific for dsRNA that contain a helicase domain, two tandem RNaselll domains, a PAZ domain, and dsRNA binding domains (Bernstein et al, 2001). The PAZ domain is so named because of three of the Drosophila proteins in which it is found: Piwi, Argo and Zwille/Pinhead (Cerutti et al, 2000) and binds to the 2nt 3 'overhang of the siRNAs (Lingel et al, 2004; Ma et al, 2004). There are four Dicer-like enzymes in the Arabidopsis genome (Denti et al, 2004). The two RNaselll domains cleave the dsRNA to form short interfering RNAs (siRNAs) (Hamilton and Baulcombe, 1999). These siRNAs were originally identified as 25 nucleotide (nt) fragments (Hamilton and Baulcombe, 1999) then classified as 21 nt fragments (Waterhouse and Helliwell, 2003) and have since been shown to include both a 21 nt and a 24 nt class with a 2-3 nt 3' overhang (Hamilton et al, 2002; Zamore, 2002; Tang et al, 2003). Furthermore, Tang et al. (2003) demonstrated that, in wheat germ extract, the 21nt and the 24 nt classes of siRNAs are generated by different members of the Dicer family. By using radiolabeled dsRNA, this study also provided the first definitive evidence that siRNAs are, in fact, generated through the cleavage of dsRNA, as previously hypothesized (Tang et al, 2003). These siRNAs are responsible for the  55 sequence specific degradation of mRNA in the cytoplasm (English et al, 1996; Hammond et al, 2000). The siRNAs are passed in an inactive state from the Dicer enzyme to an RNA induced silencing complex (RISC). RISC is a multicomponent nuclease complex that contains the siRNAs, as well as members of the Argonaute family of proteins (Hammond et al, 2001). In Arabidopsis, argonautel (agol) plants have severe developmental defects and are infertile (Bohmert et al, 1998). In addition, they are impaired in PTGS and show a decrease in transgene methylation (Fagard et al, 2000). This interesting mutant phenotype indicates a possible additional role for PTGS in the regulation of gene expression during development. Although the function of Argonaute proteins remains unknown, the presence of a PAZ domain suggests an interaction between the Dicer and Argonaute enzymes at the site of this conserved domain (Hammond et al, 2001). In the RISC complex, the ds siRNAs are separated into single stranded components and used to guide the complex to homologous mRNA (Hammond et al, 2000). RISC then cleaves the mRNA at a point corresponding to the middle of the siRNA fragment. This cleavage may result in a destabilization of the mRNA and cause further degradation (Ahlquist, 2002) . Evidence for the siRNA component mediating the sequence specific cleavage comes from the correlation between silencing and siRNA accumulation. When no silencing is observed, there is no corresponding accumulation of siRNAs (Escobar et al, 2003) . Additionally, treatment of active fractions with RNaseA before addition of mRNA eliminates the degradation whereas treatment with DNase has no affect (Bernstein et al, 2001).  56 The sequence-specific degradation of transgene and endogene mRNA has been shown to spread throughout the plant (Palauqui et al, 1997; Voinnet and Baulcombe, 1997; Crete et al, 2001; Mlotshwa et al, 2002; Mallory et al, 2003). The exact nature of this mobile signal, however, remains unclear. Boutla et al. (2002) identified RNA as playing a role in the mobile signal. RNA was isolated from GFP transgene-silenced Nicotiana benthamina and fed to GFP expressing C. elegans. Introduction of the plant 80-90 nt RNA fraction induced an RNAi response in the worms. These results both implicate  RISC  •  + ATP  RISC  •  + ATP  ^^^^^^LJ  +siRNA  Methylation m R N A degradation  Figure 1.7: Current model for t h e method of P T G S / R N A i m R N A silencing a n d D N A methylation  57  RNA in the signal transmission of RNAi and expose a possible conserved mechanism of RNAi and PTGS (Boutla et al, 2002). In contrast, Hamilton et al (2002) found differential activity between the 21nt and 24 nt classes of siRNAs. The 21 nt class of siRNAs were essential for sequence specific degradation but were not required for signalling or for DNA methylation whereas the 24 nt class showed opposite activities (Hamilton et al, 2002). Intriguing resultsfromVionnet et al (1998) showed spread of the mobile signal through cells not carrying homologous copies of the target gene. Furthermore, the phloem channels were implicated in the long-distance spread of silencing whereas plasmodesmatal connections acted to spread the signal between adjacent cells. Finally, meristems (vegetative, floral, and root) remained unsilenced, perhaps indicating an inability of the signal to penetrate the meristem (Voinnet et al, 1998). An RNA dependent RNA polymerase (RdRp) (Tang et al, 2003) has also been identified as an essential component of the RNAi/PTGS pathway in certain instances (Boutet et al, 2003; Ahlquist, 2002). However, the role of the RdRp remains enigmatic. It has been suggested that this enzyme may be required for the spread of RNAi by amplifying the mobile signal but that it may not be necessary for RNAi degradation in instances where the RNA forms a direct stem-loop structure (Boutet et al, 2003; Tang et a/., 2003). Much recent research has focused on the role of the RNAi pathway in DNA modification. Frequently PTGS is correlated with an increase in methylation of the transcribed regions of transgenes and their corresponding endogenous genes (Vaucheret  58 et al, 1998). Moreover, mutants impaired in PTGS also show a decrease or absence of this methylation (Morel et al, 2000; Chan et al, 2004). A DNA-methyltransferase (MET1/DDM2) is required for PTGS silencing in Arabidopsis as metl mutants exhibit an inhibition of PTGS (Morel et al, 2000). Whereas PTGS methylation is generally localized to the transcribed region of the gene (Morel et al, 2000), a role for the RNAi pathway in the initiation of transcriptional gene silencing (TGS) is emerging. TGS involves the methylation of promoter regions of DNA (Morel et al, 2000; Vaucheret et al, 1998; Carmichael, 2003). Correspondingly, 21 nt siRNAs have been identified in the nucleus oi Arabidopsis (Papp et al, 2003). These siRNAs could potentially act as signals to activate and direct the DNA methyltransferases and histone modifying enzymes involved in TGS (Matzke et al, 2004). The phenomenon of RNAi/PTGS can be exploited in research to generate loss of function phenotypes (Lawrence and Pikaard, 2003). Constructs designed with a transgene in inverted repeat formation are effective in initiating RNAi degradation (Waterhouse et al, 1998; Lacomme et al, 2003; Wang and Waterhouse, 2000; Smith et al, 2000). Furthermore, separating the inverted repeats by an intron has been demonstrated to enhance the generation of dsRNA required to initiate the RNAi pathway (Smith et al, 2000; Wesley et al, 2001; Brummell et al, 2003). The intron is spliced out during pre-mRNA processing. It is believed that theconditions within the splicosome that hold the RNA molecules in close proximity stabilizes the formation of the dsRNA moiety necessary for PTGS (Smith et al, 2000; Wesley et al, 2001).  59  1.8 Thesis objectives Since the initiation of this project, an understanding of the mechanisms of PTGS has grown considerably. The goal of this project was to localize the PTGS silencing of the CER6 gene to the tapetal cells of the anther, where the pollen coat is produced. This localized silencing would generate plants that were phenotypically identical to wildtype with respect to stem wax load, but were conditionally male sterile when grown under low humidity conditions. In order to localize this silencing, a promoter driving expression in this tissue at the proper temporal stages was identified and its expression patterns confirmed. Chapter 3 discusses this expression pattern confirmation and the phenotypic analysis of the conditionally sterile primary transformants. At the onset of this project there were few reports on the stability of inheritance of PTGS. Several papers reported stable inheritance of their silencing phenotypes for up to eight generations (Chuang and Meyeroitz, 2000; Stoufjesdijk et al, 2002). Since this time, anecdotal evidence suggests that the silencing phenotype is frequently lost in subsequent generations. A n increasing understanding of the role of PTGS pathway components in D N A methylation may explain the suppression of transgene-induced PTGS. Chapter 4 includes the examination of T and T plants that have reverted to 2  3  wildtype fertility, considering phenotype, CER6 transcript accumulation, aberrant transgene transcript accumulation, and D N A methylation.  60  Chapter 2: Materials and Methods 2.1 Plasmid DNA preparation , DNA gel purification, DNA ligation, and DNA sequencing  Plasmid DNA was purified using QIAprep miniprep purification kits (Qiagen) and carried out according to manufacturer protocols. DNA was excisedfromagarose gels and purified using QIAquick nucleic acid purification kits (Qiagen) as directed by the manufacturer. All ligations were carried out using T4 DNA ligase (Invitrogen, Carlsbad, CA) as specified by the manufacturer. DNA sequencing was carried out at the University of British Columbia Nucleic Acid and Protein Service (NAPS) unit by automated Prism Cycle Sequencing. 2.2 Bacterial growth conditions Chemically competent Escherichia coli cells (XL1-BLUE) were prepared by treatment with 0.1M MgCl followed by 0.1M CaCl and resuspended in 0.1M CaCl 2  2  2  with 15% glycerol. 100 uL aliquots of the cell suspension were instantly frozen on dry ice and stored at -80°C until needed. Before transformation, cells were removed from -80°C and allowed to thaw on ice. Plasmid DNA was then added, mixed, and allowed to sit on ice for 5-10 minutes. Cells were heat shocked at 42°C for 45 seconds, and incubated with shaking at 37°C for 1 hour. Various amounts of cell suspension were spread on plates containing LB medium, agar (7g/500mL), and the appropriate antibiotic for selection, and allowed to incubate overnight at 37°C. All E. coli overnight cultures were grown at 37°C with shaking for 14-16 hours. Competent Argobacterium tumefaciens cells (GV 3101; pMP90) were grown in LB medium containing 25mg/L rifampicin and 25 mg/L gentamicin, and prepared by  61 treatment with 20mM CaCb, and frozen instantly in liquid nitrogen in 100 uL aliquots. Competent cells were stored at -80°C until required. Transformation with binary vectors was carried out by adding plasmid DNA tofrozencells and thawing at 37°C for 5 minutes with periodic mixing. LB broth was added and cells were allowed to grow at 28°C for 2-3 hours with gentle shaking. Various amounts of cell suspension were plated on plates containing LB medium, agar (7g/500mL), and the appropriate antibiotics and allowed to grow for 48 hours at 28°C.  2.3 Nucleic acid analysis 2.3.1 Isolation of Arabidopsis genomic DNA Genomic DNA was isolated according to Edwards et al. (1991). Two young rosette leaves were harvested and stored at -80°C. Frozen tissues were ground into a fine powder in a 1.5mL eppendorf tube. 500uL of extraction buffer (200mM Tris-HCl pH 7.5, 250mM NaCl, 25mM EDTA pH 8.0, 0.5% SDS) was added to each sample, vortexed and centrifuged for 3 minutes at maximum speed. DNA was precipitatedfromthe supernatant by addition of 300uL isopropanol. Samples were centrifuged for 2 minutes, washed with 70% ethanol, allowed to air dry and resuspended in lOOuL TE pH 8.0. 2.3.2 Isolation of Arabidopsis total RNA Flower buds of wildtype Columbia, transgenic, and cer6 plants were staged based on their location in the inflorescence, and physical characteristics described in Table 1.1. Three buds from each of stages 9, 10, 11, 12, and 14 were removed and immediately placed in an eppendorf tube in a dry ice bath. Frozen tissues were transferredfromthe dry ice bath to -80°C for storage until RNA extraction. Total RNA was extracted from  62 these tissues using Trizol Reagent (Invitrogen, Carlsbad, CA) and isopropanol precipitation modified from manufacturer's protocol to accommodate the small amount of tissue present. Tissues were ground in eppendorf tubes placed in a dry ice bath to prevent thawing. Volumes of reagents used were scaled down because of the small amount of tissue, but ratios of reagents and incubation times remained as specified by the manufacturer. Finally, RNA was resuspended in 12uL RNase free H2O.  2.3.3 RNA quantification and Reverse Transcriptase (RT) reaction RNA was quantified using an Ultrospec 3000pro UV/visible spectrophotometer (Biochrom Ltd., Cambridge, UK). To standardize the reactions, 300ng of total RNA was used for each reverse transcriptase (RT) reaction. DNases were removed using DNase I (Invitrogen, Carlsbad, CA). Reactions were carried out with Superscript II Reverse Transcriptase (Invitrogen, Carlsbad, CA) as per manufacturer's specifications. 2.3.4 Optimization of PCR from cDNA template Because of the small amount of tissue used, there was generally a low yield of total RNA. For PCR reactions, a volume of cDNA template corresponding to 25 ng total RNA was calculated and used for all PCR reactions. PCR conditions were optimized using this volume of template from wildtype tissue with both CER6 and GAPC primers (reactions were carried out in separate tubes). GAPC encodes the cytosolic form of glyceraldehyde-3-phosphate dehydrogenase (Shih et al., 1991) and is used as a loading control. Accumulation of amplification products was examined for a range of total cycles ranging from 20-32. A protocol with 24 cycles was chosen because this resulted in  63 products within the log phase of accumulation for both CER6 and GAPC when amplified from wildtype tissue.  2.4 Polymerase Chain Reaction (PCR) All PCR reactions were carried out in 0.5 mL eppendorf tubes, using a DNA Thermal Cycler 480 (PerkinElmer Instruments, Norwalk, CT). 2.4.1 Amplification of the CER6 3'UTR The 3' untranslated region (UTR) of CER6 was amplifiedfromgenomic DNA using oligonucleotide primers containing restriction enzyme recognition sequences. The forward primer contained sequences recognized by Xbal (underlined) and EcoRI (double underlined) 5TTGGTCTAGAGAATTCGTCTTTGAACGGTTTAGTAACGGTTG3' and the reverse primer contained sequences recognized by C M (underlined) and Kpnl (double underlined) 5'GGAAATCGATGGTACCTACTGCTATCGGATCCTCAAACTGG3'. Amplification was carried out in a 50 jxL reaction with Taq DNA polymerase (Invitrogen, Carlsbad, CA) under the following conditions: initial denaturation at 94°C for 4 minutes followed by 30 cycles of denaturation at 94°C for 40 seconds, annealing at 60°C, and extension at 72°C, followed by a final extension at 72°C for 7 minutes. 2.4.2 Amplification of the MS2 promoter A 1.2 kbp region immediately upstream of the MS^.gene, hereafter referred to as the MS2 promoter, was amplifiedfromgenomic DNA using oligonucleotide primers. The forward primer contained both SstI (underlined) and Notl (double underlined) restriction enzyme cutting sites  64 5' ATAACACGTCGCGGCCGCAATTAGACAATACCCTAT3' and the reverse primer contained an Xhol cutting site (underlined) 5' GCGCCTCGAGAGAGCCTCCATCACAAGCT3'. Amplification was carried out in a 50uL reaction using Pwo DNA polymerase (Invitrogen, Carlsbad, CA) under the following conditions: Initial denaturation at 94°C for 2 minutes and 30 seconds followed by 30 cycles of denaturation at 94°C for 40 seconds, annealing at 60°C for 30 seconds, and extension at 72°C for 1 minute and 30 seconds, followed by a final extension at 72°C for 10 minutes. 