GENETIC ANALYSIS OF GLANDULAR TRICHOME DEVELOPMENT AND METABOLI SM IN CULTIVATED TOMATO By Brian St . Aubin A DISSERTATION Submitted to Michigan State University i n partial fulfillment of the requirements f or the degree of Plant Biology Doctor of Philosophy 2019 P UBLIC A BSTRACT GENETIC ANALYSIS OF GLANDULAR TRICHOME DEVELOPMENT AND METABOLI SM IN CULTIVATED TOMATO By Brian St . Aubin Plants protect themselves from being eaten by producing physical and chemical barriers. Some plants, such as cultivated tomato, are covered in cellular hairs known as trichomes that produce and store chemicals to ward off attacks. While it is known how the chemicals are produced, it is unclear how the plants restrict the production and storage of these chemicals to these special hair cells. Several mutants have been found that fail to produce the hairs or the chemicals. By analyzing the genes that have been mutated in these plants we can identify key aspects to the development of these barriers to attack. In this work I identified a gene involved in the production of a class of molecules that are important for cellular membrane function. Restoration of this gene just in the hair cells of plants lacking this activity restored chemicals important for protection of the plant from predation . The results of this and similar experiments, with other mutants included here, provide a clearer picture of hair cell deve lopment and on how plants rely on other metabolic pathways for the development of hairs and the production of chemicals necessary to defend against attack. A BSTRACT GENETIC ANALYSIS OF GLANDULAR TRICHOME DEVELOPMENT AND METABOLI SM IN CULTIVATED TOMATO By Brian St . Aubin Morphological s tructures and associated chemical compounds serve important functions during the life cycle of plants , including deter rence of herbivor es. Terpenoids represent a major class of compounds involved in def ending plant s from herbivores . T heir differential accumulat ion in specific tissues and cell types is at present not well understood . In cultivated tomato ( Solanum lycopersicum ) , terpenoids and flavonoids accumulate in type VI glandular trichomes that are found on most above ground plant organs . Due to our ability to isolate them, these glandular trichomes of tomato are an attractive model system that can help us better understand how plants regulate the b iosynthesis of terpenoids. The mutations in the odorless - 2 ( od - 2 ) and the anthocyanin free ( af ) genes produce defective type VI glandular trichomes that fail to accumulate flavonoids and volatile terpenoids . Accordingly, these mutants provide useful geneti c tools to study the relationship between the development and metabolic capacity of glandular trichomes. In this work , I determined the molecular basis of the od - 2 mutation and show ed that the wild - type Od - 2 gene encodes an enzyme (designated LCB1A for long chain base 1A ) involved in the biosynthesis of long - chain base precursors of sphingolipids . A functional LCB1A gene is n e ce ss ary for type VI trichomes to produce high levels of volatile terpenoids. Transge nic expression of the wild - type LCB1A cDNA from the Cauliflower Mosaic Virus 35S promoter complemented both the leaf developmental and sphingolipid content phenotypes of the leaflets of the mutant but failed to rescue the terpenoid deficiency. Conversely, expression of LCB1A from a trichome - specific promoter restored terpenoid production in type VI trichomes of od - 2 but did not complement the foliar sphingolipid profile or other lea f developmental defects. These results provide new insights into the tissue - and cell type - specific function of sphingolipid biosynthesis in tomato and establish an essential role for sphingolipids in trichome - borne specialized metabolism. In a similar fashion, I used the chalcone isomerase 1 (CHI1) - deficient af mutant to study the tissue and cell - type specific role of flavonoids. Specifically, I constructed transgenic lines in which CHI1 was expressed in the af mutant background from either the 35S or a trichom e - specific promoter. Detailed biochemical and physiological characterization of these lines showed that CHI1 activity in type VI glandular trichomes , but not in other cell types of the leaf, is required for abundant production of volatile terpenoids . Expre ssion of CHI1 in type VI trichomes was also required to restore resistance of af leaves to feeding by the two - spotted spider mite Tetranychus urticae . In summary, the results described in this thesis advance knowledge of how various metabolic pathways, inc luding sphingolipid and flavonoid biosynthesis , operate to promote normal development and function of glandular trichomes . iv A CKNOWLEDGEMENTS I would like to thank Edward Cahoon for the extraction and measurements of sphingolipids which were conducted in his lab at the University of Nebraska, Lincoln. Tom Sharkey and Alan McClain were essential for helping me to set up for measurement of and und erstanding of photosynthetic parameters of od orless - 2 . A significant amount of work was done by Jin - Ho Kang which allowed the finding of the causal mutation in od orless - 2 , and in developing the pCHI and p35S lines in anthocyanin free . Thanks are also due t o Koichi Sugimoto for the identification of the tomato mutant a nthocyanin f ree being susceptible to the two - spotted spider mite. Metabolic analysis were conducted primarily at the Michigan State University Mass Spectrometry Research Technology Support Facility and the time and teaching provided by the staff (especially Tony Schilmiller and Arthur Jones) were a major help in all sections of this work. Wo rking with Guo - qing Song to perform the tomato transformations taught me so much and they could not have gone better. Computational work in support of this research was performed at Michigan State University High Performance Computing Center. This publicat ion was made possible by a predoctoral training award to B rian S t. Aubin from G rant N umber T32 - GM110523 from National Institute of General Medical Sciences of the National Institutes of Health , and a M ichigan S tate U niversity College of Natural Sciences di ssertation completion fellowship . This study was supported by a National Science Foundation grant (award 1456864). I would like to thank my dissertation committee that supported and guided this research from the beginning to the end. I would especially lik e to acknowledge Gregg Howe for all the time he has spent in helping me bring the story of th is research into clear view. v T A BLE OF C ONTENTS LIST OF TABLES ................................ ................................ ................................ ................................ ...... vii LIST OF FIGURES ................................ ................................ ................................ ................................ ... viii KEY TO ABBREVIATIONS ................................ ................................ ................................ ...................... ix CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 1 DEVELOPMENTAL SPECIALIZATION OF EPIDERMAL CELLS ................................ ................... 2 BIOLOGICAL FUNCTION OF TRICHOMES ................................ ................................ ....................... 3 TRICHOME DEVELOPMENT ................................ ................................ ................................ ............... 4 Arabidopsis as a model system for development of single - celled trichomes ................................ ........ 4 Glandular trichomes ................................ ................................ ................................ ............................. 7 TRICHOMES OF CULTIVATED AND WILD TOMATO ................................ ................................ .... 8 Description of trichome types ................................ ................................ ................................ ............... 8 The type VI glandular trichome ................................ ................................ ................................ ............ 9 Mutations affecting tomato glandular trichome development and function ................................ ......... 9 GLANDULAR TRICHOME METABOLISM ................................ ................................ ...................... 11 Flav onoid biosynthesis ................................ ................................ ................................ ........................ 11 Terpenoid biosynthesis ................................ ................................ ................................ ........................ 13 Other specialized metabolic pathways in Solanum glandular trichomes ................................ ........... 15 Regulation of specialized metabolism in glandular trichomes ................................ ........................... 16 AIMS OF THE THESIS RESEARCH ................................ ................................ ................................ ... 17 CHAPTER 2 THE TOMATO ODORLESS - 2 GENE LINKS TRICHOME - BORNE SPECIALIZED METABOLISM AND LEAF DEVELOPMENT TO SPHINGOLIPID BIOSYNTHESIS ....................... 19 ABSTRACT ................................ ................................ ................................ ................................ ............ 20 INTRODUCTION ................................ ................................ ................................ ................................ .. 20 MATERIALS AND METHODS ................................ ................................ ................................ ............ 23 Plant materials and growth conditions ................................ ................................ ............................... 23 Photosynthetic parameters ................................ ................................ ................................ .................. 23 Isotopic discrimination ................................ ................................ ................................ ....................... 24 RNA sequencing ................................ ................................ ................................ ................................ .. 24 Whole genome sequencing ................................ ................................ ................................ .................. 25 Sanger sequencing ................................ ................................ ................................ .............................. 25 Plasmid construction ................................ ................................ ................................ .......................... 25 Plant and Ag robacterium transformation ................................ ................................ ........................... 26 Sphingolipid analysis ................................ ................................ ................................ .......................... 26 Leaf surface chemistry extraction ................................ ................................ ................................ ....... 26 Mass spectrometry ................................ ................................ ................................ .............................. 27 Leaf and trichome morphology analysis ................................ ................................ ............................. 29 RESULTS ................................ ................................ ................................ ................................ ............... 29 The od - 2 mutation impairs leaf development ................................ ................................ ...................... 29 Impact of od - 2 on photosynthetic performance ................................ ................................ .................. 31 The od - 2 leaf transcriptome is indicative of elevated stress and low terpenoid production ............... 31 Odorless - 2 encodes a tomato homolog of LCB1 ................................ ................................ ................. 34 vi Cell type - specific complementation of od - 2 leaf traits ................................ ................................ ....... 40 DISCUSSION ................................ ................................ ................................ ................................ ......... 43 APPENDIX ................................ ................................ ................................ ................................ ............. 51 CHAPTER 3 CELL TYPE SPEC IFIC COMPLEMENTATION OF CHALCONE ISOMERASE IN S. LYCOPERSICUM ................................ ................................ ................................ ................................ ....... 55 ABSTRACT ................................ ................................ ................................ ................................ ............ 56 INTRODUCTION ................................ ................................ ................................ ................................ .. 56 MATERIALS AND METHODS ................................ ................................ ................................ ............ 60 Plant material and growth conditions ................................ ................................ ................................ 60 Plasmid construction ................................ ................................ ................................ .......................... 60 Plant and Agrobacterium transformation ................................ ................................ ........................... 61 Anthocyanin quantification ................................ ................................ ................................ ................. 61 Metabo lite analysis ................................ ................................ ................................ ............................. 62 Mass spectrometry ................................ ................................ ................................ .............................. 62 Isolation of glandular trichomes ................................ ................................ ................................ ......... 63 Western blot analysis ................................ ................................ ................................ .......................... 63 Spider mite feeding assays ................................ ................................ ................................ .................. 63 RESULTS ................................ ................................ ................................ ................................ ............... 64 CHI1 protein expression localization ................................ ................................ ................................ . 64 Tissue - specific expression of CHI1 modulates leaf anthocyanin levels ................................ ............. 67 Effects of tissue - specific CHI1 expression on type VI trichome size and density ............................... 67 Trichome metabolite abundance in CHI1 - complemented lines ................................ .......................... 69 Restricted expression of CHI1 to type VI trichomes confers resistance to the two - spotted spider mite ................................ ................................ ................................ ................................ ............................ 69 DISCUSSION ................................ ................................ ................................ ................................ ......... 71 APPENDIX ................................ ................................ ................................ ................................ ............. 76 REFERENCES ................................ ................................ ................................ ................................ ........... 79 vii LIST OF TABLES Table A - 1: PCR primers used in this study. ................................ ................................ ................................ 52 Table B - 1: PCR primers used in this study. ................................ ................................ ................................ 78 viii L IST OF FIGURES Figure 1 - 1: The S. lycopersicum type VI trichome ................................ ................................ ..................... 10 Figure 1 - 2. Flavonoid biosynthetic pathway ................................ ................................ ............................... 12 Figure 1 - 3: Terpenoid biosynthetic pathway ................................ ................................ .............................. 14 Figure 2 - 1: Morphological phenotypes of od - 2 leaves ................................ ................................ ............... 28 Figure 2 - 2: Photosynthetic parameters in WT and od - 2 leaves ................................ ................................ .. 30 Figure 2 - 3: Potential interactions between genes that are differentially expressed i n od - 2 ........................ 32 Figure 2 - 4: Molecular identification of od - 2 as LC B1A ................................ ................................ ............. 33 Figure 2 - 5: Changes in sphingolipid composition of o d - 2 leaves are complemented by a 35S::LCB1A transgene ................................ ................................ ................................ ................................ ..................... 36 Fi gure 2 - 6: Leaf morphology of LCB1A complementation lines ................................ ................................ 37 Figure 2 - 7: Trichome morphology of LCB1A complementation lines ................................ ....................... 39 Figure 2 - 8: Leaf terpenoid and flavonoid content in LCB1A complementation line s ................................ . 41 Figure 2 - 9: Phenotypic rescue in F 1 progeny of crosses between p35S and pMKS . ................................ ... 42 Figure A - 1: Additional photosynt hetic parameters in wildtype and od - 2 leaves. ................................ ....... 53 Figure 3 - 1: CHI1 protein levels in leaves and trichomes of tomato genotypes used in this study ............. 65 Figure 3 - 2: Tissue - specific expression of CHI1 modulates anthocyanin accumulation and trichome morphology ................................ ................................ ................................ ................................ ................. 66 Figure 3 - 3: Flavonoid and terpenoid levels in type VI trichomes of CHI1 - complemented lines ............... 68 Figure 3 - 4: Trichome - specific expression of CHI1 confers resistance to the two - spotted spider mite ( Tetranychus urticae ) ................................ ................................ ................................ ................................ .. 70 Figure B - 1: Additional data of terpenoid and flavonoid levels in trich omes of CHI1 - complemented tomato lines. ................................ ................................ ................................ ................................ ................ 77 ix KEY TO ABBREVIATIONS Af A nthocyanin free CHI1 Chalcone i somerase 1 GIPC Glycosyl i nositol p hospho - c era mides GTs G landular trichomes LCB1A Long c hain b ase 1A MEP Methyl erythritol phosphate MKS Methyl ketone synthase MVA Mevalonate Od - 2 O dorless - 2 SNP S ingle nucleotide polymorphism WT Wild - type 1 C HAPTER 1 I NTRODUCTION 2 DEVELOPMENTAL SPECIALIZATION OF EPIDERM AL CELLS The epidermis mediates the plant interactions with the environment. Plant protodermal cells differentiate into a variety of shapes and sizes , each with different impacts on envi ronmental interactions. Mature pavement cells , for example, have a flat exterior surface, reticulated - interlocking edges , and are often transparent . S tomata contain a pair of guard cells that form an epidermal pore for facilitating gas exchange . Trichomes are single - or multi - celled epidermal pro jections. Each of these epidermal cell type s are characterized by unique developmental, signaling, and metabolic activities. This diversity of ce ll types arises through cell division and differentiation of protodermal cells. The formation of pavement cells occurs through the expansion and anticlinal division of p rotodermal cells. A key characteristic of pavement cells is the cuticle, which is a com plex low - permeability wax - containing layer on the outer surface of the cell (Ingram and Nawrath 2017) . Guard cells develop through a series of unequal anticlinal divisions of pavement cells (Shpak et al. 2005) . Guard cells have the unique ability to allow gas exchange across the epidermis through the stoma , control transpiration, and limit intrusion of pathogens. Th e opening and closing of guard cells is controlled by turgor pressure and actin dynamics (Yi et al. 2019) . Trichomes are formed f r om an outgrowth perpendicular to the surface of the epidermis. This process often involves at least one periclinal division and can result in a trichome consisting of many cells. Th e projection of these cells away from the epidermal surface requires additional mechanical support as a freestanding structure. Such p hysical support typically involve s the cytoskeleton , as well as the cell wall , and t urgor pressure (El - Assal et al. 2004; Folkers et al. 2002) . Another unique attribute of trichomes, and glandular trichomes in particular is the production and storage specialized metabol ites (Werker 2000) . Storage or secretion of metabolites from trichomes 3 can have a profound impact on plant biotic interactions (Glas et al. 2012; Sánchez - Hernández et al. 2006) . BIOLOGICAL FUNCTION OF TRICHOMES The unique functions that trichomes provide are often based on their shape. Long , sharp , or densely growing trichomes provide physical deter rence against herbivory (Agrawal 1999; Agren and Schemske 1993) . Dense , wispy trichomes or large umbrella - shaped trichomes can provide protect ion from excessive light exposure, or may limit water loss (Liakoura et al. 1997 ) . Some trichomes have physical characteristics that help maintain leaf dry ness (Barthlott et al. 2009) . Bulbus trichomes often store specific compounds. Within these basic classes of trichome shape there is yet more diversity. L ong and sharp trichomes , for example, may be straight , curved, twisted, or hooked. They can be as large as a raspberry p r ickle or as small as a single cell. Bulbus trichomes, typically referred to as glandular, can be single - or multi - celled structures, and can be extended away from the surface o n a stalk, held close to the surface, or even sunk in a cavity within the epidermis . Glandular trichomes store or exude a variety of compounds which are often unique to that type of trichome. Specific metaboli c pathways associated with glandular trichomes (GTs) can impart an additional layer of cellular specializ ation . Sticky residues produced by some GTs can have a negative impact on insects that encounter the se residue s (Vosman et al. 2018) . Some residues can be so thick they form a water - impermeable layer on the leaf (Dell and McComb 1977) . Some GTs store noxious compounds and then release them when disturbed. Other GTs sequester chemicals that would otherwise be harmful if left to accumulate in the other cells (Dassanayake 2017) . 