Citation
Manipulation of the Flavanoid Pathway in Citrus

Material Information

Title:
Manipulation of the Flavanoid Pathway in Citrus
Creator:
KOCA, UFUK ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Chalconoids ( jstor )
Enzymes ( jstor )
Flavanones ( jstor )
Flavonoids ( jstor )
Grapefruits ( jstor )
Plasmids ( jstor )
Polymerase chain reaction ( jstor )
RNA ( jstor )
Species ( jstor )
Transgenic plants ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Ufuk Koca. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
12/31/2005
Resource Identifier:
436098590 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

MANIPULATION OF THE FLAVONOID PATHWAY IN CITRUS By UFUK KOCA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 By Ufuk Koca

PAGE 3

This dissertation is dedicated to my parent s Mustafa and Gnl, my brother Tan and my sister Gne along with the people who supported me in various ways during my journey, and Deniz who brings out the best in me.

PAGE 4

ACKNOWLEDGMENTS I would like to express my deepest thanks to my advisor for being my mentor; she is my inspiration as a human being. Her work style, encouragement and generous attitude gave me the opportunity to learn various research techniques and to improve myself as an independent scientist. I would like to thank my committee members, Drs. C. Guy, K. Cline, B. Rathinasabapathi and F. Gmitter, for their time and helpful discussions. I am very appreciative of having Dr. Ricardo Harakava in my life as a friend and a mentor. His courage and support in various ways made me stand by my aspirations when the obstacle of research and life blocked my way. I am very grateful to my parents for their never-ending confidence in me. I would like to thank Dr. Karen Champ, Dr. Vicente Febres and Dr. Eduardo Carlos, for their helpful discussions and friendship that made the lab an exciting place to come. I thank Ms. Kim Niblett for her friendship and her help in the lab, which made me conduct my research more efficiently. I would also like to thank past and present members of the Moore lab and PMCB who in one way or another helped me to improve my work and myself as a better scientist. Finally I would like to show my gratitude to the Minister of Education of Turkish Republic for providing the scholarship for me to come to the United States to have my education in the best scientific environment. iv

PAGE 5

TABLE OF CONTENTS ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT.........................................................................................................................x CHAPTER 1 INTRODUCTION........................................................................................................1 2 LITERATURE REVIEW.............................................................................................4 Flavonoid Biosynthesis.................................................................................................4 Flavonoids in the Plant Kingdom..........................................................................4 Flavonoids in Citrus..............................................................................................7 Nonmolecular Studies That Attempted to Alter The Naringin Level in Citrus..11 Manipulation of the Flavonoid Pathway by Silencing the Flavonoid Genes via Antisense Suppression, Sense Suppression, and/or Co-suppression...............13 Difficulties in Citrus Breeding....................................................................................18 Citrus Transformation.........................................................................................19 Manipulation of the Flavonoid Pathway in Citrus..............................................25 3 GENETIC TRANSFORMATION OF CITRUS WITH FLAVONOID PATHWAY GENES.......................................................................................................................27 Introduction.................................................................................................................27 Materials and Methods...............................................................................................30 Agrobacterium strains and vector plasmids........................................................30 Plant material, transformation and regeneration.................................................33 Histochemical GUS assay...................................................................................35 Results and Discussion...............................................................................................41 4 CHARACTERIZATION OF TRANSGENIC PLANTS...........................................50 Introduction.................................................................................................................50 Materials and Methods...............................................................................................51 Southern Blot Analysis........................................................................................51 v

PAGE 6

Relative Quantitative Reverse Transcription-Polymerase Chain Reaction.........54 HPLC Analysis....................................................................................................58 Results and Discussion...............................................................................................59 Southern Blot Analyses.......................................................................................59 HPLC Analysis....................................................................................................70 5 CONCLUSIONS........................................................................................................82 APPENDIX: MOLECULAR AND CHROMATOGRAPHIC ANALYSIS OF TRANSGENIC PLANTS...........................................................................................88 LIST OF REFERENCES...................................................................................................91 BIOGRAPHICAL SKETCH.............................................................................................99 vi

PAGE 7

LIST OF TABLES Table page 2-1. The major types of flavanone glycoside(s) in different citrus species........................8 3-1. Results of Agrobacterium-mediated transformation of C. paradisi with the chalcone synthase CHS-S construct, followed by GUS histochemical assay........................45 3-2. Results of Agrobacterium-mediated transformation of C. paradisi with the chalcone synthase CHS-AS construct, followed by GUS histochemical assay.....................46 3-3. Results of Agrobacterium-mediated transformation of C. paradisi with the chalcone isomerase CHI-S construct, followed by GUS histochemical assay.......................47 3-4. Results of Agrobacterium-mediated transformation of C. paradisi with the chalcone isomerase CHI-AS construct, followed by GUS histochemical assay....................47 3-5. Results of Agrobacterium-mediated transformation of C. paradisi with the 1,2 rhamnosyl transferase hairpin (1,2RT-HP) construct, followed by PCR analysis..48 3-6. Results of Agrobacterium-mediated transformation of C. paradisi with the pME 1,2 rhamnosyl transferase (1,2RT-S) construct, followed by PCR analysis................48 3-7. Results of Agrobacterium-mediated transformation of C. paradisi with the 1,2 rhamnosyl transferase pBINplus sense construct followed by PCR analysis.........48 4-1. Southern blot analysis of CHS and CHI transgenic plants.......................................64 4-2. Southern blot analysis of 1,2 RT transgenic plants..................................................66 vii

PAGE 8

LIST OF FIGURES Figure page 2-1. Basic flavonoid chemical structures..........................................................................5 2-2. General flavonoid biosynthesis pathway...................................................................6 2-3. The basic chemical structures of citrus major flavanone and glycosides.................9 2-5. A model for RNAi, modified from Finnegan and Matzke (2003)..........................19 3-2. Two 1,2 rhamnosyl transferase transformation vectors..........................................38 3-3. The 1,2RT-sense (1,2RT-S) transformation vector.................................................39 3-5. PCR analysis of 1,2RT transgenic plants................................................................44 3-6. Morphological comparison of 1,2 RT plants to control plants...............................49 4-1. Estimation of the exponential phase for CHI transgenic plants..............................57 4-2. Estimation of 18S ribosomal RNA : Competimers ratio for CHI gene...............58 4-3. Southern blot analyses of CHS transgenic plants....................................................61 4-4. Southern blot analyses of CHS transgenic plants....................................................62 4-5. Southern blot analyses of CHI transgenic plants....................................................63 4-6. Southern blot analyses of 1,2 RT transgenic plants................................................67 4-7. Expression analyses of CHS plants.........................................................................71 4-8. Expression analyses of CHI plants..........................................................................72 4-9. Expression analysis of 1,2 RT plants......................................................................73 4-10. Naringin levels (mg/g) of control plants analyzed by HPLC..................................74 4-11. Naringin levels of CHS transgenic plants analyzed by HPLC................................75 4-12. Naringin levels of CHI transgenic plants analyzed by HPLC.................................76 viii

PAGE 9

4-14. Naringin levels of CHS-AS transgenic plants and controls....................................77 4-15. Naringin levels of CHI-S transgenic plants and controls........................................77 4-16. Naringin levels of CHI-AS transgenic plants and controls.....................................78 4-17. Naringin levels of 1,2 RT transgenic plants analyzed by HPLC.............................78 4-18. Demonstration of the transgenic plant 1,2RT 17....................................................80 4-19. Flavonoid levels of transgenic and control pants analyzed by HPLC.....................81 ix

PAGE 10

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MANIPULATION OF THE FLAVONOID PATHWAY IN CITRUS By Ufuk Koca December 2004 Chair: Gloria A. Moore Major Department: Plant Molecular and Cellular Biology Flavonoids are widely distributed diverse groups of plant secondary metabolites with important functions in protection against UV light damage, pigment production, pollen growth, male fertility, and plant-microbe signaling. Moreover, they contribute dietary health benefits to humans due to their antioxidant, anticancer, anti-inflammatory, and antithrombogenic properties. In citrus, flavanone glycosides are the main group of flavonoids that accumulate in entire plants; roots, leaves and fruits. However, their physiological role(s) in citrus plants are not known. The presence of some of these compounds affects the taste of citrus fruits, rather than influencing their color. Flavanone rutinosides are tasteless, whereas flavanone neohesperidosides, for instance naringin, give a bitter taste to fruit and fruit juice products. Since consumers’ food choice is driven by taste, the bitterness of naringin in citrus reduces the acceptability of fresh fruit and juice products in commercial markets and affects the consumers’ having benefit of them. The main objective of this research is to manipulate the citrus flavanoid biosynthetic pathway in order to reduce bitter taste by altering the x

PAGE 11

production of flavanone neohesperidosides. Molecular genetic and transformation techniques were applied to achieve this goal. In order to suppress the expression of the target gene, sense and antisense constructs containing cDNAs of chalcone synthase (CHS) and chalcone isomerase (CHI) genes, whose products catalyze the first two steps in the biosynthetic pathway, were exploited in grapefruit transformation experiments. The 1,2 rhamnosyl transferase (1,2 RT) gene, which catalyzes the last biochemical step in the formation of naringin, was also utilized for the genetic transformation of C. paradisi. A sense construct was used for overexpression and a hairpin-forming construct of this gene was employed to express dsRNA, which interferes with target gene expression. Transformation efficiency varied depending on gene transformation construct. All of the resulting transgenic plants were analyzed by HPLC for their flavonoid composition. A few transgenic plants showed lower naringin levels than the control. According to southern blot analysis most of the transgenic plants had one to three transgene copy numbers. Some of the plants that were significantly different based on statistical analyzes, were selected to analyze for their target gene transcript level. However, no direct correlation was observed between target gene transcript and naringin level. Our initial studies are with the grapefruit; however, the approach could be applied to other citrus types that have bitter tasting fruit. xi

PAGE 12

CHAPTER 1 INTRODUCTION Flavonoids are the most studied polyphenolics. The early steps of the pathway are highly similar, and the sequences of the genes are highly conserved among plants. The genes that function early in the pathway have been cloned and exploited to alter the products of the pathway in many different plant species (Elomaa et al. 1994; Koes et al. 1994; Winkel-Shirley 2001b; Forkman et al. 2001). The first step of the pathway is catalyzed by chalcone synthase (CHS), which utilizes three molecules of malonyl CoA and one molecule of p-coumaryl CoA to produce naringenin chalcone. The second step is the formation of flavanone, which occurs either spontaneously or with the help of chalcone isomerase (CHI). From this point on, the genes that predominantly function in the pathway diverge depending upon the kinds of flavonoids that accumulate in plants. Flavanones are the most abundant flavonoids in citrus (Horowitz and Gentili 1977). They tend to occur in glycoside form, in which a disaccharide is attached to the aglycone through the C-7 hydroxyl group. The disaccharide can be -rutinose, which is tasteless, or -neohesperidose, which is bitter (Horowitz and Gentili 1977). Naringin, one of the flavanone neohesperidosides, is a highly abundant bitter compound in grapefruit (Bar-Peled 1991). Although some “tartness” is a desirable specific character of grapefruit taste, the massive amount of bitterness not only effect consumers having benefits from grapefruit juice and juice products, but also decreases the commercial value of them (Drewnowski 1997; 2000). Our goal is to manipulate the flavonoid biosynthetic pathway to alter the levels and types of flavanones in citrus fruits to improve their palatability 1

PAGE 13

2 and/or pharmacological properties. The state of Florida is the main grapefruit producer in the US and world. A total of 48.5 million 85 lb box units of commercial grapefruit were produced in the US during the 2003-2004 seasons (Florida Department of Citrus http://www.fred.ifas.ufl.edu/citrus/pubs/ref/CRB2004.pdf ). A total of 38.7 million boxes of grapefruit were harvested from Florida alone. A total of 14.1 million boxes of Floridian grapefruit were utilized as fresh, 16.0 million boxes were utilized as frozen concentrated juice, 6.2 million boxes were utilized as chilled grapefruit juice, and 2.4 million boxes were recorded as non-certified and other. Manipulation of the flavanone glycosides might decrease the bitter taste either by a reduction in the concentration of bitter compounds or by increasing the concentration of tasteless compounds that can block perception of the bitter ones. Previously CHS and CHI cDNAs were isolated from Marsh grapefruit in our laboratory (Luth 1997). Sense and antisense constructs containing cDNAs of CHS and CHI genes were exploited in grapefruit transformation experiments in order to suppress the expression of the target gene. The number of experiments and produced plants were not sufficient, although use of the CHI antisense construct appeared to cause of reduction in naringin level (Luth 1997). To achieve our main goal we produced more transgenic plants by utilizing the same constructs in this project. Moreover, a sense and a hairpin forming construct of the 1,2 rhamnosyl transferase (1,2 RT) gene, which catalyses the last biochemical step in the formation of naringin, were exploited for the genetic transformation. The hairpin-forming construct expresses dsRNA, which should interfere with the target gene activity. Regenerated plants were analyzed for evidence of transformation by GUS histochemical assay and PCR techniques. Transgenic plants were furthermore characterized for their

PAGE 14

3 transgene copy numbers and steady state RNA levels in order to estimate the degree of gene transcription and were analyzed by HPLC for their naringin level and flavonoid composition.

PAGE 15

CHAPTER 2 LITERATURE REVIEW Flavonoid Biosynthesis Flavonoids in the Plant Kingdom Flavonoids are broadly distributed, low-molecular-weight phenolics in plants. More than 4000 flavonoid compounds have been isolated from different plant species (Winkel-Shirley 2001a). They are distinguished by their C6-C3-C6 carbon skeleton structure (Figure 2-1) (Horowitz and Gentili 1977). The oxidation level and structural features of the C3 determine the specific type of flavonoid, while the C6 chains are aromatic rings usually carrying hydroxyl, methoxyl and/or other constituents. There are seven main types of flavonoids (Figure2-1): chalcones, flavones, flavonols, flavanones, anthocyanidins, condensed tannins (proanthocyanidins), and aurones (Winkel-Shirley 2001a). The seventh group, namely aurones is not widely distributed among plant species. Specialized forms of flavonoids called isoflavonoids are frequently considered as a type of flavonoids too. Flavonoids have important roles in the adaptation of plants to their environments. They have been found to be developmentally regulated, tissue specific and responsive to light and stress conditions (Winkel-Shirley 2001b). Numerous flavonoids have function in variation of physiological processes in plants including in pollen germination in petunia (Mo et al. 1992), negative regulators of auxin transport in Arabidopsis (Brown et al. 2001), amelioration of heat stress in Ipomea (Caberly and Rausher 2003). Flavonoids not only function in plant growth and development but they also have effects on human health. A number of flavonoids showed antiproliferative 4

PAGE 16

5 activities in various human cancer lines (Manthey and Guthrie 2002), flavonoid aglycone hesperetin demonstrates antiallergic activity by blocking the histamine receptors (Lee et al. 2004) and they protect against cardiovascular diseases via their antioxidant activities (Ness et al. 1997). C6-C3-C6 carbon skeleton Chalcone Flavone Flavonol Flavanone Anthocyanidin Proanthocyanidin Aurone isoflavone 7 6 5 8 2 4 3 1 Figure 2-1. Basic flavonoid chemical structures. Flavonoid biosynthesis is one of the most widely studied biochemical pathway in plants. The genes of the enzymes that function early in the pathway have been cloned from numerous plant species (Elomaa and Hohan 1994; Koes et al. 1994; Winkel-Shirley 2001b; Forkman and Martens 2001). The early steps of the pathway are highly comparable among plant species, and the sequences of the genes are highly conserved.

PAGE 17

6 The first step of the pathway is catalyzed by chalcone synthase (CHS) (Figure 2-2), which utilizes three molecules of malonyl CoA and one molecule of p-coumaryl CoA to produce naringenin chalcone. The second step is the formation of flavanone which occurs either spontaneously or with the help of chalcone isomerase (CHI).Naringenin flavanone is the most common precursor of flavonoid compounds in plant species. From this point on, the genes that predominantly function in the pathway diverge depending upon the kinds of flavonoids that accumulate in specific plants. For example, based on the species, flavanone might be converted to dihydroflavonols by flavanone 3-hydroxylase (F3H). 4-Coumaryl-CoA Malonyl-CoA CHS Figure 2-2. General flavonoid biosynthesis pathway. The main types of flavonoids are placed in boxes. Chalcones CHI Flavanones Isoflavones Dihydroflavonols Leucoanthocyanidins Anthocyanidins DFR DFR FLS Aurones F3H Flavonols Condensed Tannins Flavanols

PAGE 18

7 Alternatively, flavanone can be converted to flavanols by dihydroflavonol 4-reductase (DFR). In most species, dihydroflavonols are used as substrates for a flavonol synthase (FLS) enzyme to produce flavonols or are reduced to leucoanthocyanidins that are the precursors for colored anthocyanidins. Flavonoid compounds are usually stabilized via glycosylation by UDP-glucose glucosyl transferases (UFGT), which add glucose units to a variety of positions (e.g., positions 3, 5, and sometimes 7 Figure 2-3-a). Methylation, acylation, and rhamnosylation can also appear at various positions in the molecule. Flavonoids in Citrus There are three main types of flavonoids in citrus: flavanones, flavones, and anthocyanidins (Horowitz and Gentili 1977). Flavanones are the most abundant flavonoids in citrus. They tend to occur in glycoside form, in which a disaccharide is attached to the aglycone through the C-7 hydroxyl group (Figure 2-3). The disaccharide can be -neohesperidose, which is composed of L-rhamnose linked to position 2 of D-glucose (Figure 2-3-b). Alternatively, the disaccharide can be -rutinose, where L-rhamnose is linked to position 6 of the D-glucose (Figure 2-3-c). -neohesperidosides such as naringin (Figure 2-3-d) have a bitter taste, whereas -rutinosides, such as hesperidin (Figure 2-3-e), are tasteless (Horowitz and Gentili 1977). Flavonoid biosynthesis in citrus species is very similar to that in other plant species. As in other species, CHS catalyzes the formation of naringenin chalcone, which is converted to naringenin flavanone by CHI. This step turns out to be mostly spontaneous, when the fruit becomes mature. These two steps are believed to be followed by the addition of hydroxyl and/or methyl groups to produce the different flavanones shown in Figure 2-4. The last steps involve glucosylation and rhamnosylation. UDP

PAGE 19

8 flavonoid glucosyl transferases (UFGT) add a glucosyl group to position 7 of the flavanones. Then, a rhamnosyl transferase (RT) adds a rhamnosyl group to either position 6 of the glucose, producing a rutinoside (tasteless), or the position 2, producing a neohesperidoside (bitter). The 1,2 rhamnosyl transferase enzyme that attaches the rhamnose unit to the C-2 OH group of D-glucose has been isolated from pummelo leaves (Bar-Peled et al. 1991). This enzyme cannot form the 1-6 linkage that produces rutinosides. Therefore, a different enzyme adds a rhamnose group specifically to position 6 and is responsible for the production of tasteless flavonoid compounds. Some citrus species contain only flavanone-neohesperidosides, while other types contain only rutinosides (Horowitz and Gentili 1977; Rouseff et al. 1987; Kawaii et al. 1999). Finally, there are citrus types, mostly hybrids, which contain both neohesperidosides and rutinosides. An example is grapefruit, which is a hybrid of sweet orange and pummelo. Table 2-1. The major types of flavanone glycoside(s) in different citrus species. Citrus Species Flavanone neohesperidoside Flavanone rutinoside Sour Orange (C. aurantium) naringin Pummelo (C. grandis) naringin Sweet orange (C. sinensis) hesperidin Mandarin (C. reticulata) hesperidin Lemon (C. limon) hesperidin Grapefruit (C. paradisi) naringin hesperidin

PAGE 20

9 R1X-O OH O O R2 8 7 6 5 4 2 3 a) Flavanon e OH OH OH CH2 O O OH OH O OH CH3 OH O OH CH2OH OH O O OH OH OH CH3 OH OH OH OH CH2 O O OH OH O OH CH3 OH O OH CH2OH OH O O OH OH OH CH3 OH b ) -neohes p eridose c ) rutinose OH O OH O O O CH2OH OH O O OH OH OH CH3 OH d) Naringin OH OCH3O OH O O OH OH CH2 O O OH OH O OH CH3 OH e) Hesperidin Figure 2-3. The basic chemical structures of citrus major flavanone and glycosides. A) flavanone, B) -neohesperidose (2-O--L-rhamnosyl--D-glucoside) (Horowitz and Gentili 1977), C) -rutinose (6-O--L-rhamnosyl--D-glucoside), D) naringin and E) hesperidin.

