Geminivirus Al2 and L2 Proteins Interact with and Inactivate

GEMINIVIRUS AL2 AND L2 PROTEINS INTERACT WITH AND INACTIVATE

ADENOSINE KINASE

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By Hui Wang, M.S. *****

The Ohio State University 2004

Dissertation Committee: Dr. David Bisaro, Adviser

Dr. Biao Ding Approved by Dr. Erich Grotewold

Dr. Deborah Parris ______Adviser Molecular, Cellular, and Developmental Biology

ABSTRACT

AL2 and L2 are related proteins encoded by geminiviruses of the Begomovirus and Curtovirus genera, respectively. Both are pathogenicity determinants that cause enhanced susceptibility when expressed in transgenic plants. To understand how geminiviruses defeat host mechanisms that limit infectivity, we searched for cellular proteins that interact with AL2 and L2. Here, evidence is presented which indicates that the viral proteins interact with and inactivate adenosine kinase (ADK), a nucleoside kinase that catalyzes the salvage synthesis of 5'-AMP from adenosine and ATP. We show that the AL2 and L2 proteins inactivate ADK in vitro and following coexpression in

E. coli and yeast. We also demonstrate that ADK activity is reduced in transgenic plants expressing the viral proteins and in geminivirus infected plant tissue. In contrast, ADK activity is increased following inoculation of plants with diverse RNA viruses or a geminivirus lacking a functional L2 gene. Consistent with its ability to interact with multiple cellular kinases, we also demonstrate that AL2 is present in both the nucleus and the cytoplasm of infected plant cells. To our knowledge this is the first evidence that

ADK is targeted by viral pathogens, and the first evidence that this "housekeeping"

enzyme might be a part of host defense responses.

In previous work, we showed that AL2 and L2 also interact with and inactivate

SNF1 kinase, a global regulator of metabolism that is activated by 5'-AMP. Together

these observations suggest that metabolic alterations mediated by SNF1 are an important

component of innate antiviral defenses and that inactivation of ADK and SNF1 by the

geminivirus proteins represents a dual strategy for countering this defense.

Another connection between ADK and viral pathogenesis relates to RNA

silencing suppression, which is associated with methylation. By recycling adenosine,

ADK plays a critical role in sustaining the methyl cycle and SAM-dependent

methyltransferase activity. AL2 proteins have also been shown to act as suppressors of

RNA silencing, an adaptive host defense response. In this thesis, we confirm that the

geminiviruses TGMV and BCTV induce and are targeted by RNA silencing in the course

of a normal infection, and that the AL2 and L2 proteins they encode are capable of

suppressing RNA silencing in a transient three component system. Remarkably, we

found that inhibiting ADK activity at the RNA level by expression of a dsRNA construct

directed against ADK, or at the protein level by the use of an ADK inhibitor (the

adenosine analogue RBI), also results in silencing suppression. Direct measurement of

ADK activity in tissue showing silencing suppression following infiltration with AL2,

L2, dsADK, or RBI revealed that in all cases ADK activity was significantly reduced iii

(>50%). These data provide strong evidence that AL2 and L2 suppress silencing by inhibiting ADK activity.

Taken together, the studies in this thesis first demonstrated that geminivirus AL2 and L2 proteins interact with and inactivate adenosine kinase and that this inactivation has two effects on pathogenesis: interference with plant innate defense and interference with plant adaptive defense.

Dedicated to my parents

ACKNOWLEDGMENTS

I express my deepest gratitude to my mentor Dr. David M. Bisaro for his patience and support. His enthusiasm and encouragement in pursuing novel ideas and insights were the key to the success of this project.

I extend my appreciation to my committee members Dr. Biao Ding, Dr. Erich

Grotewold and Dr. Deborah S. Parris for their time, support, guidance and advice.

I would like to express my sincere thanks to Dr. Garry Sunter and Dr. Kenneth

Buckley for their assistance in overcoming many technical difficulties and for all the useful suggestions and insightful discussions to improve my research. Especially, I would like to thank Dr. Garry Sunter for the work of AL2 protein localization. I would also like to acknowledge other members of the Bisaro lab: Janet Sunter, Dr. Linhui Hao, Duan

Wang, Xiaojuan Yang, Cody Buchmann, Nick Green and Tim Cowley, and members in

Dr. Biao Ding’s lab, Dr.Yijun Qi and Dr. Asuka Itaya for their help, friendship and stimulating discussions.

I would also like to thank the staff of the biotechnology center: Melinda Parker,

Diane Furtney, Dave Long, Scott Hines, Joe Takayama and MCDB secretary, Jan Zinich for all of their help.

Finally, I am indebted to my parents and my family for their unconditional love, continual support and encouragement. To my father, Jiwen Wang, my sister Ai Wang and my brother Yun Wang, I cannot say thank you enough. Whatever happened, they were always there with me. To my wife Xiuping and daughter Emily, I am sorry for all the inconvenience I have brought home over all the challenging time. Thank you for your understanding, constant love and support.

vii

VITA

1983 - 1987…………… B.S, Plant Pathology, Plant Protection Department, Zhejiang

Agricultural University, Hangzhou, PRC

1987 -1990…………… M.S, Phytobacteriology, Plant Pathology Department, Nanjing

Agricultural University, Nanjing, PRC

1990 - 1996…………… Researcher, Institute of Virology, Zhejiang Academy of

Agricultural Sciences, Hangzhou, PRC

1996 - 1998…………… Research associate, Molecular Virology, Plant Pathology

Department, University of Hawaii at Manoa, Honolulu, HI, USA

1998 -present………… Research associate, MCDB, The Ohio State University

PUBLICATIONS

Research Publications:

1. Hui Wang, Linhui Hao, Garry Sunter and David Bisaro (2003) Adenosine Kinase Is Inactivated by Geminivirus AL2 and L2 Proteins. Plant Cell 15(12):3020-3032 (The research published in this paper is highlighted in “In This Issue” section of the journal, as “Viral Defense and Counterdefense: A Role for Adenosine Kinase in Innate Defense and RNA Silencing”.)

viii

2. Linhui Hao, Hui Wang, Garry Sunter and David Bisaro (2003) Geminivirus AL2 and L2 Proteins Interact with and Inactivate SNF1 Kinase. Plant Cell 15(4):1034- 1048

3. Hui Wang and Ren Xinzheng (1993) Adsorption, penetration and multiplication of the pathogen (Pseudomonas solanacearum) in the roots of tomato plant. Acta Phytopathological Sinica 23 (2):143-150

4. Hui Wang and Yili Ruan (1993) Plant viral pinwheel inclusion bodies. Virol. Sinica 8(2):119-124 (review article)

5. Hui Wang and Yili Ruan (1993) Mature embryo culture of barley cultivars resistant or susceptible to Barley Yellow Mosaic Virus (BaYMV). Barley Science 23(2):17-22

6. Yili Ruan, Wanhe Zou, and Hui Wang (1995) Identification of resistance in barley varieties to BaYMV with ELISA. Acta Phytopathological Sinica 25(3):232-238

7. Yili Ruan, Wanhe Zou, and Hui Wang (1995) Evaluation of barley cultivars to BaYMV with ELISA method. Acta Agri. Zhejiangesesis 7(3):187-193

8. Yili Ruan, Wanhe Zou, Hui Wang, Chen Shi and He Chen (1994) Resistance and susceptibility reactions of barley varieties of Japan and West-Europe to Chinese BaYMV. Acta Phytophylac. Sinica 21(3):239-245

9. Yili Ruan, Wanhe Zou, Hui Wang, and Shuimiao Huang (1993) Pathogenicity of main isolates of BaYMV to barley cultivars in China. Acta Phytopathol. Sinica 23(4):348-354

10. Yili Ruan, Wanhe Zou, Hui Wang, Chen Shi and He Chen (1993) Pathogenicity of BaYMV in the provinces of Zhejiang, Jiangsu and Shanghai to parts of

BaYMV-resistant cultivars of Japan, England and Germany. Acta Agri. Zhejiangesesis 5(3):182-187

11. Jianping Chen, Majia Dong, Jinfei Tao, Yili Ruan, Hui Wang and Shengxiang Chen (1993) Detection and identification of plant viruses with different morphology in leaf-saps by colloidal gold immunosorbent electron microscopy. Acta Phytopathol. Sinica 23(2):169-175

FIELDS OF STUDY

Major Field: Molecular, Cellular, and Developmental Biology

TABLE OF CONTENTS

Abstract ………………………………………………………………………………….ii

Dedication……………………………………………………………………………… v

Acknowledgments……………………………………………………………………....vi

Vita………………………………………………………………………………….....viii

List of Figures………………………………………………………………………….xvi

Chapters:

CHAPTER 1 ...... 1

INTRODUCTION...... 1

1.1 Geminiviruses...... 1

1.1.1 Geminivirus classification...... 2

1.1.2 Geminivirus nomenclature and gene organization ...... 3

1.1.3 Geminivirus replication and cell cycle reprogramming...... 6

1.2 AL2 and L2 have multiple functions...... 12

1.2.1 AL2 is a transcriptional activator but L2 is not ...... 12

1.2.2 AL2 and L2 act as pathogenicity determinants by interfering with global

metabolism ...... 15

1.2.3 AL2 and L2 are RNA silencing suppressors...... 16

1.3 SNF1, a cellular fuel gauge, is a global metabolic regulator...... 17

1.3.1 SNF1 has multiple functions involved in stress responses ...... 17

1.3.2 SNF1 is inactivated by geminivirus AL2 and L2 proteins ...... 28

1.4 Adenosine kinase generates 5′-AMP and has potential functions in plant defense

responses ...... 31

1.4.1 ADK has important pharmacological roles in mammals...... 32

1.4.2 ADK has a critical role in methyl cycling in yeast ...... 33

1.4.3 ADK is involved in cytokinin metabolism and maintenance of

transmethylation in plants ...... 34

1.5 Plant defense responses to virus infection...... 41

1.5.1 Plant innate defense responses – the hypersensitive response and

systemic acquired resistance ...... 42

1.5.2 An adaptive plant defense response - RNA silencing...... 44

CHAPTER 2 ...... 66

GEMINIVIRUS AL2 AND L2 PROTEINS INTERACT WITH

AND INACTIVATE ADENOSINE KINASE ...... 66

2.1 Introduction ...... 66

2.2 Results ...... 69

2.2.1 AL2 and L2 interact with ADK ...... 69

2.2.2 AL2 and L2 inhibit ADK activity in E. coli ...... 71

2.2.3 AL2 and L2 inhibit ADK in vitro...... 72

xii

2.2.4 AL2 and L2 inhibit ADK activity in yeast...... 74

2.2.5 AL2 and L2 inhibit ADK activity in plants ...... 75

2.2.6 AL2 accumulates in the cytoplasm and the nucleus ...... 77

2.2.7 SNF1 inactivates adenosine kinase ...... 80

2.3 Discussion...... 82

2.4 Methods ...... 90

2.4.1 Two-hybrid analysis...... 90

2.4.2 E. coli complementation experiments...... 91

2.4.3 Yeast complementation experiments...... 92

2.4.4 Protein expression ...... 93

2.4.5 SNF1 kinase assay ...... 94

2.4.6 ADK assays...... 94

2.4.7 Immunolocalization of AL2...... 95

2.4.8 Subcellular localization of AL2 ...... 96

CHAPTER 3 ...... 120

ADENOSINE KINASE INHIBITION AND SUPPRESSION OF

RNA SILENCING BY GEMINIVIRUS AL2 AND L2

PROTEINS...... 120

3.1 Introduction ...... 120

3. 2 Results ...... 125

3.2.1 TGMV and BCTV induce RNA silencing in infected plants ...... 125

xiii

3.2.2 RNA silencing can be suppressed by AL2 and L2 proteins and by inhibiting

ADK ...... 127

3.2.3 ADK activity is reduced in tissues infiltrated with AL2/L2 protein, dsADK,

or RBI...... 129

3.3 Discussion...... 130

3.4 Methods ...... 137

3.4.1 Plant material, Agrobacterium infiltration and agroinoculation ...... 137

3.4.2 Plasmid constructs for transient expression ...... 138

3.4.3 PTGS suppression assay and GFP imaging ...... 139

3.4.4 RNA isolation and Gel blot analysis...... 140

3.4.5 Protein extraction and ADK kinase assay...... 141

CHAPTER 4 ...... 152

DISCUSSION...... 152

4.1 Inactivation of ADK by AL2 and L2 and its effects on innate defense ...... 152

4.1.1 Geminivirus AL2 and L2 proteins are pathogenicity determinants...... 152

4.1.2 Geminivirus AL2/ L2-mediated ES acts by blocking SNF1 activity...... 153

4.1.3 SNF1-mediated ES is different from other innate defense responses...... 154

4.1.4 Virus infection tends to increase ADK activity...... 155

4.1.5 Inactivation of ADK by AL2 and L2 could efficiently block SNF1-mediated

plant innate defense...... 157

4.1.6 Virus infection affects plant metabolism including sugar metabolism ...... 158

xiv

4.1.7 SNF1 activates plant defense responses possibly by increasing soluble sugars

...... 160

4.1.8 SNF1, ADK and AL2/L2 might form a complex ...... 162

4.2 Inactivation of ADK by AL2 and L2 and its effects on adaptive defense...... 164

4.2.1 Geminiviruses can induce RNA silencing and encode RNA silencing

suppressors ...... 164

4.2.4 Inactivation of ADK by AL2 and L2 suppresses gene silencing...... 166

4.3 Relationship between the AL2/L2-mediated ES phenotype and AL2/L2 mediated

silencing suppression...... 167

BIBLIOGRAPHY:...... 171

LIST OF FIGURES

Page Figure 1.1 Geminivirus genome organization...... 58

Figure 1.2 Geminivirus replication cycle...... 60

Figure 1.3 The methyl cycle ...... 62

Figure 1. 4 Steps leading to RNA silencing and suppression of RNA silencing...... 64

Figure 2.1 AL2 and L2 abolish complementation of E. coli HO4 by ADK ...... 98

Figure 2.2 ADK expression and in vitro activity...... 100

Figure 2.3 Recombinant AL2 and L2 proteins inhibit ADK activity in vitro...... 102

Figure 2.4 L2 reduces complementation of a yeast ADK deletion strain by Arabidopsis ...... 104

Figure 2.5 ADK levels are reduced in transgenic plants expressing AL2 and L2, and also during geminivirus infection...... 106

Figure 2.6 AL2 is in the nucleus and the cytoplasm of TGMV infected cells...... 108

Figure 2.7 Localization of AL2 in insect cells...... 110

Figure 2. 8 SNF1 and SNF1-KD abolish complementation of E. coli HO4 by ADK .. 112

Figure 2. 9 Potential SNF1 phosphorylation sites in ADK protein ...... 114

Figure 2. 10 ADK protein is phosphorylated by SNF1 protein kinase...... 116

Figure 2. 11 ADK activity is affected by incubation with SNF1 protein kinase ...... 118

Figure 3.1 Geminiviruses TGMV and BCTV are inducers and targets of RNA silencing ...... 142

Figure 3.2 Geminiviruses AL2/L2 proteins and dsADK suppress RNA silencing ...... 144

xvi

Figure 3.3 Geminiviruses AL2/L2 proteins, dsADK and ADK inhibitor suppress silencing ...... 146

Figure 3.4 ADK inhibitor RBI blocks ADK activity...... 148

Figure 3.5 ADK activity in infiltration zone...... 150

Figure 4.1 Working Model for AL2/L2 mediated pathogenesis...... 169

xvii

CHAPTER 1

INTRODUCTION

1.1 Geminiviruses

What is a virus? “A virus is a set of one or more nucleic acid template molecules, normally encased in a protective coat or coats of protein or lipoprotein, that is able to

organize its own replication only within suitable host cells. It can usually be horizontally

transmitted between hosts. Within such cells, virus replication is (1) dependent on the

host’s protein-synthesizing machinery, (2) organized from pools of the required materials

rather than by binary fission, (3) located at sites that are not separated from the host cell

contents by a lipoprotein bilayer membrane, and (4) continually giving rise to variants

through various kinds of change in the viral nucleic acid” (Hull, 2002).

Geminiviruses are plant viruses belonging to the family Geminiviridae, which

contains viruses with one or two circular, single-stranded DNA genomes (Bisaro, 1996).

Geminiviruses cause economically important diseases of both monocotyledonous and

dicotyledonous plants throughout the world (Hull, 2002; Mansoor et al., 2003). In fact, a

geminivirus is known to be responsible for the earliest recorded plant virus disease

(Saunders et al., 2003). 1

Geminiviruses are unusual among plant viruses in their twin icosahedral

(geminate) particles. Each paired particle encapsidates a single molecule of circular,

single-stranded DNA (ssDNA) 2.5 to 3.0 kb in length. Among plant viruses, only

geminiviruses and nanoviruses have ssDNA genomes. The ssDNA genomes are

expressed and replicated from viral double-stranded DNA (dsDNA) replicative forms

(RFs), which are organized as minichromosomes complexed with histone proteins

(Gutierrez, 1999; Hanley-Bowdoin et al., 2000; Pilartz and Jeske, 1992). Geminivirus

replication and transcription is carried out by host DNA and RNA polymerases.

1.1.1 Geminivirus classification

The family Geminiviridae currently is divided into four genera on the basis of

genome organization, host range and insect vectors (Fauquet et al., 2003). Those with a monopartite genome and transmitted by leafhopper vectors, primarily to monocotyledonous plants, are included in the genus Mastrevirus. Maize streak virus

(MSV) is the type species in this genus (Palmer and Rybicki, 1998). Viruses with monopartite genomes distinct from those of the mastreviruses and transmitted by leafhopper vectors to dicotyledonous plants are included in Curtovirus, with Beet curly top virus (BCTV) as the type species. Genus Topocuvirus is a recently designated genus by the International Committee on Taxonomy of Viruses (ICTV) (Pringle, 1999a;

Pringle, 1999b), which has been separated from the Curtovirus genus. Only one member

(also the type species), Tomato pseudo-curly top virus, represents this genus. It has a

similar genome organization to the curtoviruses but is transmitted by a treehopper vector

to dicotyledonous plants.

Begomovirus is the largest genus with more than 100 species (Fauquet et al., 2003).

They all infect dicot plants and are transmitted by whiteflies (Bemisia tabaci Genn.), with

Bean golden yellow mosaic virus (originally Bean golden mosaic virus –Puerto Rico) as

the type species. Most members of this genus have their genomes divided between two

DNA molecules of similar size (about 2.5 kb) and both DNAs are needed for infectivity

(Hamilton et al., 1983; Stanley, 1983). The nucleotide sequences of the two DNAs are

very different except for a 200-nucleotide non-coding region, the common region, which

is almost identical on each DNA. The genomes of some members in this group are monopartite. Well-characterized bipartite begomoviruses include Tomato golden mosaic

virus (TGMV), African cassava mosaic virus (ACMV), Squash leaf curl virus (SqLCV)

and Cabbage leaf curl virus (CabLCV). Well-characterized monopartite viruses include

Tomato yellow leaf curl virus (TYLCV) and Tomato leaf curl virus (ToLCV). The

curtovirus BCTV and the begomovirus TGMV are the member viruses studied in our lab

and are the focus of my thesis.

1.1.2 Geminivirus nomenclature and gene organization

Geminivirus genes are oriented in different directions on the circular genomes.

There are two nomenclature systems currently in use. In both systems, genes and gene

products are designated by numbers. One system names genes based on whether they are

specified by the virion (V) or complementary (C) sense DNA strand, whereas the other 3

system indicates genes as oriented in the rightward (R, viral sense, clockwise) or leftward

(L, complementary sense, counterclockwise) direction on the genome map (Figure 1.1).

In this thesis the R and L system will be used. In bipartite viruses, genome components are designated A and B.

All bipartite geminiviruses have a conserved common region (CR) as mentioned above. The CR is part of the non-coding intergenic region (IR). An inverted repeat within the IR can form a hairpin loop, and within the loop is a conserved sequence, 5′ -

TAATATTAC-3′ found in all geminiviruses (Lazarowitz, 1987). Interestingly, this IR and its hairpin structure are also found in the related Nanovirus genus (5′ -

TANTATTAC-3′), from the Circoviridae family. These plant viruses also have small circular, ss DNA genomes, but with 6 or 7 genomes separately encapsidated (Burns,

Harding, and Dale, 1995; Hull, 2002). The IR region and hairpin structure are important for virus replication in both the Geminiviridae and Circoviridae.

The begomoviruses can be further divided into Old World and New World groups. Most of the Old World and New World begomoviruses have two component genomes. The single-component types are native to the Old World. The two component begomoviruses encode five or six genes on DNA A genome, one or two on the virion sense strand (termed AR) and four in the complementary-sense strand (AL). DNA B encodes two proteins, one on the virion-sense strand (BR) and one on the complementary strand (BL) (Figure 1.1). The product of ORF AR1 is the coat protein. On genome A, the proteins encoded by the complementary sense of DNA A are all involved in replication and expression. AL1 is the replication initiator protein (Rep), involved in viral DNA replication. AL2 was called transcriptional activator protein when it was first identified in

our lab, thus it is also called TrAP (Sunter and Bisaro, 1991). Recently, in our lab it has

been found to have more than one function, which is the focus of this thesis. AL3 is the

replication enhancer protein as AL3 mutants typically accumulate viral DNA in reduced

amounts (Sunter et al., 1990); the function of AL4 is unknown. Genome DNA B only

encodes two genes, BR1, the virion sense gene, which is the nuclear shuttle protein, and

BL1, the complementary sense gene, which is involved in cell-to-cell movement. Both

BR1 and BL1 are required for spread of the virus in infected plants. The genome

organization of the single-component begomoviruses is essentially the same as that of

DNA A of the two-component viruses (Figure 1.1).

The BCTV genome has about 3000 nucleotides (Hormuzdi and Bisaro, 1993;

Stanley et al., 1986). This single component curtovirus genome encodes seven proteins.

The product of ORF R1 is the virion coat protein, which is required for infectivity

(Hormuzdi and Bisaro, 1993), that of ORF R2 is involved in regulation of the genomic ss

and ds DNA levels, and that of ORF R3 facilitates the cell-to-cell movement of the virus

(Hormuzdi and Bisaro, 1995; Stanley et al., 1992a). Four proteins encoded by the

complementary-sense ORFs are involved in virus replication. ORF L1 encodes the Rep

protein, L3 a protein analogous to the replication enhancer protein of begomoviruses, and

L4 a protein that can initiate cell division (Latham et al., 1997). The function of L2 was

unknown. Recently in our lab its two functions have been found to be related with plant

innate and adaptive defenses, which is also the focus of this thesis.

1.1.3 Geminivirus replication and cell cycle reprogramming

Replication is a key step for successful virus infection. In order to elucidate the details, two research methods have been of great use: agroinfection of whole plants, and transfection of protoplasts. It was shown that when the whole maize plants were

inoculated with Agrobacterium tumefaciens containing tandem repeats of MSV DNA with a modified Ti plasmid, treated plants showed symptoms of MSV infection

(Grimsley et al., 1987). This experiment not only provided a very convenient way to infect plants but also demonstrated that Agrobacterium could interact with a monocotyledonous plant. The agroinfection procedure also provides a simple and efficient assay for TGMV replication (Elmer et al., 1988). When transgenic Nicotiana benthamiana plants are inoculated with two Agrobacterium cultures, one containing

DNA A and one containing DNA B, infection usually ensues. This method is very efficient; as few as 2000 Agrobacterium cells can generate an infection (Elmer et al.,

1988).

Protoplasts are another fast and convenient way to study viral DNA replication.

Protoplasts can be prepared from leaf tissues or suspension culture cells by removing their cellulose walls. Currently, both PEG-mediated transfection and electroporation methods are used to deliver viral DNA into protoplasts. Viral replication can be observed within 24 hours after transfection (Brough et al., 1992).

How does the small geminivirus genome replicate in the host cell? As we mentioned above, geminivirus genomes are compactly organized and do not contain a polymerase gene. Therefore, geminivirus replication requires host DNA polymerase and

replication machinery. The role of viral proteins is to reprogram the host cell so that the

cellular replication machinery is expressed, to direct the machinery to the viral genome,

and initiate replication. The replication of geminiviruses is from ssDNA to ssDNA via a

dsDNA intermediate stage in the nuclei of host cells, by rolling circle replication

mechanism (RCR) (Gutierrez, 1999; Gutierrez, 2000; Hanley-Bowdoin et al., 1999)

(Figure 1.2). RCR is common in the replication of ssDNA containing bacterial viruses

(phage) and plasmids (Doermann, 1973; Novick, 1998). It is a two-step process: in the

first phase ss (+) strand is the template for the synthesis of (-) strand to generate a ds, replicative form (RF). The second step is to generate (+) ssDNA from the dsRF. This

double-stranded DNA intermediate includes both open and covalently closed

(supercoiled) circular forms, and is organized as minichromosomes complexed with

histone proteins (Bisaro et al., 1982; Pilartz and Jeske, 1992a). Thus chromatin

remodeling might be involved in controlling viral genome replication and gene

expression. The dsRF has two functions. One function is as the template for transcription.

Viral transcription from the dsRF is accomplished by host RNA polymerase II. Thus, a

geminivirus requires host cell RNA polymerase and related transcription factors to

accomplish its life cycle. Another function of the dsRF is to serve as template for (+)-

strand DNA synthesis.

The mechanism by which (+)-strand viral DNA is converted to dsRF is unknown,

but probably involves priming of (-)-strand synthesis by an RNA polymerase (primase)

activity. A small oligonucleotide complementary to the 3′ intergenic region has been

isolated from several mastreviruses (Donson et al., 1984; Hayes et al., 1988a; Morris et

al., 1992). This oligonucleotide is encapsidated with the (+)- viral DNA genome and can

be extended by DNA polymerase in vitro. Thus it may be the in vivo (-)-strand primer.

This is supported by the finding that sequences in the 3′ intergenic region of Wheat dwarf

virus (WDV) have been implicated as being involved in replication (Hayes et al., 1988b;

Kammann et al., 1991). In contrast, begomoviruses do not encapsidate such an

oligonucleotide. Two-dimensional electrophoretic analysis of ACMV replication

intermediates (Saunders, Lucy, and Stanley, 1991) indicated that (-)-strand synthesis is

primed within the 5′ intergenic region in this begomovirus.

Synthesis of viral (+)-strand is initiated in the hairpin motif within IR (Arguello-

Astorga et al., 1994). The replication begins at a specific site in the hairpin called the

origin of replication, or “ori”. At the origin, the two strands of DNA are pulled apart to

form a replication bubble. This creates a region of single stranded DNA on each side of

the bubble. The DNA polymerase machinery then can move in and begin to synthesize

the new strands of DNA, using the old strands as templates. The geminivirus genome,

like the circular DNA in prokaryotic cells, has a single origin of replication. The loop in

the hairpin contains a sequence 5′ -TAATATTAC, that is conserved among all

geminivirus genomes and is found in the (+)-strand origins of other rolling circle systems

(Baas, 1987; Rogers et al., 1986). The (+)-strand cleavage site has been mapped to the

same position in the conserved sequence in all geminiviruses.

AL1 (Rep) localizes to nuclei of infected plant cells (Nagar et al., 1995). It plays

critical roles in geminivirus DNA replication. AL1 has virus specific recognition of its

origin of replication (Choi and Stenger, 1995; Gladfelter et al., 1997; Jupin et al., 1995)

and initiates (+)-strand DNA replication (Orozco and Hanley-Bowdoin, 1996). AL1

represses its own expression at the level of transcription (Eagle, Orozco, and Hanley-

Bowdoin, 1994; Sunter, Hartitz, and Bisaro, 1993). AL1 has been found to have origin recognition and DNA binding ability, DNA cleavage or nick/ligation function, and ATP binding and ATPase function although no DNA helices activity has been reported for

AL1. Recently, it has been found to have another function, reprogramming the cell cycle by inducing proliferating cell nuclear antigen (PCNA) expression and/or interacting with pRB (retinoblastoma protein) as discussed below.

Geminiviruses can replicate in terminally differentiated cells that have shut down their DNA replication activities and do not express DNA replication enzymes. Normally, in plants only undifferentiated meristematic cells in the growing tips of shoots and roots and cells in the cambium of the stem can perform DNA replication and cell division.

Therefore geminiviruses, like many animal DNA tumor viruses, have to overcome this limitation and induce the synthesis of DNA replication enzymes in terminally differentiated cells. They accomplish this by re-programming cells back to S phase

(Cheng et al., 1995; Gutierrez, 2000). It has been reported that there is a correlation between geminivirus replication and S phase (Accotto et al., 1993). Not surprisingly, it has been found that geminiviruses can induce expression of host DNA synthesis proteins, such as PCNA in infected plants (Egelkrout, Robertson, and Hanley-Bowdoin, 2001).

PCNA is the processivity factor for DNA polymerase δ (Waseem et al., 1992). It is an essential component of the eukaryotic DNA replication complex. PCNA tethers DNA polymerases on the DNA template to accomplish processive DNA synthesis as a DNA sliding clamp. In young leaves and cycling cells, PCNA mRNA levels normally are very high, whereas in mature leaves PCNA mRNA is undetectable by Northern blot.

Geminivirus infection activates the PCNA promoter in mature leaves (Egelkrout,

Robertson, and Hanley-Bowdoin, 2001). There is a strong correlation between symptom

severity, viral DNA accumulation, PCNA promoter activity and endogenous PCNA

mRNA levels (Egelkrout, Robertson, and Hanley-Bowdoin, 2001).

The geminiviral gene product, AL1 (Rep), is responsible for this induction

(Castillo et al., 2003; Hanley-Bowdoin et al., 2000; Luque et al., 2002; Nagar et al.,

1995). AL1 is necessary and sufficient for viral DNA replication in association with host

protein factors. In non-dividing plant cells, AL1 can also induce expression of PCNA.

Furthermore, transgenic plants expressing TGMV AL1 contain detectable levels of

PCNA in differentiated cells, in which PCNA is normally undetectable as mentioned

above. This indicates that AL1 protein is sufficient for reprogramming host cells. TGMV

AL1 can directly bind to dsDNA in a sequence specific manner and may be able to regulate transcription, thus it might activate transcription of PCNA components directly.

But, AL1 may also have indirect function by interacting with plant cell cycle regulatory

components to induce replication competence in host cells. This is supported by

experimental data. First, TGMV-infected plants or transgenic plants expressing an AL1

transgene show morphological changes associated with de-differentiation of plant cells,

which suggests that AL1 can initiate the de-differentiation process (Nagar et al., 1995).

Second, WDV L1 (RepA) and TGMV AL1 can interact with pRB (retinoblastoma

protein) from human and maize (Xie et al., 1996; Xie, Suarez-Lopez, and Gutierrez,

1995). pRB is a tumor suppressor protein in mammalian cells. It binds to transcriptional

factor E2F and thereby represses the transcription of genes required for S phase

(Egelkrout, Robertson, and Hanley-Bowdoin, 2001; Herwig and Strauss, 1997). Although

there are no data demonstrating that plant pRB functions in cell cycle regulation, the

maize pRB is like the mammalian counterpart in several respects. Maize pRB contains A

and B pocket domains like mammalian pRB proteins, and can interact with SV40 large T

antigen, papillomavirus E7 and plant cyclin δ through an LXCXE motif (Ach et al., 1997;

Collin et al., 1996; Grafi et al., 1996). In animal cells, DNA tumor viruses induce host gene transcription by binding to pocket proteins and disrupting their interaction with E2F

(Weinberg, 1995). AL1 may disrupt pRb/E2F complexes that repress PCNA promoter function and promoters of other S phase proteins to induce transcription in mature leaves.

Third, AL1 can also interact with host proteins that may be involved in plant cell division and development (Hanley-Bowdoin et al., 2000).

Besides AL1, another geminivirus protein, AL3 is also required for efficient viral

DNA replication. So it is also called REn (replication enhancer). AL3 can increase viral

DNA accumulation by as much as 50-fold (Sunter et al., 1990) and can enhance viral

infection and symptoms (Hormuzdi and Bisaro, 1995; Sung and Coutts, 1995). AL3 is a

small highly hydrophobic protein of only 134 amino acids and is localized in the nucleus

(Nagar et al., 1995; Pedersen and Hanley-Bowdoin, 1994). It is very conserved among

all begomoviruses. AL3 is able to interact with AL1 (Settlage, Miller, and Hanley-

Bowdoin, 1996) and PCNA (Castillo et al., 2003), which suggests that AL3/AL1 and

PCNA form a complex during virus infection, inducing the assembly of the viral

replication complex near the viral origin of replication. But the biological consequences

of the AL1/AL3-PCNA interactions remain to be determined.

1.2 AL2 and L2 have multiple functions

1.2.1 AL2 is a transcriptional activator but L2 is not

TGMV (begomoviruses) AL2 and BCTV (curtovirus) L2 are positionally homologous between these two different virus genomes. AL2 and L2 also share limited amino acid sequence homology. Do they have the same functions? What are these functions? First of all, AL2 was originally found to be a transcriptional activator protein

(TrAP) (Sunter and Bisaro, 1991). All begomoviruses encode an allele of the AL2 gene

(or AC2; C2), which is not present in Mastreviruses. AL2 is highly conserved in begomoviruses and its transactivation function is not virus specific. AL2 transactivates virion-sense gene expression from the AR1 (coat protein) and BR1 (movement protein) promoters at the transcriptional level in a virus non-specific manner (Sunter and Bisaro,

1992; Sunter, Stenger, and Bisaro, 1994). AL2 mutants accumulate reduced amount of ssDNA and do not accumulate capsid protein (Sunter and Bisaro, 1991; Sunter et al.,

1990). Because coat protein is required for stabilizing viral ssDNA, AL2 is required for optimal accumulation of this DNA form. Since BR1 is required for systemic spread of the virus in plants, AL2 mutants also are not infectious. Therefore, AL2 mutants can replicate in plant protoplasts but cannot cause systematic infection of plants (Sunter and Bisaro,

1991; Sunter et al., 1990).