4 TRICHOME DEVELOPMENT Arabidopsis as a model system for development of single - celled trichomes Trichome development is best understood in the model system Arabidopsis thaliana . The single - cell ed branching trichomes of Arabidopsis develop through several distinct phases. The first phase involves protodermal cells that under go endoreduplication and expansion in the plane of the epidermis. T he se cells then elongate perpendicular to the local epidermal surface. Subsequently, the newly formed trichome cell branch es to produce a three - to four - pronged structure reminiscent of antlers (Marks 1994) . Among t he surfaces adorned with trichomes in Arabidopsis are adaxial juvenile lea v es, ab - and adaxial mature leaves, inflorescence stems, and the seed coat. The entry and exit of different stages of development , a s well as the placement and spacing of trichomes , is regulated by specific transcription factors, protein modification s , microRNA s , and hormones. The development of protoder mal cell s into a trichome requires the activity of various transcription factors, including: the R2R3MYB paralogs GLABROUS 1 (GL1), MYB23, or MYB82; along with a member of MYC subgroup IIIf GLABROUS 3 (GL3) or ENHANCER OF GL3 (EGL3); and a member of the tr yptophan aspartate repeat (WDR) family TRANSPARENT TESTA GLABROUS 1 (TTG1) (Pattanaik et al. 2014) . The presence of sufficient levels of these three classes of transcription factors in the same epidermal cell allows the initiation of endoreduplication , which is required for normal devel opment of trichomes . Th e MYC/R2R3MYB/WDR (MBW) regulatory complex activates other transcription factors , such as homeodomain containing transcription factor GLABROUS 2 (GL2) and the WRKY transcription factor TTG2 , which affects further development of the trichome and patterning of neighboring cells. Transcription factors of the R3MYB class ( CAPRICE , CPC ; TRIPTYCHON , TRY ; TRICHOMELESS , TCL1, and TCL 2 ; and ENHANCER OF TRY AND CPC , ETC1, ETC 2, and 5 ETC 3) limit trichome development by disrupt ing the MBW complex (Kirik et al. 2 004) . A n assembled MBW complex can recruit the mediator complex to the promoter region and induce gene expression (Çevik et al. 2012) . The expression of R3MYB transcription factors such as TRY are also induced by the MBW complex and diffuse to neighboring cells , thus enforcing trichome spacing. The trihelix transcription factor GT - 2 - LIKE 1 (GTL1) contributes to trichome size and impacts e ndoreduplication. Without early stages of endoreduplication , trichome cells do not expand and development is halted. In later stages of development (e.g., following four rounds of endoreduplication ), the fully formed trichomes are typically 32C ( i.e., havi ng 32 times the number of chromosomes) . In GTL1 mutants the average is closer to 80C and the trichomes are significantly larger (Breuer et al. 2009) . The regulation of trichome development is also impacted by several microRNA s . The distribution of microRNA miR156 impacts trichome development by target ing the SQUAM OSA - PROMOTER BINDING PROTEIN - LIKE (SPL) transcription factors (Wu et al. 2009) . SPLs induce the expression of negative regulators of trichome development, including TRY and TCL1. In this manner , low levels of miR156 result in higher level s of SPLs , which induces TRY and TCL1 , and restrict ion of protodermal cell development into trichomes (Pattanaik et al. 2014) . miR172 targets AP2 - like transcription factors TARGET OF EARLY ACTIVATION TAGGED 1 (TOE1) and TOE2. Loss of transcription factors TOE1 (also named RAP2.7) and TO E2 results in more juvenile looking leaves with fewer trichomes (Zhai et al. 20 15) . These miRNA effects are linked by SPL - 9 , which promotes the expression of miR172 (Wu et al. 2009) . Juvenile leaves have an abundance of miR172 and lo w levels of miR156 , which suppresses trichome development throughout the leaf via reduction in TOE1 and 2, and inhibition of trichomes by the abundance of TRY and TCL1 because of higher levels of SPLs. More mature leaves have an abundance of 6 miR156 , thereby inhibiting SPL l evels and limit ing the expression of miR172 and TRY/TCL1. This causes higher levels of TOE1 and TOE2, and allows the MBW complex to form (Pattanaik et al. 2014; Wu et al. 2009) . Arabidopsis ha s a characteristic non - glandular branching trichome with three to four spikes having uneven lengths . Some genes that affect trichome number also impact the degree of branching ( fewer trichomes and less - branching mutants: GL3, EGL3) ( more trichomes and more - branching mutants: TRY). However, there are additional genes c ontrolling branching. Mutants defective in t he MIXTA - like MYB transcription factor NOECK (NOK) have trichomes with extra branches and do not develop papillae (Jakoby et al. 2008) . This la t ter process of forking and cuticular development implies directed cell wall remodeling. Targeting of vesicles for cell wall remodeling and cellulose synthesis requires construction and stable attachment of tubulin to the plasma membrane. The t ubulin double mutant LEFTY1/LEFTY2 has trichomes with fewer branches including sometimes no branches at all (Abe et al. 2004) . As the names of many trichome regulatory genes imply , mutations affecting trichome development also affect other phenot ypes. One of the best example s of phenotypic pleiotropy is transparent - testa - glabrous 1 and 2. Not only do mutations in these genes confer a defect in trichome development , they also result in reduced a ccumulation of anthocyanin s in the seed testa. Paralog s and orthologs of several of the regulatory genes are also implicated in anthocyanin production. Arabidopsis TRANSPARENT TESTA 8 (TT8) is a close paralog of GL3 and EGL3. Maize R genes , which are major regulatory factors for anthocyanin production, are or tholog s of Arabidopsis GL3, EGL3, and MYC1 . There are also multiple links between cuticle and trichome development . The Shine genes (SHN1, 2 and, 3) are AP2 type transcription factors that help regulate cell wall and cuticle production. Overexpression of S HN1,2, or 3 causes a reduction in trichome density 7 ( Aharoni et al. 2004) . SH N 3 is normally expressed in the trichome - support cells, and SHN2 is expressed in trichomes (Oshima et al. 2013) . The cuticle wax content of trichomes is distinct from other epidermal cells , further strengthening a link between trichome developmental regulation and metabolism of the cuticle (Hegebarth et al. 2016) . Glandular trichomes T richome morphology across the plant kingdom varie s greatly and includes diverse m ulticellular trichomes and GTs. Among the major morphological classes of trichomes are stellate, peltate, capitate, and colleter (Payne 1978) . The ability to accumulate or secrete specialized metabolites also adds to the complexity and functional diversity of GTs . Stellate trichomes have a star - shaped, uniform branching pattern, exemplified by cotton lea f trichomes (Inamdar and Rao 1981) . In p eltate trichomes , cells are arrange d in a single plane at the top of the structure; examples include flattened cell s on trichomes of pineapple leaves or globular structures within the mint family (Maffei et al. 1986; Sakai and Sanford 1980) . Capitate trichomes have a single enlarged cell at the top of a stalk, such as the type VI trichome of tomato. Colleter trichomes contain both a multicellular glandular head, and a stalk that is multicellular in cross section , which is common in medicinal plants such as Cannabis sativa and Artemisia annua (Dayanandan and Kaufman 1976; Duke and Paul 2016) . Our current understanding of multicellular GT development lags significantly behind the extensive knowledge of single - celled trichome development in Arabidopsis . Nevertheless, several recent discoveries of regulatory factors governing multicellular trichome development are worth noting . GT density and cuticle production in leaves of Artemisia annua is regulated by homeo - d omain 1 type transcription factors; namely, AaHD1, AaHD8 , and by the MYB - type transcription factor AaMIXTA1 (Yan et al. 2018) . The biosynthesis of specialized metabolites in A. annua GTs 8 is regulated by Trichome and Artemisinin R egulator 1 (TAR1), a transcription factor homolog ous to AtSHN1 / 2 /3 (Tan et al. 2015) . Production of terpenoid and phenylpropanoid metabolites in GTs of Mentha spicata is under negative regulat ion by the transcription factor MsYABBY5 (Wang et al. 2016) . There are also non - transcription factor genes that seem to be important for controlling essential trichome functions such as the ABC transpo rter PDR2 in Petunia hybrida . Reduced expression of phPDR2 led to a decrease in the petuniasterone and petuniolide levels in trichomes , and also resulted in increased susceptibility to herbivory (Sasse et al. 2016) . TRICHOMES OF CULTIVATED AND WILD TOMATO Description of trichome types Plant spec ies within t he Solanaceae clade display a variety of trichome types. The rich genetic resources available for c ultivated Solanum lycopersicum and various wild tomato species provides an opportunity to explore the genetic basis of trichome diversity . The various trichome types in Solanaceae have been described and classified (Luckwill 1943; Snyder and Carter 1985) . According to the nomenclature of Luckwill (1943), t ype II, III, and V are multicel lular , non - glandular structures of varying lengths . T ype I and IV are glandular capitate hairs, whereas type VI and VII are glandular trichomes consisting of a multicellular head (Luckwill 1943) . Type VI is also considered a peltate structure . T he various t richome types of cultivated and wild tomato are distributed on nearly all above - ground surfaces of the plant , with species - specific and environmentally influenced di fferences in trichome distribution often observed . Generally, type I trichomes occur infrequently , but being the largest ( ~3mm ) of the tomato trichome types, is readily apparent on stems (Chang et al. 2018) . By contrast, t ype VII trichomes are small structures with bent stalk cells associated with vasculature regions on leaves. Type IV trichomes develop on juv enile leaves and are replaced by type V as the plant age s (Vendemiatti et al. 2017) . Type V and 9 VI trichomes are high ly abundan t on adult leaves , whereas the stems and petioles are dominated by type VI trichomes (Snyder and Carter 1985) . The type VI glandular trichome Type VI trichomes are structurally and biochemically unique. These peltate trichomes consists of four glandular head cells, a neck or intermediate cell, a stalk cell, a nd a base cell ( Figure 1 - 1 ). Mature type VI trichomes often have a small, clear fluid - filled cavity surrounded by the four yellow - green pigmented head cells and have a head width of 50 - 80 m (Xu et al. 2018) . The neck cell is small and connects the four head cells to the stem cell. The overall height of type VI trichomes is about 250 m (Bergau et al. 2015) . These trichomes contain significant amounts of flavo noids and terpenoids, the most abundant of which is rutin (70 ng per gland, or ~25% by volume) and - phellandrene ( 0.5 ng, or ~0. 4 % by volume) , respectively (Kang et al. 2010 a ) . Mutations affecting tomato glandula r trichome development and function Several mutations affecting trichome development in tomato have been identified and characterized . The dominant w oolly ( w o ) mutation affects the activity of an HD - ZIP transcription factor that is homologous to Arabidopsis GL2 (Yang et al. 2011) . RNAi - based silencing of Wo resulted in plants lac king type I trichomes. Conversely, o ver - expression of Wo resulted in stunted plants with curly leaves and profuse type I trichomes. Analysis of these mutants indicated that although other trichome type s were not affected by RNAi - Wo , other trichome type s we re converted to type I structures in w o heterozygous lines (Yang et al. 2011) . A nother mutat ion, nam ed hairs absent ( h ) , affects the function of a C2H2 zinc - finger transcription factor. Overexpression of H led to pro liferation of type I trichomes, wh ereas RNAi - based silencing of H cause d an absence of trichomes (Chang et al. 2018) . Homologs of H in Arabidopsis called GLABROUS INFLORESCENCE STEMS (G IS) promote trichome initiation but can also repress trichome branching (Wang et al. 2019) . The recessive hairless ( hl ) mutation of tomato impairs the function of a subunit (called SRA1) of the WAVE complex, which is involved in the control of actin filament nucleation (Kang et al. 2016) . All trichome types on hl mutant plants are misshapen, smaller, and often 10 Figure 1 - 1: The S. lycopersicum type VI trichome . (A) Image of type VI trichome. Scale bar is 50 m (B) Schematic of type VI trichome. 1 1 2 3 4 4 5 5 5 B A 11 crooked compared to trichomes on wild - type plants (Kang et al. 2010 b ) . Another recessive muta tion called anthocyanin free ( af ) does not obviously impair trichome morphology but rather reduces the size of the glandular head cells on type VI trichomes and causes a lack the characteristic terpenoid and flavonoid metabolites (Kang et al. 2014) . Kang et al. (2014) showed that the Af gene encodes Chalcone Isomerase 1 (CHI1) which catalyzes second step in the production of flavonoids in tomato. The odorless - 2 ( od - 2 ) mutant is also associated with defective trichome development and metabolic output in tomato but the corresponding gene has yet to be identified. The reported phenotypes of od - 2 plants include reduced trichome density, low abundance of flavonoids and ter penoids, defective trichome morphology, and small leaflets (Kang et al. 2010 a ) . GLANDULAR TRICHOME METABOLISM Flavonoid biosynthesis Flavonoids comprise a major class of polyphenolic compounds produced by all pla nts ( Figure 1 - 2 ). The chemical diversity of flavonoids in the plant kingdom reflects species - specific modifi cations to the core flavonoid backbone ( Figure 1 - 2 ), as well as differences in cell - type specificity , and cellular compartment ation (Agati et al. 2012) . There are , however , a core set of flavonoid biosynthetic enzymes present in most if not all higher plants. This core flavonoid biosynthetic pathway is initiated by Chalcone synthase ( CHS) , which uses para - coumaroyl - CoA , produced from phenylalanine , to form the B ring of naringenin and three subsequent malonyl - CoAs in a polyketide - like reaction that generates the two - ring structure of naringenin chalcone (Rahman et al. 2012) . This reaction is slow and is often a rate limiting step in th is process. Chalcone isomerase ( CHI ) catalyzes ring closure ( ring C) of naringenin chalcone to produce naringenin (Jez and Noe l 2002) . Flavanone 3 - hydroxylase (F3H) , the next enzyme in the pathway, catalyzes the production of dihydrokaempferol (Maloney et al. 2014; Spribille and Forkmann 1982) . In some plants, r ing A is hydroxylated by flavo - hydroxylase and/ or reduced by 12 Figure 1 - 2 . Flavonoid bio synthe tic pathway Chalcone synthase (CHS) Chalcone Isomerase (CHI) Flavone 3 - hydroxylase (F3H) - coumaroyl CoA 3x Malonyl CoA Naringenin - chalcone Naringenin 2 - Oxoglutarate Dihydroflavonols Dihydroflavonol 4 - reductase (DFR) Anthocyanin synthase (ANS) Flavonol synthase (FLS) Anthocyanins Glycosyltransferases Rutin (example) A B C 13 Flavanol Synthase (FLS) (Holton et al. 1993 a, 1993 b ) . Alternatively, d ihydroxyflavonol 4 - reductase (DFR) and anthocyanin synthase (ANS) work in combination to produce anthocyanins from flavonols (Goldsbrough, Belzile, and Yoder 1994) . Many flavonoids are gly cosylated or methylated, and some plants have earlier branch points into flavones or isoflavones (Markham et al. 1978) . Most genes involved with the synthes is of flavonoids have been identified in tomato. Terpenoid biosynthesis Terpenoids are synthesized from the isoprenoid units isopentenyl diphosphate and dimethylallyl diphosphate which , in plants, are synthesized via two di stinct pathways ( Figure 1 - 3 ). Enzymes located in the cytosol and peroxisome produce isopren oids through the mevalonate (MVA) pathway . The second pathway resides in the chloroplast and produce s isoprenoids thr ough the methylerythritol phosphate (MEP) pathway (Schwender et al. 1996) . A wide variety of terpenoid compounds are made in plants , including hormones, pigments, and defensive compounds (Tarkowská and Strnad 2018) . Isopren oid units are also attached to proteins and other metabolites via enzymes that catalyze prenylation reactions (Disch et al. 1998; Rodriguez - Concepcion et al. 1999) . Monoterpenoids made from two isoprenoid units, and sesquiterpenes made from three isoprenoid units, are typically synthesized through a two enzymatic step process. Ini tial polymerization is conducted in a head to tail manner with the retention of only one diphosphate group , such as by neryl diphosphate synthase (NDPS) (Schilmiller et al. 2009) . Th e next typical step is a free radical cyclization by a terpene synthase , such as - phellandrene synthase (PHS) (Schilmiller et al. 2009) . The re are multiple genes encoding these enzymes in Solanaceae, with cultivated tomato ha ving six cis - prenyltransferases other than NDPS and potentially 28 functional terpene synthases other than PHS (Akhtar et al. 2013; Falara et al. 2011) . In many plants, volatile m ono - and sesquiterpenes impart characteristic odors and have been implicated in various 14 Figure 1 - 3: Terpenoid biosynthe tic pathway MEP pathway DXPS 2 - phospho - 4 - - diphospho 2 - methylerythritol 1 - hydroxy - 2 - methl - 2 - butenyl 4 - diphosphate DXR MCT CMK HDR MCS HDS 2 - Methylerythritol 4 - phosphate Pyruvate Glyceraldehyde 3 - phosphate Deoxyxylulose 5 - phosphate 4 cytidine - diphospho 2 methylerythritol 2 - methylerythritol 2,4 - cyclodiphosphate Dimethylallyl - PP Acetyl - CoA Acetoacetyl - CoA 3 - Hydroxy 3 - methylglutaryl CoA Mevalonate Mevalonate - 5P Mevalonate - 5PP HMGS HMGR MK PMK DPMDC MVA pathway Isopentenyl - PP IPPI NDPS 1 PHS 1 Neryl diphosphate - phellandrene 15 defense processes (Langenheim 1994; Sánchez - Hernández et al. 2006) . Volatile terpenoids have been found to negatively impact herbivore feeding preference of flea beetles Epitrix hirtipenni s ; and presence, feeding, and egg deposition of whitefly Bemisia tabaci (Bleeker et al. 2012; Du et al. 2016; Rick et al. 1976) . Increased presence of plant derived terpenoids can also attracted enemies of herbivores the carnivorous mite P hytoseiulus persimilis and the parasitic wasp Cotesia marginiventris (Di cke et al. 1999; Turlings, Tumlinson, and Lewis 1990) . The production of high concentrations of terpenoids is often restricted to specific cell types and the chloroplast pathway (Bleeker et al. 2012; Lan ge et al. 2000) . The biosynthesis of terpenoids via the MEP pathway requires both A TP and NADPH , and is likely supported by a realignment of metabolism and photosynthesis inside the glandular trichome (Balcke et al. 2017; Banerjee and Sharkey 2014) . Other specialized metabolic pathways in Solanum glandular trichomes In addition to flavonoid and terpenoid pathways, some GTs in Solanum species contain pa thways for the biosynthesis of other specialized compounds, including acylsugars and methyl ketones. Acylsugars typically consist of a sugar unit (e.g. sucrose or glucose ) esterified to aliphatic acyl groups. The high degree of chemical diversity among acyl sugars reflec ts differences in the sugar component , as well as the chain length and branching pattern of the attached acyl groups. In the