PAGE 21

10 Although grapefruit accumulates both bitter and tasteless compounds, the bitter-tasting naringin is the most abundant flavanone glycoside (Johnson 1988). Radioimmunoassay studies on the quantitative distribution of naringin in C. paradisi Macf. cv Duncan showed that seeds, seedlings, young plants, branches, flowers and fruit contain naringin (Jourdan et al. 1985). Young shoots and coats of seeds have the highest levels of naringin, whereas roots and cotyledons contain lower amounts of naringin. In flowers, the ovary contains the highest level of naringin, as much as 11% of the fresh weight. It has been observed that the younger the tissue, the higher the concentration of naringin, which decreases as the tissue becomes older. Up to 10% of the fresh weight of young leaves of grapefruit seedlings is composed of naringin. The level of naringin in very young-immature fruits (~1cm) may reach 75% of their dry weight (Bar-Peled 1991). In a mature fruit, the naringin level is higher in the peel than the juice sacs (Jourdan et al. 1985). Although the naringin level decreases as fruits mature, mostly due to a dilution effect, the bitterness remains significant. Horowitz and Gentili (1977) studied the relationship between flavonoid chemical structures and taste. They discovered that naringin and poncirin are the most bitter compounds in citrus, followed by neohesperidin and neoeriocitrin (Figure 2-4). Naringin chalcone and especially naringin dihydrochalcone compounds were found to be very sweet; thus they can be candidates for sweeteners (Horowitz and Gentili 1977). Interestingly, naringin dihydrochalcone can be obtained from the catalytic hydrogenation of naringin in an alkaline solution. Therefore, a transgenic citrus plant that produces a high level of naringin could be used for large-scale extraction of naringin to produce naringin dihydrochalcone sweeteners commercially.

PAGE 22

11 Flavonoids that are neither bitter nor sweet can contribute to taste as well (Horowitz and Gentili 1977). When the tasteless flavone rhoifolin (an analog of naringin) is present, it increases the threshold concentration at which the bitter taste of naringin is detected. Rhoifolin may partially block the response to the bitterness of the naringin by competing for sites on the taste receptors. Accordingly, if the production of tasteless compounds is increased in transgenic plants, they could mask the bitter taste of the fruit. Enzyme activity assays and radiolabeling studies revealed that flavanone glycosides (FGs) are synthesized during cell division in growing leaves and fruits (Berhow and Vandercook 1989; Bar-Peled et al. 1993; Castillo et al. 1992; Moriguchi et al. 2001). Although expression of flavonoid pathway genes and flavonoid content slightly increases, they appear to decrease as the fruit and leaves become mature, since during maturation their size expands and thus flavonoid level gets diluted. The occurrence of this in fruit and leaves is correlated. Therefore leaves can show the relative indication of the concentration of these compounds in fruits. Since it takes 5-10 years to obtain the fruit from a grapefruit plant, characterization of transgenic lines with altered flavonoid content can be anticipated by analyzing their leaves. Nonmolecular Studies That Attempted to Alter The Naringin Level in Citrus There are studies on reducing the naringin level in commercial citrus juice products. Naringin levels can be reduced by several technologies such as applying gibberellic acid via lanolin or spraying onto immature fruit (Berhow and Vandercook 1992; Berhow 2000), adsorptive debittering of juice, chemical methods, treatment with divinyl benzene styrene (DVB) resins, and by adding the naringinase enzyme (Puri et al. 1996; Puri and Banerjee 2000). However, these methods are subject to limitations. In particular, chemical methods are nonspecific and inefficient; furthermore they may

PAGE 23

12 Figure 2-4. Putative flavonoid biosynthetic pathway in citrus (slightly modified from Rouseff, CREC, University of Florida). Pictures are the fruits, which produce corresponding compounds excessively. Grapefruit Orange Grapefruit Grapefruit Poncirus

PAGE 24

13 alter the composition of the original juice by removing some of the nutrients and the desirable flavoring components. Moreover, they may introduce batch to batch variations due to non-monitorable changes. Although the debittering of the fruit juice by naringinase enzyme is attractive, this method has some disadvantages as well. The enzyme is very costly to procure in sufficient amounts for large-scale citrus juice processes. In addition, there is no available naringinase preparation that completely hydrolyzes naringin. Manipulation of the Flavonoid Pathway by Silencing the Flavonoid Genes via Antisense Suppression, Sense Suppression, and/or Co-suppression Gene expression can be modified by silencing a gene, via transformation with an antisense or a (double-strand) dsRNA-producing construct. This results in low expression and sometimes no expression at all. Alternatively, one can attempt to increase the expression level via transformation with target genes as sense constructs. Surprisingly, however, the introduction of sense constructs may cause silencing of both the transgene and the endogenous target gene. This was first observed by the Jorgensen group (Napoli et al. 1990) and termed co-suppression. A chalcone synthase gene was introduced into petunia under the control of a strong promoter to intensify the purple flower color. However, many of the transgenic flowers appeared variegated or even white. In other studies, antisense suppression was used to manipulate the flavonoid pathway (Van der Krol et al. 1988; Colliver et al. 1997; El Euch et al. 1998; Forkman et al. 2001). In this technique, a DNA sequence homologous to the target sequence is introduced into plant cells in antisense, or inverse, orientation. In early studies, it was suggested that the transfer of the antisense DNA sequence caused the production of RNA molecules that were complementary to the target mRNA sequence. The antisense RNA

PAGE 25

14 would then form a duplex with the endogenous RNA, preventing its translation and thus the expression of the target gene. The cellular causes of gene silencing are now being elucidated, however, it is thought that transgene-induced silencing in plants and animals may occur through two different mechanisms: at the transcriptional level (transcriptional gene silencing or TGS), which involves gene-specific methylation, or at the post-transcriptional level (post-transcriptional gene silencing, or PTGS), which involves rapid RNA degradation in the cytoplasm and prevention of the accumulation of mRNA. Recent studies have demonstrated that constructs producing self-complementary hairpin RNA (hpRNA) or double stranded RNA (dsRNA) silence the target genes in plants very efficiently (Waterhouse 1998 et al.; Chuang 2000 et al.; Smith et al. 2000; Wesley et al. 2001; Stoutjesdijk et al. 2002). This phenomenon was first observed in Caenorhabditis elegans and was termed RNA interference (RNAi) (Fire and.Mello 1998). RNAi has since been observed in other organisms; co-suppression or post-transcriptional gene silencing (PTGS) is a specific type of RNAi in plants. On the basis of genetic and biochemical data, possible models for dsRNA induced gene silencing have been proposed (Vance 2001 et al.; Waterhouse 2001a; b; c; Zamore 2002). The model pathway has been modified with a growing understanding of the phenomena by means of the information provided via identification of new intermediates, and components of the complexes (Finnegan and Matzke 2003; Bartel and Bartel 2003). Comparable mechanisms were also demonstrated for RNAi in Drosophila melanogaster, C. elegans and in Arabidopsis thaliana. According to this recently proposed mechanism (Figure 2-5), aberrant single-strand (ss) RNAs, which are transcribed from a transgene, trigger the formation of

PAGE 26

15 double-strand RNA (dsRNA) by RNA dependent RNA polymerases (RdRP) and consequently the RNAi pathway is activated (Finnegan and Matzke 2003, Kusaba 2004). First, long dsRNA is recognized by a member of the RNase III family, Dicer, and digested into 21-24 nucleotide (nt) short interfering RNA (siRNA). The siRNA duplex consists of a 19 nt double strand region with 2 nt 3’ overhangs. This duplex is then unwound and one of the two strands is incorporated into the RNA induced silencing complex (RISC) often preferentially. The antisense strand of the siRNA in a RISC complex then hybridizes to mRNA as a guide, and the RISC cleaves. However, recent studies have indicated that gene silencing mechanism may be more complex than indicated by this model. Plants make two functionally discrete size classes of small RNA. A short class, 21-22 nt, has a role in mRNA degradation, as described above (Tang et al. 2003). The long size class, 24-26 nt, has been implicated in DNA methylation and systemic silencing in plants (Hamilton et al. 1999). Recently, four-dicer-like (DCL) proteins were identified in Arabidopsis (Schauer et al. 2002). DCL1 and DCL4 have one or more nuclear localization signals (NLS), which implies that there are both nuclear and cytoplasmic RNAi pathways for processing dsRNA in plants. Separate DCL activities might be necessary for the production of different classes of small RNAs (Tang et al. 2003; Papp et al. 2003). Dicers proceed in complexes with other proteins including members of the Argonaute family (Carmell et al. 2002) and HEN1 (Park et al. 2002; Boutet et al. 2003). Distinct family members of these proteins might affect the activity and function of DCL in complexes, consequently different size small RNAs might be generated.

PAGE 27

16 Micro RNAs (miRNA) are also recognized for their key role as siRNAs in the RNAi pathway. Hard work to clone size selected RNAs have revealed numerous miRNAs in C.elegans, A. thaliana and mice (Lagos-Quintana et al. 2001, Lagos-Quintana et al. 2002; Lau et al. 2001; Park et al. 2002; Ambros et al. 2003). miRNAs are derived from pre-miRNA that are transcribed from non-protein-coding genes of plants and animals (Carrington and Ambros 2003). The pre-miRNAs form self-complementary stem-loop structures (Figure 5). These 70-200 nt stem-loop structures are also recognized and processed with a Dicer to generate 21-22 nt miRNAs. In Arabidopsis, DCL1 processes miRNA precursors in the nucleus. The other miRNA precursors might be processed by a Dicer in the cytoplasm (Lee et al. 2002). miRNAs are natural negative regulators of specific target mRNAs that have roles in implementing developmental programs in animals and plants (reviews, Bartel and Bartel 2003; Carrington and Ambros 2003; Denli and Hannon 2003; Hunter and Poethig 2003). miRNAs and siRNAs silence at the posttranscriptional level; despite the differences in their origin, their functions are interchangeable (Carrington and Ambros 2003). siRNAs associated with an endonuclease-containing complex (RISC) cause degradation of cognate mRNAs in plants; miRNAs associated with a RISC-like complex can either basepair with the 3’UTR of mRNAs and block translation (Figure 2-5) or act in the manner of siRNAs and guide mRNA degradation. The choice between miRNA or siRNA paths is probably determined by the degree of complementarity between a given miRNA/siRNA and its target mRNA. However, plant miRNAs complementary to 3’UTR can probably guide a translational repression. Additionally, various members of the Argonaute (Ago) family might determine whether siRNA or miRNA will be the substrates for RISC. Cloning of small

PAGE 28

17 RNA has recovered 19 unique miRNAs that are encoded by 41 miRNA genes in plants (Bartel and Bartel 2003). The majority of them are complementary to transcription factors, which control fate-decisions in the shoot meristem, whereas the others regulate the intensity of DCL1 mRNA (Xie et al. 2003). Some small RNAs identified in plants might target native promoters of endogenous genes (Figure 2-5), which reveals that RNAi can be lethal when the critical regions in promoters are targeted (Park et al. 2002). RNAi is a highly efficient knockdown technology for genome-wide analysis of gene functions and improvement of economically important crops. In plants RNAi is often achieved by a transgene that produces hairpin RNA (hpRNA). (Chuang and Meyerowitz 2000; Smith e al. 2000; Wesley et al. 2001; Helliwell and Waterhouse 2003). Chuang and Meyerowitz (2000) reported that dsRNA-generating construct caused 87-99% of silencing in flower development-genes of Arabidopsis. The Waterhouse group designed different hpRNA constructs to obtain efficient silencing (Wesley et al. 2001). Transformation of hp constructs containing sense/antisense fragments ranging from 98 nt to 853 nt presented efficient silencing in studied plant species. Addition of an intron as a spacer in between sense and antisense fragment in transformation constructs, consistently enhanced the transfer efficiency from 55% to 90-100% (Wesley et al. 2001). The first commercial crop successfully produced by using the RNAi technology is low glutelin content-1 (LGC1) rice (Kusaba et al. 2003). It is valuable for patients with kidney disease who can have only limited amounts of protein. In this case hpRNA is produced from an inverted repeat for a glutelin gene leading to lower glutelin protein levels in rice, The efficiency of the hpRNA producing construct was 95 %, whereas the efficiency was ~50% with the sense construct.

PAGE 29

18 In summary, there are multiple ways to silence a target gene via transgene introduction. Whether a particular method would be effective in a given situation may depend on a variety of factors, including the type of a construct/constructs employed, the transgene-promoter, and the type of a gene/genes. Designing sense, antisense, or dsRNA-generating constructs of selected genes that could silence or reduce the expression of the genes in the naringin biosynthetic pathway might be an effective way to decrease the bitter taste of grapefruit. In particular, constructs that would give rise to dsRNA, particularly hpRNA generation, may be especially promising, given the recent reports in different plants species other than citrus. Difficulties in Citrus Breeding Most citrus types reproduce as facultative apomicts via nucellar embryony (Hodgson 1967; Grosser and Gmitter 1998; Moore 2001). This apomixes prevents gene exchange, which is required to produce new varieties and combine desired traits using conventional breeding techniques. Present citrus cultivars grown as scions or rootstocks mainly arose by chance seedling, limb or bud sport mutations. They were then propagated vegetatively. Despite the great genetic diversity in the Citrus genus, variability is limited in commercially important “species” such as grapefruit and sweet orange, which arose as interspecific hybrids. Variation within these types must occur via somatic mutation and then plants with desirable traits are propagated as cultivars. The presence of high heterozygosity in the genus also makes it difficult to understand the genetic control of significant traits. As a result, a cross between complementary parents often fails to produce favorable recombinants. Crosses may also yield weak/unhealthy progeny because of inbreeding depression. Moreover, long juvenile phases and a large tree size increase the cost of hybrid evaluation.

PAGE 30

19 Figure 2-5. A model for RNAi, modified from Finnegan and Matzke (2003). For all of these reasons, it is difficult to improve citrus varieties by conventional breeding practices. On the other hand, molecular genetic techniques such as genetic transformation can be very useful. Gene cloning and genetic transformation allow the significant characteristics associated with important citrus types to be maintained, while adding a single new favorable trait of commercial value. Additionally, unfavorable traits such as bitter taste can be suppressed using antisense or co-suppression strategies. Citrus Transformation Genetic transformation of citrus began with the introduction of bacterial plasmid DNA into protoplasts isolated from orange (Citrus sinensis Osb.) cell suspension cultures

PAGE 31

20 (Kobayashi and Uchimiya 1989). Protoplasts were suspended in mannitol and then mixed with the bacterial plasmid pCT2T3. Subsequently, polyethylene glycol solution was added to the protoplast-DNA mixture. Four-week cultured protoplasts were transferred to the first selection medium containing 50mg/l kanamycin sulfate. After transfer to the second selection medium, which contain 25 mg/l kanamycin sulfate, eight clones survived, two of which were stable transformants, while three of them showed weak signals with Southern blot hybridization. In this study no transgenic plants were regenerated. Transformation efficiency was estimated as 1.0 x10-6. Consecutively, pCAP212 plasmid DNA was transferred to rough lemon (C. jambhiri Lush) protoplasts (Vardi et al. 1990) by using a similar method. This time the selectable marker Kan was replaced with paromomycin sulfate (PAR), which was reported as more efficient than Kan for some plant systems. A total of 21 clones were isolated, 9 clones were stable transformants, and 2 clones were regenerated into transgenic plants. Transformation efficiency was not reported. Since the Agrobacterium host range is limited, a protoplast system was used to introduce DNA into citrus at the beginning. However, an increase in reports of transgenic work on woody plants motivated the studies on Agrobacterium-mediated transformation to citrus. The first Agrobacterium-mediated transformation was performed using suspension culture cells of citrus species (Hidaka et al. 1990). Embryo callus lines of Washington navel orange (C. sinensis Osbeck), Ohta ponkan and Kara mandarin (C. reticulata Blanco), and a pollen embryoid callus of Trovita orange (C. sinensis) were cocultivated with two strains of Agrobacterium tumefaciens A415 or GV3010. A415 harbors a binary vector, which includes a neomycin phosphotransferase (nptII) gene and

PAGE 32

21 GV3010 harbors a cointegrative vector, which contains a hygromycin phosphotransferase (hpt) gene. Cocultivated cell colonies were selected based on their resistance to kanamycin sulfate or hygromycin sulfate. After the second selection, embryoids were transplanted on plant regeneration media. One transgenic plant was regenerated from Washington navel orange. Transformation efficiency was estimated as 7.0 x10-3. Production of embryo derived callus is very difficult or impossible for some of the citrus cultivars (Gmitter and Moore 1986). Therefore, requirement of citrus embryogenic callus limits these transformation methods. A direct gene transformation method, which utilizes seedling internodal stem segments as explants, was reported by Moore in 1992 (Moore et al. 1992). Several citrus cultivars were tested for their ability to regenerate from different explants and for their susceptibility to various A. tumefaciens strains. Seedlings from Carrizo citrange [C. sinensis (L.) Osb. X Poncirus trifoliata (L.) Raf.]; Swingle citrumelo (C. paradisi Macf. X P. trifoliata); Key lime [C. aurantifolia (Christm.) Swing], and Hamlin sweet orange [C. sinensis (L.) Osb.] were grown in sterile tubes for 2-4 months in 16h light conditions. Seedlings were cut ~1cm in length and inserted into cocultivation medium with one of the ends protruding. The protruding ends were inoculated with a few drops of engineered A. tumefaciens strain EHA101 harboring either a pGA472 or a pMON9793 binary plasmid. After a two to three day coculture period, the cut stems were transferred to shoot generating medium containing 100 mg/L kanamycin sulfate. Of all the cultivars, Carrizo citrange and Swingle citrumelo were the most regenerable, Key lime was intermediate, and Hamlin sweet orange regenerated very poorly. Only two plants were regenerated into soil, because of the difficulties in rooting. Transformation efficiency was estimated as 4