2.4.3 PCR from cDNA template Amplification of a region of CER6 and GAPC for expression analysis by RT-PCR was carried out using oligonucleotide primers CER6 cDNA forward 5'ATCTAGCCCGCGACTTGCTC3'and reverse 5' CACTTGAAACCACTCCCAAACG3' (Hooker et al., 2002) and GAPC forward 5' TCAGACTCGAGAAAGCTGCTAC 3' and reverse 5' GATCAAGTCGACCACACGG 3' (Western et al, 2004). PCR amplification was carried out in a 20 uL reaction using Taq DNA polymerase under the following conditions: an initial denaturation of 30 seconds at 94°C followed by 24 cycles of 30 second denaturation at 94°C, 30 second annealing at 60°C, and 45 second extension at 72°C, followed by a final extension at 72°C for 7 minutes. Amplification products were run on a 1% (w/v) agarose gel, stained with ethidium bromide and visualized in a Multilmage Light Cabinet (Alpha Innotech Corporation, San Leandro, CA). Band fluorescence was quantitated using Alphalmager 1200 software.  2.4.4 Amplification of aberrant transcripts in T2 and T3 generations Primers were designed to test for the presence of aberrant RNAi transcript. These primers, forward 5' GATCGAACATGTAAGATGA 3' and reverse 5' ATTCATGTTCGACTAATTC 3' amplified across the pdk intron present in pHannibal and separating the CER6 3'UTR inverted repeats. PCR amplification was carried out in a 20uL reaction using Taq DNA polymerase (Invitrogen, Carlsbad, CA) under the following conditions: an initial denaturation of 30 seconds at 94°C followed by 30 cycles of 30 second denaturation at 94°C, 30 second annealing at 50°C, and 1 minute 30 second extension at 72°C, followed by a final extension at 72°C for 7 minutes. 2.4.5 Amplification to test for methylation in T and T generations 2  3  Primers were designed to test for methylation of the transgene. Genomic DNA from selected plants was digested with Mbol or iSa«3AI (New England Biolabs, Beverly, MA). These two enzymes both recognize the sequence 5'GATC3' but differ in their sensitivity to DNA methylation. Mbol cleaves DNA regardless of the presence/absence of methylation whereas Sau3Al does not cleave methylated DNA. DNA was digested for 3 hours at 37°C as per manufacturer's protocol. To test for methylation within the transgene sequence, primers from section 2.4.4 to amplify the pdk intron across an MboVSau3AI recognition sequence. To test for methylation within the promoter sequence, primers were designed to amplify a section of the promoter and insertion that contained an Mbol/Sau3Al recognition sequence. The forward primer, 5 'TATAATTTTTATGATGACC AT3' hybridizes 800bp upstream of the ATG start codon and the reverse primer, 5'AAAGTCAGCCTCCATCA3' hybridizes at the promoter/insertion border. PCR amplification for both primer sets was carried out in a  66  20uL reaction using Taq DNA polymerase (Invitrogen, Carlsbad, CA) under the following conditions: an initial denaturation of 30 seconds at 94°C followed by 40 cycles of 30 second denaturation at 94°C, 30 second annealing at 52°C, and 1 minute extension at 72°C, followed by a final extension at 72°C for 7 minutes. 2.5 Construct design 2.5.1 Generation of RNAi construct The PCR products from the amplification of the MS2 promoter and the CER6 3'UTR were run on a 1% agarose gel. Both amplification products, the 397 base pair CER6 3'UTR and the 1250 base pair MS2 promoter, were excisedfromthe gel and purified (Figure 2.1 A) as stated in section 2.1. The purified CER6 3'UTR amplicon was then digested with EcoRl and Kpnl (Invitrogen. Carlsbad, CA) and ligated into the corresponding sites of the pHannibal vector (Figure 2.1 B; Wesley et al, 2001). The resulting construct (Figure 2.1 C) was transformed into E.coli and selected for using resistance to the antibiotic ampicillin (lOOmg/L). Selected colonies showing resistance to ampicillin were grown overnight in 3mL cultures containing ampicillin (lOOmg/L) at 37 °C, the plasmid isolated, and presence of the insert was verified by test digestion with Noil. A plasmid containing the CER6 3'UTR in the forward orientation was selected for further cloning. The sequence and orientation of this, and all subsequent clones were verified by DNA sequencing. Both this plasmid and the purified (uncut) CER6 3'UTR amplicon were digested with Clal and Xbal (Invitrogen, Carlsbad, CA) and the products run on a 1% agarose gel. The digestion products were excised, purified and ligated in the antisense orientation using the Clal and Xbal sites of the vector as outlined above (Figure 2.1 D).  67  Xbal EcoR  P"\ C/al  Notl  K  Ssf  CER6 3'UTR  B  Xhol  MS2 promoter  Xhol  Ssf  EcoR\  C a M V 3 5 S promoter  Ssf  Kpnl  C/al  Xbal  N  o  f  i  Not\  Xhol  Clal  EcoR\  Not\  •  Xbal  Notl  Kpnl  CER6 3'UTR  C a M V 3 5 S promoter  OCS  pdk intron  pdk intron  OCS  D Xhol  Notl  Ssf  Kpnl  C a M V 3 5 S promoter  Xbol Ssfl.  EcoRl  CER6 3'UTR  MS2 promoter  pdk intron  Notl  EcoRl  Xbal  CER6 3'UTR  OCS  EcoRl Kpnl  Notl  cm  Kpnl  CER6 3'UTR  Clal pdk intron  EcoRl  CER6 3'UTR  X  b  a  \  OCS  Figure 2.1: Cloning strategy for generation of the CER6 3'UTR RNAi construct. 397 bp of the CER6 3 'UTR and 1.2 kbp of the MS2 promoter were amplified by PCR using oligonucleotide primers (Table 2.1) that contained restriction sites (A) to allow cloning into the pHannibal vector (B). First the CER6 3'UTR was inserted in the sense orientation (C) followed by the antisense orientation (D). Finally the MS2 promoter was subcloned into pHannibal (E). The entire cassette was excised using the Notl sites and cloned into the pART27 binary vector for plant transformation.  Notl  68 Again, transformed E.coli were selected by resistance to ampicillin, grown in overnight cultures, and the presence of the CER6 3'UTR in both forward and reverse orientations was verified by test digestion. A plasmid containing the CER6 3'UTR in both orientations was selected for further cloning. Both the MS2 promoter amplicon and the pHannibal 2xCER6 UTR plasmid were digested with SstI andXhol (Invitrogen, Carlsbad, CA), run on a 1% agarose gel and purified as previously outlined. The resulting fragments were ligated using T4 DNA ligase (Invitrogen, Carlsbad, CA), transformed into E.coli and selected for based on their resistance to ampicillin. The presence of all three inserts were verified by test digestion. Figure 2.1 E shows the completed RNAi construct with the CER6 3'UTR in both forward and reverse orientations separated by the pdk intron contained in the pHANNIBAL vector, driven by the 1.2 kbp fragment of the MS2 promoter. 2.5.2 Generation of MS2 promoter::GUS construct  The plasmid generated in the previous section was digested with Xbal and Xhol to remove the inverted repeat of the CER6 3'UTR. The digestion products were run on a 1% agarose gel and the band corresponding to the pHannibal iproMS^ construct was excised and purified. Into this construct, a 1876 bp fragment containing the Bglucuronidase (GUS) gene was ligated. The resultant plasmid was transformed into chemically competent E.coli and selected for based on resistance to ampicillin (lOOmg/L). Insertion of the GUS gene was verified using test digestion. 2.5.3 Generation of vectors for plant transformation  Both the RNAi and the GUS vectors were digested with Not! to excise the respective cassettes. These cassettes were gel purified and ligated into the binary vector  69 pART27 (Wesley et al, 2001). The resulting plant transformation vectors were transformed into chemically competent E.coli and selected for based on resistance to the antibiotic spectinomycin (30mg/L). Selected colonies were grown in overnight cultures, followed by plasmid isolation and restriction digests to verify the presence of the insert. Purified plasmid, each containing either the proMS2::CER6 3'UTR RNAi or the proMS2::GUS cassettes were transformed into Agrobacterium tumefaciens. Transformed Agrobacteria were selected for based on resistance to the antibiotics rifampicin, spectinomycin, and gentamycin. Cultures were grown for 48 hours at 28°C and the presence of the correct insert was verified by plasmid purification and test digestion.  2.6 Plant growth conditions, transformation, and selection 2.6.1 Plant growth conditions  Seeds oi Arabidopsis thaliana were placed on plates containing AT medium, agar (3.5g/500mL), and the appropriate antibiotic for selection (for transgenic seeds only). The seeds were stratified at 4°C for 3 days and then germinated at 20°C under continuous light (100//E m' s photosynthetically active radiation). After fourteen days seedlings 2  1  were transplanted into soil (Sunshine Mix 5, SunGro, Kelowna BC) supplemented with AT medium. Plants were grown at 20°C under continuous light until maturity. High humidity conditions were generated by covering the upper portion of the plants with plastic bags that were secured around the stem. Plants were bagged individually in an attempt to reduce mold and mildew formation on both the soil and the plant. After three days slits were cut in the bag, or the bottom of the bag was opened to slowly reduce the humidity and prevent mold growth.  70 2.6.2 Floral dip Arabidopsis plants were transformed using the floral dip method (Clough and Bent, 1998). A 300mL culture of Agrobacteria was centrifuged and the resulting pellet resuspended in a 5% sucrose solution containing 0.05% silwet-L77. This bacterial suspension was placed in a shallow container. Pots containing Arabidopsis plants were inverted and the green tissues of the plants were immersed in the bacterial suspension for 5-10 seconds with gentle swirling. Pots were then placed on their sides and covered to generate low light conditions and left overnight. Pots were then returned to their upright positions and the covering removed. Plants were grown as usual until maturity and senescence, at which point the seeds were harvested. 2.6.3 Screening for transgenic Arabidopsis in the TI and further generations Harvested seed were germinated on AT medium, agar (3.5g/500mL), arid lOOmg/L kanamycin, and seedlings containing the transgene were selected based on their resistance to this antibiotic. Resistant seedlings were selected after fourteen days germination and transplanted to soil. Along with transgenic seed, wildtype Columbia and cer6 (EW8 allele) were always planted as controls for comparison. As plants matured, they were visually examined for stem wax load. cer6 plants have a greatly reduced stem wax load and thus appear bright green and glossy. Transgenic plants were compared against both these glossy green stems and the glaucous white stems of wildtype Arabidopsis and scored as either glossy or waxy. Male sterility was determined visually by comparison to the srriall sterile siliques produced by the cer6 plants. Frequently thefirstsilique produced by wildtype plants was very small, so thefirstsilique of each plant was not included in determining sterility. To  71 allow the siliques sufficient time to mature, plants were evaluated for sterility after they had produced at least 6 siliques. Plants showing small sterile siliques were transferred to high humidity growth conditions to evaluate the reversibility of this sterility. Plants were grown at high humidity for three days and then returned to regular growth cabinet humidity conditions. Siliques produced during this period of high humidity were allowed to mature and evaluated for seed set based on the size of the siliques (as compared to both cer6 and wildtype plants). Upon senescence, seed was harvested from all plants. 2.6.4 Segregation of antibiotic resistance in T and T 2  3  Seeds harvested from the primary transformants (Ti) were germinated on AT medium, agar (3.5g/500mL), and lOOmg/L kanamycin. The number of seedlings resistant to the antibiotic and the number of seedlings susceptible to the antibiotic were counted and recorded. These numbers were evaluated for their fit to the 3:1 segregation ratio expectedfrominsertion of a single copy of the transgene using the Chi-squared goodness of fit test. The Chi-squared value was calculated using the formula A, =I(observed-expected) /expected and this value was compared against the critical 2  2  value of A,(o.o5)=3.841. Plants with a value less than the critical value were accepted as 2  fitting the expected ratios whereas plants with a value greater than the critical value did not fit the expected 3:1 segregation ratios.  2.7 GUS histochemical assays Tissues containing the GUS transgene were removedfromthe mature plant and immersed in GUS staining buffer (50mM NaP0 , 0.5 mM K Fe(CN) , 0.5 mM 4  3  6  K4Fe(CN) , 2mM X-GLUC, and 0.05% Triton-X) and air was removedfromthe tissues 6  72 by vacuum infiltration. Infiltrated tissues were incubated in the staining buffer for 4 hours, or overnight. Stem and leaf tissues were cleared of chlorophyll by overnight immersion in 95% ethanol. Stained flower buds were placed on glass microscope slides and their petals and sepals were moved to allow visualization of the stamens. The chlorophyll was removed by overnight incubation in a clearing solution (72% choral hydrate, 11% glycerol, 17% water). Stained and cleared samples were examined visually with the naked eye and microscopically with an Axioskop 2 plus Routine Microscope (Zeiss, Gottingen, Germany).  2.8 Pollen microscopy  Pollen grains were extracted from mature, senesced flowers. Anthers were excisedfromflowers at stage 13 of development, placed in 1.5 mL eppendorf tubes containing water and vortexed for 1 minute. After vortexing, the remaining anther tissues were removed from the water and tubes were centrifuged for 2 minutes at maximum speed to pellet the pollen. Supernatant was decanted and the pollen was suspended in a Nile red (Sigma, St. Louis, MO) staining solution. Nile red is a lipid-specific stain that fluoresces when excited with light at wavelength of 450-500 nm and generates emissions at 515 nm (Greenspan and Fowler, 1985;Greenspan et al, 1985). A 10 mg/mL acetone dissolved stock was diluted 1/1000 in a 10% v/v glycerol solution. Pollen grains were incubated for 15 minutes at room temperature before being transferred to a glass microscope slide. Samples were viewed using a Leica Light Microscope (Leica, DMR), 13 filter for fluorescence at 450-490 nm, and photographed using the QCAM (Q-Imaging) and Improvision Openlab software.  73  Chapter 3: Results and Discussion - Analysis of TI Generation  3.1 Promoter::GUS fusion confirms anther specificity of the MS2 promoter Transgenic plants carrying the proMS^: ^-glucuronidase (GUS) fusion were generated as outlined in Chapter 2. Plants showing resistance to the antibiotic kanamycin were selected for further analysis. Of these, twelve plants were subjected to GUS activity assays overnight and scored both visually and microscopically. All lines examined showed similar patterns of blue colouration as a result of the GUS expression and thus proMS2::GUS #2 was selected for detailed analysis. Visual inspection of GUS stained transgenic and wildtype Arabidopsis plants showed no detectable staining in tissues other than inflorescences (Figure 3.1). In the transgenic inflorescences GUS activity was  Figure 3.1: Wildtype (A) and transgenic (B) p r o M S 2 : : G U S plants treated with X-gluc show no blue G U S staining in stems (left) or in cauline leaves (right).  