Consistent with its transactivation function, TGMV AL2 was found to contain a

C-terminal acidic-type transcriptional activation domain (Hartitz, Sunter, and Bisaro,

1999a). AL2 binds ssDNA relatively strongly but dsDNA weakly in a sequence- nonspecific manner (Hartitz, Sunter, and Bisaro, 1999a; Noris et al., 1996; Sung and 12

Coutts, 1996). In addition, AL2 also binds zinc (Hartitz, Sunter, and Bisaro, 1999). The biological significance of the zinc, ssDNA and dsDNA binding activities are presently unclear. So how does AL2 function as a transcriptional activator? It is known that the C- terminal region of AL2 (amino acids 115-129 of the 15 kDa TGMV protein) is a minimal activation domain, which is functional in mammalian, yeast and plant cells (Hartitz,

Sunter, and Bisaro, 1999). Recent studies suggest that AL2 is likely recruited to responsive promoters by protein-protein interactions (Sunter and Bisaro, 2003), similar to adenovirus E1A and herpesvirus VP16. These transactivators also do not bind specific

DNA strongly, but they are targeted by interaction with cellular factors. These factors recognize specific sequences within responsive promoters (Gerster and Roeder, 1988;

Kristie, LeBowitz, and Sharp, 1989; Lillie and Green, 1989; Liu and Green, 1994;

Triezenberg, Kingsbury, and McKnight, 1988; Triezenberg, LaMarco, and McKnight,

1988). Thus AL2 is recruited to the promoter region by interacting with other proteins, which can bind a specific sequence in the promoter.

Interestingly, it has been shown that AL2 can both activate and de-repress the coat protein gene promoter in mesophyll and phloem cells (Sunter and Bisaro, 1997; Sunter and Bisaro, 2003). In protoplasts, regulation of a minimal coat protein promoter also involves both activation and derepression by AL2 (Sunter and Bisaro, 2003). Further study has shown that a sequence located between –125 to –107 (element B) and another between –96 and –60 (element C/R) from the coat protein transcription start site are both required for AL2-mediated activation. Element C/R seems to interact with an unknown host repressor, because deletion of this sequence increases basal promoter activity in the absence of AL2. Titration of the repressor by using excess –107 to –60 sequence also

leads to the increase of basal activity in protoplasts (Sunter and Bisaro, 2003).

Furthermore, a linker scanning experiment within C/R region has identified a 6-

nucleotide sequence (K11; AATCTA; -73 to -68) which when replaced gives high basal

expression. This mutant promoter also cannot be activated by AL2. Further a mutant

TGMV with this 6-nuclotide replacement elicits mild symptoms, like coat protein

mutants (Buckley, Sunter, Buchmann and Bisaro, unpublished results).

Although L2 is a positional homologue of AL2, L2 is not a transcription factor.

L2 has no obvious activation domain. L2 mutants can systematically infect plants and

the coat protein gene expression is normal (Hormuzdi and Bisaro, 1995; Stanley et al.,

1992b). However, plants infected with BCTV L2 mutants have an abnormal phenotype in some circumstances. The phenotype is described as “enhanced recovery”, since most

Nicotiana species also show varying degrees of recovery from wild-type BCTV infection

(Hormuzdi and Bisaro, 1995). A large proportion of N. benthamiana plants inoculated

with BCTV L2 mutants appeared to recover from the infection (Hormuzdi and Bisaro,

1995). In plants exhibiting recovery, the curly top symptom appeared early in infection,

but new shoots showed a clear reduction in curly top symptoms. Recovered tissue

contained little or no viral DNA. These observations suggested that L2 may affect BCTV

replication indirectly or that L2 might play a direct role in viral pathogenesis.

1.2.2 AL2 and L2 act as pathogenicity determinants by interfering

with global metabolism

A second function of AL2 is as pathogenesis determinant that suppresses plant stress and defense responses during virus infection (Hao et al., 2003; Sunter, Sunter, and

Bisaro, 2001). Because full-length AL2 protein is toxic to plants and no transgenic plants

expressing AL2 can be generated (Sunter and Bisaro, unpublished results), transgenic N.

benthamiana and N. tabacum var Samsun plants expressing a truncated TGMV AL2

transgene (AL21-100, lacking the activation domain) were generated. Several lines were

established, and these were challenged by inoculation with TGMV, BCTV and tobacco

mosaic virus (TMV), an unrelated RNA virus. Surprisingly, a novel enhanced

susceptibility (ES) phenotype was observed in these transgenic plants. ES is mainly

characterized by a reduction in mean latent period (from 1 to 9 days depending on the

virus tested) and by a decrease in the inoculum concentration required to infect these

transgenic plants (ID50 reduced 6- to 60-fold). However, disease symptoms and virus

replication levels were not significantly increased. Interestingly, transgenic plants

expressing full length L2 showed the same ES phenotype as transgenic plants expressing

AL21-100 (Sunter, Sunter, and Bisaro, 2001). Therefore TGMV AL21-100 and BCTV L2

share the ability to suppress a host stress or defense response so that transgenic plants

expressing AL2 or L2 are more susceptible to infection by both DNA and RNA viruses.

The transgenes (AL21-100 and L2) expressed in these plants do not have transcription

activation activity, therefore defense suppression and transcriptional activation are

separate activities. Thus AL2 has transcriptional activation and pathogenesis functions while L2 shares this second pathogenesis function with AL2.

1.2.3 AL2 and L2 are RNA silencing suppressors

A third function for AL2 has also been recognized. Begomovirus AC2 (AL2) and

C2 proteins have been shown to act as suppressors of RNA silencing (Brigneti et al.,

1998; van Wezel et al., 2001; van Wezel et al., 2002; Van Wezel et al., 2003; Voinnet et

al., 1998). Briefly, RNA silencing is an adaptive host defense response. Viruses are both

the triggers and targets of the defense system, and most viruses encode RNA silencing

suppressors as a counterdefense (Brigneti et al., 1998; Voinnet et al., 1998).

Earlier studies showed that the AL2 homologues of ACMV and TYLCV are

silencing suppressors (Hamilton et al., 2002a; van Wezel et al., 2001; van Wezel et al.,

2002; Van Wezel et al., 2003; Voinnet, Pinto, and Baulcombe, 1999). Using a 35S

promoter (from Cauliflower mosaic caulimovirus) mediated transient expression system,

we found that TGMV AL2 and BCTV L2 are also silencing suppressors (Chapter 3).

Therefore, AL2 and L2 share another function. How AL2 and L2 exert their silencing

suppression function is still unclear. Studies of TGMV AL2 and BCTV L2 silencing suppression activity are discussed in Chapter 3.

1.3 SNF1, a cellular fuel gauge, is a global metabolic regulator

The SNF1 (Sucrose Non-Fermenting 1) protein kinase family comprises SNF1

itself in the yeast Saccharomyces cerevisiae, the AMP-activated protein kinase (AMPK) in mammals, and the SNF1-related protein kinases (SnRKs) in higher plants. These serine/threonine kinases are highly conserved and are essential for metabolic regulation in response to stresses in fungi, plants and animals (Hardie, Carling, and Carlson, 1998).

The physiological roles of SNF1 family members are better known in yeast and animals.

From the yeast and animal studies, SNF1 function is believed to defend cells against nutritional and environmental stresses by regulating both metabolism and gene expression.

1.3.1 SNF1 has multiple functions involved in stress responses

1.3.1.1 Yeast SNF1 is a sensor for carbon metabolism

The SNF1 gene was originally defined from mutation studies in yeast. Some

mutants that would not grow on glycerol, maltose, ethanol, sucrose, or raffinose turned

out to be alleles of the same gene, SNF1. SNF1 is required for derepression of essentially

all glucose-repressed genes (Ronne, 1995). In S. cerevisiae, the SNF1 gene encodes a 72

kD kinase with an N-terminal protein kinase domain and a C-terminal regulatory domain.

It is required for expression of glucose repressed genes in response to glucose starvation

(Celenza and Carlson, 1986; Estruch et al., 1992; Gancedo, 1992), such as the SUC2 gene

which encodes invertase (Carlson and Botstein, 1982). SNF1 mutants are unable to utilize

sucrose, galactose, maltose, melibiose, or other nonfermentable carbon sources (Estruch

et al., 1992). The SNF1 kinase complex comprises the catalytic subunit Snf1p ( subunit

of AMPK), Snf4p ( subunit of AMPK), and one of three subunits of

SIP1/SIP2/GAL83 subfamily. The AL2 and L2 interact with Snf1p subunit. SNF1 is

physically associated with a 36 kD protein, SNF4 (Celenza and Carlson, 1989) under

derepressing conditions (i.e. low to medium glucose). This interaction is necessary for

SNF1 activity. Mutations in the SNF4 gene have essentially the same phenotype as those

in the SNF1 gene (Woods et al., 1994). SNF4 stimulates kinase activity by counter-acting

auto-inhibition by the SNF1 regulatory domain (Jiang and Carlson, 1997). The subunit,

a member of SIP1/SIP2/GAL83 subfamily, forms a ‘scaffold’ on which the SNF1 and

SNF4 proteins assemble (Jiang and Carlson, 1997). It regulates the subcellular

localization of the kinase (Vincent et al., 2001) and mediates interactions with

downstream targets (Vincent and Carlson, 1999). SNF1 is activated by phosphorylation

of the activation-loop threonine residue (Estruch et al., 1992; McCartney and Schmidt,

2001; Wilson, Hawley, and Hardie, 1996; Woods et al., 1994), as is also the case for

AMPK (Hawley et al., 1996). Activation of the kinase in response to glucose limitation is accompanied by a conformational change of the kinase complex. When cells are grown under high glucose medium, SNF1 and SNF4 do not interact directly, but indirectly through the third subunit (scaffold protein). The C-terminal regulatory domain of SNF1 interacts with its own kinase domain and inhibits the kinase activity. When glucose is removed from the medium, the whole complex undergoes a conformational change.

SNF4 interacts directly with the regulatory domain of SNF1, exposing the kinase domain,

and threonine 210 of SNF1 becomes phosphorylated (Estruch et al., 1992), which makes

SNF1 kinase domain accessible to substrates.

How is yeast SNF1 activated? Removal of glucose in the yeast growth medium

causes a rapid and dramatic increase in AMP/ATP ratio (increased >200-fold within 5 min) (Wilson, Hawley, and Hardie, 1996). Although AMP does not allosterically activate

SNF1 in yeast (Woods et al., 1994), it may indirectly activate SNF1 by affecting other

metabolites that activate SNF1. Some genes appear to act upstream of SNF1 because

disruption of these genes activates SNF1 (Ludin, Jiang, and Carlson, 1998; Sanz et al.,

2000). One of these is HXK2 (HEX1) which encodes the major isoform of hexokinase

(Hardie, 1999). Disruption of this gene interferes with ATP production and thus elevates

AMP through the adenylate kinase reaction. Adenylate kinase is a ubiquitous enzyme that

contributes to the homeostasis of adenine nucleotides in eukaryotic and prokaryotic cells.

It catalyzes the reversible reaction of phosphorylation of AMP with ATP to generate

ADP.

SNF1 activation by stresses that deplete cellular ATP requires phosphorylation at

a conserved threonine residue within the activation loop of the kinase domain, but

identifying the upstream kinase(s) responsible for this has been a problem. Recently,

three yeast kinases, Pak1p, Tos3p, and Elm1p have been shown to activate SNF1 kinase

in vitro and in vivo (Hong et al., 2003; Sutherland et al., 2003). All three kinases

phosphorylate recombinant SNF1 on the activation-loop threonine. Tos3p phosphorylates

mammalian AMPK on the equivalent residue and activates the enzyme, suggesting

functional conservation of the upstream kinases between yeast and mammals. Expression

of activated, truncated Elm1p from its own promoter gave a constitutive pseudohyphal

growth phenotype that was rescued by deletion of SNF1, showing that SNF1 was acting

downstream of Elm1p. However, deletion of ELM1 does not give SNF1- mutant

phenotype. Triple deletion of the cognate genes causes a SNF1- mutant phenotype and

abolishes SNF1 catalytic activity. Therefore Elm1p, Pak1p, and Tos3p are upstream

kinases for the SNF1 complex that have partially redundant functions (Hong et al., 2003;

Sutherland et al., 2003).

What are the downstream targets for SNF1? SNF1 has many target proteins

including the metabolic enzyme acetyl-CoA carboxylase and transcriptional factors

MIG1, SIP4 and CAT8 (Hardie, Carling, and Carlson, 1998; Vincent and Carlson, 1999).

Acetyl-CoA carboxylase plays a key role in fatty acid synthesis (Woods et al., 1994). The acetyl-CoA carboxylase is inactivated when it is phosphorylated by SNF1. The MIG1

(multicopy inhibitor of GAL gene expression) gene product is a C2H2 zinc finger transcription factor binding to the promoters of glucose-repressible genes, such as GAL1,

GAL4, SUC2, CAT8 and MAL genes (Klein, Olsson, and Nielsen, 1998). When MIG1 is

phosphorylated by SNF1, it moves to the cytoplasm and loses its repression function. For

gluconeogenic structural genes, the carbon source-responsive element (CSRE) has been

identified as a common upstream regulatory motif. Gene activation by a CSRE motif

requires a functional CAT8 gene, which encodes a protein with a binuclear zinc cluster

domain. A similar function was shown for the SIP4 gene product containing a zinc

cluster with significant similarity to the CAT8 CSRE-binding domain. CAT8 and SIP4

function as transcriptional activators under derepressing growth conditions but not in the

presence of high concentrations of a fermentable sugar (Hiesinger et al., 2001; Roth,

Kumme, and Schuller, 2003). The ability of CAT8 or SIP4 to activate transcription is

regulated by glucose and requires a functional SNF1 protein kinase. CAT8 or SIP4 proteins are phosphorylated and activated in vivo in response to glucose starvation and

SNF1 is necessary for this phosphorylation (Lesage, Yang, and Carlson, 1996). SNF1 can also directly interact with the RNA polymerase II holoenzyme C-terminal domain and thus control a large number of genes (Kuchin, Treich, and Carlson, 2000). Yeast HSF

(heat shock transcription factor) protein responds to heat stress and glucose starvation.

HSF is an evolutionarily conserved protein that mediates eukaryotic transcriptional

responses to stress. SNF1 interacts with HSF in vivo and directly phosphorylates it in

vitro (Hahn and Thiele, 2003). Phosphorylation of HSF by SNF1 leads to activation of

HSF. Expression of a subset of HSF targets by glucose starvation was dependent on

SNF1 and the HSF carboxyl-terminal activation domain (Hahn and Thiele, 2003).

1.3.1.2 Mammalian SNF1 is activated by 5′ -AMP

The mammalian homologue of SNF1 is the AMP-activated protein kinase

(AMPK) (Carling et al., 1994; Mitchelhill et al., 1994). AMPK is a heterotrimeric

complex composed of three subunits, catalytic α subunit, accessory β and γ subunits

(Davies et al., 1992; Mitchelhill et al., 1994). The α subunit is closely related to SNF1

gene product from yeast. It contains a typical protein serine/threonine kinase domain

within the N-terminal domain. The C-terminal domain is less closely related to yeast

SNF1 than the kinase domain. The C-terminal domain of α subunit interacts with β/γ

subunits. The AMPK β subunit is related to the products of yeast SIP1/SIP2/GAL83

subfamily, which in the SNF1 complexes interacts with SNF1 and SNF4 and act as a

‘scaffold’ protein. The γ subunit of AMPK is closely related to the SNF4 gene product

from yeast. Therefore, structurally the mammalian AMPK is very similar to the yeast

SNF1 complex.

Furthermore, functionally mammalian AMPK is also like SNF1, because AMPK

is activated by AMP in response to stress conditions (Carling et al., 1989; Hawley et al.,

1995). Activation of AMPK has been found in response to a variety of conditions of

stress in mammalian cells which lead to high AMP and low ATP (Corton, Gillespie, and

Hardie, 1994; Moore, Weekes, and Hardie, 1991; Sato, Goldstein, and Brown, 1993).

AMP activates AMPK through multiple mechanisms to allow the system to respond in a

highly sensitive manner to an increase in AMP concentration. The activation of AMPK

by AMP can be accomplished by allosteric activation of AMPK by binding AMP

directly, or by allosteric activation of an upstream regulatory kinase. In addition, binding of AMP to AMPK blocks the dephosphorylation of AMPK by protein phosphatase 2A

(PP2A) or protein phosphatase 2C (PP2C) (Davies et al., 1995). AMPK is also phosphorylated and activated by an upstream kinase, AMPK kinase (AMPKK) (Hawley et al., 1996). AMPKK is also an AMP-activated protein kinase and binding of AMP to

AMPK promotes AMPK phosphorylation by AMPKK (Hardie and Carling, 1997).

Activation of AMPK by AMP is antagonized by high (mM) concentration of

ATP. A high AMP:ATP ratio is symptomatic of low cellular energy level. Therefore,

AMPK is a cellular ‘fuel gauge’, monitoring cellular energy status by measuring the

AMP and ATP levels, and being switched on if it detects a rise in the AMP:ATP ratio

(Hardie and Carling, 1997). When activated, AMPK acts to conserve ATP by

phosphorylating and inactivating regulatory enzymes involved in ATP-consuming, biosynthetic pathways such as acetyl-CoA carboxylase (fatty acid synthesis) (Dyck et al.,

1999; Winder et al., 1997; Witters and Kemp, 1992) and HMG-CoA reductase

(sterol/isoprenoid synthesis) (Carling et al., 1989; Gillespie and Hardie, 1992). AMPK also promotes alternative pathways of ATP production, such as fatty acid oxidation

(Hardie, 1999).

Under normal physiological ATP concentration, AMPK is not activated. Healthy cells maintain a high ratio of ATP to ADP at 10:1 and an ATP/AMP ratio of 100:1

(Hardie and Hawley, 2001). Since almost all cellular processes are coupled to ATP breakdown, it is critical for the cell to keep appropriate ratios of ATP: ADP and ATP:

AMP. The depletion of ATP by some adverse condition may result in an increased

ADP/ATP ratio, and the AMP/ATP ratio varies as the square of the ADP/ATP ratio, which represents a sensitive indicator of cellular energy status. An increase in AMP/ATP ratio switches on the AMPK system. Thus, as mentioned above, the AMPK system works like a cellular ‘fuel gauge’ which is activated by low energy status and initiates protective pathways to conserve ATP or promote alternative methods of ATP regeneration. Under environmental and nutritional stresses, ATP could be depleted and AMP levels would be elevated. These stresses include heat shock, metabolic poisons which inhibit the tricarboxylic acid cycle, glucose deprivation (hypoglycemia), oxygen deprivation

(hypoxia), and interruption of the blood supply. All of these cause ATP depletion and the classical cellular stress response (Corton, Gillespie, and Hardie, 1994; Salt et al., 1998).

Although it is unclear how AMPK induces stress protein expression, it is believed that activation of AMPK is a central component in the response of cells to stresses. Therefore,

the signals that activate the AMPK kinase cascade are high AMP concentration coupled

with low ATP concentration (Davies et al., 1995; Hawley et al., 1995).

1.3.1.3 Plant SNF1 is a bridge between carbon metabolism and

stress defense response in plants

The plant SNF1-related protein kinases, referred to as SnRKs, are related to the

AMPKs in animals and the SNF1 kinase in yeast (Alderson et al., 1991; Halford and

Hardie, 1998). AMPKs and SNF1 regulate many cellular responses to environmental and nutritional stresses (Halford and Hardie, 1998; Halford et al., 2003; Hardie, Carling, and

Carlson, 1998). Plants possess yeast SNF1 homologs that can complement the SNF1

deletion mutation in yeast (Alderson et al., 1991; Bhalerao et al., 1999). There are three

subfamilies of plant SNF1-related protein kinases (SnRK): SnRK1, SnRK2 and SnRK3,

among which SnRK1 is most homologous to mammalian AMPK and yeast SNF1

(Halford et al., 2003). SnRK2 and SnRK3 proteins show 42-45% amino acid sequence

identity with SnRK1, SNF1 and AMPK in the catalytic domain. SnRK2 and SnRK3 sub-

families are relatively large and diverse compared to SnRK1 sub-family. Therefore, they

are significantly less similar to SNF1 and AMPK than SnRK1 is. In Arabidopsis there are

2 SnRK1 genes, 10 SnRK2, and 29 SnRK3 genes. Two isoforms, Arabidopsis protein

kinase 10 (AKIN10) and AKIN11 in SnRK1 subfamily have been experimentally identified in Arabidopsis (Bhalerao et al., 1999). They share 89% amino acid identity in

the N-terminal portion which contains the kinase catalytic domain and 64% identity in

the C-terminal regulatory domain. A potential third isoform, At5g39440, which is closely

related to AKIN11, was identified by sequence homology comparison (Wang, Harper,

and Gribskov, 2003). AL2 and L2 have been shown to interact with AKIN10 and AKIN

11, both SnRK1 type kinases. Therefore this discussion is focused on SnRK1, which is

also most related to SNF1 in yeast and AMPK in mammals. SnRK1 kinases from

different plant species have been shown to complement yeast SNF1- mutant strains and allow growth on non-fermentable carbon sources (Hardie, Carling, and Carlson, 1998), which suggests that SnRK1 is functionally and structurally related to yeast SNF1.

RKIN1, isolated from a rye endosperm cDNA library was the first SnRK1 kinase identified (Alderson et al., 1991). RKIN1 encodes a 57.71 kDa protein of 502 amino acid residues with 48% overall amino acid sequence identity with yeast SNF1 and AMPK, and 62-64% identity in the kinase catalytic domains. AMPK is slightly larger than

RKIN1, with a size of 63 kDa, while SNF1 is the largest with 72 kDa. The size difference is due to the variation in the regulatory C-terminal regions.

How is plant SnRK1 activity is regulated? Members of SnRK1 subfamily of protein kinases are higher plant homologues of mammalian AMPK and yeast SNF1 protein kinases. Based on analogies with the mammalian system, it has been found that the spinach SnRK1 kinases is regulated by phosphorylation on a threonine, equivalent to

Thr175 in Arabidopsis thaliana SnRK1 (AKIN10), within the 'T loop'. Inactivation of two spinach SnRK1 kinases is associated with changes in the phosphorylation state of this threonine. Dephosphorylation of this threonine by protein phosphatases, and consequent inactivation, is inhibited by low concentrations of 5'-AMP, via AMP binding to the substrate (i.e. the SnRK1 kinase). Thus like yeast SNF1, plant SnRKs do not seem

to be activated directly by AMP, but AMP appears to inhibit dephosphorylation and the concomitant inactivation of spinach SnRK activity at physiological concentrations

(Sugden et al., 1999a). However, it is not yet known if an upstream kinase can be activated by AMP and then activates SnRKs. Nevertheless, AMP:ATP ratio is important to regulate plant SnRKs.

Other mechanisms can also be involved in regulating plant SnRK activity. For example, glucose-6-phosphate may inhibit SnRK1 activity (Toroser, Plaut, and Huber,

2000). From its substrates and the genes it regulates, it has been suggested that SnRK1 is activated in response to high intracellular sucrose and/or low intracellular glucose levels

(Slocombe et al., 2002), in addition to high AMP:ATP ratios. Thus SnRK1 is regulated in a complex manner by cellular energy charge (AMP level) and glucose levels.

There is a convenient assay to measure SnRK1 activity because SnRK1 has been shown to phosphorylate the SAMS peptide (His Met Arg Ser Ala Met Ser Gly Leu His

Leu Val Lys Arg Arg), a synthetic peptide based on the sequence in the primary phosphorylation site for AMPK on rat acetyl-CoA carboxylase (Mackintosh et al., 1992).

SnRK2 and SnRK3 appear to have slightly different recognition sequences (Halford et al., 2003)

SnRK can phosphorylate and inactivate many key metabolic enzymes in plants,

The first identified substrate for SnRK1 in Arabidopsis is HMG-CoA reductase (Dale et al., 1995). The phosphorylation site is Ser-577 and phosphorylation causes inactivation of the enzyme. SnRK1 has also been shown to phosphorylate and inactivate a bacterially- expressed recombinant Arabidopsis HMG-CoA reductase (Ball et al., 1994). Two other important enzymes have also been demonstrated to be substrates for spinach SnRK1.

They are sucrose phosphate synthase (SPS) and nitrate reductase (NR) (Sugden et al.,

1999b). The phosphorylation site is Ser-158 for SPS and Ser-543 for NR (Bachmann et al., 1996b; Douglas, Morrice, and MacKintosh, 1995). In both cases, phosphorylation leads to inactivation of the enzymes (Bachmann et al., 1996a; Moorhead et al., 1996).

Transgenic potatoes expressing antisense SnRK1 show up to a 79% reduction in SAMS

peptide kinase activity and a reduction of sucrose synthase activity of up to 64% (Purcell,

1998). Increased invertase gene expression also has been observed in transgenic

Arabidopsis over-expressing SnRK1 (Halford et al., 2003). Transient expression of

antisense SnRK1 results in a repression of α-amylase promoter activity in wheat embryos

(Halford et al., 2003). Besides its function in metabolism, SnRK1 is also involved in other aspects of cell growth. For example, over-expression of SnRK1 in yeast results in reduction of yeast cell size, which suggests that SnRK1 might play a role in cell cycle signaling (Dickinson, 1999). In plants, SnRKs play an important role in carbon metabolism by directly phosphorylating and inactivating the above biosynthetic key

enzymes. Thus SnRK1 is a key player in controlling carbon metabolism (Halford and

Hardie, 1998; Halford et al., 2003).

There is also reason to believe that SnRK1 is involved in host defense. Elevated

sugar concentrations in plants can condition resistance to pathogen attack or other

environmental stresses by inducing the expression of stress-related genes such as the PR

(pathogenesis-related) genes. Expression of yeast invertase in plants leads to increased

levels of monosaccharides and induction of PR proteins (Herbers et al., 2000; Roitsch,

1999; Roitsch et al., 2003). It has been proposed that elevated hexose levels during virus

infection in tobacco plants may enhance the expression of defense-related functions and

might possibly explain the phenomenon of high sugar resistance in plants (Herbers et al.,

1996; Herbers et al., 2000; Roitsch, 1999; Roitsch et al., 2003). Cell wall invertase- expressing transgenic tobacco plants were found to be resistant to potato virus Y (PVY)

(Herbers et al., 1996). After PVY infection, tobacco leaves started to accumulate soluble sugars and leaf photosynthesis decreased. The accumulation of soluble sugars was accompanied by a gradual decrease in the sucrose-to-hexose ratio, and increase of transcripts encoding PR-proteins (Herbers et al., 2000). Many environmental stresses lead to major alterations in carbohydrate metabolism, and it is possible that the cross talk between sugar signaling pathways and stress pathways regulated by SnRK1 is a part of the defense response in plants. This is relevant to our finding that geminivirus AL2 and

L2 inactivate SNF1 activity, and that this causes enhanced susceptibility to virus infection (Hao et al., 2003), which is discussed in detail in the following section. Based on these observations, viruses might interfere with plant metabolism to counter a plant defense system.

1.3.2 SNF1 is inactivated by geminivirus AL2 and L2 proteins

The experiments presented in this section describe a project that I was involved

in, especially in the infectivity experiments. These experiments suggest that the SNF1-

mediated metabolic stress response plays a critical role for plant defense response to virus infection, and that geminivirus AL2 and L2 proteins inactivate SNF1, leading to an enhanced susceptibility to virus infection.

1.3.2.1 AL2 and L2 interact with SNF1

To further study the ES phenotype caused by AL2 and L2 (Sunter, Sunter, and

Bisaro, 2001), host proteins that can interact with both AL2 and L2 were sought using

yeast two-hybrid system with AL21-83 as bait. By screening a yeast two-hybrid

Arabidopsis cDNA library, SNF1 kinase was isolated. BCTV L2 also interacts with

SNF1. The interaction was also observed when SNF1 was used as bait with full-length

AL2 or L2 as prey. AL2 and L2 also show weak interaction with yeast SNF1 but not with other control proteins including CDK2, p53 and TGMV BL1, BR1 and coat protein. This

interaction is not a kinase-substrate relationship because both AL2 and L2 lack a

consensus SNF1 phosphorylation site (Weekes et al., 1993). The major interaction region

was defined to a small conserved region of AL2 (amino acid 30 to 43) and L2 (amino acid 66-79), which contains a putative zinc finger (Hartitz, Sunter, and Bisaro, 1999).

The isolated SNF1 is identical to Arabidopsis protein kinase 11 (AKIN11) (Hao et al.,

2003).

1.3.2.2 AL2 and L2 condition pathogenesis by inactivating SNF1

Several lines of evidence indicate that AL2 and L2 cause ES by inactivating

SNF1. Expression of antisense SNF1 phenocopies the expression of the viral proteins

AL2 and L2 (Hao et al., 2003). Transgenic N. benthamiana plants expressing an

antisense SNF1 construct are more susceptible to TGMV and BCTV infection than wild- type non-transgenic plants, as measured by a reduction in mean latent period (time to first

appearance of symptoms) and ID50 (amount of virus inoculum required to cause 50% of

plants to show symptoms). The disease symptoms and virus load in transgenic plants

expressing antisense SNF1 or expressing viral AL21-100 and L2 are not increased. In

contrast, constitutive expression of a sense SNF1 transgene results in enhanced resistance

to geminivirus infection (Hao et al., 2003).

The AL2-mediated ES phenotype depends on the ability of AL2 to interact with

SNF1, because transgenic plants expressing SNF1 interaction-defective version of AL2

(AL21-32;44-114) do not exhibit enhanced susceptibility. The mutant AL2 protein (deletion of 33-44 residues) has a reduced ability to interact with SNF1, as determined in the yeast

two-hybrid system (Hao et al., 2003).

AL2 and L2 inhibit SNF1 kinase activity in vitro. Bacterial expressed

recombinant SNF1 kinase domain (SNF1-KD) polypeptide has autophosphorylation

activity. But addition of purified AL2 or L2 proteins to the kinase assay resulted in a

strong inhibition of SNF1-KD activity, with a nearly complete inhibition observed at a

2:1 molar ratio of AL2 to SNF1-KD. However, addition of AL2 mutant protein (∆33-44)

caused weak inhibition of SNF1-KD activity (Hao et al., 2003).

Arabidopsis SNF1 or SNF1-KD can complement yeast SNF1- mutants. However,

when the yeast SNF1- mutants were cotransfected with a SNF1 or SNF1-KD expression

plasmid and a plasmid expressing L2, the complementation was abolished and cells were

unable to grow on the glycerol-containing medium. L2 itself does not affect growth of the

same cells on medium containing glucose as a carbon source. Therefore, functional L2-

SNF1 interaction can occur in vivo in a eukaryotic cell (Hao et al., 2003).

Geminivirus AL2 and L2 proteins cause enhanced susceptibility when expressed

in transgenic plants. The above genetic and biochemical evidence proved that enhanced susceptibility is attributable to the interaction of AL2 and L2 with SNF1 kinase, a global regulator of metabolism. Specifically, AL2 and L2 inactivate SNF1 in vitro and in vivo, and expression of an antisense SNF1 transgene in N. benthamiana plants causes enhanced susceptibility similar to that conditioned by the AL2 and L2 transgenes, whereas SNF1 overexpression leads to enhanced resistance. Furthermore, transgenic plants expressing an AL2 protein that lacks a significant portion of the SNF1 interaction domain do not display enhanced susceptibility. Together, these observations suggest that the metabolic alternations mediated by SNF1 are a component of innate antiviral defense and that SNF1 inactivation by AL2 and L2 is a counterdefensive measure. These data also indicate that geminiviruses are able to modify host metabolism to their own advantage, and provide a molecular link between metabolic status and inherent susceptibility to viral pathogens

(Hao et al., 2003).

1.4 Adenosine kinase generates 5′-AMP and has potential

functions in plant defense responses

As mentioned above, in order to further understand the functions of AL2 and L2

involved in pathogenesis, we looked for host proteins that can interact with both AL2 and

L2 in yeast two-hybrid system, initially using AL21-100 as bait. In addition to SNF1, we

found that ADK (Adenosine kinase) interacts with viral AL2 and L2 proteins. Adenosine

kinase (ADK, ATP:adenosine 5′ phosphotransferase; EC 2.7.1.20) is an abundant housekeeping enzyme in eukaryotes. ADK activity has not been detected in prokaryotes

(Hove-Jensen and Nygaard, 1989).

ADK is constitutively expressed and catalyzes the transfer of the γ-phosphate from

ATP to adenosine to generate 5′ -AMP. ADK is a key component in the salvage pathways of both adenine and adenosine. These salvage pathways could prevent the possible inhibitory concentrations of these purines and efficiently recycle adenine and adenosine into the adenylate pools. Therefore ADK has a very important biological role in nucleotide metabolism, including the synthesis of nucleotide co-factors and nucleic acids.

In addition, the AMP generated by ADK could be a significant regulator of SNF1 activity. Thus the known roles of ADK in mammalian, yeast, and plant metabolism will be discussed.

1.4.1 ADK has important pharmacological roles in mammals

In mammals, adenosine kinase plays an important role in the regulation of intra-

and extracellular levels of adenosine (Berne, 1993; Chang et al., 1983). Adenosine has

widespread effects on the cardiovascular, nervous, respiratory, and immune systems (Fox

and Kelley, 1978). For example, elevated levels of adenosine in heart and brain are

associated with the attenuation of ischemic injury and in animal models the inhibition of

adenosine kinase has been shown to markedly increase the therapeutic effect of adenosine

(Jiang et al., 1997; Kowaluk, Bhagwat, and Jarvis, 1998; Martin et al., 1997; Tatlisumak

et al., 1998). Thus, since the first human adenosine kinase gene was cloned in 1996 32

(Singh et al., 1996; Spychala et al., 1996), most studies of ADK have been focused on its inhibitors (McGaraughty et al., 2001; McGaraughty, Cowart, and Jarvis, 2001; Wiesner et al., 1999; Zhu et al., 2001). ADK inhibitors may play important pharmacological roles by increasing intravascular adenosine concentrations and acting as anti-inflammatory agents.

1.4.2 ADK has a critical role in methyl cycling in yeast

In yeast, inactivation of the single adenosine kinase gene (ADO1) leads to a slow- growth phenotype. ADK does not play a major role in adenine utilization in yeast.

However, ADK-deficient mutants of S. cerevisiae accumulate S-adenosyl homocysteine

(SAH) and thus are affected in utilization of S-adenosyl methionine (Iwashima et al.,

1995) (Figure 1.3). Because adenosine is not efficiently synthesized from adenine in yeast, the physiological role of ADK in S. cerevisiae could primarily be to recycle adenosine produced by methyl cycle (Lecoq et al., 2001). Thus, the slow-growth phenotype of ADO1 mutants is likely due to a reduced cellular transmethylation activity.