PAGE 33

22 8% with this method. Similar methods with some modifications were used to transform Poncirus trifoliata which is a citrus relative [Kaneyoshi (Hiramatsu) et al. 1994]. In this method, seedlings were grown for 20 days in dark to obtain epicotyl segments which produce adventitious buds with a high frequency (Hiramatsu et al. 1987). Epicotyl segments, were cut ~1cm in length, immersed for 15 min with A. tumefaciens strain LB4404 harboring pBI121 or pBI101-012-p1 binary vectors and placed horizontally onto coculture medium. After a 3 day co-culture period in hormone free medium, segments were transferred to selection medium containing 100 mg/L kanamycin sulfate. More than a hundred transformed plants were regenerated. The transformation efficiency was reported as 55.4-87.7% for P. trifoliata, whose hybrids are more efficiently regenerable than the other citrus cultivars (Moore et 1992; Gutierrez et al. 1997). The technique of using stem segments has become widely accepted for citrus Agrobacterium-mediated transformation, because of its simple and efficient features. Further studies have concentrated on increasing the efficiency of transformation in various citrus cultivars by using this method. Pea et al. (1995) increased the exposure time of stem segments to selection agent, in order to increase the transformation efficiency. Five-week old Carrizo citrange [C. sinensis (L.) Osb. X P. trifoliata (L.) Raf.] seedlings were cut and soaked in A. tumafaciens EHA105 harboring p35GUSINT for two days. After two days, segments were placed horizontally on shoot regeneration medium that contained 100 mg/L kanamycin sulfate. Stem segments were maintained in the dark for 8 weeks and then transferred to light conditions. Segments were transferred every four weeks to fresh medium. After 1-4 months, shoots regenerated, which were harvested and assayed for GUS activity. Only 2 out of 32 shoots were GUS positive which reflects the

PAGE 34

23 transformation efficiency being 6.25%. Nonetheless, 1-2 months later the same segments developed new shoots, and a high frequency of GUS positive shoots was obtained (55.1%). Stem segments that were exposed to selection media longer produced more GUS positive shoots by decreasing the frequency of escapes. In order to avoid losing shoots during rooting (Moore et al. 1992), GUS positive and chimeric shoots were grafted onto Troyer citrange seedlings. Although 100% graft success was obtained, 46.8% of the regenerated shoots were morphologically abnormal. Such shoots showed abscission, bleached and died within 2-3 weeks of grafting. After 3 weeks the remaining grafted plants were grafted on to 5-month old rough lemon (C. jambhiri Lush) seedlings because of their vigorous growth. At the end of the experiments transformation efficiency was estimated as 20.6%. A similar procedure was used to establish an efficient Agrobacterium-mediated transformation protocol for Mexican (Key) lime (C. aurantifolia Swing.), which is a commercially important citrus variety (Pea et al. 1997). Six to twelve month old greenhouse-grown seedlings were cut to have ~1cm long stem segments. The stem pieces were immersed in A. tumefaciens (Pea et al. 1995) for 15-30 minutes and placed horizontally either on MSB1 medium (Murashige and Skoog with White vitamins) or tomato feeder plates supplemented with acetosyringone, which induces the virulence of A. tumefaciens and wound response to wounding in lime stem segments. After 3 day cocultivation in dark, explants were transferred to selection media that contains 100mg/L kanamycin sulfate and they were transferred to 16h light condition after they were kept in dark for 2-4 days. Using tomato feeder plates and dark condition for cocultivation, increases the callus formation and thus transformation frequency. Feeder plates showed 82% transformation efficiency on kanamycin selection media,

PAGE 35

24 whereas MSB1 plates showed only 42.2%. After a period on the selection media, explants produced shoots that were shown to be transgenic. These transgenics were then grafted onto Troyer citrange as described previously (Pea et al. 1995). A total of 16 GUS positive plants were successfully produced as a result of these experiments (Pea et al. 1997). The first report on integration and expression of a foreign gene in commercially citrus varieties came from the Moore laboratory (Gutierrez et al. 1997). Four citrus types were chosen as explant sources for Agrobacterium-mediated transformation. These were namely Carrizo citrange, which was easy to transform; Key lime which has a short juvenile period; and sour orange and Pineapple sweet orange which were economically important. The objectives of this study were to decrease the number of escapes and to increase the frequency of rooting. Stem segments were inoculated with Agrobacterium strain EHA101 harboring pGA482GG or pGA482-CTVCP (CTV coat protein). After 1-3 day cocultivation in dark, they were transferred to selection media which contains 100 mg/L kanamycin. Placing stem segments horizontally on plates and 100 mg/L kanamycin was effective to decrease the number of escapes. Although adding 1mg/L BA and alternatively, transferring GUS positive shoots directly to soil demonstrated increased frequency of rooting in Carizzo citrange shoots, rooting still remained to be a problem. The transgenic plants, which were GUS positive, were tested for CTV-CP protein expression by immunoblot analysis. Some of them effectively showed the CTV-CP expression. To avoid the rooting problem in Agrobacterium-mediated citrus transformation, micrografting of transgenic shoots onto seedling rootstocks was commonly applied by different groups (Pea et al. 1995, Bond and Roose 1998, Yang et

PAGE 36

25 al. 2000). Meanwhile, studies to establish transformation protocols for different citrus varieties, and to improve available citrus transformation specifically by increasing the rooting have been conducted. First grapefruit (C. paradisi cv. Duncan) transformation was demonstrated by Luth (1999). A previously established transformation protocol was performed on 4-6 week old grapefruit seedlings (Gutierrez et al. 1997; Moore et al.1993). In order to improve the rooting efficiency, shoots were kept for three weeks in hormone free medium to eliminate the affects of BA which is included in selection medium. After 3 weeks shoots were transferred to the MS medium containing 5 M NAA, which seems to increase the rooting. A rather high overall transformation efficiency of 43.5% was reported for the grapefruit. This was the highest transformation efficiency that was reported for grapefruit without having rooting problem. Therefore we followed the same protocol with slight modifications in our transformation experiments. Manipulation of the Flavonoid Pathway in Citrus The flavonoid biosynthetic pathway has been manipulated successfully in other plant species, mostly to affect the flower color of ornamental plants (Dixon and Steel 1999; Davies 2000; Forkman and Martens 2001). It has also been reported that transferring heterologous CHI into tomato plants increases the antioxidant activity in peels via increasing the production of flavonol compounds (Muir et al. 2001). Our goal is to manipulate the flavonoid biosynthetic pathway to alter the levels and types of flavanones in citrus fruits for improving their palatability and/or pharmacological properties. Many citrus varieties could be improved by the manipulation of the flavonoid pathway, but grapefruit is probably the most appropriate type to begin with, because of its bitter and tasteless flavonoid compounds and high productivity. The

PAGE 37

26 state of Florida is the main grapefruit producer in the world (FAO, FASS 1970-2003). Manipulation of the flavanone glycosides might decrease the bitter taste either by a reduction in the concentration of bitter compounds or by increasing the concentration of tasteless compounds that can block perception of the bitter ones. If this strategy is successful, other bitter citrus varieties can be improved with this method as well. Sour orange produces attractive, nice colored and good size fruits, but because of their bitterness, this species is used primarily as a rootstock (Rouseff et al. 1987). Poncirus trifoliata (trifoliate orange) fruit are extremely bitter and so rich in naringin and poncirin that the tree is also used only as a rootstock (Hodgson 1967). However, this interfertile citrus relative is very cold hardy and is resistant to phytopathogens like citrus tristeza virus that plague citrus. Crosses made between P. trifoliata and Citrus to move these valuable characteristics into edible types have resulted in hybrids that frequently are cold hardy and disease resistant but produce bitter fruits. Decreasing the bitter taste in P. trifoliata or in its hybrids via molecular techniques might even extend citrus fruit production to cooler zones of the world.

PAGE 38

CHAPTER 3 GENETIC TRANSFORMATION OF CITRUS WITH FLAVONOID PATHWAY GENES Introduction Citrus is a fruit crop broadly grown in tropical and subtropical regions of the earth. Despite the extensive cultivation and consumption of citrus by humans, conventional breeding is hampered in improving citrus varieties for a number of reasons. Most citrus types reproduce as facultative apomicts (Hodgson 1967; Grosser and Gmitter 1998; Moore 2001); their seeds contain almost entirely asexual embryos, which prevents gene exchange, hence the production of new varieties. Present citrus cultivars predominantly arose by chance seedling, limb or bud sport mutations, subsequently, desirable traits were propagated vegetatively. High heterozygosity in the genus also complicates the genetic control of significant traits. Consequently, a cross between complementary parents often fails to produce favorable recombinants, and furthermore may yield weak/unhealthy progeny due to inbreeding depression. Moreover, long juvenile periods increase the costs of hybrid assessment. Alternative methods such as genetic transformation can be useful to improve the existing varieties (Gutierrez et al. 1997; Pea et al. 2001; Febres et al. 2002). Gene cloning and genetic transformation allow the significant characteristics associated with important citrus types to be maintained, while adding a single new favorable trait of commercial value. Additionally, unfavorable traits such as bitter taste can be suppressed by antisense or co-suppression strategies with the help of transformation techniques. 27

PAGE 39

28 Genetic transformation of citrus was initiated with the introduction of plasmid DNA into citrus protoplasts with the help of PEG (Kobayashi and Uchimiya 1989, Vardi et al. 1990) electroporation (Hidaka and Omura 1993), and particle bombardment methods (Yao et al. 1996). Successful reports on transformation of woody plants encouraged the studies on Agrobacterium-mediated transformation of citrus. The first Agrobacteriummediated transformation was performed using embryo cell lines of selected citrus varieties (Hidaka et al. 1990). The requirement for citrus embryogenic callus limits this transformation method, since the production of embryo derived callus is very difficult or even impossible for some citrus cultivars (Gmitter and Moore 1986). Alternatively, seedling internodal stem segments have been utilized for transformation experiments. The technique of using stem segments in Agrobacterium-mediated citrus transformation was predominantly accepted because of its simple and efficient features (Gutierrez et al. 1997; Kaneyoshi et al. 1994; Moore et al. 1992; Pea et al. 1995 a; b; Pea et al. 1997). Further studies concentrated on increasing the efficiency of transformation and improving rooting success for the recovery of transgenic plants (Luth and Moore 1999). On the other hand, various groups have utilized micrografting methods by grafting transgenic shoots onto seedlings in vitro or in vivo to avoid the rooting problem (Pea et al. 1995; Bond and Roose 1998; Yang et al. 2000). The first successful rooting protocol was reported by Luth and Moore (1999). In this laboratory, a previously established transformation protocol was performed on 4-6 week old grapefruit seedlings (Luth and Moore 1999). In order to improve the rooting efficiency, regenerated shoots were kept 3 weeks in hormone free medium to eliminate the affects of BA (benzyladenine), which is included in the selection medium. The regenerated shoots were

PAGE 40

29 then transferred to MS (Murashige and Skoog) medium containing 5M NAA (-naphthalene acetic acid), which appears to increase rooting. The overall transformation efficiency with this protocol was 43.5%, which is the highest reported for grapefruit thus far. Therefore, in this research, we followed the same transformation protocol to transfer our genes of interest. Previously, chalcone synthase (CHS) and chalcone isomerase (CHI) genes were isolated from grapefruit. Sense and antisense constructs of the genes were transferred to grapefruit with the goal of reducing the naringin level in transgenic plants (Luth 1997). According to the preliminary HPLC results, one of the transgenic plants containing a CHI antisense construct displayed an almost 20-fold reduced level of naringin compared to control plants. To confirm these results, a higher number of transformants and more analysis were needed. Thus, the objective of this study was to obtain more transgenic plants, utilizing the previously established transformation method and flavonoid pathway genes of interest, to draw a better conclusion about the effects of the flavonoid pathway genes on the levels of flavonoids in transgenic plants, particularly the reduction of naringin levels. To begin the initial pilot experiments, CHS and CHI cDNAs were cloned from Marsh grapefruit previously but their sequences were not submitted to Genbank (Luth 1997). Later, two family members of the CHS gene (CitCHS1AB009350 and CitCHS2AB009351) and one CHI (AB011794) gene were cloned from C. sinensis (Moriguchi et al.. 1999). CitCHS2 is almost (98%) identical to the CHS that was previously cloned by Luth (1997). Therefore, to prevent complications in the future, here the original CHS clone that was isolated in this laboratory and was used in these experiments is referred to as CHS2.

PAGE 41

30 To accomplish our main objective we produced more transgenic plants by utilizing the previously designed CHS2 and CHI sense / antisense constructs. Moreover, we transferred constructs of a recently cloned citrus 1,2 rhamnosyl transferase gene (AY048882). Although the enzyme was purified earlier (Bar-Peled 1991), cloning the gene was hindered due to the high sequence similarity and great number of glucosyltransferases in plants (Vogt and Jones 2000). The product of the gene catalyzes the last step in the naringin biosynthesis pathway. Therefore manipulation of this gene should be a more specific to the alteration of naringin production. The transformation method, the plasmid vectors that have been employed, and the confirmation of transgenic plants via GUS histochemical assays and PCR will be discussed for each construct. Analysis of transgenic plants will be presented in the next chapter. Materials and Methods Agrobacterium strains and vector plasmids A total of seven transformation plasmids were utilized for our gene transformation experiments. Previously four vector plasmids, chalcone synthase in both the sense (CHS-S) and antisense (CHS-AS) directions and chalcone isomerase in both sense (CHI-S) and antisense (CHI-AS) directions had been constructed by former PhD student Diane Luth (Luth 1997). Three 1,2 rhamnosyl transferase (1,2 RT) plasmids were constructed by our colleague Dr. Yoram Eyal’s group. These constructs are pBinplus1,2RT-HP (1,2 RT hairpin or pan-handle), pBinplus1,2 RT-S (1,2RT sense) and pME1,2RT-S. Additionally, two empty vector plasmids pGA482/GG and pBinplus were exploited in transformation experiments for control purposes. Details of preparations of all the constructs were presented below.

PAGE 42

31 To construct the CHS and CHI plasmids (Figure 3-1) Diane used a paired system, which utilizes pUC18-exp17 and pGA482/GG vectors, was used (Slightom 1991). This system was chosen because of its availability and verified effectiveness in citrus transformation experiments (Gutierrez et al. 1995). The 1.4 kb CHS and 0.9 kb CHI cDNAs were isolated from C. paradisi and cloned into the pUC18-exp17 vector, which contains an expression cassette that includes the cauliflower mosaic caulimovirus (CaMV) 35S promoter, a 70 nucleotide 5’ untranslated region (5’UTR) from the cucumber mosaic cucumovirus (CMV) to enhance translation, an Nco I cloning site, and a CaMV polyadenylation (poly-A) signal. The sense and antisense orientations of the clones were verified. Then, pUC18-exp17 plasmids that contained the genes of interest were digested with Hind III and ligated with Hind III digested pGA482/GG plasmid (Figure 3-1). pGA482/GG contains right (RB) and left (LB) T-DNA borders. Between the borders there is a neomycin phosphotransferase resistance gene (NPTII) to provide the kanamycin resistance for selection of transformants; a Hind III-containing multiple cloning site, and a GUS ( -glucuronidase) reporter gene. Outside the borders there is a gentamycin resistance gene for maintaining the plasmid in Agrobacterium strains and a tetracycline resistance gene for growth in E. coli cell lines. After confirmation for the size and orientation of the inserts, the four transformation vectors were individually transformed into Agrobacterium strain EHA101 (Hood et al. 1986). EHA101 contains a disarmed pTiBo542 that is a hypervirulent tumor inducing plasmid, which has the vir components required for infection and integration into plant cell nuclei. To construct pBinplus1,2RT-HP, 1,2 rhamnosyl transferase cDNA was isolated from grapefruit (Frydman unpublished). A sense fragment was amplified using a primer

PAGE 43

32 set which consisted of forward (5’-CTTGTCATGAATACCAAGCATCAAGATAAG-3’) and reverse (5’GGATCCTTATTCAGATTTCTTGACAAGC-3’) primers and was then cloned into the pGEM-T Easy vector (Promega). The fragment was digested with Nco I and BamH I and cloned into the vector pRTL2 by replacing the GUS region. The antisense fragment was amplified using a primer set which consisted of forward (5’-GTTGAGATCTGATACCAAGCATCAAGATAAGCCAA) and reverse (5’CTACAGATCTAGCCTGCGGAACCCAACCTTGTAC-3’) primers, and was then cloned into pGEM-T Easy. The fragment was digested with Bgl II and cloned in the antisense orientation into the BamH I site, which is downstream of the cloned sense fragment. The entire insert of the pRTL2 plasmid (35S promoter-1,2RT sense-antisense-35S terminator) was isolated by Sph I digestion. The ends were filled by using Klenow fragment, and the fragment was cloned into the Sma I site of the pBINplus (Engelen van et al. 1995) transformation plasmid (Figure 3-2). To compare the transformation efficiency of the 1,2RT-HP and 1,2RT-S plasmids, which contain the gene of interest in different orientations, 1,2RT-S constructs were cloned into 2 plasmids. Due to its availability, first, 1,2RT-S was cloned into the pME504 (Figure 3-3) binary plasmid (Dr. Eyal Group). The 1,2RT was amplified by the set of the forward (5’GCTCAGGTCTCTAGAGAACATGGATACCAAC-3’) and reverse (5’-GGTCGAGAGCTCTTATTCAGATTTCTTGAG-3’) primers, and a proofreading polymerase enzyme. The amplified fragment was cloned into pGEM-T Easy, and subcloned into the Xba I and Sac I enzyme sites of the pME504 vector plasmid under the control of the 35S promoter. However, for a more direct comparison with the 1,2RT-HP, the 1,2RT sense fragment was cloned into pBINplus. The 1,2RT-S cDNA was amplified