confined to the stamens. It was primarily detected in the anthers, with occasional staining in the filaments. GUS activity in the filament was only observed in young buds of stages 8 and 9, and only where staining of the anthers was particularly intense. Aarts et al.  74 (1997) reported MS2 expressionfromthe stage of free microspore release until the completion of exine development, corresponding to stages 10 through 11 of floral development as outlined in Chapter 1. My experiments confirmed that the MS2 promoter does indeed drive GUS expression during stages 10 and 11, however the duration of expression is extended beyond these developmental stages. GUS activity is already detectable in late stage 8 and persists until mid stage 12 (Figure 3.2). Similar results were observed for all twelve transgenic plants examined. This discrepancy in GUS activity could be due to the differences in length of promoterfragmentsused in the two experiments. Aarts et al. (1997) used a 1077 base pairfragmentimmediately upstream of the MS2 coding region for their expression analysis, whereas a 1.2 kbp region was used for the current analysis. It is possible that this additional 120 bp of the 5' promoter region contains enhancer elements necessary for full expression in Arabidopsis. Blue colouration caused by GUS activity appeared to be localized mainly in the inner tissues of the anther and diminished as the tapetum degenerated (Figure 3.3). Occasionally, GUS activity was also detected in the epidermal cells of young anthers. GUS staining was never observed in treated wildtype pollen, although activity was occasionally detected in transgenic pollen. When pollen staining was observed, it was not uniform as both stained and unstained pollen grains were detected in the same anther. This result is consistent with the observation by Aarts et al. (1997) that mature pollen stained in approximately a 1:1 ratio. Although GUS activity was detected in maturing pollen grains, mature pollen in this assay had no blue colouring. All evidence of GUS activity disappeared by the end of stage 12 of floral development, immediately preceding anther dehiscence and floral bud opening.  The observed colouration of the pollen grains  o  co 0  E  J3  CO  0  0  CO ••-»  OI CO  (fl  0  0  (fl  CD -t—'  .Q CO  0)  5 S g <o -C  c  > N  -Q  •o 0 I—  CO 0 CL CL C0  d E <D JO  I  0)  o T—  I  0  CT CO +-I  (fl  CM  0  =  3  CO  W  X "co I  CO TO J=  E "  CD C  £ 0  CO _Q  CD  CD CO ZJ  •4-1  CL '  CO  0  CM  o 0  u CO  (fl  CNJ  -  0  (3  0 I  'c  'co  -c  OI CO  (fl  CO  cl "D  I  I  CN  0  O) CO  (fl  i2  CO  CO  CO a ^ •4—  1  •5.  .2  Q.  •g 'E  « "2 ^ >CD CO £ —  co co qs CO  —  CO  c  =3  CL CO  0  ZD  0  CO =3 • C  0  c  01 cB o o ^  0  U, CO  (fl  CO  o I CM  0  CO CO  (fl  o c o  M -  CO  o  -..  CO <->  (fl  (fl  T3  -Q  0 JO  I 0  Ui CO  (fl  >  ^ CO  O  q=  |  co  «  gp  00 0  CO  CO  U—I  CO CO  0 0  CO  co CO CO c c  •c  CO  X!  C>O CO c b c CO  X  o  CD  C  CN  C  CO  CL  0  Cfl  2  0  CL  O  c  oo jS  15 co 9- ^ COZ)  CO LU  c CO  c  c ° C O CO CO 0 c  oo 0  CL  2CL °0  il O  CO t+=  0  10  CT) CO  0  CT ro CO  ro •  CM  0  ro to  ro  => ro 0c - Q <=> E 0CT)TO 0 l_ —  0 0 TJ CO 0 0  Q.  ro £  CL CL  o o :» . c 0 ro o +- 3 x : c CT CO to — o "C0T i 2 CO N CT) —CM ro X. X o CD C f O > 0 0 I |  > nS  CO  CN  0 CT  ro  *-> CO  0 CT)  ro CO  c o *J 0 **  « —  CO  E g ro 0  -fc i-  §>|c0  Q.  CO  CL  c ro ro .E to  CM  ro  CM  -E Cl 0 TO o CO CO  C 05 0c CO  CT) 0 ro  O 0 X CD ro co CO <^ « 0  §? CN T— CO  0  CT ro  +J  co  0  ro co  g8  -t—I  CL CL CO CO '"D  ^ CT) O CO i _ H—< Q . CO  O a) "S - C CO  £ 1  15 ° .9. ro  > ro  LU  CO  o w CO CO  c  CO  1_  TO o CL  ro  3  ro  d ro CT)C O  0 c to HE ' c >*- •F CO o c _> ro ty)  co  CO co c  CO CO CL  f ®  ro +•>  co  CO CO -t—  1  co  -*—'  _ 0  1 i CN  0 CT)  0  CO CO  E ' cc 0 ^ o 0 "O 0 (/) C L -Q Q. CO X. 00 vu E iO— O ra 13 CM 0 H—* 0 o CO c CO "C  cd "C O 0  C  ro CL  *"*  = c o) ro LL to  0  o  £1 o  '-P  77 could indicate sporadic activity of the MS2 promoter in pollen grains, or, as is more likely, an uptake of GUS enzymes from the locule into the pollen grain. The contrast between these results and those reported by Aarts et al. (1997) could be due to the slightly different times at which the pollen grains were examined. Aarts et al. (1997) examined 'mature' pollen grains inside the anther before anther dehiscence, at a time when the pollen grains are still undergoing desiccation and are not yet fully mature. Here mature, dessicated pollen was examined in anthers that had undergone dehiscence, and no blue colouration was ever observed. Potentially, GUS enzymes are taken up from the locule into the pollen grain and degraded as the pollen proceeds through the final stages of maturation and desiccation.  3.2 Transgenic plants expressing the RNAi construct are conditionally male sterile Eighty transgenic plants carrying the RNAi construct targeted against CER6 expression in the tapetum were selected on the basis of their resistance to the antibiotic kanamycin and examined visually for stem waxiness and male sterility as indicated by small siliques (Dellaert et al, 1979; Koornneef et al, 1989). All plants examined showed levels of stem glaucousness similar to that of wildtype Columbia plants. Representative stem glaucousness of an RNAi transgenic line in comparison to wildtype and cer6 wax-deficient mutant plant is shown in Figure 3.4a. Of the eighty transgenic plants analyzed, sixteen had short undeveloped siliques (Figure 3.4b), suggesting male sterility, and an additional twelve plants developed both long wildtype-looking and short siliques. A selection of primary transformants RNAi#72 through to RNAi #80 are shown in Figure 3.5. Many of the transgenic plants are phenotypically similar to wildtype,  78 indicating they are fertile. The siliques of RNAi#75, however, are phenotypically similar to that of the cer6 control plant. A combination of small sterile and large fertile siliques is seen in RNAi#79. In this case, there appears to be one revertant branch showing fertility. The siliques of RNAi#76 appear smaller than those of the wildtype controls but still larger than those of cer6. Close examination of these siliques indicated that seed formation occurred and they were thus not counted as sterile plants.  Figure 3.4: RNlAi plant #7 as representative of stem waxiness (A) a n d male sterility (B).  The development of both fertile and sterile siliques on RNAi expressing transgenic plants is consistent with the reported observations of Preuss et al. (1993) that cer6 plants often produce one or two fertile siliques and that only the extremely severe cer6-2 allele is completely sterile. Furthermore, it has been noted that fertility is more easily restored in mutants lacking CER6 activity than is stem waxiness, indicating that a lower threshold of CER6 activity is required for pollen coat formation (Fiebig et al, 2000). Dehio and Schell (1994) described that post-transcriptional gene silencing of a RolB transgene in Arabidopsis generated a range of intermediate phenotypes from fully silenced to fully expressed. This range of silencing efficiency typically obtained in RNAi silencing experiments coupled with the low threshold of CER6 expression required for pollen fertility could explain why fifty-two of the eighty transgenic plants appeared to have wildtype siliques. Additionally, investigations into the technique of PTGS indicate that there may be an initial period of transgene expression followed by an increase transcript reduction (Waterhouse et al, 2001). This initial lag time of transgene expression coupled with the low threshold of activity required for pollen fertility could result in the occasional fertile silique. A short burst of CER6 activity may provide sufficient pollen coat lipids for fertility of the pollen grains. However, as all transgenic plants have waxy stems, the silencing signal does not appear to have spread throughout the plant, making silencing in each stamen an independent event. Thus it would be possible, if not expected, to find both sterile and fertile siliques on the same plant.  80  Figure 3.5: Examples of siliques of primary transformants RNAi#72 through RNAi#80 w h e n grown at low humidity. R N A i # 7 5 has short undeveloped siliques similar to those of cer6 and RNAi#79 has both sterile and fertile siliques.  Reversibility of the male sterile phenotype was examined by transferring the plants to conditions of high humidity. Figure 3.6 shows representatives of plants where fertility was restored (Figure 3.6A) and where fertility was not restored (Figure 3.6B). Of the twenty-eight plants showing sterile siliques, treatment with high humidity did not restore fertility in four of the plants. The siliques of these plants remained small and empty and no seed was formed. The results of the preliminary analysis suggest that PTGS mediated silencing of the CER6 condensing enzyme is restricted to the anther and generates conditionally male sterile plants. The waxy stems observed in all transgenic plants indicate that silencing of  81 CER6 is localized to the anthers of transgenic plants. Thus, the reported mobile signal involved in PTGS silencing is not observed in these plants. This lack of a spread of silencing could be due to the short developmental time span in which the inverted repeat transgene is expressed. Indeed, GUS analysis (see section 3.1) indicates that the transgene would only be expressedfromlate stage 8 of floral development to mid stage 12. According to the stages of floral development outlined by Smyth et al. (1990) this time accounts for less than half of the time of bud development and approximately one quarter of the timefromfloral organ initiation to floral organ senescence. Moreover, the tapetal tissue in which expression of the silencing transgene is expressed undergoes programmed cell death at its maturity, effectively killing any signal produced therein. Thus it is conceivable that the time span and the tissue in which the silencing construct are expressed are such that the mobile signal either cannot develop or cannot spread. The appearance of both conditionally sterile and fertile plants in the primary transformants indicates that there is a threshold level of CER6 expression above which plants are fertile and below which plants are sterile. This hypothesis is supported by the incomplete penetrance of the sterility phenotype in many of the cer6 alleles. It is possible that even low activity of CER6 condensing enzyme allows for production of a pollen coat sufficient to initiate hydration and maintain fertility. Because the fertility can be restored in a manner identical to cer6 plants implies that the generated sterility is indeed due to the lack of CER6 activity in the anthers of the transgenic plants.  82  Figure 3.6: Examples of primary transformants RNAi#4 (A), RNAi#13 (B), and RNAi#30 (C) w h e n grown at high humidity. RNAi#4 and RNAi#30 s h o w restored fertility w h e r e a s fertility in R N A i # 1 3 is not restored.  3.3 R T - P C R of CER6  transcript in wildtype buds  RNA was isolated from the buds of wildtype Columbia flowers at stages 9, 10, 11, 12, and 14 of floral development (Smyth et al, 1990; Table 1.2) and analyzed by RTPCR as outlined in Chapter 2. Resultsfromwildtype buds show high expression of CER6 through all developmental stages examined, with slightly greater transcript levels detected at stages 12 and 14 (Figure 3.7). Stages 9 and 10 of floral development  83 correspond to the deposition of the callose wall around the MMCs and the release of the microspores from the tetrad respectively. CER6 expression has been localized to the epidermal cells of the anther until the time of microspore release from the tetrads (Hooker et ai, 2002). Additionally, tapetal degeneration begins during stages 11-12 of floral development and the tapetum is completely absent at stage 14. Thus, RT-PCR analysis of CER6 transcript levels at stages 9 and 10 may indicate epidermal, not tapetal expression. Similarly, expression of CER6 at stage 14 cannot be attributed to tapetal expression. At this stage of floral development, the flower is completely open and the stigma may be beginning to extend. CER6 expression at this stage may be attributable to the developing silique. Indeed, the CER6 promoter has been reported to drive expression of the GUS reporter gene in the developing silique (Hooker et al, 2002). For these reasons, buds at stages 11 and 12 of development were chosen for examination of CER6 expression in transgenic plants.  Figure 3.7: R T - P C R analysis of CER6 expression in flower buds of wildtype C o l u m b i a Arabidopsis thaliana at different stages of floral development. GAPC is an internal standard. Expression is strongest at stages 12 and 14.  84 3.4 CER6 transcript accumulation is reduced in conditionally sterile transgenic plants  RNA was isolated from from wildtype and transgenic RNAi Arabidopsis flower buds at stages 11 and 12 and analyzed by RT-PCR. Representative plants showing both fertility and conditional male sterility are shown in Figure 3.8(A). In RNAi#30 and RNAi#40 plants, siliques were short, indicating male sterility and this sterility was reversed by exposure to high humidity. These phenotypic observations suggested that a suppression of CER6 activity was the cause of the conditionally male sterile phenotype. This hypothesis was confirmed with the RT-PCR analysis indicating that CER6 transcript levels were significantly reduced or absent in the sterile plants at stage 12 of floral development. Figure 3.8B shows the relative intensity of the band corresponding to CER6 transcript as compared to the band corresponding to GAPC transcript. While CER6 transcript is reduced at stage 12 in the silenced plants, transcript is still present at stage 11. Examination of the proMS^GUS stamens reveals that it is not until stage 11 that expression is localized to the tapetum. If transgene expression is initiated in the tapetum at sage 11 of floral development, and there is a period of transgene expression (and thus CER6 transcript accumulation) before silencing is initiated (Waterhouse et al, 2001) then an accumulation oiCER6 transcript in transgenic buds at stage 11 could occur. Moreover, as flower development is a continuous process, the high levels of CER6 transcript seen in RNAi#40 at stage 11 could be the result of CER6 transcription in flower buds early in stage 11 before silencing was initiated. The hypothesis that the lack of CER6 transcript results in conditional male sterility is further supported by the accumulation of CER6 transcript to levels similar to wildtype in the fertile plant RNAi#35.  85  A RNAi #30  Plant  RNAi  , RNAi  Stage of floral development CER6  GAPC  B  11  12 wt  11  12  11  12  12  RNAi  RNAi  RNAi  #30  #40  #35  Plant and staqe o f floral d e v e l o p m e n t Figure 3.8: R T - P C R analysis of CER6 transcript levels of wildtype, R N A i # 3 0 , RNAi#40, and RNAi#35 primary transformants at stages 11 and 12 of floral d e v e l o p m e n t (A). A reduction of CER6 transcript w a s observed in stage 12 of plants showing conditional male sterility (#30 and # 4 0 ) and w a s not reduced in the fertile plant # 3 5 . 6>4PC is a loading control. Relative intensity of CER6 bands compared to GAPC bands (B).  