1.4.3 ADK is involved in cytokinin metabolism and maintenance

of transmethylation in plants

1.4.3.1 ADK in nucleotide metabolism and its relationship with

SNF1

The first plant ADK was cloned from the moss Physcomitrella patens in 1998 by

a functional complementation of an E. coli purine auxotrophic strain (von

Schwartzenberg et al., 1998). Arabidopsis ADK genes have also been cloned. Two

isoforms, ADK1 and ADK2, sharing 92% amino acid identity were identified (Moffatt et

al., 2000). Both ADKs are expressed constitutively in Arabidopsis plants with the highest

ADK mRNA levels in stem and root, but generally ADK1 transcript level is higher than

that of ADK2. ADK enzyme activity levels are correlated with protein expression levels,

with high activity in stems, leaves, and flowers, and low activity in root, siliques and dry

seeds.

ADK activity in plants has roles in the synthesis of a variety of biomolecules

including nucleotide cofactors and nucleic acids. ADK is involved in the salvage

pathways of both adenine and adenosine. ADK activity decreases the intracellular level

of adenosine (directly) and adenine (indirectly) that may otherwise affect the activity of

other enzymes.

ADK activities may also convert cytokinin bases and ribosides to their

corresponding nucleotides. Cytokinin (isopentenyladenosine) is the primary hormone regulating plant cell division, which occurs only in specialized regions and tissues, such 34

as meristems, developing embryos, leaf primordia, and the vascular cambium. Cytokinins

have also been found to participate in regulating the following processes: initiation of

chloroplast development, delay of leaf senescence, enhancement of cotyledon cell

expansion, mobilization of nutrients and initiation of buds (Kakimoto, 2003). Cytokinin

bases and possibly ribosides are the active forms of cytokinins, whereas phosphorylated

forms are generally inactive. Thus cytokinin interconversion by phosphorylation-

dephosphorylation may be important in regulating the active level of this hormone in

plants. Moss ADK was first suggested to be involved in cytokinin metabolism because it

can transfer phosphate from ATP to cytokinin to form isopentenyladenosine

monophosphate (von Schwartzenberg et al., 1998). However, the real mechanism of

cytokinin regulation has not yet been established.

The basic function of ADK is to transfer a phosphate group from ATP or GTP to

its substrate adenosine to generate 5′-AMP. Both Arabidopsis ADKs have a Km for

adenosine of 0.3 to 0.5 µM and a Km for riboside N6 (isopentenyl) adenosine of 3 to 5

uM. Thus adenosine is the preferred substrate for both ADK enzymes to generate 5′-

AMP. As discussed above, SNF1 can be activated by 5′-AMP. Mammalian AMPK is an

AMP activated kinase which directly binds AMP and gets activated. Plant SnRK also binds AMP, and both yeast SNF1 and plant SnRKs are indirectly activated by AMP.

AMP is generated by ATP depletion during stress conditions, but AMP can also be produced by ADK. Therefore, ADK might play an important role in regulating SNF1 activity.

1.4.3.2 ADK in transmethylation and its relationship with gene

silencing

Maintaining flux through the methyl cycle is critical for cells to sustain

transmethylation reactions that use S-adenosyl-L-methionine (SAM) as a cofactor (Figure

1.3). S-adenosyl-L-homo-cysteine (SAH), an inhibitor of transmethylases, is generated as a result of SAM-dependent methylation reactions. Hydrolysis of SAH is carried out by

SAH hydrolase (SAHH), yielding homocysteine and adenosine. This reaction is reversible, and equilibrium lies in the direction of SAH synthesis. Thus it is critical that either homocysteine or adenosine is removed in order to prevent pathway inhibition

(Figure 1.3). In most cases, the affinity of a methyltransferase for SAH is higher than for

its substrate SAM. Therefore it is proposed that transmethylation activity is regulated by

the ratio of SAM to SAH (Duerre and Briske-Anderson, 1981). Thus adenosine kinase

activity is critical for maintaining flux through the methyl cycle and sustaining

transmethylation reactions. In support of this idea, it has been shown that Arabidopsis

SAHH activity in vitro can be inhibited up to 25% by 0.5 uM adenosine and up to 70%

by 5 µM adenosine (Moffatt et al., 2002). In addition, in vitro, SAH, 3-deazaadenosyl-

homo-cysteine, and sinefugin, a naturally occurring analog of adenosyl-homo-cysteine,

were found to be competitive inhibitors for guanidoacetate methyltransferases purified

from pig liver with Ki values of 16, 39, 18 µM, respectively. This indicates that

methyltransferases with SAM as a methyl donor are inhibited by SAH (Im, Chiang, and

Cantoni, 1979).

One way to examine whether changes in adenosine level could cause SAH level changes in plants is to make transgenic plants expressing antisense or sense ADK. ADK

deficient Arabidopsis lines were generated using antisense ADK1 genes driven by

enhanced 35S promoter (Moffatt et al., 2002). Because of the high nucleotide sequence

identity of the two ADK genes (89%), using antisense ADK1 gene could reduce

expression of both ADK genes. In the 4 antisense lines examined, all plants were very

small and contained lower ADK protein and activity levels than wild type plants (Moffatt et al., 2002). Interestingly, attempts to over-express ADK in transgenic plants containing

sense ADK1 gene driven by 35S promoter were not successful. In 8 examined sense lines, the same small plant phenotype and lower ADK activity was observed as in antisense lines. In these lines, accumulation of adenosine because of reduced ADK activity is expected to result in an increased SAH level (Figure 1.3). As expected, in ADK-deficient lines (4 sense ADK lines), SAH levels were elevated. For example, in lines with less than

20% ADK activity, SAH levels were increased about 40-fold relative to wild-type plants.

In lines with 35% ADK activity, SAH levels were about 14 times higher than wild type.

SAM levels were correspondingly decreased in these lines. Other studies in mammals and yeast have also found that increased levels of adenosine due to ADK deficiency also result in increasing levels of SAH (Iwashima et al., 1995; Kredich and Martin, 1977;

Lecoq et al., 2001; Moffatt et al., 2002).

To assess SAM-dependent methylation, the level of pectin methylation in the mucilage of ADK deficient and wild type seeds was compared by staining with ruthenium red (RR). RR staining is an indicator of methylation because methyl- esterification can reduce the intensity of pectin staining. The RR staining data indicated

that the pectin in seed mucilage of the ADK deficient lines is less methyl-esterified than that of wild type seeds (Moffatt et al., 2002). The most deficient line tested (sADK 7-4, a sense line), which has about 7% residual ADK activity, had 56% less methyl- esterification in its seed mucilage than wild type (Moffatt et al., 2002). These data suggest that ADK deficient transgenic lines have a defect in methyl recycling and transmethylation activities due to adenosine inhibition of SAHH activity.

ADK deficiency in the ADK sense or antisense lines was apparently due to ADK gene silencing (transcriptional or posttranscriptional). However, because maintenance of gene silencing depends on methylation of DNA and protein (see below), these ADK deficient plants were not able to maintain ADK gene silencing. Therefore, although ADK deficient plants have abnormal phenotypes, such as stunting, wrinkled leaves, and shorter stamen filaments, at each generation about 10% to 15% of the plants appeared to revert to a wild type phenotype with normal ADK activity and protein levels (Moffatt et al., 2002).

This indicates that ADK activity is important in transmethylation and maintenance of silencing, as discussed later.

In addition to the requirement of ADK activity in maintenance of gene silencing, stress or developmental processes that generate methyl demand also require increased

ADK activity. For example, under salt stress, spinach (Spinacia oleracea) and sugar beet

(Beta vulgaris) have increased level of the osmolyte glycine betaine. Synthesis of glycine betaine requires SAM-dependent methylation of its precursor, phosphocholine. In contrast, tobacco (N. tabacum) and canola (Brassica napus) do not accumulate glycine betaine under salt stress. When spinach, sugar beet, tobacco and canola plants were treated with high salt, ADK activity and SAHH activity were increased dramatically in

extracts from the leaves of spinach and sugar beet plants but not in extracts from tobacco and canola (Weretilnyk et al., 2001). This suggests that under conditions requiring methylation (‘methyl stress’), ADK levels are increased in order to sustain flux through the methyl cycle.

DNA methylation is an important DNA modification accomplished by DNA methyltransferases which catalyze the transfer of a methyl group to bases within the

DNA helix (for review, see Finnegan and Kovac, 2000). Cytosine-5-methyltransferases play major roles in DNA methylation by transferring a methyl group from SAM to carbon 5 of cytosine residues. Methylcytosine can occur in any sequence context in plant

DNA, but is most common in cytosines within CpG or CpNpG sequences (Gruenbaum et al., 1981). Methylation patterns are established by de novo methyltransferases and transmitted through DNA replication by maintenance methyltransferases (Bird, 1978).

Arabidopsis MET1 methyltransferase is a major maintenance methyltransferase. It restores the parental pattern of cytosine methylation in CpG sequences to the newly replicated daughter strands. The DNA methylation locus DDM1 (decrease in DNA methylation) has been suggested to be involved in a chromatin remodeling step that allows methyltransferases to access DNA (Finnegan and Kovac, 2000).

DNA methylation is important in regulating gene expression during development and is associated with transcriptional gene silencing (TGS) and posttranscriptional gene silencing (PTGS). Methylation of the transgene promoter correlates with TGS (Razin,

1998), whereas methylation of the coding sequence is associated with PTGS

(Baulcombe, 1996). Both ddm1 and met1 mutants exhibit impaired TGS and impaired

PTGS, which correlates with a decrease in transgene methylation (Jeddeloh, Bender, and

Richards, 1998; Morel et al., 2000; Steimer et al., 2000). In addition, DNA methylation is

tightly correlated with histone H3 methylation and methylation of histone H3 at lysine 9

(H3K9) is considered to be crucial for heterochromatin assembly, whereas methylation of

H3 at lysine 4 (H3K4) is generally associated with actively transcriptional chromatin

(Lachner et al., 2001; Rice and Allis, 2001; Richards and Elgin, 2002).

Unlike MET1 or DDM1, ADK is indirectly involved in DNA or histone

methylation by sustaining methyl supply. Methyl supply can be blocked by increasing

SAH levels, which can be achieved by reducing SAHH levels by expression of an

antisense SAHH transgene in plants or by reducing ADK activity to cause accumulation

of adenosine and indirectly accumulate SAH (Figure 1.3).

To examine whether DNA methylation is affected by blocking methyl supply,

transgenic tobacco plants expressing antisense SAHH driven by the 35S promoter were

generated (Tanaka et al., 1997). The overall methylation status of genomic DNAs from these transgenic or wild type tobacco plants was examined by digesting their genomic

DNA with methylation-sensitive or -insensitive restriction endonucleases. To help

observation, a repetitive element, HRS60.1, which is a highly repeated sequence in

tobacco comprising about 2% of the total genome sequence (Koukalova et al., 1989), was

used as a probe in the genomic Southern blot. As expected, the digestion pattern showed

that cytosine methylation in CpNpG or CpG sequences was reduced in SAHH antisense

tobaccos as compared with wild type plants (Tanaka et al., 1997). Interestingly, half of

more than 50 transgenic plants showed distinct morphological changes, with the most

significant change being dwarfing. It is likely that reducing ADK activity has a similar

effect on DNA methylation.

1.5 Plant defense responses to virus infection

All viruses, whether the viral genome consists of DNA or RNA, and whether it is single or double-stranded, have a basic strategy of invading the host cell in order to exploit the macromolecular and enzymatic machinery for their own purposes. Because viral infections are normally deleterious to invaded host cells, hosts have evolved a variety of mechanisms to counter viruses. In plants, a general barrier to viral infection is provided by the cell walls. Plant epidermis also is covered by wax. In addition to this passive defense, hosts have developed various active cellular defense strategies to combat viruses once they have reached the cytoplasm. There are two kinds of plant resistance: one is nonhost resistance and the other is host resistance. Nonhost resistance, in which one pathogen can infect one plant but not other plants, is still poorly understood. Host resistance is most studied and some mechanisms are beginning to emerge. These can be placed in two general classes: innate defense responses and adaptive defense responses.

Innate defense refers to virus-nonspecific defense mechanisms that a host uses immediately or within several hours after exposure to a pathogen. This is a general defense response against a broad range of pathogens and it is the initial response by plants to prevent infection. Adaptive (acquired) defense refers to specific defense mechanisms that take longer to become protective and are designed to limit or remove a specific pathogen.

1.5.1 Plant innate defense responses – the hypersensitive response

and systemic acquired resistance

Some pathogens kill host tissue to extract nutrients (necrotrophs), while others are

highly co-evolved and require living host tissue to complete their life cycles (biotrophs).

The innate immune response is the first line of defense against these infections. The

innate immune system responds to shared features of different pathogens. Plant-microbe

interactions begin with specific recognition. As well as a large array of recognition gene

functions, a number of subsequent signal transduction steps are necessary to generate a

completely effective resistant phenotype. These steps include calcium signaling, an

oxidative burst, kinase cascades, and activation of a complex defense response consisting

of a variety of antimicrobial proteins (Dangl and Jones, 2001; Lam, Kato, and Lawton,

2001).

Genetic analyses have demonstrated that recognition functions are provided by dominant disease resistance genes (or R-genes) in the plant whose action is triggered, either directly or indirectly, by the product of a pathogen avirulence (avr) gene. This is a

genetic explanation for interactions between plants and all classes of pathogens: fungal,

bacterial, viral, and insect. In the absence of either the plant R allele, or the corresponding

pathogen avr gene, recognition is lost and disease results. A variety of R genes were

recently cloned from various plant species. R gene products all share structural features, such as leucine-rich repeats (LRR) and nucleotide binding sites (NBS). Accumulating evidence suggests that diversifying selection acts on a subset of residues within LRR domains, which thus might act as ligand interaction domains. 42

Many different pathogen-encoded avr genes have been identified, but their

sequences do not readily predict function. When corresponding R and avr genes are

present in both host and pathogen, the result is disease resistance. If either is inactive or

absent, disease results (Flor, 1971). A simplest mechanistic model is that the avr gene encodes a ligand that is recognized by the product of the matching R gene, which then triggers a hypersensitive response (HR) and disease resistance (Bent, 1996). HR is one type of programmed cell death (PCD) associated with the induction of local and systemic

defense responses. It is normally controlled by direct or indirect interactions between

pathogen avirulence gene products and those of plant resistance genes leading to rapid

cell death at the immediately surrounding infection sites (Agrios, 1997).

In addition to this immediate HR response, plants have other general resistance

mechanisms induced after HR or during a successful infection to combat secondary

infection from a broad spectrum of pathogens or to prevent an existing infection from

spreading further. One of such innate defense mechanism is known as systemic acquired

resistance (SAR), an inducible disease defense system. SAR is induced in many species

upon local infection by necrogenic pathogens and by the hypersensitive response (Ryals

et al., 1996). During SAR, an increased ability to resist attack from a wide array of

pathogens is systemically induced, which lasts several weeks to several months after

initiation. Induced expression of a subset of PR genes, called SAR genes, is highly

correlated with SAR (Uknes et al., 1992; Ward et al., 1991). SAR is believed to be a

result of the concerted activation of PR genes (Uknes et al., 1992; van Loon and van

Kammen, 1970; Ward et al., 1991; Yalpani et al., 1991). In Arabidopsis, the PR-1 gene is

the most reliable molecular marker for SAR. SAR requires the signal molecule salicylic

acid (SA), which accumulates in plants before the onset of SAR. Over-expression of a

single PR gene only provides limited protection to the plants (Alexander et al., 1993; Liu

et al., 1994). It has become evident that plants utilize multiple pathways to transduce

pathogenic signals to activate HR, SAR, and other resistance responses. Besides SA- mediated SAR, jasmonic acid (JA) and ethylene are also important signals in the

induction of broad resistance against pathogens (Dong, 1998; Ellis, Karafyllidis, and

Turner, 2002; Fluhr and Kaplan-Levy, 2002; Nurnberger and Scheel, 2001; Ryals et al.,

1995; Thomma et al., 2001; Vijayan et al., 1998).

Another interesting observation is that defense gene activation is very often

correlated with high levels of soluble sugars (Johnson and Ryan, 1990; Tsukaya et al.,

1991). It has been found that plant invertase mRNA levels increase rapidly after pathogen

inoculation (Sturm and Chrispeels, 1990). Furthermore, transgenic tobacco plants that

over-express yeast invertase show a condition that resembles systemic acquired resistance

(Herbers et al., 1996). So there is a correlation between metabolic status and resistance

states, but so far little is known. The inactivation of SNF1 by viral AL2 and L2 combined

with the ES phenotype caused by over-expression of AL2/ L2 or antisense SNF1 further

suggests that metabolic changes are a component of the innate defense response.

1.5.2 An adaptive plant defense response - RNA silencing

Recently, a new plant defense against virus infection was found, which is RNA

silencing, also known as co-suppression, RNA interference, post-transcriptional gene

silencing (PTGS) and quelling. RNA silencing was first discovered in transgenic plants 44

expressing transgenes driven by strong promoter, such as the 35S promoter. In such

plants the highly expressed mRNA was inactivated or degraded in a sequence-specific

manner in the cytoplasm (Napoli, Lemieux, and Jorgensen, 1990; van der Krol et al.,

1990). Later on it was found that RNA silencing is a universal phenomenon in

eukaryotes. It is highly conserved in plants, in fungi such as Neurospora, and in animals

such as Caenorhabditis elegans and Drosophila (Ahlquist, 2002; Grishok and Mello,

2002; Hannon, 2002; Pickford et al., 2002; Zamore, 2002). In the laboratory, RNA silencing is a convenient and reliable method to knock down gene expression at the mRNA level. It causes sequence specific degradation of RNA in the cytoplasm

(Baulcombe, 1999; Matzke et al., 2001; Vance and Vaucheret, 2001; Waterhouse, Wang, and Lough, 2001). One role of RNA silencing is to act as a host adaptive defense mechanism targeted against invasive or mobile RNA elements such as viruses, or transposable retro-elements. In plants, RNA silencing is also known as posttranscriptional gene silencing (PTGS) because transcription in the nucleus is normal.

There is another type of silencing, called transcriptional gene silencing (TGS), in which no functional transcription occurs. Recent data indicate that RNA silencing and TGS share common features, including DNA methylation.

1.5.2.1 Experimental methods to trigger RNA silencing in plants

Two convenient methods have been developed to trigger RNA silencing and

knock out (or knock down) a target gene in plants. One uses a virus-based vector (VIGS

system, virus-induced gene silencing system) and another is an Agrobacterium based

transient expression system. VIGS has been used as a genetic method to suppress

endogenous gene expression by infecting plants with a recombinant virus vector (VIGS

vector, such as recombinant potato virus X or tobacco rattle virus) carrying a host-derived

sequence (Dalmay et al., 2000; Dinesh-Kumar et al., 2003; Voinnet, Lederer, and

Baulcombe, 2000). Thus, VIGS has offered a convenient means of gaining insight into gene function in plants. But VIGS is limited by virus host range and has the undesirable side effect of virus-caused symptoms.

Another method used to trigger RNA silencing involves an Agrobacterium- mediated transient expression system. In this case, Agrobacterium cells harboring a Ti

(tumor inducing) plasmid expressing a gene of interest from a strong promoter will cause

silencing of the introduced gene and the corresponding endogenous gene. A variation of

this method uses an inverted repeat sequence that is transcribed into hairpin dsRNA form.

Such dsRNA constructs contain an intron and polyadenylation signals, which presumably

facilitate entry of the transcripts into the mRNA export pathway. The dsRNA produced

by such constructs is an exceptionally strong silencing trigger (Brigneti et al., 1998;

Chuang and Meyerowitz, 2000; Llave, Kasschau, and Carrington, 2000; Schweizer et al.,

2000; Smith et al., 2000; Voinnet et al., 1998; Waterhouse, Graham, and Wang, 1998).

1.5.2.2 RNA silencing mechanisms

RNA silencing is efficiently triggered by double stranded RNA (dsRNA)

(Baulcombe, 2002; Hamilton et al., 2002; Hamilton and Baulcombe, 1999; Mourrain et al., 2000; Tenllado and Diaz-Ruiz, 2001). These dsRNAs are proposed to be synthesized

from aberrant transgene mRNAs by the action of a plant encoded, or virus encoded RNA-

dependent RNA polymerase (RdRP) (Dalmay et al., 2000b; Mourrain et al., 2000). The only RNA molecules normally found in the cytoplasm of a cell are molecules of single-

stranded mRNA. If the cell finds molecules of dsRNA, it uses an RNase III-like dsRNA-

specific ribonuclease complex (the one in Drosophila has been named Dicer) to cut them

into fragments containing 21–25 base pairs (~2 turns of a double helix) corresponding to

both sense and antisense strands of the target gene (Bernstein et al., 2001; Ketting et al.,

2001; Nykanen, 2001). Because of their action, these fragments of RNA have been

named "short (or small) interfering RNA" (siRNA). siRNA is produced during virus

infection or expression of a transgene designed to form dsRNA after transcription.

These siRNAs associate with and guide another nuclease complex (RNA-induced

silencing complex, RISC) to target homologous ssRNA (mRNA, or viral RNA) for

degradation or inhibition of translation (Bernstein et al., 2001; Elbashir, Lendeckel, and

Tuschl, 2001; Elbashir et al., 2001; Nykanen, 2001). During this process, RISC

undergoes an ATP-dependent activation step to unwind the ds siRNA. An as yet

unidentified endoribonuclease cleaves the target RNA. RISC might mediate both

cleavage and translational control (Hutvagner and Zamore, 2002; Mourelatos et al.,

2002). The mechanism of translational control is not known, although some have

proposed that it is mediated by the same core protein complex as target cleavage.

Each RISC contains only one of the two strands of the siRNA duplex as a guide to

identify complementary RNAs, but both siRNA strands may be able to direct RNA

interference (Martinez et al., 2002). Prototypical siRNA duplexes are 21 nt in length,

containing 19 base pairs with 2 nt 3′ overhanging ends (Tang et al., 2003). Active

siRNAs contain 5′ phosphate and 3′ hydroxyl groups (Chiu and Rana, 2003; Mallory et al., 2002). When a small RNA is fully complementary to its mRNA target, the mRNA is cleaved at a single phosphodiester bond located near the center of the sequence that is complementary to the siRNA. In contrast, if small RNAs do not pair with their mRNA targets near the center of the siRNA:mRNA helix, small RNA will direct translational repression without destabilizing the mRNA (Slack et al., 2000; Zeng, Wagner, and

Cullen, 2002). The antisense strand of an siRNA can direct cleavage of a corresponding sense RNA target, whereas the sense siRNA strands direct cleavage of an antisense target

(Martinez et al., 2002). Recent evidence suggests that the two strands of a siRNA duplex are not equally assembled into RISC (Schwarz et al., 2003). Stabilities of the base pairs at the 5′ ends of the two siRNA strands with the complementary ssRNA determine which strand is in the RNAi pathway. Finally, the cleaved RNA is probably further degraded by exoribonucleases and may function as siRNA again (Hammond et al., 2000) (Figure 1.4).

Besides siRNA, another type of small RNA called microRNA (miRNA) with 21 nt in length also triggers RNA silencing and can affect plant development because some miRNAs are expressed in a developmental stage-specific or tissue-specific manner and miRNA targets are often transcripts of regulatory genes (Ambros, 2003; Carrington and

Ambros, 2003; Kasschau et al., 2003; Llave et al., 2002; Palatnik et al., 2003; Reed et al.,

2003; Xie, Kasschau, and Carrington, 2003).

There are some similarities and differences between siRNA and miRNA. Both are generated from the cleavage of dsRNA precursors (Park et al., 2002). However, siRNAs are normally processed from long double-stranded RNAs whereas miRNAs are generated from transcripts of endogenous, non-protein coding genes. The precursor miRNA can

form small stem loops from which mature miRNAs are cleaved. Recent data also indicate that single-stranded miRNAs are initially generated as siRNA-like duplexes, but their structures determine which strand enters RISC (Khvorova, Reynolds, and Jayasena,

2003; Schwarz et al., 2003). Dicer is required for the production of both siRNAs and miRNAs (Bernstein et al., 2001; Grishok and Mello, 2002; Knight and Bass, 2001). siRNAs are 21 to 26 nt long and miRNAs are about 21 nt in length. miRNAs are usually single-stranded, whereas siRNAs are believed to be predominantly double-stranded, with

2 nt 3′-overhangs formed by dicer cleavage. The predominant constitutive 21-nt cellular

RNA species are miRNAs, whereas siRNAs are normally induced after viral infection or when dsRNAs are introduced into cells artificially, although active transposons also appear to lead to siRNA production (Hutvagner and Zamore, 2002). siRNAs and miRNAs are functionally interchangeable in cleaving mRNA or suppressing translation depending on the degree of complementarity between the small RNA and its target

(Doench, Petersen, and Sharp, 2003; Zeng, Yi, and Cullen, 2003). RNAi-related mechanisms suppress transposon activity, viral infection and overexpression of some genes, whereas miRNAs have roles in development. Nevertheless, there is significant overlap between cellular pathways using siRNAs and miRNAs.

1.5.2.3 RNA silencing and protein/DNA methylation

Plant RNA silencing is frequently accompanied by DNA cytosine methylation of

the silenced gene that corresponds to the target RNA (Bender, 2001; Wassenegger,

2000). In tobacco, the first evidence for RNA directed DNA methylation (RdDM) has

come from studies with a plant RNA viroid, which replicates in nuclei (Wassenegger et

al., 1994). In transgenic plants containing a viroid sequence fragment, after viroid infection, DNA methylation was found in the intergrated viroid DNA sequence.

Methylation of genomic DNA is also observed even when the silencing is induced by an

RNA virus that replicates exclusively in the cytoplasm (Jones, Thomas, and Maule,

1998), which suggests the communication of an RNA silencing signal between the

cytoplasm and the nucleus. The signal is likely RNA, but it is not known if long dsRNA,

siRNA, or ssRNA is responsible for RNA directed, DNA methylation.

When dsRNA has sequence homology to a promoter, dsRNA can induce

transcriptional gene silencing associated with DNA methylation in the homologous

promoter region (Jones, Ratcliff, and Baulcombe, 2001; Matzke, Matzke, and Kooter,

2001). In contrast, in RNA silencing, DNA methylation is usually confined to the region

of homology between the dsRNA trigger and the target transcript (i.e. the coding region).

Why is DNA methylation associated with RNA silencing? One possibility is that

RdDM provides a way of amplifying the proportion of aberrant RNA generated during

transcription to enhance the RNA silencing process (Bender, 2001; Jones et al., 1999).

Increased aberrant RNAs produce more dsRNA. So far, RdDM has been described only

in plants.

Surprisingly, it has recently been shown that RNA silencing plays a central role in

the chromatin silencing (Figure 1.4). Eukaryotic chromatin has high densities of

reiterated sequences. Transposable elements, and these heterochromatic regions are

correlated with increased levels of DNA methylation. Twelve small (~20 base pair)

dsRNAs with homology to the centromeric repeats have been found in the fission yeast,

Schizosaccharomyces pombe (Reinhart and Bartel, 2002). Interestingly, transcriptionally

silent transgenes within the centromeric heterochromatin of S. pombe were activated in

mutants lacking Dicer or RdRP (Volpe et al., 2002), which indicates that siRNA is

critical in TGS. These mutants contained overlapping `forward´ and `reverse´

centromeric transcripts homologous to the previously mentioned, 12 small dsRNAs,

whereas the wild type contained `reverse´ transcripts only, and these appeared to be

turned over rapidly (Volpe et al., 2002). In addition, RdRP is physically associated with

centromeric heterochromatin (Volpe et al., 2002). These data indicate that RNA silencing

machinery and transcripts encoded by centromeric DNA are involved in the

transcriptional silencing of centromeric heterochromatin.

Methylation at lysines 9 and 27 in histone H3 is associated with transcriptionally

silent chromatin, whereas methylation at position 4 is associated with transcriptionally active chromatin (Cao et al., 2002; Strahl and Allis, 2000). Methylation of H3K9 is central to chromatin silencing in S. pombe, and HP1 (heterochromatin protein 1) binding

leads to silencing (Hall et al., 2002). The HP1 chromodomain targets the protein to

chromatin by interacting specifically with histone H3 methylated at lysine 9 (H3K9)

(Lachner et al., 2001). Interestingly, methylation of H3K27 facilitates the binding of

Polycomb, another heterochromatin-associated protein (Cao et al., 2002). Arabidopsis

HP1 binds methylated H3K9, and this interaction is required for the binding and activity of the CpNpG-specific DNA methyltransferase, chromomethylase 3 (Jackson et al.,

2002a). This evidence suggests a direct link between histone methylation and DNA methylation. Once methylated, DNA is bound by methylcytosine-binding proteins, such as MeCP1 and MeCP2, which are components of the histone deacetylase complex

(Shahbazian et al., 2002). Deacetylation of histone H4 then enables the chromatin to take on a more compact configuration, and completes the heterochromatic transition (Strahl and Allis, 2000).

The Arabidopsis mutant decreased DNA methylation 1 (ddm1) has highly reduced levels of DNA methylation, and also shows transposon re-activation and numerous developmental defects that accumulate over successive generations. Interestingly, histone methylation patterns are altered in ddm1 heterochromatin. Most importantly, H3K9 methylation is largely replaced by H3K4 methylation (Gendrel et al., 2002). Thus, the function of the DDM1 protein may be to enable histone methyltransferases to gain access to chromatin. Loss of this activity would affect the distribution of methylated histones and, in turn, the binding of DNA methyltransferases (Gendrel et al., 2002). Thus protein methylation and DNA methylation are correlated and both might be critical for gene silencing.

1.5.2.4 RNA silencing and plant viruses

RNA silencing can act as a virus-specific defense mechanism in plants. In an early study, transgenic plants expressing an untranslatable viral transgene showed resistance to tobacco etch virus (TEV), in which all RNAs with homology to the transgene were degraded, including the sequences of invading ssRNA virus (Lindbo and Dougherty,

1992a; Lindbo and Dougherty, 1992b; Lindbo et al., 1993). Further study found that the plant virus itself could also induce RNA silencing, and this virus-induced gene silencing can be targeted against the virus itself and a transgene that shares sequence homology

with the virus (Ruiz, Voinnet, and Baulcombe, 1998). This VIGS system has been used to

study gene functions as mentioned above. RNA silencing can explain both the

mechanisms of plant recovery from certain viral infections, and why recovered plants are

resistant to reinfection by the initial virus or closely related viruses (Covey, 1997;

Ratcliff, Harrison, and Baulcombe, 1997). Interestingly, RNA silencing can be induced

locally and then spread throughout the organism (Palauqui et al., 1997; Voinnet and

Baulcombe, 1997). But the mechanism of systemic spread of silencing is one of the key

unresolved questions in this field. The moving signal might be dsRNA or siRNA.

Plants have developed effective RNA silencing against virus infection, but many

plant viruses have evolved silencing-suppressor proteins that block cell-autonomous or

systemic silencing (Li et al., 2001; Voinnet, Pinto, and Baulcombe, 1999). Silencing

suppressors have been described in diverse DNA and RNA viruses of plants (Voinnet,

Pinto, and Baulcombe, 1999). The first viral suppressor of silencing characterized was

the helper component-proteinase (HC-Pro) encoded in the potyviral genome. HC-Pro has

been shown to interfere with both transgene-induced RNA silencing (Kasschau and

Carrington, 1998) and VIGS (Brigneti et al., 1998). HC-Pro suppresses the maintenance

of RNA silencing in tissues where silencing has already been established (Llave,

Kasschau, and Carrington, 2000). The 2b protein of cucumoviruses is also a silencing

suppressor. However, unlike HC-Pro, 2b is unable to reverse already established RNA

silencing, but prevents its initiation at the growing points of the plant by inhibiting the long-range activity of the silencing signal (Brigneti et al., 1998; Guo and Ding, 2002).

Like the 2b silencing suppressor, the p25 movement protein of potexviruses, such as potato virus X (PVX), also affect RNA silencing by preventing spreading of the silencing

signal (Voinnet, Lederer, and Baulcombe, 2000). Other known viral silencing suppressors include the P1 protein from sobemoviruses, p19 from tombusviruses, 130K replication

protein of Tomato mosaic virus (ToMV), and B2 of flock house virus (FHV) (Carrington,

Kasschau, and Johansen, 2001; Kubota et al., 2003; Li, Li, and Ding, 2002; van Wezel et

al., 2001; van Wezel et al., 2002; Voinnet, 2001). B2 from FHV is so far the only known

silencing suppressor from animal viruses.

Silencing suppression enhances viral pathogenicity. For example, potyviral

synergistic interactions with other virus have been shown to result from the suppression

of RNA silencing by HC-Pro. PVX and potato virus Y (PVY) can co-infect tobacco to

cause a classic synergistic disease. Synergistic viral diseases are characterized by

dramatic increases in symptoms and in accumulation of one of the co-infecting viruses. In

PVX and potyviral synergism, increased pathogenicity and accumulation of PVX are

mediated by the expression of potyviral HC-Pro. Co-infection with PVY also enhances

the pathogenicity and accumulation of two other heterologous viruses: cucumber mosaic

virus and tobacco mosaic virus (Pruss et al., 1997; Vance, 1991).

Mechanisms of viral silencing suppression are not well understood, although it is

clear that different suppressors target different steps in the silencing pathway. Further,

while silencing is virus specific, silencing suppression is not. This indicates that silencing

suppression is targeted at the general silencing machinery. HC-Pro suppresses both VIGS

and transgene-induced RNA silencing (Anandalakshmi et al., 1998). Furthermore, it even

can reverse already established RNA silencing of a transgene (Brigneti et al., 1998). It

has been found that suppression by HC-Pro and ToMV 130K protein does not reduce the

amount of 21 to 25 nucleotide dsRNA production and may even increase its

accumulation, suggesting that the major target of these suppressors is downstream from the production of siRNAs in the RNA silencing pathway (Wang and Bisaro, unpublished data, Johansen and Carrington, 2001) (Figure 1.4).