PAGE 44

33 via forward (5’-GTCATGAATACCAAGCATCAAGATAAG-3’) and reverse (5’-GGATCCTTATTCAGATTTCTTGACAAGC-3’) primers and cloned into the Nco I and BamH I sites of the pRTL2 plasmid. The pRTL2-1,2RT construct was digested with Sph I and the fragment containing the 35S promoter-1,2RT cDNA-35S terminator was isolated, filled by Klenow, and cloned into the Sma I site of the pBINplus transformation vector (Figure 3-2B). For genetic transformation experiments all of the 1,2RT plasmids were transferred to Agrobacterium strain EHA105 because of its suitable antibiotic resistance gene for this particular plasmid. Plant material, transformation and regeneration C. paradisi Macf. cv.Duncan (grapefruit) seeds were locally purchased. Seed coats were removed by forceps, and the seeds were sterilized in 70% ethanol for 3 min, followed by 0.525% sodium hypochlorite, in which 0.05% (1-2 drops) Tween 20 was added, for 10 min, then the seeds were rinsed with sterile water 3 times for 1 min each (Luth and Moore 1999). After the last rinsing step was completed, the seeds were placed individually into 15x20 cm tubes, which contained in vitro germination medium (Figure 3-4). The tube racks were kept in the dark for 4-6 weeks to obtain etiolated seedlings. Germinated-etiolated seedlings were employed in transformation experiments. The EHA101 Agrobacterium cultures that harbored pGA482/GG with the genes of interest were grown in YEP (yeast extract-peptone) medium with 60 mg/L gentamycin (Sigma Chemicals, St Louis, MO) and 50 mg/L kanamycin ( Fisher Biotech Fairlawn, NJ) to an OD620= 0.6-1.0. The 1,2RT-HP constructs were harbored by EHA 105 Agrobacterium culture and they were grown in YEP medium in the presence of rifampycin 50 mg/L and kanamycin 50 mg/L. Cultures were centrifuged at 3500 rpm for

PAGE 45

34 5 min. The pellet was resuspended to a cell density of 5x108 5x1010 in MS medium with freshly made 100 M acetosyringone added (Kaneyoshi et al. 1994). Epicotyls of the etiolated seedlings were cut into 1 cm segments (Kaneyoshi et al. 1994; Luth and Moore 1999). The segments were immersed in Agrobacterium resuspended solution for 1 min, then excess liquid was blotted on a sterile paper towel. Sixteen segments were placed horizontally onto a 100x15 mm petri plate that contained (MSA) medium (100 M acetosyringone plus MS). The plates were kept in darkness for 2-3 days at 28oC to allow Agrobacterium infection to take place, then segments were transferred to selection medium under 16h cool-white fluorescent light conditions at 27-28 oC to induce shooting. The control segments were kept in MSBC media; its composition was the same as selection medium except without kanamycin. The selection medium (MSBCK) contains Murashige and Scoog basal salts (MS) 4.33 g/L, (Phytotechnology Laboratories, Shawnee Mission, KS) benzyladenine (BA) 0.5 mg/L, kanamycin monosulfate 75 mg/L (Fisher Biotech Fair Lawn, NJ), and claforan 500 mg/L (Abbott Laboratories, Abbot Park, IL). The segments were transferred to fresh selection medium every 4 weeks. When regenerated shoots became 0.3 0.5 cm in length, they were removed from the segments and a very little portion was cut from their basal end for a histochemical test to confirm whether if they were GUS positive or not (Moore 1992). GUS-positive shoots were transferred to NAA (-naphthalene acetic acid, Sigma Chemicals, St. Louise, MO) containing (5 mg/L) rooting medium to induce roots. Only 6 control shoots were transferred to rooting medium. After the roots appeared and became at least 0.5 cm long, the rooted plants were transferred to soil cups containing MS medium that was mixed with Metromix 300 soil (Scotts Marysville, OH). The cups

PAGE 46

35 were covered with plastic wrap to maintain moisture in the cups and kept in a growth chamber at 28-30 oC, 16h light/8h dark cycle. A few days later, small holes were opened in the wraps to water the plants and the number of holes was increased over time to decrease the extra humidity slowly. When the plantlets became well established in soil, they were retested with a histochemical GUS assay. GUS positive plantlets were transferred to pots. When transgenic and control plants were approximately 20 cm in height, they were transferred to the greenhouse. Histochemical GUS assay A small section from the basal end of each regenerated shoot was cut and placed in a well of a round bottom 96-well microtiter plate (Corning Incorporated Corning, NY 14831) containing 30 l of x-gluc (1 mg/ml 5-bromo-4-chloro-3-indolyl--D-glucuronide in 0.1 M sodium phosphate buffer, pH 7.0 with Na2EDTA) (Moore et al. 1992) per well. The plate was incubated at 37 oC up to overnight. Then the tissue was washed and fixed with 50 l 95% ethanol: glacial acetic acid (3:1 v/v). Thirty minutes later, blue sectors were scorable under the microscope. A second GUS assay was performed on leaf segments that were collected from the rooted plants. PCR Analysis DNA was extracted from the leaves of GUS positive plants by using the DNAzolES method (Molecular Research Center, Cincinnati, OH). Only 1 l DNA extract was employed for each 50 l PCR reaction. PCR reaction buffer, MgCl2 and Taq polymerase were provided by Promega (Promega Madison, WI). The primer concentration was 1ng for each primer individually in a 50 l PCR reaction unless otherwise changed for a specific reaction. For the analysis of CHS and CHI gene transferred plants, multiplex PCR was performed by using NPTII forward 5’

PAGE 47

36 TCACTGAAGCGGGAAGGGACT-3’ and reverse 5’CATCGCCATGGGTCACGACGA -3’ primers, which give a ~380 bp PCR product and GUS forward 5’CAACGAACTGAACTGGCAG -3’ and reverse 5’– CATCACCACGCTTGGGTG -3’ primers, which give a 780 bp PCR product. The cycles were as follows: 94 oC, 2 minutes, to denature the DNAs; then 29 cycles, 94 oC, 30 seconds; 55 oC, 30 seconds; 72 oC, 50 seconds; afterwards, 72 oC 5 minutes. Twelve l PCR products were applied into a 2 % agarose gel (Gibco-BRL). For the analysis of 1,2RT-HP and 1,2RT-S putative transgenic plants, PCR was performed utilizing the NPTII forward and reverse primers only. Most of the transformants were positive for the NPTII gene. Therefore, to distinguish between NPTII positive only and NPTII positive plus target gene 1,2RT transformed plants, we analyzed the same plants utilizing a different primer set, 35S promoter forward primer 5’-TATATTAGATCTCCGGAAACCTCCTCGGATTCCAT3’ and 1,2 RT reverse primer 5’ATACACGACTGAACGAGGCTC -3,’ which give a 1300 bp PCR product for the 1,2RT-HP insert and 1200 bp products for both 1,2RT-Sense constructs. The cycles were as follows: 94 oC, 2 minutes, to denature the DNAs; then 32 cycles, 94 oC, 30 seconds; 55 oC, 30 seconds; 72 oC, 1 minute: afterwards, 72 oC, 5 minutes. Twelve l PCR products were applied into a 2 % agarose gel.

PAGE 48

37 Figure 3-1. Transformation constructs were used to transfer full length chalcone synthase (1.4 kb) and chalcone isomerase (0.9 kb) cDNAs into citrus. Four transformation plasmids were constructed by using the paired vector system (Slightom 1991). Hind III expression cassette (top) from pUC 18-exp 17 includes the gene of interest; A) Chalcone synthase-sense (CHS-S), B) Chalcone synthase-antisense (CHS-AS), C) Chalcone isomerase-sense (CHI-S), D) Chalcone isomerase-antisense (CHI-AS), cloned into the Hind III site of pGA482/GG transformation vector (bottom). CP AMV 35S romoter CAMV PolyA CMV 5’UTR HindIII HindIII CHS-S CAMV 35S Promoter CAMV PolyA CMV 5’UTR CAMV 35S Promoter CAMV PolyA CMV 5’UTR HindIII HindIII CHS-AS C AMV 35S Promoter CAMV PolyA CMV 5’UTR HindIII CHI-S CAMV 35S Promoter CAMV PolyA CMV 5’UTR HindIII CAMV 35S Promoter CAMV PolyA CMV 5’UTR HindIII HindIII CHI-AS D) C) B) A)

PAGE 49

38 EcoRI XhoI KpnI Figure 3-2. Two 1,2 rhamnosyl transferase transformation vectors. A) The full length 1,2RT sense (S) and antisense (AS) fragments were cloned separately, then they were cloned next to each other in the pRTL2 vector.The entire insert was digested and cloned into the SmaI polylinker site of the pBinplus binary vector to construct the 1,2RT-HP (1,2 rhamnosyl transferase hairpin) clone. B) The full length 1,2RT sense fragment cloned into the pRTL2 vector. The entire insert digested and cloned into the SmaI polylinker site of pBinplus binary vector to construct 1,2RT-S. These two binary vectors do not contain GUS as a reporter. CAMV 35S Promoter 1,2 RT-AS 1,2 RT-S CAMV 35S Terminator 1360bp 1000bp PstI PstI HincII EcoRI XhoI 1,2RT-S 1,2RT-HP A) B) CAMV 35S Terminator 1,2 RT-S CAMV 35S Ptomoter

PAGE 50

39 EcoRI EcoRI PstI Figure 3-3. The 1,2RT-sense (1,2RT-S) transformation vector. 1,2 rhamnosyl transferase cDNA was subcloned into Xba I and Sac I sites of the pME504 binary vector to construct pME1,2RT-S. This vector does not contain a GUS gene as a reporter. HincII XhoI PstI CAMV 35S Ptomoter 1,2 RT-S CAMV 35S Terminator PstI PstI XhoI HincII 1,2 RT-S CAMV 35S Terminator CAMV 35S Ptomoter

PAGE 51

40 1239117812461413 10 5 1239117812461413 10 5 Figure 3-4. Agrobacterium-mediated citrus transformation tissue culture chart;1) Seedy Duncan grapefruit; 2) Peeling the seeds 3) Sterilization of the seeds, 4) Growing etiolated seedling in dark, 5) 3-4 week grown etiolated seedling, 6) Growing Agrobacterium strain 7) Incubation of seedling segments in Agrobacterium, at 28 oC, 8) Incubation of stem segments in kanamycin selection medium, 9) Cutting and placing of regenerated shoots into rooting medium and preperation of GUS histochemical assay, 10) Result of GUS histochemical assay, 11) Rooted shoots, 12) A rooted GUS positive shoot transferred to a pot containing soil mixture, 13) A GUS positive plantlet transferred to a bi gg er p ot g rowin g in g rowth chamber, 14 ) Survivin g GUS p ositive trans g enic and control p lants in g reenhouse.

PAGE 52

41 Results and Discussion A sequence of Agrobacterium-mediated transformation experiments was accomplished with four pGA482/GG constructs; CHS-S, CHS-AS, CHI-S and CHI-AS over four years to obtain a sufficient number of transgenic plants to analyze in detail and to draw a reliable conclusion. In addition, some experiments have been performed with the more recently isolated 1,2RT gene in the transformation constructs pBinplus1,2RT-HP, pME1,2RT-S and pBinplus1,2RT-S as well. All of the constructs were competent for the transfer of the target genes, resulting in regenerated shoots with some differences in transformation efficiency (Tables 3-1 to 3-7). All of the putative transgenic plants that were transformed with CHS and CHI gene constructs were analyzed by GUS assay at least two times followed by PCR analysis to confirm the presence of the transgene. We utilized GUS and NPTII primers to confirm the transgenic plants. Transformation efficiency was estimated for each experiment based on the number of GUS positive shoots over the number of total shoots produced [(number of GUS positive shoots/number of total shoots produced) x 100]. Estimated transformation efficiency varied from experiment to experiment; while some experiments revealed efficiencies up to 67 % (Table 3-2), some did not produce any positive shoots (Table 3-1). Average transformation efficiencies for CHS and CHI gene transformation constructs were 21 % to 34 % with the exception of CHS-S construct which yielded surprisingly low transformation efficiency (2.3%, Table 3-1). Although escapes were observed, the highest number of regenerated shoots and GUS positive shoots were produced by the CHI gene constructs. Moreover, most of the regenerated shoots successfully survived and grew to be plantlets. The CHS gene transformation constructs had a propensity of regenerating low numbers of shoots, sometimes not at all, compared

PAGE 53

42 to the CHI constructs. The putatively positive shoots for the CHS constructs were smaller than the control explant shoots and they often showed morphological abnormalities. Transformants grew slower than control shoots, they displayed abnormal leaf curling, and most of them died in the rooting medium. Even if they survived and had roots, they died in a short time after they were transferred to soil cups. Similar results were reported previously by Luth (1997) and by Kaplan (1999) in their dissertations. There are currently three surviving transgenic CHS-S plants in the greenhouse. There are no apparent morphological differences between control and the surviving transgenic plants. There are 12 transgenic CHS-AS plants in the greenhouse. PCR analysis revealed that they have both the GUS and NPTII genes. DNA of plants transformed with 1,2 RT constructs were first amplified with NPTII primers in a PCR reaction. Most of the plants died before they were transferred to soil. Remaining plants were analyzed by PCR. The majority of the plants were positive for the NPTII gene. These PCR positive plant DNAs were amplified with a primer set (35S promoter forward, and 1,2 RT reverse) to investigate the full length transgene in vivo. Only a few plants gave positive bands with this primer set (Figure 3-5). This might be due to the construct which might not be very stable. The hairpin constructs are less efficient than the intron included hairpin constructs (Wesley et al. 2001). The 1,2RT-HP construct was designed and was utilized before the efficiency of intron included constructs was known. Therefore, the 1,2RT-HP constructs might not be as efficient to silence the target gene as recent hairpin constructs. A possible reason that most of the plants died would be the deleterious effects of 1,2RT gene. Under or over expression of the 1,2RT gene might be detrimental in citrus. These few 1,2RT transgenic plants also grew slower than the control plants (Figure 3-6), and their leaves

PAGE 54

43 were smaller in size compared to controls. It has been shown that inhibition of glucosyltranferase expression by antisense mRNA under the control of its own promoter was lethal in pea and alfalfa (Woo et al. 1999, 2003). Transformation of glucosyltransferase mRNA as antisense construct to pea, alfalfa and Arabidopsis resulted in slow growth and altered life cycle, suggesting that glucosyltransferase might be required for normal growth and development of those studied plant species. Woo et al. (2003) hypothesized that glucosyltransferase has a substrate that controls cell cycle. This endogenous substrate is called factor controlling cell cycle (FCC), which might regulate functional activity of auxin by modulating its transport. Pilot studies on FCC indicate that its features are consistent with flavonoids. Flavonoids had been implicated in regulating endogenous auxin levels by affecting polar auxin transport (Jacobs and Rubery 1988; Muphy et al. 2000; Brown et al. 2001; Peer et al. 2001; Buer 2004). Glucosyltransferase might control auxin levels by modulating its substrate FCC through conjugation. Inhibition, over or under expression of the conjugating enzyme would result in deregulated auxin transport/uptake, which could cause diverse changes to regulation of growth and development (Woo et al. 2003). The 1,2RT gene might have a similar role in growth and development of citrus. Probably the level of the 1,2RT is tightly controlled in citrus, therefore under or over expression of the gene causes deregulation of axuin which in turn affects the growth and development of the transgenic grapefruit plantlets.

PAGE 55

44 1,2 RT-HP Sense Anti-sense 35Spromoter 35Sterminator 1.3 kb 1,2RThp plasmid 38 52 61 17 1.4kb Figure 3-5. PCR analysis of 1,2RT transgenic plants. Transgenic plants 38, 52, 61 were transformed with 1,2RT-HP construct and transgenic plant number 17 was transformed with 1,2RT-S construct. Primers were designed to amplify 1.3 kb region on the 1,2RT-HP plasmid vector above. 1.2kb

PAGE 56

45 Table 3-1. Results of Agrobacterium-mediated transformation of C. paradisi with the chalcone synthase CHS-S construct, followed by GUS histochemical assay Exp. DateTotal number of segmentsTotal number of shoots producedPercent of shoots producedNumber of GUS positive shootsPercent of shoots GUS positive1/24/200014081258.88%64.80%2/17/20001104534.80%23.77%2/21/2000672416.10%37.32%3/7/2000896424.69%24.76%3/12/2000768293.78%00.00%3/15/2000368277.34%00.00%3/17/2000912232.52%00.00%3/22/200051250.98%00.00%4/3/20001600422.63%00.00%4/12/2000680202.94%00.00%4/14/2000544397.17%00.00%4/21/20007687710.03%00.00%5/5/2000706314.39%00.00%5/16/200060860.99%00.00%5/17/2000672152.23%00.00%Total122185754.71%132.26%Total vector control46241513.27%2443% The number of surviving CHS-S transgenic plants in the greenhouse 3

PAGE 57

46 Table 3-2. Results of Agrobacterium-mediated transformation of C. paradisi with the chalcone synthase CHS-AS construct, followed by GUS histochemical assay. Exp. DateTotal number of segmentsTotal number of shoots producedPercent of shoots producedNumber of GUS positive shootsPercent of shoots GUS positive5/2/199725662.34%466.67%5/24/199751230.59%133.33%5/31/199741092.20%111.11%6/7/199730651.63%360.00%10/14/199756030.54%133.33%10/28/1997656365.49%822.22%10/21/199760817228.29%8348.26%11/13/199770400.00%00.00%11/18/199749600.00%00.00%12/2/199711221.79%150.00%12/16/199727200.00%00.00%4/15/1998592335.57%412.12%4/22/1998418337.89%13.03%6/15/1998608345.59%38.82%7/20/1998663304.52%413.33%7/21/199838420.52%150.00%7/22/199840030.75%00.00%7/1/1999816121.47%18.33%7/7/199975281.06%10.00%7/22/19991056363.41%140.00%Total5319236.736584.45%4.2056421.03%Total vector control46241513.27%2443% The number of surviving CHS-AS transgenic plants in the greenhouse 12

PAGE 58

47 Table 3-3. Results of Agrobacterium-mediated transformation of C. paradisi with the chalcone isomerase CHI-S construct, followed by GUS histochemical assay. Exp. DateTotal number of segmentsTotal number of shoots producedPercent of shoots producedNumber of GUS positive shootsPercent of shoots GUS positive7/7/199748081.67%562.50%7/8/1997500234.60%1252.17%10/6/199986412013.89%2621.67%10/8/1999544325.88%721.88%10/20/19992566826.56%3044.11%10/29/1999384318.07%1341.94%10/14/2000576335.73%618.18%1/21/2000124216713.45%2414.37%Total386645111.67%12334.60%Total vector control46241513.27%2443% The number of surviving CHI-S transgenic plants in the greenhouse is 21 Table 3-4. Results of Agrobacterium-mediated transformation of C. paradisi with the chalcone isomerase CHI-AS construct, followed by GUS histochemical assay Exp. DateTotal number of segmentsTotal number of shoots producedPercent of shoots producedNumber of GUS positive shootsPercent of shoots GUS positive7/16/1997624203.21%945.00%7/23/1997192136.77%430.77%8/19/199793621723.18%9342.86%3/12/19983843910.16%923.08%3/13/19981152746.42%1618.18%3/13/199814410.69%1100.00%3/17/199825651.95%240.00%3/20/199838410.26%00.00%3/24/199812832.34%266.67%4/1/1998188189.57%15.56%4/6/199822410.45%00.00%4/8/199819221.04%150.00%6/16/1998464357.54%25.71%7/13/1998736385.16%513.16%7/14/199844461.35%116.67%8/7/19986049615.89%3334.38%Total70525698.07%17930.75%Total vector control46241513.27%2443% The number of surviving CHI-AS transgenic plants in the greenhouse is 17

PAGE 59

48 Table 3-5. Results of Agrobacterium-mediated transformation of C. paradisi with the 1,2 rhamnosyl transferase hairpin (1,2RT-HP) construct, followed by PCR analysis Exp. DateTotal number of segments Total number of shoots producedPercent shoots producedNumber of NPTII positivePercent of NPTII positiveNumber of RT 1,2 (transgene) positive 2/1/2001160050.31%120.00%02/6/20011504151.00%00.00%02/15/2001512112.15%218.18%03/15/20014806814.17%1319.12%23/20/2001896343.79%514.71%13/29/20011040242.31%520.83%04/5/20011104353.17%720.00%14/26/20012112180.85%633.33%3Total92482102.27%3918.57%7Total vector control100435134.96%6819.37% The number of surviving 1,2RT-HP transgenic plants in the greenhouse is 3 Table 3-6. Results of Agrobacterium-mediated transformation of C. paradisi with the pME 1,2 rhamnosyl transferase (1,2RT-S) construct, followed by PCR analysis Exp. DateTotal number of segments Total number of shootsPercent shoots Number of NPTII positive plants Percent NPTII positiveNumber of RT 1(transgene) positive plants03/20/021248211.7%523.8%3 The number of surviving 1,2RT-S transgenic plants in the greenhouse is 1 Table 3-7. Results of Agrobacterium-mediated transformation of C. paradisi with the 1,2 rhamnosyl transferase pBINplus sense construct followed by PCR analysis. Exp. DateTotal number of segments Total number of shoots producedPercent shoots producedNumber of NPTII positive plant pools*2/25/200350411623.02%52/27/200340430274.75%123/14/2003102053852.75%15Total192895649.59%32 *Small leaf tissues of 10 plant shoots were combined together, which is called a pool and DNA was extracted from the 10 plant altogether for the PCR analysis. The plants were not analyzed further due to the insufficient amount of tissue. They will be analyzed in the future.