Solanaceae , esterification is catalyzed by a family of BAHD acyltransferases (Schilmiller et al. 2012) . There have been several duplications of the BAHDs and this has allowed the diversification of activities , with the closest sequence similarity between genes of different species coding for proteins that use different acyl sugar substrates or attach acyl groups to different locations (Moghe et al. 2017) . This example of BAHDs illustrate s the potential diversity of compounds that can be derived from a single set of enzymes and metabolic precursors. This chemical diversity is further compounded by differential expression and activities of acyl sugar 16 acylhydrola s es, in the different S olanum species, which selectively remove acyl groups from acyl sugars (Schilmiller et al. 2016) . Another class of compounds, m ethyl - ketones are toxic compounds found in trichomes of a subset of the Solanaceae despite the wider presence of the genes (Yu and Pichersky 2014) . This presence of genes but expressional suppression of the meta bolism they catalyze in some species can further increase diversity and shows the great range of potential metabolism in trichomes. Regulation of specialized metabolism in glandular trichomes Investigations into the regulation of specialized metabolism in tomato type VI trichomes has resulted in the identification of two transcription factors that coordinate the expression of associated biosynthetic genes . The recent characterization of tomato MYC1 d emonstr ates a major role for the transcription factor in the regulation of type VI trichome development (Xu et al. 2018) . Loss of MYC1 resulted in an absence of type VI GTs and reduction in MYC1 expression was associated with smaller and less dense type VI trichomes . Xu et al. (2018) detect ed lower expression of several terpene synthases and lower production of monoterpenes in the leaves and trichom es of plants in which MYC1 was subject to RNAi - based silencing and in the h et erozygous MYC1 loss of function plants. Interestingly , the levels of sesquiterpenes and expression levels of TPS12 were higher in the knockdown plant stem trichomes . This could im ply that MYC1 is regulating the sesquiterpene synthases in an opposite manner, or suppression of mono - terpene synthesis results in higher levels of sesquiterpenes. Tomato MYC2 is a JA - responsive transcription factor that is homologous to Arabidopsis MYC2 . Silencing of MYC2, however , does not appear to affect trichomes (Du et al. 2017) . Interestingly , tomato MYC1 and MYC2 are two paralogs that share significant sequence similarity to Arabidopsis MYC2, MYC3, MYC4 in clade IIIe bHLHs. Th e rol e of clade IIIe bHLH s in regulating GT development and function contrast s the situation in 17 Arabidopsis , where development of non - glandular trichomes is dependent on the clade IIIf bHLHs GL3, EGL3, and MYC1 (Zhao et al. 2012) . The ability of MYCs to induce transcription in response to JA requires JA and COI1 - dependent degradation of JAZ repressors (Howe et al. 2018; Li et al. 2004) . Because application of JA causes a proliferation of type VI GTs and increased accumulation of specialized metabolites inside of them on newly developing tomato leaves (Boughton, Hoover, and Felton 2005; Chen et al. 2018; Li et al. 2004) , it seems likely that the JA - COI1 - JAZ signaling module participates in development of this epidermal structure. It remains unclear, however, whether MYC1 activity is directly controlled by JAZ repressors. MYC1 does not act alone in controlling the devel opment and metabolic activity of type VI GTs . A zinc - finger transcription factor Expression of terpenoids 1 (EOT1) was shown to induce expression of the lin a lool synthase gene, named terpene synthase 5 (TPS5) , in type VI trichomes through direct binding t o the TPS promoter (Spyropoulou et al. 2014) . A HD - ZIP mutant named cutin deficient 2 ( cd - 2 ) has significantly reduced levels of anthocyanins in the epidermis, a lower density of type VI trichomes, and lower abundance of volatile terpenoids (Nadakuduti et al. 2012) . In summary , t richome development in tomato is known to be controlled by multiple transcription factors , including MYC1, C D - 2, W O , H and EOT1 . Several other proteins acting outside the nucleus, including CHI1/ A f ; and SRA/ H l , also modulate the development of these struc tures . Despite this progress , much remains to be learned about the genetic basis of type VI trichome development and function . AIM S OF THE THESIS RESEARCH M any genes and metabolic pathways are likely to be critical for proper development of glandular tric home s . Identifying additional trichome mutants of tomato, and characterizing their role in trichome development , will add to our understanding of this complex structure. At a 18 practical level, this knowledge may facilitate biotechnological efforts aimed at improv ing cro p resilience and the production of valuable plant - based compounds . At the start of this research , the gene defined by the odorless - 2 ( od - 2 ) mutant in tomato had yet to be identified; given the unique development and metabolic phenotypes of this mutant (Kang et al. 2010 a ) , a major aim of this thesis research was to identify the Od - 2 gene and characterize its role in trichome development . To ward this goal, RNA - seq was used to determine how the od - 2 mutation affects gene expression and thus other metabolic pathways . The location of the recessive mutation was previously mapped to a position on chromosome 11 (Kang et al. 2010 a ) . Whole - genome sequencing of the mutant and the progenitor allowed alignment of reads to the newly published genome (The Tomato Genome Consortium 2012) f or SNP analysis and identification the causal mutation. Molecular i dentification of the Od - 2 gene was then validated with the wild - type gene . This work is described in Chapter 2. The anthocyanin free ( af ) mutant of tomato displays a unique pleiotropic phen otype that links flavonoid production to the development and metabolic activity of type VI trichomes. Specifically, loss of function CHI1 in the af mutant results not only in a flavonoid deficiency but also in loss of terpenoid accumulation in type VI tric homes (Kang et al. 2014 ) . Given the unique nature of this phenotype, a more thorough investigation of the function of CHI1 and associated flavonoids was warranted. Accordingly, the second major aim of this thesis was to investigate the tissue - and cell - type specific function of flavonoids through the construction and characterization of transgenic tomato lines in which CHI1 is expressed under the control of various promoters . Prior to this work , it was unclear if flavonoid production in the trichome w as linked to terpenoid biosynthesis, or whether flavon oids in the leaf were responsible for a signa l impacting terpenoid accumulation in the trichome . Chapter 3 describes the investigation of this question . 19 C HAPTER 2 THE TOMATO ODORLESS - 2 GENE LINKS TRICHOME - BORNE SPECIALIZED METABOLISM AND LEAF DE VELOPMENT TO SPHINGOLIPID BIOSYNTHESIS 20 ABSTRACT Glandular trichomes are hair - like epidermal structures that produce volatile terpenoids and other specialized metabolites implicated i n plant anti - herbivore defense. The odorless - 2 ( od - 2 ) mutant of cultivated tomato ( Solanum lycopersicum ) is severely impaired in the accumulation of terpenoids in type VI glandular trichomes and exhibits defects in leaf development. The mechanism by which od - 2 confers these pleiotropic effects remains unknown. Here, we used whole - genome sequencing to screen for nucleotide polymorphisms associated with the od - 2 locus. We identified a single nucleotide deletion in Solyc11g019950, which is predicted to encode LCB1A, a subunit of the serine palmitoyltransferase complex, involved in the biosynthesis of long - chain base precursors of sphingolipids. Consistent with this finding, the sphingolipid profile of od - 2 leaves was indicative of a biosynthetic defect in long - chain sphingoid bases. Transgenic expression of the wild - type LCB1A cDNA driven by the Cauliflower Mosa ic Virus 35S promoter complemented both the leaf developmental and the bulk sphingolipid phenotypes of the mutant but failed to rescue the terpenoid deficiency. Conversely, expression of LCB1A from a trichome - specific promoter restored terpenoid production in type VI trichomes of od - 2 but did not complement the foliar sphingolipid profile of other developmental defects. These results provide new insights into the tissue - and cell type - specific function of sphingolipid biosynthesis in tomato. INTRODUCTION Terpenoids are a diverse class of biologically active compounds originating from the five - carbon subunits isoprene pyrophosphate and dimethylallyl pyrophosphate (Paduch et al. 2007; Schwender et al. 1996) . Because of their volatile nature, many terpenoids (e.g. monoterpenes) contribute to the characteristic scents and flavor of plant s , such as lemon, mint, and pine. Terpenoids have many practical uses in human societies . Terpenoid accumulation in plants is 21 r egulated at multiple levels. Two recurring themes of terpenoid biology in plants is the ir production in specific tissues, cell types, or anatomical structures (Yazaki et al. 2017) , and their biosynthesis in response to herbivory or other types of biotic stress (Turlings et al. 1990) . Given the ecological and biotechnological importance of terpenoids, it would be useful to identify regulatory mechanisms that increase or otherwise modulate terpenoid production. To date ther e has been limited success in achieving this goal (Shen et al. 2016; Spyropoulou et al. 2014) ; therefore, efforts to better understand the fundamental cellular and biochemical process es required for terpenoid biosynthesis and storage are warranted. In cultivated tomato ( Solanum lycopersicum ), the production of terpenoids and other secondary metabolites in glandular trichomes (GTs) is part of a multi - layered defense strategy for protection against insect herbivores . The major terpenoid produced in type VI GTs is - phellandrene. Genetic defects resulting in decrea sed production of - phellandrene are associated with increased susceptibi lity to various insect pests (Bleeker et al. 2012; Kang et al. 2014; Kang et al. 2010 a ; Rick et al. 1976) . Exogenous application of terpenoid enhances defense s against persistent pests , such as whitefly (Du et al. 2016) confirming the role of these compounds in herbivore defense . Genetic analysis in tomato have shown that genes required for normal development of GT s are also needed for synthesis and accumulation of various se condary metabolites suggesting that investigating the molecular basis of these mutants c ould lead to a better understanding of how GTs and secondary metabolites are regulated. Previous advances in understanding tomato trichome for mation have found increased production of type I glandular trichome s by the dominant mutation w oolly ( w o ) (Young and Ma cArthur 1947) . The wo mutation , in a HD - bZIP transcription factor, is lethal when homozygous indicating a function outside of trichome development, and RNAi - 22 based silencing of the WT allele results in plants with no trichomes (Yang et al. 2011) . The hairless ( hl ) tomato mutant , with profuse but highly deformed trichomes, provides an example of the add itional structural support required for normal trichome developme n t compared to many other cell types . The genetic identity of the hl mutation is a deletion of the last exon of SRA1 a member of the WAVE complex responsible for regulating and stabilizing actin dynamics (Kang et al. 2016) . Mutation s affecting the WAVE complex in Arabidopsis have also been shown to result in deformities of trichomes (Brembu et al. 2004) . These actin dynamics are important for vesicle trafficking and structural connections to the plasma membrane (Takenawa and Suetsugu 2007) . The odorless - 2 ( od - 2 ) mutant was originally identified based on an altered leaf aroma phenotype, and subsequently shown to have structurally defective trichomes that fail to produce significant quantities of rutin and - phellandrene . Type VI GT development in the od - 2 mutant is severely impaired and recognizable GTs a re sparse on od - 2 leaves (Kang et al. 2010 a ) . In testing the type VI GT from od - 2 Kang et al, ( 2010 a ) found no detectable terpenoids , which leads to the hypothesis that there is some regulatory restriction on the production of terpenoids in od - 2 even if the cells reach typical type VI GT shape (Kang et al. 2010 a ) . Importantly, growth of od - 2 in field experiments showed that this mutation compromises resistance to natural populations of insect herbivores, such as flea beetle ( Epitrix cucumeris ), and Colorado potato beetle larvae ( Leptinotarsa decemlienata ) (Kang et al. 2010 a ) . Another distinguishing characteristic of od - 2 was the small leaflets but otherwise normal leaf lengths (Kang et al. 2010 a ) . Silver damage from Western flower thrips ( Fran kliniella occidentalis ) wa s found to be much higher on od - 2 along with higher levels of JA - Ile and lower levels of ABA (Escobar - Bravo et al. 2018) . The od - 2 mutant has become a useful too l for testing trichome mediated herbivore and pathogen responses due to its otherwise normal seeming induction of other defense compounds (Escobar - Bravo et al. 2018; 23 Escoba r - Br avo et al. 2019 a , 2019 b ; Kang et al. 2010 a ) . The od - 2 locus was mapped to chromosome 11 (Kang et al. 2010 a ) but the molecular identity of this gene remain ed unknown . In th is study, we employed whole - genome sequencing to identify candidate mutations responsible for the pleiotropic phenotypes of od - 2 . We identified a small deletion in a gene (Solyc11g019950) predicted to encode an enzyme (termed LCB1A) involved in the biosyn the sis of long - chain base precursors of sphingolipids. Sphingolipid profiling of od - 2 leaves validated a biosynthetic defect in long - chain sphingoid bases. Transgenic expression of the wild - type LCB1A cDNA from the Cauliflower Mosaic Virus 35S promoter comple mented the leaf developmental and bulk sphingolipid phenotypes of the mutant but, interestingly, failed to rescue the terpenoid deficiency. Expression of LCB1A from a trichome - specific promoter restored terpenoid production in type VI trichomes of od - 2 but did not complement the foliar sphingolipid profile or other leaf developmental defects. These data provide new insights into the tissue - and cell type - specific function of sphingolipid biosynthesis in tomato . MATERIALS AND METHODS Plant materials and growth conditions We grew our plants in growth chambers with 16 hr, 250 E, 28° C day and 18° C, 8 hr night cycles. Wild - type tomato seeds originated from A.L. Castle Inc. (variety C astlemart II) . The seed source for od - 2 w as previously describ ed (Kang et al. 2010 a ) . The seeds were bleach sterilized for 20 minutes with 20% bleach ( 2.5% sodium hypochlorite ) prior to use, germinated in J7 Jiffy peat pellets , and transferred to 6 cm x 6 cm x 8 cm pots containing soil at three weeks. Photosynthetic parameters F ive - week - old plants were assessed with a LI - Cor 6800 (Lantz et al. 2019) . A leaflet from the fourth leaf of each plant was monitored for CO 2 - 2 s - 1 24 m - 2 s - 1 of light. Inlet air temperature was maintained at 28 o C . VPD was kept between 0.8 and 1. Chlorophyll levels were measured using a separate leaflet from the same set of plants used for gas exchange measurements . An 8 mm diameter hole - punch was used to obtain four leaf disks from each leaf. The leaf disks were weighed and t hen extracted with 1 mL methanol by briefly beating in a TissueLyser II (Quiagen) with two 3 mm stainless beads and then incubating at 4 o C with gentle shaking for fou r hours. Debris was removed by centrifugation for 1 min at 13 ,000 x g and transfer red to new tubes. All extractions were diluted to 1 : 10 with methanol and analyzed for absorption at 652 nm, 665 nm, and 750 nm (Porra et al. 1989) . Isotopic discrimination A single leaflet from the fourth leaf was lyophilized and analyzed for 13 C discrimination and carbon and nitrogen content at the S table I sotope R atio F acility for E nvironmental R esearch , University of Utah (Farquhar et al. 1989; Guo et al. 2018). RNA seq uencing RNA was purified from leaflets of five - week - old Castlemart and od - 2 plants. Each sequenc ed sample consisted of pooled cDNA from three plants with three replicated samples for each genotype. Illumina TruSeq sequencing was performed on a single lane to obtain 125 bp paired - end reads from the six libraries. Library construction and sequencing wa s performed in the MSU Genomics core R esearch T echnology S upport F acility . The unprocessed sequencing data was deposited in the National Center for Biotechnology Information Short Read Archive ( ncbi.nlm.nih.gov/sra/PRJNA580508 ). Sequencing r eads were mapped to ITAG2.4 and SL2.50 available through solgenomics.net. Analysis of differential gene expression was conducted with the Bioconductor.org package EdgeR , using a threshold value of 0.05 from common dispersion false discovery rate (FDR), and greater than one log 2 fold change ( LFC, positive for up regulation , 25 or negative for down regulation) . A ll genes with greater than one LFC had an FDR less than 0.02. Gene Ontology terms were obtained from pantherdb.org (Mi et al. 2017) and searched for sta tistically overrepresented terms using the list of differentially expressed genes. For Gene Ontology listing, t he Panther14.1 program includes Benjamini Hochberg FDR correction for multiple testing. The s tring diagram was generated using a list of genes that are differentially expressed in od - 2 to string - db.org. Whole genome sequencing Illumina sequencing was used to obtain 100 bp paired - end reads which were aligned to the SL2.5 genome and annotation ITAG2.4 developed by the International Tomato Annotation Group (Fernandez - Pozo et al. 2015). Nucleotide polymorphisms between CM and od - 2 were investigated by scanning for non - synonymous SNPs in the Integrated Genome Viewer, broadinstitute.org (Thorvaldsdóttir et a l. 2013). Sanger sequencing Primers 1 R 2 L ( Appendix, Table A - 1 ) were used to PCR amplify a region of LCB1A from genomic DNA from wildtype and od - 2 . The thermocycled PCR reaction was verified to contain a single band on agarose gel and submitted to the Michigan State University Research Technology Support Facility Genomics Core for sanger Sequencing. Plasmid construction A fragment containing the Solanum habrochaites MKS1 promoter was PCR amplified with primers 70L and 71R ( Appendix, Table A - 1 ) from the MKS1 - GFP expression vector ( Akhtar et al. 2013) and digested with HindIII and XbaI. The 35S promoter in pGWB - 402 was digested using HindIII and XbaI . T he 1644 bp PCR prod uct containing the MKS1 promoter was ligated in t he place of the 35S promoter, resulting in pGWB - 40x_MKS1. A cDNA obtained from S. 26 lycopersicum cv. C astle M art was amplified with primers 26L 105L ( Appendix , Table A - 1 ). The resulting 1522 bp fragment was used in an LR reaction with either pGWB - 402 or pGWB - 40x_MKS1 resulting in expression vectors 35S : : LCB1A and MKS : : LCB1A , respectively. The steps taken in the LR reaction were performed with LR ClonaseTM II ( ThermoFische r ) as described by the manufacturer . Plant and Agrobacterium t ransformation The MKS : :LCB1A and 35 S: :LCB1A plasmids were introduced into Agrobacterium tumefaciens strain AGL0 using the heat shock method (Weigel and Glazebrook 2002). Transformation of od - 2 was conducted in the Plant Biotechnology Resource and Outreach Center of Michigan State University (http://www.ptc.msu.edu/index.html). Resulting plants were tran sferred to soil for seed propagation in the greenhouse . Sphingolipid analysis Leaflets from leaves three , four , and five were harvested from five - week - old plants and lyophilized. One plant was used per sample and five samples were obtained per genotype. O ne sample of od - 2 was lost during processing. Measurement of sphingolipids was performed as previously described (Markham et al. 2006) . Leaf surface chemistry extraction E xtracted leaf surface metabolites following a similar protocol to that used by Kang et al (2010 a ). O ne leaflet from leaf seven of each five - week - old plant was weighed and washed with 1 - 1 tetradecane and 1 - 4 - hydroxybenzoate internal standards for 2 minutes with gentle shaking. A 100% methano l extraction was used to prevent non - enzymatic isomerization of naringenin - chalcone to naringenin despite 100% methanol being a 27 non - ideal extraction buffer for flavonoids and terpenoids. Wash effluent as stored in individual sealed 1.5 mL glass vials at - 2 0°C until analysis. Mass spectrometry Volatile terpenoids were quantified by injecting o ne quadrupole (Agilent) by a CombiPAL (CTC Analytics) . The column used was a 40m VF5MS, 30 m with 10 diameter (Agilent) . Operating conditions were as follows: 280°C inlet temperature, 88.35 psi inlet pressure, 1:10 split ratio, 45 - 350 mass scan starting at 3.5 min. Ramp temperatures used initial 40° C for one min, to 90° C at 40° C min - 1 , to 110° C at 15° C min - 1 , to 250° C at 25° C min - 1 , to 320°C at 40° C min - 1 , then hold 320° C for 2 minutes. Tota l run time was 12.93 minutes. Flavonoid abundance was measured by