86 Interestingly, at stage 12 of floral development, CER6 transcript was abundant in RNAi#75, which had all sterile siliques, but absent in RNAi#79 despite this plant having both sterile and fertile siliques (Figure 3.5 for plant phenotypes, Figure 3.9 for transcript levels). The presence of CER6 transcript in RNAi#75 indicates that the observed sterility is due to a factor other than the disruption of CER6 activity. Potentially the site of transgene insertion disrupted the function of another gene whose expression is necessary for seed set and silique maturation. Selected pollen grains were extracted from developing anthers, stained with lipid-specific Nile Red stain and viewed at 63x magnification. The results confirm that the pollen of RNAi#75 shows no appreciable differencefromthat of fertile transgenic or wildtype plants (Figure 3.10). The lack of CER6 transcript in RNAi#79 could be due to the small sampling size of three buds per RNA extraction. Since there was one revertant, fertile shoot, it is possible that none of the flower buds selected for RNA extraction camefromthis shoot, thereby excluding the buds producing CER6 transcript. Alternatively, the CER6 transcript could be degraded and the observed fertile siliques could have resulted from varying growth chamber conditions. This latter hypothesis is unlikely because the fertile siliques were observed on only one shoot of this plant. Predictably, if the fertility of the plant were due to an increase of humidity, the effect would be general and not limited to one shoot. Thus, it is likely that buds from this shoot were not accounted for in the RNA extraction and consequently there is an absence of CER6 transcript at stage 12 of floral development of RNAi#79.  87  A  Plant Stage of floral development  wt 11  12 »  CER6 GAPC  RNAi#75 11 •MM*  12  11  12  mmm  m  H I  RNAi#79  M  m •*  B >  'co c c  _ro  o wt  11 12 RNAi#75  11 12 RNAi#79  Plant a n d stage o f floral d e v e l o p m e n t Figure 3.9: R T - P C R analysis of CER6 transcript levels of wildtype, RNAi#75, and RNAi#79 primary transformants at stages 11 a n d 12 of floral development (A). A reduction of CER6 transcript w a s observed in stage 12 of both plants with complete elimination in RNAi#79. GAPC is an internal control. Relative intensity of CER6 bands compared to GAPC bands (B).  88  Figure 3.10: Microscopic analysis of wildtype Columbia and transgenic T1 plants expressing the CER6 silencing transgene. Pollen grains were stained with Nile R e d , and viewed at 63x magnification. Pollen grains are shown under normal light microscopy in picture A for each sample. All fluorescent pictures were taken after a 30 second fade period. Transgenic plants do not show a decrease in fluorescence over wildtype controls. Wildtype picture C is an unstained sample to control for pollen autofluorescence. Scale bar = 11.013 urn.  89 3.5 S u m m a r y  The MS2 promoter drives expression of the GUS reporter gene from stage 8 through mid stage 12 of floral development (Figures 3.2 and 3.3). This expression can be seen throughout the anther until the onset of stage 11 when expression appears localized in the tapetum. Expression of an inverted repeat of the 3'UTR fragment ofCER6 under the control of the MS2 promoter (as outlined in Chapter 2) results in transgenic plants producing siliques that phenotypically resemble the siliques of cer6 plants and the stem wax load of wildtype Columbia plants (Figure 3.4). This conditional sterility has been shown by RT-PCR to be correlated to the reduction of CER6 transcript accumulation in developing flower buds (Figure 3.8).  90  Chapter 4 : Results and Discussion - Analysis of T and T 2  3  4.1 Segregation fits the 3:1 ratio expected f r o m a single insertion of the transgene  Seventy of the eighty primary transgenic plants (Ti) were allowed to self-fertilize and generate seeds (T ). T seeds were germinated on AT medium containing kanamycin, 2  2  to examine the segregation of resistant: susceptible pheriotypes in the seedlings. The T plants grown are designated by the name of the Ti parent and the number of the T  2  2  individual (ex. RNAi#l-4). Of the original Ti plants, three (RNAi #3, 11, and 70) either did not produce seeds, or produced few seeds that did not germinate and could not be analyzed further. Since the NPTII gene conveying resistance to the antibiotic kanamycin is contained within the T-DNA of the binary vector (pART27), segregation analysis of resistant: susceptible plants provides an indication of the number of transgene inserts contained within the genome of the plant. Of the initial twenty-eight Ti plants showing small, sterile siliques, the segregation ratios generated from twenty-four plants of the T  2  progeny were examined. Despite the observation that high humidity did not restore fertility sufficient to generate large siliques, plants RNAi#13 and RNAi#25 produced a small number of seeds sufficient for further analysis. The results of this segregation analysis are presented in Table 4.1. Of these twenty-four Ti plants, the T segregation of 2  nineteen plants show adherence to the 3:1 single insertion ratio.  91 Table 4.1: Segregation analysis of T seedlings generated by self-fertilization of TI conditional male sterile plants. Calculated x values were compared against a critical X (o.05) 3.841. Accept indicates a statistical fit to the expected 3:1 segregation ratio for a single transgene insertion whereas Reject indicates the observed data do not fit the 3:1 ratio. 2  2  2  =  Plant  RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi RNAi  #2 #4 #7 #12 #13 #16 #17 #19 #21 #23 #25 #26 #30 #36 #40 #41 #42 #46 #49 #52 #53 #55 #64 #66  Number Resistant  34 33 53 101 40 32 35 19 39 30 17 45 41 27 106 32 200 100 111 32 100 33 33 23  Number Susceptible  16 17 21 40 10 18 15 31 11 20 1 5 9 23 43 18 62 38 30 18 32 17 17 27  Expected Resistant  37.5 37.5 55.5 105.75 37.5 37.5 37.5 37.5 37.5 37.5 13.5 37.5 37.5 37.5 111.75 37.5 196.5 103.5 105.75 37.5 99 37.5 37.5 37.5  Expected Susceptible  12.5 12.5 18.5 35.25 12.5 12.5 12.5 12.5 12.5 12.5 4.5 12.5 12.5 12.5 37.25 12.5 65.5 34.5 35.25 12.5 33 12.5 12.5 12.5  X Value 2  1.307 2.160 0.450 0.853 0.667 3.227 0.667 36.507 0.240 6.000 3.630 6.000 1.307 11.760 1.183 3.227 0.249 0.473 1.043 3.227 0.040 2.160 2.160 22.427  Fit 3:1 Ratio  Accept Accept Accept Accept Accept Accept Accept Reject Accept Reject Accept Reject Accept Reject Accept Accept Accept Accept Accept Accept Accept Accept Accept Reject  92 4.2 Phenotypes of T plants - fertility is restored 2  As seen in the primary transformants, all of the transgenic plants analyzed in the second generation had visibly waxy stems. Selected representatives of plants from this second generation are shown in Figure 4.1. Initially twenty-four T progeny plants were 2  sownfromeach T i parent that showed small siliques. The plants shown in Figure  Figure 4 . 1 : Selected plants from the T generation. Plants derived from the self-fertilization of RNAi#30 (A) all show large, fully developed siliques. Plants derived f r o m the selffertilization of RNAi#52 (B) have fertile (RNAi#52-25, # 5 2 - 2 8 , and # 5 2 - 2 0 ) or both fertile and sterile siliques (RNAi#52-26 and #52-27). 2  4.1 are from a later planting and are not of the initial twenty-four T progeny. All plants 2  showed large, fully developed siliques. Occasionally a plant would have one small silique (as shown in Figure 4.1, plant RNAi#52-26) or several skinny, curled siliques (not shown). Initially it was thought that the restoration of fertility was due to a change in  93 growth conditions between the Ti and T generation. When the T] plants were grown, 2  they were placed in a large, walk-in growth cabinet with very few other plants. The T  2  plants were grown in the same growth cabinet, however the cabinet was now full of many plants. The greater number of plants growing in the cabinet noticeably increased the humidity. Control cer6 plants also showed larger than expected siliques and seed set. Thus it was hypothesized that this change in environment between the Ti and T  2  generations resulted in the observed restoration of fertility of these plants. To test this hypothesis, seeds from selected lines were re-planted. In an attempt to control humidity, these plants were grown in a smaller growth chamber. Additionally, seeds were harvested from T plants and these T3 seeds from selected lines were germinated and 2  grown to maturity.  4.3 Phenotypes of T and T 3 plants - fertility is restored 2  4.3.1 Phenotypic analysis of transgenic plants  The T plants shown in Figure 4.1 are from the second planting of T seed. Five 2  2  T seeds were plantedfromeach Ti plant showing male sterility. Their phenotype was 2  identical to thosefromthe first T planting, showing full siliques that indicate a 2  restoration of fertility. Indeed, while a reduction of CER6 transcript was seen in RNAi#30 and the segregation of T seed fit the expected 3:1 ratio for a single transgene 2  insertion, no transcript reduction was seen in T progenyfromthis plant (RNAi#30-25 to 2  30-29, Figure 4.2). Thus the restoration of fertility in these plants is not caused by an increase in ambient humidity, but rather by the restored accumulation of CER6 transcript in the flowers.  94  Stage of floral development  11  11  12  CER6 ~  GAPC  MMp  RNAi #30-25  RNAi #30  wt  Plant  •  12  ••J I P *  M  11  12  RNAi #30-27 11  RNAi #30-13-1  12  11  —  —  12  •gum  I iPNpsM^I'  ^MIM^MI  Plant and stage o f floral d e v e l o p m e n t Figure 4.2: R T - P C R analysis of CER6 transcript levels of wildtype, RNAi#30 (Ti parent), R N A i # 3 0 - 2 5 and #30-27 ( T representatives), and RNAi#30-13-1 ( T representative) at stages 11 a n d 12 of floral d e v e l o p m e n t (A). A reduction of CER6 transcript w a s observed stage 12 of R N A i # 3 0 but transcript levels are similar to those of wildtype in all subsequent generations e x a m i n e d . GAPC is a loading control. Relative intensity of CER6 bands c o m p a r e d to GAPC bands (B). 2  3  95 T3 seed harvested from T plants showing an occasional small or skinny curled 2  silique was germinated and the fertility of the resulting T3 plants was evaluated (Table 4.2). Although the primary Ti transformant parent RNAi#19 did not fit the expected 3:1 segregation ratio (indicating multiple insertions of the transgene), T 3 seed from the T  2  plant RNAi# 19-3 segregated 3:1 (resistantsusceptible). Representative siliques of mature plants are shown in Figure 4.3. All T3 plants resulting from the self-fertilization of parent RNAi#16-3 (Figure 4.3 A) appeared to have wildtype levels of fertility when compared against wildtype Columbia and the cer6 mutant. This fertility was correlated with no change in levels of CER6 transcript accumulation between stages 11 and 12 of floral development of the fertile plants (Figure 4.4). Although transcript accumulation is reducedfromthat observed in the wildtype plants, it is not eliminated, and the observed reduction is not sufficient to generate the conditionally sterile phenotype.  Table 4.2: Segregation analysis of T3 seedlings generated by self-fertilization of T plants with occasionally altered siliques. Calculated x values were compared against a critical x(o.05) 3.841. Accept indicates a statistical fit to the expected 3:1 segregation ratio for a single transgene insertion whereas Reject indicates the observed data do not fit the 3:1 2  2  2  Plant RNAi # 16-3 RNAi #19-3 RNAi #30-13 RNAi #52-24  =  Number Resistant 26 30 14 3  Number Expected Susceptible Resistant 8 25.5 26.25 5 5 14.25 7 7.5  Expected Susceptible 8.5 8.75 4.75 2.5  x  1  Value 0.039 2.142 0.018 10.8  Fit 3:1 Ratio Accept Accept Accept Reject  96  Three of the five T 3 plants resulting from the self-fertilization of RNAi#19-3 showed large, fully developed siliques whereas two of the five exhibited both small, sterile and fully developed siliques (Figure 4.3 B). Interestingly, the small siliques in RNAi#19-3-3 are accompanied by a reduction in CER6  wt Col  B\/r  v  wtCol  Y  wtCol  D  16-3-1  16-3-2  transcript accumulation whereas  16-3-3  16-  L  16-3-5  \«  ,\j  cer6  '«?  •*  f  L  19-3-1  IT  r  •\  19-3-2  1 1  V  N  P 1 V 19-3-4  19-3-3  s  19-3-5  ;  > cei6  -f. JT/ T i  30-13-1 30-13-2 30-13-3 30-13-4 30-13-5 cer6  y  ^ ^ ^ ^ •  wt Col  -24-1  52-24-2  52-24-3 cer6  Figure 4 . 3 : Selected plants f r o m the T generation. T parents that had several small, underdeveloped siliques were allowed to self-fertilize. Plants generated f r o m parent RNAi#16-3 (A) all appeared indistinguishable f r o m wildtype. T h o s e derived f r o m RNAi#19-3 (B) parent were fertile (RNAi#19-3-2, #19-3-4, and # 1 9 3-5) or were mostly sterile with the occasional fertile silique (RNAi#19-3-1 a n d #19-3-3). Plants derived from the self-pollination of RNAi#30-13 (C) were all fertile, despite the reduction of CER6 transcript seen in RNAi#30. Finally, T 3 plants derived f r o m RNAi#52-24 (D) all appeared fertile w h e n compared to wildtype Columbia and cer6 mutant plants. 3  2  97 there is no such observed reduction in RNA extracted from RNAi# 19-3-1 (Figure 4.5). This discrepancy could be due to the small window of development in which the silencing would be detectable. Flower buds at early stage 12 may be in the process of initiating the PTGS pathway and thus still contain detectable levels of transcript. Alternatively, the siliques of flower buds at the end of stage 12 entering stage 13 may have already begun to elongate. Since CER6 is expressed in developing siliques (Hooker et al., 2002), flower buds with maturing siliques may also contain detectable levels of CER6 transcript. Consequently, while the absence of CER6 amplification indicates that there is a reduction of CER6 transcript accumulation in the selected floral buds, detection of CER6 transcript by PCR does not necessarily eliminate the possibility that the transgene is degraded at the desired stage of development. A strategy to circumvent this discrepancy and detect if the observed sterile phenotype is indeed correlated with a reduction of CER6 activity is discussed in Chapter 5. T progeny from the self-fertilization of RNAi#30-13 and RNAi#52-24 all 3  exhibited large, fully developed siliques (Figure 4.3 C and 4.3 D respectively). Despite the Ti parent RNAi#30 showing a marked decrease in CER6 transcript accumulation, it is clear that the observed fertile phenotype in the T and this T3 generation is due to an 2  accumulation of CER6 transcript (Figure 4.2) and a presumed corresponding restoration of CER6 activity in the tapetal cells. The phenotypes observed in these T and T3 generations presented a perplexing 2  problem. Was the loss of the conditionally male sterile phenotype a result of a defect in the transgene, or has transgene expression itself been silenced in these plants? Two strategies were adopted to address these questions. The first involved examination of the  98  Plant  wt  Stage of floral development  11  RNAi  RNAi  #16-3-4  #16-3-5  11  12  12  11  12  CER6 .,3,.  GAPC  >  .  'tO  _  4 .  (5 I  0 > •-  1  —>  L  J  11  1  12  x  w  t  1  1  i - „ i t , 1 1—,—1— 11 12 11 12 RNAi RNAi #16-3-4 #16-3-5  Plant a n d stage o f floral d e v e l o p m e n t Figure 4 . 4 : R T - P C R analysis of CER6 transcript level in wildtype, R N A i # 1 6 - 3 - 4 a n d #16-3-5 ( T representatives) at stages 11 a n d 12 of floral d e v e l o p m e n t (A). No c h a n g e in accumulation of CER6 transcript w a s observed between stage 11 and stage 12 of T plants. G A P C is a loading control. Relative intensity of CER6 bands c o m p a r e d to GAPC bands (B). 3  3  99  Plant  wt  Stage of floral development  11  12  RNAi #19-3-1  RNAi #19-3-3  11  12  11  11  12 RNAi #19-3-1  12  CER6  GAPC  B  in  e  s~  .