PAGE 60

49 A) 1,2RT-HP 52 1,2RT –HP 38 Control B) 1,2RT-S 17 Soil: 7-29-02 1,2RT-HP 61 Soil:7-8-01 1,2RT-HP 52 Soil:7-8-01 1,2RT-HP 38 Soil: 7-8-01 Nontransformed Control. Soil:7-29-01 Figure 3-6. Morphological comparison of 1,2 RT plants to control plants. Picture A was taken on 4-10-02, note that all three plants were transferred to the soil at almost the same time. Plant pictures in B were taken individually on the same day (4-24-04) and they were grouped for relative comparison. The label shows the transfer date of the plants to soil. It is clearly seen that the control plant is taller than the transgenics. All the plants were planted in same size pots with the exception of plant 17 which is in slightly smaller pot.

PAGE 61

CHAPTER 4 CHARACTERIZATION OF TRANSGENIC PLANTS Introduction The flavonoid pathway is the most highly studied biochemical pathway in the plant kingdom. The early genes in the pathway have been cloned and exploited to alter the products of the pathway in various plant species (Elomaa et al. 1994; Koes et al. 1994; Winkel-Shirley 2001b; Forkman et al. 2001). Flavanones are the most abundant flavonoids in citrus (Horowitz and Gentili 1977). They tend to accumulate in a glycoside form, in which a disaccharide is attached to the aglycone through the C-7 hydroxyl group. The disaccharide can be a tasteless -rutinose, or a bitter -neohesperidose (Horowitz and Gentili 1977). Naringin, one of the flavanone neohesperidosides, is a highly abundant bitter compound, which reduces the acceptability of grapefruit by consumers (Bar-Peled 1991, Drenowski 1997; 2000). Our goal is to manipulate the flavonoid biosynthetic pathway to alter the levels and types of flavanones in citrus fruits for improving their palatability and/or pharmacological properties. We have chosen several genes from the flavonoid pathway to explore strategies that have the potential to accomplish our aim. To begin the initial pilot experiments, the CHS and CHI cDNAs were cloned from Marsh grapefruit in this laboratory (Luth 1997). Later, two family members of CHS gene (CitCHS1AB009350 and CitCHS2AB009351) were cloned from citrus species by another group (Moriguchi et al. 1999). The nucleotide sequences of these two CHS genes are significantly different from each other. CitCHS2 is almost (98%) identical to the CHS that was previously cloned by Luth (1997). Therefore, to prevent complications in the future, in the text, 50

PAGE 62

51 CHS2 refers to original CHS, which was cloned and utilized by Luth. In the present experiments constructs that contained the sense and antisense cDNAs of CHS2 and CHI genes were utilized in grapefruit transformation experiments in order to suppress or overexpress the target genes. The number of experiments and produced plants in the original experiments by Luth were too small for further analysis, although use of the CHI antisense construct appeared to cause a significant reduction in the naringin level in at least one plant, which did not survive (Luth 1997). To accomplish our main objective we produced more transgenic plants by utilizing the previously prepared CHS2 and CHI sense / antisense constructs. In addition, a sense and a hairpin forming construct of the 1,2 rhamnosyl transferase (1,2 RT) gene, which catalyses the last biochemical step in the production of naringin, were used in genetic transformation experiments. The hairpin-forming construct expresses dsRNA, which should interfere with the target gene expression and decrease or eliminate it efficiently. Regenerated plants following Agrobacterium-mediated transfer of the constructs were analyzed for evidence of transformation by using GUS and/or PCR techniques. Transgenic plants were characterized for their transgene copy numbers via Southern blot analysis and for their steady state RNA levels via relative quantitative RT-PCR in order to estimate the degree of gene transcription relative to 18S ribosomal RNA transcription. Furthermore, they were analyzed by HPLC for their naringin level and flavonoid composition over time. Materials and Methods Southern Blot Analysis Leaf tissues were collected from transgenic and control plants grown in the greenhouse. Leaves were washed to eliminate any residues that might have come from their growth environment. The tissues were ground in the presence of liquid nitrogen

PAGE 63

52 using a mortar and pestle. Genomic DNA was extracted from roughly 250 mg of ground tissue by using 800 l DNAzolES reagent (Molecular Research Center, Inc.). The manufacturer’s instructions were followed to extract DNA. At the last step, the DNA pellet was dissolved in 30 l TE buffer. The DNA concentration was estimated by a spectrophotometer measuring sample absorbance at 260 nm. Additionally DNA samples were electrophoresed on a 1% agarose gel to make sure they were not degraded and were of good quality. To digest the DNAs from CHI and CHS sense/antisense transformed plants, Hpa I (Promega) was selected. This enzyme does not have a site in either of the genes and it cuts only once in the pGA482 T-DNA (Figure 4.3). On the other hand, EcoR I and Sst I were selected to digest the DNAs from transgenic plants transformed with 1,2RT constructs and wild type control citrus plant. Genomic DNA (15 g) was digested with 4 l (10 U/l) selected enzyme in a 50 l reaction containing manufacturer’s buffer and 1 l RNAse. Digestion was conducted overnight at 37oC which is the temperature that all of the enzymes utilized here were 100% active. The digested DNA samples were electrophoresed for approximately 12 hour in a 1% agarose TAE (40 mM Tris-acetate pH 7.6, 1 mM Na2EDTA) gel in the presence of 1X TAE buffer. The gel was then stained in 0.25g/ml ethidium bromide for 10 minutes and the DNAs were visualized with a UV transilluminator. Before transferring the DNAs to a membrane, the gel was incubated in depurination (0.25 M HCL for 10 minutes), in denaturation (0.5 N NaOH, 1.5 M NaCl, 15 minutes 2 times ), and in neutralization (0.5 M Tris-HCL pH 7.5, 3 M NaCl, 15 minutes 2 times) solutions respectively. Finally, the gel was incubated at least 10 minutes in 20X SSC and the DNAs were transferred to a positively charged nylon membrane in the presence of 20X SSC. Capillary transfer lasted approximately 20 hours. At the end of the transfer, the DNAs were immobilized onto the membrane by crosslinking in a UV Stratalinker 1800 at 120 mJ (Stratagene).

PAGE 64

53 Labeled probes were prepared by PCR amplification in the presence of Dig-l1-dUTP (Roche Molecular Biochemicals). Gene transfer plasmids were used as templates to amplify the target. The PCR reaction conditions were as follows: 2.5 mM MgCl2, 1X PCR buffer (10 mM Tris-HCL pH 9.0, 50 mM KCL and 0.1 % Triton X-100), dNTP mix (200 M dATP, 200 M dGTP, 200 M dCTP, 130 M dTTP, 70 M DIG-11-UTP) and 200 nM each forward and reverse primers. NPTII primers, forward 5’-TCACTGAAGCGGGAAGGGACT-3’and reverse 5’-CATCGCCATGGGTCACGACGA-3’ were used in the PCR reaction to prepare a 300 bp NPTII probe. 1,2 RT primers forward 5’-GCATTTGAAGACGCAAAACCT-3’ and reverse 5’-ATACACGACTGAACGAGGCTC-3’were utilized to make a 500 bp 1,2 RT probe. GUS primers forward 5’-CAACGAACTGAACTGGCAG-3’ and reverse 5’-CATCACCACGCTTGGGTG-3’ were used to prepare a 780 bp GUS probe. At the end of the PCR reaction, 3 l PCR product was electrophoresed on a 2 % agarose gel to confirm the labeling efficiency. Incorporation of DIG-11-UTP resulted in a slow migration rate compared to unlabeled PCR product. The rest of the total PCR product was electrophoresed, and bands were cut from the gel and purified with Qiagen Gel Purification kit. The manufacturer’s procedure was followed, and at the end of the procedure the product was eluted with 30 l elution buffer to obtain a concentrated probe. Reconstituted DIG Easy-Hyb granules (Roche Molecular Biochemicals) were preheated to 42 oC prior to usage for both prehybridization and hybridization. Prehybridization was performed for 1 hour with 10 ml/100 cm2 DIG Easy-Hyb solution. Meanwhile the probe was denatured in boiling water for 7 minutes and was added (2 l/ml) to preheated DIG Easy-Hyb which was immediately poured onto the membrane. Hybridization was performed for 16-18 hours. The next day the membrane was washed twice for 5 minutes using low stringency (2X SSC, 0.1% SDS) conditions. A second wash was done twice for

PAGE 65

54 15 minutes at 65oC under high stringency (0.5X SSC, 0.1% SDS) conditions. The manufacturer’s procedure was followed for detection (Roche Diagnostics). CSPD (disodium 3-[-4-methoxypiro {1,2-dioxetane-3,2’-[5’-chloro]tricycle[3.3.1.13,7]decan}-4-yl]phenyl phosphate) ready to use was utilized. CSPD is a chemiluminescent substrate for alkaline phosphatase that enables detection of biomolecules by producing light which is captured with a film or with an imaging system. Upon addition of CSPD, the membrane was incubated at 37 oC for 10 minutes to enhance the luminescent reaction. Signal was detected after exposing the membrane to X-ray (Kodak) film for at least 1 hour. Relative Quantitative Reverse Transcription-Polymerase Chain Reaction Steady state RNA levels of target genes in transgenic plants were estimated with the relative quantitative reverse transcription-polymerase chain reaction (RT-PCR) method. Sample Collection: Leaf samples were collected in the morning from 10.00 am to 11.30 am. At each collection time, three to four fully expanded similar-size leaf samples were collected from each greenhouse grown transgenic and control plant to be tested. The leaves were cut vertically into two equal halves with a clean blade. One half of the leaves was dried for HPLC analysis and the other half was frozen in liquid nitrogen immediately. Frozen samples were kept at -80 oC until they were utilized for RNA extraction. RNA extraction: RNA extraction was performed according to the manufacturer’s protocol using Trizol (Gibco-BRL). Approximately 200 mg ground leaf tissue was utilized. At the last step of the extraction, the RNA pellet was dissolved in 30 l DEPC water. The concentration of total RNA was estimated by measuring absorbance at 260 nm with a UV spectrophotometer. Additionally, 1 l of each RNA sample was

PAGE 66

55 electrophoresed in an ethidium bromide stained 2% agarose gel to make sure that it was not degraded and was of good quality. Total RNA was treated with DNAse I (Ambion) to remove residual DNA contamination prior to the reverse transcriptase reaction. RT-PCR: All reaction components for relative quantitative RT-PCR were purchased from Ambion and the manufacturer’s instructions were followed to perform the procedure. Total RNA (2.5 g) was utilized, along with 2 l random primers (50 M), 4 l dNTP mix (2.5 mM each), 100 units Moloney Murine Leukemia Virus reverse transcriptase (M-MLV RT) enzyme and manufacturer’s buffer in a total of 20 l reverse transcriptase reaction, which was performed at 42 oC for 1 hour. Interactions of the reaction components give the correct molarity. The produced cDNAs from selected samples were amplified with a pair of gene specific primers as was a control band produced by Universal 18S Internal rRNA primers and Competimers mix (Ambion). Although 18S rRNA is the best internal control since its level does not change much from tissue to tissue or between cell types, the abundance of the 18S rRNA limits its application. Competimers are also 18S rRNA but they are modified at their 3’ end to block extension during the PCR reaction. Mixing 18S primers with Competimers reduces the overall amplification efficiency of 18S, thus producing a better control in the reaction. Determination of exponential phase: For each target gene, the best cycle number was determined by selecting the cycle when the product accumulation was still at the exponential phase (Figure 4.1). In other words, the best cycle is one of the cycle numbers before the product accumulation reaches its plateau. For each gene, two cDNA samples were selected. One of the cDNAs had the highest naringin level and the second one was the vector control. The rationale was that, if transgenic plants that had a low naringin level were selected and used to determine the cycle number, high expressing samples would pass the plateau, which would make comparisons impossible. However, some of

PAGE 67

56 the samples that have low naringin levels were also tested in order to have an idea of the range of the exponential phase to select an optimum cycle number. Regular PCR reactions were conducted for each specific gene. The annealing temperature was previously determined according to the primer pairs’ Tm. The PCR program to determine the exponential phase for target CHS2 cDNA was performed as follows: 94 oC 2 minutes, 36 cycles of 94 oC 30 seconds, 58 oC 30 seconds, 72 oC 1 minute, and final extension at 72 oC for 10 minutes; for target CHI cDNA was performed as follows: 94 oC 2 minutes, 36 cycles of 94 oC 30 seconds, 59 oC 30 seconds, 72 oC 1 minute, and final extension at 72 oC C for 10 minutes; for 1,2 rhamnosyl transferase cDNA was performed as follows: 94 oC 2 minutes, 36 cycles of 94 oC 30 seconds, 59 oC 30 seconds, 72 oC 30 seconds, and final extension at 72 oC for 10 minutes. One l cDNA was used as a template with 1ng gene specific primer pairs (CHS2 forward 5’-AGTGTCAACCAGGCTGATTATC-3’and reverse 5’-GAATCAAGATGAGTATCGGCAG-3’; CHI forward 5’-GGTCGAGAACGTCACTTTCAC-3’and reverse 5’-CTTTGCTGCAGGAGAAACACC-3’; 1,2 RT forward 5’-GCATTTGAAGACGCAAAACCT-3’ and reverse 5’-ATACACGACTGAACGAGGCTC-3’) in a 50 l reaction. The PCR master mixture was distributed to 9 identical tubes and they were placed into a PCR machine. At the end of cycle 20, 22, 24, 26, 28, 30 32, 34 and 36 one tube was taken from the PCR machine (Figure 4-1). PCR products (20 l) were electrophoresed into 2% agarose gel to compare and determine the exponential phase. The best cycle number determined was 30 for CHS2, it was 28 for CHI (Figure 4.1), and it was 27 for 1,2 RT. Determination of 18S primers: Competimers ratio: This ratio was also determined for each specific gene. A regular PCR reaction was conducted with determined cycles for each specific gene. Increasing 18S: Competimer ratios (1:9, 2:8, 3:7) were added to the

PAGE 68

57 reaction in separate tubes. Two control tubes, one containing only the gene specific primer pair and the other one containing only 18S: Competimers (3:7) were placed in the PCR apparatus along with the other tubes. PCR products (20 l) were electrophoresed in a 2% agarose gel (Figure 4-2). The best ratio was where the gene specific primer band intensity was equal to the 18S band intensity for the same cDNA sample. Transgenic lines that have the lowest and the highest naringin levels (according to the HPLC analysis data) and controls were selected. These selected cDNA samples from each group of transgenic plants were amplified in a total of 50 l PCR reactions, which contain optimized 18S: Competimers to quantify the steady state RNA level. PCR products (20 l) were electrophoresed in a 2 % agarose gel. After the gel was stained in ethidium bromide and the picture was taken for confirmation, bands were transferred to a nylon membrane for approximately 16 hours. The chemiluminescent technique was used for more sensitive detection of the bands. Steady state RNA level was measured with the VersaDoc 1000 imaging system (Bio-Rad). Graphics were drawn by obtaining gene specific bands’ signal value and normalizing them with 18S rRNA bands’ signal values. Resulting values were presented as relative units or folds. 20 22 24 26 28 30 32 34 36 20 22 24 26 28 30 32 34 36 PCR cycle numbers 600bp 500b p CHI Figure 4-1. Estimation of the exponential phase for CHI transgenic plants. cDNA samples of # 63 and # 45 (2-20) were utilized in the PCR reaction. Cycle 28 was selected for further PCR reactions.