LC - MS. One into a n Acquity Binary Solvent Manager and Column Manager (Waters) by a 2777C Sample manager (Waters) . A Xevo G2 - XS QTof (Waters) mass spectrometer was o perat ed under the following conditions: capillary voltage , 2 kV ; sample cone , 40 V ; source temperature , 100° C; desolvation temperature , 350° C; desolvation flow rate , 600 L hr - 1 ; and electrospray ionization , negative sensing mode. MS e was used with several functions which split each sampling timepoint : 1. No collision energy (to gather whole molecule masses) ; 2. Ramp of 20 - 60 V collision energy (to analyze sample molecule fragmentation patterns) ; and 3. Ion l ock spray (for calibration of exact mass) . A 10 cm x 2.1 mm, 2.7 m Ascentis® Express C - 18 HPLC column ( supelco ) held at 40° C was used for all separations . De - ionized w ater containing 0.1% formic acid was used for solvent A, and methanol was used for solvent B. Using a continuous flow rate of 0.3 mL min - 1 , the gradient profile started at 5% solvent B and ramped to 60% solvent B over 3 minutes using the curve 6 setting on the Waters sol vent manager and then ramped to 100% 28 Figure 2 - 1: Morphological phenotypes of od - 2 leaves . (A) Representative images of 7 th leaf from 6 - week - old plants ( WT , left ; and od - 2 , right ) . Scale bar is 2 cm. (B) Dissecting microscope image of leaf surface ( WT , left ; and od - 2 , right ) . Scale bar is 0.1 mm. (C) Micrographs of leaf cross sections stained with hematoxylin and eosin ( WT , left ; and od - 2 , right ) . Scale bar is 100 m. A WT od - 2 C B 29 solvent B over 4 min using curve 6, held at 100% solvent B until the 5.5 min point, dropped to 5% solvent B at 5.51 minutes, then held at 5% solvent B until 6.5 minutes. Leaf and trichome morphology analysis The 7th leaf of six - week - old plants w as used to measure leaflet length and area. Trichomes were imaged using a Leica M16 dissecting microscope at 20x and 80x magnification. RESULTS The od - 2 mutation impairs leaf development C omparison of wild - type (WT) and od - 2 plants showed that the mutant has smaller compoun d leaves and a smaller petiole width (Figure 2 - 1A) , as previously noted (Kang et al. 2010 a ) . In contrast to WT leaflets, which typically exhibit varying degrees of curling, od - 2 leaflets are relatively small and fl at. The small size of od - 2 leaflets is associated with epidermal surface abnormali ties, most notably a sparseness of both glandular and non - glandular leaf trichomes (Figure 2 - 1B) . To determine whether the small size of od - 2 leaflets is associated with corresponding changes in cell size, we used light microscopy to visualize the cellular organization in leaf cross sections. WT leaflets show ed regularly spaced epidermal cells, rows of palisade cells , and uniform size of mesophyll cells ( Figure 2 - 1C ) . By contrast, the epidermal cell layer of od - 2 leaflets contained very large cells having a high degree of size and shape irregularity ( Figure 2 - 1C ). These defects in the size and general o rganization of epidermal cells in the mutant extended to other cell types in the leaf, including palisade and mesophyll cells. Measurements of total leaf thickness derived from the cross - sectioned images showed that the thickness of od - 2 leaflets is signif icantly greater than that of WT ( Figure 2 - 1C , 2 - 6 ). 30 Figure 2 - 2: Photosynthetic parameters in WT and od - 2 lea ves . Comparison of parameters related to photosynthesis in 5 - week - old WT and od - 2 p lants C WT od - 2 0 .0 0.5 1 .0 1.5 2 .0 2.5 3 .0 Mesophyll conductance (g m , mol m - 2 s - 1 ) A p < 0.005 WT od - 2 0 10 20 30 40 50 Dry leaf mass per area (LMA, g m - 2 ) p < 0.005 WT od - 2 0 50 100 150 200 250 Water content (g m - 2 ) p < 0.005 D WT od - 2 0.0 0.5 1.0 1.5 2.0 Chlorophyll a+b abundance (nmol mm - 2 ) p < 0.05 F B p = 0.4 WT od - 2 0.0 0.1 0.2 0.3 0.4 0.5 Total c onductance (g t , mol m - 2 s - 1 ) E WT od - 2 0 1 2 3 4 % nitrogen in DW p < 0.001 31 Impact of od - 2 on photosynthetic performance The increased thickness and level of cell disorganization (e.g., larger cells) in od - 2 leaves raised the possibility that the mutant is altered in photosynthetic performance (Ren, Weraduwage, and Sharkey 2019) . To address this question, we compared various photosynthetic parameters in WT and od - 2 leaves (Lantz et al. 2019) . Gas exchange performed - 2 s - 1 ) and - 2 s - 1 ) light intensity revealed that od - 2 leaves have lower mesophyll conductance ( Figure 2 - 2A ). The rates of stomatal conductance and total conductance, however, were not significantly affected in od - 2 ( Figure 2 - 2B , Appendix , Figure A - 1 E ). Independent measurements of conductance by analysis of 12 CO 2 / 1 3 CO 2 dis crimination at rib u lose - 1,5 - bis phosphate c arboxylase (rubisco) showed that the overall rates of conductance were similar in od - 2 and WT leaves ( Appendix , Figure A - 1 K ). The increased leaf thick ness in od - 2 prompted us to investigat e other physiological parameters of the leaf. Dry mas s per area (LMA) is widely used as a key indicator of a variety of leaf traits that influence plant performance in, and adaptation to, variable environments (Poorter et al. 2009) . We found that, despite increased thickness of od - 2 leaves, the LMA of the mutant was significantly less than WT ( Figure 2 - 2C ). Consistent with leaf traits that correlate with LMA (Poorter et al. 2009) , od - 2 also had a higher water content and increased total leaf nitrogen in comparison to WT ( Figure 2 - 2D ). We also found that chlorophyll abundance on a per leaf area basis, was higher in od - 2 . The t hicker leaves of od - 2 contain more water as may be expected ( Figure 2 - 2C ), however there was a lower dry leaf mass per area (LMA) in od - 2 ( Figure 2 - 2D ). The od - 2 leaf transcriptome is indicative of elevated stress and low terpenoid production To better understand the impact of od - 2 on genome - wide processes in the leaf we used mRNA sequencing (RNA - seq) to compare the transcript profile of WT and od - 2 leaves. Analysis 32 Figure 2 - 3: Potential interactions between genes that are differentially expressed in od - 2 . (A) Down - regulated genes in od - 2 lea ves . (B) Up - regulated genes in od - 2 lea ves . Image made by entering differentially expressed genes in od - 2 ( threshold: FDR < 0.05, logFC > 1 or < - 1 for (A) and (B) respectively ) to string - db.org. Line thickness indicates confiden ce level of the interaction . See S upplemental table 2 - 1 and 2 - 2 for gene identifiers and expression differences . A B 33 Figure 2 - 4: Molecular identification of od - 2 as LCB1A . LCB1A_SOLLC LCB1_ARATH LCB1B_SOLLC SPTC1_HUMAN Q6NUL7_HUMAN LCB1_YEAST SPTC2_HUMAN SPTC3_HUMAN LCB2_YEAST K4C4I2_SOLLC LCB2B_SOLLC LCB2A_SOLLC LCB2A_ARATH LCB2B_ARATH F4JF53_ARATH KBL_HUMAN A8K228_HUMAN Q2QKD2_ARATH Q9LZ63_ARATH K4C4N9_SOLLC Q8GW43_ARATH HEM1_YEAST LCB1 LCB2 KBL B C SlLCB1A TEYYNVPVDKIDIITAAMGHALATEGGFC 328 SlLCB1B TEHCKVPSDKIDIITAAMGHALATEGGFC 353 AtLCB1 AEHHSVPIEKIDVVTAAMGHALATEGGFC 325 SlLCB2A CELLGVDTADVEIMMGTFT K SFGSCGGYI 320 SlLCB2B CELLGVDTADVDIMMGTFT K SFGSCGGYI 320 AtLCB2A CELLGVDTSDVDIMMGTFT K SFGSCGGYI 320 A tLCB2B CELLGVDTADVDVMMGTFT K SFGSCGGYI 320 * : .:::: .:: :::.: **: A WT .. cta aaa gat att cag ggg tag gaa ata atg agc gat .. od - 2 .. cta aaa gat att cag gg t agg aaa taa tga gcg at .. WT .. L K D I Q G M E I M S D .. od - 2 .. L K D I Q G W K * * A .. Solyc11g019950.1 1 Kb 34 of the sequencing data identified a total of 18,705 expressed genes in either WT or od - 2 . In comparing gene expression levels between the two genotypes, the threshold for differential expression was defined as log 2 FC > 1 or < - 1 with an FDR cutoff of 0.05 . This analysis identified 1924 and 1059 genes that were up - regulated and down - regulated , respectively, in od - 2 relative to WT (Supplemental File Table 2 - 1, and 2 - 2) . G ene ontology (GO) analysis of differentially regulate the biological processes most significantly overrepresented in genes expressed to lower levels in od - 2 than WT leaves (Supplemental File Table 2 - 3) . In the case of genes that were up - regulated in enriched GO categories ( Supplemental File Table 2 - 4 ). We next used a recursive search (string - db.org) to identify groups of differentially expressed genes that potentially interact within a n etwork (Crosara et al. 2018) . As depicted in Figu re 2 - 3 results from this analysis further suggest that gene s down - regulated in od - 2 and involved terpenoid and flavonoid synthesis are well connected , and genes up - regulated in od - 2 involved in various stress responses, including phenylpropanoid synthesis, jasmonate response, and sphingolipid synthesis are also well connected . Odorless - 2 encodes a tomato homolog of LCB1 O d - 2 was previously mapped to a 6 - centimorgan region on chromosome 11 (Kang et al. 2010 a ) . In order to identify the causal mutation in od - 2 , we performed whole - genome sequencing of both the homozygous ( od - 2 / od - 2 ) mutant and its WT parental line (cv Castlemart). Comparison of the target interval for nucleotide polymorphisms identified a single ba se - pair deletion at position 9,815,979 bp of the short arm of chromosome 11. This mutation is located in exon 11 of Solyc11g019950 and is predicted to generate a frameshift followed by a pre - mature stop codon 35 ( Figure 2 - 4A ). The mutation was verified by san ger sequencing a fragment of Solyc11g019950 from 3 independent od - 2 plants and 2 wildtype plants. Solyc11g019950 is predicted to en code a protein called Long Chain Base 1 (LCB1) . The predicted protein has a high degree of similarity to LCB1 from Arabidopsi s thaliana (75% sequence ident ity ), Saccharomyces cerevisiae (36% sequence identi ty ), and a Homo sapiens family member called SPT1 (46% sequence identi ty ). The pre - mature stop codon from the od - 2 mutation is predicted to truncate 93 amino acids from the C - terminus of the 486 - amino - acid full - length protein ( Figure 2 - 4A ). LCB1 and a related protein, LCB2 , have been extensively characterized as interacting subunits of the serine palmitoyl transferase complex involved in the biosynthes is of sphingolipids (Chen et al. 2006; Gable et al. 2010; Lynch and Fairfield 1993; Nagiec et al. 1994) . Phylogenetic analysis of the LCB family place d th e tomato Od - 2 / LCB1 gene within the LCB1 subclade ( Figure 2 - 4B ) . BLAST analysis of the tomato genome with Od - 2 / LCB1 as a query identified a closely related paralog, Solyc08g066020. We hereafter refer to this second gene (Solyc08g066020) as LCB1B and designate Od - 2/LCB1 (Soly c11g019950) as LCB1A (Figure 2 - 4B). A major difference between members of the LCB1 and LCB2 subclades is the ability to bind the coenzyme - phosphate (vitamin B6) within the LCB1 - LCB2 heterodimeric complex (Chen et al. 2006) . Sp ecifically, LCB2s have a characteristic lysine that is required for binding the co - factor, whereas LCB1 proteins lack lysine at this position. The absence of this signature residue in tomato LCB1A is consistent with its identity as a member of the LCB1 sub clade (Figure 2 - 4C). To test whether the identified mutation in LCB1A affects sphingolipid synthesis, we measured the sphingolipid content in lyophilized leaves from 4 - week - old WT and od - 2 plants. The most abundant sphingolipid in WT leaves was d18:2 gluc osyl ceramide ( Figure 2 - 5A ), consistent with previous studies (Markham et al. 2006) . In od - 2 leaves, both the total level of glucosyl 36 Figure 2 - 5: Changes in sphingolipid composition of od - 2 lea ve s are complemented by a 35S::LCB1A transgene . Sphingolipid profiling was performed with leaves from WT, od - 2 , p35S , and pMKS plants . b b b C b b a b a b b a b b a c b c B C a c b d A a b a b WT 37 Figure 2 - 6 : Leaf morphology of LCB1A complementation lines A WT od - 2 p35S pMKS B C a a b b 0.2 0.3 0.4 0.5 Leaf b lade t hickness (mm) 38 ceramides and the amount of the abundant C16:0 acyl species ( Figure 2 - 5B ) was reduced to approximately 66% and 58%, respectively, of WT levels. Leaves of od - 2 , however, were not deficient in all classes of sphingolipids. The most obvious difference was an increase in the amount of N - acetyl glucosyl - inositol - phosphoryl - ceramides (NacGIPC) in th e mutant relative to WT ( Figure 2 - 5 C ). These findings demonstrate that the od - 2 mutation is associated with changes in leaf sphingolipid content and composition. To determine whether the mutation identified in LCB1A was responsible for the altered sphingolipid composition of od - 2 leaves, we performed genetic complementation experiments in which the od - 2 mutant was transformed with the WT LCB1A cDNA. Repeated attempts were unsuccessful to PCR amplify and clone or have manufactured the LCB1A native promoter for use in this experiment, likely due to high AT content and many long repeats of this region. As an alternative approach, we chose to express LCB1A from the strong constitutive 35S promoter and, given the multiple trichome - related phenotypes of the mutant (Kang et al. 2010 a ) , a type VI trichome - specific promoter from the S. habrochaites METHYL KETONE SYNTHASE 1 ( MKS1 ) gene as well (Akhtar et al. 2013) . Agrobacterium mediated transformation was used to introduce the 35S and MKS promoter driven LCB1A constructs into the od - 2 genetic background. Multiple independent lines were generated for each construct and, following self - pollination for two generations, T3 lines homozygous for either 35S : :LCB1A ( p35S ) or MKS : : LCB1A ( pMKS ) were obtained. Sphingolipid profiling showed that b oth the glucosyl ceramide and NacGIPC content of p35S leaves largely resembled that of the WT ( Figure 2 - 5A - C ). These data demonstrate that the sphingolipid defect in od - 2 leaves results from mutation of LCB1A . 39 0 5 10 15 Type VI trichome density (trichomes mm - 2 ) a b b c C A WT od - 2 pMKS p35S 0 25 50 75 Type VI trichome width ( m) a c b b B Figure 2 - 7 : Trichome morpholog y of LCB1A complementation lines 40 Cell type - specific complementation of od - 2 leaf traits We next addressed the question of whether the pleiotropic phenotypes of od - 2 leaves (Kang et al. 2010 a ) , including changes in leaf morphology and defective trichom e development, are also caused by mutation of LCB1A . Among 6 independent 35S :: LCB1A lines tested, all displayed an overall leaflet size and thickness that was comparable to WT ( Figure 2 - 6 ). By contrast, all MKS : :LCB1A lines tested ( n= 4) had small, thick leaves that were indistinguishable from the od - 2 mutant (Figure 2 - 6). Thus, genetic c omplementation of the leaf morphology by p35S (but not pMKS ) correlates with the biochemical rescue of the sphingolipid profile ( Figure 2 - 5 ). Measurements of type VI trichome morphology (i.e., width of four - celled glandular head) showed that this trait was partially rescued in both p35S and pMKS lines ( Figure 2 - 7 ). We noted, however, that the frequent occurrence of atrophied a nd discolored type VI trichomes on od - 2 leaves was occasionally observed in p35S lines, whereas the general appearance of type VI GTs in pMKS lines was indistinguishable from WT ( Figure 2 - 7 A - B ). Measurements of type VI trichome density showed that this leaf trait was largely (but not completely) restored to that obser ved in WT ( Figure 2 - 7 C ). The od - 2 mutant was originally identified on the basis of its altered leaf aroma, which was associated with a defect in terpenoid and flavonoid accumulation in type VI trichomes (Kang et al. 2010 a ) . To determine the extent to which these chemical phenotypes were rescued in the complemented lines, we compared the levels of the most abundant trichome - derived terpenoids from WT, od - 2 , p35S , and pMKS plants. As shown in Figure 2 - 8A , GC - MS analysis of leaf surface extracts showed that p35S plants fail to accumulate virtually all terpenoids identified in the WT. The terpenoid profile of extracts obtained from pMKS plants, by contrast, was nearly identical to WT. Qu antitative data for - phellandrene, the major terpenoid in type VI trichomes, is shown in 41 Figure 2 - 8 : Leaf t erpenoid and flavonoid content in LCB1A complementation lines 4.00 4.50 5.00 5.50 Retention time (minutes) WT pMKS p35S od - 2 A - phellandrene - phellandrene - pinene 2 - carene - pinene 2 - carene B C WT od - 2 p35S pMKS 0 50 100 200 300 - phellandrene content ( g g - 1 FW) WT od - 2 p35S pMKS 0 200 400 600 800 1000 Rutin content ( g g - 1 FW) a a b b a a b b Relative abundance 42 Figure 2 - 9: Phenotyp ic rescue in F 1 progeny of crosses between p35S and pMKS . p35S x pMKS WT B WT od - 2 p35S pMKS p35S x pMKS A C a a b b a c c 0 50 100 200 300 - phellandrene content ( g g - 1 FW) D a b bc b a d cd 0 100 200 300 400 500 600 Rutin content ( g g - 1 FW) 43 Figure 2 - 8B . Similar results were obtained from LC - MS analysis of rutin, which is the major flavonoid compound in type VI trichomes ( Figure 2 - 8C ). These data provide evidence that cell type - specific expression of LCB1A in type VI trichomes is sufficient to support th e synthesis and accumulation of specialized metabolites produced in this cell type. The inability of 35S :: LCB1A to restore terpenoid and flavonoid production in the od - 2 genetic background further suggests that the 35S promoter is not expressed to sufficie nt levels in type VI trichomes. To further test whether the distinct LCB1A - complemented phenotypes of the p35S and pMKS lines results from cell type - specific expression of the respective promoters, we analyzed the leaf morphology and trichome - associated ph enotypes of F 1 progeny derived from reciprocal crosses between p35S and pMKS lines, and in which the 35S::LCB1A and MKS::LCB1A transgenes are hemizygous. In contrast to the partial (i.e., trait - specific) complementation phenotypes of the two parental lines, F 1 progeny from reciprocal crosses exhibited full recovery of normal leaf morphology, trichome morphology, and leaf surface chemistry ( Figure 2 - 9 ). DISCUSSION Type VI GTs of tomato provide an attractive experimental system in which to study specialized metabolic pathways for terpenoid and flavonoid biosynthesis in a single cell type (Schilmiller et al. 2008; Tissier 2012) . Over the past 15 years, characterization of various mutants affected in the development and chemical composition of type VI glands supports the view that these epidermal st ructures are an important part of a multilayered defense system for resistance to insect herbivores (Bleeker et al. 2012; Kang et al. 2014, 2016; Kang et al. 2010a , 2010b ; Li et al. 2004; Xu et al. 2018) . Much remains to be learned, however, about the specific cellular and biochemical processes required to coordinate the developmental and metabolic fate of glandular trichomes in tomato and other plant species. With this goal in mind, we set out to identif y the 44 molecular basis of the defect in type VI trichome development in the tomato od - 2 mutant (Kang et al. 2010 a ) . The od - 2 mutation is predicted to truncate 93 amino acids (~20% of the full - length protein) from th e C - terminus of LCB1A. Previous structure - function studies in yeast provide evidence that the C - terminal end of LCB1 and LCB2 are important for regulation of SPT activity or substrate specificity of the enzyme (Monaghan et al. 2002) . Thus, it is possible that production of truncated LCB1A protein in the od - 2 mutant alters, but does not eliminate, the activity of the SPT complex. This, as well as the observation that null mutations in LCB1 are lethal in Arabidopsis and yeast ( Saccharomyces cerevisiae ) (Buede et al. 1991; Chen et al. 2006; Shi et al. 2007) , provides a unique opportunity to investigate the importance of sphingolipid synthesis in pla nts and eukaryotes in general. It is also possible that the LCB1B paralog of LCB1A contributes to sphingolipid synthesis in the od - 2 mutant by compensating for the loss of LCB1A. This hypothesis is supported by our RNA - seq data showing that LCB1A is expres sed to about 8 - fold higher levels than LCB1B in WT leaves, and that the absolute level of LCB1B mRNA abundance in od - 2 leaves increases nearly 7 - fold relative to WT levels. Future studies using CRISPR - Cas9 technology to generate null mutations in tomato LC B1A and LCB1B should help to resolve this question. Ceramides, glycosyl inositol phospho - ceramides (GIPC s ), and other classes of s phingolipids perform a variety of basic cellular functions in eukaryotes, including vesicle trafficking, signaling, and modulation of membrane structure (Michaelson et al. 2016; Simons and Ikonen 1997) . These functions directly impact processes such as necrosis, cell differentiation, cell cyc le, and cell survival (Berkey et al. 2012) . In p lant s, sphingolipids have been implicated in the maintenance of turgor pressure through the activation of calcium channels in guard cells and pollen tubes (Chen et al. 2015; Ng et al. 2001). Sphingolipid production is required for normal salicylic 45 acid (SA) production , and SA - mediated plant immunity (Riv as - San Vicente et al. 2013 ; Wang et al. 2008) , as well as pollen development and cell adhesion (Bown et al. 2017; Msanne et al. 2015) . Sphingolipid species containing very - long - chain fat ty acids are required for plasma membrane stability, organization and cytoskeletal connections which , for example, impact the localization of auxin transport proteins (Markham