In the study of RNA silencing, using different systems to trigger and suppress

RNA silencing may give different results and possible different mechanisms of the suppression of RNA silencing of the same suppressor. For example, as discussed above,

HC-Pro can accumulate small RNA level in the suppression tissue when Agrobacterium- mediated transient expression system is used. But it has been found that transgenic plants expressing HC-Pro, in which PTGS of GUS transgene has been suppressed by HC-Pro, fail to accumulate the GUS small RNAs associated with silencing. The GUS transgene locus in these plants remains methylated. This data indicates that HC-Pro functions downstream of transgene methylation and upstream of accumulation of the siRNAs

(Mallory et al., 2001). This opposite result of accumulation of siRNAs might be due to the different expression system used. What cellular proteins interact with HC-Pro and how that causes silencing suppression are still unknown. But using a yeast two-hybrid system, a cellular suppressor of PTGS, a calmodulin-related protein (termed rgs-CaM) was identified that interacts with HC-Pro. Interestingly, the rgs-CaM, like HC-Pro itself, suppresses gene silencing (Anandalakshmi et al., 2000a).

Plants use RNA silencing as a defense against RNA virus infection and RNA viruses encode RNA silencing suppressors to counter this defense. The above mentioned silencing suppressors are all encoded by RNA viruses. What about DNA viruses?

Geminiviruses replicate in nucleus and do not produce dsRNA as a replication

intermediate (Figure 1.2). Thus the viral DNA genome could not be the direct inducer or the target of RNA silencing.

Nevertheless, geminiviruses can cause transgene RNA silencing when they are used as episomal vectors, and have also been used to silence a chromosomal gene

(Kjemtrup et al., 1998).

In addition, geminiviruses can induce and be the target of viral RNA silencing in infected plants. For example, small RNAs have been found in tomato plants agroinoculated or naturally infected with Tomato yellow leaf curl Sardinia virus

(TYLCSV, closely related to TYLCV), in which one band was detected just above a 20- nt-long oligo nucleotide DNA marker, hybridizing with both plus- and minus-strand AL1 probes in Northern blot of total RNA extracted from virus infected plants (Lucioli et al.,

2003). Our data also suggest that geminiviruses can be both inducers and targets of RNA silencing (see Chapter 3).

Besides RNA silencing induced by geminiviruses, transcriptional gene silencing of geminiviral promoter-driven transgenes following homologous virus infection has been reported (Seemanpillai et al., 2003). For example, transgenic plants expressing GUS transgene driven by TLCV promoter were silenced after TLCV infection (Seemanpillai et al., 2003).

Geminiviruses can not only cause RNA silencing, but also encode gene silencing suppressors, including AC2 (also known as AL2) protein of ACMV and the C2 protein of tomato yellow leaf curl virus (TYLCV) (van Wezel et al., 2001; Voinnet, 2001). The cysteine rich domain and the nuclear localization domain in C2 protein may be required for suppression of RNA silencing (Dong et al., 2003; Van Wezel et al., 2003; van et al.,

2002). In this thesis, data will be presented which indicates that TGMV AL2 and BCTV

L2 are also RNA silencing suppressors, and that the interaction with ADK is responsible for this activity (Chapter 3).

Figure 1.1 Geminivirus genome organization

The diagram depicts the double-stranded replicative forms of Maize streak virus (MSV,

Mastrevirus), Beet curly top virus (BCTV, Curtovirus), Tomato yellow leaf curl virus

(TYLCV, monopartite, Begomovirus), and Tomato golden mosaic virus (TGMV, A and

B, bipartite, Begomovirus). The solid arrows indicate the positions of viral genes with the approximate molecular mass of each encoded protein given in kD. Viral genes are designated by number and the direction of transcription: leftward (L, complementary sense) or rightward (R, viral sense). Certain viral genes are also indicated by name, including Rep (replication initiator protein), TrAP (transcriptional activation protein),

REn (replication enhancer), and CP (coat or capsid protein). The position of the conserved hairpin is indicated by an asterisk within the intergenic region (IR). The common region (CR), a sequence of ~230 bp that is nearly identical in TGMV DNAs A and B, is indicated by a hatched box (Bisaro, 1996) .

Figure 1.1 Geminivirus genome organization

Figure 1.2 Geminivirus replication cycle

Geminivirus DNA replication occurs in two stages. First, the ssDNA is converted into

dsDNA. The dsDNA serves as template for viral gene expression. Secondly, the dsDNA

serves as template for rolling circle replication cycle to produce new ssDNA products.

Both stages require cellular factors. Progeny ssDNA can re-enter the DNA replication

pool, associate with coat protein, or be transported outside the nucleus and to the

neighboring cells with the help of the viral movement protein BR1 (Figure provided by

D.M. Bisaro).

Figure 1.2 Geminivirus replication cycle 61

Figure 1.3 The methyl cycle

S-adenosyl-methionine (SAM) is the methyl donor for many transmethylation reactions.

The product, S-adenosyl-homocysteine (SAH), is a potent inhibitor of methylases. It is

converted to homocysteine and adenosine by S-adenosyl-homocysteine hydrolase

(SAHH). Phosphorylation of adenosine by adenosine kinase (ADK) is critical because the

SAHH reaction is reversible and the equilibrium lies in the direction of SAH synthesis.

Figure 1.3 The methyl cycle

Figure 1. 4 Steps leading to RNA silencing and suppression of RNA silencing

Cellular or geminiviral genes are normally expressed via mRNA (ssRNA). So called

“aberrant RNA (abRNA)” may also be produced by an unknown mechanism. These

ssRNA or abRNA might be converted to double-stranded RNA (dsRNA) by cellular

RNA dependent RNA polymerase (RdRP) and other factors. These dsRNAs (or viral

dsRNAs generated by RNA viruses) are converted to small interfering RNAs (siRNA) by

DICER-like dsRNase. Following incorporation into the RNA-induced silencing complex

(RISC), siRNA can direct mRNA or viral RNA degradation, or inhibition of translation.

The siRNA and ssRNA duplex could also be converted to dsRNA by RdRP and amplify

the silencing signal (probably siRNA). RNA silencing suppressors block different steps

in the pathway. For example, potyvirus help-component protease (HC-Pro) is a strong

silencing suppressor working downstream of the siRNA, while tombusvirus p19 binds

and prevents its incorporation into RISC. Geminivirus AL2 or L2 proteins block a

maintenance step, apparently by interfering with methylation. Reducing adenosine kinase

activity through the action of the geminivirus proteins, by using a dsADK construct, or an

ADK inhibitor (RBI) has the same effect (Chapter 3).

Figure 1. 4 Steps leading to RNA silencing and suppression of RNA silencing

CHAPTER 2

GEMINIVIRUS AL2 AND L2 PROTEINS INTERACT WITH

AND INACTIVATE ADENOSINE KINASE

2.1 Introduction

Geminiviruses are single-stranded DNA viruses that infect a wide range of plant species and cause considerable losses of food and fiber products. These relatively simple pathogens amplify their genomes in the nuclei of host cells by a rolling circle replication mechanism that employs double-stranded DNA intermediates as replication and transcription templates (Bisaro, 1996; Gutierrez, 1999; Hanley-Bowdoin et al., 1999). As geminiviruses do not encode a DNA or RNA polymerase, viral replication and transcription depend almost entirely on cellular machinery. However, in members of the genus Begomovirus (e.g. Tomato golden mosaic virus; TGMV) the product of the AL2 gene is additionally required for the expression of late viral genes, including coat protein and the BR1 nuclear shuttle protein that is necessary for spread of the virus in infected plants (Sunter and Bisaro, 1992). Stimulation of the coat protein promoter in phloem and

mesophyll cells involves both activation and derepression by AL2 (Sunter and Bisaro,

1997; Sunter and Bisaro, 2003). The 15 kDa AL2 (also known as transcriptional

activator protein; TrAP) has a C-terminal activation domain that is functional in plant,

yeast, and mammalian cells (Hartitz, Sunter, and Bisaro, 1999a).

Beet curly top virus (BCTV; genus Curtovirus) encodes a positional homologue

of AL2 called L2. However, the L2 protein is not required for late viral gene expression and shares only limited homology with AL2 (Hormuzdi and Bisaro, 1995; Stanley et al.,

1992a). The most conserved region encompasses about 20 amino acids near the middle

of both proteins and contains a series of cysteine and histidine residues that might form a

zinc binding structure. Despite their differences, the two proteins share a common

function that does not involve transcriptional activation. When expressed from

transgenes in Nicotiana benthamiana and N. tabacum var. Samsun plants, both AL21-100

(lacking the activation domain) and L2 condition a unique enhanced susceptibility

phenotype. More specifically, TGMV, BCTV, and the RNA-containing Tobacco mosaic

virus (TMV) exhibit greater infectivity on plants expressing AL21-100 or L2 than they do

on non-transgenic plants. Enhanced susceptibility is characterized by a reduction in

mean latent period (time to first appearance of symptoms) and by a decrease in the

inoculum concentration required to elicit infection, without significant enhancement of disease symptoms or an increase in virus replication levels (Sunter, Sunter, and Bisaro,

2001a).

In a recent study, we showed that the enhanced susceptibility phenotype is due to

the interaction of AL2 and L2 with SNF1 kinase, a global regulator of metabolism. In

response to nutritional or environmental stresses that deplete ATP, SNF1 turns off energy consuming biosynthetic pathways and turns on alternative ATP-generating systems. AL2 and L2 inactivate SNF1 both in vitro and in vivo. Further, reducing SNF1 activity in transgenic plants by antisense expression causes enhanced susceptibility similar to that conditioned by AL2 and L2 transgenes, whereas SNF1 over-expression leads to enhanced resistance (Hao et al., 2003). Thus, metabolic alterations mediated by SNF1 appear to be a component of innate antiviral defenses and SNF1 inactivation by AL2 and L2 can be viewed as a counterdefensive measure.

One of the most exciting developments in the field of virus-host interactions in recent years is the recognition that RNA silencing can act as an adaptive response to limit virus replication (Ahlquist, 2002; Baulcombe, 1999; Carrington, Kasschau, and Johansen,

2001; Vance and Vaucheret, 2001; Waterhouse, Wang, and Lough, 2001). RNA silencing (also known as posttranscriptional gene silencing, RNA interference, or gene quelling) occurs in most eukaryotes and involves the induction, by dsRNA, of a multi- step process that leads to RNA directed, sequence-specific gene silencing. Studies in several model organisms have led to the identification of silencing machinery components and a partial understanding of the silencing process (Bass, 2000; Plasterk,

2002; Sharp, 2001; Tang et al., 2003).

Most plant viruses, including geminiviruses, are both inducers and targets of RNA silencing, and many actively counter this defense by encoding proteins with antisilencing activity (Brigneti et al., 1998; Kasschau and Carrington, 1998; Kjemtrup et al., 1998;

Voinnet, Pinto, and Baulcombe, 1999). Here the most relevant examples are the AL2

(also called AC2 or C2) proteins from the begomoviruses African cassava mosaic virus 68

(ACMV) and Tomato yellow leaf curl virus (van Wezel et al., 2002; Voinnet, Pinto, and

Baulcombe, 1999). Recent experiments in David Bisaro’s laboratory indicate that

TGMV AL2 and BCTV L2 also have antisilencing activity (unpublished results). Thus,

AL2 and L2 appear to disable both innate (SNF1-mediated) as well as adaptive (RNA silencing-mediated) defense pathways.

In this chapter, we demonstrate that geminivirus AL2 and L2 proteins interact with and inactivate ADK both in vitro and in vivo, that ADK activity is reduced in transgenic plants expressing these proteins, and that ADK activity is reduced in geminivirus infected plants. In contrast, we show that ADK levels are increased in plants infected with diverse RNA viruses or with a BCTV mutant lacking a functional L2 protein. Finally, we demonstrate that AL2 protein is present in both the nuclear and cytoplasmic compartments of infected cells. We propose that ADK inactivation helps to suppress SNF1-mediated responses and may also play a role in suppressing RNA silencing.

2.2 Results

2.2.1 AL2 and L2 interact with ADK

Because AL2 and L2 condition enhanced susceptibility when expressed in

transgenic plants (Sunter, Sunter, and Bisaro, 2001), we were interested in identifying

cellular proteins that interact with these related geminivirus proteins. The yeast two- hybrid system employed in the search relies on strain Y190, which harbors both HIS3 and 69

lacZ reporter genes (Durfee et al., 1993; Harper et al., 1993). The system was used in an earlier screen to identify the interactions between AL21-83 and L2 with SNF1 (Hao et al.,

2003). Continued screening using AL21-115 (lacking the activation domain) and proteins

expressed from an Arabidopsis cDNA-activation domain fusion library identified a

protein corresponding to the C-terminal residues 145 to 345 of adenosine kinase (ADK),

which was designated ADK-C. An Arabidopsis genomic sequence containing ADK-C

(GenBank accession number AL162751) was mapped to chromosome 5 by BLAST

search. From this a full-length mRNA sequence was deduced, which in turn was used to

design a forward PCR primer including the predicted start codon and a reverse primer

containing the stop codon. Using an Arabidopsis cDNA library as template, these

primers generated a 1038 bp DNA fragment which contained a full-length ADK sequence

(345 amino acids; 37.9 kDa) (data not shown). Arabidopsis has two ADK genes, ADK1

and ADK2, which are 92% identical at the amino acid level (Moffatt et al., 2000). The

amino acid sequence encoded by the cDNA identified in our screen is identical to ADK2.

The interaction of full-length ADK2 with TGMV AL2 and BCTV L2 was

confirmed in the two-hybrid system. Yeast growth in selective medium lacking histidine

(indicative of interaction) was supported when cells expressed both full-length ADK2

and either AL21-115 or L2. These cells were also positive in β-galactosidase filter assays,

and similar results were obtained regardless of whether the proteins were expressed as

bait or prey (data not shown). Interaction was also observed with the AL2 (AC2) protein

of ACMV and with AL2 from Cabbage leaf curl virus (CaLCuV), which infects

Arabidopsis. However, no growth was evident when ADK2 was co-expressed with non-

interacting, negative control proteins including p53, CDK2, lamin, TGMV AL1, AL3, 70

and coat protein, and BCTV coat protein. It was concluded that both AL2 and L2 specifically interact with ADK2.

Not surprisingly, we subsequently found that AL2 and L2 also interact specifically with ADK1 (data not shown). All further work was done with ADK2 which for convenience is simply referred to as ADK.

2.2.2 AL2 and L2 inhibit ADK activity in E. coli

E. coli strain HO4 was employed in an in vivo assay to verify that the ADK cDNA encoded a functional protein. Prokaryotes do not encode ADK and therefore cannot carry out direct phosphorylation of adenosine. Instead, a two step salvage pathway is used in which adenosine is first converted to adenine by purine nucleoside phosphorylase (deoD) and then ribophosphorylated to AMP by adenine phosphoribosyltransferase (APT). E. coli HO4 (purE, purF, deoD, apt) is deficient in both purine biosynthesis and adenosine salvage (Hove-Jensen and Nygaard, 1989).

However, strain HO4 can survive in minimal medium if provided with adenosine and

ADK activity. ADK cDNA and truncated ADK cDNAs (N-terminal deletion derivatives

ADK-∆N13 and ADK-∆N88) were inserted into the expression vectors pMAL or pBSK, and E. coli HO4 cells transformed with these constructs were plated on complete medium and minimal medium containing adenosine. Expression from pMAL produces maltose binding protein fusions, whereas pBSK constructs express proteins fused to the β- galactosidase α-peptide. As shown in Figure 2.1A, none of the constructs had an adverse effect on cell growth in complete medium. However, only cells containing plasmids that 71

expressed full-length ADK (pBSK-ADK or pMAL-ADK) were able to grow in minimal

medium containing adenosine. Thus the ADK cDNA encodes a protein with adenosine kinase activity which is not affected by fusion with the α-peptide or maltose binding

protein, and deletion of as few as 13 N-terminal amino acids renders the protein inactive.

To determine the consequences of AL2 and L2 interaction with ADK, the kinase

and the viral proteins were co-expressed in HO4 cells. To perform these experiments,

full-length AL2 and L2, as well as a control protein that does not interact with ADK

(chloramphenicol acetyl transferase; CAT), were expressed from pDHK29 (Phillips,

Park, and Huber, 2000). The replication origin of this plasmid is compatible with those

of the pMAL and pBSK vectors. As demonstrated in Figure 2.1B, co-expression of ADK

with CAT, AL2 or L2 had no effect on the growth of E. coli HO4 in complete medium.

And as before, expression of ADK alone (or with CAT) permitted growth on minimal

medium containing adenosine. Remarkably, however, co-expression of either AL2 or L2

with ADK prevented growth on the supplemented minimal medium. These results

confirm that AL2 and L2 interact with ADK, and indicate that these interactions result in

inhibition of ADK activity.

2.2.3 AL2 and L2 inhibit ADK in vitro

As an additional confirmation of functional interaction, partially purified proteins

were tested in vitro. Histidine-tagged ADK, deletion derivatives (ADK-∆N13, ADK-

∆N24, ADK-∆N88), and CAT were expressed in E. coli BL21 cells and partially purified

by nickel-NTA chromatography (Figure 2.2A and 2.2B). For unknown reasons, cells 72

expressing full-length ADK yielded less total protein, and consequently less His-ADK,

than those expressing the deletion derivatives. AL2 and L2 proteins fused to glutathione-

S-transferase (GST-AL2 and GST-L2) were likewise expressed in BL21 cells and

partially purified by glutathione-agarose chromatography (Hartitz, Sunter, and Bisaro,

1999). Histidine tagged AL2 (His-AL2) was expressed in insect (Sf9) cells from a

baculovirus vector (not shown).

ADK, and deletion derivatives, were added to reaction mixtures containing the

32 substrates adenosine and [γ- P]ATP, and conversion of adenosine to AMP was

monitored by thin layer chromatography. No activity was detected with the truncated

ADK proteins or with CAT (negative control) in this assay (data not shown). However,

AMP was generated in reactions containing full-length ADK, and the amount of product

was proportional to the amount of added protein over a range of 1 to 70 ng (Figure 2.2C).

Addition of EDTA to 10 mM completely inhibited activity, confirming the magnesium

dependence of the reaction (data not shown).

In reaction mixtures containing 10 ng of ADK, pre-incubation with AL2 or L2 at

a 3:1 molar ratio prior to the addition of ATP resulted in a substantial reduction in ADK

activity (Figure 2.3A). By contrast, pre-incubation with the control proteins CAT or GST

at a 7:1 molar ratio did not reduce the amount of AMP generated. In experiments

containing 10 ng ADK and varying amounts of AL2, ADK activity was reduced nearly

90% at a 1:1 molar ratio of AL2:ADK (Figure 2.3B and 3C).

These results conclusively demonstrate that AL2 and L2 directly interact with

ADK and inhibit its activity. The stoichiometry of inhibition is also consistent with a

mechanism involving direct interaction between the viral proteins and the cellular kinase. 73

2.2.4 AL2 and L2 inhibit ADK activity in yeast

L2 was co-expressed with ADK in yeast to determine whether the proteins might

functionally interact in the context of a eukaryotic cell. The BCTV protein was used in preference to AL2 in this study because expression of the latter can have adverse effects

on yeast cell growth (Hartitz, Sunter, and Bisaro, 1999).

We first asked whether Arabidopsis ADK protein can complement a yeast ADK

deletion strain (ado1), which displays a slow growth phenotype (Lecoq et al., 2001). As

shown in Figure 4A, growth of the ado1 mutant is indeed much reduced compared to the

ADO1 parent strain, but growth was not further reduced by expression of L2. However, the extent and rate of growth was reproducibly complemented to levels slightly greater than the parent strain by the introduction of a plasmid expressing Arabidopsis ADK.

Thus the conservation of yeast and plant ADK proteins apparent at the amino acid level

(39% identity) reflects a functional conservation which allows the Arabidopsis protein to complement the growth defect of the yeast ADK deletion strain. However, when the deletion strain was cotransfected with expression plasmids containing ADK and L2, a significant reduction in growth, relative to that observed with expression of ADK alone, was observed. Using the in vitro ADK assay, comparison of crude extracts obtained from the complemented ado1 mutant and from ado1 cells expressing both ADK and L2 confirmed that the presence of L2 resulted in a significant reduction in ADK activity

(Figure 2.4B and 2.4C). Cells of the complemented mutant strain (ado1 + ADK)

contained nearly twice as much ADK activity as cells also expressing the L2 protein

(ado1 + ADK + L2).

We concluded from these experiments that Arabidopsis ADK is capable of complementing the growth defect of the yeast ado1 deletion strain, and that complementation is reduced by L2 with a corresponding reduction in ADK activity.

These studies provide strong evidence for functional, in vivo interaction between L2 and

ADK.

2.2.5 AL2 and L2 inhibit ADK activity in plants

Two approaches were employed to investigate whether AL2 and L2 might inhibit

ADK activity in plant cells. One approach took advantage of previously constructed, transgenic N. benthamiana lines expressing AL21-100 or L2 from the constitutive 35S promoter (Sunter, Sunter, and Bisaro, 2001). Comparable stem pieces (~3 mm) were obtained from just below the shoot apex from transgenic plants (AL2 line 472-1 and L2 line CTL2-6) and comparable non-transgenic plants approximately 4 weeks post- germination. Stem pieces were used because stem extracts contain considerably more

ADK activity than comparable amounts of leaf extract. Each sample consisted of pooled tissue from 3 randomly selected plants, and 3-5 samples were analyzed for each treatment in each of 6 independent experiments.

It was first determined that crude stem extracts from non-transgenic plants displayed ADK activity in the in vitro assay, and that the level of activity was proportional to the amount of extract added over a range of 100 to 500 ng (data not shown). Within this range, extracts from transgenic plants expressing AL2 or L2

consistently showed a 15-30% reduction in ADK activity compared to extracts from non-

transgenic plants (Figure 2.5A). These reductions were significant (P < 0.05) as

determined by Student's t test. Further, they do not appear to be due to reductions in

SNF1 activity (which is also reduced by AL2 and L2), since ADK activity was normal in antisense SNF1 plants (line AS-12; data not shown). Because the non-transgenic and

transgenic N. benthamiana plants are isogenic except for the presence of an AL2 or L2 transgene, we concluded that AL2 and L2 interaction with ADK, with consequent inhibition of kinase activity, can occur in planta.

In the second approach, the in vitro ADK assay was used to ask whether kinase activity in stem extracts obtained from TGMV and BCTV infected N. benthamiana plants is reduced relative to extracts from mock inoculated plants and plants infected with

BCTV containing a frameshift mutation that inactivates L2. The BCTV L2-2 mutant was

chosen for this study because it replicates to nearly wild-type levels and generates

symptoms similar to wild-type virus (Hormuzdi and Bisaro, 1995). TGMV AL2 mutants

could not be used in this experiment because they do not express the BR1 movement

protein and for this reason are not infectious (Sunter and Bisaro, 1992). ADK activity

was also measured in plants infected with the unrelated, RNA-containing Cucumber

mosaic virus (CMV) and Potato virus X (PVX). Stem pieces, which showed no obvious

necrosis, were obtained from comparable locations near the shoot apex of infected and

mock inoculated plants. Stem extracts are appropriate for this study because all viruses

tested here infect both phloem and mesophyll cells except for BCTV, which is phloem-

limited. Two to three experiments were performed with each virus, and each experiment

included three to four replicate samples which consisted of pieces from three randomly

chosen plants. Plants were infected four weeks post-germination and tissue was harvested 14 days post-inoculation.

Consistent with previous results, extracts from BCTV or TGMV infected plants contained 1.5-fold and 3.5-fold less ADK activity than extracts from mock inoculated plants, respectively (Figure 2.5B). By contrast, ADK activity levels in plants infected with the BCTV L2 mutant virus were 1.3-fold higher than in mock-inoculated plants.

This outcome indicates that functional interaction resulting in suppression of ADK activity occurs during geminivirus infection, and suggests that if the responsible viral protein (L2 in this experiment) is not present, geminivirus infection would otherwise cause an increase in ADK activity. That ADK activity was also increased 1.4-fold to 3.5- fold in PVX and CMV-infected plants supports the idea that virus infection can induce an increase in ADK activity.

2.2.6 AL2 accumulates in the cytoplasm and the nucleus

AL2 is a phosphoprotein that binds ssDNA and zinc and has an acidic, C-terminal activation domain (Hartitz, Sunter, and Bisaro, 1999). In keeping with its role in transcription activation, AL2-GFP fusion protein has been shown to localize to the nucleus (van Wezel et al., 2001). However, because AL2 appears to interact with and inhibit cytoplasmic ADK in vivo, it was important to demonstrate that some AL2 protein can be found in the cytoplasm.

Several different approaches were employed to localize TGMV AL2. First, we observed considerable GFP signal in both the nucleus and the cytoplasm of N. 77

benthamiana or N. tabacum BY-2 protoplasts following transfection with expression plasmids producing AL2 fused with multiple GFP moieties at either the N- or C-terminus

(AL2-GFP-GFP and GFP-GFP-AL2; data not shown). However, given the considerable cytoplasmic and nuclear signal seen in control transfections with GFP-GFP and GFP-

GFP-GFP fusion proteins, we concluded that it was difficult to localize AL2 with this method.

Using AL2 antiserum and a fluorescently-labeled (FITC) secondary antiserum, we observed that cells from transgenic N. benthamiana plants expressing truncated TGMV

AL21-100 display a widespread fluorescence which suggested that the protein is present

throughout the cell (data not shown). However, because localization of native protein

might be different, and additionally might be influenced by other viral proteins or the

infection process, we attempted to identify AL2 in TGMV infected cells. Epidermal

peels taken from infected N. benthamiana plants or mock-inoculated plants were

examined under comparable conditions using the TGMV AL2 antiserum. As shown in

Figure 2.6, AL2 was found to accumulate in the nucleus of infected cells, as verified by

DAPI staining. However, in nearly all infected cells a significant amount of fluorescence

indicating the presence of AL2 was also observed in the cytoplasm, where it appeared in

small to large aggregates. At this time we do not know if these aggregates are associated

with specific cellular or virus-induced structures. In contrast, little or no fluorescence

was observed in cells from mock inoculated plants (Figure 2.6), and no fluorescence was

observed in mock inoculated or infected cells when the primary AL2 antibody was

omitted (data not shown). We concluded that AL2 is present in both the nuclear and

cytoplasmic compartments in infected cells.

Cell fractionation was another approach used to localize AL2. Native AL2 protein was expressed in insect (Sf9) cells from a baculovirus vector as described previously (Hartitz, Sunter, and Bisaro, 1999), and nuclear and soluble cytoplasmic (S-

100) fractions were obtained by differential centrifugation. Extracts were prepared 48 hours after cells were infected with the baculovirus vector, at which time nuclei remained intact. As shown in Figure 2.7, AL2 accumulated to significant amounts in both the cytoplasmic and nuclear fractions, and a comparison of relative band intensities suggests that as much as a third of the AL2 protein is cytoplasmic. It is interesting that the nuclear fraction contained at least three forms of AL2. The two slower-migrating species observed in SDS-PAGE were previously assumed to correspond to phosphorylated forms of the protein (Hartitz, Sunter, and Bisaro, 1999). Only trace amounts of these forms were present in the cytoplasm, and treatment of nuclear fractions with phosphatase resulted in their disappearance and a concomitant increase in the intensity of the fastest- migrating form, which co-migrates with cytoplasmic AL2 (Figure 2.7). Phosphatase treatment of cytoplasmic extracts had no effect on the migration of AL2 protein, further suggesting that this faster-migrating form is not phosphorylated (data not shown).

Given the differential accumulation of phosphorylated and non-phosphorylated forms in the nucleus and cytoplasm, it is unlikely that the presence of AL2 protein in the cytoplasm is an artifact caused by the loss of nuclear envelope integrity. And since similar phosphorylated AL2 forms can be observed in whole Sf9 cell extracts (Hartitz,

Sunter, and Bisaro, 1999), it is unlikely that their relative absence in the S-100 fraction is the result of a compartmentalized phosphatase activity. As a further test, we found no evidence for a phosphatase activity directed against AL2 after re-mixing the S-100 and

nuclear fractions (data not shown). Thus, as in infected plant cells, it was concluded that

AL2 protein is present in the nucleus and the cytoplasm. In insect cells, phosphorylated

AL2 appears to accumulate almost exclusively in the nucleus, whereas non- phosphorylated AL2 is found in both compartments. Because phosphorylation and localization events often occur with high fidelity in insect cells, it is likely that this is also

the case in plant cells, although this point remains to be investigated.

2.2.7 SNF1 inactivates adenosine kinase

AL2 and L2 interact with both SNF1 and ADK, so we wanted to know whether

ADK also interacted with SNF1. First we found that they interact with each other in the

yeast two-hybrid system. Yeast strain Y190 was able to grow on selective medium

lacking histidine (indicative of interaction) only when cells expressed both full-length

ADK and either SNF1 full-length or SNF kinase domain (SNF-KD). These cells were

also positive in β-galactosidase filter assays, and similar results were obtained regardless

of whether the proteins were expressed as bait or prey (data not shown). However, no

growth was observed when ADK was co-expressed with non-interacting, negative control

proteins including p53, CDK2, lamin, TGMV AL1, AL3, and coat protein, and BCTV

coat protein. Therefore, both SNF1 and SNF1-KD specifically interact with Arabidopsis

ADK.

As above, E. coli strain HO4 was employed to study whether the interaction

between ADK and SNF1 could happen in bacterial cells and what the consequence of this

interaction is to the growth of HO4 which is complemented by ADK. As stated above, strain HO4 can survive in minimal medium only if provided with adenosine and ADK activity. To perform these experiments, full-length SNF1 or SNF1-KD, as well as a control protein (CAT) that does not interact with ADK were expressed from pDHK29.

ADK cDNA was inserted into the expression vectors pMAL or pBSK, and E. coli HO4 cells transformed with these constructs were plated on complete medium and minimal medium containing adenosine. As shown in Figure 2.8, co-expression of ADK with CAT had no effect on the growth of E. coli HO4 in complete medium. However, co-expression of either SNF1 or SNF1-KD (not shown) with ADK prevented E. coli HO4 growth on the supplemented minimal medium. These results indicate that SNF1 and SNF1-KD interact with ADK in vivo and these interactions cause inhibition of ADK activity.

SNF1 is a kinase and has a signature sequence in its substrates. ADK amino acid sequence was scanned for the potential SNF1 phosphorylation sites. Two sites were found in the N-terminal region (Figure 2.9). Next we did the SNF1 kinase assay using

ADK as its substrate. ADK is a kinase but does not phosphorylate itself. Therefore partially purified SNF1-KD and ADK proteins were tested in vitro in order to examine the possibility of ADK phosphorylation by SNF1. The SNF1-KD purification was described as before (Hao et al., 2003). The SNF1 kinase reaction was set up as before

(Hao et al., 2003) with adding ADK as potential SNF1 substrate. SNF1-KD has autophosphorylation function (Hao et al., 2003). Unfortunately, the full length ADK has a size similar to SNF1-KD, so the phosphorylation of full length ADK is difficult to differentiate from SNF1-KD on the gel. But all the ADK mutant proteins with deletion of

13, 24, 88 amino acids from the N terminal start site were clearly phosphorylated by

SNF1-KD. The control CAT protein is not phosphorylated by SNF1-KD (Figure 2.10).

ADK and its mutants do not have auto-phosphorylation activity (Figure 2.10). This preliminary result suggested that ADK was phosphorylated by SNF1 kinase (Figure

2.10).

ADK activity was also examined in vitro when SNF1 was present. ADK and

SNF1-KD were added to reaction mixtures containing [γ-32P]ATP but lacking the substrate adenosine and incubated for 10 min at 30 °C. Then adenosine was added to start the ADK kinase reaction. In reaction mixtures containing 10 ng of ADK, pre-incubation with SNF1 at about 0.5:1, 1:1.and 2:1 molar ratio (SNF1-KD:ADK) prior to the addition of adenosine, resulted in a substantial reduction in ADK activity (Figure 2.11). By contrast, pre-incubation with the control proteins CAT at a 7:1 molar ratio did not reduce the amount of AMP generated. These results indicate that SNF1 directly interacts with

ADK and inhibits its activity. It will be interesting to find whether the inhibition is due to the physical binding, or phosphorylation or both.

2.3 Discussion

We previously showed that the TGMV AL2 and BCTV L2 proteins are pathogenicity determinants that cause enhanced susceptibility when expressed in transgenic plants (Sunter, Sunter, and Bisaro, 2001). Here we demonstrate that these two proteins interact with and inactivate ADK in vitro and in vivo, and that ADK activity is

reduced in geminivirus infected tissue in a manner that depends, in the case of BCTV, on the presence of a functional L2 protein. We also found that ADK activity is increased in response to infection with CMV, PVX, and a BCTV L2 mutant, suggesting that increasing the activity of this enzyme is part of the host response to infection by RNA and DNA viruses.

Experiments presented here show that AL2 and L2 interact with ADK, and in a previous study we demonstrated that these same viral proteins also interact with and inactivate SNF1 (Hao et al., 2003). Both ADK and SNF1 are kinases. However, it is unlikely that the AL2/L2 interactions with ADK and SNF1 are mediated solely by a motif conserved among different kinases, as the viral proteins do not interact with cyclin dependent kinase 2 (CDK2) in the yeast two-hybrid system (Hao et al., 2003). In addition, under conditions similar to those used to demonstrate inhibition of SNF1, AL2 and L2 do not inhibit the in vitro autophosphorylation activity of casein kinase II (data not shown). Finally, AL2 is clearly a substrate for phosphorylation by an as yet unidentified cellular kinase(s) (Hartitz, Sunter, and Bisaro, 1999; Figure 2.7). Thus we conclude that the interactions with ADK and SNF1 are both specific and significant, and that kinase inactivation is not the only possible consequence of interaction with AL2 and

L2.

Our observation, by immunofluorescence microscopy, that AL2 is located in both the nucleus and the cytoplasm of infected N. benthamiana cells provides direct evidence that AL2 has the opportunity to interact with ADK, which is believed to be primarily located in the cytoplasm (Figure 2.6). At this time it is not clear whether the localization of AL2 is influenced by other viral proteins, but we have not observed a direct interaction 83

between AL2 and any other TGMV protein in the yeast two-hybrid system (data not

shown). The results of cell fractionation studies using extracts from Sf9 cells infected

with a baculovirus expressing native AL2 support the conclusion that this protein is

present in the nucleus and the cytoplasm, and further suggest that phosphorylated forms

accumulate predominantly in the nucleus (Figure 2.7). If confirmed in plant cells, this

observation would imply that the functional localization and multiple activities of AL2

are at least in part regulated by cellular protein kinases.