PAGE 69

58 Figure 4-2. Estimation of 18S ribosomal RNA : Competimers ratio for CHI gene. cDNA of sample # 45 (2-20) was utilized in the PCR reaction. The best ratio was 2:8, where the CHI and 18S band intensities are similar. 18S CHI 1:9 2:8 3:7 HPLC Analysis The sample collection for the HPLC analysis was described above under the RT-PCR section. Collected leaf samples were dried in a 70oC oven for 3 days (Luth 1997). Dried samples were sent to our colleague Dr. Mark Berhow for further HPLC analysis (Berhow et al. 1998). Flavonoid extraction was conducted by him. Dried samples were crumbled in vials. Approximately 250 mgs was removed and weighed. Extraction buffer (3 ml) (80% MeOH, 20% DMSO) was added to the vial. The samples were sonicated at 50oC for 15 minutes, allowed to stand at room temperature for 3 days, and then sonicated again for 5 minutes. The liquid was filtered through a 0.45 micron nylon filter (Alltech Associates, Deerfield, IL) into 1.5 ml autoinjector vials. Samples were analyzed with a binary gradient on a Shimadzu (Columbia, MD) 10A system (SIL10A auto injector, two LC10AT pumps, CTO-10A column oven, SCL10A controller) with a Hewlett Packard 1040A diode-array detector (running under Chemstation A.02.05 software). The column was an Inertsil C18 reverse phase 4.6 mm x 250 mm column (5 micron, ODS 3, Varian, Torrance, CA). The initial column conditions were 20% 0.01 M phosphoric acid and 85% methanol at a flow rate of 1 ml/minute with the column oven set at 40 oC . The effluent was monitored at 285 nm. After 2 minutes at the initial conditions, the column was developed with a linear gradient to 100% methanol over 50 minutes. Standard curves 2:8

PAGE 70

59 were prepared from stock solutions prepared from pure naringin for the flavanones and rhoifolin for flavones. Peaks were identified by characteristic spectra and relative retention times from comparison to standards and/or LC-MS analysis. Results and Discussion Southern Blot Analyses Transgenic plants were previously confirmed by GUS histochemical tests and/or PCR for their transgene. Since the genes of interest were already present endogenously in grapefruit, GUS and NPTII gene primers were used in PCR reactions to verify that the plants were stably transformed. Then, PCR positive plants were analyzed by Southern blot analysis to estimate transgene copy number. Analysis of Chalcone Synthase and Chalcone Isomerase Transgenic Plants In pGA482GG, the transgene of interest was inserted between the NPTII and GUS genes. Therefore, NPTII (300 bp) and GUS (780bp) probes were selected to estimate the transgene number of CHS S/AS and CHI S/AS transgenic plants. Hpa I was chosen as a restriction enzyme, because it cuts the T-DNA once in between the borders. Moreover, neither the CHS nor CHI cDNAs have a site for this enzyme. As a result, hybridizing bands show the transferred T-DNA copy numbers and those of endogenous genes. Southern blot analysis was completed on a total of 47 transgenic plants; 13 CHS transgenic plants and 34 CHI transgenic plants (Table 4-1). Only the pictures of a few Southern blots (Figure 4-3, 4-4, 4-5) are displayed in this dissertation as examples. Probing the blots with both genes indicated that 1-3 copies of the transgenes were integrated into most of the transgenic plants. Just a few plants displayed multiple bands for NPTII integration into transgenic plants. The number of bands for the NPTII gene and the GUS gene were the same in most cases. Since the target gene was placed in between GUS and NPTII gene, having same number of gus and nptII bands reflects that T-DNA

PAGE 71

60 was transferred inside plant cell. A few plants displayed different numbers of bands for the NPTII and the GUS genes (Table 4-1). This may be due to the partial transfer of the T-DNA. On the other hand, in the case of plant 17, the band probed with GUS, which is an intense band, might in fact be two bands of the same size. Transgenic plants 24 and 33 (Figure 4-3 A and B) did not display significant bands when the blot was probed with GUS. These two plants showed chimeric staining in the GUS assay. Although the bands in Figure 4-4 A, B and C looked like doublets, they are most probably thick single bands. Comparison of the same transgenic plants such as plant 31 and plant 56 in other Southern blots hybridized with the same probes (e.g. Figure 4-3 A and B) confirms that these double looking bands are single bands. The three Southern blots in Figure 4-4 were generated by stripping and reprobing the same DNA membrane; therefore the bands looked like doublets in all three blots. This might have been due to a problem during the capillary transfer of the DNA to the membrane. Table 4.1 was generated by counting the visual bands on the DNA membranes hybridized with GUS and NPTII probes. For some of the transgenic plants, copy numbers were not estimated, because either the plant died too soon and/or the tissue was not adequate to extract DNA or the plant was chimeric for the GUS or NPTII and the probe did not find the target. Multiple indicates a number of bands more than three.

PAGE 72

61 HpaI 24 56 57 26 27 28 29 Cont 31 32 33 DIG VIII CHS-AS CHS-AS CHS-S A) Probe: NPTII 857674276106489936392799HpaI Transgene Nos NPTII 35S GUS RB LB 8576742761064899 36392799HpaI 3 24 56 57 26 27 28 29 Cont 31 32 3 DIG VIII CHS-AS CHS-AS CHS-S B) Probe: GUS Figure 4-3. Southern blot analyses of CHS transgenic plants. Genomic DNAs (15 g) were digested with Hpa I which does not have a site in the gene and cuts the plasmid once in the T-DNA. Hybridized with A) NPTII (300 bp) probe. B) GUS (780 bp) probe. Cont. is a negative control DNA which is extracted from a non-transformed grapefruit plant.

PAGE 73

62 HpaI 35S-GUS Transgene Nos-NPTII RB LB HpaI HpaI 24 27 31 33 36 56 Cont58 CHS-AS CHS-S CHS-AS CHS-S 24 27 31 33 36 56 Cont 58 CHS-AS CHS-S CHS-AS CHS-S A) Probe: CHS B) Probe: NPTII HpaI 24 27 31 33 36 56 Cont 58 CHS-AS CHS-S CHS-AS CHS-S Figure 4-4. Southern blot analyses of CHS transgenic plants. Genomic DNAs (15 g) were digested with HpaI which does not have a site in the gene and cuts once in the T-DNA. Hybridized with A) CHS2 (500 bp) probe. B) NPTII (300 bp) probe. C) GUS (780 bp) probe. Cont. is a negative control DNA which is extracted from a non-transformed grapefruit plant. C) Probe: GUS

PAGE 74

63 HpaI 35S-GUS Transgene Nos-NPTII RB LB HpaI 16 17 19 Cont 43 * CHI-S CHI-AS HpaI * 16 17 19 Cont 43 CHI-AS CHI-S B) Probe: NPTII A) Probe: CHI 16 17 19 Cont 43 * CHI-S * HpaI CHI-AS C) Probe: GUS Figure 4-5. Southern blot analyses of CHI transgenic plants. Genomic DNAs (15 g) were digested with HpaI, which does not have a site in the gene and cuts once in T-DNA. Hybridized with A) CHI (600 bp) probe and B) NPTII (300 bp) probe. C) GUS (780 bp) probe. Cont. is a negative control DNA which is extracted from a nontransformed grapefruit plant. There is an insufficient amount of DNA in lane 5 (*) that did not hybridize with any of the probe.

PAGE 75

64 Table 4-1. Southern blot analysis of CHS and CHI transgenic plants. Construct Plant Identity NPT II Probe (300bp) Estimated Copy Number GUS Probe (780bp) Estimated Copy number CHS-S 24 1 1 CHS-S 56 1 1 CHS-S 57 1 1 CHS-AS 26 1 2 CHS-AS 27 2 2 CHS-AS 28 1 2 CHS-AS 29 1 1 CHS-AS 31 1 1 CHS-AS 32 2 2 CHS-AS 33 2 1 CHS-AS 34 1 NE* CHS-AS 36 2 2 CHS-AS 37 1 2 CHS-AS 58 1 2 CHI-S 2 2 NE* Dead CHI-S 8 1 1 Dead CHI-S 38 1 1 CHI-S 39 2 2 CHI-S 40 1 1 CHI-S 41 2 2 CHI-S 43 1 1 CHI-S 45 3 3 CHI-S 46 1 1 CHI-S 47 3 3 CHI-S 48 1 1 CHI-S 49 1 1 CHI-S 50 1 1 CHI-S 51 1 1 CHI-S 53 1 1 CHI-S 54 1 1 CHI-S 55 1 1 CHI-AS 5 3 3 CHI-AS 6 1 2 CHI-AS 9 1 NE* Dead CHI-AS 10 2 1 CHI-AS 11 3 NE* CHI-AS 12 3 NE* CHI-AS 14 1 1 CHI-AS 15 multiple 2 CHI-AS 16 3 3 Dead CHI-AS 17 2 1 CHI-AS 19 3 3 CHI-AS 20 NE* 3 CHI-AS 62 1 2 CHI-AS 63 2 2 *Copy numbers of genes were not estimated (NE).

PAGE 76

65 Analysis of 1,2 Rhamnosyl Transferase Transgenic Plants Given that the 1,2 RT transfer plasmid does not have a GUS gene, only the NPTII probe was used to hybridize to the DNA blot. A 1,2 RT (500bp) gene specific probe was also used in DNA blot analysis to have more information about the transgene copy numbers. Sst I was selected as a restriction enzyme because it does not have a site in the cDNA, whereas it only cuts once in the T-DNA. Most of the PCR positive plants for the NPTII gene died before they were analyzed for their RT transgene copy number. Surviving NPTII positive plants were PCR amplified with a primer set which consists of 35S promoter forward and 1,2RT gene reverse primers to check for the intact hairpin construct. Only 4 plants (1,2RT 38, 52, 61, and 69 (dead) were positive for that primer set. This might be because the hairpin construct is not very stable; it can easily recombine itself (Febres V, personal communication). It has been shown that hairpin producing constructs that do not have a spacer between sense and antisense fragments were not affective to lead silencing as the constructs that have spacers. The PCR positive NPTII plant Southern blot membranes were probed with 1,2RT and NPTII cDNA probes (Figure 4-6 A and B). Probing the DNA membrane with 1,2 RT demonstrated up to 7 copy numbers, which represent transgene and endogenous gene copy number. Nontransformed control plants displayed more than one band. There are probably other glucosyl and rhamnosyl transferase genes that have high homology. Plant 17 demonstrated doublet band although it looks very dark on the picture. DNA amount which is loaded to the gel might be much higher than the estimated amount, which may due to the misreading with the spectrophotometer. Plants 38, 53 displayed 2 and plant 61 displayed 3 copies of the NPTII gene.

PAGE 77

66 Table 4-2. Southern blot analysis of 1,2 RT transgenic plants. Construct Plant Identity NPT II Probe (300pb) Estimated Copy Number 1,2RTProbe (500bp) Estimated Copy Number 1,2RT-S 17 2 NE 1,2RT-HP 38 2 6 1,2RT-HP 52 2 6 1,2RT-HP 61 3 7

PAGE 78

67 EcoRI A) 1,2 RT-HP transgenics 38 Cont 52 Cont 61 1315148285767427610648993639279919521882A) Probe: 1,2 RT 35S prom. SstI Nos-NPTII 35S term. RB LB 1,2RTantisense 1,2RT-Sense 17 38 52 61 1,RT-S 1,2 RT-HP SstI B) Probe: NPTII Figure 4-6. Southern blot analyses of 1,2 RT transgenic plants. Genomic DNAs (15 g) were digested with A) EcoR I which does not have a site in the gene, and B) Sst I which cuts the plasmid once, but does not cut the target gene. Digested DNAs were electrophoresed and transferred to a membrane. Hybridized with A) 1,2 RT (500 bp) probe. B) NPTII (300 bp) probe. Cont. is a negative control DNA which is extracted from a nontransformed grapefruit plant.

PAGE 79

68 Relative Quantitative Reverse Transcription-Polymerase Chain Reaction Relative quantitative RT-PCR was completed on selected transgenic plants at four different time points (Figures 4-7, 4-8, 4-9). Gene specific primers were selected to amplify transferred genes CHS2, CHI and 1,2 RT to have information about the transcripts of both transferred and endogenous genes in each transgenic plant. Expression levels of each gene were normalized with the 18S rRNA expression level to estimate the expression of the combined endogenous and transgenes and compare them with the vector control (BL2, BL5, P2, P3) and nontransformed plants (C6). CHS Plants: The plants 24, 56 and 57 were transformed with the CHS-S construct, and the plants 32, 33, 34 and 37 were transformed with the CHS-AS construct (Figure 4-7). We do not have data for the plant 32 on 10-23-03 (Figure 4-7 A) and the plants 24, 56 and 57 on 12-19-03 (Figure 4-7 B), due to the insufficient amount of tissue for RNA extraction. Gene expression levels decreased on 12-19-03 for the analyzed plants. This might due to the physiological state of the plants and environmental factors. According to our observations plants are not as productive in the months of November and December; although they do not shed their leaves, they rarely produce fresh leaves. Since the flavonoids are produced primarily during exponential growth, during winter probably the genes were not expressed as much as in other seasons. When the plants began to produce more fresh leaves starting in January, CHS2 expression also began to increase. Although there is some fluctuation between time points, each plant maintained a relatively consistent expression profile. For example, plant 24 always displayed a lower expression profile than the other CHS-S plants 56 and 57 at each time point. Plant 37 consistently had a low expression level, except for the first time point, compared to plants 32, 33 and 34. Vector control plant BL2 had a consistently lower expression level than BL5. Control plant C6 had lower transcript level compared to CHS sense and antisense transgenic

PAGE 80

69 plants (Figure 4-7). The transcript level of C6 was either close (Figure 4-8 a, c) or lower (Figure 4-8 d) than the CHI transgenic plants. CHI plants: Plants 45 and 47 were transformed with the CHI-S construct, and plants17, 19, 62 and 63 were transformed with the CHI-AS construct (Figure 4-8). The data from plant 19 could not be obtained on the 2-20-04 (Figure 4-8 D) collection time point due to insufficient amount of tissue for RNA extraction. Expression levels of the CHI gene for each plant were significantly reduced in December, and begun to increase again in January. This was not surprising, because flavonoids are produced primarily during the exponential growth of the plant (Jourdan et al. 1985), and growth of the plants was very slow during the winter time. It was almost impossible to find fully expanded but not old leaves for the analyses. Flavonoids genes are environmentally regulated. Expression of some of the genes might be low during the winter time due to short daylight or less light and cold. Despite the changes in expression levels, the expression profiles did not change for most of the plants at different time points. For instance, plant 47 always had lower expression levels than plant 45, except at the first time point; plant 17 consistently had the lowest expression levels among the transgenic plants. Plant 62 almost always had higher expression levels than the plant 63. Vector control plants BL2 and BL5 had similar expression levels at each time points. Control plant C6 almost always had the lowest expression level compared to other analyzed plants. 1,2 RT plants : Plant 17 was transformed with the 1,2 RT sense construct and plants 38, 52 and 61 were transformed with the1,2 RT hairpin construct. Since the plants 17 and 52 were very small there was not enough leaf tissue for the expression analyses respectively on 12-19-03 (Figure 4-9 B) and on 10-23-04 (Figure 4-9 A). Again, overall the transgenic plants displayed similar profiles at different time points, in spite of the changes in expression levels. For example, the sense transgenic plant 17 consistently had the highest

PAGE 81

70 expression level at all of the collection points. The hairpin transgenic plants 38 and 52 consistently had lower expression than plant 61. Expression levels of vector control plants (P2 and P3) fluctuated between time points (Figure 4-9). This might be due to the physiological state of the plants, environmental factors or the poor sample collection and RNA preparation, and will be discussed further below. HPLC Analysis Leaf tissue of all of the surviving transgenic and control plants was collected at sequential time points on 10-23-03, 12-19-03, 1-20-04, and 2-20-04 in the morning from 10.00 am to 11.30 am. The samples were extracted and analyzed by Dr. Berhow for their major flavonoid compounds (Figures 4.10, 4.11, 4.12, 4.13). Some of the transgenic plants had been previously analyzed in different times by HPLC, since they were placed into soil (data not shown). Naringin represents by far the highest component among the major flavonoids. Thus the figures show only naringin levels (mg per grams of sample) of control (Figure 4-10) and groups of transgenic plants (4-11, 4-12, and 4-13) at different collection times. Naringin levels of control plants fluctuated greatly, between 3 and 26 mg/g, although there was no CHS/CHI transgenic plant that had a naringin level over 17 mg/g (Figures 4-11 and 4-12). The three CHS-S transgenic plants displayed similar naringin levels except for plant 56, which showed a spike of naringin content on 1-20-04 (Figure 4-11 A). A corresponding peak in the RT-PCR assay was not observed. Overall HPLC analyses showed that CHS-AS transgenic plants consistently had lower naringin levels compared to CHS-S plants (Figure 4-11 A and B). CHS-AS plants 32, 33, 34 and 37 have the lowest naringin levels among all of the CHS transgenic plants. Their naringin levels did not fluctuate significantly, they were always under 8 mg/g at each time point. These plants had significantly and consistently lower naringin levels than almost all of the control plants (Figure 4-14). CHI-S plants 45 and 47, and CHI-AS plant 17 also

PAGE 82

71 CHS 1-2000.511.522.5324565732333437BL2BL5C6Plant identity number CHS 2-2000.511.522.5324565732333437BL2BL5C6Plant identity number CHS 10-2300.511.522.53245657333437BL2BL5C6Plant identity number CHS 12-1900.511.522.53245657323437BL2BL5C6Plant identity number A) B) Relative Unit (CHS / 18SrRNA) C) D) Figure 4-7. Expression analyses of CHS plants. Plants 24, 56, 57 are CHS-S, 32, 33, 34, 37 are CHS-AS construct transformed. BL2 and BL5 are vector controls, C6 is a nontransformed control, which comes from the tissue culture. Fold increase is calculated relative to 18S rRNA for each sample. Leaf samples were collected on A) 10-23-04, B) 12-19-04, C) 1-20-04. D) 2-20-04.