£  2  >  1  OJ  0  11  12 wt  11  12 RNAi #19-3-3  Plant a n d staqe o f floral development Figure 4.5: R T - P C R analysis of CER6 transcript levels of wildtype, RNAi#19-3-1, a n d RNAi#19-3-3 ( T representatives) at stages 11 a n d 12 of floral development (A). A reduction in CER6 transcript w a s observed between stage 11 a n d stage 12 of RNAi#19-3-3 but not #19-3-1. GAPC is a loading control. Relative intensity of CER6 bands c o m p a r e d to GAPC bands (B). 3  100 cDNA fraction for the presence of aberrant forms of the transgene transcript. Here an absence of PCR products would indicate that the loss of silencing phenotype was not due to a defect in the transgene. The second strategy required the examination of the genomic DNA for methylation. Transcriptional gene silencing (TGS), which would prevent the formation of the transgene silencing transcript and thus allow accumulation of the CER6 transcript, is correlated with an increase in DNA methylation within the promoter region of the transgene (Morel et al, 2000).  4.3.2 RT-PCR analysis reveals no aberrant transcripts in T and T3 fertile plants 2  To investigate the possibility that the observed reversion of the transgenic conditional sterile phenotype was due to a defect in transcription of the transgene, primers were designed to detect the presence of transgene mRNA (section 2.4.4). If a defect in transgene transcription prevented formation of the dsRNA hairpin structure then silencing of the transgene and the endogene (CER6) would not occur. The primer set would detect the pdk intron from the pHannibal vector and should not generate any product if the transgene is being transcribed and processed properly. The intron should be spliced out of the mRNA to facilitate dsRNA formation and initiate PTGS. Analysis of cDNA generatedfromboth flower buds at both stage 11 and stage 12 of wildtype Columbia (fertile), RNAi#75 (Ti, some small siliques), RNAi#79 (Ti, sterile), RNAi#30 (Ti, sterile), RNAi#30-25 (T , fertile), RNAi#30-27 (T , fertile) 2  2  and RNAi #30-13-1 (T3, fertile) did not reveal the presence of the pdk intron in the mRNAfraction(data not shown). This absence of the intron suggests that either the transgene is being transcribed and processed correctly, or that the transgene is not being transcribed. The former case is likely for the TI plants showing phenotypic conditional  101 sterility and a reduction in CER6 transcript. For the T and T3 plants that have lost the 2  conditional male sterility and are accumulating CER6 transcript, neither case can be ruled out. It is possible that the transgene is still being transcribed and processed properly, and the loss of silencing is due to a mechanism downstream of dsRNA formation. Alternatively, it is possible that the transgene is not longer transcribed, thus allowing wildtype levels of CER6 transcript accumulation resulting in fertility. 4.3.3 P C R analysis reveals methylation of transgene pdk intron  In Neurospora, the presence of unpaired transgene DNA sequences in the genome can initiate the meiotic silencing of unpaired DNA (MSUD) (Shiu et al, 2001). An RNA-dependent RNA polymerase homologous to SDE1 from Arabidopsis was implicated in this process, suggesting a connection between the mechanisms of PTGS/RNAi and MSUD (Shiu et al, 2001). Such a strategy may protect plants against both viral infection and the movement of transposable elements (Waterhouse et al, 2001). Indeed, Lippman et al. (2003) observed the involvement of DNA methylation in silencing transposons in Arabidopsis. PTGS in Arabidopsis has been associated with the methylation of the transgene coding region (Morel et al, 2000), and the 24nt products of the PTGS pathway have been implicated in DNA methylation (Hamilton et al, 2002). To examine the methylation of the CER6 silencing transgene, genomic DNA was extractedfromselected T and T3 fertile plants and subjected to parallel digestions with 2  methylation insensitive (Mbol) and methylation sensitive (Sau3AT) restriction enzyme isoschizomers. The digestions were PCR amplified using oligonucleotide primers specific to the pdk intron and the results are shown in Figure 4.6. As expected, no amplification was observed in the wildtype samples. The presence of ampiiconsfromthe  102 samples digested with Sau3 AI and their corresponding absence in the Mbol digestions clearly indicates the presence of transgene DNA methylation in these plants.  wildtype S  M  U  RNAi #16-3-4  RNAi #16-3 S  M  U  S  M  U  RNAi #30-26  RNAi #30-25 S  M  U  S  M  U  Figure 4.6: P C R analysis of methylation of the pdk intron present in the transgene insertion. Genomic DNA cut with Sau3AI (S), Mbol (M), or uncut (U) w a s amplified. Sau3AI will not cut methylated DNA, allowing for the amplification of methylated introns. Wildtype D N A does not contain the transgene, therefore no amplification product is produced. Uncut DNA is a control template.  The PTGS-directed methylation of transgene coding region has been implicated in the maintenance of the silencing phenotype (Correa et al, 2004) and Arabidopsis mutants defective in PTGS also lack the ability to methylate transgene DNA (Chan et al, 2004). This correlation between transgene methylation and PTGS maintenance makes the relationship between PTGS/RNAi and MSUD unclear. Therefore the observed presence  103 of methylation at the Sau2> AI recognition sequence in the pdk intron does not explain the fertile phenotype of the T and T plants. The presence of methylation in this sequence 2  3  could be an artifact of the PTGS silencing seen in the Ti parents of these plants. 4.3.4 PCR analysis reveals methylation of transgene MS2 promoter sequence The fertile phenotype seen in the T and T3 generations could be a result of 2  transcriptional gene silencing (TGS) of the transgene itself. If this were the case, the transgene would not be transcribed and no silencing of CER6 would be initiated. TGS is associated with methylation of the promoter region of the transgene (Morel et al, 2000). To test for the presence of transgene promoter methylation, genomic DNA from selected plants was digested with methylation sensitive and insensitive restriction enzyme isoschizomers as in section 4.3.3, and a region of the transgene promoter amplified by PCR. The results of this analysis (Figures 4.7 and 4.8) clearly indicate an inability of the methylation-sensitive 6aw3AI to cleave the genomic DNA of these plants. Plant RNAi#16-3-4 has large, fertile siliques (Figure 4.3) associated with an increase in CER6 transcript accumulation (Figure 4.4). These two phenotypes can be linked to the presence of DNA methylation within the transgene promoter sequence (Figure 4.7) and transcriptional gene silencing. This methylation is present in both the RNAi#16-3 and the RNAi#19-3 parents (Figure 4.7), corresponding with the onset of the fertile phenotype in these lines.  104  RNAi #16-3  wt S  M U  S  M U  RNAi #16-3-4 S  M U  RNAi #19-3 S  M U  RNAi #19-3-5 S  M U  Figure 4.7: P C R analysis of methylation of the MS2 promoter in the transgene insertion. G e n o m i c DNA digested with Sau3AI (S), Mbo\ (M), or uncut (U) w a s amplified. Presence of a band in t h e lane designated S indicates no cleavage by the methylation-sensitive Sau3A\. Wildtype D N A does not contain the transgene, therefore no amplicon is produced. Uncut D N A is a control template.  Similarly, while the RNAi#30 plant exhibited conditional male sterility associated with an absence of CER6 transcript accumulation in flower buds at stage 12 of development, subsequent generations appeared fertile (Figures 4.1 and 4.3) and accumulated CER6 transcript (Figure 4.2). Expression of the silencing transgene in these plants appears to be prevented by methylation of the transgene promoter sequence, initiating TGS. Indeed, many recent papers have begun to link the machinery involved in PTGS with DNA methylation and the initiation of TGS. Mette et al. (2000) observed the initiation of TGS by dsRNA triggered promoter sequence methylation. The presence of dsRNA in the nucleus can initiate TGS and heterochromatin formation (Carmichael, 2003). Additionally, the PTGS machinery has also been linked to TGS (Denti et al, 2004)  105  RNAi #30-25  wt S  M U  S  M U  RNAi #30-26 S  M U S  RNAi #30-28 M U S  RNAi #30-13 M U  RNAi #30-13-2 S  M U  Figure 4.8: P C R analysis of methylation of the MS2 promoter in the transgene insertion. Genomic DNA digested with Sau3AI (S), Mbo\ (M), or uncut (U) w a s amplified. Presence of a band in t h e lane designated S indicates no cleavage by the methylation-sensitive Sau3A\. Wildtype D N A does not contain t h e transgene, therefore no amplicon is produced. Uncut DNA is a control template.  and the long 24nt class of siRNA is required for DNA methylation (Hamilton et al, 2002). Matzke et al. (2004) found that RNA signals in the nucleus interact with site specific methyl transferases to methylate the transgene promoter and initiate TGS. Furthermore, Arabidopsis mutants defective in PTGS also exhibit a decrease in DNA methylation (Chan et al, 2004), further substantiating the relationship between PTGS and TGS. 4.4 S u m m a r y  The conditionally male sterile phenotype associated with the reduction of CER6 transcript accumulation seen in the TI generation was lost in all subsequent generations. Plants in the T and T3 generations were fully fertile and accumulated CER6 transcript in 2  the flower buds. No aberrant transcript was detected in the mRNA fraction, indicating either the PTGS machinery was not functioning or the silencing transgene was no longer transcribed. The latter hypothesis was confirmed as the loss of the PTGS phenotype is  106  correlated with methylation of the transgene promoter region. Methylation of the transgene promoter results in the transcriptional gene silencing of the transgene. Because the transgene is not transcribed, PTGS cannot be initiated and CER6 transcript levels are unaffected.  107  Chapter 5: Conclusions and Future Directions  5.1 Analysis of silencing using GFP fluorescence While the examination of the CER6 transcript accumulation by RT-PCR in silenced plants provides an indication of the severity of the silencing and of the level of transcript that remains, it does not provide information about the CER6 protein levels in the cell. A fusion of the gene encoding green fluorescent protein (GFP) with the CER6 gene in transgenic plants would provide a tool to examine the effects of PTGS targeted against CER6 at the protein level. GFP fluorescence would allow visualization of CER6 protein levels present in the cell before and after PTGS induction. Furthermore, if transcription of this gene fusion were driven by the CER6 promoter, GFP fluorescence would provide an accurate indication of endogenous CER6 expression patterns and the loss of this fluorescence would indicate where PTGS has been induced. Plants homozygous for this pxoCER6::GFP:CER6+y\]TK  could be used as the male parent in a  cross with the conditionally sterile RNAi lines showing silencing of CER6 in the anther. 5.1.1 Generation of GFP.CER6  + 3'UTR construct  A construct containing a GFP :CER6 transcriptional fusion driven by the CER6 promoter was available in the lab, but did not include the CER6 3'UTR, the region used to induce PTGS. An Sstl site was identified in the coding region of CER6 and in the vector sequence between the CER6 cassette and the NOS terminator. Oligonucleotide primers were designed with Sstl cutting sites on the 5' end to amplify a region corresponding to 300 bp of the 3' end of the CER6 coding region and 397 bp of the CER6 3'UTR (Figure 5.1). The oligonucleotide sequence of the primers was: forward  108 5' AATCGAGCTCCAAAAGAATCTACAACTA3', and reverse 5'AATAGAGCTCACTGCTATGCCATCCT3'. The fragment was amplified from cDNA generated from the flower buds of wildtype Columbia plants using the following conditions: an initial denaturation at 94°C followed by 30 cycles of a 30 second denaturation at 94"C, a 30 second annealing at 55°C, and 1 minute and 30 seconds of extension at 72°C, followed by one final extension for 7 minutes at 72°C. The PCR reaction was run on a 1% agarose gel and the product, corresponding to an approximately 700bp fragment was excised, purified, and ligated into the GFP vector (Figure 5.1).  Sstl • •  CER6  promoter  GFP  CER6  NOST  \  —; , , _  Sst\  B Sst\ QER6  'Ssf I  Ssfl  \  CER6  promoter  GFP  CER6  NOST  3'UTR Ssfl  Figure 5.1: Cloning strategy used for generation of the proCER6:: GFP: CER6+3 'UTR construct. Sstl sites were identified in the existing binary vector (A) and a 700bp fragment corresponding to the 3' terminus of the CER6 coding region and the CER6 3'UTR was amplified and ligated into the vector.  109 Because the fragment was inserted using one restriction enzyme, it was possible for insertion to occur in either the forward or the reverse orientation. The orientation of the insertions was verified using test digestions and DNA sequencing. Two clones were chosen, one containing the insert in the desired forward orientation and one containing the insert in the reverse orientation. Both clones were transformed into wildtype Columbia plants and cer6 mutant plants. 5.1.2 Plant transformation, selection, a n d phenotypic analysis  Plants containing the transgene were selected based on their resistance to hygromycin (3Qmg/L). Wildtype plants containing the forward orientation of the insert were allowed to mature and their seed harvested. One further generation of growth, harvesting, and selection is required to determine which, of these transformant lines is homozygous for the GFP transgene. While homozygosity for the GFP transgene is not an absolute requirement for the further experiments described in section 5.1.3, it would eliminate the need to select plants using two antibiotics cer6 plants containing the forward orientation of the insert were compared against cer6 plants containing the reverse orientation, and examined for stem waxiness and fertility. Those containing the forward orientation of the insert showed varying levels of stem glaucousness. None appeared as waxy as the wildtype control, however all were visibly whitish, indicating that their stem wax load was increased over that of the cer6 mutants. Furthermore, all transformed plants in this category showed a restoration of fertility (Figure 5.2). This result is consistent with the observations of Fiebig et al. (2000) that fertility is more easily restored than stem waxiness. None of the cer6 plants transformed with the insert in the reverse orientation showed any visible increase in stem  110  wax load or any restoration of fertility. They appeared indistinguishable from the cer6 mutant control plants (data not shown).  Figure 5.2: Fertility is restored in plants containing the cassette with the 3'UTR in the forward orientation w h e n grown under normal humidity. Transgenic siliques appear large and similar to wildtype.  The restoration of fertility and the partial restoration of stem glaucousness in the cer6 plants containing the forward orientation of the 3'UTR insert indicate that this insert does indeed contain a functional copy of CER6. It is predicted that plants homozygous for this transgene would show complete restoration to a wildtype phenotype. 