et al. 2011). The od - 2 mutation does not eliminate sphingolipid synthesis but ra ther alters the composition of specific classes of these lipids . The lower abundance of d18:1 glucosyl ceramides in od - 2 is reminiscent of the altered sphingolipid composition resulting from RNAi - based suppression of LCB1 in Arabidopsis (Chen et al. 2006) . In this context, it has also been shown that pharmacological down - regulation of LCB1 activity in yeast le ads to lower abundance of sphingolipids, including inositol phosphoryl ceramides (Huang et al. 2012) . W e a lso found that expression of the WT LCB1A cDNA from the 35S promoter rescued most of the changes in sphingolipid content in od - 2 plants, demonstrating a functional role for LCB1A in sphingolipid metabolism. These changes in sphingolipids are apparently res ponsible for the pleiotropic phenotypic changes seen in od - 2 . Given the many known roles of sphingolipids in eukaryotic cells and their importa nce in cellular differentiation, it is perhaps not surprising that homozygous od - 2 plants exhibit a wide spectrum of abnormalities. For example, t he disorganization of multiple cell layers in od - 2 leaves appears to underlie the observed changes in leaf thickn ess, with larger mesophyll cells potentially accounting for the decreased mesophyll conductance of the mutant. Sphingolipid involvement with control of the cell cycle and polar localization of signaling factors could impact tissue development processes suc h as cell - size regulation or asymmetric cell division (Cheng et al. 2001; Humphries et al. 2011) . It is possible that c ell size regulation through changes in ploidy contribute to cell morphological phenotypes in od - 2 , as has been seen in mutants of Arabidopsis (Breuer et al. 2 009; 46 Sugimoto - Shirasu and Roberts 2003) . It has also been reported inhibition of SPT activity in human cells results in cell cycle arrest prior to mitosis at the G2/M transition (Lee et al. 2011) . Aanalysis of ploidy levels in the various cells of od - 2 , would provide a test of this hypothesis. The full recovery of normal leaf morphology, including mesophyll patterning, in p35S but not pMKS lines indicates that production of sphingolipids are critically required for normal blade expansion and photosynthetic capacity. The full recovery of type VI trichomes in the pMKS lines but not the p35S lines further shows the importance of sphingolipid s i n the development of this specialized structure. The use of cell - and tissue - type promoters for these experiments highlights the cell - autonomous role of sphingolipid s in trichome development and metabolic function (e.g., volatile terpenoid production) as well . I t is conceivable that sphingolipid deficiency disrupts one or more processes involved in the t ransport and storage of terpen oids and other specialized metabolites . Unidirectional (polar) transport of metabolites in glandular cells is known to depen d on targeting of proteins and vesicles to specific region s of the plasma membrane . Polar localization of such components commonly involves lipid rafts , which are characterized by a high abundance of sphingolipids (Simons and Ikonen 1997). Long distance transport of other lipids has been demonstrated in species across the plant kingdom (Guelette et al. 2012; Hoffmann - Benni ng et al. 2002) . Our data, however, suggests that sphingolipids act cell autonomously in gland cells of the type VI trichome. Future studies directed at understanding the precise effects of altered sphingolipid synthesis on the accumulation of terpenoid s and other specialized metabolites in type VI trichomes is warranted. Some trichome phenotypes of od - 2 are similar to those of another tomato mutant, anthocyanin free ( af ), but there are some key differences. The production of trichome - associated 47 terpenoi ds and flavonoids is severely compromised in both od - 2 and af . In af plants, low flavonoid levels are caused by a mutation that abolishes the activity of chalcone isomerase 1 (CHI1), an enzyme that is required for the biosynthesis of flavonoids (Kang et al. 2014) . Although the mechanism by which flavonoids promote terpenoid accumulation in type VI trichomes remains to be determined, these findings raise the possibility that the terpenoid deficiency in od - 2 trichomes is related to low flavonoid production. This hypothesis is supported by the very low levels of rutin and other flavonoids in od - 2 glandular trichomes, even though this mutant accumulates normal anthocyanin levels in other cell types of the leaf . Of particular relevance to the trichome chemical phenotypes of od - 2 is the low expression of terpenoid and flavonoid biosynthetic genes . T he final steps of - phellandrene biosynthesis are catalyzed by neryl - diphosphate synthase 1 (NDPS1) and phellandrene synthase (PHS) (Schilmiller et al. 2009). Interestingly, the genes encoding these two enzymes are among the most strongly down - regulated in od - 2 leaves. Reduced expression of these genes may explain the low abundance of volatile terpenoids in od - 2 . A similar correlation was observed between low expression of flavonoid biosynthetic genes and the flavonoid deficiency of od - 2 type VI trichomes . Previous studies showed o ver - expression of the transcription factor MYB75 ( encoded by Solyc10g086250) promotes the accumulation of f lavonoids (Jian et al. 2019) . The low expression of this gene in od - 2 leaves relative to WT suggests that altered MYB75 activity may contribute to the metabolic phenotypes of od - 2 . Another MYB transcription factor, MYB101 ( encoded by Solyc07g052300) , is also down regulated in od - 2 . Increased activity of MYB101 homologs in Arabidopsis and Solanum tuberosum homologs via reduc ed miR159 expression caused an increase in trichome density (Pieczynski et al. 2013) , wh ereas overproduction of miR159 caused a reduction of trichome abundance and enlarge d mesophyll cells in Arabidopsis (Alonso - Peral et al. 2010) . These 48 collective observations support the notion that changes in the activity of specific transcription factors may account for the pleiotropic metabolic phenotypes of the od - 2 mut ant. The developmental phenotypes of od - 2 leaves suggests that differentiation of protodermal cells into trichomes or stomata may also be influenced by changes in sphingolipid content. For example, we found that the tomato homolog (Solyc08g079980) of the Arabidopsis STOMATA CELL DENSITY DISRUPTION 1 ( SDD1 ) gene is strongly upregulated in od - 2 leaves . Previous studies showed that o verexpression of SDD1 in Arabidopsis resulted in lower stomata density and developmental arrest of stom ata (Von Groll et al. 2002) . Moreover, ssd1 mutants display altered trichome branching , consistent with a role for SSD1 in trich ome development (Lawson et al. 2014) . Similar to the ssd1 mutant , Arabidopsis mutants defective in MPK3 have paired stomata. We found that the tomato homolog of At MPK3 (Solyc06g005170) is also up regulated in od - 2 leaves . Down - regulation of an Arabidopsis ERECTA - LIKE (ERL1 and ERL 2) homolog of tomato (Solyc03g007050) may contribute to the large size and disorganization of cells in od - 2 leaves , which is reminiscent of the cell morphology defects observed in the corresponding loss - of - function mutants in Arabidopsis (Shpak et al. 2003) . We also o bserved d own regulation of two tomato homologs (Solyc01g104210 and Solyc07g018240) of the Arabidopsis FATTY ACID ELONGATION 1 and 2 genes ( ELO1 and ELO 2 ) in od - 2 . Leaves of the Arabidopsis elo1 and elo 2 mutants have i ncreased cell size, smaller blade area, and reduced numbers of palisade cell s (Falcone e t al. 2007; Nelissen et al. 2005) . We also noted that homologs (Solyc03g096770, Solyc03g096780, and Solyc06g072990) of the Arabidopsis LOW SULFUR (LSU) genes are all down regulated in od - 2 ; the fourth (Solyc03g096760) was not within the - 1 LFC cutoff (L FC - 0.97 8 , FDR 0.0005). In Arabidopsis, LSU expression is induced in response to low sulfur, low iron, basic pH, excess copper, or high salt conditions (Garcia - Molina et al. 2017) . It has also been 49 shown that elevated LSU expression promote s the production of H 2 O 2 from O 2 - during various stress conditions by activating superoxide dismutase, whereas reduced expression of LSU is results in low H 2 O 2 levels (Garcia - Molina et al. 2017) . Another sulfur gene homolog (Solyc06g007970) of the Arabidopsis SULFUR DEFICIENCY INDUCED 1 (SDI1) is also down regulated. SDI1 is involved in the utilization of SO 4 2 - while under sulfur depravation, and low expression is associated with sulfur replete conditions (Howarth et al. 2009) . Further investigation of sulfur metabol ism will be required to verify if there is a deficiency in od - 2 . A non - targeted metabolomics study of od - 2 could additionally identify other deficiencies or metabolic imbalances in the mutant. Restoration of type VI trichome size and density in both the p35S - and pMKS - c omplementation lines demonstrates the importance of sphingolipids in development of this trichome type . However, the precise spatial and temporal requirement for these lipids in development and initiation of type VI trichomes from the epide rmal pavement cells remains to be determined . Assuming th at sphingolipids function cell autonomously, as our data suggest, it is possible that weak expression of LCB1A in trichomes of p35S plants, as well as weak expression of LCB1A in epidermal cells of p MKS plants, is sufficient to allow normal specification and differentiation of the type VI structure . T he presence of atrophied type VI trichomes in the p35S but not pMKS lines , however, suggests that later stages of trichome development (e.g., gland filli ng) require a certain threshold of sphingolipid synthesis for normal trichome function. In c ompari ng the extent to which the 35S: : LCB1A and MKS : :LCB1A transgenes complemented the various phenotypes of od - 2 , this study provides new insight into the tissue - spec i fic functions of sphingolipids in tomato. Whereas 35S : :LCB1A restored leaf and trichome morphology , it did not re scue flavonoid and terpenoid production in type VI GTs. Co nversely, MKS : : LCB1A did not restore leaf morphology but did fully res to re all trichome - related 50 phenotypes . It may be possible to extend this approach to additional tissue - or cell type - specific promoters, such as mesophyll - or epidermal cell - specific promoters, to better understand the cell type - specific roles of sphingolipid s in plants. 51 APPENDIX 52 Primer name Sequence Solyc ID Target gene 1R CCTGGTATGGTTTCAGGTCTAA Solyc11g019950 LCB1A 2L CTTCAAGACCTGCAATGACAAT Solyc11g019950 LCB1A 103L cagaaagcttCCGGTGAAGATTGTCGTTC NA pMKS1 104R agtggtctagaTTTTGAACAACTACCACAAAG NA pMKS1 26L ggggaccactttgtacaagaaagctgggtcctaGCTCTGATCTGTCAATAGCG Solyc11g019950 LCB1A 105L ggggacaagtttgtacaaaaaagcaggctagATGGATACAACAATATCAGTTC Solyc11g019950 LCB1A Table A - 1: PCR primers used in this study. 53 Figure A - 1 : Additional photosynthetic parameters in wildtype and od - 2 leaves. WT od - 2 0.0 0.2 0.4 0.6 Stomatal c onductance (g sw , mol m - 2 s - 1 ) p = 0.726 WT od - 2 6 8 10 12 Triose p hosphate u tilization (TPU, mol m - 2 s - 1 ) p = 0.67 WT od - 2 0 50 100 150 200 Electron t ransport r ate (J, mol m - 2 s - 1 ) p = 0.43 WT od - 2 0 5 10 15 20 25 30 Carbon a ssimilation (A, mol m - 2 s - 1 ) p = 0.831 p = 0.08 WT od - 2 0 50 100 200 300 (Vc max , mol m - 2 s - 1 ) A B C D E 54 CO 2 partial pressure at r u b is co (Cc, Pa) WT od - 2 0 10 20 30 40 50 p = 0.23 WT od - 2 0 1 2 3 4 5 15 N AIR (%/100) p < 0.05 WT od - 2 - 40 - 30 - 20 - 10 0 13 C VPDB (0/00) WT od - 2 0 5 10 15 20 C:N ratio p < 0.005 Figure A - WT od - 2 0 1 2 3 4 Chlorophyll a/b ratio p = 0.1 WT od - 2 0 10 20 30 40 50 % carbon in DW p < 0.05 F G H I J K p = 0.231 55 CHAPTER 3 CELL TYPE SPECIFIC COMPLEMENTATION OF CHALCONE ISOMERASE IN S. LYCOPERSICUM 56 ABSTRACT Plants produce an array of structures and compounds that deter herbivory. Unfortunately, analysis of a mutation which affects herbivory is often confounded by ple i otropic phenotypes. Some specialized structures , such as glandular trichomes , not only physical ly deter herbivory but are also the source of defense compounds that deter herbivory. Complementatio n of the tomato mutant anthocyanin free ( af ) by expressing chalcone isomerase 1 (CHI1) with the Solanum habrochaites methyl ketone synthase 1 promoter resulted in expression of CHI1 specifically in type VI glandular trichomes, restored type VI trichome fla vonoid and terpenoid levels, increased resistance to herbivory by the two - spotted spider mite, Tetranychus urticae , but did not restore anthocyanin production. Complementation of the tomato mutant af by expressing CHI1 with the Cauliflower Mosaic virus 35S promoter resulted in expression of CHI1 in leaf tissue, restored type VI trichome flavonoid and leaf anthocyanin levels, but did not restore type VI trichome terpenoids, and had no effect on herbivory by the two - spotted spider mite. Deterrence of spider m ite - herbivory correlated with restoration of terpenoid production and expression of CHI1 in type VI trichomes. INTRODUCTION Investigations of plant mutants that are compromised in resistance to herbivore deterrence have identifi ed many genes that control plant development and responses to the environment. T hese genes are often identified through the study of mutat ions that ha ve ple i otropic phenotypes . One example is the jasmonate insensitive 1 - 1 ( jai1 - 1 ) mutant in tomato , which dis rupts the gene CORONATINE INSENSITIVE 1 (COI1) . COI1 codes for an F - box subunit that is responsible for targeting transcriptional repressors for degradation in the presence of jasmonate (Chini et al. 2007; Thines et al. 2007) . The jai1 - 1 mutant has fewer type VI glandular trichomes (GT s ), reduced levels of volatile terpenoids, lower levels of anti - insect proteins such as protease inhibitors, reduced 57 sensitivity to jasmonate, reduced fertility, and minimal amounts of anthocyanin (Li et al, 2004). Other plant defense genes have been id entified through profiling various tissues and treatments for induction of genes and those are then assigned a function based on in - vivo activity , such as polyphenol oxidase (Haruta et al, 2001). These and other investigations laid the groundwork for under standing the gen es and molecular phenomena that impact plant - herbivore interactions. What is not clear from these findings is for which cell types these genes hold importan ce , and what specific metabolites or structures confer the herbivore deterrence seen . For example , Li et al (2004) associate increased spider mite feeding on jai1 - 1 with reduced abundance of GTs and volatile - terpenoids, however there are also lower levels of protease inhibitor II and polyphenol oxidase. Without separati ng each of these pleotropic mutant phenotypes into different plants that share a common genetic background , theories about the loss of herbivore deterrence due to a specific compound or structure cannot be tested. An example of separating out pleotropic phenotypes was given by Jewell and Browse (2016) where they used a variety of promoters to attempt to restore infertility to the mutant of the COI1 homolog in Arabidopsis. In Arabidopsis the coi1 - 1 mutant is male ste rile ; the stamens do not reach their wildtype length, the pollen is non - viable, and the pollen does not dehisce (Feys et al. 1994) . Restoration of COI1 activity in the tapetal layer of coi1 - 1 by using the A9 promoter resulted in n o recovery of fertility phenotypes (Jewell and Browse 2016) . Use of the cell specific promoters for the filament by the DEFECTIVE ANTHER DEHISCENCE 1 promoter , or the anther epidermis by the WUSHEL promoter resulted in partial recovery of the coi1 - 1 fertility phenotypes (Jewell and Browse 2016) . However, use of the epidermal specific L IPID T RANSFER P ROTEIN 1 promoter to express COI1 in the coi1 - 1 mutant restored all fertility deficiencies normally seen in coi1 - 1 (Jewell and Browse 2016) . This series of experiments clearly showed that 58 sensing of jasmonate is required in the epidermis of filaments and anthers to provide normal male fertility and identified the cell types that require COI1 protein for proper development of stamen and pollen (Ishiguro et al. 2001; Sanders et al. 2000) . The S. lycopersicum mutant an thocyanin free ( af ) contains a frameshift mutation in the gene encoding chalcone isomerase 1 (CHI1) . As a consequence, CHI1 protein does not accumulate in the mutant (Kang et al. 2014) . CHI1 catalyzes the second step in flavonoid production by isomeriz ing naringenin chalcone to naringenin. In addition to expected mutant phenotypes such as decreased anthocyani n and other flavonoid levels , the af mutant displays sev eral ple i otropic phenotypes that are more difficult to interpret based on a defect in CHI1 . Of particular interest is the impaired production of type VI trichome volatile terpenoids and the increased susceptibility to insect herbivores (Kang et al. 2014; Rick et al. 1976) . The difficulty in interpreti ng this phenotype is related to the fact that terpenoids and flavonoids are derived from largely independent pathways that operate in distinct subcellular compartments: anthocy anin and flavonoid s ynthesis occur in the cytoplasm and ER , whereas isoprenoid synthesis via the methyl - erythritol phosphate ( MEP ) pathway occurs in the chloroplast , which is the source of most volatile terpenoids (Balcke et al. 2017; Banerjee and Sharkey 2014) . The flavonoid biosynthesis substrates are malonyl - CoA and - coumaroyl - CoA , w hile biosynthesis of terpenoids via the MEP pathway utilizes glyceralde hyde - 3 - phosphate and pyruvate. The different substrates utilized for the flavonoid and terpenoid pathways makes metabolic feedback an unlikely source of regulation in type VI trichomes. Yet r estoration of all phenotypes, including terpenoid accumulation, w as achieved by expressing wild - type CHI1 from the native promoter ( pCHI ) in the af mutant (Kang et al. 2014) . Synthesis of mono - terpenoids in type VI trichomes originates in the plastids through the MEP pathway . The pathway culminat es in the production of isopentenyl pyrophosphate (IPP) and 59 d imethylallyl pyrophosphate (DMAP). As an example, neryl diphosphate synthase 1 (NDPS 1 ) combines two isoprenoid units to form neryl diphosphate ( N P P ), then phellandrene synthase 1 (PHS1) cyclizes N P P to form - phellandrene . P lants also produce isoprenoid units from the mevalonic acid ( MVA ) pathway in the cytosol and ER. Th e MVA pathway is not considered the major source of substrates for volatile terpenoid production in tomato , but may be a source of sesquiterpenes (Sallaud et al. 2009; Schilmiller et al. 2009) . Many genes have highly specific expression patterns that reflect their function (Hause et al. 2000; Mohan, Vijayan, and Kolattukudy 1993; Pattison et al. 2015) . Publicly available expression data, h owever, indicates that CHI1 (Solyc05g010320) has a broad expression pattern excluding ripe fruit (bar.utoronto.ca/eplant_tomato) (The Tomato Genome Consortium 2012) . This pattern correlates well with visual accumulation of anthocyanins in most above ground tissues, and the low levels of flavonoids in fruits except naring e n i n chalcone in fruit skin due to a lack of chalcone - isomerase activity (España et al. 2014; Muir et al. 2001) . The broad expression pattern of CHI1 makes it difficult to attribute specific phenotypes in af to deficienc y of CHI1 expression in specific cell types , such as trichomes. T o better understand the function of CHI1 in particular cell type s of tomato , I adopted an experimental approach in which general and tissue specific promoters were used to drive the expression of CHI1 in the af mutant background. The native CHI1 promoter was used to drive expression of CHI1 ( pCHI1 ) to verify phenotypes associated with a normal CHI1 expression pattern (Kang et al. 2014) . The cauliflower mos aic virus 35S promoter was used to induce expression of CHI1 ( p35S ) throughout much of the plant. The 35S promoter has been used extensively to promote over - expression in a number of species, especially in leaves (Sanders et al. 1987) . The 35S promoter is specific to particular cell types within the tissues it is expressed 60 (Benfey and Chua 1990; Zhang et al. 1995) . The S. lyc opersicum type VI trichome specific methy l ketone synthase 1 ( MKS1 ) promoter from Solanum habrochaites w as used to drive expression of CHI1 ( pM KS ) (Akhtar et al. 2013) . Transgenic lines in which CHI1 is expressed in specific tissues were used to study the ple i otropic phenotypes of the af mutation separately, and to better define the connection between the loss of flavonoid production and the loss of terpenoid production in type VI trichomes . MATERIALS AND M ETHODS Plant material and growth conditions Plants were grown in a single growth chamber operated at 28 ° C/18 ° C and 16 hr / 8 hr, 2 5 0 E light/dark. S eed for anthocyanin free ( af ) (LA1049 ) and the pCHI1 line (line CHI4) w ere previously described (Kang et al. 2014) . Seed from the homozygous p35S line (line D3 ) was used throughout and seed from a he mi zygous (line H4 ) plant was used in Appendix Figure B - 1 . Seed from th e pMKS line (line 50) was used for all experiments. Seeds were bleach sterilized for 20 minutes with 20% bleach (2.5% Sodium hypochlorite) prior to use, germinated in Jiffy 7 Peat Pellets 42 mm, and transferred to 6 cm x 6 cm x 8 cm pots with soil prior to week t hree . Plasmid construction 35S: : CHI1 expression vector (pBI - 35S::CHI1) was previously published (Kang et al. 2014) . A fragment containi ng the S. habrochaites methyl ketone synthase 1 ( MKS1 ) promoter (see GenBank: GU987105.1) (Akhtar et al. 2013) w as PCR amplified with primers 103L and 104 R ( Appendix , Table B - 1 ) using the MKS1 - GFP expression vector as a template. The 1644 b p product was digested with HindIII and XbaI. The 35S promoter in pGWB - 402 was released by digest ion with HindIII and XbaI , and the pMKS1 fragment was inserted in its place resulting in pGWB - 40x_MKS1 . A CHI1 cDNA from S. lycopersicum cv. Castlemart was amplified with 61 primers 34R and 102L ( Appendix , T able B - 1 ). The resulting 906 bp fragment was used in an LR reaction with pGWB - 40x_MKS1 , resulting in expression vector MKS 1 : : CHI1 . LR reaction s were per formed with LR Clonase TM II ( ThermoFischer ) as described by the manufacturer . Plant and Agrobacterium t ransformation The 35S::CHI1 and MKS1 : : CHI1 expression vectors w ere introduced into Agrobacterium tumefaciens strain AGL0 using the heat shock method from Weigel and Glazebrook (2002). Transformation of the af mutant was conducted in the Plant Biotechnology Resource and Outreach Center of Michigan State University ( http://www.ptc.msu.edu/index.html ). For each expression vector, c otyledons from 1 00 - 15 0 one - week old af seedlings were excised and placed into minimal growth medium. Two days later , the explants were infected with A. tumefaciens harboring 35S:: CHI1 or MKS1::CHI1 plasmid s . Explants were transferred to agar plates containing 100 mg L - 1 kanamycin to select for resistant plant cells contain ing on of the plasmid s . Resulting calli were regenerated into plants and transferred to soil for seed propagation in the greenhouse . Anthocyanin quantification Anthocyanins were extracted from weighed leaflet s with 480 L methanol:HCl ( 99:1 ) for about eight hours with slow shaking at 4 ° C in the dark. The following day , 320 L deionized water and 320 L chloroform:isoamyl - alcohol ( 24:1 ) was added to each sample . The extracts were vortexed and centrifuged at 10 ,000 x g for 10 min at 4 ° C. The aqueous layer (top) containing anthocyanin pigment s was used to determine the abs orption at 530 nm on a S pectr a max M2e plate reader (Molecular Devices ) and adjusted for chlorophyll interference based on (Spitzer - Rimon et al. 2010) . Absorbance r eadings were normalized to leaflet fresh weight. 