PAGE 83

72 CHI 10-23055454717196263BL2BL5C6Plant identity number Figure 4-8. Expression analyses of CHI plants. Plants 45, 47 are CHI-S, 17, 19, 62 and 63 are CHI-AS construct transformed lines. BL2 and BL5 are vector controls, C6 is a nontransformed control which comes from the tissue culture. Fold increase is calculated relative to 18S rRNA for each sample. Leaf samples were collected on A) 10-23-04, B) 12-19-04, C) 1-20-04. D) 2-20-04. A) 2 1. 1 0. CHI 12-1900.10.20.30.40.5454717196263BL2BL5C6Plant identity number CHI 1-2000.511.52454717196263BL2BL3C6Plant identity number CHI 2-2000.511.52454717196263BL2BL5C6Plant identity number B) Relative Unit (CHI / 18SrRNA) C) D)

PAGE 84

73 Figure 4-9. Expression analysis of 1,2 RT plants. Plant 17 is 1,2 RTsense, 38,52,and 61 are 1,2 RThp construct transformed. P2 and P3 are vector controls. Fold increase is calculated relative to 18S rRNA for each sample. Leaf samples were collected on A) 10-23-04, B) 12-19-04, C) 1-20-04. D) 2-20-04. A) B) D) C) Relative Unit (1,2RT / 18SrRNA) RT 4-2100.5 1 1.5217385261P2P3Plant identity number RT 2-2000.511.5217385261P2P3Plant identity number RT 10-2300.511.5217385261p2p3Plant identity number RT 12-1900.511.5217385261p2p3Plant identity number

PAGE 85

74 displayed consistently low naringin levels compared to other transgenic plants and controls (Figure s 4.15 and 4.16). CHI-AS 62 and 63 displayed the highest naringin level compared to the other transgenic plants, but they did not have higher levels of naringin than the controls. Only Control Plants0.05.010.015.020.025.0(10/23/03)(12/19/03)(1/20/04)(2/20/04)Sample Collection TimeNrg mg/g C1 C5 C6 C8 C10 C11 BL1 BL2 BL5 Figure 4-10. Naringin levels (mg/g) of control plants analyzed by HPLC. Nontransformed controls (C), and vector controls (BL) came from tissue culture. The 1,2 RT hairpin construct transformed plants did not have naringin levels over 10 mg/g at any time point (Figure 4.17). The sense 1,2 RT transformed plant 17 also had naringin level below 10 mg/g except at one time point (2-20-04) where the level spiked. Plant 17 is two years old, and is significantly dwarfed with small leaves compared to control plants. Although we are careful about the pest level in the greenhouse, plant 17 has severe pest damage on its leaves (Figure 4-18 A). Thus we consistently did not have sufficient plant material of this particular plant for assays. Flavonoids are known to respond to wounding, and we might have collected the wounded leaves for the 2-20-04

PAGE 86

75 time point. The plant 1,2 RT 61 had the lowest level of naringin among the transgenic and control plants, consistently producing a low level of naringin at each time point. CHS-S Transgenic Plants0.05.010.015.020.025.0(10/23/03)(12/19/03)(1/20/04)(2/20/04)Sample Collection TimeNrg mg/g 24 56 57 A) CHS-AS Transgenic Plants0.05.010.015.020.025.0(10/23/03)(12/19/03)(1/20/04)(2/20/04)Sample Collection TimeNrg mg/g 25 26 27 28 29 31 32 33 34 35 36 37 58 B) Figure 4-11. Naringin levels of CHS transgenic plants analyzed by HPLC. A) CHS-S transgenic plants, B) CHS-AS transgenic plants

PAGE 87

76 CHI-S Transgenic Plants0.05.010.015.020.025.0(10/23/03)(12/19/03)(1/20/04)(2/20/04)Sample Collection TimeNrg mg/g 1 3 4 38 39 40 41 42 43 44 45 46 47 48 49 50 51 54 55 A) CHI-AS Transgenic Plants0.05.010.015.020.025.0(10/23/03)(12/19/03)(1/20/04)(2/20/04)Sample Colletion TimeNrg mg/g 5 7 9 10 12 14 15 17 19 23 60 61 62 63 B) Figure 4-12. Naringin levels of CHI transgenic plants analyzed by HPLC. A) CHI-S transgenic plants, B) CHI-AS transgenic plants

PAGE 88

77 CHS-AS Transgenic Plants Compared to Controls0.02.04.06.08.010.012.014.016.018.020.022.024.026.0(10/23/03)(12/19/03)(1/20/04)(2/20/04)Sample Collection TimeNrg (mg/g) 32 33 34 35 37 C1 C5 C6 C8 C10 C11 BL1 BL2 BL5 Figure 4-14. Naringin levels of CHS-AS transgenic plants and controls; nontransformed (C1, C5, C6, C8, C10, and C11) and vector controls (BL1, BL2, BL5) were analyzed by HPLC. CHI-S Transgenic Plants Compared to Control Plants0.02.04.06.08.010.012.014.016.018.020.022.024.026.0(10/23/03)(12/19/03)(1/20/04)(2/20/04)Sample Collection TimeNRG (mg/g) 45 47 C1 C5 C6 C8 C10 C11 BL1 BL2 BL5 Figure 4-15. Naringin levels of CHI-S transgenic plants and controls; nontransformed (C) and vector controls (BL) were analyzed by HPLC. Plant number 45 and 47 have lower naringin levels than the control plants.

PAGE 89

78 CHI-AS Transgenic Plants Compared to Control Plants0.02.04.06.08.010.012.014.016.018.020.022.024.026.0(10/23/03)(12/19/03)(1/20/04)(2/20/04)Sample Collection TimeNrg mg/g 17 62 63 C1 C5 C6 C10 C11 BL1 BL2 BL5 Figure 4-16. Naringin levels of CHI-AS transgenic plants and controls; nontransformed (C) and vector controls (BL) were analyzed by HPLC. Plant number 17 and has lower naringin levels than the controls. 0.02.04.06.08.010.012.014.016.018.020.022.024.0(10/23/03)(12/19/03)(1/20/04)(2/20/04)Sample Collection TimeNrg mg/g 17 38 52 61 p2 (control) p3 (control) 1,2 RT Transgenic Plants Compared to Control Plants 0.02.04.06.08.010.012.014.016.018.020.022.024.0(10/23/03)(12/19/03)(1/20/04)(2/20/04)Sample Collection TimeNrg mg/g 17 38 52 61 1,2 RT Transgenic Plants Figure 4-17. Naringin levels of 1,2 RT transgenic plants analyzed by HPLC. Plant 17 is transformed with the 1,2 RTsense construct, the rest are transformed with 1,2RT-HP construct. P2 and P3 are vector controls. Although only naringin levels are shown above, total flavonoid composition was determined for each plant and time point. In most cases, controls and transgenic plants showed a common profile, illustrated by control plants in Figure 4-19 A and B, although absolute levels fluctuated from time point to time point. However, comparison of the major flavonoid composition of transgenic plant 1,2RT 61 to the vector control plants showed a surprising profile. Overall the level of major flavonoids was significantly low

PAGE 90

79 in the transgenic plant when they were compared to the vector control plants (Figure 4-19). Although as stated above, the levels changed slightly at different time points, the ratio of the major flavonoids was stable in the control plants (Figure 4-19 A and B), whereas the ratio fluctuated in transgenic plants 61 and 17. Specifically, in plant 1,2 RT 61, the narirutin, rhoifolin, and isorhoifolin concentrations fluctuated and were often as great or greater than the naringin concentration. Based on our observations and the literature (Del Rio et al. 2004) transient changes occur in flavonoid composition in citrus plants when they are attacked by fungi or pests, but we observed no significant attack to plant 61. The changes in the flavonoid composition were observed in plant 17 returned to almost the same levels as before in the next analysis (Figure 4-19 C). On the other hand, plant 61 maintained changes for over 3 months. In an effort to find out any correlation between RT-PCR and HPLC results, we plotted the expression level of the target gene and the naringin and narirutin content (mg/g) on an excel graph (data not shown). No direct correlation was observed between naringin level and the target gene transcript levels of the analyzed transgenic plants. Since the enzyme levels were not measured, it is not known that whether the enzyme levels were altered in each of the transgenic plants. It is necessary to measure the levels of enzymes in the transgenic plants to have information about how the flavonoid pathway and the flavonoid compound production were controlled in citrus.

PAGE 91

80 Figure 4-18. Demonstration of the transgenic plant 1,2RT 17. A) Plant 17 is transformed with the 1,2 RT-S construct. B) Plant 17 next to plasmid only transferred control plant that almost same age with the transgenic plant. 15 c m Control 17 A) B) 60 c m

PAGE 92

81 Major Flavonoid Composition of Vector Control pBin20.02.04.06.08.010.012.014.016.018.020.0ApDiGNRTNRGNHSIRFRFNNMNDSPONTotalFlav.Major Flavonoidsmg/g 10/23/2003 12/19/2003 1/20/2004 2/20/2004 Major Flavonoid Composition of Vector Control pBin30.02.04.06.08.010.012.014.016.018.020.0ApDiGNRTNRGNHSIRFRFNNMNDSPONTotalFlav.Major Flavonoidsmg/g 10/23/2003 12/19/2003 1/20/2004 Major Flavonoid Composition of 1,2 RT Transgenic Plant 170.02.04.06.08.010.012.014.016.018.020.0ApDiGNRTNRGNHSIRFRFNNMNDSPONTotalFlav.Major Flavonoidsmg/g 10/23/2003 12/19/2003 1/20/2004 Major Flavonoid Composition of 1,2RT Transgenic Plant 610.02.04.06.08.010.012.014.016.018.020.0ApDiGNRTNRGNHSIRFRFNNMNDSPONTotalFlav.Major Flavonoidsmg/g 10/23/2003 12/19/2004 1/20/2004 2/20/2004 A) B) C) D) Figure 4-19. Flavonoid levels of transgenic and control pants analyzed by HPLC. Major Flavonoid composition of vector controls A) pBin2, B) pBin 3, plants compared to 1,2 RT transgenic plants C) 1,2RT-HP Plant 17 and D) 1,2RT-S Plant 61. NRG (naringin), RF (rhoifolin), NRT (narirutin), IRF (Isorhoifolin), PON (Poncerin).

PAGE 93

CHAPTER 5 CONCLUSIONS Naringin, the most abundant bitter tasting flavonoid compound in grapefruit, not only decreases the consumption of the fruit but also reduces the economical value of the fruit and juice products. The objective of this dissertation was to attempt to reduce the bitter taste of naringin and/or increase the pharmacological value of grapefruit. It has been proven that manipulating the flavonoid pathway in other plant species can alter the expression of secondary compounds. To accomplish our objective, we have investigated several strategies. CHS2 and CHI cDNAs were cloned from grapefruit previously in the Moore Lab. The enzymes of CHS and CHI catalyze the first steps of the flavonoid biosynthetic pathway. In the present study, grapefruit were transformed by utilizing sense and antisense constructs of these genes. Despite the effort to produce as many transgenic plantlets as possible with CHS-S and CHS-AS constructs, many of the plantlets died either just after they were transferred to soil or a short later in the growth chamber. Transformation efficiency for CHS-S was estimated as 2.3%, although for CHS-AS it was much higher (21%). As of August 2004, only three transgenic plants survived from CHS-S transformation experiments, whereas 12 transgenic plants survived from CHS-AS transformation experiments. Transformation efficiency was much higher with the CHI-S (34.6%) and CHI-AS (30.8%) constructs than with the CHS constructs. Thus there does not seem to be any doubt that overexpression, and to a lesser extent, antisense expression, of CHS was deleterious in grapefruit. This was quite unexpected, and, to our knowledge, has not been reported in any other species before. 82

PAGE 94

83 Later in the project, the 1,2 RT cDNA was cloned first from C. maxima (AY048882) then C. paradisi by a colleague and given to us for use in transformation experiments. The high number and the similarity of the glucosyl transferase sequences in citrus, as in other species, hindered the cloning of the rhamnosyl transferase gene. Hairpin and sense constructs of the 1,2 RT gene were transferred to grapefruit seedlings. We obtained many plantlets which were positive for the NPTII gene but not for the intact hairpin construct. This might due to the instability of the hairpin construct. Only a few plants survived which tested positive for the hairpin and sense constructs of 1,2RT gene. All of the transgenic plants were characterized for their transgene copy numbers via Southern blot analysis. The three surviving CHS-S transgenic plants each only have a single copy of the transgene. We take this as further evidence that multiple copies of the transgene would likely be deleterious and so lethal. Analyzed CHS-AS transgenic plants either have one or two copies of the transgene. The majority of the CHI-S transgenic plants have one or two copy whereas nearly all of the CHI-AS transgenic plants have more than one copy. Analyzed 1,2RT transgenic plants mainly have two and one plant (1,2RT61) has three copies of the transgene. Thus none of the surviving transgenic plants possesses large multiples of the transgene. Semi quantitative RT-PCR was selected to measure the steady state RNA level of the target genes to obtain information about the expression levels. Leaf tissues of the transgenic and control plants were collected for four sequential time points (10-23-03, 12-19-03, 1-20-04, and 2-20-04). For each sample half of the leaf samples were frozen immediately for RNA extraction and the other half was utilized for HPLC analysis. Analyzed CHS and CHI plant samples displayed low expression levels in December

PAGE 95

84 before the levels began to increase by January and February. As a matter of fact, the CHI expression dropped to almost zero in the CHI transgenic plants. In research by others (Horowitz and Gentili 1977; Moriguchi et al. 2001) the CHI enzyme was detected when the plants are young or during the exponential growth of the cells; when the plant or tissue became older the enzyme diminished. Flavonoids are also regulated by light. During the months of November and December, the plants in the greenhouse do not produce growth flushes as much as during the other months of a year. Additionally, the plants in the greenhouse do not have light in winters as much as the other months of the year. Probably, these are the reasons why the expression levels of CHI were so low in the plants. The expression profile in individual plants was very similar between time points, although there were some changes in the expression levels; some of the transgenic plants had almost always low expression levels compared to the other transgenics and some always had high. Nontransformed control plant almost always had a low expression level for CHS and CHI genes at different time points. The data suggest that there is no direct correlation between expression level and naringin level, since the naringin levels of these transgenic plants are significantly lower than the control plants. The data might also suggest that there is a post-translational regulation of flavonoid biosynthesis in citrus. Expression profile of 1,2RT transgenic was consistent between collection points. Plant 52 had almost always the lowest expression level compared to other transgenic and the control plants. The plant 17 which was transformed with 1,2RT-S construct always had the highest expression level in each collection points. This particular plant does not look normal either. Although it is a 2 year old plant, it is only 15 cm tall; its leaves are small (~5 cm) compared to control plant leaves (~13 cm). Remarkably the leaves are the

PAGE 96

85 favorite of snails in the greenhouse. 1,2RT gene is a recently cloned gene, and we had the opportunity to utilize its constructs in our Agrobacterium-mediated gene transformation experiments. Therefore there is no literature about the effects of rhamnosyl transferase gene transformation on the plants. This plant will be analyzed further to have more information about the effects of the transformation. All of the transgenic and the control plant samples collected at selected time points were analyzed by HPLC for their major flavonoid levels. Naringin levels of vector and nontransformed control plants fluctuated between 3.5 mg/g and 26 mg/g levels at different collection time points. A major difference between control and transgenic plants was the degree of fluctuation in flavonoid content. Transformation with any construct tended to reduce fluctuation in flavonoid levels and produce a consistent flavonoid level, although this level varied from plant to plant. Thus, the presence of the transgene appeared to have a “buffering” effect on the plant. Flavonoids have different roles in plants including protection against UV. Probably the transgenic plants having more flavonoid gene expression can produce different flavonoids other then major citrus flavonoids, which would be utilized in different stress conditions. Having excess flavonoid gene expression might be an advantage for the transgenic plants over control plants. CHS transgenic plants, specifically CHS-AS plants, had consistently lower levels of naringin compared to controls (Figure 4-11, 4-14). Total flavonoid levels also decreased in these plants. Thus, as measured in leaves in young plants, these experiments look promising for the reduction of naringin. The naringin and total flavonoid levels were reduced and remained consistently reduced over a considerable period of time. There

PAGE 97

86 were a few CHI transgenic plants (CHI-S 45, CHI-S 47 and CHI-AS 17) also had much lower naringin levels than the controls, while the rest of the CHI plants had naringin levels closer to those of the controls. This data might suggest that even though both genes have influences on the flavonoid pathway of citrus, the CHS gene/genes might have a stronger influence on the naringin pathway than CHI. Only a very few plants transgenic for the RT gene were produced, in spite of a large effort with the hairpin construct. Only limited numbers of experiments were completed with the 1,2RT-S construct, and only two transgenic plants were obtained. Therefore we do not have enough data to compare the transformation results of 1,2RT-HP and 1,2RT-S construct and to draw a conclusion. Two 1,2 RT-HP plants (38 and 52) had similar naringin levels to the controls, but plant 61 displayed significantly lower levels of naringin than the controls at each time point. We also analyzed the major flavonoid composition of all of the transgenic plants and the controls. Plants RT–S 17 and RT-HP 61 showed different flavonoid composition than the controls, not just varying flavonoid levels, at different time points. Naringin content was always the highest flavonoid in the controls, whereas in the transgenic plants 17 and 61, narirutin, which is a tasteless compound, became the highest. HPLC analysis showed that the plant 61 consistently kept this altered composition for a few months. We observed that changes in the levels of flavonoids and compositions can occur when they are attacked by pest or fungi but the composition comes back to the original in a short time. However, plant 61 was not attacked by a fungus or pest, and it kept the same profile over an extended period of time. This particular plant might have the altered flavonoid composition with a lower

PAGE 98

87 naringin level as a result of our manipulation in the flavonoid pathway. It will be analyzed further for the effects of transformation. According to our expression and HPLC analysis data no direct correlation was observed between transcript level of the target genes and the naringin levels in grapefruit plants. Although control plants had higher levels of naringin, the transcript levels of the target genes were lower in controls than the analyzed transgenic plants, which had almost equal or higher transcript levels but lower naringin levels. Analysis of enzyme levels in the transgenic plants with regards to control would be helpful to have more information about the regulation of flavonoid pathway in citrus.

PAGE 99

APPENDIX MOLECULAR AND CHROMATOGRAPHIC ANALYSIS OF TRANSGENIC PLANTS List of the Transgenic Plants Analyzed by Molecular Techniques Construct Plant Identity NPT II Probe (300bp) Estimated Copy Number GUS Probe (780bp) Estimated Copy number Expression Analysis HPLC Analysis CHS-S 24 1 1 x x CHS-S 56 1 1 x x CHS-S 57 1 1 x x CHS-AS 26 1 2 x CHS-AS 27 2 2 x CHS-AS 28 1 2 x CHS-AS 29 1 1 x CHS-AS 31 1 1 x CHS-AS 32 2 2 x x CHS-AS 33 2 1 x x CHS-AS 34 1 NE* x x CHS-AS 36 2 2 x CHS-AS 37 1 2 x x CHS-AS 58 1 2 x CHI-S 2 2 NE* Dead CHI-S 8 1 1 Dead CHI-S 38 1 1 x CHI-S 39 2 2 x CHI-S 40 1 1 x CHI-S 41 2 2 x CHI-S 43 1 1 x CHI-S 45 3 3 x x CHI-S 46 1 1 x CHI-S 47 3 3 x x CHI-S 48 1 1 x CHI-S 49 1 1 x CHI-S 50 1 1 x CHI-S 51 1 1 x CHI-S 53 1 1 x CHI-S 54 1 1 x CHI-S 55 1 1 x CHI-AS 5 3 3 x CHI-AS 6 1 2 x 88

PAGE 100

89 CHI-AS 9 1 NE* Dead x CHI-AS 10 2 1 x CHI-AS 11 3 NE* x CHI-AS 12 3 NE* x CHI-AS 14 1 1 x CHI-AS 15 multiple 2 x CHI-AS 16 3 3 Dead x CHI-AS 17 2 1 x x CHI-AS 19 3 3 x x CHI-AS 20 NE* 3 x CHI-AS 62 1 2 x x CHI-AS 63 2 2 x x Construct Plant Identity NPT II Probe (300pb) Estimated Copy Number 1,2RTProbe (500bp) Estimated Copy Number Expression Analysis HPLC Analysis 1,2RT-S 17 2 NE x x 1,2RT-HP 38 2 6 x x 1,2RT-HP 52 2 6 x x 1,2RT-HP 61 3 7 x x

PAGE 101

90 Selected HPLC Chromatograms of Control and Transgenic Plant 1,2RT61 ADCG NRT NRG NHS IRF RFN N6M DID PON NAR Major flavanones in a grapefruit leaf Vector control leaf NRG NRT IRF RFN Modification of NRT and IRF levels in the transgenic p lant 1,2RT 61 leaf N R G N RT IRF R F N ADCG Apiginin-C-glycoside NRT Narirutin NRG Naringin NHS Neohesperidin IRF Isorhoifolin RFN Rhoifolin N6M Naringin-6-Malonate DID Didymin PON Poncerin NAR Naringenin