5.1.3 Future strategies  Despite the successful generation of plants containing a XiroCER6::GFP:CER6+y\]TR  cassette, and the impending isolation of homozygous  lines, the loss of silencing in the T and T3 generations of the RNAi CER6 silenced lines 2  Ill prevents the crossing of the two lines to examine silencing. Nevertheless, GFP remains a powerful tool to examine the effects of PTGS on protein accumulation. The complications derived from the loss of silencing in later generations could be circumvented by co-transforming both the RNAi and GFP constructs into wildtype Columbia plants. Transformants selected for based on their resistance to both kanamycin and hygromycin would contain both transgenes. These plants would show GFP fluorescence in the epidermal cells of the stem, where the CER6 promoter drives expression, but would not show fluorescence in the anthers because of the silencing of GFP by the RNAi construct. An alternative to co-transformation is the generation of homozygous lines carrying the TproCER6::GFP:CER6+3'XJTR cassette and transformation of these lines with the RNAi vector. This strategy would eliminate the complications of co-transformation and would also allow for verification of GFP fluorescence in the anthers before transformation with the RNAi vector. Indeed, if GFP fluorescence were lost in the anthers after transformation with the CER6 3'UTR RNAi cassette, it would provide direct evidence of the silencing of CER6 in this location.  5.2 Conclusion  The mechanism of PTGS/RNAi is a powerful tool for generating loss-of function mutants and examining gene function. The current work has shown that silencing of the CER6 gene in Arabidopsis can be triggered by a transgene containing the 3'UTR in an inverted repeat orientation. Furthermore, this silencing can be localized to the anthers and is correlated with a reduction of CER6 transcript at in flower buds at stage 12 of development. Plants showing this reduction have a wildtype stem wax load but are  112 conditionally male sterile. The silencing, however, is not stably inherited in the subsequent generations. Despite several reports of the stable inheritance of PTGS silencing in Arabidopsis for 5-8 generations (Chuang and Meyerowitz, 2000; Stoutjesdijk et al, 2002), these reports may be the exception rather than the norm. Recent evidence implicates the PTGS/RNAi machinery in DNA methylation and the induction of transcriptional gene silencing (TGS) of both transgenes and their homologous endogenes (Chan et al, 2004; Lippman et al, 2003; Matzke et al, 2004; Mette et al, 2000). It is therefore possible that the PTGS machinery induced by the RNAi vector may be the same machinery involved in suppression of the silencing phenotype. Indeed, these reports linking PTGS with DNA methylation and the initiation of TGS are supported by the current finding that loss of the silencing phenotype is correlated with methylation in the transgene promoter region. While our understanding of the mechanisms of PTGS has grown exponentially within the last few years, there remains much to be explored. This method is useful in the analysis of gene function but the current results suggest that it may not be sufficiently stable for the generation of conditionally male sterile lines for use in breeding systems.  113  References  Aarts,M.G.M., Dirkse,W.G., Stiekema,W.J., and Pereira,A. (1993) Transposon Tagging of A Male-Sterility Gene in Arabidopsis. Nature 363:715-717. Aarts,M.G.M., Hodge,R., Kalantidis,K., FIorack,D., Wilson,Z.A., MuIligan,B.J., Stiekema,W.J., Scott,R., and Pereira,A. (1997) The Arabidopsis MALE STERILITY 2 protein shares similarity with reductases in elongation/condensation complexes. Plant Journal 12:615-623. Aarts,M.G.M., Keijzer,C.J., Stiekema,W.J., and Pereira,A. (1995) Molecular characterization of the CER1 gene of Arabidopsis involved in epicuticular wax biosynthesis and pollen fertility. Plant Cell 7:2115-2127. Ahlquist,P. (2002) RNA-dependeht RNA polymerases, viruses, and RNA silencing. Science 296:1270-1273. / Altman,A. (1999) Plant biotechnology in the 21st century: the challenges ahead. EJB Electronic Journal of Biotechnology 2:51-55. Ariizumi,T., Amagai,M., Shibata,D., Hatakeyama,K., Watanabe,M., and Toriyama,K. (2002) Comparative study of promoter activity of three antherspecific genes encoding lipid transfer protein, xyloglucan endotransglucosylase/hydrolase and polygalacturonase in transgenic Arabidopsis thaliana. Plant Cell Reports 21:90-96. Ariizumi,T., Hatakeyama,K., Hinata,K., Sato,S., Kato,T., Tabata,S., and Toriyama,K. (2003) A novel male-sterile mutant of Arabidopsis thaliana, faceless pollen-1, produces pollen with a smooth surface and an acetolysissensitive exine. Plant Molecular Biology 53:107-116. Barnes, S. H. and Blackmore, S. Some functional features in pollen development. Blackmore, S. and Ferguson, I. K. 12, 71-80. 1986. London, Academic Press Inc. Pollen and Spores: Form and Function. Ref Type: Conference Proceeding Baud,S., Guyon,V., Kronenberger,J., Wuilleme,S., MiqueLM., Caboche,M., Lepiniec,L., and Rochat,C. (2003) Multifunctional acetyl-CoA carboxylase 1 is essential for very long chain fatty acid elongation and embryo development in Arabidopsis. Plant Journal 33:75-86.  114 Beals,T.P. and Goldberg,R.B. (1997) A novel cell ablation strategy blocks tobacco anther dehiscence. Plant Cell 9:1527-1545. Beaudoin,F., Gable,K., Sayanova,0., Dunn,T., and Napier,J.A. (2002) A Saccharomyces cerevisiae gene required for heterologous fatty acid elongase activity encodes a microsomal beta-keto-reductase. Journal of Biological Chemistry 277:11481-11488. Bedinger,P. (1992) The Remarkable Biology of Pollen. Plant Cell 4:879-887. Bernstein,E., Caudy,A.A., Hammond,S.M., and Hannon,G.J. (2001) Role for a bidentate ribonuclease in the initiation step of RNA interference. Nature 409:363366. Bessoule,J.J., Lessire,R., and Cassagne,C. (1989) Partial-Purification of the Acyl-CoA Elongase of Allium porrum Leaves. Archives of Biochemistry and Biophysics 268:475-484. Bohmert,K., Camus,L, Bellini,C, Bouchez,D., Caboche,M., and Benning,C. (1998) AGOl defines a novel locus of Arabidopsis controlling leaf development. EMBO Journal 17:170-180. Boutet,S., Vazquez,F., Liu,J., Beclin,C, Fagard,M., Gratias,A., Morel,J.B., Crete,P., Chen,X.M., and Vaucheret,H- (2003) Arabidopsis HEN1: A genetic link between endogenous miRNA controlling development and siRNA controlling transgene silencing and virus resistance. Current Biology 13:843-848. Boutla,A., KaIantidis,K., Tavernarakis,N., Tsagris,M., and Tabler,M. (2002) Induction of RNA interference in Caenorhabditis elegans by RNAs derived from plants exhibiting post-transcriptional gene silencing. Nucleic Acids Research 30:1688-1694. BrummelLD.A., Balint-KurtLP.J., Harpster,M.H., Palys,J.M., Oeller,P.W., and Gutterson,N. (2003)Tnverted repeat of a heterologous 3 '-untranslated region for high-efficiency, high-throughput gene silencing. Plant Journal 33:793-800. Budar,F. and Pelletier,G. (2001) Male sterility in plants: occurrence, determinism, significance and use. Comptes Rendus de 1 Academie des Sciences Serie IiiSciences de la Vie-Life Sciences 324:543-550. Canales,C, Bhatt,A.M., Scott,R, and Dickinson.il. (2002) EXS, a putative LRR receptor kinase, regulates male germline cell number and tapetal identity and promotes seed development in Arabidopsis. Current Biology 12:1718-1727.  115 CarmichaeLG.G. (2003) Antisense starts making more sense. Nature Biotechnology 21:371-372. Cassagne,C., Lessire,R., Bessoule,J.J., Moreau,P., Creach,A., Schneider,F., and Sturbois,B. (1994) Biosynthesis of Very Long-Chain Fatty-Acids in HigherPlants. Progress in Lipid Research 33:55-69. Cecchetti,V., Pomponi,M., Altamura,M.M., Pezzotti,M., Marsilio,S., D'Angeli,S., Tornielli,G.B., Costantino,P., and Cardarelli,M. (2004) Expression of rolB in tobacco flowers affects the coordinated processes of anther dehiscence and style elongation. Plant Journal 38:512-525. Cerutti,L., Mian,N., and Bateman,A. (2000) Domains in gene silencing and cell differentiation proteins: the novel PAZ domain and redefinition of the Piwi domain. Trends in Biochemical Sciences 25:481-482. Chan,S.W.L., Zilberman,D., Xie,Z.X., Johansen,L.K., Carrington,J.C, and Jacobsen,S.E. (2004) RNA silencing genes control de novo DNA methylation. Science 303:1336. Chaudhury,A.M. (1993) Nuclear Genes-Controlling Male-Fertility. Plant Cell 5:12771283. Chen,X.B., Goodwin,S.M., Boroff,V.L., Liu,X.L., and Jenks,M.A. (2003) Cloning and characterization of the WAX2 gene of Arabidopsis involved in cuticle membrane and wax production. Plant Cell 15:1170-1185. Chen,Y.C.S. and McCormick,S. (1996) Sidecar pollen, an Arabidopsis thaliana male gametophytic mutant with aberrant cell divisions during pollen development. Development 122:3243-3253. Cho,H.J., Kim,S., Kim,M., and Kim,B.D. (2001) Production of transgenic male sterile tobacco plants with the cDNA encoding a ribosome inactivating protein in Dianthus sinensis L. Molecules and Cells 11:326-333. Chuang,C.F. and Meyerowitz,E.M. (2000) Specific and heritable genetic interference by double-stranded RNA in Arabidopsis thaliana. Proceedings of the National Academy of Sciences of the United States of America 97:4985-4990. Clough,S.J. and Bent,A.F. (1998) Floral dip: a simplified method for Agrobacteriummediated transformation of Arabidopsis thaliana. Plant Journal 16:735-743.  116 Clough,R.C, Matthis,A.L., Barnum,S.R., and Jaworski,J.G. (1992) Purification and Characterization of 3-Ketoacyl-Acyl Carrier Protein Synthase-Iii from Spinach A Condensing Enzyme Utilizing Acetyl-Coenzyme-A to Initiate Fatty-Acid Synthesis. Journal of Biological Chemistry 267:20992-20998. Correa,R.L., Gomes,L.L., Margis,R., and Vaslin,M.F.S. (2004) Suppression of posttranscriptional gene silencing by callus induction and virus infection reveals the existence of aberrant RNAs. Plant Science 167:159-164. Costa,M.A., Collins,R.E., AnteroIa,A.M., Cochrane,F.C, Davin,L.B., and Lewis,N.G. (2003) An in silico assessment of gene function and organization of the phenylpropanoid pathway metabolic networks in Arabidopsis thaliana and limitations thereof. Phytochemistry 64:1097-1112. Crane, P. R. Form and function in wind dispersed pollen. Blackmore, S. and Ferguson, I. K. 12, 179-202. 1986. London, Academic Press Inc. Pollen and Spores: Form and Function. Ref Type: Conference Proceeding Crete,P., Leuenberger,S., Iglesias,V.A., Suarez,V., Schob,H., HoltorLH., van Eeden,S., and Meins,F. (2001) Graft transmission of induced and spontaneous post-transcriptional silencing of chitinase genes. Plant Journal 28:493-501. Dahl, A. O. Observation on pollen development in Arabidopsis under gravitationally controlled environments. Blackmore, S. and Ferguson, I. K. 12, 49-59. 1986. London, Academic Press Inc. Pollen and Spores: Form and Function. Ref Type: Conference Proceeding Dawson,J., Sozen,E., Vizir,L, Van Waeyenberge,S., Wilson,Z.A., and Mulligan,B.J. (1999) Characterization and genetic mapping of a mutation (ms35) which prevents anther dehiscence in Arabidopsis thaliana by affecting secondary wall thickening in the endothecium. New Phytologist 144:213-222. de Graaf,B.H.J., Derksen,J.W.M., and Mariani,C. (2001) Pollen and pistil in the progamic phase. Sexual Plant Reproduction 14:41-55. Deblock,M., Debrouwer,D., and Moens,T. (1997) The development of a nuclear male sterility system in wheat. Expression of the barnase gene under the control of tapetum specific promoters. Theoretical and Applied Genetics 95:125-131. Dehio,C. and Schell,J. (1994) Identification of Plant Genetic-Loci Involved in A Posttranscriptional Mechanism for Meiotically Reversible Transgene Silencing. Proceedings of the National Academy of Sciences of the United States of America 91:5538-5542.  117 Dellaert,L., Van Es,J., and Koornneef,M. (1979) Eceriferum mutants in Arabidopsis thaliana (L.) Heynh.: II. Phenotypic and genetic analysis. Arabid Inf Serv 16:1026. Denti,M.A., Boutla,A., Tsagris,M., and Tabler,M. (2004) Short interfering RNAs specific for potato spindle tuber viroid are found in the cytoplasm but not in the nucleus. Plant Journal 37:762-769. Dickinson,H.G., Elleman,C.J., and Doughty,J. (2000) Pollen coatings - chimaeric genetics and new functions. Sexual Plant Reproduction 12:302-309. Dickinson, H. G. and Sheldon, J. M . The generation of patterning at the plasma membrane of the young microspore of Lilium. Blackmore, S. and Ferguson, I. K. 12, 1-17. 1986. London, Academic Press Inc. Pollen and Spores: Form and Function. Ref Type: Conference Proceeding Doughty,J., Mccubbin,A., Hedderson,F., Elleman,C, and Dickinson,H. (1992) The Role of the Pollen Grain Coating in Pollination and Self-incompatibility in Brassica oleracea. In Angiosperm Pollen and Ovules, E.Ottaviano, D.L.Mulcahy, M.Sari Gorla, and G.Bergamini Mulcahy, eds (New York: Springer-Verlag), pp. 55-65. Edwards,K., Johnstone,C, and Thompson,C. (1991) A Simple and Rapid Method for the Preparation of Plant Genomic DNA for PCR Analysis. Nucleic Acids Research 19:1349.  Eglinton,G. and Hamilton,R.J. (1967) Leaf Epicuticular Waxes. Science 156:1322-&. Eigenbrode,S.D. (1996) Plant surface waxes and insect behaviour. In Plant Cuticles, G.Kerstiens, ed (Oxford: BIOS Scientific Publishers Ltd.), pp. 201-221. Elleman,C.J. and Dickinson,H.G. (1996) Identification of pollen components regulating pollination-specific responses in the stigmatic papillae of Brassica oleracea. New Phytologist 133:197-205. Elleman,C.J. and Dickinson,H.G. (1990) The Role of the Exine Coating in PollenStigma Interactions in Brassica oleracea 1. New Phytologist 114:511-518. Elleman,C.J., Franklintong,V., and Dickinson,H.G. (1992) Pollination in Species with Dry Stigmas - the Nature of the Early Stigmatic Response and the Pathway Taken by Pollen Tubes. New Phytologist 121:413-424.  118 English, J.J., Mueller,E., and Baulcombe,D.C. (1996) Suppression of virus accumulation in transgenic plants exhibiting silencing of nuclear genes. Plant Cell 8:179-188. Escobar,M.A., Civerolo,E.L., Polito,V.S., Pinney,K.A., and Dandekar,A.M. (2003) Characterization of oncogene-silenced transgenic plants: implications for Agrobacterium biology and post-transcriptional gene silencing. Molecular Plant Pathology 4:57-65. Evans,D.E., Taylor,P.E., Singh,M.B., and Knox,R.B. (1992) The Interrelationship Between the Accumulation of Lipids, Protein and the Level of Acyl Carrier Protein During the Development of Brassica napus 1 Pollen. Planta 186:343-354. Fagard,M., Boutet,S., Morel,J.B., BeIlini,C, and Vaucheret,H. (2000) AGOl, QDE2, and RDE-1 are related proteins required for post-transcriptional gene silencing in plants, quelling in fungi, and RNA interference in animals. Proceedings of the National Academy of Sciences of the United States of America 97:11650-11654. Federle,W., Rohrseitz,K., and Holldobler,B. (2000) Attachment forces of ants measured with a centrifuge: Better 'wax-runners' have a poorer attachment to a smooth surface. Journal of Experimental Biology 203:505-512. Fehling,E., Lessire,R., Cassagne,C, and Mukherjee,K.D. (1992) Solubilization and Partial-Purification of Constituents of Acyl-Coa-Elongase from Lunaria annua. Biochimica et Biophysica Acta 1126:88-94. Fei,H.M. and Sawhney,V.K. (1999) MS32-regulated timing of callose degradation during microsporogenesis in Arabidopsis is associated with the accumulation of stacked rough ER in tapetal cells. Sexual Plant Reproduction 12:188-193. Ferreira,M.A., Engler,J.D., Miguens,F.C, Vanmontagu,M., Engler,G., and deOHveira,D.E. (1997) Oleosin gene expression in Arabidopsis thaliana tapetum coincides with accumulation of lipids in plastids and cytoplasmic bodies. Plant Physiology and Biochemistry 35:729-739. Feys,B.J.F., Benedetti,C.E., Penfold,C.N., and Turner,J.G. (1994) Arabidopsis Mutants Selected for Resistance to the Phytotoxin Coronatine Are Male-Sterile, Insensitive to Methyl Jasmonate, and Resistant to A Bacterial Pathogen. Plant Cell 6:751-759. Fiebig,A., Mayfield,J.A., Miley,N.L., Chau,S., Fischer,R.L., and Preuss,D. (2000) Alterations in CER6, a gene identical to CUT1, differentially affect long-chain lipid content on the surface of pollen and stems. Plant Cell 12:2001-2008.  