How might a reduction in ADK activity be related to enhanced susceptibility?

There are several possibilities, two of which will be discussed here. In eukaryotes, ADK

is responsible for recycling adenosine and maintaining intracellular AMP levels. This

can be related to our recent finding that AL2 and L2 also interact with and inactivate

SNF1 kinase and that reduction of SNF1 activity in transgenic plants, by expression of an

antisense SNF1 transgene, results in enhanced susceptibility (Hao et al., 2003). SNF1 (in plants and yeast) and its mammalian homolog, AMP-activated protein kinase (AMPK),

play a central role in the regulation of metabolism in response to stresses that deplete

ATP (the cellular stress response) (Halford and Hardie, 1998; Hardie, Carling, and

Carlson, 1998; Hardie and Hawley, 2001; Johnston, 1999). These serine/threonine

kinases exist in heterotrimeric complexes along with two other subunits that serve

scaffolding and regulatory functions. When AMP:ATP ratios are elevated, SNF1/AMPK

complexes turn off energy-consuming systems and turn on alternative ATP-generating

systems. In plants for example, SNF1 has been shown in vitro to phosphorylate and

inactivate sucrose phosphate synthase (sucrose synthesis), nitrate reductase (nitrogen for

nucleic acid and protein synthesis), and HMG CoA reductase (steroid and isoprenoid

synthesis) (Sugden et al., 1999b).

The available evidence indicates that 5'-AMP binds and activates the plant and

mammalian kinases by multiple mechanisms (Davies et al., 1995; Hardie and Hawley,

2001; Hawley et al., 1995; Sugden et al., 1999a). That the AL2 and L2 proteins interact with and inactivate both SNF1 and ADK leads us to advance the proposal that SNF1- mediated metabolic responses are an important component of innate antiviral defense, and that geminiviruses have evolved a dual approach to disabling these responses; i.e. by inactivating both SNF1 and ADK (Hao et al., 2003). Inherent in this model is the idea that AMP generated by ADK is an early activator of SNF1, although it is recognized that

SNF1 also responds to AMP which accumulates as a result of ATP depletion. However,

ATP depletion caused by the stress of virus replication might occur too slowly to generate a useful AMP signal. Our recent observation that ADK and SNF1 interact in the yeast two-hybrid system provides some support for this view (Chapter 3), and we speculate that ADK and SNF1 might exist in a complex for the purpose of generating a rapid metabolic response. The model also predicts that ADK activity should increase following infection with a virus that does not encode a protein capable of interacting with

ADK, as is the case with the BCTV L2 mutant and may be the case for PVX and CMV

(Figure 2.5). The apparent absence of an ADK inactivating protein could mean that these

RNA viruses have not evolved a mechanism to disable this pathway, or that they target a

downstream step.

Precisely how SNF1-mediated responses act to limit viral infectivity is not clear,

but the ability of an organism to mount an effective defense against an invading pathogen 85

likely depends on the ability of its cells to maintain a positive energy balance. SNF1

shut-off of biosynthetic pathways might also deprive the virus of essential precursors

needed for replication. Whatever the mechanism, that geminiviruses target and inactivate

both SNF1 and its potential activator ADK underscores the importance of metabolic

responses and provides a molecular link between host metabolic status and susceptibility

to pathogens. Given the conservation of ADK and SNF1/AMPK, the possibility that

metabolic responses are an important feature of pathogen defense in other eukaryotes

warrants further investigation.

Another possible connection between ADK and viral pathogenicity is the observation that AL2 (also known as AC2 or C2) from African cassava mosaic virus

(ACMV), and Tomato yellow leaf curl virus (TYLCV) can suppress RNA silencing

(posttranscriptional gene silencing) (van Wezel et al., 2002; Voinnet, Pinto, and

Baulcombe, 1999). We have found that TGMV AL2 and BCTV L2 are also silencing

suppressors (Chapter 3). In plants, viruses are both initiators and targets of RNA

silencing, which acts to limit the extent of virus infection (Ahlquist, 2002; Baulcombe,

1999; Carrington, Kasschau, and Johansen, 2001; Vance and Vaucheret, 2001;

Waterhouse, Wang, and Lough, 2001). It has been noted that the maintenance of both

RNA silencing and transcriptional gene silencing is associated with methylation of target

gene sequences (Ahlquist, 2002; Baulcombe, 1999; Bender, 2001; Carrington, Kasschau,

and Johansen, 2001; Jones, Ratcliff, and Baulcombe, 2001; Paszkowski and Whitham,

2001; Vance and Vaucheret, 2001; Waterhouse, Wang, and Lough, 2001). RNA

silencing has also been associated with specific histone methylation, particularly histone

H3 at lysine 9, and histone and DNA methylation may be coupled (Jackson et al., 2002a; 86

Volpe et al., 2002). In the case of the geminiviruses, which produce mRNAs from a

dsDNA template that assembles into a minichromosome (Pilartz and Jeske, 1992),

methylation of the viral genome and chromatin components must be considered as a

potential mechanism to limit viral gene expression.

ADK activity is critical for sustaining the methyl cycle and SAM-dependent

transmethylation. In yeast, the available evidence suggests that a primary role of ADK is

to recycle adenosine produced by the methyl cycle (Lecoq et al., 2001). This may also be

true in plant cells, where ADK activity has been observed to increase (~1.5- to 3-fold) in response to methyl demand (Weretilnyk et al., 2001). That transgenic, ADK-deficient

Arabidopsis plants have reduced transmethylation activity supports this view (Moffatt et al., 2002). Thus, the increase in ADK activity we observed in infected N. benthamiana

plants might also reflect an increased demand for transmethylation activity in response to

virus challenge (Figure 2.5). We recognize that SAM is an important cofactor that could

impact many aspects of cellular metabolism relevant to virus replication. However, considering the role of methylation in reinforcing silencing pathways, we speculate that

by inhibiting ADK, AL2 and L2 might indirectly suppress silencing by interfering with

methylation. In support of this idea, transmethylation deficiency has been cited as a

possible explanation for the frequent reversion of ADK-silenced plants (see below;

Moffatt et al., 2002).

Unfortunately, our attempts to produce transgenic plants that would permit a

direct examination of the relevance of ADK to geminivirus infection have proved

unsuccessful. After examining a combined total of 49 transgenic N. benthamiana and

Arabidopsis plants containing constructs designed to constitutively over-express the 87

kinase (ADK sense) or under-express the kinase (ADK antisense and ADK RNAi) we have obtained only one line with altered ADK activity, as determined by the in vitro assay (data not shown). The exceptional Arabidopsis line (At-RNAi-4) produces two types of progeny: one is extremely stunted and has about 25% of wild-type ADK activity. The second type is less stunted and has about 60% of wild-type activity. We cannot explain why only one of the transgenic events caused a difference in ADK activity, despite high levels of transgene expression in the case of the sense and antisense lines. But in several respects our experience has been similar to that of Moffat et al.

(Moffatt et al., 2002), who previously constructed transgenic Arabidopsis ADK lines.

These investigators found that all lines examined (8 containing sense and 4 containing antisense ADK constructs) had reduced ADK activity suggesting varying degrees of silencing, and that reversion to wild-type was frequent (10 to 15% of plants in each generation). That is, they did not obtain plants with increased activity, and lines with reduced activity were unstable. In addition, plants with less than 50% of wild-type ADK activity had a characteristic dwarf phenotype, the severity of which correlated with the amount of residual activity. Based on these results, we speculate that ADK activity is tightly regulated and that attempts to artificially alter activity are overcome by post- translational mechanisms, by epigenetic events, or by some combination of these factors.

This may in part explain our observation that transgenic AL2 and L2 plants display only modest reductions (15-30%) in ADK activity (Figure 2.5).

Although we have been unable to directly examine the relevance of ADK activity to virus infection using transgenic approaches, there are several reasons why we are confident that inhibition of ADK activity by AL2 and L2 occurs and is an aspect of

geminivirus pathogenesis. First, we have demonstrated that AL2 and L2 interact with

and inhibit ADK in vitro. Using intact E. coli and yeast cells as well as extracts from

geminivirus infected plants and transgenic plants expressing AL2 and L2, we have also

shown that functional interaction occurs in vivo. Second, the interaction is not virus

specific. ADK interacts with AL2 from TGMV and CaLCuV (both New World

begomoviruses), with AL2 from ACMV (an Old World begomovirus), and with the more distantly related L2 from BCTV (a curtovirus). That the interaction is conserved across the AL2/L2 protein family suggests it is not fortuitous. Third, ADK presents an interesting target as it produces 5'-AMP which can activate SNF1. That AL2 and L2 inactivate both SNF1 and ADK seems more than coincidental. In addition, ADK's role in

sustaining the methyl cycle suggests another possible reason for targeting this nucleoside

kinase. Finally, we observed that ADK activity is increased following infection of plants

with an L2 deficient geminivirus as well as RNA viruses belonging to two different

families. This provides strong evidence that increasing ADK activity is a component of

the host response to virus challenge, and thus inhibition of ADK activity by the

geminivirus AL2 and L2 proteins is most likely a deliberate counter-response.

In conclusion, ADK inhibition by geminivirus AL2 and L2 proteins may be a

mechanism to suppress the induction of SNF1-mediated responses, to suppress RNA

silencing or other cellular defense pathways that depend on methylation, or both. It will

be interesting to learn more about the roles of ADK and SNF1 in host defense and about

the roles of AL2 and L2 in viral pathogenesis.

2.4 Methods

2.4.1 Two-hybrid analysis

The yeast two-hybrid system was used to identify interactions between AL21-115

and proteins expressed from an Arabidopsis cDNA library (Durfee et al., 1993; Harper et

al., 1993). The cDNA library was obtained from the ABRC (Ohio State University,

Columbus). TGMV AL2 was obtained by PCR (forward primer: 5'-

GCGGGCGCCATGCGAAATTCGTCTTCC; reverse primer: 5'-

GCGGAGCTCCTAAAGTTGAGAAATGCC) and cloned into the NcoI site of pAS2 to produce pAS-AL21-115. NcoI cleaved pAS2 was previously rendered blunt-ended by

treatment with the Klenow fragment. BCTV-Logan L2 was obtained by PCR (forward

primer: 5’-GCGCCATGGAAAACCACGTG; reverse primer: 5'-

GCGGATCCTTATCCAAGTATATCTC) and inserted as an NcoI-BamHI fragment into

pAS2 and pACT2 to generate pAS-L2 and pACT-L2. ADK cDNA larger than the

original 605 bp partial fragment was obtained by PCR using the Arabidopsis cDNA

library as template with the reverse primer 5'-

CAGCTCGAGAAGCTTAGTTAAAGTCGGGCTTCTCAGGC and forward primers 5'-

GGGTACCTCTAGAATGGCTTCTTCTTCTAA (full-length ADK) or 5'-

GGGTACCTCTAGAATGGGTAACCCACTCCTC (ADK-∆N13). PCR products,

digested with XbaI and XhoI, were inserted into pFastBacHTb (Life Technologies,

Carlsbad, CA) to generate pHT-ADK and pHT-ADK-∆N13. The ADK insert was

excised from pHT-ADK as an NcoI-XhoI fragment and inserted into pAS2 or pACT to

generate pAS-ADK and pACT-ADK. Positive interaction was indicated by the ability of

Y190 cells cotransformed with bait and prey constructs (expressed in pAS2 or pACT2,

respectively) to grow on medium lacking histidine and containing 50 mM 3-amino

triazole. The medium also lacked leucine and tryptophan to ensure the maintenance of

expression plasmids. Additional confirmation of interactions was obtained by assessing

β-galactosidase activity using a filter lift assay.

2.4.2 E. coli complementation experiments

ADK, AL2, and L2 were expressed in the purine deficient strain HO4 (purE,

purF, deoD, apt) (Hove-Jensen and Nygaard, 1989), and complementation was assessed

by the ability of cells to grow on selective minimal medium (M9 containing 10 µg/ml

methionine, 1 µg/ml thiamin, 0.2% mannitol, 1 mM IPTG, 100 µM adenosine, and 0.1

µM 2-deoxycoformycin, which prevents deamination of adenosine to inosine). Media

also contained antibiotics appropriate for maintenance of expression plasmids: ampicillin

(100 µg/ml) for pBSK and pMAL, and kanamycin (50 µg/ml) for pDHK29. ADK

sequence was obtained from pHT-ADK as an SpeI-XhoI fragment and inserted in

pBluescript SK (pBSK; Stratagene) to create pBSK-ADK. Inserts from pHT-ADK and

pHT-ADK∆N13 were obtained by digestion with XbaI and HindIII and cloned into

pMAL-c2x (New England Biolabs) to generate pMAL-ADK and pMAL-ADK∆N13. A

BamHI-HindIII fragment obtained from pHT-ADK was inserted into pMAL to create

pMAL-ADK∆N88. Full-length TGMV AL2 was obtained by PCR (upstream primer: 5'-

CGCAGATCTGAATTCATGCGAAATTCGTCTTCCTCA; downstream primer: 5'- 91

GCGGAGCTCCTATTTAAATAAGTTCTC) and inserted as a BamHI-EcoICRI

fragment into the BamHI/EcoRV sites of pDHK29 (Phillips, Park, and Huber, 2000) to

create pDHK-AL2. BCTV L2 was removed from pAS-L2 as an NcoI-BamHI fragment and inserted into pDHK29 to generate pDHK-L2. Yeast construct pACT-SNF1, pACT-

SNF1-KD were made as described (Hao et al., 2003). To generate constructs pDHK-

SNF1 and pDHK-SNF1-KD, first NcoI-EcoRI SNF1 fragment from pGEX-2T-SNF1

(Hao et al., 2003) was cloned into pRSET-B NcoI-EcoRI site to produce pRSET-SNF1.

Then BamHI-EcoRI fragment from pRSET-SNF1 was cloned into pDHK29 BamHI-

EcoRI site to generate pDHK-SNF1. The BamHI-XbaI fragment from pRSET-SNF1 was

cloned to pDHK30 BamHI-XbaI site to generate pDHK-SNF1-KD.

2.4.3 Yeast complementation experiments

The Saccharomyces cerevisiae ado1 deletion strain (record number 2583) and the

BY4741 parent strain were obtained from Research Genetics, Huntsville, AL. (BY4741:

ADO1, MATa, his3D1, leu2D0, met15D0, ura3D0.) The ADK sequence was excised

from pHT-ADK as a Stu1-XhoI fragment and ligated with SmaI-SalI cleaved YEpL to create YEpL-ADK. YEpU-L2 has been previously described (Hao et al., 2003). Growth experiments were carried out by inoculating synthetic defined medium lacking uracil and leucine with equal numbers of cells from overnight cultures. Cells were incubated at

30°C. In all cases, cells contained expression plasmids providing leucine prototrophy

(YEpL-ADK2 or YEpL) and uracil prototrophy (YEpU-L2 or YEpU). The YEpL and

YEpU plasmids express the LacZ α-peptide, whose coding sequence was replaced by

ADK and L2, respectively.

2.4.4 Protein expression

Histidine-tagged ADK, various ADK deletion derivatives, and CAT were

expressed from pRSET vectors (Invitrogen) in E. coli BL21 (DE3) pLysS (Stratagene)

and partially purified using nickel-NTA resin (Qiagen). Inserts from pHT-ADK and

pHT-ADK∆N13 were obtained by digestion with NcoI and HindIII and cloned into

pRSET-B (Invitrogen) to generate pRSET-ADK and pRSET-ADK∆N13. SalI-HindIII

and BamHI-HindIII fragments obtained from pRSET-ADK were inserted into pRSET-B

to create pRSET-ADK∆N24 and pRSET-ADK∆N88, respectively. Expressed proteins

were visualized by staining with Coomassie Brilliant Blue R250 following

polyacrylamide gel electrophoresis, and protein concentrations were estimated by

comparing band intensities with bovine serum albumin standards. Gel blot analysis was

performed using an anti-histidine tag antibody. AL2 and L2 proteins fused to

glutathione-S-transferase (GST-AL2 and GST-L2) were likewise expressed in BL21 cells

and partially purified by glutathione-agarose chromatography. Histidine tagged AL2

(His-AL2) was expressed in insect (Sf9) cells using a baculovirus vector. Conditions for

expression and purification of the viral proteins have been previously described (Hartitz,

Sunter, and Bisaro, 1999).

2.4.5 SNF1 kinase assay

SNF1 kinase assay with ADK as substrate was performed primary as before with some modifications (Hao et al., 2003). His-SNF1-KD about 100ng was mixed with

100ng His-ADK, 250 ng His-ADK-∆N13, 200ng His-ADK-∆N24, 200ng His-ADK-

∆N88 or 200 ng His-CAT, respectively, in SNF1 kinase buffer for 30 min at 30˚C. The reaction mixtures were separated on 15% SDS polyacrylamide gel and transferred to a nitrocellulose membrane. The membrane was examined with phosphoimager.

2.4.6 ADK assays

Reactions were performed in a total volume of 15 µl and contained 50 mM Tris-

32 HCl (pH 7.6), 5 mM MgCl2, 1 µM adenosine, 5 µCi [γ- P]ATP (3,000 Ci/mmol) with

10-70 ng ADK. After incubation at 37˚C for 20 min, reactions were stopped by addition of EDTA to 30 mM. Products were resolved by thin layer chromatography (TLC) on polyethyleneimine-cellulose plates developed with 1 M acetic acid. To measure ADK activity affected by SNF1, ADK (10 ng) and SNF1-KD (0, 4, 10, 20 ng) were added to reaction mixtures containing [γ- 32P]ATP except the substrate adenosine and incubated for 10 min at 30˚C. Then adenosine was added to start the ADK kinase reaction for 30 min at 30˚C. Product of AMP was monitored by TLC as above.

To measure ADK activity in plants, stem pieces (3 mm) were obtained from transgenic N. benthamiana plants (AL2 line 472-1 and L2 line CTL2-6) (Sunter, Sunter, and Bisaro, 2001) and comparable non-transgenic plants approximately 4 weeks post-

germination. Crude extracts were obtained from 25 mg of stem tissue in 500 µl 50 mM

HEPES, 10 mM MgCl2, 1 mM EDTA, 1 mM EGTA, 1 mM benzamidine, 5 mM DTT,

50 mM NaF, and 0.1 mM PMSF. Crude yeast extracts were obtained using the same

buffer. Total protein concentration was estimated using the Bradford assay. In

independent experiments, equal amounts of extracted protein from each sample (150-400

ng for plant experiments, 300 ng for yeast experiments) were added to the in vitro ADK

assay. Labeled AMP generated was quantitated using a phosphorimager. Extracts were

also obtained from plants infected 4 weeks post-germination, and tissue was harvested 14

days post-inoculation. PVX and CMV were introduced by standard mechanical

inoculation of infected sap. Plants were mechanically inoculated with TGMV DNA or agroinoculated with wild-type BCTV or BCTV L2 mutant virus as previously described

(Sunter, Sunter, and Bisaro, 2001). Mock inocula consisted of buffer or Agrobacterium

lacking viral DNA.

2.4.7 Immunolocalization of AL2

N. benthamiana plants were agroinoculated with TGMV or mock-inoculated with

A. tumefaciens containing the binary vector alone as described previously (Sunter, Sunter,

and Bisaro, 2001). Systemically infected leaf tissue from TGMV inoculated plants, or

comparable asymptomatic leaf tissue from mock inoculated plants, was used for

immunolocalization. Epidermis was peeled from the lower surface of several leaves and

the tissue fixed in 4% paraformaldehyde in phosphate buffered saline (PBS)/0.1% Triton

X-100 for 10 minutes on ice. Fixation was stopped with 3X PBS for 3 minutes on ice, 95

followed by three washes in PBS/0.1% Triton X-100. The tissue was stored in the same

buffer for 3 hours prior to immunolabeling.

Antibody labeling and washing steps were carried out at 4°C. Fixed tissue was

incubated with rabbit-derived primary antibody prepared against full-length TGMV AL2

protein (Cocalico) at a 1:200 dilution in PBS overnight with shaking. After at least 6

washes with PBS/0.1% Triton X-100 with shaking overnight, tissues were incubated

overnight with shaking with FITC-conjugated goat anti-rabbit IgG secondary antibody

(1:1000 dilution in PBS/0.1 % Triton X-100). Following at least 6 washes with

PBS/0.1% Triton X-100 for 6 hours, the tissues were treated with DAPI (10 µg/ml) for nuclear staining. Tissue samples were examined under a fluorescence microscope

(Axioskope, Carl Zeiss), and bright field images were taken with differential interference

contrast (DIC). FITC fluorescence was detected using a filter set with excitation at 485

nm and emission at 515 nm. DAPI fluorescence was detected using a filter set with

excitation at 390 nm and emission at 460 nm.

2.4.8 Subcellular localization of AL2

Native AL2 protein was expressed in insect cells using the Bac-to-Bac

baculovirus expression system (Invitrogen) essentially as described (Hartitz, Sunter, and

Bisaro, 1999). The intact TGMV AL2 ORF was obtained from pTGA26 (Sunter et al.,

1990) by PCR and cloned into the baculovirus donor vector, pFastBac. Following

selection, the recombinant bacmid containing the AL2 ORF (pTGA800) was transfected

into Spodoptera frugiperda Sf9 cells. At 48 hours post-inoculation, infected cells were 96

counted, harvested by centrifugation, and resuspended in nuclear isolation buffer (NIB;

10 mM Hepes, pH 7.9, 1.5 mM MgCl2, 10 mM KCl, 0.2 mM PMSF, 0.5 mM DTT) at a

rate of 10 x 106 cells/ml. Resuspended cells (100 x 106) were subjected to Dounce

homogenization to release nuclei followed by centrifugation at 3,300 x g to pellet nuclei.

The supernatant was removed and subjected to centrifugation at 100,000 x g for 1 hour to

generate a soluble cytoplasmic protein fraction (S-100). Pelleted nuclei were washed

with an equal volume of ice-cold 10% glycerol, centrifuged at 3,300 x g, and resuspended

in a final volume of 5 ml 10% glycerol. Nuclei were further purified by centrifugation

through a 2.1 M sucrose pad at 120,000 x g for 90 min. Purified nuclei were resuspended

in 2 ml NIB containing 10% glycerol and stored at –80°C.

Soluble cytoplasmic and nuclear fractions (equivalent to 5 x 105 cells) containing

AL2, or chloramphenicol acetyltransferase (CAT) as a control, were electrophoresed

through 4-20% Tris-Glycine-SDS polyacrylamide gels (Invitrogen) in TGS buffer (25

mM Tris, 192 mM glycine, 0.1% (w/v) SDS, BioRad). Following electrophoresis,

proteins were transferred to nitrocellulose and native AL2 was detected by Western blot

using a primary antibody (1:10,000 dilution) directed against the full-length AL2 protein,

and a secondary alkaline phosphatase-conjugated goat-anti-rabbit antibody (1:10,000

dilution; Santa Cruz Biotechnology). NBT and BCIP (Promega) were used for color

development.

Figure 2.1 AL2 and L2 abolish complementation of E. coli HO4 by ADK

(A) Complementation of E. coli HO4 growth by ADK. Cells were transformed with the

indicated expression plasmids and streaked on complete medium (LB) or minimal

medium containing adenosine. Growth was evaluated following incubation at 30˚C for 4 days. pMAL and pBSK indicate empty plasmid vectors.

(B) Inhibition of HO4 complementation upon co-expression of ADK with AL2 or L2.

Cells were transformed with the indicated expression plasmids. pDHK29 indicates empty plasmid vector, and pDHK-CAT a plasmid expressing chloramphenicol acetyl transferase (negative control protein).

Figure 2. 1 AL2 and L2 abolish complementation of E. coli HO4 by ADK

Figure 2.2 ADK expression and in vitro activity

His-tagged ADK, the indicated ADK deletion derivatives, and CAT were expressed in E. coli BL21 cells and partially purified using nickel-NTA resin.

(A) Coomassie stained polyacrylamide gel showing the partially purified proteins. ADK protein concentrations were estimated by comparing band intensities with bovine serum albumin standards.

(B) Immunoblot of the same gel probed with anti-histidine tag antibody (anti-his tag).

(C) Demonstration of in vitro ADK activity. Reactions contained the substrates adenosine and ATP with 10-70 ng ADK. Products were resolved by thin layer chromatography. An autoradiograph is shown. Control ATP, ADP, AMP, and adenosine

(spotted individually to the right) were visualized with ultraviolet light.

100

Figure 2.2 ADK expression and in vitro activity

101

Figure 2.3 Recombinant AL2 and L2 proteins inhibit ADK activity in vitro

(A) Autoradiograph of a chromatogram showing AMP generated in ADK reactions preincubated with the indicated proteins prior to ATP addition, as described in the text.

(B) ADK activity is proportionally reduced by preincubation with increasing amounts of

AL2. Reactions contained 10 ng ADK and varying amounts of His-AL2.

increasing His-AL2:ADK ratio. ADK activity (AMP/AMP+ATP) in each reaction was

calculated following phosphorimager quantitation of radioactivity in individual spots.

102

Figure 2.3 Recombinant AL2 and L2 proteins inhibit ADK activity in vitro 103

Figure 2.4 L2 reduces complementation of a yeast ADK deletion strain by Arabidopsis

ADK2. A representative experiment (one of three replicates) is shown

(A) Complementation of a yeast ADK deletion strain (ado1) by ADK2, and inhibition by

L2. Growth of the ado1 mutant strain and the parent strain (BY4741; ADO1, ura, leu) expressing the indicated proteins was assessed in synthetic complete medium lacking uracil and leucine. In all cases, cells were cotransfected with expression plasmids providing leucine prototrophy (YEpL-ADK2 or YEpL) and uracil prototrophy (YEpU-L2 or YEpU). The YEpL and YEpU plasmids express the LacZ α-peptide.

(B) ADK activity is reduced by L2. Crude extracts (300 ng protein) were obtained from the indicated yeast cells and examined using the in vitro ADK assay described previously.

(C) Relative ADK activity. The graph shows relative ADK activity (from B) in extracts from ado1 cells expressing ADK2 alone or ADK2 + L2, relative to BY4741 cells

(ADO1). Labeled AMP was quantitated using a phosphorimager.

104

Figure 2.4 L2 reduces complementation of a yeast ADK deletion strain by Arabidopsis

ADK2 105

Figure 2.5 ADK levels are reduced in transgenic plants expressing AL2 and L2, and also

during geminivirus infection

(A) ADK activity is reduced in transgenic N. benthamiana plants expressing AL21-100 and

L2. Crude extracts obtained from comparable stem samples from just below the shoot apex were obtained and analyzed for ADK activity as described in the text, using equal

amounts of protein extract from each sample (150-400 ng in different experiments).

Labeled AMP produced was quantitated using a phosphorimager. The observed

differences between non-transgenic and transgenic plants were statistically significant, as determined by Student's t-test (P < 0.05). Graph values represent the mean +/- SE.

(B) ADK activity is reduced in TGMV and BCTV infected plants. Comparable stem

samples from just below the shoot apex were obtained from mock-inoculated N.

benthamiana plants, and plants infected with wild-type BCTV, a BCTV L2 mutant,

TGMV, PVX, or CMV, and analyzed as described above and in the text. In all cases, the

observed differences between virus infected and control plants proved statistically

significant, as determined by Student's t-test (P < 0.05). Graph values represent the mean

+/- SE.

106

Figure 2.5 ADK levels are reduced in transgenic plants expressing AL2 and L2, and also during geminivirus infection

107

Figure 2.6 AL2 is in the nucleus and the cytoplasm of TGMV infected cells Epidermal cells from TGMV infected plants (top three rows) or mock inoculated plants

(bottom row) were examined under bright field illumination using differential

interference contrast (DIC). Nuclei were located by DAPI staining (blue; first column) and AL2 protein was detected using AL2 primary antiserum followed by FITC- conjugated secondary antiserum (green; second column). A DAPI + FITC merge is shown in the third column. The irregular cells shown in the top row were originally positioned over mesophyll, whereas the elongated cells shown in lower rows were associated with veins.

108

Figure 2.6 AL2 is in the nucleus and the cytoplasm of TGMV infected cells

109

Figure 2.7 Localization of AL2 in insect cells Fractionated protein extracts from Sf9 cells infected with recombinant baculoviruses expressing either native TGMV AL2 or CAT (negative control protein) are shown.

Cytoplasmic (S100) and nuclear fractions were prepared 48 hours post-inoculation and equivalent amounts (representing 5 x 105 cells) were electrophoresed on 4-20% polyacrylamide-SDS gels and subjected to Western blot analysis using an AL2-specific antibody (anti-AL2). AL2 samples were also incubated at 37° for 3 hours with and without calf intestinal alkaline phosphatase (CIAP) prior to electrophoresis. The positions of non-phosphorylated and phosphorylated AL2 are indicated to the right. The two phosphorylated species are indicated by an asterisk.

110

Figure 2.7 Localization of AL2 in insect cells

111

Figure 2. 8 SNF1 and SNF1-KD abolish complementation of E. coli HO4 by ADK Inhibition of HO4 complementation upon co-expression of ADK with SNF1 protein kinase. Cells were transformed with the indicated expression plasmids. Empty plasmid vector control is pDHK29, and CAT protein expressed from pDHK-CAT plasmid is a negative control protein. SNF1-KD also has the same inhibition of cell growth complemented by ADK as SNF1.

112

Figure 2. 8 SNF1 and SNF1-KD abolish complementation of E. coli HO4 by ADK

113

Figure 2. 9 Potential SNF1 phosphorylation sites in ADK protein

Underlined amino acid sequences were potential SNF1 phosphorylation sites based on the SNF1 signature sequence. Amino acid S56 and T105 are phosphorylated. Restriction enzyme sites indicate the different ADK truncated proteins.

114

Figure 2. 9 Potential SNF1 phosphorylation sites in ADK protein

115

Figure 2. 10 ADK protein is phosphorylated by SNF1 protein kinase

ADK protein phosphorylation by SNF1 kinase was performed primary as before (Hao et al., 2003). His-SNF1-KD about 100 ng and 100 ng His-ADK, 250 ng His-ADK-∆N13,

200 ng His-ADK-∆N24, 200 ng His-ADK-∆N88 or 200 ng His-CAT were mixed with

SNF1 kinase buffer.

116

Figure 2. 10 ADK protein is phosphorylated by SNF1 protein kinase 117

Figure 2. 11 ADK activity is affected by incubation with SNF1 protein kinase

ADK kinase assay was done as described in Methods. ADK (10 ng) and SNF1-KD (0, 4,

10, 20 ng) were added to reaction mixtures containing [γ-32P]ATP except the substrate

adenosine and incubated for 10 min at 30˚C. Then adenosine was added to start the ADK

kinase reaction for 30 min at 30˚C. Product of AMP was monitored by thin layer

chromatography (TLC). CAT protein was used as a negative control. Relative ADK

activity was also shown in the right graph.

118

AMP

ATP

Figure 2. 11 ADK activity is affected by incubation with SNF1 protein kinase

119

CHAPTER 3

ADENOSINE KINASE INHIBITION AND SUPPRESSION OF

RNA SILENCING BY GEMINIVIRUS AL2 AND L2

PROTEINS

3.1 Introduction

In eukaryotic cells, homology-dependent gene silencing that operates at the RNA level is a widespread phenomenon involved in a number of fundamental processes, including cellular defense against foreign gene expression (from viruses, transgenes, and transposons), developmental gene regulation via microRNAs, certain transcriptional gene silencing events, and the establishment of heterochromatic regions in chromosomes

(Hutvagner et al., 2001; Ketting et al., 2001a; Ketting et al., 1999; Voinnet, 2001;

Voinnet, 2002; Wu-Scharf et al., 2000). All of these processes involve the appearance of double-stranded RNA (dsRNA) homologous to a silenced gene, chromosomal region, or target RNA. The dsRNA may be generated from single-stranded RNA (ssRNA)

120

templates by host or virus-encoded RNA dependent RNA polymerase (RdRP) (Dalmay et

al., 2000; Mourrain et al., 2000), by transcription of inverted repeats by DNA dependent

RNA polymerases (Hutvagner et al., 2001; Ketting et al., 2001a), or by sense and

antisense transcription from converging promoters (Aravin et al., 2001). Emerging

molecular evidence suggests that various phenomena referred to as RNA interference in

animals, posttranscriptional gene silencing plants, and RNA quelling in fungi are

mechanistically related, so we will use the generic term RNA silencing.

A hallmark of RNA silencing is the appearance of 21-26 nucleotide (nt) double-

stranded RNA (dsRNA) homologous to the silenced gene or input RNA, referred to as

short interfering RNA (siRNA) (Hamilton and Baulcombe, 1999). These siRNAs are

produced from larger dsRNA by the action of an RNaseIII-like enzyme called Dicer

(Bernstein et al., 2001; Zamore et al., 2000). The siRNAs in turn guide another RNase-

containing complex, the RNA induced silencing complex (RISC), to homologous ssRNA

targets for degradation (Elbashir, Lendeckel, and Tuschl, 2001; Hammond et al., 2000).

In addition, siRNAs can provide primers to convert target mRNA into new dsRNA,

thereby amplifying the production of siRNA (Lipardi, Wei, and Paterson, 2001).

DNA methylation is also associated with the RNA silencing process. In plants,

RNA directed DNA methylation (RdDM) has been observed in integrated viroid cDNA

sequences, in silenced transgenes, and in transposons (Ingelbrecht et al., 1994;

Wasseneger et al., 1994). Virus induced gene silencing (VIGS), a method of infecting

plants with viruses carrying sequences homologous to a transgene (or an endogenous

gene), also results in silencing and methylation of transgenes and can limit accumulation

of the RNA virus vector (English, Mueller, and Baulcombe, 1996; Jones et al., 1999).

121

RdDM associated with RNA silencing occurs in the region of homology between the

inducing and target RNA (usually coding regions), and can result in dense methylation at

symmetrical and non-symmetrical cytosine residues (Pelissier et al., 1999). The role of

methylation in RNA silencing is presently unclear, although it has been suggested that it might indirectly amplify siRNA production by increasing the abundance of aberrant transcripts that might serve as template for RdRP (Bender, 2001).