62 Metabolite analysis Extraction of trichome metabolites was conducted with 1 L methanol with 1 M tetradecane and 1 g L - 1 propyl - 4 - hydroxybenzoate as internal standards was used to extract each pre - measured leaflet (Kang et al. 2010 a ) . After two minutes of extraction with gentle shaking the solvent was transferred to 1.5 mL sealed glass containers for analysis by gas chromatography - mass spectroscopy ( GC - MS ) and liquid chromatography - mass spectroscopy ( LC - MS ) . Mass spectrometry Volatile terpenoids were quantified by injecting one L into a GC 6890N/5975B quadrupole ( Agilent ) by a CombiPAL (CTC Analytics) . The column used is a 40m VF5MS, 30 m with 10 m EZGuard, 0.25 m film and 0.25 mm inner diameter (Agilent) . Operating conditions were as follows: 280° C inlet temperature, 88.35 psi inlet pressure, 1: 10 split ratio, and 45 - 350 mass scan starting at 3.5 min. Ramp temperatures used initial 40° C for one min, 90° C at 40° C min - 1 , 110° C at 15° C min - 1 , 250° C at 25° C min - 1 , 320° C at 40° C min - 1 , and finally h eld at 320° C for 2 minutes. Total run time was 12.93 minutes. Flavonoids were analyzed by injecting o ne were injected into an Acquity Binary Solvent Manager and Colum n Manager (Waters) by a 2777C Sample manager (Waters) . A Xevo G2 - XS QTof (Waters) mass spectrometer was o perat ed under the following conditions: capillary voltage , 2 kV ; sample cone , 40 V; source temperature , 100° C ; desolvation temperature , 350° C ; desolvation flow rate , 600 L hr - 1 ; electrospray ionization in negative sensing mode. MS e was used with several functions which split each sampling timepoint : 1. No collision energy (to gather wh ole molecule masses) ; 2. Ramp of 20 - 60 V collision energy (to analyze sample molecule fragmentation patterns) ; and 3. Ion lock spray (for calibration of exact mass) . A 10 cm x 2.1 mm, 2.7 m Ascentis® Express C - 18 HPLC column (supelco) held at 40° C was used for all 63 separations . De - ionized w ater containing 0.1% formic acid was used for solvent A, and methanol was used for solvent B. Using a continuous flow rate of 0.3 mL min - 1 , the gradient profile was started at 5% solvent B and ramped to 60% solven t B over 3 minutes using curve 6 setting on the Waters solvent manager and then ramped to 100% solvent B over 4 min using curve 6, held at 100% solvent B until the 5.5 min point, dropped to 5% solvent B at 5.51 minutes, then held at 5% solvent B until 6.5 minutes. Isolation of glandular t richome s T ype VI glandular t richomes were picked by hand using a pulled glass Pasteur pipet (Kang et al. 2014) , or washed from leaf lets (McDowell et al. 2011) . For chemical analysis of picked trichomes, 200 trichomes were p laced in to 100 L 100% methanol and the resulting samp les were analyzed by either GCMS or LCMS. A pproximately 50 mg w ashed trichomes were used for protein extraction. Western blot analysis Protein was e xtract ed from leaf tissue (50 mg) or isolated trichome s as previously described (Schilmiller et al. 2010) . Protein concentration was quantified by Bradford assay using Bio - Rad Western blot analysis of CHI1 protein levels was per formed as previously described (Kang et al. 2014) using 13 g of total protein for SDS - PAGE. Spider mite feeding assays Tetranychus urticae (two - spotted spider mites) were used in both choice and no - choice assays. Spider mites were obtained from tomato and alfalfa plants growing in the greenhouse and reared on af leaflets at room temperature for longer than 10 days prior to use in the assays. The choice assay used a toothpick bridge and a 10 - minute time course in which five adult females were placed in the center of a toothpick th at was balanced between two leaflets of either the same host 64 genotype or two different genotypes. The n umber of spi der mites on each leaflet w as scored after 10 minutes. Several r eplicates were conducted with different leaflets from different plants and with different sets of spider mites. T he mites typically settled on a leaflet within the 10 - minute time frame of the assay , and mites were rarely observed to move once a leaf was reached. For t he no - choice assay , mites were placed onto leaflets (three leaflets per genotype per plate) in cubated on a petri plate containing a moist filter paper. Each genotype was tested mul tiple times in independent plates . Each leaflet was challenged with three adult female spider mites. Leaf d amage caused by mite feeding was assessed as previously described (Santamaria et al. 2012) . Leaf images were obtained with a photocopier after 7 days for feeding, and area determined using Inkscape (inkscape.org). RESULTS CHI 1 protein expression localization To better understand the function of chalcone isomerase in respect to type VI trichomes of tomato ( Figure 3 - 1A ) , we used three different promoters to express the CHI1 cDNA in the af mutant, which lacks CHI1 protein in all tissues (Kang et al. 2014) . We reasoned that ea ch promoter, namely the native CHI1 promoter, the CaMV 35S promoter, and the type VI - specific MKS promoter, would generate distinct CHI1 expression domains in the resulting CHI1::CHI1 ( pCHI ), 35S::CHI1 ( p35S ), and MKS::CHI1 ( pMKS ) transgenic lines respecti vely. To test this, we used Immunoblot analysis to determine the level of CHI1 protein in whole leaves, purified trichomes, and leaves in which trichomes were removed by shaving from each of the three tissue - specific transgenic lines, as well as the af mut ant (null control). In pCHI plants, which are evaluated as the phenotypic equivalent to the WT control, accumulated CHI1 in total leaves, trichomes, and shaved leaves lacking trichomes ( Figure 3 - 1B ). In pMKS plants, CHI1 was detected in trichomes but not 65 Figure 3 - 1: C HI 1 protein levels in leaves and trichomes of tomato genotypes used in this study . ( A) Light micrograph of a tomato leaf t ype VI trichome. Scale bar = 50 m. ( B) Western blot analysis of CHI1 protein levels in various leaf tissues from af mutant and transgenic lines expressing CHI1 from the CHI1 , MKS1 and CaMV 35S promoters. For each genotype, total p rotein was isolated from whole leaves (upp er panel), purified type VI trichomes (middle panel), or leaves in which trichomes were physically removed by shaving (lower panel). C oomassie blue staining shows r u b is c o large subunit as loading control. Shaved leaf Type VI trichomes Leaf 66 Figure 3 - 2: Tissue - specific expression of CHI1 modulates anthocyanin accumulation and trichome morphology . (A) Light micrographs of the leaf surface of the indicated genotypes. T ype VI GTs (black arrows) and anthocyanin accumulation (white arrows) in minor veins . S ca le bar = 100 m. (B) Leaf a nthocyanin levels in indicated genotypes. n = 10 samples per genotype. Each sample was from a single plant . (C) Width of the four - celled glandular head of type VI t richome s in the indicated genotypes. 5 0 trichomes per genotype . (D) Density of type VI t richome s in the indicated genotypes. n = 5 plants per genotype . Different letters within each analysis indicate significant differences based on the results of an ANOVA an d Tukey HSD test , p < 0.05. 0 5 10 15 20 Trichome density (Trichomes mm - 2 ) ab a ab b 0 20 40 60 80 100 Trichome size ( m) b a a a 0 1 2 3 4 Anthocyanin content (A530 g - 1 FW) b a b a af pCHI pMKS p35S 67 in total leaf or shaved leaf samples. p35S plants accumulated CHI1 in both the total and shaved leaf samples, with very low levels of protein detected in isolated trichomes. These findings indicate that p35S and pMKS plants express CHI1 primarily in foliar (i.e. non - trichome) and trichome cells, r espectively, and demonstrate the feasibility of using different gene promoters to manipulate the tissue - specific expression of CHI1 in tomato leaves. Tissue - specific expression of CHI1 modulates leaf anthocyanin levels Cultivated tomato plants grown under normal conditions accumulate anthocyanins which can be visualized in the ma jor veins and the abaxial surface of leaves (Von Wettstein - Knowles 1968) . Th e af mutant fails to accumulate anthocyanins in all parts of the leaf (Rick et al. 1976) . This deficiency is completely restored in pCHI1 plants, which expresses CHI1 from the native CHI1 prom oter (Kang et al. 2014) ( Figure 3 - 2 ) . Leaf anthocyanin levels were also fully restored in p35S plants, whereas pMKS plants displayed an anthocyanin - deficient phenotype similar to that of the af mutant ( Figure 3 - 2A B ). These findings are consistent with the CHI1 protein expression pattern in these lines, and further validate the utility of this approach for producing flavonoids in a tissue - specific manner. Effects of tissue - specific CHI1 expression on type VI t richome s ize and density T he size of the four - celled glandular head of t ype VI trichome s in the af mutant wa s slightly but significantly smaller than t hat of trichomes produced by the pCHI complemented line ( Figure 3 - 2C ) , consistent with previous reports (Kang et al. 2014) . A similar increase in trichome size was observed in the p35S and pMKS lines (Figure 3 - 2C). Measurements of the type VI trichome density on the adaxial leaf surface in each genotype showed no consistent differences between the lines ( Figure 3 - 2D ). 68 Figure 3 - 3: Flavonoid and terpenoid levels in type VI trichomes of CHI1 - complemented lines . Various flavonoids (A - C) and terpenoids (D - F) were measured from type VI glands isolated (hand - picked) from the indicated genotypes. Different letters within each analysis indicate significa nt differences based on the results of an ANOVA and Tukey HSD test, p < 0.05. n = at least 5 plants per genotype. Nari ngenin c halcone 0.0 0.1 0.2 0.3 0.4 g 100 - 1 trichomes a b b b Rutin 0 2 4 6 8 g 100 - 1 trichomes b a a a 0 5 10 15 Kaempferol t riglucoside Relative abundance 100 - 1 trichomes a b a b c - pinene 0.0 0.5 1.0 1.5 2.0 2.5 n g 100 - 1 trichomes b a a a 2 - carene 0 1 2 3 4 n g 100 - 1 trichomes c a b c - phellandrene 0 10 20 30 40 50 n g 100 - 1 trichomes b a a b 69 Trichome metabolite abundance in CHI1 - complemented lines We performed targeted metabolite analysis to determine how the restriction of CHI1 expression to specific leaf tissues affects flavonoid and terpenoid production in the type VI trichomes. For these studies, the glandular head of type VI trichomes was isola ted using a fine glass needle (Kang et al. 2010 a ) to minimize potential contamination by other cell types of the leaf. As previously observed (Kang et al. 2014) , type VI trichomes f rom af plants accumulated relatively high levels of naringenin chalcone (the substrate for CHI1) but had very low levels of both rutin, the major flavonoid product of CHI1 in this trichome, and terpenoids ( Figure 3 - 3 ). Expression of CHI1 from the MKS promo ter in pMKS plants largely restored the flavonoid and terpenoid content to that observed in pCHI plants . In the p35S plants flavonoids are recovered as evidenced by levels of rutin and other prevalent flavonoids ( Figure 3 - 3, Appendix , Figure B - 1 ). The p35S line did not recover the major terpenoids like - phellandrene and 2 - carene ( Figure 3 - 3D, E ) , however , some of the low abundance mono terpenes such as - pinene were found at levels comparable to those found in the pCHI line ( Figure 3 - 3F ) . Restricted expression of CHI1 to type VI trichomes confers resistance to the tw o - spotted spider mite During the growth of af and pCHI plants in the greenhouse, we consi stently observed that leaves of af plants were infested with two - spotted spider mites ( Tetranychus urticae ), whereas leaves of neighboring pCHI plants were not (Sugimoto and Howe, unpublished results). This observation suggested that CHI1 is an important f actor in tomato leaf resistance to two - spotted spider mites, and also raised the possibility that resistance results from CHI1 expression in a specific tissue type. To test this hypothesis, we perform ed no - choice feeding assays in which detached af , pCHI , pMKS , p35S leaves were infested with spider mites and leaf damage was 70 Figure 3 - 4: Trichome - specific expression of CHI1 confers resistance to the two - spotted spider mite ( Tetranychus urticae ) . Three adult female spider mites were placed on the adaxial surface of detached leaflets of the indicated genotype. Following 7 days of infestation, leaf damage was assessed by quantifying the amount of leaf surface area devoid of chlorophyll. (A) Photograph of representative leaves showing feeding damage at the end of the feeding trial . S cale bars correspond to 1 cm. (B) Quantification of leaf area dama ged by feeding . Lower case letters indica te significant differences based on Kruskal - Wallis followed by Nemenyi test ( p < 0.05 ). n = 10 leaflets per genotype . af pCHI pMKS p35S a a b b af pCHI pMKS p35S 0.00 0.05 0.10 0.15 Spider mite damage (mm 2 ) 71 assessed seven days after challenge. Spider mites caused significant damage to af leaves but did little or no damage to pCHI leaves during the feeding assay (Figure 3 - 4), consistent with the preliminary observations in the green house. Interestingly, the l evel of damage inflicted to infested pMKS leaves was similar to pCHI leaves, whereas the amount of damage to p35S leaves was similar to the af control ( Figure 3 - 4 ). Close monitoring of spider mite behavior further showed that mites were either immobilized on pCHI or pMKS leaves or had moved off the leaf completely during the feeding trial. In contrast, spider mites placed onto af and p35S leaves often laid eggs and contained newly hatched juvenile mites ( Figure 3 - 4A ). These findings establish an important role for CHI1 in resistance of tomato to the two - spotted spider mite, and further demonstrate that expression of CHI1 in type VI trichomes is sufficient for resistance. DISCUSSION Separation of pleiotropic phenotyp es by cell - type - specific expression allows proper attribution of mutant phenotypes to deficiencies in specific cell types and testing the hypothesis that there is a transmittable signal responsible for a phenotype . This approach has previously been used to address the question of whether genes act in a cell autonomous or non - autonomous manner . Differential development of chloroplasts in C4 plants exhibiting K ranz anatomy was demonstrated through cell - type specific expression of the G olden2 - like gene (Waters et al . 2008) . Complementation of Arabidopsis bri1 mutants with cell - type specific promoters showed significant morphological differences due to recovery of brassinosteroid sensitivity in specific cell types (Hategan et al. 2014) . In the case of jasmonate signaling f or male fertility, this approach was used to show that COI1 activity in the epidermis is necessary for jasmonate sensing and proper de velopment of the stamen and pollen (Jewell and Browse 2016) . Likewise, cell type specific production of bilin chromophores was used to show that mesophyll cells play a critical role in 72 certain respon ses to high light (Oh and Montgomery 2019) . Flavonoids have been implicated as signals for a variety of processes, including pollen development , pollen germination, bending of roots during development, and development of root hairs and lateral roots (Brown et al. 2001; Maloney et al. 2014; Mo et al. 1992) . A potential role for flavonoids as a signal in trichome development and metabolism has been proposed but remains to be demonstrated (Kang et al. 2014) . In this study, we dissected the various ple i otropic phenotypes of the af mutant by expressing the wildtype CHI1 gene from tissue - specific promoters in the af background . As previously reported by Kang et al. (2014), af contains no detectable CHI1 protein in le aves and no d etectable anthocyanins, indicating that this single member of the tomato CHI gene family plays a major, if not sole, role in anthocyanin production . The af mutant thus provides a genetic background in which to restore CHI1 activity in any tissue or cell ty pe for which a suitable promoter is available. As proof of concept, we compared af and pCHI1 plants to transgenic lines that express CHI 1 from either the 35S or trichome - specific MKS promoter. Anthocyanin recovery in leaves by both the pCHI1 and p35S lines , but not the pMKS line , lends support to the idea that CHI1 is expressed in the leaf lamina of the pCHI and p35S lines but not in the pMKS line s . CHI1 protein was readily detectable in type VI trichomes of the pMKS but not the p35S lines . In the p35S lines , CHI1 accumulated to high levels in the leaf lamina , consistent with the general pattern of 35S expression . Unexpectedly , however, rutin (the major type VI trichome flavonoid) production was restored in p35S whereas - phellandrene (the major terpenoid) was not. The most straightforward interpretation of this finding is that the level of CHI1 in trichomes of p35S plants is sufficient to support the biosynthesis of rutin from naringenin but is not sufficient to support terpenoid biosynthes is. In p35S lines , there was some accumulation of minor terpenoids 73 such as - pinene and limonene, but no recovery of the major terpenoids including - phellandrene and 2 - carene. Although we detected low levels of CHI1 protein in type VI GTs in the p35S line , the plants are likely able to recover flavonoid levels due to the h igh activity of the CHI protein (Jez and Noel 2002) . Species within t he mint family also accumulate both flavonoids and volatile terpenoids in glandular trichomes , although the levels of flavonoids in this case is much lower than that seen in tomato (Lange et al. 2000; Voirin et al. 1993) . In the closely related S olanum habrochaites , high - level production of terpenoids does not strictly depend on the production of large quantities of flavonoids (Balcke et al. 2017) . These ex amples indicate that the production of volatile terpenoids in tomato trichomes may also not depend on the production of large quantities of flavonoids . One potential explanation for th e unequal restoration of terpene synthesis in the p35S line trichomes is that the activity of a subset of terpene synthases is sensitive to metabolic changes resulting from the loss of CHI1 in af trichomes. Additional studies are needed to clarify the connection between CHI1 and terpenoid biosynthesis in glandular trichomes of tomato and other plants. With the abundance of flavonoid biosynthesis mutants and the identification of damage - induced volatile terpenoids (Kroes et al. 2017) , Arabidopsis could potentially provide an experimental system in which to investigate the metabolic linkage between flavonoid s and terpenoid s . Several factors may contribute to t he low expression by the 35S promoter in type VI trichomes. One possibility is that glandular