119 Fire,A., Xu,S.Q., Montgomery,M.K., Kostas,S.A., Driver,S.E., and Mello,C.C. (1998) Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391:806-811. Fuerstenberg,S.I., Bucciaglia,P.A., and Smith,A.G. (2000) Molecular characterization of an anther-specific gene from tobacco shows sequence similarity to a tapetumspecific genefromtomato. Sexual Plant Reproduction 12:250-252. Furness,C.A. and RudalLP.J. (2004) Pollen aperture evolution - a crucial factor for eudicot success? Trends in Plant Science 9:154-158. Furness,C.A. and RudalLP.J. (2001) The tapetum in basal angiosperms: Early diversity. International Journal of Plant Sciences 162:375-392. Gable,K., Garton,S., Napier,J.A., and Dunn,T.M. (2004) Functional characterization of the Arabidopsis thaliana orthologue of Tscl3p, the enoyl reductase of the yeast microsomal fatty acid elongating system. J.Exp.Bot. 55:543-545. Goetz,M., Godt,D.E., Guivarc'h,A., Kahmann,U., Chriqui,D., and Roitsch,T. (2001) Induction of male sterility in plants by metabolic engineering of the carbohydrate supply. Proceedings of the National Academy of Sciences of the United States of America 98:6522-6527. Goldberg,R.B., Beals,T.P., and Sanders,P.M. (1993) Anther Development - Basic Principles and Practical Applications. Plant Cell 5:1217-1229. Greenspan,P. and Fowler,S.D. (1985) Spectrofluorometric Studies of the Lipid Probe, Nile Red. Journal of Lipid Research 26:781-789. Greenspan,P., Mayer,E.P., and Fowler,S.D. (1985) Nile Red - A Selective Fluorescent Stain for Intracellular Lipid Droplets. Journal of Cell Biology 100:965-973. Hamilton,A., Voinnet,0., ChappelLL., and Baulcombe,D. (2002) Two classes of short interfering RNA in RNA silencing. EMBO Journal 21:4671 -4679. Hamilton,A.J. and Baulcombe,D.C. (1999) A species of small antisense RNA in posttranscriptional gene silencing in plants. Science 286:950-952. Hammond,S.M., Bernstein,E., Beach,D., and Hannon,G.J. (2000) An RNA-directed nuclease mediates post-transcriptional gene silencing in Drosophila cells. Nature 404:293-296.  120 Hammond,S.M., Boettcher,S., Caudy,A.A., Kobayashi,R., and Hannon,G.J. (2001) Argonaute2, a link between genetic and biochemical analyses of RNAi. Science 293:1146-1150. Han,G.S., Gable,K., Kohlwein,S.D., Beaudoin,F., Napier,J.A., and Dunn,T.M. (2002) The Saccharomyces cerevisiae YBR159w gene encodes the 3ketoreductase of the microsomal fatty acid elongase. Journal of Biological Chemistry 277:35440-35449. Hannoufa,A., Mcnevin,J., and Lemieux,B. (1993) Epicuticular Waxes of Eceriferum Mutants oi Arabidopsis thaliana. Phytochemistry 33:851-855. Hannoufa,A., Negruk,V., Eisner,G., and LemieuXjB. (1996) The CER3 gene of Arabidopsis thaliana is expressed in leaves, stems, roots, flowers and apical meristems. Plant Journal 10:459-467. Harwood,J.L. (1996) Recent advances in the biosynthesis of plant fatty acids. Biochimica et Biophysica Acta-Lipids and Lipid Metabolism 1301:7-56. Hernandez-Pinzon,L, Ross,J.H.E., Barnes,K.A., Damant,A.P., and Murphy,D.J. (1999) Composition and role of tapetal lipid bodies in the biogenesis of the pollen coat of Brassica napus. Planta 208:588-598. Heslop-Harrison,J. (1968a) Pollen Wall Development. Science 16L230-&. Heslop-Harrison,J. (1968b) Tapetal Origin of Pollen-Coat Substances in Lilium. New Phytologist 67:779-&. Heslop-Harrison,J. and Heslop-Harrison,Y. (1994) Intracelular movement and pollen physiology: progress and prospects. In Molecular and Cellular Aspects of Plant Reproduction, R.J.Scott and A.D.Stead, eds (Cambridge: Cambridge University Press), pp. 191-201. Heslop-Harrison,Y. and Shivanna,K.R. (1977) Receptive Surface of Angiosperm Stigma. Annals of Botany 4L1233-&. Hesse, M . Nature, form and function of pollen-connecting threads in Angiosperms. Blackmore, S. and Ferguson, I. K. 12, 109-118. 1986. London, Academic Press Inc. Pollen and Spores: Form and Function. Ref Type: Conference Proceeding Hird,D.L., WorraLD., Hodge,R., Smartt,S., Paul,W., and Scott,R. (1994) Characterisation of Arabidopsis thaliana anther-specific gene which shares  121 sequence similarity with beta-1, 3-glucanases. In Molecular and Cellular Aspects of Plant Reproduction, R.J.Scott and A.D.Stead, eds (Cambridge: Cambridge University Press), pp. 137-158. Hong,H.P., Ross,J.H.E., Gerster,J.L., Rigas,S., Datla,R.S.S., Hatzopoulos,P., ScoIes,G., Keller,W., Murphy,D.J., and Robert,L.S. (1997) Promoter sequencesfromtwo different Brassica napus tapetal oleosin-like genes direct tapetal expression of beta-glucuronidase in transgenic Brassica plants. Plant Molecular Biology 34:549-555. Hooker,T.S., Millar,A.A., and Kunst,L. (2002) Significance of the expression of the CER6 condensing enzyme for cuticular wax production in Arabidopsis. Plant Physiology 129:1568-1580. Horner,H.T. and Palmer,R.G. (1995) Mechanisms of Genie Male-Sterility. Crop Science 35:1527-1535. Hulskamp,M., Kopczak,S.D., Horejsi,T.F., KihLB.K., and Pruitt,R.E. (1995) Identification of Genes Required for Pollen-Stigma Recognition in Arabidopsis thaliana. Plant Journal 8:703-714. Ikeda,S., Nasrallah,J.B., Dixit,R., Preiss,S., and Nasrallah,M.E. (1997) An aquaporinlike gene required for the Brassica self-incompatibility response. Science 276:1564-1566. Irish,V.F. and Yamamoto,Y.T. (1995) Conservation of Floral Homeotic Gene-Function Between Arabidopsis and Antirrhinum. Plant Cell 7:1635-1644. Ishiguro,S., Kawai-Oda,A., Ueda,J., Nishida,I., and Okada,K. (2001) The Defective in Anther Dehiscence! Gene Encodes A Novel Phospholipase A l Catalyzing the Initial Step of Jasmonic Acid Biosynthesis, Which Synchronizes Pollen Maturation, Anther Dehiscence, and Flower Opening in Arabidopsis. Plant Cell 13:2191-2209. James,D.W., Lim,E., Keller,J., Plooy,L, Ralston,E., and Dooner,H.K. (1995) Directed Tagging of the Arabidopsis Fatty-Acid Elongation-1 (Fael) Gene with the Maize Transposon Activator. Plant Cell 7:309-319. Jenkins,E.S., PauLW., Craze,M., Whitelaw,C.A., Weigand,A., and Roberts,J.A. (1999) Dehiscence-related expression of an Arabidopsis thaliana gene encoding a polygalacturonase in transgenic plants of Brassica napus. Plant Cell and Environment 22:159-167.  122 Jenks,M.A., Tuttle,HA., Eigenbrode,S.D., and Feldmann,K.A. (1995) Leaf Epicuticular Waxes of the Eceriferum Mutants in Arabidopsis. Plant Physiology 108:369-377. Keijzer,C.J. (1987a) The Processes of Anther Dehiscence and Pollen Dispersal .1. the Opening Mechanism of Longitudinally Dehiscing Anthers. New Phytologist 105:487-&. Keijzer,C.J. (1987b) The Processes of Anther Dehiscence and Pollen Dispersal .2. the Formation and the Transfer Mechanism of Pollenkitt, Cell-Wall Development of the Loculus Tissues and A Function of Orbicules in Pollen Dispersal. New Phytologist 105:499-&. ; Kishitani,S., Yomoda,A., Konno,N., and Tanaka,Y. (1993) Involvement of Phenylalanine Ammonia-Lyase in the Development of Pollen in Broccoli (Brassica oleracea L). Sexual Plant Reproduction 6:244-248. Kolattukudy,P.E., Buckner,J.S., and Liu,T.Y.J. (1973) Biosynthesis of Secondary Alcohols and Ketones from Alkanes. Archives of Biochemistry and Biophysics 156:613-620. KoItunow,A.M., Truettner,J., Cox,K.H., Wallroth,M., and Goldberg,R.B. (1990) Different Temporal and Spatial Gene-Expression Patterns Occur During Anther Development. Plant Cell 2:1201-1224. KoornneeLM., Hanhart,C.J., and Thiel,F. (1989) A Genetic and Phenotypic Description of Eceriferum (Cer) Mutants in Arabidopsis thaliana. Journal of Heredity 80:118-122. Kunst,L., Taylor,D.C, and UnderhilLE.W. (1992) Fatty-Acid Elongation in Developing Seeds of Arabidopsis thaliana. Plant Physiology and Biochemistry 30:425-434. Kuranouchi,T., KawaguchLK., and Tanaka,M. (2000) Male sterility in sugar beet induced by cooling treatment and its application to cross-pollination for breeding. Breeding Science 50:283-289. Kurata,T., Kawabata-Awai,C, SakuradanLE., Shimizu,S., Okada,K., and Wada,T. (2003) The YORE-YORE gene regulates multiple aspects of epidermal cell differentiation in Arabidopsis. Plant Journal 36:55-66.  123 Lacomme,C, Hrubikova,K., and Hein,I. (2003) Enhancement of virus-induced gene silencing through viral-based production of inverted-repeats. Plant Journal 34:543-553. Lalanne,E., Honys,D., Johnson,A., Borner,G.H.H., Lilley,K.S., Dupree,P., Grossniklaus,lL, and Twell,D. (2004) SETHI and SETH2, two components of the glycosylphosphatidylinositol anchor biosynthetic pathway, are required for pollen germination and tube growth in Arabidopsis. Plant Cell 16:229-240. Lardizabal,K.D., Metz,J.G., Sakamoto,T., Hutton,W.C, Pollard,M.R., and Lassner,M.W. (2000) Purification of a jojoba embryo wax synthase, cloning of its cDNA, and production of high levels of wax in seeds of transgenic Arabidopsis. Plant Physiology 122:645-655. Lawrence,R.J. and Pikaard,C.S. (2003) Transgene-induced RNA interference: a strategy for overcoming gene redundancy in polyploids to generate loss-offunction mutations. Plant Journal 36:114-121. Lee,Y.H., Chung,K.H., Kim,H.U., Jin,Y.M., Kim,H.L, and Park,B.S. (2003) Induction of male sterile cabbage using a tapetum-specific promoter from Brassica campestris L. ssp pekinensis. Plant Cell Reports 22:268-273. LingeLA., Simon,B-, Izaurralde,E., and Sattler,M. (2004) Nucleic acid 3 '-end recognition by the Argonaute2 PAZ domain. Nature Structural & Molecular Biology 11:576-577. Lippman,Z., May,B., Yordan,C, Singer,T., and Martienssen,R. (2003) Distinct mechanisms determine transposon inheritance and methylation via small interfering RNA and histone modification. Plos Biology 1:420-428. Liu,D.H. and PostbeittenmiIIer,D. (1995) Discovery of An Epidermal Stearoyl-Acyl Carrier Protein Thioesterase - Its Potential Role in Wax Biosynthesis. Journal of Biological Chemistry 270:16962-16969. Lolle,S.J. and Pruitt,R.E. (1999) Epidermal cell interactions: a case for local talk. Trends in Plant Science 4:14-20. Loussaert,D. (2004) Trihalogenated methylsulfonamides as specific male gametocides. Sexual Plant Reproduction 16:299-307. Ma,J.B., Ye,K.Q., and PateLD.J. (2004) Structural basis for overhang-specific small interfering RNA recognition by the PAZ domain. Nature 429:318-322.  124 Mallory,A.C, Mlotshwa,S., Bowman,L.H., and Vance,V.B. (2003) The capacity of transgenic tobacco to send a systemic R N A silencing signal depends on the nature of the inducing transgene locus. Plant Journal 35:82-92. Marais,G.F., Botes,W.C, and Louw,J.H. (2000) Recurrent selection using male sterility and hydroponic tiller culture in pedigree breeding of wheat. Plant Breeding 119:440-442. Mariani,C, Gossele,V., Debeuckeleer,M., Deblock,M., Goldberg,R.B., Degreef,W., and Leemans,J. (1992) A Chimeric Ribonuclease-Inhibitor Gene Restores Fertility to Male Sterile Plants. Nature 357:384-387. Matsuda,N., Tsuchiya,T., Kishitani,S., Tanaka,Y., and Toriyama,K. (1996) Partial male sterility in transgenic tobacco carrying antisense and sense P A L c D N A under the control of a tapetum-specific promoter. Plant and Cell Physiology 37:215-222. Matzke,M., Aufsatz,W., Kanno,T., Daxinger,L., Papp,L, Mette,A.F., and Matzke,A.J.M. (2004) Genetic analysis of RNA-mediated transcriptional gene silencing. Biochimica et Biophysica Acta-Gene Structure and Expression 1677:129-141. Mayfield,J.A., Fiebig,A., Johnstone,S.E., and Preuss,D. (2001) Gene families from the Arabidopsis thaliana pollen coat proteome. Science 292:2482-2485. Mayfield,J. A. and Preuss,D. (2000) Rapid initiation of Arabidopsis pollination requires the oleosin-domain protein GRP17. Nature Cell Biology 2:128-130. McConn,M. and Browse,J. (1996) The critical requirement for linolenic acid is pollen development, not photosynthesis, in an Arabidopsis mutant. Plant Cell 8:403-416. McCormick,S. (1993) Male Gametophyte Development. Plant Cell 5:1265-1275. Mcnevin,J.P., Woodward,W., Hannoufa,A., Feldmann,K.A., and Lemieux,B. (1993) Isolation and Characterization ofEceriferum (Cer) Mutants Induced by T-DNA Insertions in Arabidopsis thaliana. Genome 36:610-618. Mette,M.F., Aufsatz,W., van der Winden,J., Matzke,M.A., and Matzke,A.J.M. (2000) Transcriptional silencing and promoter methylation triggered by doublestranded R N A . E M B O Journal 19:5194-5201.  125 Metz,J.G., Pollard,M.R., Anderson,L., Hayes,T.R., and Lassner,M.W. (2000) Purification of a jojoba embryo fatty acyl-coenzyme A reductase and expression of its cDNA in high erucic acid rapeseed. Plant Physiology 122:635-644. Metzlaff,M., OdelLM., Cluster,P.D., and Flavell,R.B. (1997) RNA-mediated RNA degradation and chalcone synthase A silencing in Petunia. Cell 88:845-854. Millar,A.A., Clemens,S., Zachgo,S., Giblin,E.M., Taylor,D.C, and Kunst,L. (1999) CUT], an Arabidopsis gene required for cuticular wax biosynthesis and pollen fertility, encodes a very-long-chain fatty acid condensing enzyme. Plant Cell 11:825-838. Millar,A.A. and Kunst,L. (1997) Very-long-chain fatty acid biosynthesis is controlled through the expression and specificity of the condensing enzyme. Plant Journal 12:121-131. Millar,A.A., Wrischer,M., and Kunst,L. (1998) Accumulation of very-long-chain fatty acids in membrane glycerolipids is associated with dramatic alterations in plant morphology. Plant Cell 10:1889-1902. Mlotshwa,S., Voinnet,0., Mette,M.F., Matzke,M., Vaucheret,H., Ding,S.W., Pruss,G., and Vance,V.B. (2002) RNA silencing and the mobile silencing signal. Plant Cell 14:S289-S301. MoreI,J.B., Mourrain,P., Beclin,C, and Vaucheret,H. (2000) DNA methylation and chromatin structure affect transcriptional and post-transcriptional transgene silencing in Arabidopsis. Current Biology 10:1591-1594. Murphy,D.J. (1996) Engineering oil production in rapeseed and other oil crops. Trends in Biotechnology 14:206-213. Murphy,D.J. (2001) The biogenesis and functions of lipid bodies in animals, plants and microorganisms. Progress in Lipid Research 40:325-438. Murphy,D.J. and Ross,J.H.E. (1998) Biosynthesis, targeting and processing of oleosinlike proteins, which are major pollen coat components in Brassica napus. Plant Journal 13:1-16. Napoli,C, Lemieux,C, and Jorgensen,R. (1990) Introduction of A Chimeric Chalcone Synthase Gene Into Petunia Results in Reversible Co-Suppression of Homologous Genes in Trans. Plant Cell 2:279-289.  