PAGE 102

LIST OF REFERENCES Ambros V, Lee R, Lavanway A, Williams P, and Jewell D (2003) MicroRNAs and other tiny endogenous RNAs in C. elegans. Curr Biol 13: 807-818 Aufsatz V, Mette M F, van der Winden J, Matzke M, Matzke AJM (2002) HDA6 a putative histone deacytylase need to enhance DNA methylation induced by double-stranded RNA. EMBO J 21: 6832-6841 Bar-Peled M, Lewinsohn E, Fluhr R, Gressel J (1991) UDP-rhamnose:flavanone-7-0-glucoside-2’-0-rhanmosyltransferase. Purification and characterization of an enzyme catalyzing the production of bitter compounds in citrus. J Biol Chem 266: 20953-20959 Bar-Peled M, Fluhr R, Gressel J (1993) Juvenile-specific localization and accumulation of a rhamnosyl transferase and its bitter flavonoid in foliage, flowers and young citrus fruits. Plant Physiol 103:1377-1384 Bartoshuk LM (2000) Comparing sensory experiences across individuals: Recent biochemical advances illuminate genetic variation in taste perception. Chem Senses 25:447-460 Bartel B, Bartel D (2002) MicroRNas in plants. Genes Dev. 16:1616-1626 Bartel B, Bartel D (2003) MicroRNAs: at the root of plant development. Plant Physiol 132: 709-717. Beclin C, Boutet S, Waterhouse P, Vaucheret H (2002) A branched pathway for transgene-induced RNA silencing in plants. Curr Biol 12: 684-688 Berhow MA, Vandercook CE (1989) Biosynthesis of naringin and prunin in detached grapefruit. Phytochem 28: 627-1630 Berhow MA, Vandercook CE (1992) The reduction of naringin content of grapefruit by applications of gibberellic acid. Plant Growth Regulation 11: 75-80 Berhow MA, Tisserat B, Kanes K and Vandercook CE (1998) Survey of phenolic compounds produced in Citrus. USDA, ARS, Technical Bulletin No. 1856, U.S. Government Printing Office, Wash. DC Berhow MA (2000) Effects of early plant growth regulator treatments on flavonoid levels in grapefruit. Plant Growth Regulation 30: 225-232 Baulcombe DC (2000) Unwinding RNA silencing. Science 290:1108-1109 Brantl S (2002) Antisense-RNA regulation and RNA interference. Biochimica et Biophysica Acta-Gene Structure and expression 1575:15-25 91

PAGE 103

92 Brown DE, Rashotte AM, Murphy AS, Normanly J, Tague BW, Peer WA, Taiz L, Muday GK (2001) Flavonoids act as negative regulators of auxin transport in vivo in Arabidopsis thaliana. Plant Physiol 126: 524-535 Buer CS, Muday GK (2004) The transparent testa4 mutation prevents flavonoid synthesis and alters auxin transport and response of Arabidopsis roots to gravity and light. The Plant Cell 16: 1191-1205 Carmell M, Xuan Z, Zhang M, and Hannon, G (2002) The Argonaute family: tentacles that reach into RNAi, developmental control, stem cell maintenance, and tumorigenesis. Genes Dev 16: 2733-2742 Carrington J, and Ambros V (2003) Role of microRNAs in plant and animal development. Science 301: 336-338 Castillo J, Benavente O, del Rio JA (1992) Naringin and neohesperidin levels during development of leaves, flower buds, and fruits of C. aurantium. Plant Physiol 99: 67-73 Chuang CF, Meyerowitz EM (2000) Specific and heritable genetic interference by double-stranded RNA in Arabidopsis thaliana. PNAS 97: 4985-4990 Coberly LC, Rausher MD (2003) Analysis of a chalcone synthase mutant in Ipomea purpurea reveals a novel function for flavonoids: amelioration of heat stress. Molecular Ecology 12: 1113-1124 Cogoni C, Macino G (2000) Post-transcriptional gene silencing across kingdoms. Current Opinions in Genetics and Development 10: 638-643 Colliver SP, Morris P, Robbins, MP (1997) Differential modification of flavonoid and isoflavonoids biosynthesis with an antisense chalcone synthase construct in transgenic Lotus corniculatus. Plant Molecular Biology 35: 509-522 Covey NS (2000) Silencing genes silencing genes. Trends Plant Sci 5:10: 405-406 Del Rio JA, Gomez P, Baidez AG, Arcas MC, Botia JM, and Ortuo A (2004) Changes in the levels of polymethoxyflavones and flavanones as part of the defense mechanism of Citrus sinensis (Cv.Valencia Late) fruits against Phytophthora citrophthora. J Agric Food Chem 52: 1913-1917 Denli, A, and Hannon, G (2003) RNAi: ever growing puzzle. Trends Biochem Sci 28: 196-201 Dixon RA, Steele CL (1999) Flavonoids and isoflavonoids-a gold mine for metabolic engineering. Trends Plant Sci 4: 394-400

PAGE 104

93 Drewnowski A, Henderson SA, Shore AB (1997) Taste responses to naringin, a flavonoid, and the acceptance of grapefruit juice are related to genetic sensitivity to the 6n-propylthouracil. A J Clin Nutr 66: 391-397 Drewnowski A, Gomez-Carneros C (2000) Bitter taste, phytonutrients and the consumer: a review. A J Clin Nutr 72: 1424-1435 Elomaa P, Holton T (1994) Modification of flower color using genetic engineering. Biotechnology and Genetic Engineering Reviews 12: 63-88 El Euch C, Jay-Alllemand C, Pastuglia M, Doumas M, Doumas P, Charpentier JP, Capalli P, Jouanin L (1998) Plant Mol Biol 38:3: 467-479 Engelen van FA, Molthoff JW, Conner AJ Nap JP, Pereira A, and Stiekema WJ (1995) pBINPLUS: an improved plant transformation vector based on pBIN19. Transgenis Research 4: 288-290 Fagard M, Vauchert H (2000) (Trans)gene silencing in plants: How many mechanisms? Annu Rev Plant Physiol Plant Mol Biol 51: 167-194 Florida Department of Citrus. Economic and market research department. Citrus reference book. http://www.fred.ifas.ufl.edu/citrus/pubs/ref/CRB2004.pdf Revised in May 2004, last accessed August 16, 2004. Forkman G, Martens S (2001) Metabolic engineering and application of flavonoids. Curr Biotech 12: 155-160 Grosser JW, Gmitter FG Jr. (1998) Protoplast fusion and Citrus improvement. In: Dudley JW, Ryder EJ, Dinus RJ, Janic J. (eds) Plant Breeding Reviews Vol.8, Timber Press, Inc. Portland, Oregon pp 339-343 Gutirrez-E MA, Luth D and Moore GA (1997) Factors affecting the Agrobacterium-mediated transformation in Citrus and production of sour orange (Citrus aurantium L.) plants expressing the coat protein gene of citrus tristeza virus. Plant Cell Rep 16: 745 Hamilton JA, Baulcombe DC (1999) A species of small antisense RNA in posttranscriptional gene silencing in plants. Science 286: 950-952 Hasegawa S, Berhow MA, and Fong CH (1996) Analysis of bitter principles in citrus. In Linskens HF and Jackson JF (eds) Modern Methods of Plant Analysis, Vol. 18 Fruit Analysis, SpringerVerlag Berlin Heidelberg, pp 59-80 Helliwell, C, Waterhouse, P (2003) Constructs and methods of high-throughput gene silencing in plants. Methods 30: 289-295

PAGE 105

94 Hidaka T, Omura M, Ugaki M, Tomiyama M, Kato A, Ohshima M and Motoyoshi F (1990) Agrobacterium-mediated transformation and regeneration of Citrus spp. from suspension cells. Jpn. Breed 40: 199 Hodgson RW (1967) Horticultural varieties of citrus. In: Reuther W, Batchelor LD, Webber HJ (eds) The Citrus Industry, Vol 1, University of Claifornia Press, Berkeley, pp431-591 Holton TA, Brugliera, Tanaka Y (1993) Cloning and expression of flavonol synthase from Petunia hybrida. Plant J 4: 1003-1010 Horowitz RM, Gentili B (1969) Taste and structure in phenolic glycosides. J Agr Food Chem 17: 696-700 Horowitz RM, Gentili B (1977) Flavonoid constituents of Citrus. In:Nagy S, Shaw PE, Veldhuis MK (eds) Citrus Science and Technology, AVI Publishing, Westport, CT, pp 397-426 Ishige K, Schubert D, Sagara Y (2001) Flavonoids protect neuronal cells from oxidative stress by three distinct mechanisms. Free Radical Biology and Medicine 4: 433-446 Jacobs M, Rubery PH (1988) Naturally occurring auxin transport regulators Science 24: 346-349 Jackson JJ, Lindroth AM, Cao XF, and Jacobsen SE (2002) Control of CpNpG DNA methylation by KRIPTONITE histone H3 methyltransferase. Nature 416: 556-560 Jones L, Ennos AR, Turner SR (2001) Cloning and characterization of irregular xylem4 (irx4): a severely lignin-deficient mutant of Arabidopsis. The Plant Journal 26: 2:205-216 Jourdan P, McIntosh CA, and Mansell RL (1985) Naringin levels in citrus tissues. Plant Physiology 77: 903-908. Kaneyoshi (Hiramatsu) J, Kobayashi S, Nakamura Y, Shigemota N and Doi Y (1994) A simple and efficient gene transfer system of trifoliate orange (Poncirus trifoliata Raf.). Plant Cell Rep 13: 541 Kawaii S, Tomono Y, Katase E, Ogawa K, Yano M (1999) Quantittation of flavonoid constituents in citrus fruits. J.Agric. Food Chem. 47: 3565-3571 Koes RE, Uattrocchio F, Mol JNM (1994) The flavonoid biosynthetic pathway in plants:function and evolution. Bioessay 16:2: 123-132 Kusaba M, Miyahara K, Iida S, Fukuoka H, Takano, T, Sassa H, Nishimura M, Nishio T (2003) Low glutelin content 1: a dominant mutation that suppresses the glutelin multigene family via RNA silencing in rice. Plant Cell 15: 1455-1467

PAGE 106

95 Lagos-Quintana M, Rauhut R, Lendeckel W, and Tuschl T (2001) Identification of novel genes coding for small expressed RNAs. Science 294: 853-858 Lagos-Quintana M, Rauhut R., Yalcin A, Meyer J, Lendeckel W, and Tuschl T (2002) Identification of tissue-specific microRNAs from mouse. Curr Biol 12: 735-739 Lau N, Lim L, Weinstein E, and Bartel D (2001) An abundant class of tiny RNAs with probable regulatory roles in Caenorhabditis elegans. Science 294: 858-862 Lawrence RJ, Pkaard CS (2003) Transgene-induced RNA interference: A strategy for overcoming gene redundancy in polyploids to generate loss-function mutations. Plant J 36: 114-121 Lee NK, Choi SH, Park SH, Park EK, Kim DH (2004) Antiallergic activity of hesperidin is activated by intestinal microflora. Pharmacology 71: 174-180 Lewinsohn E, Britsch L, Mazur Y, Gressel J (1989) Flavanone glycoside biosynthesis in Citrus-Chalcone synthase, UDP-glucose:flavanone-7-0-glucosyltransferase and rhamnosyl transferase acticities in cell-free extracts. Plant Physiology 91: 1323-1328 Lewinsohn E, Berman E, Mazur Y, Gressel J (1989) Glucosylation and (1,6) rhamnosylation of exogenous flavanones by undifferentiated citrus cell cultures. Plant Sci 61: 23-28 Luth D (1997) Transgenic approaches to modulating naringin content in Citrus paradisi. PhD. thesis, University of Florida USA. Luth D (1999) Transgenic grapefruit plants obtained by Agrobacterium tumefaciens-mediated transformation. Plant Cell, Tissue and Organ Culture 57: 219-222 Manthey JA, Buslig BS (1998) Flavonoids in the Living System. Plenum Press, New York Manthey JA, Guthrie N (2002) Antiproliferative activities of citrus flavonoids against six human cancer cell lines. J Agric Food Chem 50: 5837-5843 Mo Y, Nagel C, Taylor LP (1992) Biochemical complementation of chalcone synthase mutants defines role for flavonols in functional pollen. Proc Natl Acad Sci USA 89: 7213-7217 Moore GA Jacaano CC, Neidigh J, Lawrence SP, Cline K (1992) Agrobacterium-mediated transformation of citrus stem segments and regeneration of transgenic plants. Plant Cell Rep 11: 238-242 Moore GA (2001) Oranges and lemons: clues to the taxonomy of citrus from molecular markers. Trends Genetics 17: 536-540

PAGE 107

96 Moriguchi T, Kita M, Tomano Y, Endo T, Omura M (2001) Gene expression in flavonoid biosynthesis: correlation with flavonoid accumulation in developing citrus fruit. Phisiologia Plantarum 111: 66-74 Moriguchi T, Kita M, Tomano Y, Endo T, Omura M (1999) One type of Chalcone Synthase expressed during embryogenesis regulates the flavonoid accumulation in citrus cell cultures. Plant Cell Physiol 40: 6:651-655 Murphy A, Peer WA, Taiz L (2000) Regulation of auxin transport by aminopeptidases and endogenous flavonoids. Planta 211: 315-324 Ness AR, Powles JW (1997) Fruits and vegetables, and cardiovascular disease: a review. Int. J Epidemiol 26: 1-13 Napoli C, Lemieux C, Jorgensen R (1990) Introduction of a chimeric chalcone synthase gene into petunia results in reversible co-suppression of homologous genes in trans. The Plant Cell 2: 279-289 Pea L, Cervera M, Jurez J, Ortega C, Pina JA, Duran-Vila N and Navarro L (1995a) High efficiency Agrobacterium-mediated transformation and regeneration of citrus. Plant Science 104: 183 Pea L, Cervera M, Jurez J, Navarro L, Pina JA and Navarro L (1997) Genetic transformation of lime (Citrus aurantifolia Swing.): Factors affecting transformation and regeneration. Plant Cell Rep 17: 731 Pea L, Cervera M, Jurez J, Navarro A, Pina JA, Duran-Vila N and Navarro L (1995b) Agrobacterium-mediated transformation of sweet orange and regeneration of transgenic plants. Plant Cell Rep 14: 616 Puri M, Marwaha SS, Kathari RM (1996) Studies on the applicability of alginate-entarpped naringinase for the debittering of kinnow juic. Enzyme and Microbial Technology 18: 281-285 Puri M, Banerjee UC (2000) Production, purification,and characterization of the debittering enzyme naringinase. Biotechnology Advances 18: 207-217 Rouseff RL, Martin SF, Youtsey CO (1987) Quantitative survey of narirutin, naringin, hesperidin and neohesperidin in citrus. J Agric Food Chem 35: 1027-1030 Slightom JL (1991) Custom polymerase chain reaction engineering of a plant expression vector. Gene 100: 251-255 Stoutjesdijk PA, Singh SP, Liu Q, Huristone CJ, Waterhouse PA, Green AG (2002) hpRNA-mediated targeting of the Arabidopsis FAD2 gene gives highly efficient and stable silencing. Plant Physiology 129: 1723-1731

PAGE 108

97 Tang G, Reinhart B, Bartel D, and Zamore P (2003) A biochemical framework for RNA silencing in plants. Genes Dev 17: 49-63 Van der Krol AR, Lenting PE, Veenstra J, Van der Meer IM, Koes RE, Gerats IGM, Mol JNM, Stuitje AR (1988) An anti-sense chalcone synthase gene in transgenic plants inhibits flower pigmentation. Nature 333: 866-869 Van der Krol AR, Mur LA, Beld M, Mol JMN, Stuitje AR (1990) Flavonoid genes in petuniaAddition of a limited number of gene copies may lead to a suppression of gene expression. Plant Cell 2: 291-299 Vance V, Vaucheret H (2001) RNA Silencing in plants-defense and counterdefense. Science 292: 2277-2280 Vaucheret H, Vazquez F, Crata P, and Bartel D (2004) The action of ARGONAUTE1 in the miRNA pathway are crucial for plant development. Genes Dev 18: 1-11 Verniere C, Hartung JS, Pruvost OP, Civerolo EL, Alvarez AM, Maestri P, Luisetti J (1998) Characterization of phenotypically distinct strains of Xathomonas axonopodis pv. citri from Southwest Asia. European Journal of Plant Pathology 104: 477-487 Vougt T, Jones P (2000) Glucosyltransferases in plant natural product synthesis: characterization of a supergene family. Trends in Plant Science 5: 380-386 Zamore PD (2002) Ancient pathways programmed by small RNAs. Science 296: 1265-126 Zilberman, D, Cao X, and Jacobsen S (2003) ARGONAUTE4 control of locus-specific siRNA accumulation and DNA and histone methylation. Science 299: 716-719. Waterhouse PM, Graham MW, Wang MB (1998) Virus resistance and gene silencing in plants can be induced by simultaneous expression of sense and antisense RNA. Proc Natl Acad Sci USA 95: 13959-13964. Waterhouse PM, Wang MB, Finnegan EJ (2001a) Role of short RNAs in gene silencing. Trends Plant Sci 6: 297-301 Waterhouse PM, Wang MB, Lough T (2001b) Gene silencing as an adaptive defense against viruses. Nature 411: 834-842 Waterhouse PM (2001c) Application of gene silencing in plants. Curr Plant Biol 5: 146-150 Wesly SV, Helliwell CA, Smith NA, Wang MB, Rouse DT, Liu Q, Gooding PS, Singh SP, Abbott D, Stoutjesdijk PA, Robinson SP, Gleave AP, Green AG, Waterhouse PM (2001) Construct design for efficient, effective and high-throughput gene silencing in plants. Plant J 27: 581-590

PAGE 109

98 Winkel-Shirley B (2001a) Flavonoid biosynthesis. A colorful model for genetics, biochemistry, cell biology, and biotechnology. Plant Physiology 126: 485-493 Winkel-Shirley B (2001b) It takes a garden. How work on diverse plant species has contributed to an understanding of flavonoid metabolism. Plant Physiology 127: 1399-1404.

PAGE 110

BIOGRAPHICAL SKETCH Ufuk Koca was born in Ankara, Turkey. Ufuk attended the Pharmacy Faculty at Gazi University, where she earned her diploma to be Pharmacist in 1993. Following her graduation, she began her master’s studies at the Pharmacognosy Department of the Pharmacy Faculty. While she was conducting her research, she also became a teaching assistant at the department. In 1995, Ufuk earned a scholarship to come to the USA for her master’s and PhD studies, after having successfully passed a nation-wide exam. This scholarship was provided by the Ministry of Education of The Republic of Turkey. Ufuk came to Pullman, Washington, in 1996 where she was accepted by the Biochemistry department of Washington State University. She worked in Dr. Norman Lewis’s Laboratory at the Institute of Biological Chemistry where she earned her master’s degree. Ufuk began her PhD degree in 1999 at the University of Florida in Plant Molecular and Cellular Biology program, which is at the Horticultural Sciences Department. Upon completion of her degree, she will join the Pharmacy Faculty in Erciyes University in Turkey as the founding chair of the Pharmacognosy Department. 99