cells of the type VI trichome lack the activity of factors that initiate transcription from the 35S promoter . In support of this hypothesis, previous s tudies have shown that the 35S promoter is express ed in stalk cells of tobacco floral trichome s but not in the glandular secreting cells (Benfey and Chua 1990) . Members of the bZIP TGA - like transcription factor family have been implicated as positive regulators of the 35S promoter (Benfey 74 and Chua 1990) , but whether these factors are active in type VI trichomes remains to be determined . It is also possible that gland cells of the trichome express transcription factors that negatively regulate the 35S promoter in the glandular cells (Lee and Seo 2018) . Analysis of epigenetic marks in type VI trichomes coul d help clarify regulatory processes governing transcriptional control in the type VI trichomes . M etabolic flux through the phenylpropanoid pathway impacts the production of multiple classes of compounds, am ino acids , quinones, flavonoids, anthocyanin s , and lign i n subunits (Soubeyrand et al. 2018; Suzuki et al. 2005; Wang and Maeda 2018) . In addition to specific transcription factors that bind to the promoter regions of genes encoding phenylpropanoid biosynthetic enzymes, recent studies have shown that the trans cription co - activator complex, Mediator, also plays an important role in control ling metabolic flux through the phenylpropanoid pathway (Dolan et al. 2017) . In general, t he multi - subunit Mediator complex is thought to integrate external signals such as light , temperature , and biotic stress with transcription fac tors that coordinate the balance between primary and specialized metaboli sm (Çevik et al. 2012; Kazan 2017) . Given the likely role of the Mediator complex in controlling metabolic processes in specific cell - types (Kagey et al. 2010) , it will be interesting to determine the extent to which Mediator controls specialized metabolism in glandular trichomes . Crop d amage caused by the two - spotted spider mite is a significant problem in greenhouses and other high intensity crop - p roduction systems . Spider mite infestation of cotton and strawberry, for example, resulted in 84% and 40% losses, respectively (Walsh et al. 2002; Wilson et al. 1991) . Greenhouse grown tomato is also susceptible to spider mites and can result in yield losses of up to 50% (Stacey et al. 1985) . Severa l defense - related compounds have been implicated in tomato resis tance to spider mites , including p roteinase inhibitors, acylsugars, and sesquiterpenes (Bleeker 75 et al. 2012; Leckie et al. 2016; Santamaria et al. 2012) . Considerable genetic evidence has further established an important role for jasmonate in promoting tomato resistance to this pest (Li et al. 2002, 2004) . However, because jasmonate promotes the expression of a wide range of chemical (e.g., proteinase inhibitors) and physical (e.g., trichome s) defense traits, it remains to be determined which specific jasmonate - mediated defense traits are most effective for resistance to spider mites. Observations by a previous lab member (Koichi Sugimoto) that the af muta nt is compromised in resistance to tw o - spotted spider mites provides new insight into this question. Previous work showed that the af mutation does not impair the production of several tomato defensive compounds , including the glycoalkaloid tomatine , acylsugars , and proteinase inhibitors (Kang et al. 2014) . The low level of spider mite feeding on pCHI and pMKS lines (relative to the af mutant) correlate d with high levels of trichome - borne - phellandrene and other volatile terpenoids. By contrast, p35S lines that produce d low levels of - phellandrene remained susceptible to spider mite feeding. Th e s e data suggest that CHI1 - mediated production of terpenoids in type VI trichomes contributes to the resistance of tomato to the two - spotted spider mite. In summary, d ata presented in the Chapter show that CHI1 expression in glandular trichomes compr i se s one component of a multi - layered defense system employed by cultivated tomato for defense against arthropod herbivores . In addition to its canonical role in flavonoid biosynthesis, CHI1 appears to exert broad effects on trichomes development and metabolism, which is consistent with the proposed function of flavonoids in other aspects of plant growth and development (Brown et al. 2001; Maloney et al. 2014; Muir et al. 2001) . The use of t issue - and cell - type - specific promoters as described here offe rs an attractive approach to further dissect the complex roles of flavonoids in the plant life cycle. 76 APPENDIX 77 E 0 1 2 3 4 (ng 100 - 1 trichomes) b a b a b b b a Figure B - 1 : Additional data of terpenoid and flavonoid levels in trichomes of CHI1 - complemented tomato lines. 0 10 20 30 40 50 (ng 100 - 1 trichomes) b a b a b b b a 0 2 4 6 8 ( g 100 - 1 trichomes) c ab b ab c b c a 0.0 0.2 0.4 0.6 ( g 100 - 1 trichomes) b c c c b c a c - phellandrene Rutin Naringenin - chalcone 2 - carene 0.0 0.5 1.0 1.5 2.0 2.5 (ng 100 - 1 trichomes) c ab ab bc c a c ab - pinene A B C D 78 Primer name Sequence Solyc ID Target gene 103L cagaaagcttCCGGTGAAGATTGTCGTTC NA pMKS1 104R agtggtctagaTTTTGAACAACTACCACAAAG NA pMKS1 34R ggggaccactttgtacaagaaagctgggtcctaGTGGGACCTTATTGCACACT Solyc05g010320 CHI 1 102L ggggacaagtttgtacaaaaaagcaggcttcACTAATAGATCTTAAATGG Solyc05g010320 CHI 1 Table B - 1: PCR primers used in this study. 79 REFERENCES 80 R EFERENCES Abe, T . , Thitamadee , S. , and Hashimoto , T d efects and c ell m orphogenesis in the Lefty1 L efty2 h ibulin m utant of Arabidopsis thaliana Plant and Cell Physiology 45:211 20. Agati, G . , Azzarello , E. , Pollastri , S. , and Tattini , M a ntioxidants in p lants: l ocation and f unctional s Plant Science 196:67 76. Agrawal, A . r esponses to h erbivory in w ild r adish: e ffects on s everal h erbivores and p lant f Ecology 80:1713 23. Agren, J ., and Schemske , D.W c ost of d efense a gainst h a n e xperimental s tudy of t richome p roduction in Brassica rapa The American Naturalist 141:338 50. Aharoni, A . , Dixit , S. , Jetter , R. , Thoenes , E. , van Arkel , G. , and Pereira , A c lade of AP2 d omain t ranscription f actors a ctivates w ax b iosynthesis, a lters c uticle p roperties, and c onfers d rought t olerance w hen o The Plant Cell 16:2463 80. Akhtar, T . A., Matsuba , Y. , Schauvinhold , I. , Yu , G. , Lees , H.A. , Klein , S.E. , and Pichersky , E . t omato c is - p renyltransferase g ene f Plant Journal 73:640 52. Alonso - Peral, M . M., Li , J. , Li , Y. , Allen , R.S. , Schnippenkoetter , W. , Ohms , S. , White , R.G. , a nd Millar , A.A m icroRNA159 - r egulated G a MYB - like g enes i nhibit g rowth and p romote p rogrammed c ell d Plant Physiology 154:757 71. Balcke, G . U., Bennewitz , S. , Bergau , N. , Athmer , B. , Henning , A. , Majovsky , P. , Jiménez - Gómez , J.M. , Hoehenwarter , W. , and Tissier , A - o mics of t omato g landular t richomes r eveals d istinct f eatures of c entral c arbon m etabolism s upporting h igh p roductivity of s pecialized m The Plant Cell 29:960 83. Banerjee, A. , and Sharkey , T.D - p hosphate (MEP) p athway m etabolic r Natural Product Reports 31:1043 55. Barthlott, W . , Wiersch , S. , Z. , and Koch , K t richome t ypes within s pecies of the w ater f ern Salvinia , and o ntogeny of the e ggbeater t Botany 87:830 36. Benfey, P . N. , and Chua , N c auliflower m osaic v irus 35S p c ombinatorial r egulation of t ranscription in p Science 250:959 66. Bergau, N . , Bennewitz , S. , Syr owatka , F. , Hause , G. , and Tissier , A d evelopment of t ype VI g landular t richomes in the c ultivated t omato Solanum lycopersicum and a r elated w ild s pecies S. habrochaites BMC Plant Biology 15:1 15. Berkey, R . , Bendigeri , D. , and Xiao , S p lant d efense/ d isease: t he 81 d c onnection and b Frontiers in Plant Science 3:68. Bleeker, P . M., Mirabella , R. , Diergaarde , P.J. , VanDoorn , A. , Tissier , A. , Kant , M.R. , Prins , M. , De Vos , M. , Haring , M.A. , and Schuurink , R.C h erbivore r esistance in c ultivated t omato with the s esquiterpene b iosynthetic p athway from a w ild r Proceedings of the National Academy of Sciences of the United States of America 109:20124 29. Boughton, A . J., Hoove r , K. , and Felton , G.W j asmonate a pplication i nduces i ncreased d ensities of g landular t richomes on t omato, Lycopersicon esculentum Journal of Chemical Ecology 31:2211 16. Bown, L . , Kusaba , S. , Goubet , F. , Codrai , L. , Dale , A.G. , Zhang , Z. , Yu , X. , Morris , K. , Ishii , T. , Evered , C. , Dupree , P. , and Jackson , S e ctopically p arting c ells 1 - 2 (Epc1 - 2) m utant e xhibits an e xaggerated r esponse to a bscisic a Journal of Experimental Botany 58:1813 23. Brembu, T . , Winge , P. , Seem , M. , and Bones , A.M e ncode s ubunits of a p utative WAVE r egulatory p rotein c omplex i nvolved in p lant c ell m Plant Cell 16:2335 49. Breuer, C . , Kawamura , A. , Ichikawa , T. , Tominaga - Wada , R. , Wada , T. , Kondou , Y. , Muto , S. , Matsui , M. , and Sugimoto , K t rihelix t ranscription f actor GTL1 r egulates p loidy - d ependent c ell g rowth in the Arabidopsis t The Plant Cell 21:2307 22. Brown, D . E., Rashotte , A.M. , Murphy , A.S. , Normanly , J. , Tague , B.W. , Peer , W.A. , Taiz , L. , Muday , G.K. a ct as n egative r egulators of a uxin t ransport in v ivo in Plant Physiology 126:524 35. Buede, R., Rinker - Schaffer , C. , Pinto , W.J. , Lester , R.L. , and Dickson , R.C c haracterization of LCB1, a Saccharomyces g ene r equired for b iosynthesis of the l ong - c hain b ase c omponent of s Journal of Bacteriology 173:4325 32. Çevik, V . , Kidd , B.N. , Zhang , P. , Hill , C. , Kiddle , S. , Denby , K.J. , Holub , E.B. , Cahill , D.M. , Manners , J.M. , Schenk , P.M. , Beynon , J. , and Kazan , K a cts as an i ntegrative h ub for the r egulation of j asmonate - r esponsive g ene e Plant Physiology 160:541 55. Chang, J . , Yu , T. , Yang , Q. , Li , C. , Xiong , C. , Gao , S. , Xie , Q. , Zheng , F. , Li , H. , Tian , Z. , Yang , C. , and Ye , Z e ncoding a s ingle C2H2 z inc - f inger p rotein, r egulates m ulticellular t richome f ormation in t Plant Journal 96:90 102. Chen, G . , Klinkhamer , P.G.L. , Escobar - Bravo , R. , and Leiss , K.A . 201 g landular t richome d ensity and t heir d erived v olatiles a re d ifferently i nduced by j asmonic a cid in d eveloping and f ully d eveloped t omato l eaves: i mplications for t hrips r Plant Science 276:87 98. Chen, M., Han , G. , Dietrich , C.R. , Dunn , T.M. , and Cahoon , E.B e ssential n ature of 82 s phingolipids in p lants as r evealed by the f unctional i dentification and c haracterization of the Arabidopsis LCB1 s ubunit of s erine p The Plant Cell Online 18:3576 93. Cheng, J., Park , T.S. , Fischl , A.S. , and Ye , X.S c ycle p rogression and c ell p olarity r equire s phingolipid b iosynthesis in Aspergillus nidulans Molecular and Cellular Biology 21:6198 6209. Chini, A., Fonseca , S. , Fernández , G. , Adie , B. , Chico , J.M. , Lorenzo , O. , García - Casado , G. , López - Vidriero , I. , Lozano , F.M. , Ponce , M.R. , Micol , J.L. , and Solano , R f amily of r epressors i s the m issing l ink in j asmonate s Nature 448:666 71. Crosara, K . T . B . , Moffa , E.B. , Xiao , Y. , an d Siqueira , W.L i n - s ilico and in v itro s alivary p rotein c omplex p artners u sing the STRING d atabase: a t Journal of Proteomics 171:87 94. m aking p lants b reak a s weat: t he s tructure, f unction, and e volution of p lant s alt g Frontiers in Plant Science 8:1 20. Dayanandan, P. , and Kaufman , P.B Cannabis sativa American Journal of Botany 63:578. Dell, B. , and McComb , A.J h air f o rmation and r esin s ecretion in Eremophila fraseri Protoplasma 92:71 86. Dicke, M . , Gols , R. , Ludeking , D. , and Posthumus , M.A a cid and h erbivory d ifferentially i nduce c arnivore - a ttracting p lant v olatiles in l ima b ean p Journal of Chemical Ecology 25:1907 22. Disch, A ., Hemmerlin , A. , Bach , T.J. , and Rohmer , M. - d erived i sopentenyl d iphosphate i s the b iosynthetic p recursor of u biquinone p renyl s ide c hain in t obacco BY - 2 c el Biochemical Journal 331:615 21. Dolan, W . L., Dilkes , B.P. , Stout , J.M. , Bonawitz , N.D. , and Chapple , C c omplex s ubunits MED2, MED5, MED16, and MED23 g enetically i nteract in the r egulation of p henylpropanoid b Plant Cell 29:3269 85. Du, M . , Zhao , J. , Tzeng , D.T.W. , Liu , Y. , Deng , L. , Yang , T. , Zhai , Q. , Wu , F. , Huang , Z. , Zhou , M. , Wang , Q. , Chen , Q. , Zhong , S. , Li , C.B. , and Li , C o rchestrates a h ierarchical t ranscriptional c ascade t hat r egulates j asmonate - m ediated p lant i mmunity in t Plant Cell 29:1883 1906. Du, W . , Han , X. , Wang , Y. , and Qin , Y p rimary s creening and a pplying of p lant v olatiles as r epellents to c ontrol w hitefly Bemisia tabaci (Gennadius) on t Scientific Reports 6:22140. Duke, S . O. , and Paul , R.N f ine s tructure of the g landular t richomes of Artemisia annua L International Journal of Plant Sciences 154:107 18. 83 El - Assal, S . E . D . , Le , J. , Basu , D. , Mallery , E.L. , and Szymanski , D.B e ncodes an ARPC2 s ubunit of the p utative Arabidopsis ARP2/3 c Plant Journal 38:526 38. Escobar - Bravo, R . , Chen , G. , Grosser , K. , Van Dam , N.M. , Leiss , K.A. , and Klinkhamer , P.G.L . 2019 a r adiation e nhances s alicylic a cid - m ediated d efense s ignaling and r esistance to p seudomonas s yringae DC3000 in a j asmonic a cid - d eficient t omato m Plant Signaling & Behavior 14:e1581560. Escobar - Bravo, R . , Chen , G. , Kim , H.K. , Grosser , K. , Van Dam , N.M. , Leiss , K.A. , and Klinkhamer , P.G.L . 2019 b r adiation e xposure t ime and i ntensity m odulate t omato r esistance to h erbivory through a ctivation of j asmonic a cid s Journal of Experimental Botany 70:315 27. Escobar - Bravo, R . , Ruijgrok , J. , Kim , H.K. , Grosser , K. , Van Dam , N.M. , Klinkhamer , P.G.L. , and Leiss , K.A i ntensity - m ediated i nduction of t richome - a ssociated a llelochemicals i ncreases r esistance a gainst t hrips in t Plant & Cell Physio logy 59:2462 75. España, L . , Heredia - Guerrero , J.A. , Reina - Pinto , J.J. , Fernández - Muñoz , R. , Heredia , A. , and Domínguez , E s ilencing of CHALCONE SYNTHASE during f ruit r ipening m odifies t omato e pidermal c ells and c uticle p Plant Physiology 166:1371 86. Falara, V., Akhtar , T.A. , Nguyen , T.T.H. , Spyropoulou , E.A. , Bleeker , P.M. , Schauvinhold , I. , Matsuba , Y. , Bonini , M.E. , Schilmiller , A.L. , Last , R.L. , Schuurink , R.C. , and Pichersky , E . t omato t erpene s ynthase g ene f Plant Physiology 157:770 89. Falcone, A . , Nelissen , H. , Fleury , D. , Van Lijsebettens , M. , and Bitonti , M.B i nvestigations of the Arabidopsis thaliana E LO 1 m utant g ive n ew i nsights into l eaf l ateral g rowth and e longator f Annals of Botany 100:261 70. Feys, B.J.F., Benedetti , C.E. , Penfold , C.N. , and Turner , J.G m utants s elected for r esistance to the p hytotoxin c oronatine a re m ale s terile, i nsensitive to m ethyl j asmonate, an d r esistant to a b acterial p Plant Cell 6:751 59. Folkers, U., Kirik , V. , Schobinger , U. , Falk , S. , Krishnakumar , S. , Pollock , M.A. , Oppenheimer , G.D. , Day , I. , Reddy , A.R. , Jurgens , G. , and Hulskamp , M c ell m orphogenesis g ene ANGUSTIFOLIA e ncodes a CtBP/BARS - like p rotein and i s i nvolved in the c ontrol of the m icrotubule c EMBO Journal 21:1280 88. Gable, K . , Gupta , S.D. , Han , G. , Niranjanakumari , S. , Harmon , J.M. , and Dunn , T.M d isease - c ausing m utation in the a ctive s ite of s erine p almitoyltransferase c auses c atalytic p Journal of Biological Chemistry 285:22846 52. Garcia - Molina, A . , Altmann , M. , Alkofer , A. , Epple , P.M. , Dangl , J.L. , and Falter - Braun , P . n etwork h ubs i ntegrate a b iotic and b iotic s tress r esponses via i nteraction with the s uperoxide d Journal of Experimental Botany 68:1185 97. 84 Glas, J . J., Schimmel , B.C.J. , Alba , J.M. , Escobar - Bravo , R. , Schuurink , R.C. , and Kant , M.R . g landular t richomes as t argets for b reeding or e ngineering of r esistance to h International Journal of Molecular Sciences 13:17077 103. Goldsbrough, A., Belzile , F. , and Yoder , J.I t omato a nthocyanin without (Aw) m utant u sing the d ihydroflavonol 4 - r eductase g Plant Physiology 105:491 96. Von Groll, U . , Berger , D. , and Altmann , T s ubtilisin - l ike s erine p rotease SDD1 m ediates c ell - to - c ell s ignaling during Arabidopsis s tomatal d The Plant Cell 14:1527 39. Guelette, B . S., Benning , U.F. , and Hoffmann - Benning , S l ipids and l ipid - b inding p roteins in p hloem e xudates from Arabidopsis thaliana Journal of Experimental Botany 63:3603 16. Hategan, L . , Godza , B. , Kozma - Bognar , L. , Bishop , G.J. , and Szekeres , M e xpression of the b rassinosteroid r eceptor - e ncoding BRI1 g Planta 239:989 1001. Hause, B., Stenzel , I. , Miersch , O. , Maucher , H. , Kramell , R. , Ziegler , J. , and Wasternack , C . - s pecific o xylipin s ignature of t omato f lowers: a llene o xide c yclase Is h ighly e xpressed in d istinct f lower o rgans and v ascular b Plant Journal 24:113 26. Hegebarth, D . , Buschhaus , C. , Wu , M. , Bird , D. , and Jetter , R c ompos ition of s urface w ax on t richomes of Arabidopsis thaliana d iffers from w ax on o ther e pidermal c Plant Journal 88:762 74. Hoffmann - Benning, S . , Gage , D.A. , McIntosh , L. , Kende , H. , and Zeevaart , J.A.D . 2002. p eptides in the p hloem s ap of f lowering and n on - f lowering p erilla and l upine p lants u sing m icrobore HPLC f ollowed by m atrix - a ssisted l aser d esorption/ i onization t ime - of - f light m ass s Planta 216:140 47. Holton, T . A., Brugliera , F. , Lester , D.R. , Tanaka , Y. , Hyland , C,D. , Menting , J.G.T. , Lu , C.Y. , Farcy , E. , Stevenson , T.W. , and Cornish , E.C . 1993 a e xpression of c ytochrome P450 g enes c ontrolling f lower c Nature 366:276 79. Holton, T . A., Brugliera , F. , and Tanaka , Y . 1993 b c loning and e xpression of f lavonol s ynthase from p etunia h The Plant Journal 4:1003 10. Howarth, J . R., Parmar , S. , Barraclough , P.B. , and Hawkesford , M.J s ulphur d eficiency - i nduced g ene, S DI 1, i nvolved in the u tilization of s tored s ulphate p ools under s ulphur - l imiting c onditions h as p otential as a d iagnostic i ndicator of s ulphur n utritional s Plant Biotechnology Journal 7:200 209. Howe, G . A., Major , I.T. , and Koo , A.J j asmonate s igna ling for m ultistress r Annual Review of Plant Biology 69:387 415. 85 Huang, X . , Liu , J. , and Dickson , R.C - r egulating s phingolipid s ynthesis i ncreases y east l PLoS Genetics 8 :16 . Humphries, J . A., Vejlupkova , Z. , Luo , A. , Meeley , R.B. , Sylvester , A.W. , Fowler , J.E. , and Smith , L.G a ct with the r eceptor - like p rotein PAN1 to p olarize a symmetric c ell d ivision in m Plant Cell 23:2273 84. Inamdar, J.A. , and Rao , V.S o ntogeny, c lassifi cation, t axonomic s ignificance of t richomes and e n ectaries in c ultivars of c Feddes Repertorium 92:551 56. Ingram, G ., and Nawrath , C r oles of the c uticle in p lant d evelopment: o rgan a dhesions and b Journal of Experimental Botany 68:5307 21. Ishiguro, S., Kawai - Oda , A. , Ueda , J. , Nishida , I. , and Okada , K d efective in a nther d ehiscence1 g ene e ncodes a n ovel p hospholipase A1 c atalyzing the i nitial s tep of j asmonic a cid b iosynthesis, w hich s ynchronizes p ollen m aturation, a nther d ehiscence, and f lower o Plant Cell 13:2191 2209. Jakoby, M . J., Falkenhan , D. , Mader , M.T. , Brininstool , G. , Wischnitzki , E. , Platz , N. , Hudson , A. , Hülskamp , M. , Larkin , J. , and Schnittger , A T ranscriptional p rofiling of m ature A rabidopsis t richomes r eveals t hat NOECK e ncodes the MIXTA - Like t ranscriptional r Plant Physiology 148:1583 1602. . , Stukkens , Y. , Degand , H. , Purnelle , B. , Marchand - Brynaert , J. , and Boutry , M . p lant p lasma m embrane ATP b inding c assette - t ype t ransporter i s i nvolved in a ntifungal t erpenoid s The Plant Cell 13:1095 1107. Jewell, J . B. , and Browse , J. j asmonate p erception i s s ufficient for a ll a spects of j asmonate - m ediated m ale f ertility in a The Plant Journal 85:634 47. Jez, J . M. , and Noel , J.P m echanism of c halcone i somerase: P h d ependence, d iffusion c ontrol, and p roduct b inding d Journal of Biological C hemistry 277:1361 69. Jian, W . , Cao , H. , Yuan , S. , Liu , Y. , Lu , J. , Lu , W. , Li , N. , Wang , J. , Zou , J. , Tang , N. , Xu , C. , Cheng , Y. , Gao , Y. , Xi , W. , Bouzayen , M. , and Li , Z n MYB - t ype t ranscription f actor, p romotes a nthocyanin a ccumulation and e nhances v olatile a roma p roduction in t omato f Horticulture Research 6 :15 . Kagey, M . H., Newman , J.J. , Bilodeau , S. , Zhan , Y. , Orlando , D.A. , van Berkum , N.L. , Ebmeier , C.C. , Goossens , J. , Rahl , P.B. , Levine . S.S. , Taatjes , D.J. , Dekker , J. , and Young , R.A . c ohesin c onnect g ene e xpression and c hromatin a Nature 467:430 35. Kang, J., McRoberts , J. , Shi , F. , Moreno , J.E. , Jones , A.D. , and Howe , G.A f lavonoid b iosynthetic e nzyme c halcone i s omerase m odulates t erpenoid p roduction in g landular t richomes of t Plant Physiol 164:1161 74. 