126 Negruk,V., Yang,P., Subramanian,M., Mcnevin,J.P., and Lemieux,B. (1996) Molecular cloning and characterization of the CER2 gene of Arabidopsis thaliana. Plant Journal 9:137-145. NiebeLF.D.C, Frendo,P., Vanmontagu,M., and Cornelissen,M. (1995) Posttranscriptional Cosuppression of Beta-1,3-Glucanase Genes Does Not Affect Accumulation of Transgene Nuclear Messenger-RNA. Plant Cell 7:347-358. Nonomura,K.L, Miyoshi,K., Eiguchl,M., Suzuki,T., Miyao,A., Hirochika,H., and Kurata,N. (2003) The MSP1 gene is necessary to restrict the number of cells entering into male and female sporogenesis and to initiate anther wall formation inrice.Plant Cell 15:1728-1739. Ohlrogge,J.B. and Jaworski,J.G. (1997) Regulation of fatty acid synthesis. Annual Review of Plant Physiology and Plant Molecular Biology 48:109-136. Owen,H.A. and Makaroff,C.A. (1995) Ultrastructure of Microsporogenesis and Microgametogenesis in Arabidopsis thaliana (L) Heynh Ecotype Wassilewskija (Brassicaceae). Protoplasma 185:7-21. PalauquLJ.C, Elmayan,T., Pollien,J.M., and Vaucheret,H. (1997) Systemic acquired silencing: Transgene-specific post-transcriptional silencing is transmitted by graftingfromsilenced stocks to non-silenced scions. EMBO Journal 16:47384745. Papp,L, Mette,M.F., Aufsatz,W., Daxinger,L., Schauer,S.E., Ray,A., van der Winden,J., Matzke,M., and Matzke,A.J.M. (2003) Evidence for nuclear processing of plant micro RNA and short interfering RNA precursors. Plant Physiology 132:1382-1390. Paul,W., Hodge,R., Smartt,S., Draper,J., and Scott,R. (1992) The Isolation and Characterization of the Tapetum-Specific Arabidopsis thaliana A9 Gene. Plant Molecular Biology 19:611-622. Paxson-Sowders,D.M., Dodrill,C.H., Owen,H.A., and Makaroff,C.A. (2001) DEX1, a novel plant protein, is required for exine pattern formation during pollen development in Arabidopsis. Plant Physiology 127:1739-1749. Piffanelli,P. and Murphy,D.J. (1998) Novel organelles and targeting mechanisms in the anther tapetum. Trends in Plant Science 3:250-253. Piffanelli,P., Ross,J.H.E., and Murphy,D.J. (1998) Biogenesis and function of the lipidic structures of pollen grains. Sexual Plant Reproduction 11:65-80.  127 PiffanellLP., Ross,J.H.E., and Murphy,D.J. (1997) Intra- and extracellular lipid composition and associated gene expression patterns during pollen development in Brassica napus. Plant Journal 11:549-562. Post-BeittenmiIler,D. (1996) Biochemistry and molecular biology of wax production in plants. Annual Review of Plant Physiology and Plant Molecular Biology 47:405430. Preuss,D., Lemieux,B-, Yen,G., and Davis,R.W. (1993) A Conditional Sterile Mutation Eliminates Surface Components from Arabidopsis Pollen and Disrupts Cell Signaling During Fertilization. Genes & Development 7:974-985. Preuss,D., Rhee,S.Y., and Davis,R'\V. (1994) Tetrad Analysis Possible in Arabidopsis with Mutation of the Quartet (Qrt) Genes. Science 264:1458-1460. Pruitt,R.E., Vielle-Calzada,J.P., Ploense,S.E., Grossniklaus,U., and Lolle,S.J. (2000) FIDDLEHEAD, a gene required to suppress epidermal cell interactions in Arabidopsis, encodes a putative lipid biosynthetic enzyme. Proceedings of the National Academy of Sciences of the United States of America 97:1311-1316. Ratcliff,F., Harrison,B.D., and Baulcombe,D.C. (1997) A similarity between viral defense and gene silencing in plants. Science 276:1558-1560. Rawsthorne,S. (2002) Carbon flux and fatty acid synthesis in plants. Progress in Lipid Research 41:182-196. Regan,S.M. and Moffatt,B-A- (1990) Cytochemical Analysis of Pollen Development in Wild-Type Arabidopsis and A Male-Sterile Mutant. Plant Cell 2:877-889. Rhee,S.Y., Osborne,E., Poindexter,P.D., and Somerville,C.R. (2003) Microspore separation in the quartet 3 mutants of Arabidopsis is impaired by a defect in a developmentally regulated polygalacturonase required for pollen mother cell wall degradation. Plant Physiology 133:1170-1180. Rhee,S.Y. and Somerville,C.R. (1998) Tetrad pollen formation in quartet mutants of Arabidopsis thaliana is associated with persistence of pectic polysaccharides of the pollen mother cell wall. Plant Journal 15:79-88. Rittscher,M. and Wiermann,R. (1983) Occurrence of Phenylalanine Ammonia-Lyase (Pal) in Isolated Tapetum Cells of Tulipa Anthers. Protoplasma 118:219-224.  128 Rookes,J.E. and Cahill,D.M. (2003) A PALI gene promoter-green fluorescent protein reporter system to analyse defence responses in live cells of Arabidopsis thaliana. European Journal of Plant Pathology 109:83-94. r  Rowley,J.R., Skvarla,J.J., and El Ghazaly,G. (2003) Transfer of material through the microspore exine - from the loculus into the cytoplasm. Canadian Journal of Botany-Revue Canadienne de Botanique 81:1070-1082. RubinellLP., Hu,Y., and Ma,H. (1998) Identification, sequence analysis and expression studies of novel anther-specific genes of Arabidopsis thaliana. Plant Molecular Biology 37:607-619. Sanders,L.C. and Lord,E.M. (1992) The Extracellular Matrix in Pollen tube Growth. In Angiosperm Pollen and Ovules, E.Ottaviano, D.L.Mulcahy, M.Sari Gorla, and G.Bergamini Mulcahy, eds (New York: Springer-Verlag), pp. 238-244. Sanders,P.M., Bui,A.Q., Weterings,K., McIntire,K.N., Hsu,Y.C, Lee,P.Y., Truong,M.T., Beals,T.P., and GoIdberg,R.B. (1999) Anther developmental defects in Arabidopsis thaliana male-sterile mutants. Sexual Plant Reproduction 11:297-322. Sanders,P.M., Lee,P.Y., Biesgen,C, Boone,J.D., Beals,T.P., Weiler,E.W., and GoIdberg,R.B. (2000) The Arabidopsis DELAYED DEHISCENCE 1 gene encodes an enzyme in the jasmonic acid synthesis pathway. Plant Cell 12:10411061. Schneider-Belhaddad,F. and Kolattukudy,P. (2000) Solubilization, partial purification, and characterization of a fatty aldehyde decarbonylasefroma higher plant, Pisum sativum. Archives of Biochemistry and Biophysics 377:341-349. Schreiber,D.N., Bantin,J., and DresseIhaus,T. (2004) The MADS box transcription factor ZmMADS2 is required for anther and pollen maturation in maize and accumulates in apoptotic bodies during anther dehiscence. Plant Physiology 134:1069-1079. Scott,R.J. (1994) Pollen exine - the sporopollenin enigma and the physics of pattern. In Molecular and Cellular Aspects of Plant Reproduction, R.J.Scott and A.D.Stead, eds (Cambridge: Cambridge University Press), pp. 49-81. Scott,P. and Lyne,R.L. (1994) The Effect of Different Carbohydrate Sources Upon the Initiation of Embryogenesis from Barley Microspores. Plant Cell Tissue and Organ Culture 36:129-133.  129 Shih,M.C, Heinrich,P., and Goodman,H.M. (1991) Cloning and Chromosomal Mapping of Nuclear Genes Encoding Chloroplast and Cytosolic Glyceraldehyde3-Phosphate-Dehydrogenase from Arabidopsis thaliana. Gene 104:133-138. Shimakata,T. and Stumpf,P.K. (1982) Isolation and Function of Spinach Leaf BetaKetoacyl-[Acyl-Carrier-Protein] Synthases. Proceedings of the National Academy of Sciences of the United States of America-Biological Sciences 79:5808-5812. Shiu,P.K.T., Raju,N.B., Zickler,D., and Metzenberg,R.L. (2001) Meiotic silencing by unpaired DNA. Cell 107:905-916. Shivanna,K.R. (2003) Pollen Biology and Biotechnology. Enfield, New Hampshire: Science publishers, Inc. Smith,N.A., Singh,S.P., Wang,M.B., Stoutjesdijk,P.A., Green,A.G., and Waterhouse,P.M. (2000) Gene expression - Total silencing by intron-spliced hairpin RNAs. Nature 407:319-320. Smyth,D.R., Bowman,J.L., and Meyerowitz,E.M. (1990) Early Flower Development in Arabidopsis. Plant Cell 2:755-767. Stoutjesdijk,P.A., Singh,S.P., Liu,Q., Hurlstone,C.J., Waterhouse,P.A., and Green,A.G. (2002) hpRNA-mediated targeting of the Arabidopsis FAD2 gene gives highly efficient and stable silencing. Plant Physiology 129:1723-1731. Tang,G.L., Reinhart,B.J., BarteLD.P., and Zamore,P.D. (2003) A biochemical framework for RNA silencing in plants. Genes & Development 17:49-63. Taylor,P.E., Glover,J.A., Lavithis,ML, Craig,S., Singh,M.B., Knox,R.B., Dennis,E.S., and Chaudhury,A.M. (1998) Genetic control of male fertility in Arabidopsis thaliana: structural analyses of postmeiotic developmental mutants. Planta 205:492-505. Taylor, T. N. and Zavada, M . S. Developmental and functional aspects of fossil pollen. Blackmore, S. and Ferguson, I. K. 12, 165-177. 1986. London, Academic Press Inc. Pollen and Spores: Form and Function. Ref Type: Conference Proceeding Thanikaimoni, G. Pollen apertures: forma and function. Blackmore, S. and Ferguson, I. K. 12, 119-135. 1986. London, Academic Press Inc. Pollen and Spores: Form and Function. Ref Type: Conference Proceeding  130 Todd,J., Post-Beittenmiller,D., and JaworskLJ.G. (1999) KCS1 encodes a fatty acid elongase 3-ketoacyl-CoA synthase affecting wax biosynthesis in Arabidopsis thaliana. Plant Journal 17:119-130. TwelLD. (1994) The diversity and regulation of gene expression in the pathway of male gametophyte development. In Molecular and Cellular Aspects of Plant Reproduction, R.J.Scott and A.D.Stead, eds (Cambridge: Cambridge University Press), pp. 83-135. van der Meer,I.M., Stam,M.E., van Tunen,A.J., Mol,J.N.M., and Stuitje,A.R. (1992) Inhibition of Flavanoid Biosynthesis in Petunia Anthers by Antisense RNA: A Novel Way to Engineer Nuclear Male Sterility. In Angiosperm Pollen and Ovules, E.Ottaviano, D.L.Mulcahy, M.Sari Gorla, and G.Bergamini Mulcahy, eds (New York: Springer-Verlag), pp. 22-27. Vaucheret,H., Beclin,C, Elmayan,T., Feuerbach,F., Godon,C, MoreLJ.B., Mourrain,P., Palauqui,J.C, and Vernhettes,S. (1998) Transgene-induced gene silencing in plants. Plant Journal 16:651-659. Vioque,J. and KoIattukudy,P.E. (1997) Resolution and purification of an aldehydegenerating and an alcohol-generating fatty acyl-CoA reductasefrompea leaves (Pisum sativum L). Archives of Biochemistry and Biophysics 340:64-72. Voinnet,0. and Baulcombe,D.C. (1997) Systemic signalling in gene silencing. Nature 389:553. Voinnet,0., Vain,P., Angell,S., and BauIcombe,D.C. (1998) Systemic spread of sequence-specific transgene RNA degradation in plants is initiated by localized introduction of ectopic promoterless DNA. Cell 95:177-187. von Aderkas,P. and Dawkins,M.D. (1993) Haploid embryogenesis in trees. Iri Micropropagation of Woody Plants, M.R.Ahuja, ed (The Netherlands: Kluwer Academic Press), pp. 57-65. von Wettstein-Knowles,P. (1987) Genes, Elongases and Associated Enzyme Systems in Epicuticular Wax Synthesis. In The metabolism, structure, and function of plant lipids., J.Stumpf, B.Mudd, and W.D.Nes, eds (New York: Plenum Press), pp. 489498. Wang,A.M., Xia,Q., Xie,W.S., DatIa,R., and Selvaraj,G. (2003) The classical Ubisch bodies carry a sporophytically produced structural protein (RAFTIN) that is essential for pollen development. Proceedings of the National Academy of Sciences of the United States of America 100:14487-14492.  131 Wang,M.B. and Waterhouse,P.M. (2000) High-efficiency silencing of a betaglucuronidase gene in rice is correlated with repetitive transgene structure but is independent of DNA methylation. Plant Molecular Biology 43:67-82. Waterhouse,P.M., Graham,H.W., and Wang,M.B. (1998) Virus resistance and gene silencing in plants can be induced by simultaneous expression of sense and antisense RNA. Proceedings of the National Academy of Sciences of the United States of America 95:13959-13964. Waterhouse,P.M. and HelliwelkC.A. (2003) Exploring plant genomes by RNA-induced gene silencing. Nature Reviews Genetics 4:29-38. Waterhouse,P.M., Wang,M.B., and Lough,T. (2001) Gene silencing as an adaptive defence against viruses. Nature 411:834-842. Wesley,S.V., HelliwelkC.A., Smith,N.A., Wang,M.B., Rouse,D.T., Liu,Q., Gooding,P.S., Singh,S.P., Abbott,D., Stoutjesdijk,P.A., Robinson,S.P., Gleave,A.P., Green,A.G., and Waterhouse,P.M. (2001) Construct design for efficient, effective and high-throughput gene silencing in plants. Plant Journal 27:581-590. Western,T.L., Young,D.S., Dean,G.H., Tan,W.L., Samuels,A.L., and Haughn,G.W. (2004) MUCILAGE-MODIFIED4 encodes a putative pectin biosynthetic enzyme developmentally regulated by APETALA2, TRANSPARENT TESTA GLABRA1, and GLABRA2 in the Arabidopsis seed coat. Plant Physiology 134:296-306.  Wheeler,M.J., Franklin-Tong,V.E., and Franklin,F.C.H. (2001) The molecular and genetic basis of pollen-pistil interactions. New Phytologist 151:565-584. Wilson,Z.A., Morroll,S.M., Dawson,J., Swarup,R, and Tighe,P.J. (2001) The Arabidopsis MALE STERILITY 1 (MSI) gene is a transcriptional regulator of male gametogenesis, with homology to the PHD-fmger family of transcription factors. Plant Journal 28:27-39. Wu,H.M. and Cheung,A.Y. (2000) Programmed cell death in plant reproduction. Plant Molecular Biology 44:267-281. Wu,S.S.H., Platt,K.A., Ratnayake,C., Wang,T.W., Ting,J.T.L., and Huang,A.H.C. (1997) Isolation and characterization of neutral-lipid-containing organelles and globuli-filled plastids from Brassica napus tapetum. Proceedings of the National Academy of Sciences of the United States of America 94:12711-12716.  132 Xia,Y.J., Nicolau,B.J., and Schnable,P.S. (1996) Cloning and characterization of CER2, an Arabidopsis gene that affects cuticular wax accumulation. Plant Cell 8:1291-1304. Xia,Y.J., Nikolau,B.J., and Schnable,P.S. (1997) Developmental and hormonal regulation of the Arabidopsis CER2 gene that codes for a nuclear-localized protein required for the normal accumulation of cuticular waxes. Plant Physiology 115:925-937. Xie,D.X., Feys,B.F., James,S., Nieto-Rostro,M., and Turner,J.G. (1998) COI1: An Arabidopsis gene required for jasmonate-regulated defense and fertility. Science 280:1091-1094. Xu,X.J., Dietrich,C.R., Delledonne,M., Xia,Y.J., Wen,T.J., Robertson,D.S., Nikolau,B.J., and SchnabIe,P.S. (1997) Sequence analysis of the cloned glossy8 gene of maize suggests that it may code for a beta-ketoacyl reductase required for the biosynthesis of cuticular waxes. Plant Physiology 115:501-510. Xu,X.J., Dietrich,C.R, Lessire,R, Nikolau,B.J., and Schnable,P.S. (2002) The endoplasmic reticulum-associated maize GL8 protein is a component of the acylcoenzyme A elongase involved in the production of cuticular waxes. Plant Physiology 128:924-934. YamamotOjY., Nishimura,M., Hara-Nishimura,L, and Noguchi,T. (2003) Behavior of vacuoles during microspore and pollen development in Arabidopsis thaliana. Plant and Cell Physiology 44:1192-1201. Yang,W.C. and Sundaresan,V. (2000) Genetics of gametophyte biogenesis in Arabidopsis. Current Opinion in Plant Biology 3:53-57. Yang,S.L., Xiea,L.F., Mao,H.Z., Puah,C.S., Yang,W.C, Jiang,L.X., Sundaresan,V., and Ye,D. (2003) TAPETUM DETERMINANT 1 is required for cell specialization in the Arabidopsis anther. Plant Cell 15:2792-2804. Yephremov,A., Wisman,E., Huijser,P., Huijser,C, Wellesen,K., and SaedIer,H. (1999) Characterization of the FIDDLEHEAD gene of Arabidopsis reveals a link between adhesion response and cell differentiation in the epidermis. Plant Cell 11:2187-2201. Yui,R., Iketani,S., Mikami,T., and Kubo,T. (2003) Antisense inhibition of mitochondrial pyruvate dehydrogenase E l alpha subunit in anther tapetum causes male sterility. Plant Journal 34:57-66.  133 Zamore,P.D. (2002) Ancient pathways programmed by small RNAs. Science 296:12651269. Zhang,C.H., Guinel,F.C, and Moffatt,B.A. (2002) A comparative ultrastructural study of pollen development in Arabidopsis thaliana ecotype Columbia and male-sterile mutant aptl-3. Protoplasma 219:59-71. Zhang,Y.H., Shewry,P.R., Jones,H., Barcelo,P., Lazzeri,P.A., and Halford,N.G. (2001) Expression of antisense SnRKl protein kinase sequence causes abnormal pollen development and male sterility in transgenic barley. Plant Journal 28:431441. Zhao,D.Z., Wang,G.F., SpeaLB., and Ma,H. (2002) The Excess Microsporocytesl Gene Encodes A Putative Leucine-Rich Repeat Receptor Protein Kinase That Controls Somatic and Reproductive Cell Fates in the Arabidopsis Anther. Genes & Development 16:2021-2031. Zheng,Z.F., Xia,Q., Dauk,M., Shen,W.Y., Selvaraj,G., and Zou,J.T. (2003) Arabidopsis AtGPATl, a member of the membrane-bound glycerol-3 -phosphate acyltransferase gene family, is essential for tapetum differentiation and male fertility. Plant Cell 15:1872-1887.  Zinkl,G.M. and Preuss,D. (2000) Dissecting Arabidopsis pollen-stigma interactions reveals novel mechanisms that confer mating specificity. Annals of Botany 85:1521. Zinkl,G.M., ZwiebeLB.L, Grier,D.G., and Preuss,D. (1999) Pollen-stigma adhesion in Arabidopsis: a species-specific interaction mediated by lipophilic molecules in the pollen exine. Development 126:5431-5440.  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items