It is clear that in plants and at least some animals, RNA silencing can act as an adaptive defense against virus infection (Li, Li, and Ding, 2002; Voinnet, 2001;

Waterhouse, Wang, and Lough, 2001). Viruses are both the inducers and targets of RNA silencing, and thus determine its specificity. An additional feature of this remarkable defense mechanism is that RNA silencing is systemically propagated in plants by a sequence specific signal (Palauqui et al., 1997; Voinnet and Baulcombe, 1997).

However, as surely as RNA silencing is an effective antiviral defense, viruses have evolved proteins to counter it (Voinnet, Pinto, and Baulcombe, 1999). Silencing suppressors have been found in virtually all plant virus groups and, as might be expected from the diversity of viruses that infect plants, many appear to affect different steps in the silencing pathway (Li and Ding, 2001). For example, the helper component protease

(HC-Pro) of Tobacco etch virus (TEV) interferes with a maintenance step, as it can reverse previously established silencing (Anandalakshmi et al., 1998; Brigneti et al.,

1998; Kasschau and Carrington, 1998; Llave, Kasschau, and Carrington, 2000).

Surprisingly, HC-Pro interacts with a cellular protein (rgs-CaM) that is itself a silencing suppressor, suggesting that it acts by stimulating an endogenous, regulatory pathway

(Anandalakshmi et al., 2000). In contrast, the p25 movement protein of Potato virus X

122

(PVX) interferes with an initiation step and prevents spread of the silencing signal, possibly by blocking the conversion of ssRNA to dsRNA (Voinnet, Lederer, and

Baulcombe, 2000). The coat protein (CP) of Turnip crinkle virus (TCV) achieves a similar effect, but appears to interfere with the production of siRNAs from dsRNA (Qu,

Ren, and Morris, 2003), whereas the p19 protein of Tomato bushy stunt virus (TBSV) and related tombusviruses acts by specifically binding siRNAs (Silhavy et al., 2002;

Vargason et al., 2003; Ye, Malinina, and Patel, 2003). The 2b protein of Cucumber mosaic virus (CMV) inhibits systemic transport of the silencing signal by an unknown mechanism (Guo and Ding, 2002). These and other suppressor proteins allow viruses to successfully infect plants, and at the same time provide a number of useful reagents with which to probe the mechanism of RNA silencing.

Geminiviruses are single-stranded DNA (ssDNA) viruses that infect a wide range of plants and cause considerable crop losses. These agents amplify their genomes in the host cell nucleus through double-stranded (dsDNA) intermediates that are packaged into minichromosomes (Bisaro, 1996; Gutierrez, 1999; Hanley-Bowdoin et al., 1999).

Because geminiviruses do not encode polymerases, they are entirely dependent on host machinery for DNA replication and transcription. However, viral proteins are required to provide a cellular environment favorable to replication, to potentate spread through the host, to initiate specific steps in replication and/or transcription, and to suppress host defenses. For example, the AL2 protein of Tomato golden mosaic virus (TGMV; genus

Begomovirus) is a transcription factor required for the expression of late viral genes

(Sunter and Bisaro, 1991; Sunter and Bisaro, 1992; Sunter and Bisaro, 1997; Sunter and

Bisaro, 2003). The 15 kDa AL2 protein (also known as AC2, C2, or transcriptional

123

activator protein; TrAP) has a C-terminal activation domain that is functional in plant,

yeast, and mammalian cells (Hartitz, Sunter, and Bisaro, 1999). It also has additional

functions in pathogenesis, and the AL2 proteins of African cassava mosaic virus and

Tomato yellow leaf curl virus (TYLCV) have been shown to be silencing suppressors

capable of reversing previously established silencing (van Wezel et al., 2002; Voinnet,

Pinto, and Baulcombe, 1999).

In addition, we have demonstrated that TGMV AL2, and the related L2 protein of

Beet curly top virus (BCTV; genus Curtovirus) are pathogenicity factors that condition a unique and non-specific enhanced susceptibility phenotype when expressed in transgenic plants (Sunter, Sunter, and Bisaro, 2001). This enhanced susceptibility is attributable to their ability to interact with and inactivate SNF1 kinase (Hao et al., 2003). Because

BCTV L2 (which does not appear to be a transcriptional activator (Hormuzdi and Bisaro,

1995; Stanley et al., 1992) and TGMV AL2 lacking its C-terminal activation domain can both condition enhanced susceptibility, this function is clearly independent of transcriptional activation. We have also found that AL2 and L2 interact with and inactivate adenosine kinase (ADK), a nucleoside kinase that phosphorylates adenosine to produce 5' AMP (Chapter 2). Because AMP is known to activate SNF1 by a number of direct and indirect mechanisms, we have proposed that inactivation of SNF1 and ADK by

AL2/L2 represents a dual mechanism to counter SNF1-mediated responses.

ADK is involved in adenine and adenosine salvage pathways and is important for the synthesis of nucleotides and nucleotide cofactors. It may also regulate the interconversion of cytokinin bases and ribosides (von Schwartzenberg et al., 1998).

However, by recycling adenosine, ADK also plays a major role in sustaining the methyl

124

cycle and S-adenosylmethionine (SAM)- dependent methyltransferase activity in yeast

and plants (Lecoq et al., 2001; Moffatt et al., 2002; Weretilnyk et al., 2001). In yeast,

methylation deficiency appears to be the primary defect of ADK-null mutants. ADK

deficiency reduces methyltransferase activity in plants, and observational evidence

suggests that it also compromises the ability to maintain RNA silencing (Moffatt et al.,

2002; Chapter 2). Thus ADK might also be targeted by AL2/L2 for the purpose of RNA

silencing suppression.

In this chapter, we confirm that begomoviruses (TGMV) and curtoviruses

(BCTV) are targeted by RNA silencing, and that their respective AL2 and L2 proteins are

silencing suppressors. We further demonstrate that reducing ADK activity results in

RNA silencing suppression, whereas reducing SNF1 activity has no apparent or direct

effect on this process. We propose that AL2/L2 suppress RNA silencing by reducing

ADK activity and that AL2 has at least three distinct functions in geminivirus replication: transcriptional activation, suppression of a SNF1-mediated, innate defense response by

inhibition of SNF1 and ADK, and suppression of RNA silencing (adaptive defense) by

inhibition of ADK.

3. 2 Results

3.2.1 TGMV and BCTV induce RNA silencing in infected plants

Geminiviruses do not have a genomic RNA phase and do not require RdRP for

replication. Thus, unlike RNA viruses, they do not generate dsRNA directly and their

125

genomes are not ribonuclease substrates. Rather, they produce mRNA from dsDNA templates by the action of the host RNA polymerase II machinery, and it is these transcripts that are the likely targets of the RNA silencing system. The origin of the dsRNA signal that triggers RNA silencing following geminivirus infection is not known, but it may be produced by transcriptional read-through of the polyadenylation signals of convergent transcripts (Sunter and Bisaro, 1989), or by the action of RdRP on aberrant or over-abundant transcripts, or both. In any event, that geminiviruses can be used as vectors to silence endogenous genes and transgenes (Kjemtrup et al., 1998; Peele et al.,

2001), that geminiviruses replication can be inhibited by RNA silencing approaches

(Vanitharani, Chellappan, and Fauquet, 2003), that siRNAs homologous to the viral Rep gene can be detected in infected plants, and that geminiviruses encode silencing suppressors strongly indicates that a dsRNA signal is generated and that geminiviruses are in fact targeted by RNA silencing.

We confirmed that the begomovirus TGMV and the curtovirus BCTV induce and are targeted by RNA silencing by showing that siRNAs diagnostic of the process are present in infected N. benthamiana plants. As illustrated in Figure 3.1, siRNAs hybridizing to both sense and antisense probes corresponding to the entire genome or to the coat protein gene can be found in extracts from plants infected with TGMV or BCTV.

The siRNAs represent at least two distinct size classes consisting of a smaller 21-22 nt species and a larger 24-26 nt species. The small class has been associated with mRNA degradation while the larger correlates with systemic silencing and RdDM (Hamilton et al., 2002).

126

To our knowledge this is the first formal demonstration that curtoviruses are

targeted by RNA silencing. The results further indicate that, in addition to the early Rep gene, late viral late gene expression is also targeted by RNA silencing.

3.2.2 RNA silencing can be suppressed by AL2 and L2 proteins and by inhibiting ADK

We set out to confirm that TGMV AL2, like its begomovirus counterparts ACMV

and TYLCV, is a silencing suppressor and to determine whether the related BCTV L2

protein also has this activity. This study employed a three component, transient system

that involves agroinfiltration of N. benthamiana leaves with plasmids that express: (1)

green fluorescent protein (GFP), (2) an inverted repeat RNA as a strong silencing trigger

(dsGFP), and (3) a test protein or construct (Johansen and Carrington, 2001). Test constructs expressing the known viral silencing suppressors HC-Pro, p19, and AL2

(AC2) from ACMV were used as positive controls, and a construct expressing β-

glucuronidase (GUS) was used as a negative control. The results of these experiments

clearly showed that TGMV AL2 and BCTV L2 can in fact suppress silencing, as judged

by the appearance of green color following illumination of infiltration zones with

ultraviolet light (Figure 3.2) and by the accumulation of GFP mRNA (Figure 3.3A).

AL21-100, which lacks the activation domain, also suppressed silencing, confirming that

transcriptional activation and RNA silencing suppression are separate activities. Further,

like the p19 and AC2 positive controls, AL2, AL21-100, and L2 inhibited the accumulation

of GFP-specific siRNAs of both size classes (Figure 3.3B). In contrast, accumulation of 127

GFP mRNA was not supported and GFP-specific siRNAs were found to accumulate in tissue infiltrated with the GUS expression plasmid. While the HC-Pro control clearly

allowed the accumulation of GFP mRNA, it nevertheless failed to strongly suppress

siRNAs. Both classes of siRNA, and particularly the larger class, were observed to

accumulate by 5 days post-infiltration. The results obtained with HC-Pro and p19 are in general agreement with the previous findings of others (Hamilton et al., 2002; Johansen

and Carrington, 2001; Silhavy et al., 2002).

We next asked whether reducing expression of the two kinases inhibited by AL2

and L2 had any effect on GFP silencing. We adopted an RNA silencing approach to this

question, based on the observation that components of the RNA silencing machinery, and

by analogy other proteins that might play a supporting role, can be at least partially

silenced (Bernstein et al., 2001). Intron-containing, inverted repeat constructs were

prepared for Arabidopsis ADK2 (dsADK) and the Arabidopsis SNF1 kinase AKIN11

(dsSNF1) (Smith et al., 2000). ADK and SNF1 genes and proteins are highly conserved

among plants, animals, and yeast, and both plant proteins can functionally complement

the corresponding yeast mutants (Hao et al., 2003; Wang et al., 2003). Thus it was

considered likely that the Arabidopsis sequences would be effective in N. benthamiana.

Remarkably, addition of the dsADK construct as the third component (GFP + dsGFP + dsADK) suppressed silencing and allowed GFP expression while inhibiting the accumulation of both classes of GFP siRNAs (Figure 3.3). The dsSNF1 construct had no apparent effect on the silencing triggered by dsGFP.

The results obtained with the dsADK construct prompted us to search for other means of inhibiting ADK. Using an in vitro ADK assay that measures phosphorylation

128

of adenosine in the presence of γ-32P-ATP (Chapter 2), it was determined that the adenosine analogue A-134974 (RBI), an inhibitor of mammalian ADK, also inhibits

ADK activity when added to extracts from wild-type N. benthamiana plants over a range

of 0.01 to 10 µM (Figure 3.4). In subsequent experiments, infiltration of RBI (10 µM) along with GFP and dsGFP constructs (GFP + dsGFP + RBI) also allowed accumulation

of GFP mRNA and prevented accumulation of GFP-specific siRNAs (Figure 3.3).

We concluded from these experiments that AL2 and L2 are both RNA silencing suppressors and that silencing suppression does not require the AL2 activation domain.

We further concluded that inhibition of ADK activity, either at the RNA level by

expression of dsRNA construct directed against ADK mRNA, or at the protein level by

addition of a specific ADK inhibitor, also results in RNA silencing suppression with a

similar outcome in the transient system. Specifically, GFP mRNA is observed to

accumulate and both classes of siRNAs are suppressed.

3.2.3 ADK activity is reduced in tissues infiltrated with AL2/L2 protein, dsADK, or RBI

We previously showed that AL2 and L2 are effective inhibitors of ADK activity

both in vitro and in vivo, including geminivirus infected tissue (Chapter 2). Thus we

asked whether suppression of RNA silencing could be correlated with reductions in ADK

activity. ADK activity levels in extracts from tissues infiltrated with various test

constructs were determined using the in vitro ADK assay. As expected, tissues receiving

(GFP + dsGFP) + AL2, L2, dsADK, or RBI displayed RNA silencing suppression and 129

showed reductions in ADK activity averaging 50 to 90% (Figures 3.3 and 35). In contrast, tissues infiltrated with (GFP + dsGFP) + GUS, dsSNF1 or HC-Pro did not display significant reductions in ADK activity relative to control tissue infiltrated with vector DNA, regardless of whether the constructs did (HC-Pro) or did not (GUS, dsSNF1) suppress RNA silencing.

We concluded that suppression of RNA silencing correlates with significant reductions in ADK activity in the case of the AL2, L2, dsADK, and RBI samples. In contrast, even though HC-Pro is a very efficient silencing suppressor, it has no effect on

ADK activity levels.

3.3 Discussion

We demonstrated previously that the AL2 and L2 proteins interact with and inactivate ADK in vitro and in vivo, and that ADK activity is reduced in TGMV and

BCTV-infected tissue in a manner that depends on the presence of functional AL2/L2 protein. We also found that ADK activity is increased in response to infection with PVX,

CMV, and a BCTV L2 mutant, suggesting that increasing the activity of this enzyme is part of the host response to virus infection (Chapter 2).

ADK plays a key role in the adenosine salvage pathway, and is important for the synthesis of nucleic acids, nucleotide co-factors, and the regulation of cellular energy charge. In this context, it is interesting that elevated AMP:ATP ratios, which occur under conditions of nutritional and environmental stress, result in the activation of SNF1

(Hardie, Carling, and Carlson, 1998; Sugden et al., 1999). That AL2 and L2 inactivate 130

both SNF1 and ADK reinforces our proposal that SNF1-mediated responses are an

important component of innate defenses, and further suggests that inactivation of the two kinases represents a dual strategy to counter this defense (Hao et al., 2003). Implicit in this model is the idea that AMP generated by ADK activates SNF1 in response to virus challenge. Such a mechanism could provide a more rapid and effective AMP signal than

ATP depletion resulting from virus replication.

By recycling adenosine, ADK also plays a critical role in sustaining the methyl cycle and SAM-dependent methyltransferase activity (Figure 1.3), and ADK-deficient yeast and Arabidopsis exhibit reduced methylation (Lecoq et al., 2001; Moffatt et al.,

2002; Weretilnyk et al., 2001). Thus another connection between ADK and viral

pathogenesis relates to RNA silencing suppression, which is associated with methylation.

In this light, our observation of increased ADK activity following virus infection (in the

absence of AL2/L2 proteins) may reflect increased methyl demand brought about by the

activation of silencing mechanisms. Further, we and others have had difficulty

generating transgenic plants with altered ADK activity. The few silenced lines with

reduced ADK activity were unstable and reverted at high frequency, suggesting they

were not able to maintain the silenced state (Moffatt et al., 2002). Thus we have also

proposed that AL2 and L2 may suppress RNA silencing by inhibiting ADK (Wang et al.,

2003).

In this chapter, we confirm that the geminiviruses TGMV and BCTV induce and

are targeted by RNA silencing in the course of a normal infection, and that the AL2 and

L2 proteins they encode are capable of suppressing RNA silencing in a transient three

component system. We also examined the effects on RNA silencing of inhibiting the two

131

kinases known to be targeted by AL2 and L2. Remarkably, we found that inhibiting

ADK activity at the RNA level by expression of a dsRNA construct directed against

ADK, or at the protein level by the use of an ADK inhibitor (the adenosine analogue

RBI), also results in silencing suppression. Direct measurement of ADK activity in tissue

showing silencing suppression following infiltration with AL2, L2, dsADK, or RBI

revealed that in all cases ADK activity was significantly reduced (>50%). These data

provide strong evidence that AL2 and L2 suppress silencing by inhibiting ADK activity.

A comparison of the suppression mechanisms employed by the proteins used in

this study is informative. The p19 protein cannot reverse previously established silencing

(Silhavy et al., 2002). HC-Pro and AL2 can both reverse silencing, indicating that they

suppress a step required for maintenance (Anandalakshmi et al., 1998; Brigneti et al.,

1998; Kasschau and Carrington, 1998; Llave, Kasschau, and Carrington, 2000).

However, all AL2 proteins tested here (TGMV AL2 and AL21-100, ACMV AC2, and

BCTV L2) effectively suppressed siRNA accumulation over the 5 day course of the

experiment, whereas the large siRNA class in particular was observed to accumulate in

the presence of HC-Pro by 5 days post-infiltration (Figure 3.3B). In addition, the

AL2/L2 proteins inhibited ADK activity while HC-Pro did not. Thus reduced ADK

activity is not a general effect of RNA silencing suppression, but rather is specifically

caused by the AL2/L2 proteins, dsADK, and RBI. We conclude that the AL2/L2

proteins, HC-Pro, and p19 suppress silencing by different mechanisms, and that ADK

inhibition best explains the ability of AL2 and L2 proteins to suppress RNA silencing.

We hypothesize that reducing ADK activity suppresses silencing because it

inhibits methyltransferase activity required for RdDM. DNA methylation is important

132

for RNA silencing, since silencing is partially reversed by mutations that globally reduce

methylation, including met1 (methyltransferase 1, a maintenance methylase) and ddm1

(DNA demethylation 1, a homologue of yeast SWI2/SNF2 homologue) (Morel et al.,

2000). Partial reversion has also been noted following 5 azacytidine-induced hypomethylation (Kovarik et al., 2000). These observations suggest that methylation has a role in maintenance. Certainly there are many examples in which RNA silencing and

DNA methylation are strongly correlated (English, Mueller, and Baulcombe, 1996;

Ingelbrecht et al., 1994; Jones, Thomas, and Maule, 1998; Jones et al., 1999). However, there are also instances where correlation is lacking, and RNA silencing and methylation can be uncoupled by the HC-Pro silencing suppressor (Jones et al., 1999; Mallory et al.,

2001). Thus the relationship between DNA methylation and RNA silencing, and the role of DNA methylation in this process, remains unclear. However, the identification of

ADK as the target of silencing suppressor proteins of DNA viruses provides yet another link between silencing suppression and methylation, and suggests that DNA methylation

(and/or specific histone methylation; see below) may be more important for this process than is generally recognized.

In addition to RdDM, there is a growing body of evidence that dsRNA can also alter histone methylation. The histone code hypothesis suggests that modification of particular amino acids in highly conserved histone tails strongly influences chromatin structure and gene expression (Jenuwein and Allis, 2001). Of particular interest are histone H3 methylation on lysine 4 (H3mK4), which is associated with active chromatin, and methylation of lysine 9 (H3K9), which is enriched in inactive chromatin. In addition, the latter modification is required for binding of heterochromatic protein 1 (HP1) to DNA

133

to establish silenced chromatin regions (Bannister et al., 2001; Jacobs et al., 2001). In

Schizosaccharomyces pombe, which lacks DNA methylation, it has been shown that

deletion of RNA silencing machinery components results in the loss of heterochromatic

silencing and H3K9 methylation (Volpe et al., 2002). This has led to the proposal that an

RNA signal may trigger sequence specific H3K9 methylation, which in turn might guide

DNA methyltransferases in higher organisms. In Neurospora crassa, H3K9 methylation

precedes and is required for DNA methylation, since mutation of dim5 (a histone H3K9

methyltransferase) or mutation of H3K9 results in a complete loss of DNA methylation

(Tamaru and Selker, 2001). Additional links between histone and DNA methylation have

been established recently in Arabidopsis (Gendrel et al., 2002; Jackson et al., 2002;

Johnson, Cao, and Jacobsen, 2002). The kryptonite gene (kyp) was isolated in a screen

for suppressors of an allele silenced by cytosine methylation (Jackson et al., 2002b). Its

product is a methyltransferase specific for H3K9. Loss of function kyp alleles have the

same phenotype as loss of function alleles of the DNA methyltransferase

chromomethylase 3 (cmt3). Mutation of either gene results in a loss of CpNpG

methylation but not CpG methylation (Bartee, Malagnac, and Bender, 2001; Jackson et

al., 2002; Lindroth et al., 2001). Because CMT3 does not directly bind DNA, but does

bind an HP1 homologue that recognizes H3mK9, CMT3 is likely directed by the same

modified histone residue that recruits HP1. Thus H3K9 methylation occurs first,

followed by cytosine methylation. While this progression remains to be conclusively

proved, it is clear that there is cross-talk between these epigenetic marks, and both could

be important for RNA silencing.

134

The geminivirus genome is transcribed in the nucleus from dsDNA templates that assemble into minichromosomes (Pilartz and Jeske, 1992), so it is likely subject to epigenetic modifications similar to those that affect nuclear gene expression. Our working model for AL2/L2 silencing suppression implies that the geminivirus genome is a target for DNA methyltransferases and possibly histone H3K9 methyltransferases. In support of this idea, we have shown previously that in vitro methylation of TGMV DNA severely impairs its ability to replicate on tobacco protoplasts. However, we also found that TGMV DNA isolated from infected N. benthamiana plants is not appreciably methylated (Brough et al., 1992). In light of the results reported here, it will be interesting to compare the DNA and H3 methylation patterns of wild-type TGMV and

BCTV and corresponding AL2 and L2 mutants, and to examine methylation in hosts that may be less permissive for replication.

Our results show that in the transient silencing system, expression of a dsRNA construct directed against SNF1 does not result in silencing suppression. However, the

SNF1 kinase targeted in these experiments (AKIN11) is a representative of only one of three subfamilies of SNF1-related kinases in plants, which may well have some redundant function (Chikano et al., 2001; Halford and Hardie, 1998). So at this time we cannot definitively rule out a role for SNF1-related kinases in RNA silencing. However, it is interesting that AL21-100 and L2 do not behave as silencing suppressors when constitutively expressed in transgenic plants (Sunter, Sunter, and Bisaro, 2001). A silencing suppressor would be expected to affect virulence, resulting in increased disease symptoms and virus accumulation (Pruss et al., 1997). In contrast, the enhanced susceptibility phenotype seen in these transgenic lines primarily affects infectivity:

135

plants can be infected with lower inoculum doses, but increases in symptom severity and

virus load are not observed. We believe this is due to the fact that AL2 and L2 inactivate

both SNF1 and ADK. Because ADK deficiency can result in a dwarf phenotype, low level expression of the transgenes is favored, and the modest reductions (15-30%) we

observed in our transgenic lines supports this contention (Moffatt et al., 2002). The

residual ADK activity in transgenic plants expressing AL21-100 or L2 is apparently

sufficient to support silencing, and the more subtle enhanced susceptibility phenotype due

to SNF1 interaction is observed instead (Hao et al., 2003).

In conclusion, our results to date support the idea that TGMV AL2 has three

distinct functions. First, AL2 stimulates late viral gene expression at the level of

transcription in a manner that depends on a functional activation domain. Second, AL2

causes enhanced susceptibility by suppressing SNF1-mediated responses. This function

requires SNF1 interaction, but does not require the activation domain. Inactivation of

ADK may contribute to inhibition of SNF1. Third, AL2 suppresses RNA silencing

through its interaction with ADK, and this also does not require the activation domain.

The latter two functions are shared with the BCTV L2 protein.

136

3.4 Methods

3.4.1 Plant material, Agrobacterium infiltration and agroinoculation

Transgenic Nicotiana benthamiana plant line 16c, expressing GFP protein

(35S:GFP), was provided by Dr. David C. Baulcombe (Sainsbury Laboratory) and was

described previously (Brigneti et al., 1998; Ruiz, Voinnet, and Baulcombe, 1998).

Transgenic seeds were germinated on MSO medium containing kanamycin (300 mg/ml)

before seedlings were transferred to soil.

Agrobacterium infiltration of leaves was performed as described (Johansen and

Carrington, 2001; Llave, Kasschau, and Carrington, 2000; Voinnet, Lederer, and

Baulcombe, 2000). For co-infiltration, equal volumes of the indicated A. tumefaciens

cultures (OD600 = 1) were mixed before infiltration. A 1-cc syringe was used to infiltrate tissue from the underside of leaves. The ADK inhibitor, RBI (A-13497, sigma catalog

number A2846), was mixed with the bacterium culture just before infiltration at a final

RBI concentration of 100uM. Infiltrated leaves were harvested after 3 to 7 days and used

for RNA and protein isolation.

Agroinoculation of healthy wild type N. benthamiana plants with TGMV and

BCTV was performed as described (Sunter, Sunter, and Bisaro, 2001). Plants were

inoculated at 30 to 40 days after germination by injecting 10 µL of an A. tumefaciens

9 culture (OD600 = 1 = 10 cells/mL) into the base of each of three leaves using a

Hamilton syringe. A total inoculum volume of 30 µL was used per plant. Cells containing

137

TGMV component A and B were mixed in a 1:1 ratio. Infected plants were harvested

after 14 and 30 days post inoculation and used for RNA isolation.

3.4.2 Plasmid constructs for transient expression

Transient gene expression was driven by the cauliflower mosaic virus 35S

promoter and terminator. The base vector for most constructs, pRTL2 (Restrepo, Freed,

and Carrington, 1990), contained an enhanced 35S promoter from cauliflower mosaic

virus, the TEV 5'-non-translated sequence and the 35S terminator.

pRTL2-GUS, pRTL2-GFP, pRTL2-dsGFP and pRTL2-HC-Pro were provided by

Dr. James Carrington (Johansen and Carrington, 2001). pRTL2-AC2, pRTL2-AL2,

pRTL2-AL21-100 , and pRTL2-L2 were generated by cloning NcoI-BglII fragments of

AC2, AL2, AL21-100 and L2 into NcoI-BamHI digested pRTL2.

Binary vector based constructs pBI-GUS, pBI-GFP, pBI-dsGFP, pBI-HC-Pro,

pBI-AC2, pBI AL2, pBI- AL21-100 and pBI-L2 were generated by excising the expression cassette from each pRTL2-based construct with HindIII and BamHI and inserting it into

HindIII-BamHI digested pKJB5033, a plant transformation vector (provided by Dr.

Kenneth Buckley), derived from pBI121.

An Arabidopsis ADK fragment of approximately 500 bp (nt 306 to 814) from an

ADK2 cDNA was PCR amplified from an Arabidopsis cDNA library (Durfee et al.,

1993; Harper et al., 1993). The cDNA library was obtained from the ABRC (Ohio State

University, Columbus). The ADK-RNAi-5' primer contained SpeI (italic) and AscI

(underlined) restriction sites, 5'-

138

GAGACTAGTGGCGCGCCGGATGCTACAGCAGCTGG, and the ADK-RNAi-3'

primer contained BamHI (italic) and SwaI (underlined) restriction sites,

GAGGGATCCATTTAAATCCACTGGATCAGCGCCCTG-3’. The ADK PCR

fragments (SpeI-BamHI, AscI-SwaI) were sequentially cloned into pFGC1008

(http://ag.arizona.edu) binary vector to yield pFGC-dsADK which contains an inverted

repeat of the ADK sequence driven by a 35S promoter.

pBI-dsSNF1 (provided by Dr. Kenneth Buckley) contains an inverted repeat of

SNF1 N-terminal 1kb sequence driven by a 35S promoter. pBin61-p19 containing the strong silencing suppressor gene p19 from tombusvirus TBSV was provided by Dr.

Baulcombe (Voinnet et al., 2003).

Each of these plasmids was introduced into Agrobacterium tumefaciens strain

GV3101 by electroporation.

3.4.3 PTGS suppression assay and GFP imaging

The PTGS suppression assay was carried out as described (Voinnet, Pinto, and

Baulcombe, 1999). Visual detection of GFP fluorescence at 3, 5 and 7 days post

infiltration was performed using a 100 W, hand-held, long-wave ultraviolet (UV) lamp

(Blak-Ray model B 100YP; UV products, Upland, CA). Plants were photographed with a

Nikon Coolpix 990 digital camera (Nikon, Tokyo) mounted with both UV and yellow

filters. The images were processed electronically using Adobe Photoshop.

139

3.4.4 RNA isolation and Gel blot analysis

Total RNA was obtained using Trizol reagent (Gibco BRL) according to the

manufacturer's instructions. The RNA pellet was dissolved in cold formamide. Total

RNA concentration was determined by measuring OD260 in a UV spectrophotometer.

To detect mRNA, approximately 15 µg of total RNA was fractionated on a formaldehyde agarose gel, then transferred to a nitrocellulose membrane. To detect small

RNAs, approximately 45 µg of total RNA were separated by electrophoresis in a 15% polyacrylamide gel containing 7 M urea in 45 mM Tris-borate (pH 8.0), 1 mM EDTA.

The gel was blotted to Hybond-NX membrane (Amersham Pharmacia) and UV crosslinked (1200 µJ, Stratalinker; Stratagene).

32P-UTP radioactively labeled in vitro transcript riboprobes were synthesized

using Ambion’s Strip-EZ RNA in vitro transcription kit. Sense and antisense GFP probes were used for northern analyses of small RNA, while antisense probe was used for mRNA detection. Sense and antisense riboprobes of the complete BCTV genome or sense riboprobe of the TGMV coat protein gene were used for northern analysis of virus small RNA.

Membranes were prehybridized in 50% (vol/vol) formamide, 5× Denhardt's

solution, 0.5 mg/ml sheared salmon sperm DNA, 0.1% SDS, 5× SSC and 20 mM

phosphate buffer (pH 7.4) at 38°C for at least 2 h. Gels were stained with ethidium bromide before blotting to confirm equal loading of RNA samples. Membranes were hybridized with the 32P-UTP labeled riboprobes for approximately 16 hours. Membranes

were washed twice in a 2xSSC, 0.1%SDS solution at room temperature for 30 minutes.

140

For mRNA detection, membranes were further washed 1x with 0.1x SSC, 0.1%SDS at

65˚C for 15 to 30 minutes. For small RNA detection, the temperature was lowered to

50˚C.

3.4.5 Protein extraction and ADK kinase assay

Fresh tissue cut from the agroinfiltrated region of a N. benthamiana leaf was

ground in an ice cold motor and pestle. Total cellular protein was extracted and the ADK

kinase assay was performed as described in Chapter 2.

141

Figure 3.1 Geminiviruses TGMV and BCTV are inducers and targets of RNA silencing

N. benthamiana plants were infected by TGMV and BCTV with agroinoculation. At 14 and 30 days post inoculation (dpi), total RNAs were isolated from the plants. Total RNA

(45 µg) was fractionated on a 15% polyacrylamide gel containing 7M urea and hybridized with mixed sense and antisense BCTV genome riboprobes or TGMV sense coat protein riboprobe.

142

Figure 3.1 Geminiviruses TGMV and BCTV are inducers and targets of RNA silencing

143

Figure 3.2 Geminiviruses AL2/L2 proteins and dsADK suppress RNA silencing

N. benthamiana GFP transgenic line 16c leaf tissues were infiltrated with Agrobacterium strains delivering the expression plasmids 35S-GFP, 35S-dsGFP, and the indicated constructs. After 5 days GFP expression was examined by a UV lamp. Photos were processed by Adobe Photoshop program.

144

Figure 3.2 Geminiviruses AL2/L2 proteins and dsADK suppress RNA silencing

145

Figure 3.3 Geminiviruses AL2/L2 proteins, dsADK and ADK inhibitor suppress silencing

N. benthamiana leaf tissues were infiltrated with Agrobacterium strains delivering the expression plasmids 35S-GFP, 35S-dsGFP, and the indicated constructs, or the ADK inhibitor RBI. Total RNA was isolated from infiltration zones 3 and 5 days post- infiltration (dpi) and subjected to Northern blot analysis. (A) Analysis of GFP mRNA.

Agarose gel-fractionated total RNA (15 µg) was hybridized with a labeled GFP antisense riboprobe. (B) GFP siRNAs. Total RNA (45 µg) was fractionated on a 15% polyacrylamide gel containing 7M urea and hybridized with mixed sense and antisense

GFP riboprobes. AC2 is AL2 from ACMV. A representative experiment (4 trials) is shown.

146

Figure 3.3 Geminiviruses AL2/L2 proteins, dsADK and ADK inhibitor suppress silencing

147

Figure 3.4 ADK inhibitor RBI blocks ADK activity

Total protein extracts (about 400 ng) from N. benthamiana stem tissues were used in examination of the ADK inhibitor RBI function. RBI was dissolved in distilled water to a stock 100mM concentration and examined at the indicated concentrations.

148

Figure 3.4 ADK inhibitor RBI blocks ADK activity

149

Figure 3.5 ADK activity in infiltration zone N. benthamiana leaves were infiltrated with 35S:GFP + 35S:dsGFP + the indicated constructs, or RBI, and crude extracts prepared at 5 days post-infiltration. ADK assays were performed as described in Chapter 2. Samples (250 ng) were incubated with adenosine and labeled ATP, and reaction mixtures were fractioned by thin layer chromatography. Labeled AMP was quantitated by phosphorimaging. Signal intensities were compared relative to control extracts from tissue infiltrated with vector alone (V +

V + V). The range bar is the range of two experiment data.

150

Figure 3.5 ADK activity in infiltration zone

151

CHAPTER 4

DISCUSSION

4.1 Inactivation of ADK by AL2 and L2 and its effects on innate defense

4.1.1 Geminivirus AL2 and L2 proteins are pathogenicity determinants

Viruses are biotrophs. They replicate in living hosts by pirating host nutrients,

including nucleotides and amino acids, for their own reproduction. Viral DNA, RNA and

protein synthesis requires host biosynthetic machinery, as well as host-generated energy

in the form of ATP. During evolution of virus-host interactions, hosts have elaborated

defense systems in order to protect nutrients and biosynthetic systems. In response,

viruses have acquired “pathogenicity factors” to counter host defenses. Geminivirus AL2

and L2 proteins are pathogenicity factors. Transgenic plants expressing AL2 or L2 show

enhanced susceptibility (ES) to virus infection. This ES phenotype is characterized by a

decrease in mean latent period and reduced ID50 after challenge with geminiviruses or unrelated RNA virus such as TMV (Sunter, Sunter, and Bisaro, 2001). ES does not need 152

AL2 transcriptional activation function because it can be conditioned by an AL21-100 transgene lacking the C-terminal activation domain. L2 also conditions ES although it does not have transactivation function. Thus, AL2 has at least two separate functions during infection, and how these are regulated is unknown. That ES is not virus specific suggests that AL2 and L2 block an innate plant defense response that acts against wide range of pathogens. However, whether this particular response also is effective against other plant pathogens, including bacteria and fungi, is not yet known.