86 Kang, J . , Campos , M.L. , Zemelis - durfee , S. , Al - haddad , J.M. , Jones , A.D. , Telewski , F.W. , Brandizzi , F. , and Howe , G.A c loning of the t omato h airless g ene i mplicates a ctin d ynamics in t richome - m ediated d efense and m echanical p roperties of s tem t Journal of Experimental Botany 67:5313 24. Kang, J . , Liu , G.H. , Shi , F. , Jones , A.D. , Beaudry , R.M. , and Howe , G.A . 2010 a t omato Odo rless - 2 m utant i s d efective in t richome - b ased p roduction of d iverse s pecialized m etabolites and b road - s pectrum r esistance to i nsect h Plant Physiology 154:262 72. Kang, J . , Shi , F. , Jones , A.D. , Marks , M.D. , and Howe , G.A . 2010 b t richome m orphology by the h airless m utation of t omato a ffects l eaf s urface c Journal of Experimental Botany 61:1053 64. m Frontiers in Plant Science 8:1 11. Kirik, V . , Simon , M. , Wes ter , K. , and Schiefelbein , J (ETC2) r eveals r edundancy in the r egion - s pecific c ontrol of t richome d evelopment of Plant Molecular Biology 55:389 98. Kroes, A . , Weldegergis , B.T. , Cappai , F. , Dicke , M. , and van Loon , J.J.A b iosynthesis in a rabidopsis a ttacked by c aterpillars and a phids: e ffects of a phid d ensity on the a ttraction of a c aterpillar p Oecologia 185:699 712. Lange, B.M . , Wildung , M.R. , Stauber , E.J. , Sanchez , C. , Pouch nik , D. , and Croteau , R . 2000. e ssential o il b iosynthesis and s ecretion by f unctional e valuation of e xpressed s equence t ags from m int g landular t Proceedings of the National Academy of Sciences of the United States of America 97:2934 39. Langenheim, J . p lant t erpenoids: a p hytocentric o verview of t heir e cological r Journal of Chemical Ecology 20:1223 80. Lantz, A . T., Solomon , C. , Gog , L. , McClain , A.M. , Weraduwage , S.M. , Cruz , J.A. , and Sharkey , T.D s uppression by CO2 i s n ot d ue to t riose p hosphate u tilization (TPU) l Frontiers in Forests and Global Change 2:1 13. Lawson, S . S., Pijut , P.M. , and Michler , C.H a rabidopsis s tomatal d ensity m utants i ndicat es v ariation in w ater s tress r esponses and p otential e pistatic e Journal of Plant Biology 57:162 73. Leckie, B . , D.A. , Chappell , T.M. , Halitschke , R. , De Jong , E.M. , Kessler , A. , Kennedy , G.G. , and Mutschler , M.A s ynergistic f unctionality of a cylsugars in s uppressing o viposition by i nsect h PLoS ONE 11:1 19. Lee, K ., and Seo , P.J e pigenetic c hanges during p lant r Trends in Plant Science 23:235 47. Le e, Y.S., Choi , K.M. , Choi , M.H. , Ji , S.Y. , Lee , S. , Sin , D.M. , Oh , K.W. , Lee , Y.M. , Hong , 87 J.T. , Yun , Y.P. , and Yoo , H.S p almitoyltransferase i nhibitor m yriocin i nduces g rowth i nhibition of B16F10 m elanoma c ells through G2/M p hase a Cell Proliferation 44:320 29. Li, C . , Williams , M.M. , Loh , Y.T. , Gyu , I.L. , and Howe , G.A c ultivated t omato to c ell c ontent - f eeding h erbivores i s r egulated by the o ctadecanoid - s ignaling p Plant Physiology 130:494 503. Li, L . , Zhao , Y. , Mccaig , B.C. , Wingerd , B.A. , Wang , J. , Whalon , M.E. , Pichersky , E. , and Howe , G.A t omato h omolog of CORONATINE - INSENSITIVE1 i s r equired for the m aternal c ontrol of s eed m aturation , j asmonate - s ignaled d efense r esponses, and g landular t richome d The Plant Cell 16:126 43. Liakoura, V., Stefanou , M. , Manetas , Y. , Cholevas , C. , and Karabourniotis , G d ensity and i ts UV - B p rotective p otential a re a ffected by s hading and l e af p osition on the c Environmental and Experimental Botany 38:223 29. Luckwill, L. 1943. The g enus Lycopersicon: h istorical, b iological, and t axonomic s urvey of the w ild and c ultivated t omatoes . Aberdeen, Scotland: Aberdeen University Press. Lynch, D.V. , and Fairfield , S.R l ong - c hain b ase s ynthesis in p lants ( c haracterization of s erine p almitoyltransferase a ctivity in s quash f ruit m Plant Physiology 103:1421 29. Maffei, M., Gallino , M. , and Sacco , T ndular t richomes and e ssential o ils of d eveloping l eaves in Mentha viridis lavanduliodora Planta Medica 3:187 93. Maloney, G.S., DiNapoli , K.T. , and Muday , G.K a nthocyanin r educed t omato m utant d emonstrates the r ole of f lavonols in t omato l ateral r oot and r oot h air d Plant Physiology 166:614 - U254. Markham, J . E., Li , J. , Cahoon , E.B. , and Jaworski , J.G i dentification of m ajor p lant s phingolipid c lasses from l Journal of Biological Chemistry 281:22684 94. Markham, K.R., Ternai , B. , Stanley , R. , Geiger, H., and Mabry , T.J - 13 NMR s tudies of f lavonoids - III n aturally o ccuring f lavonoid g lycosides and t heir a cylated d Tetrahedron 34:1389 97. Marks, D . d evelopment: t he m aking of a p lant h Current Biology 4:621 23. McDowell, E . T., Kapteyn , J. , Schmidt , A. , Li , C. , Kang , J. , Descour , A. , Shi , F. , Larson , M. , Schilmiller . A.L. , An , L. , Jones , A.D. , Pichersky , E. , Soderlund , C.A. , and Ga ng , D.R . f unctional g enomic a nalysis of S olanum g landular t richome t Plant Physiology 155:524 39. Michaelson, L . V., Napier , J.A. , Molino , D. , and Faure , J.D s phingolipids: t heir i mportance in c ellular o rganization and a Biochimica et Biophysica Acta - 88 Molecular and Cell Biology of Lipids 1861:1329 35. Mo, Y., Nagel , C. , and Taylor , L.P c omplementation of c halcone s ynthase m utants d efines a r ole for f lavonols in f unctional p ol Proceedings of the National Academy of Sciences of the United States of America 89:7213 17. Moghe, G . D., Leong , B.J. , Hurney , S.M. , Jones , A.D. , and Last , R.L r outes to b iochemical i nnovation r evealed by i ntegrative a nalysis of a p lant - d efense r elated s pecialized m etabolic p ELife 6:1 33. Mohan, R . , Vijayan , P. , and Kolattukudy , P.E t issue - s pecific e xpression of a t omato a nionic p eroxidase (Tap1) g ene by a m inimal p romoter, with w ound and p athogen i nduction by an a - f lanking r Plant Molecular Biology 22:475 90. Monaghan, E . , Gable , K. , and Dunn , T s ubunit of s erine p almitoyltransferase e liminate the r equirement for the TSC3 g ene in Saccharomyces cerevisiae Yeast 19:659 70. Msanne, J . , Chen , M. , Luttgeharm , K.D. , Bradley , A.M. , Mays, E .S. , Paper , J.M. , Boyle , D.L. , Cahoon , R.E. , Schrick , K. , and Cahoon , E.B i s c ritical for c el l - t ype d ifferentiation and o rganogenesis, but n ot for c ell v Plant Journal 84:188 201. Muir, S.R., Collins , G.J. , Robinson , S. , Hughes , S. , Bovy , A. , De Vos , C.H.R. , van Tunen , A.J. , and Verhoeyen , M.E p etunia c halcone i somerase in t omato r esults in f ruit c ontaining i ncreased l evels of f Nature Biotechnology 19:470 74. Nadakuduti, S.S., Pollard , M. , Kosma , D.K. , Allen , C. , Ohlrogge , J.B. , and Barry , C.S . 2012. ropic p henotypes of the s ticky p eel m utant p rovide n ew i nsight into the r ole of CUTIN DEFICIENT2 in e pidermal c ell f unction in t Plant Physiology 159:945 60. Nagiec, M. M. , Baltisberger , J.A. , Wells , G.B. , Lester , R.L. , and Dickson , R.C g ene of Saccharomyces and the r elated LCB1 g ene e ncode s ubunits of s erine p almitoyltransferase, the i nitial e nzyme in s phingolipid s Proceedings of the National Academy of Sciences of the United States of America 91:7899 7902. N elissen, H., Fleury , D. , Bruno , L. , Robles , P. , De Veylder , L. , Traas , V.J. , Micol , J.L. , Van Montagu , M. , Inze , D. , and Van Lijsebettens M e longata m utants i dentify a f unctional e longator c omplex in p lants with a r ole in c ell p roliferation dur ing o rgan g Proceedings of the National Academy of Sciences 102:7754 59. Oh, S ., and Montgomery , B.L - s pecific p hytochromes i mpact c hlorophyll l ight - h arvesting c omplexes (LHCs) and n on - p hotochemical q Plant Signaling & Behavior 14:1609857. Oshima, Y . , Shikata , M. , Koyama , T. , Ohtsubo , N. , Mitsuda , N. , and Ohme - Takagi , M . 2013. - Like t ranscription f actors and WAX INDUCER1/SHINE1 c oordinately r egulate 89 c uticle d evelopment in Arabidopsis and Torenia fournie ri The Plant Cell 25:1609 24. Paduch, R . , Kandefer - Szersze , M. , Trytek , M. , and Fiedurek , J s ubstances u seful in h uman h Archivum Immunologiae et Therapiae Experimentalis 55:315 27. Pattanaik, S . , Patra , B. , Singh , S.K. , and Yuan , L o verview of the g ene r egulatory n etwork c ontrolling t richome d evelopment in the m odel p Frontiers in Plant Science 5:1 8. Pattison, R . J., Csukasi , F. , Zheng , Y. , Fei , Z. , van der Knaap , E. , and Catalá , C . 2015. mprehensive t issue - s pecific t ranscriptome a nalysis r eveals d istinct r egulatory p rograms during e arly t omato f ruit d Plant Physiology 168:1684 1701. Payne, W . g lossary of p lant h air t Brittonia 30:239 55. Pieczynski, M . , Marczewski , W. , Hennig , J. , Dolata , J. , Bielewicz , D. , Piontek , P. , Wyrzykowska , A. , Krusiewicz , D. , Strzelczyk - Zyta , D. , Konopka - Postupolska , D. , Krzeslowska , M. , Jarmolowski , A. , and Szweykowska - Kulinska , Z - r egulation of CBP8 0 g ene e xpression as a s trategy to e ngineer a d rought - t olerant p Plant Biotechnology Journal 11:459 69. Poorter, H . , Niinemets , Ü . , Poorter , L. , W right , I.J. , and V illar , R C auses and c onsequences of v ariation in l eaf m ass per a rea (LMA): a m eta - a New Phytologist 182:565 88. Porra, R.J., Thompson , W.A. , and Kriedemann , P.E a ccurate e xtinction c oefficients and s imultaneous e quations for a ssaying c hlorophylls a and b e xtracted with f our d ifferent s olvents: v erification of the c oncentration of c hlorophyll s tandards by a tomic a bsorption s Biochimica et Biophysica Acta - Bioenergetics 975:384 94. Rahman, R . N . Z . R . A . , Zakaria , I.I. , Salleh , A.B. , and Basri , M p roperties and m utational s tudies of c halcone s ynthase from Physcomitrella patens International Journal of Molecular Sciences 13:9673 91. Ren, T. , Weraduwage , S.M. , and Sharkey , T.D e nhancing l eaf p hotosynthetic c apacity by m anipulating m esophyll c ell m Journal of Experimental Botany 70:1153 65. Rick, C . M., Quiros , C.F. , Lange , W.H. , and S tevens , M.A c ontrol of r esistance in the t omato to the t obacco f lea b eetle: p robable r epellance by f oliage v Euphytica 25:521 30. Rodriguez - Concepcion, M . , Yalovsky , S. , Zik , M. , Fromm , H. , and Gruissem , W p renylation s tatus of a n ovel p lant c almodulin d irects p lasma m embrane or n uclear l ocalization of the p Th e EMBO Journal 18:1996 2007. Sakai, W.S. and Sanford , W.G w ater - a bsorbing t richomes of 90 p ineapple ( Ananas comosus , Bromeliaceae Annals of Botany 46:7 11. Sallaud, C . , Rontein , D. , Onillon , S. , Jabes , F. , Duffe , P. , Giacalone , C. , Thoraval , S. , Escoffier , C. , Herbette , G. , Leonhardt , N. , Causse , M. , and Tissier , A n ovel p athway for s esquiterpene b iosynthesis from Z,Z - f arnesyl p yrophosphate in the w ild t omato Solanum habrochaites The Plant Cell 21:301 17. S ánchez - Hernández, C . , López , M.G. , and Délano - Frier , J.P l evels of v olatile e missions in j asmonate - d eficient Spr2 t omato m utants f avour o viposition by i nsect h Plant, Cell and Environment 29:546 57. Sanders, P.R., Winter , J.A. , Barnason , A.R. , Rogers , S.G. , and Fraley , R.T of c auliflower m osaic v irus 35S and n opaline s ynthase p romoters in t ransgenic p Nucleic Acids Research 15:1543 58. Sanders, P . M., Lee , P.Y. , Biesgen , C. , Boone , J.D. , Beals , T.P. , Weiler , E.W. , and Goldberg , R.B g ene e ncodes an e nzyme in the j asmonic a cid s ynthesis p Plant Cell 12:1041 61. Santamaria, M . E . , Cambra , I. , Martinez , M. , Pozancos , C. , González - Melendi , P. , Grbic , V. , Castañera , P. , Ortego , F. , and Diaz , I p yramiding of p eptidase i nhibitors e nhances p lant r esistance to the s pider m ite Tetranychus urticae PLoS ONE 7:2 11. Sasse, J . , Schlegel , M. , Borghi , L. , Ullrich , F. , Lee , M. , Liu , G.W. , Giner , J.L. , K ayser , O. , Bigler , L. , Martinoia , E. , and Kretzschmar , T h ybrida PDR2 i s i nvolved in h erbivore d efense by c ontrolling s teroidal c ontents in t Plant Cell and Environment 39:2725 39. Schilmiller, A. L., Schauvinhold , I. , Larson , M. , Xu , R. , Charbonneau , A.L. , Schmidt , A. , Wilkerson , C. , Last , R.L. , and Pichersky , E. g landular t richomes of t omato a re s ynthesized from a n eryl d iphosphate p recursor r ather than g eranyl d Proceedings of t he National Academy of Sciences of the United States of America 106:10865 70. Schilmiller, A . L., Charbonneau , A.L. , and Last , R.L a cetyltransferase t hat p roduces p rotective a cyl s ugars in t omato t Proceedings of the National Academy of Sciences of the United States of America 109:16377 82. Schilmiller, A . L., Gilgallon , K. , Ghosh , B. , Jones , A.D. , and Last , R.L a cylhydrolases: c arboxylesterase - c atalyzed h ydrolysis of a cylsugars in t omato t Plant Physiology 170:1331 44. Schilmiller, A . L., Last , R.L. , an d Pichersky , E. p lant t richome b iochemistry for the p roduction of u seful c The Plant Journal 54:702 11. Schilmiller, A . L., Miner , D.P. , Larson , M. , Mcdowell , E. , Gang , D.R. , Wilkerson , C. , Last, R.L . b iochemical f t omato t richome d eep e xpressed s equence t ag s equencing and p Plant Physiology 153:1212 23. 91 Schwender, J . , Seemann , M. , Lichtenthaler , H.K. , and Rohme r , M i soprenoids ( c arotenoids, s terols, p renyl s ide - c hains of c hlorophylls and p lastoquinone) via a n ovel p yruvate/ g lyceraldehyde 3 - p hosphate n on - m evalonate p athway in the g reen a lga Scenedesmus obliquus The Biochemical Journal 316:73 80. Shen, Q . , Lu , X. , Yan , T. , Fu , X. , Lv , Z. , Zhang , F. , Pan , Q. , Wang , G. , Sun , X. , and Tang , K . j asmonate - r esponsive AaMYC2 t ranscription f actor p ositively r egulates a rtemisinin b iosynthesis in Artemisia annua New Phytologist 210:1269 81. Shi, L . , Bielawski , J. , Mu , J. , Dong , H. , Teng , C. , Zhang , J. , Yang , X. , Tomishige , N. , Hanada , K. , Hannun , Y.A. , and Zuo , J s phingoid b ases in m ediating r eactive o xygen i ntermediate p roduction and p rogrammed c ell d Cell Research 17100:1030 40. Shpak, E. D., Lakeman , M.B. , and Torii , K.U - n egative r eceptor u ncovers r edundancy in the Arabidopsis ERECTA l eucine - r ich r epeat r eceptor l ike k inase s ignaling p athway t hat r egulates o rgan s The Plant Cell 15:1095 1110. Shpak, E . D., McAbee , J.M. , Pillitteri , L.J. , and Torii , K.U s cience: s tomatal p atterning and d ifferentiation by s ynergistic i nteractions of r eceptor k Science 309:290 93. Simons, K. , and Ikonen , E r afts in c ell m Nature 387:569 72. Snyder, J . C. , and Carter , C.D l eaves of Lycopersicon hirsutum , L. esculentum and t heir h Euphytica 34:53 64. Soubeyrand, E . , Johnson , T.S. , Latimer , S. , Block , A. , Kim , J. , Colquhoun , T.A. , Butelli , E. , Martin , C. , Wilson , M.A. , and Basset , G.J p eroxidative c leavage of k aempferol c ontributes to the b iosynthesis of the b enzenoid m oiety of u biquinone in p The Plant Cell 30:291 0 21. Spitzer - Rimon, B . , Marhevka , E. , Barkai , O. , Marton , I. , Edelbaum , O. , Masci , T. , Prathapani , N.K. , Shklarman , E. , Ovadis , M. , and Vainstein , A g ene e ncoding a f lower - s pecific r egulator of p henylpropanoid v olatiles b iosynthesis in p Plant Cell 22:1961 76. Spribille, R. , and G. Forkmann , G s ynthesis and h ydroxylation of f lavonoids in - p osition with e nzyme p reparations from f lowers of Dianthus caryophyllus L. ( c Planta 155:176 82. S pyropoulou, E . A., Haring , M.A. , and Schuurink , R.C t erpenoids 1, a g landular t richome - s pecific t ranscription f actor from t omato t hat a ctivates the t erpene s ynthase 5 p Plant Molecular Biology 84:345 57. Stacey, D.L., Wyatt , I.J. , and Chambers , R.J e ffect of g lasshouse r ed s pider m ite d amage on the y ield of t Journal of Horticultural Science 60:517 23. 92 Sugimoto - Shirasu, K ., and Roberts , K i e ndoreduplication and c ell - s ize c ontrol in p Current Opinion in Plant Biology 6:544 53. Suzuki, T . , Honda , Y. , and Mukasa , Y - B r adiation, c old and d esiccation s tress on r utin c oncentration and r utin g lucosidase a ctivity in t artary b uckwheat ( Fagopyrum tataricum ) l Plant Science 168:1303 7. Takenawa, T ., and Suetsugu , S - WAVE p rotein n etwork: c onnecting the m embrane to the c Nature Reviews Molecular Cell Biology 8:37 48. Tan, H . , Xiao , L. , Gao , S. , Li , Q. , Chen , J. , Xi ao , Y. , Ji , Q. , Chen , R. , Chen , W. , and Zhang , L . i s r equired for t richome d evelopment and a rtemisinin b iosynthesis in Artemisia annua Molecular Plant 8:1396 1411. Tarkowská, D ., and Strnad , M . 2018. - d erived p lant s ignaling m olecules: b iosynthesis and b iological i Planta 247:1051 66. t omato g enome s equence p rovides i nsights into f leshy f ruit e Nature 485:635 41. Thines, B . , Katsi r , L. , Melotto , M. , Niu , Y. , Mandaokar , A. , Liu , G. , Nomura , K. , He , S.Y. , Howe , G.A. , and Browse , J r epressor p roteins a re t argets of the SCFCOI1 c omplex during j asmonate s Nature 448:661 65. s pecific e xpression: p romoters and t heir a InTech 17:357 378 . Turlings, A . T.C.J., Tumlinson , J.H. , an d Lewis , W.J h erbivore - i nduced p lant o dors by h ost - s eeking p arasitic w Science 250:1251 53. Vendemiatti, E . , Zsögön , A. , e. Silva , G.F.F. , de Jesus , F.A. , Cutri , L. , Figueiredo , C.R.F. , Tanaka , F.A.O. , Nogueira , F.T.S. , and Peres L.E.P t ype - IV g landular t richomes i s a h eterochronic t rait in t omato and c an b e r everted by p romoting j Plant Science 259:35 47. Voirin, B . , Bayet , C. , and Colson , M t hat f lavone a glycones a ccumulate in the p eltate g lands of Mentha x Piperita l Phytochemistry 34:85 87. Vosman, B . , Westende , W.P.C. , Henken , B. , van Eekelen , H.D.L.M. , de Vos , R.C.H. , and Voorrips , R.E s pectrum i nsect r esistance and m etabolites in c lose r elatives of the c ultivated t Euphytica 214 :14 . Walsh, D . B., Zalom , F.G. , Shaw , D.V. , and Larson , K.D r eduction c aused by t wospotted s pider m ite f eeding in an a dvanced - c ycle s trawberry b reeding p Journal of the American Society for Horticultural Science 127:230 37. Wang, M ., and Maeda , H.A a mino a cid a minotransferases in p 93 Phytochemistry Reviews 17:131 59. Wang, Q . , Reddy , V.A. , Panicker , D. , Mao , H.Z. , Kumar , N. , Rajan , C. , Venkatesh , P.N. , Chua , N.H. , and Sarojam , R e ngineering of t erpene b iosynthesis in p lants u sing a t richome - s pecific t ranscription f actor MsYABBY5 from s pearmint ( Mentha spicata Plant Biotechnology Journal 1619 32. Wang, Z . , Yang , Z. , and Li F pdates on m olecular m echanisms in the d evelopment of b ranched t richome in Arabidopsis and n onbranched in c Plant Biotechnology Journal 1706 22. Waters, Mark T., Elizabeth C. Moylan, and Langdale , J.A t ranscription f actors r egulate c hloro plast d evelopment in a c ell - a utonomous m Plant Journal 56:432 44. d iversity and d Advances in Botanical Research 31:35. Von Wettstein - a ffecting a nthocyanin s ynthesis in the t omato I. g enetics, h Hereditas 60:317 46. Wilson, L.T., Trichilo , P.J. , and Gonzalez , D m ite (Acari: Tetranychidae) i nfestation r ate and i nitiation: e ffect on c otton y Journal of Economic Entomology 84:593 600. Wu, G . , Park , M.Y. , Conway , S.R. , Wang , J.W. , Weigel , D. , and Poethig , R.S s equential a ction of MiR156 and MiR172 r egulates d evelopmental t Cell 138 : 750 59. Xu, J . , Van Herwijnen , Z.O. , Dräger , D.B. , Sui , C. , Haring , M.A. , and Schuurink , R.C . 2018. r egulates t ype VI g landular t richome f ormation and t erpene b iosynthesis in t omato g landular c Plant Cell 30:2988 3005. Yan, T . , Li , L. , Xie , L. , Chen , M. , Shen , Q. , Pan , Q. , Fu , X. , Shi , P. , Tang , Y. , Huang , H. , Huang , Y. , Huang , Y. , and Tang , K n ovel HD - ZIP IV/MIXTA c omplex p romotes g landular t richome i nitiation and c uticle d evelopment in Artemisia annua New Phytologist 218:567 78. Yang, C . , Li , H. , Zhang , J. , Luo , Z. , Gong , P. , Zhang , C. , and Li , J r egulatory g ene i nduces t richome f ormation and e mbryo l ethality in t Proceedings of the National Academy of Sciences of the United States of America 108:11836 41. Yazaki, K . , Arimura , G.I. , and Ohnishi , T t erpeno ids in p lants: t heir b iosynthesis, l ocalization and e cological r Plant and Cell Physiology 58:1615 21. Yi, Hojae, Yintong Chen, James Z. Wang, Puri , V.M. , and Anderson , C.T s tomatal f lexoskeleton: h ow the b iomechanics of g uard c ell w alls a nimate an e lastic p ressure v Journal of Experimental Botany 70:3561 72. Young, P. A. , and MacArthur , J.W c haracters of t Texas 94 Agricultural Experiment Station 698:61. Yu, G ., and Pichersky , E, s e xpression of Methylketone Synthase1 and Methylketone Synthase2 l eads to p roduction of m ethylketones and m yristic a cid in t ransgenic p Plant Physiology 164:612 22. Zhai, Q . , Zhang, X., Wu, F., Feng, H., Deng, L., Xu , L. , Zhang, M., Wang, Q., and Li , C . 2015. m echanism of j asmonate r eceptor COI1 - m ediated d elay of f lowering t ime The Plant Cell 27:1 16. Zhang, B., Chen, W., Foley, R.C., Buttner , M. , and Singh , K.B d istinct t ypes of D NA b inding p roteins e nhance b inding to O CS e lement p romoter s The Plant Cell 7:2241 52. Zhao, H . , Wang, X., Zhu, D., Cui, S., Li, X., Cao, Y., and Ma , L s ingle a mino a cid s ubstitution in IIIf s ubfamily of b asic h elix - l oop - h elix t ranscription f actor AtMYC1 l eads to t richome and r oot h air p atterning d efects by a bolishing i ts i nteraction with p artner p roteins Journal of Biological Chemistry 287:14109 21.