4.1.2 Geminivirus AL2/ L2-mediated ES acts by blocking SNF1 activity

To further study how AL2 and L2 act as pathogenicity determinants, we screened an Arabidopsis cDNA library in the yeast two-hybrid system with AL21-100 as bait. We found that both AL2 and L2 can interact with SNF1 kinase. Further study showed that

AL2 and L2 physically interact with SNF1 and inactivate its kinase activity in vitro and in vivo. In addition, co-expression of L2 can efficiently block complementation of a yeast

SNF1 deletion mutant by Arabidopsis SNF1. The SNF1 interaction domain was mapped to a conserved region shared by AL2 and L2. Mutant AL2 lacking a portion of this domain is a poor inhibitor of SNF1 kinase in vitro. Transgenic plants expressing this mutant AL2 do not display the ES phenotype. Further, transgenic plants expressing antisense SNF1 phenocopy ES shown by transgenic plants expressing AL2 or L2, while transgenic plants over-expressing sense SNF1 have enhanced resistance to virus infection

153

(Hao et al., 2003). Therefore, AL2 and L2 inactivate a SNF1-mediated metabolic response which otherwise would lead to enhanced disease resistance.

4.1.3 SNF1-mediated ES is different from other innate defense responses

SNF1-mediated resistance appears to be an innate defense response. However, it is different from another general innate defense mechanism, systemic acquired resistance

(SAR). SAR can apparently be induced in several ways. Usually, the induction of SAR requires the signal molecule salicylic acid (SA). SA is accumulated in plants before the onset of SAR. Transgenic plants expressing salicylate hydroxylase, an enzyme encoded by the bacterial NahG gene that degrades SA, cannot establish SAR (Gaffney et al.,

1993). Conversely, treating wild-type plants with SA or its functional analogs, such as

2,6-dichloroisonicotinic acid (INA) and benzol(1,2,3)thiadiazole-7-carbothioic acid S- methyl ester (BTH), induces SAR (Chivasa and Carr, 1998; Gorlach et al., 1996).

Pathogenesis-related gene (PR1) expression is a convenient marker for monitoring SAR

(Dong, 1998). SAR can also be induced by enhanced hexose concentrations. For example, increased cell wall invertase activity in transgenic plants expressing yeast cell wall invertase leads to an accumulation of soluble sugar, and a higher resistance towards viral attack. Although the plant response to an invertase transgene resembles SAR in several respects, it is differently induced because it does not depend on SA. Transgenic

NahG plants expressing invertase, which lack the SA pathway, showed the same

154

resistance to PVY infection as wild type plants expressing invertase (Herbers et al.,

2000). But both types of SAR induce PR gene expression. In contrast, PR1 gene

expression is not induced in transgenic plants that over-express SNF1 and show enhanced

resistance to virus infection (Wang and Bisaro, unpublished data). Thus, the SA and

sugar-induced SAR pathways are biochemically different from SNF1-mediated responses. However, it is likely that the disparate pathways communicate with each other to co-ordinate defense. It will be especially interesting to know how SA-mediated defense responses affect geminivirus replication and infection.

4.1.4 Virus infection tends to increase ADK activity

SNF1 kinase and its mammalian homologue, AMPK, are global regulators of

energy and carbon metabolism. Nutritional or environmental stresses (such as virus

infection) deplete ATP and the resulting ADP is rapidly converted to AMP. Thus the

AMP:ATP ratio is increased inside cells. The increased ratio of AMP:ATP activates

SNF1 kinase or AMPK by multiple mechanisms. Plant SNF1 and mammalian AMPK are

even directly bound and activated by 5′-AMP (Davies et al., 1995; Hawley et al., 1995;

Sugden et al., 1999a; Sugden et al., 1999b).

Virus infection is a stress to plants. It would not be a surprise that viral infection triggers plant defense by many approaches, one of which is to activate SNF1-mediated metabolic responses leading to disease resistance. Generally, under stress conditions,

ATP is depleted leading to generation of AMP. However, AMP can also be directly produced from adenosine by ADK activity. We found that virus infection tends to

155

increase ADK activity, which could rapidly generate AMP. We have not yet examined

the time course of this ADK activity increase, or whether it is a local or systemic

phenomenon. But this AMP might play a more important role in SNF1 activation than

AMP generated by ATP depletion, which might be a relatively slower process. To

establish a successful infection, the early stage usually is critical. We found that the ES phenotype in our transgenic plants expressing AL2, L2 or antisense SNF1 is

characterized by reduction of ID50 and mean latent period with no significant change in

symptom severity or virus load. This is consistent with the idea that SNF1 mediated responses affect an early stage of the virus-host interaction. After successful establishment of an infection, or after passing a certain threshold, viruses can cause normal symptoms, as in non-transgenic plants. This suggests that rapid signaling must occur following virus inoculation. It may be that increased ADK activity is part of this

signaling pathway. Once activated, SNF1 could effectively limit virus replication in

initially infected cells, reducing the probability of virus spread to neighboring cells to an extent sufficient to establish a systemic infection. Further study will be required to determine whether ADK-generated AMP is an activator of SNF1.

In yeast, activated SNF1 turns on downstream targets such as invertase, which is relatively easy to monitor by measuring the glucose concentration in cells. In addition,

SAMS peptides containing the SNF1 phosphorylation signature sequence can be used to measure SNF1 kinase activity directly. How can we use these tools in plants? To avoid nonspecific reaction from other kinases, transgenic plants overexpressing Myc-tagged

SNF1 can be treated with different stresses such as virus infection or high salt or drought.

Myc-SNF1 protein could be partially purified with anti-Myc antibody from these

156

transgenic plants and used to measure invertase activity or SAMS peptide phosphorylation.

4.1.5 Inactivation of ADK by AL2 and L2 could efficiently block SNF1-mediated plant innate defense

The battle between an invading virus and a host plant is often dependent on speed.

The outcome is determined by how fast the virus can replicate to large numbers, as compared to how quickly the host plant can respond to this infection by activating its defense systems. The elaboration of virus counter-defenses contributes to this process as well. Our result that AL2 and L2 interact with and inactivate both adenosine kinase and

SNF1 kinase in vitro and in vivo is consistent with this idea, because increased ADK activity following virus infection may generate an early signal to rapidly initiate SNF1- mediated defense, and geminiviruses might attempt to block this early signal by inhibiting ADK activity.

ADK is considered a housekeeping enzyme in eukaryotes. It generates 5′-AMP by transferring a phosphate from ATP or GTP to its substrate adenosine. Reported here are the first findings that ADK activity might respond to pathogen attack and is targeted by viral proteins. First, using an in vitro kinase assay we demonstrated that purified

Arabidopsis ADK activity is blocked by adding purified AL2 or L2. The inhibition is nearly stoichiometric with a 1 to 1 molar ratio. Secondly, it was demonstrated that expression of ADK complements the growth defect of purine deficient strain E. coli HO4 on minimal medium supplemented with adenosine. However, ADK activity and 157

complementation are blocked by co-expression of AL2 or L2. Thirdly, a yeast ADO1

strain with a deletion in the ADK gene can be complemented by Arabidopsis ADK.

However, the growth rate of the complemented strain was dramatically reduced by co-

expression of L2. Furthermore, transgenic plants expressing AL2 or L2 have significantly reduced ADK activity as compared with wild type plants. More importantly, plants infected by CMV, PVX or a BCTV mutant virus with an L2 frameshift mutation have significantly higher ADK activity than mock inoculated plants. For example, infection by

BCTV L2- mutant virus results in a 1.3 fold increase in ADK activity, whereas plants

infected with wild type TGMV or BCTV, with functional AL2 or L2 protein, have

significantly lower ADK activity (about 1.3 to 3 fold) than mock inoculated plants. ADK

generates 5′-AMP. SNF1 is a major metabolic regulator and it is activated by 5′-AMP.

Thus we propose that inactivation of ADK contributes to inactivation of SNF1.

4.1.6 Virus infection affects plant metabolism including sugar metabolism

It has been known for long time that virus infection affects plant metabolism. It

has been reported that plant virus infection can affect metabolism of plant

macromolecules, such as nucleic acids, proteins, and carbohydrates. For example, DNA

synthesis was decreased in the terminal 1 mm of French bean roots when they were

infected by Tobacco ringspot virus (TRSV) (Atchison, 1973). In TMV-infected leaves,

viral RNA may represent about 75% of total nucleic acids and cause a reduction in 16S

158

and 23S chloroplast ribosomal RNAs (Fraser, 1987). Virus coat protein may represent up to 50% of the total protein. TMV infection can also reduce host protein synthesis by up to

75% during the period of virus replication (Fraser, 1987). For example, the amount of the

most abundant host protein, ribulose bisphosphate carboxylase-oxygenase (rubisco), is

reduced, which may lead to mosaic and yellowing symptoms (White, 1983). These

observations suggest that during infection, the raw materials and energy necessary to

synthesize nucleic acids and proteins may become limiting.

Virus infection also affects sugar metabolism in plants. Virus infected leaf tissues in general have elevated levels of carbohydrate (Watson, 1951). During virus infection, the levels of many enzyme activities involved in sugar metabolism are altered. Activity of

ADP-glucose pyrophosphorylase (an enzyme of starch synthesis) is decreased in

Cucumber mosaic virus (CMV) infected Cucurbita pepo L. cotyledons, while activities of sucrose synthase and α-amylase (for starch degradation) are increased in infected cotyledons compared to uninfected ones (Tecsi, 1994; Tecsi et al., 1996). In addition, activities of enzymes involved in the oxidative pentose-phosphate pathway, glycolysis, the tricarboxylic acid cycle, and the oxidative electron transport pathway are also changed during CMV infection. In general, these changes in carbohydrate metabolism result in generation of more soluble sugar at the expense of structural carbohydrates and starch (Hull, 2002; Shalitin and Wolf, 2000). Viruses sequester raw materials for their own reproduction and thus can cause host cells to be deficient for energy and resources, mimicking a nutritional stress. This in turn may induce disease symptoms. This competition for nutrients could influence the outcome of infection.

159

4.1.7 SNF1 activates plant defense responses possibly by increasing soluble sugars

Soluble sugars can modulate gene expression in primary metabolism. For example,

soluble sugar is reported to repress photosynthetic genes in maize protoplasts (Jang and

Sheen, 1994). Sugar-dependent repression of α-amylase gene expression is found in rice

(Lu, 1996). α-amylases are endo-amylolytic enzymes, which catalyze the hydrolysis of α-

1,4-linked glucose polymers and play an important role in the degradation of starch and

glycogen in higher plants, animals, and many microorganisms.

In addition, soluble sugars have also been found to activate defense-related genes

(Herbers et al., 1995; Johnson and Ryan, 1990; Tsukaya et al., 1991). PR transcripts, such as PR-Q and PAR-1 (photoassimilate-responsive 1), were found to be inducible by glucose, fructose and sucrose but not by salicylic acid (SA). Interestingly, expression of defense-related genes, such as PR-Q, PAR-1, SAR8.2 and PR-1b, peroxidase activity, and SA levels were found to be increased in these plants. A state resembling systemic acquired resistance (SAR) is induced in transgenic sugar-accumulating plants expressing

yeast cell wall invertase (Herbers et al., 1996). The transgenic sugar-accumulating plants

also showed increased resistance to potato virus Y infection (Herbers et al., 1996;

Herbers et al., 1995). In addition, it has been shown that cell wall invertase levels can be

rapidly increased during virus infection (Scholes, 1994; Sturm and Chrispeels, 1990).

Thus increasing the activity of cell wall invertase in transgenic plants or virus-infected

plants led to an accumulation of soluble sugar, and a higher resistance towards viral

attack. This condition is independent of SA, because the introduction of a second

160

transgene (NahG) which abolishes the SA pathway did not abolish resistance to PVY

infection (Herbers et al., 2000). Therefore, these data suggest that soluble sugars play a

dual role in the plant as fuel for metabolism and as a signal for defense-related gene

expression.

SNF1 is a global metabolic regulator and activation of SNF1 might lead to an

increase in soluble sugars in plants. In response to stress, the AMP-activated SNF1

kinases or AMPKs shut down biosynthetic pathways by phosphorylating and inactivating

many enzymes, such as HMG-CoA reductase, nitrate reductase, and sucrose phosphate

synthase (Sugden et al., 1999b). They also turn on energy generating pathways. For

example, in yeast, SNF1 is required for activation of glucose-repressed genes and is

known to upregulate invertase activity to increase soluble sugar levels. When yeast is

grown in low glucose medium, SNF1 is activated and one outcome is the activation of

invertase gene expression, which allows the yeast to grow on sucrose. Because SNF1 has conserved functions between yeast and plants, SNF1 is believed to alter metabolism in a similar manner that also would lead to increases of plant soluble sugar levels, such as by upregulating invertase and α-amylase gene expression (Halford et al., 2003). This could lead to activation of plant defense pathways. Our observation that overexpression of a

SNF1 transgene results in resistance to virus infection but reducing SNF1 activity by expressing antisense SNF1 leads to enhanced susceptibility could be due to soluble sugar level changes in transgenic plants (Hao et al., 2003). However, this has not yet been investigated.

In order to counter plant defense responses, geminiviruses directly (physically) and indirectly (by inactivating ADK) block SNF1 activity to cause the ES phenotype, to

161

the benefit of the virus. This indicates that SNF1 must have an important role in the plant

defense response.

4.1.8 SNF1, ADK and AL2/L2 might form a complex

Recent studies indicate that AL2 is a cytoplasmic and a nuclear protein, depending on its phosphorylation status. AL2 must be localized to the nucleus to exert its transcriptional activator function. Its N-terminal region might contain a nuclear localization signal, as has been suggested for the analogous C2 protein of TYLCV (van

Wezel et al., 2001).

Because AL2 can interact with cytoplasmic ADK, it was important to prove that

AL2 is localized in the cytoplasm as well as the nucleus. We used several methods to localize AL2 protein in plant cells and found that the protein is found in both the nucleus and the cytoplasm. The methods included immunolabeling techniques to detect AL2 in tissues from TGMV-infected plants or transgenic plants expressing AL21-100. We also

used GFP fused AL2, such as 35S-AL2-GFP-GFP and 35S-GFP-GFP-AL2, to examine

localization in tobacco protoplasts transfected with these fusion constructs. Further

experiments need to be done to determine whether AL2 and ADK are co-localized. Our

nuclear fractionation results from insect cells expressing native AL2 protein suggested

that AL2 localization might be determined by its phosphorylation status (this report;

(Hartitz, Sunter, and Bisaro, 1999). At least two phosphorylated AL2 forms, determined

by migration on polyacrylamide gel and their absence after treatment with phosphatase,

are found almost exclusively in nucleus. In contrast, nonphosphorylated forms are present

162

in both the cytoplasm and the nucleus. Protein band intensities indicate that about a third

of the AL2 protein is cytoplasmic. Because ADK is a cytoplasmic protein, it might be

the unphosphorylated AL2 form that interacts with and inactivates ADK. It will also be

interesting to determine which kinase(s) phosphorylate AL2 and how the phosphorylation

event is regulated, as this may control the multiple biological functions of AL2.

Interestingly, ADK also interacts with SNF1 in yeast two-hybrid system. ADK has

two potential SNF1 phosphorylation sites near its N-terminus. Our in vitro data suggests

that ADK is phosphorylated by the SNF1 kinase domain and that ADK activity was

reduced this phosphorylation. In addition, ADK growth complementation of purine

deficient strain HO4 can be blocked by co-expressing SNF1 or SNF1-KD. Because AL2

and L2 can interact with both SNF1 and ADK, it is likely that three of them may form a complex in the cell to efficiently fulfill their functions (Figure 4.1). Furthermore, inactivation of ADK activity by SNF1 might indicate a feedback loop between ADK and

SNF1 (Figure 4.1). It is important to know how this loop is regulated and its biological significance. In addition, AMP generated from ADK could activate SNF1 more efficiently if they formed a complex in the cell.

163

4.2 Inactivation of ADK by AL2 and L2 and its effects on adaptive defense

4.2.1 Geminiviruses can induce RNA silencing and encode RNA silencing suppressors

RNA silencing can act as a specific defense response. RNA viruses are both

initiators and targets of RNA silencing, probably because during replication they form dsRNA intermediates, which strongly trigger the silencing system. Geminiviruses are

DNA viruses that have no dsRNA component. Nevertheless, it has been shown that when

used as an episomal vector to carry sequences homologous to an endogenous gene,

geminiviruses can sponsor silencing of the endogenous, chromosomal gene by a

mechanism similar to transgene-induced RNA silencing (Kjemtrup et al., 1998). Further a

transgene expressing L1 (Rep) of TYLCV driven by the 35S promoter can be silenced by

TYLCV infection (Lucioli et al., 2003).

Do geminiviruses produce a dsRNA signal? If so, how is it produced? In the geminivirus genome, there is divergent transcription from the IR, and read-through of

termination signals could produce dsRNA (Figure 1.1). In addition, the geminivirus DNA

genome, like plant genome, may produce “aberrant RNA” by an unknown mechanism.

“Aberrant RNA” is a term used to describe an intermediate RNA which causes gene

silencing. It could be excess mRNA, antisense RNA or truncated transcripts. Possibly,

"aberrant RNA” has unusual structural features which may be targeted by plant RNA-

dependent RNA polymerase (RdRP). Therefore “aberrant RNA” could serve as template

164

for RdRP to generate dsRNA. Thus, even though they do not encode an RdRP, it is very likely that geminiviruses are themselves initiators and targets of RNA silencing, and our work indicates that this is the case (Figure 3.1).

On the other hand, geminiviruses can also cause TGS of a GUS transgene driven by Tomato leaf curl virus (TLCV)-derived promoter sequences. In these experiments,

TGS was observed following inoculation of transgenic tobacco with TLCV. TGS was accompanied by cytosine hypermethylation in the TLCV-derived promoter sequence region. However, the relevance of this type of silencing to virus infection is unclear, as it is observed 50 day post inoculation (Seemanpillai et al., 2003). But, because geminiviruses exist as minichromosomes in plant cells, it is possible that geminiviruses are the triggers and targets of TGS as well. Thus RNA silencing and TGS might both be important for plants to defend against geminivirus infection.

Like many RNA viruses, geminiviruses encode RNA silencing suppressors, such as

AC2 from ACMV and C2 from TYLCV. Both suppressors can suppress PTGS (Dong et al., 2003; van Wezel et al., 2001; van Wezel et al., 2002; Van Wezel et al., 2003;

Voinnet, Pinto, and Baulcombe, 1999). Here we demonstrated that both TGMV AL2 and

BCTV L2 are RNA silencing suppressors as well. Although mutational studies suggest that the zinc finger motif and nuclear localization signal of TYLCV C2 protein likely are required for silencing suppression, the mechanism of suppression is still unclear (van

Wezel et al., 2002; Van Wezel et al., 2003).

It is clear that AL2/L2 proteins interfere with a unique step in the silencing pathway. AL2 is functionally different from HC-Pro because AL2 reduces siRNA generation while HC-Pro has no effect on siRNA production but may even increase

165

siRNA accumulation in a transient silencing system. AL2 is also different from p19

because p19 protein cannot reverse an established silencing state while AL2 protein can.

Our data suggest that AL2/L2 block the methylation pathway by reducing ADK activity.

HC-Pro and p19 have no effect on ADK activity. Therefore, different suppressors may

target different steps in the silencing pathway.

4.2.4 Inactivation of ADK by AL2 and L2 suppresses gene silencing

Gene silencing, including RNA silencing and TGS, is associated with DNA

methylation, and ADK activity is required for transmethylation. Interestingly, we found that AL2 and L2 can inactivate ADK activity, which might indicate that the gene silencing suppression by AL2 and L2 is due to this property. When we used dsADK or an

ADK inhibitor (RBI, an analog of adenosine) as a third component in the three component, transient silencing system, we found that silencing was suppressed (Figure

4.1). Thus it is clear that maintenance of silencing suppression does in fact require ADK activity (Figure 1.3). C-terminal transactivation domain of AL2 is not required for suppression of gene silencing, because AL2 mutant protein with deletion in this region

(AL21-100) and L2, which has no transactivation activity, also have silencing suppression

function.

While the observations that inhibition of ADK activity can suppress silencing

strongly suggest that this is the mechanism of silencing suppression by AL2/L2. But this

166

has not yet been proved. The use of AL2 mutants that cannot interact with ADK might

help to prove whether the inhibition of ADK activity by AL2 is responsible for silencing suppression.

It will also be important to determine whether DNA methylation is affected in

ADK deficient plants and whether AL2 and L2 can alter DNA or protein methylation on chromatin, plasmids or the virus minichromosome. Changes in methylation could be

specific or global. We expect that AL2/L2 and RBI will alter methylation in a cell

autonomous manner, whereas dsADK construct may have global affects.

4.3 Relationship between the AL2/L2-mediated ES phenotype and AL2/L2 mediated silencing suppression

Transgenic plants expressing AL2 and L2 have enhanced susceptibility to virus

infection. AL2 and L2 are also gene silencing suppressors. Is the ES phenotype due to

silencing suppression? Probably not. The ES phenotype is different from the RNA

silencing suppression. There are several data to support this. First, ES does not result in

increased disease symptoms or a higher level of virus accumulation, as is the case in

plants expressing gene-silencing suppressors, such as HC-Pro or P1 protein (Pruss et al.,

1997). These plants show severe disease symptoms and increased virus load due to the

suppression of RNA silencing. Second, the mutant viruses TGMV al2- or BCTV l2- do

not have significant replication defects in transfected protoplasts (Hormuzdi and Bisaro,

1995), while certain TEV mutants with defects in HC-Pro show dramatic replication

defects (Kasschau, Cronin, and Carrington, 1997). Other data to support that ES is 167

different from RNA silencing suppression comes from our recent finding. Antisense

SNF1 transgenes phenocopy ES seen in AL2 or L2 transgenic plants, while transgenic

plants over-expressing sense SNF1 have enhanced resistance to virus infection (Hao et al., 2003). In addition, the silencing pathway is not defective, since SNF1 silencing was

observed in one line containing a sense SNF1 transgene (Hao et al., 2003).

Why don’t AL2 and L2 transgenes show a phenotype more consistent with

silencing suppression? AL2 and L2 can interact and inactivate ADK activity. They

execute their silencing suppression probably by this function. However, ADK is a

housekeeping gene and is tightly regulated. Plants cannot tolerate over-expression or

under-expression of ADK. For example, it is difficult to obtain transgenic plants over-

expressing ADK or under-expressing ADK. We recovered only one transgenic plant

using RNAi technique to knock down ADK and progeny plants are severely stunted and

frequently revert to normal ADK levels. Further, although AL2 and L2 inhibit ADK

activity, transgenic plants expressing AL2 or L2 only show a 15 to 30% reduction of

ADK activity. Thus plants can only tolerate a small reduction in ADK activity, and this is

not sufficient to affect its maintenance of RNA silencing function. Therefore

constitutively expressed AL2 and L2 transgenes do not behave as silencing suppressors,

allowing the more subtle ES phenotype to be observed.

168

Figure 4.1 Working Model for AL2/L2 mediated pathogenesis

Geminivirus AL2/L2 proteins interact with and inhibit ADK and SNF1 kinases. ADK

also interacts with SNF1. Thus, AL2/L2, ADK, and SNF1 might form a complex. AMP

generated by depleting ATP during stress can activate SNF1. AMP generated from ADK

activity could be an early signal to activate SNF1. Virus challenge normally increases

ADK activity. A feedback loop might exist between ADK and SNF1 because SNF1 could phosphorylate ADK and inhibit ADK activity. SNF1 is a global metabolic regulator and mediates an innate defense response. ADK is also required to maintain

RNA silencing via its role in DNA and protein methylation. Therefore, geminivirus AL2 and L2 proteins function as pathogenesis determinants in two ways. One is to block

SNF1 directly by physical interaction or indirectly by blocking ADK activity to counter the SNF1-mediated innate defense response. Another is to block RNA silencing (adaptive defense response) by inhibiting ADK activity.

169

Figure 4.1 Working Model for AL2/L2 mediated pathogenesis

170

BIBLIOGRAPHY:

Accotto, G. P., Mullineaux, P. M., Brown, S. C., and Marie, D. (1993). Digitaria streak geminivirus replicative forms are abundant in S-phase nuclei of infected cells. Virology 195(1), 257-9.

Ach, R. A., Durfee, T., Miller, A. B., Taranto, P., Hanley-Bowdoin, L., Zambryski, P. C., and Gruissem, W. (1997). RRB1 and RRB2 encode maize retinoblastoma-related proteins that interact with a plant D-type cyclin and geminivirus replication protein. Mol Cell Biol 17(9), 5077-86.

Agrios, G. N. (1997). Plant Pathology (4th Ed.). Academic Press, San Diego.

Ahlquist, P. (2002). RNA-dependent RNA polymerases, viruses, and RNA silencing. Science 296, 1270-1273.

Alderson, A., Sabelli, P. A., Dickinson, J. R., Cole, D., Richardson, M., Kreis, M., Shewry, P. R., and Halford, N. (1991). Complementation of snf1, a mutation affecting global regulation of carbon metabolism in yeast, by a plant protein kinase cDNA. Proc. Natl. Acad. Sci. USA 88, 8602-8605.

Alexander, D., Goodman, R. M., Gut-Rella, M., Glascock, C., Weymann, K., Friedrich, L., Maddox, D., Ahl-Goy, P., Luntz, T., Ward, E., and et al. (1993). Increased tolerance to two oomycete pathogens in transgenic tobacco expressing pathogenesis-related protein 1a. Proc Natl Acad Sci U S A 90(15), 7327-31.

Ambros, V. (2003). MicroRNA pathways in flies and worms: growth, death, fat, stress, and timing. Cell 113(6), 673-6.

Anandalakshmi, R., Marathe, R., Ge, X., Herr, J. M., Jr., Mau, C., Mallory, A., Pruss, G., Bowman, L., and Vance, V. B. (2000). A calmodulin-related protein that suppresses posttranscriptional gene silencing in plants. Science 290(5489), 142-4.

Anandalakshmi, R., Pruss, G. J., Ge, X., Marathe, R., Mallory, A. C., Smith, T. H., and Vance, V. B. (1998). A viral suppressor of gene silencing in plants. Proc Natl Acad Sci U S A 95(22), 13079-84. 171

Aravin, A. A., Naumova, N. M., Tulin, A. V., Vagin, V. V., Rozovsky, Y. M., and Gvozdev, V. A. (2001). Double-stranded RNA-mediated silencing of genomic tandem repeats and transposable elements in the D. melanogaster germline. Curr. Biol. 11, 1017-1027.

Arguello-Astorga, G. R., Guevara-Gonzalez, R. G., Herrera-Estrella, L. R., and Rivera- Bustamante, R. F. (1994). Geminivirus replication origins have a group-specific organization of iterative elements: a model for replication. Virology 203(1), 90- 100.

Atchison, B. A. (1973). Division, expansion and DNA synthesis in meristematic cells of French bean (Phaseolus vulgaris L.) root-tips invaded by tobacco ringspot virus. Physiol. Plant Pathol 3, 1-8.

Baas, P. D. (1987). Mutational analysis of the bacteriophage phi X174 replication origin. J Mol Biol 198(1), 51-61.

Bachmann, M., Huber, J. L., Liao, P. C., Gage, D. A., and Huber, S. C. (1996). The inhibitor protein of phosphorylated nitrate reductase from spinach (Spinacia oleracea) leaves is a 14-3-3 protein. FEBS Lett 387(2-3), 127-31.

Bachmann, M., Shiraishi, N., Campbell, W. H., Yoo, B. C., Harmon, A. C., and Huber, S. C. (1996). Identification of Ser-543 as the major regulatory phosphorylation site in spinach leaf nitrate reductase. Plant Cell 8(3), 505-17.

Ball, K. L., Dale, S., Weekes, J., and Hardie, D. G. (1994). Biochemical characterization of two forms of 3-hydroxy-3-methylglutaryl-CoA reductase kinase from cauliflower (Brassica oleracia). Eur J Biochem 219(3), 743-50.

Bannister, A. J., Zegerman, P., Partridge, J. F., Miska, E. A., Thomas, J., Allshire, R. C., and Kouzarides, T. (2001). Selective recognition of methylated lysine 9 on histone H3 by the HP1 chromo domain. Nature 410, 120-124.

Bartee, L., Malagnac, F., and Bender, J. (2001). Arabidopsis cmt3 chromomethylase mutations block non-CG methylation and silencing of an endogenous gene. Genes Dev. 15, 1753-1758.

Bass, B. (2000). Double-stranded RNA as a template for gene silencing. Cell 101, 235- 238.

Baulcombe, D. (2002). RNA silencing. Curr Biol 12(3), R82-4.

Baulcombe, D. C. (1996). RNA as a target and an initiator of post-transcriptional gene silencing in transgenic plants. Plant Mol. Biol. 32, 79-88.

172

Baulcombe, D. C. (1999). Fast forward genetics based on virus-induced gene silencing. Curr. Opin. Plant Biol. 2, 109-113.

Bender, J. (2001). A vicious cycle: RNA silencing and DNA methylation in plants. Cell 106, 129-132.

Bent, A. F. (1996). Plant Disease Resistance Genes: Function Meets Structure. Plant Cell 8(10), 1757-1771.

Berne, R. M. (1993). Adenosine--a cardioprotective and therapeutic agent. Cardiovasc Res 27(1), 2.

Bernstein, E., Caudy, A., Hammond, S. M., and Hannon, G. J. (2001). Role for a bidentate ribonuclease in the initiation step of RNA interference. Nature 409, 363-366.

Bhalerao, R. P., Salchert, K., Okresz, L., Szabados, L., Muranaka, T., Machida, Y., Schell, J., and Konz, C. (1999). Regulatory interaction of PRL1 WD protein with Arabidopsis SNF1-like protein kinases. Proc. Natl. Acad. Sci. USA 96, 5322- 5327.

Bird, A. P. (1978). Use of restriction enzymes to study eukaryotic DNA methylation: II. The symmetry of methylated sites supports semi-conservative copying of the methylation pattern. J Mol Biol 118(1), 49-60.

Bisaro, D. M. (1996). Geminivirus replication. In "DNA Replication in Eukaryotic Cells" (M. DePamphilis, Ed.), pp. 833-854. Cold Spring Harbor Laboratory Press, Cold Spring Harbor.

Bisaro, D. M., Hamilton, W. D. O., Coutts, R. H. A., and Buck, K. W. (1982). Molecular cloning and characterisation of the two DNA components of tomato golden mosaic virus. Nucleic Acids Res. 10(16), 4913-4922.

Brigneti, G., Voinnet, O., Li, W.-X., Ji, L.-H., Ding, S.-W., and Baulcombe, D. C. (1998). Viral pathogenicity determinants are suppressors of transgene silencing in Nicotiana benthamiana. EMBO J. 17, 6739-6746.

Brough, C. L., Sunter, G., Gardiner, W. E., and Bisaro, D. M. (1992). Kinetics of tomato golden mosaic virus DNA replication and coat protein promoter activity in Nicotiana tabacum protoplasts. Virology 187(1), 1-9.

Burns, T. M., Harding, R. M., and Dale, J. L. (1995). The genome organization of banana bunchy top virus: analysis of six ssDNA components. J Gen Virol 76 ( Pt 6), 1471-82.

173

Cao, R., Wang, L., Wang, H., Xia, L., Erdjument-Bromage, H., Tempst, P., Jones, R. S., and Zhang, Y. (2002). Role of histone H3 lysine 27 methylation in Polycomb- group silencing. Science 298(5595), 1039-43.

Carling, D., Aguan, K., Woods, A., Verhoeven, A. J., Beri, R. K., Brennan, C. H., Sidebottom, C., Davison, M. D., and Scott, J. (1994). Mammalian AMP-activated protein kinase is homologous to yeast and plant protein kinases involved in the regulation of carbon metabolism. J Biol Chem 269(15), 11442-8.

Carling, D., Clarke, P. R., Zammit, V. A., and Hardie, D. G. (1989). Purification and characterization of the AMP-activated protein kinase. Copurification of acetyl- CoA carboxylase kinase and 3-hydroxy-3-methylglutaryl-CoA reductase kinase activities. Eur J Biochem 186(1-2), 129-36.

Carlson, M., and Botstein, D. (1982). Two differentially regulated mRNAs with different 5' ends encode secreted with intracellular forms of yeast invertase. Cell 28(1), 145-54.

Carrington, J. C., and Ambros, V. (2003). Role of microRNAs in plant and animal development. Science 301(5631), 336-8.

Carrington, J. C., Kasschau, K. D., and Johansen, L. K. (2001). Activation and suppression of RNA silencing by plant viruses. Virology 281, 1-5.

Castillo, A. G., Collinet, D., Deret, S., Kashoggi, A., and Bejarano, E. R. (2003). Dual interaction of plant PCNA with geminivirus replication accessory protein (Ren) and viral replication protein (Rep). Virology 312(2), 381-94.

Celenza, J. L., and Carlson, M. (1986). A yeast gene that is essential for release from glucose repression encodes a protein kinase. Science 233, 1175-1180.

Celenza, J. L., and Carlson, M. (1989). Mutational analysis of the Saccharomyces cerevisiae SNF1 protein kinase and evidence for functional interaction with the SNF4 protein. Mol. Cell. Biol. 9, 5034-5044.

Chang, C. H., Cha, S., Brockman, R. W., and Bennett, L. L., Jr. (1983). Kinetic studies of adenosine kinase from L1210 cells: a model enzyme with a two-site ping-pong mechanism. Biochemistry 22(3), 600-11.

Cheng, S., Schmidt-Grimminger, D. C., Murant, T., Broker, T. R., and Chow, L. T. (1995). Differentiation-dependent up-regulation of the human papillomavirus E7 gene reactivates cellular DNA replication in suprabasal differentiated keratinocytes. Genes Dev 9(19), 2335-49.

174

Chikano, H., Ogawa, M., Ikeda, Y., Koizumi, N., Kusano, T., and Sano, H. (2001). Two novel genes encoding SNF1-related protein kinases from Arabidopsis thaliana: differential accumulation of AtSR1 and AtSR2 transcripts in response to cytokinins and sugars, and phosphorylation of sucrose synthase by AtSR2. Mol. Gen. Genet. 264, 674-681.

Chiu, Y. L., and Rana, T. M. (2003). siRNA function in RNAi: a chemical modification analysis. Rna 9(9), 1034-48.

Chivasa, S., and Carr, J. P. (1998). Cyanide restores N gene-mediated resistance to tobacco mosaic virus in transgenic tobacco expressing salicylic acid hydroxylase. Plant Cell 10(9), 1489-98.

Choi, I. R., and Stenger, D. C. (1995). Strain-specific determinants of beet curly top geminivirus DNA replication. Virology 206(2), 904-12.

Chuang, C. F., and Meyerowitz, E. M. (2000). Specific and heritable genetic interference by double-stranded RNA in Arabidopsis thaliana. Proc Natl Acad Sci U S A 97(9), 4985-90.

Collin, S., Fernandez-Lobato, M., Gooding, P. S., Mullineaux, P. M., and Fenoll, C. (1996). The two nonstructural proteins from wheat dwarf virus involved in viral gene expression and replication are retinoblastoma-binding proteins. Virology 219, 324-329.

Corton, J. M., Gillespie, J. G., and Hardie, D. G. (1994). Role of AMP-activated protein kinase in the cellular stress response. Curr. Biol. 4, 315-324.

Covey, S. N. A.-K., N. S.; Langara A.; Turner D. S. (1997). Plants combat infection by gene silencing. Nature 385, 781.

Dale, S., Wilson, W. A., Edelman, A. M., and Hardie, D. G. (1995). Similar substrate recognition motifs for mammalian AMP-activated protein kinase, higher plant HMG-CoA reductase kinase-A, yeast SNF1, and mammalian calmodulin- dependent protein kinase I. FEBS Lett 361(2-3), 191-5.

Dalmay, T., Hamilton, A., Mueller, E., and Baulcombe, D. C. (2000a). Potato virus X amplicons in arabidopsis mediate genetic and epigenetic gene silencing. Plant Cell 12(3), 369-79.

Dalmay, T., Hamilton, A., Rudd, S., Angell, S., and Baulcombe, D. C. (2000b). An RNA- dependent RNA polymerase gene in Arabidopsis is required for posttranscriptional gene silencing mediated by a transgene but not by a virus. Cell 101(5), 543-53.

175

Dangl, J. L., and Jones, J. D. (2001). Plant pathogens and integrated defence responses to infection. Nature 411(6839), 826-33.

Davies, S. P., Carling, D., Munday, M. R., and Hardie, D. G. (1992). Diurnal rhythm of phosphorylation of rat liver acetyl-CoA carboxylase by the AMP-activated protein kinase, demonstrated using freeze-clamping. Effects of high fat diets. Eur J Biochem 203(3), 615-23.

Davies, S. P., Helps, N. R., Cohen, P. T. W., and Hardie, D. G. (1995). 5'-AMP inhibits dephosphorylation, as well as promoting phosphorylation, of the AMP-activated protein kinase. Studies using bacterially expressed human protein phosphatase-

2Cα and native bovine protein phosphatase-2Ac. FEBS Lett. 377, 421-425.

Dickinson, J. R., Cole D, Halford N. G. (1999). A cell cycle role for a plant sucrose nonfermenting-1-related protein kinase (SnRK1) is indicated by expression in yeast. Plant Growth Regulation 28, 169-174.

Dinesh-Kumar, S. P., Anandalakshmi, R., Marathe, R., Schiff, M., and Liu, Y. (2003). Virus-induced gene silencing. Methods Mol Biol 236, 287-94.

Doench, J. G., Petersen, C. P., and Sharp, P. A. (2003). siRNAs can function as miRNAs. Genes Dev 17(4), 438-42.

Doermann, A. H. (1973). T4 and the rolling circle model of replication. Annu Rev Genet 7, 325-41.

Dong, X. (1998). SA, JA, ethylene, and disease resistance in plants. Curr Opin Plant Biol 1(4), 316-23.

Dong, X., van Wezel, R., Stanley, J., and Hong, Y. (2003). Functional characterization of the nuclear localization signal for a suppressor of posttranscriptional gene silencing. J Virol 77(12), 7026-33.

Donson, J., Morris-Krsinich, B. A. M., Mullineaux, P. M., Boulton, M., and Davies, J. W. (1984). A putative primer for second-strand synthesis of maize streak virus is virion-associated. EMBO J. 3, 3069-3073.

Douglas, P., Morrice, N., and MacKintosh, C. (1995). Identification of a regulatory phosphorylation site in the hinge 1 region of nitrate reductase from spinach (Spinacea oleracea) leaves. FEBS Lett 377(2), 113-7.

Duerre, J. A., and Briske-Anderson, M. (1981). Effect of adenosine metabolites on methyltransferase reactions in isolated rat livers. Biochim Biophys Acta 678(2), 275-82.

176

Durfee, T., Becherer, K., Chen, P.-L., Yeh, S.-H., Yang, Y., Kilburn, A. E., Lee, W.-H., and Elledge, S. J. (1993). The retinoblastoma protein associates with the protein phosphatase type 1 catalytic subunit. Genes Dev. 7, 555-569.

Dyck, J. R., Kudo, N., Barr, A. J., Davies, S. P., Hardie, D. G., and Lopaschuk, G. D. (1999). Phosphorylation control of cardiac acetyl-CoA carboxylase by cAMP- dependent protein kinase and 5'-AMP activated protein kinase. Eur J Biochem 262(1), 184-90.

Eagle, P. A., Orozco, B. M., and Hanley-Bowdoin, L. (1994). A DNA sequence required for geminivirus replication also mediates transcriptional regulation. Plant Cell 6(8), 1157-70.

Egelkrout, E. M., Robertson, D., and Hanley-Bowdoin, L. (2001). Proliferating cell nuclear antigen transcription is repressed through an E2F consensus element and activated by geminivirus infection in mature leaves. Plant Cell 13(6), 1437-52.

Elbashir, S. M., Lendeckel, W., and Tuschl, T. (2001). RNA interference is mediated by 21- and 22-nucleotide RNAs. Genes Dev 15(2), 188-200.

Elbashir, S. M., Martinez, J., Patkaniowska, A., Lendeckel, W., and Tuschl, T. (2001). Functional anatomy of siRNAs for mediating efficient RNAi in Drosophila melanogaster embryo lysate. Embo J 20(23), 6877-88.

Ellis, C., Karafyllidis, I., and Turner, J. G. (2002). Constitutive activation of jasmonate signaling in an Arabidopsis mutant correlates with enhanced resistance to Erysiphe cichoracearum, Pseudomonas syringae, and Myzus persicae. Mol Plant Microbe Interact 15(10), 1025-30.

Elmer, J. S., Sunter, G., Gardiner, W. E., Brand, L., Browning, C. K., Bisaro, D. M., and Rogers, S. G. (1988). Agrobacterium-mediated inoculation of plants with tomato golden mosaic virus DNAs. Plant Mol. Biol. 10, 225-234.

English, J. J., Mueller, E., and Baulcombe, D. C. (1996). Suppression of virus accumulation in plants exhibiting silencing of nuclear genes. Plant Cell 8, 179- 188.

Estruch, F., Treitel, M. A., Yang, X., and Carlson, M. (1992). N-terminal mutations modulate yeast SNF1 protein kinase function. Genetics 132(3), 639-50.

Fauquet, C. M., Bisaro, D. M., Briddon, R. W., Brown, J. K., Harrison, B. D., Rybicki, E. P., Stenger, D. C., and Stanley, J. (2003). Revision of taxonomic criteria for species demarcation in the family Geminiviridae, and an updated list of begomovirus species. Arch Virol 148(2), 405-21.

177

Finnegan, E. J., and Kovac, K. A. (2000). Plant DNA methyltransferases. Plant Mol Biol 43(2-3), 189-201.

Flor, H. H. (1971). Current Status of the Gene-For-Gene Concept. Annual Review of Phytopathology 9, 275-296.

Fluhr, R., and Kaplan-Levy, R. N. (2002). Plant disease resistance: commonality and novelty in multicellular innate immunity. Curr Top Microbiol Immunol 270, 23- 46.

Fox, H. I., and Kelley, N. W. (1978). The role of adenosine and 2'-deoxyadenosine in mammalian cells. Ann. Rev. Biochem. 47, 655-686.

Fraser, R. S. S. (1987). Biochemistry of Virus-Infected Plants. Research Studies Press(Letchworth, UK.).

Gaffney, T., Friedrich, L., Vernooij, B., Negrotto, D., Nye, G., Uknes, S., Ward, E., Kessmann, H., and Ryals, J. (1993). Requirement of salicylic acid for the induction of systemic acquired resistance. Science 261, 754-756.

Gancedo, J. M. (1992). Carbon catabolite repression in yeast. Eur J Biochem 206(2), 297- 313.

Gendrel, A. V., Lippman, Z., Yordan, C., Colot, V., and Martienssen, R. A. (2002). Dependence of heterochromatic histone H3 methylation patterns on the Arabidopsis gene DDM1. Science 297(5588), 1871-3.

Gerster, T., and Roeder, R. G. (1988). A herpes trans-activating protein interacts with transcription factor OTF-1 and other cellular proteins. Proc. Natl. Acad. Sci. USA 85, 6347-6351.

Gillespie, J. G., and Hardie, D. G. (1992). Phosphorylation and inactivation of HMG- CoA reductase at the AMP-activated protein kinase site in response to fructose treatment of isolated rat hepatocytes. FEBS Lett 306(1), 59-62.

Gladfelter, H. J., Eagle, P. A., Fontes, E. P. B., Batts, L., and Hanley-Bowdoin, L. (1997). Two domains of the AL1 protein mediate geminivirus origin recognition. Virology 239, 186-197.

Gorlach, J., Volrath, S., Knauf-Beiter, G., Hengy, G., Beckhove, U., Kogel, K.-H., Oostendorp, M., Staub, T., Ward, E., Kessmann, H., and Ryals, J. (1996). Benzothiadiazole, a novel class of inducers of systemic acquired resistance, activates gene expression and disease resistance in wheat. Plant Cell 8, 629-643.

178

Grafi, G., Burnett, R. J., Helentjaris, T., Larkins, B. A., Decaprio, J. A., Sellers, W. R., and Kaelin, W. G. (1996). A maize cDNA encoding a member of the retinoblastoma protein family: Involvement in endoreduplication. Proc. Natl. Acad. Sci. USA 93, 8962-8967.

Grimsley, N., Hohn, T., Davies, J. W., and Hohn, B. (1987). Agrobacterium-mediated delivery of infectious maize streak virus into maize plants. Nature 325, 177-179.

Grishok, A., and Mello, C. C. (2002). RNAi (Nematodes: Caenorhabditis elegans). Adv Genet 46, 339-60.

Gruenbaum, Y., Naveh-Many, T., Cedar, H., and Razin, A. (1981). Sequence specificity of methylation in higher plant DNA. Nature 292(5826), 860-2.

Guo, H. S., and Ding, S. W. (2002). A viral protein inhibits the long range signaling activity of the gene silencing signal. Embo J 21(3), 398-407.

Gutierrez, C. (1999). Geminivirus DNA replication. CMLS, Cell. Mol. Life Sci. 56, 313- 329.

Gutierrez, C. (2000). DNA replication and cell cycle in plants: learning from geminiviruses. Embo J 19(5), 792-9.

Hahn, J. S., and Thiele, D. J. (2003). Activation of the S. cerevisiae heat shock transcription factor under glucose starvation conditions by Snf1 protein kinase. J Biol Chem.

Halford, N., and Hardie, D. G. (1998). SNF1-related protein kinases: global regulators of carbon metabolism in plants? Plant Mol. Biol. 37, 735-748.

Halford, N. G., Hey, S., Jhurreea, D., Laurie, S., McKibbin, R. S., Paul, M., and Zhang, Y. (2003). Metabolic signalling and carbon partitioning: role of Snf1-related (SnRK1) protein kinase. J Exp Bot 54(382), 467-75.

Hall, I. M., Shankaranarayana, G. D., Noma, K., Ayoub, N., Cohen, A., and Grewal, S. I. (2002). Establishment and maintenance of a heterochromatin domain. Science 297(5590), 2232-7.

Hamilton, A., and Baulcombe, D. C. (1999). A species of small antisense RNA in posttranscriptional gene silencing in plants. Science 286, 950-952.

Hamilton, A., Voinnet, O., Chappell, L., and Baulcombe, D. (2002). Two classes of short interfering RNA in RNA silencing. Embo J 21(17), 4671-9.

179

Hamilton, W. D. O., Bisaro, D. M., Coutts, R. H. A., and Buck, K. W. (1983). Demonstration of the bipartite nature of the genome of a single-stranded DNA plant virus by infection with the cloned DNA components. Nucleic Acids Res. 11(21), 7387-7396.

Hammond, S. M., Bernstein, E., Beach, D., and Hannon, G. J. (2000). An RNA-directed nuclease mediates post-transcriptional gene silencing in Drosophila cells. Nature 404(6775), 293-6.

Hanley-Bowdoin, L., Settlage, S., Orozco, B. M., Nagar, S., and Robertson, D. (1999). Geminiviruses: Models for plant DNA replication, transcription, and cell cycle regulation. Crit. Rev. Plant Sci. 18, 71-106.

Hanley-Bowdoin, L., Settlage, S. B., Orozco, B. M., Nagar, S., and Robertson, D. (2000). Geminiviruses: models for plant DNA replication, transcription, and cell cycle regulation. Crit Rev Biochem Mol Biol 35(2), 105-40.

Hannon, G. J. (2002). RNA interference. Nature 418(6894), 244-51.

Hao, L., Wang, H., Sunter, G., and Bisaro, D. M. (2003). Geminivirus AL2 and L2 proteins interact with and inactivate SNF1 kinase. Plant Cell 15(4), 1034-48.

Hardie, D. G. (1999). Roles of the AMP-activated/SNF1 protein kinase family in the response to cellular stress. Biochem Soc Symp 64, 13-27.

Hardie, D. G., and Carling, D. (1997). The AMP-activated protein kinase--fuel gauge of the mammalian cell? Eur J Biochem 246(2), 259-73.

Hardie, D. G., Carling, D., and Carlson, M. (1998). The AMP-activated/SNF1 protein kinase subfamily: metabolic sensors of the eukaryotic cell? Annu Rev Biochem 67, 821-855.

Hardie, D. G., Carling, D., and Carlson, M. (1998). The AMP-activated/SNF1 protein kinase subfamily: Metabolic sensors of the eukaryotic cell? Ann. Rev. Biochem. 67, 821-855.

Hardie, D. G., and Hawley, S. A. (2001). AMP-activated protein kinase: the energy charge hypothesis revisited. Bioessays 23(12), 1112-9.

Harper, J. W., Adami, G. R., Wei, N., Keyomarsi, K., and Elledge, S. J. (1993). The p21 Cdk-interacting protein Cip1 is a potent inhibitor of G1 cyclin-dependent kinase. Cell 75, 805-816.

180

Hartitz, M. D., Sunter, G., and Bisaro, D. M. (1999). The geminivirus transactivator (TrAP) is a single-stranded DNA and zinc-binding phosphoprotein with an acidic activation domain. Virology 263, 1-14.

Hawley, S. A., Davison, M., Woods, A., Davies, S. P., Beri, R. K., Carling, D., and Hardie, D. G. (1996). Characterization of the AMP-activated protein kinase kinase from rat liver and identification of threonine 172 as the major site at which it phosphorylates AMP-activated protein kinase. J Biol Chem 271(44), 27879-87.

Hawley, S. A., Selbert, M. A., Goldstein, E. G., Edelman, A. M., Carling, D., and Hardie, D. G. (1995). 5'-AMP activates the AMP-activated protein kinase cascade, and Ca2+/calmodulin-dependent protein kinase I cascade, via three independent mechanisms. J. Biol. Chem. 270, 27186-27191.

Hayes, R. J., Brough, C. L., Prince, V. E., Coutts, R. H. A., and Buck, K. W. (1988). Infection of Nicotiana benthamiana with uncut cloned tandem dimers of tomato golden mosaic virus DNA. J. Gen. Virol. 69, 209-218.

Hayes, R. J., Macdonald, H., Coutts, R. H. A., and Buck, K. W. (1988). Priming of complementary DNA synthesis in vitro by small DNA molecules tightly bound to virion DNA of wheat dwarf virus. J. Gen. Virol. 69, 1345-1350.

Herbers, K., Meuwly, P., Frommer, W. B., Metraux, J.-P., and Sonnewald, U. (1996). Systemic acquired resistance mediated by ectopic expression of invertase: Possible hexose sensing in the secretory pathway. Plant Cell 8, 793-803.

Herbers, K., Monke, G., Badur, R., and Sonnewald, U. (1995). A simplified procedure for the subtractive cDNA cloning of photoassimilate-responding genes: isolation of cDNAs encoding a new class of pathogenesis-related proteins. Plant Mol Biol 29(5), 1027-38.

Herbers, K., Takahata, Y., Melzer, M., Mock, H.-P., Hajirezaei, M., and Sonnewald, U. (2000). Regulation of carbohydrate partitioning during the interaction of potato virus Y with tobacco. Molecular Plant Pathology 1(1), 51-59.

Herwig, S., and Strauss, M. (1997). The retinoblastoma protein: a master regulator of cell cycle, differentiation and apoptosis. Eur J Biochem 246(3), 581-601.

Hiesinger, M., Roth, S., Meissner, E., and Schuller, H. J. (2001). Contribution of Cat8 and Sip4 to the transcriptional activation of yeast gluconeogenic genes by carbon source-responsive elements. Curr Genet 39(2), 68-76.

Hong, S. P., Leiper, F. C., Woods, A., Carling, D., and Carlson, M. (2003). Activation of yeast Snf1 and mammalian AMP-activated protein kinase by upstream kinases. Proc Natl Acad Sci U S A 100(15), 8839-43.

181

Hormuzdi, S. G., and Bisaro, D. M. (1993). Genetic analysis of beet curly top virus: evidence for three virion sense genes involved in movement and regulation of single- and double-stranded DNA levels. Virology 193(2), 900-9.

Hormuzdi, S. G., and Bisaro, D. M. (1995). Genetic analysis of beet curly top virus: examination of the roles of L2 and L3 genes in viral pathogenesis. Virology 206(2), 1044-54.

Hove-Jensen, B., and Nygaard, P. (1989). Role of guanosine kinase in the utilization of guanosine for nucleotide synthesis in Escherichia coli. J. Gen. Microbiol. 135, 1263-1273.

Hull, R. (2002). Matthew's Plant Virology. Academic Press 4th Edition.

Hutvagner, G., McLachlan, J., Pasquinelli, A. E., Balint, E., Tuschl, T., and Zamore, P. D. (2001). A cellular function for the RNA-interference enzyme Dicer in the maturation of the let-7 small temporal RNA. Science 293, 834-838.

Hutvagner, G., and Zamore, P. D. (2002). A microRNA in a multiple-turnover RNAi enzyme complex. Science 297(5589), 2056-60.

Im, Y. S., Chiang, P. K., and Cantoni, G. L. (1979). Guanidoacetate methyltransferase. Purification and molecular properties. J Biol Chem 254(21), 11047-50.

Ingelbrecht, I., Van Houdt, H., Van Montagu, M., and Depicker, A. (1994). Posttranscriptional silencing of reporter transgenes in tobacco correlates with DNA methylation. Proc. Natl. Acad. Sci. USA 91, 10502-10506.

Iwashima, A., Ogata, M., Nosaka, K., Nishimura, H., and Hasegawa, T. (1995). Adenosine kinase-deficient mutant of Saccharomyces cerevisiae. FEMS Microbiol Lett 127(1-2), 23-8.

Jackson, J. P., Lindroth, A. M., Cao, X., and Jacobsen, S. E. (2002). Control of CpNpG DNA methylation by the KRYPTONITE histone H3 methyltransferase. Nature 416(6880), 556-60.

Jacobs, S. A., Taverna, S. D., Zhang, Y., Briggs, S. D., Li, J., Eissenberg, J. C., Allis, C. D., and Khorasanizadeh, S. (2001). Specificity of the HP1 chromo domain for the methylated N-terminus of histone 3. EMBO J. 20, 5232-5241.

Jang, J.-C., and Sheen, J. (1994). Sugar sensing in higher plants. Plant Cell 6, 1665-1679.

Jeddeloh, J. A., Bender, J., and Richards, E. J. (1998). The DNA methylation locus DDM1 is required for maintenance of gene silencing in Arabidopsis. Genes Dev 12(11), 1714-25.

182

Jenuwein, T., and Allis, C. D. (2001). Translating the histone code. Science 293, 1074- 1080.

Jiang, N., Kowaluk, E. A., Lee, C. H., Mazdiyasni, H., and Chopp, M. (1997). Adenosine kinase inhibition protects brain against transient focal ischemia in rats. Eur J Pharmacol 320(2-3), 131-7.

Jiang, R., and Carlson, M. (1997). The Snf1 protein kinase and its activating subunit, Snf4, interact with distinct domains of the Sip1/Sip2/Gal83 component in the kinase complex. Mol Cell Biol 17(4), 2099-106.

Johansen, L. K., and Carrington, J. C. (2001). Silencing on the spot: Induction and suppression of RNA silencing in the Agrobacterium-mediated transient expression system. Plant Physiol. 126, 930-938.

Johnson, L., Cao, X., and Jacobsen, S. E. (2002). Interplay between two epigenetic marks. DNA methylation and histone H3 lysine 9 methylation. Curr. Biol. 12, 1360-.

Johnson, R., and Ryan, C. A. (1990). Wound-inducible potato inhibitor II genes: Enhancement of expression by sucrose. Plant Mol. Biol. 14, 527-536.

Johnston, M. (1999). Feasting, fasting, and fermenting: Glucose sensing in yeast and other cells. Trends in Genetics 15, 29-33.

Jones, A. L., Thomas, C. L., and Maule, A. J. (1998). De novo methylation and co- suppression induced by a cytoplasmically replicating plant RNA virus. Embo J 17(21), 6385-93.

Jones, L., Hamilton, A. J., Voinnet, O., Thomas, C. L., Maule, A. J., and Baulcombe, D. C. (1999). RNA-DNA interactions and DNA methylation in post-transcriptional gene silencing. Plant Cell 11(12), 2291-301.

Jones, L., Ratcliff, F., and Baulcombe, D. C. (2001). RNA-directed transcriptional gene silencing in plants can be inherited independently of the RNA trigger and requires Met1 for maintenance. Curr. Biol. 11, 747-757.

Jupin, I., Hericourt, F., Benz, B., and Gronenborn, B. (1995). DNA replication specificity of TYLCV geminivirus is mediated by the amino-terminal 116 amino acids of rep protein. FEBS Lett. 362, 116-120.

Kakimoto, T. (2003). Perception and signal transduction of cytokinins. Annu Rev Plant Biol 54, 605-27.

183

Kammann, M., Schalk, H.-J., Matzeit, V., Schaefer, D., Schell, J., and Gronenborn, B. (1991). DNA replication of wheat dwarf virus, a geminivirus, requires two cis- acting signals. Virology 1991, 786-790.

Kasschau, K. D., and Carrington, J. C. (1998). A counterdefensive strategy of plant viruses: Suppression of posttranscriptional gene silencing. Cell 95, 461-470.

Kasschau, K. D., Cronin, S., and Carrington, J. C. (1997). Genome amplification and long-distance movement functions associated with the central domain of tobacco etch potyvirus helper component-protease. Virology 228, 251-262.

Kasschau, K. D., Xie, Z., Allen, E., Llave, C., Chapman, E. J., Krizan, K. A., and Carrington, J. C. (2003). P1/HC-Pro, a viral suppressor of RNA silencing, interferes with Arabidopsis development and miRNA unction. Dev Cell 4(2), 205- 17.

Ketting, R., Fischer, H. E. J., Bernstein, E., Sijen, T., Hannon, G. J., and Plasterk, R. H. A. (2001). Dicer functions in RNA interference and in synthesis of small RNA involved in developmental timing in C. elegans. Genes Dev. 15, 2654-2659.

Ketting, R., Haverkamp, T., van Luenen, H., and Plasterk, R. (1999). mut-7 of C. elegans, required for transposon silencing and RNA interference, is a homolog of Werner syndrome helicase and RNaseD. Cell 99, 133-141.

Khvorova, A., Reynolds, A., and Jayasena, S. D. (2003). Functional siRNAs and miRNAs exhibit strand bias. Cell 115(2), 209-16.

Kjemtrup, S., Sampson, K. S., Peele, C. G., Nguyen, L. V., Conkling, M., Thompson, W. F., and Robertson, D. (1998). Gene silencing from plant DNA carried by a geminivirus. Plant J. 14, 91-100.

Klein, C. J., Olsson, L., and Nielsen, J. (1998). Glucose control in Saccharomyces cerevisiae: the role of Mig1 in metabolic functions. Microbiology 144(Pt1), 13- 24.

Knight, S. W., and Bass, B. L. (2001). A role for the RNase III enzyme DCR-1 in RNA interference and germ line development in Caenorhabditis elegans. Science 293(5538), 2269-71.

Koukalova, B., Reich, J., Matyasek, R., Kuhrova, V., and Bezdek, M. (1989). A BamHI family of highly repeated DNAsequences of Nicotiana tabacum. Theor Appl Genet 78, 77–80.

184

Kovarik, A., Van Houdt, H., Holy, A., and Depicker, A. (2000). Drug-induced hypomethylation of a posttranscriptionally silenced transgene locus of tobacco leads to partial release of silencing. FEBS Lett. 467, 47-51.

Kowaluk, E. A., Bhagwat, S. S., and Jarvis, M. F. (1998). Adenosine kinase inhibitors. Curr Pharm Des 4(5), 403-16.

Kredich, N. M., and Martin, D. V., Jr. (1977). Role of S-adenosylhomocysteine in adenosinemediated toxicity in cultured mouse T lymphoma cells. Cell 12(4), 931- 8.

Kristie, T. M., LeBowitz, J. H., and Sharp, P. A. (1989). The octamer-binding proteins form multi-protein-DNA complexes with the HSV αTIF regulatory protein. EMBO J. 8, 4229-4238.

Kubota, K., Tsuda, S., Tamai, A., and Meshi, T. (2003). Tomato Mosaic Virus Replication Protein Suppresses Virus-Targeted Posttranscriptional Gene Silencing. J Virol 77(20), 11016-11026.

Kuchin, S., Treich, I., and Carlson, M. (2000). A regulatory shortcut between the Snf1 protein kinase and RNA polymerase II holoenzyme. Proc. Natl. Acad. Sci. USA 97, 7916-7920.

Lachner, M., O'Carroll, D., Rea, S., Mechtler, K., and Jenuwein, T. (2001). Methylation of histone H3 lysine 9 creates a binding site for HP1 proteins. Nature 410(6824), 116-20.

Lam, E., Kato, N., and Lawton, M. (2001). Programmed cell death, mitochondria and the plant hypersensitive response. Nature 411(6839), 848-53.

Latham, J. R., Saunders, K., Pinner, M. S., and Stanley, J. (1997). Induction of plant cell division by beet curly top virus gene C4. Plant Journal 11, 1273-1283.

Lazarowitz, S. G. (1987). The molecular characterization of geminiviruses. Plant Mol. Biol. Rep. 4, 177-192.

Lecoq, K., Belloc, I., Desgranges, C., and Daignan-Fornier, B. (2001). Role of adenosine kinase in Saccharomyces cerevisiae: identification of the ADO1 gene and study of the mutant phenotypes. Yeast 18, 335-342.

Lesage, P., Yang, X., and Carlson, M. (1996). Yeast SNF1 protein kinase interacts with SIP4, a C6 zinc cluster transcriptional activator: a new role for SNF1 in the glucose response. Mol Cell Biol 16(5), 1921-8.

185

Li, H., Li, W. X., and Ding, S. W. (2002). Induction and suppression of RNA silencing by an animal virus. Science 296(5571), 1319-21.

Li, Q., Lawrence, C. B., Xing, H. Y., Babbitt, R. A., Bass, W. T., Maiti, I. B., and Everett, N. P. (2001). Enhanced disease resistance conferred by expression of an antimicrobial magainin analog in transgenic tobacco. Planta 212(4), 635-9.

Li, W. X., and Ding, S. W. (2001). Viral suppressors of RNA silencing. Curr. Opin. Biotech. 12, 150-154.

Lillie, J. W., and Green, M. R. (1989). Transcription activation by the adenovirus E1A protein. Nature 338, 39-44.

Lindbo, J. A., and Dougherty, W. G. (1992a). Pathogen-derived resistance to a potyvirus: Immune and resistance phenotypes in transgenic tobacco expressing altered forms of a potyvirus coat protein nucleotide sequence. Mol. Plant-Microbe Interact. 5, 144-153.

Lindbo, J. A., and Dougherty, W. G. (1992b). Untranslatable transcripts of the tobacco etch virus coat protein gene sequence can interfere with tobacco etch virus replication in transgenic plants and protoplasts. Virology 189, 725-733.

Lindbo, J. A., Silva-Rosales, L., Proebsting, W. M., and Dougherty, W. G. (1993). Induction of a Highly Specific Antiviral State in Transgenic Plants: Implications for Regulation of Gene Expression and Virus Resistance. Plant Cell 5(12), 1749- 1759.

Lindroth, A. M., Cao, X., Jackson, J. P., Zilberman, D., McCallum, C. M., Henikoff, S., and Jacobsen, S. E. (2001). Requirement of CHROMOMETHYLASE3 for maintenance of CpXpG methylation. Science 292, 2077-2080.

Lipardi, C., Wei, Q., and Paterson, B. M. (2001). RNAi as random degradative PCR: siRNA primers convert mRNA into dsRNAs that are degraded to generate new siRNAs. Cell 107(3), 297-307.

Liu, D., Raghothama, K. G., Hasegawa, P. M., and Bressan, R. A. (1994). Osmotin overexpression in potato delays development of disease symptoms. Proc Natl Acad Sci U S A 91(5), 1888-92.

Liu, F., and Green, M. R. (1994). Promoter targeting by adenovirus E1a through interaction with different cellular DNA-binding domains. Nature 368, 520-525.

Llave, C., Kasschau, K. D., and Carrington, J. C. (2000). Virus-encoded suppressor of posttranscriptional gene silencing targets a maintenance step in the silencing pathway. Proc. Natl. Acad. Sci., USA 97, 13401-13406.

186

Llave, C., Kasschau, K. D., Rector, M. A., and Carrington, J. C. (2002). Endogenous and silencing-associated small RNAs in plants. Plant Cell 14(7), 1605-19.

Lu, C.-A., E.-K. Lim (1996). Sugar Response Sequence in the Promoter of a Rice a- Amylase Gene Serves as a Transcriptional Enhancer. THE JOURNAL OF BIOLOGICAL CHEMISTRY 273(17), 10120-10131.

Lucioli, A., Noris, E., Brunetti, A., Tavazza, R., Ruzza, V., Castillo, A. G., Bejarano, E. R., Accotto, G. P., and Tavazza, M. (2003). Tomato yellow leaf curl Sardinia virus rep-derived resistance to homologous and heterologous geminiviruses occurs by different mechanisms and is overcome if virus-mediated transgene silencing is activated. J Virol 77(12), 6785-98.

Ludin, K., Jiang, R., and Carlson, M. (1998). Glucose-regulated interaction of a regulatory subunit of protein phosphatase 1 with the Snf1 protein kinase in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A 95, 6245-6250.

Luque, A., Sanz-Burgos, A. P., Ramirez-Parra, E., Castellano, M. M., and Gutierrez, C. (2002). Interaction of geminivirus Rep protein with replication factor C and its potential role during geminivirus DNA replication. Virology 302(1), 83-94.

Mackintosh, R. W., Davies, S. P., Clarke, P. R., Weekes, J., Gillespie, J. G., Gibb, B. J., and Hardie, D. G. (1992). Evidence for a protein kinase cascade in higher plants. 3-Hydroxy-3-methylglutaryl-CoA reductase kinase. Eur J Biochem 209(3), 923- 31.

Mallory, A. C., Ely, L., Smith, T. H., Marathe, R., Anandalakshmi, R., Fagard, M., Vaucheret, H., Pruss, G., Bowman, L., and Vance, V. B. (2001). HC-Pro suppression of transgene silencing eliminates the small RNAs but not transgene methylation or the mobile signal. Plant Cell 13(3), 571-83.

Mallory, A. C., Reinhart, B. J., Bartel, D., Vance, V. B., and Bowman, L. H. (2002). A viral suppressor of RNA silencing differentially regulates the accumulation of short interfering RNAs and micro-RNAs in tobacco. Proc Natl Acad Sci U S A 99(23), 15228-33.

Mansoor, S., Briddon, R. W., Zafar, Y., and Stanley, J. (2003). Geminivirus disease complexes: an emerging threat. Trends Plant Sci 8(3), 128-34.

Martin, B. J., McClanahan, T. B., Van Wylen, D. G., and Gallagher, K. P. (1997). Effects of ischemia, preconditioning, and adenosine deaminase inhibition on interstitial adenosine levels and infarct size. Basic Res Cardiol 92(4), 240-51.

187

Martinez, Y., Gomez, O., Quinones, M., Pinon, M., and Fonseca, D. (2002). [Detection of begomovirus in accessions of Lycopersicon.]. Meeting of the American Phytopathological Society, Caribbean Division 92(6 Supplement), S126.

Matzke, M., Matzke, A. J., and Kooter, J. M. (2001). RNA: guiding gene silencing. Science 293(5532), 1080-3.

Matzke, M. A., Matzke, A. J., Pruss, G. J., and Vance, V. B. (2001). RNA-based silencing strategies in plants. Curr Opin Genet Dev 11(2), 221-7.

McCartney, R. R., and Schmidt, M. C. (2001). Regulation of Snf1 kinase. Activation requires phosphorylation of threonine 210 by an upstream kinase as well as a distinct step mediated by the Snf4 subunit. J Biol Chem 276(39), 36460-6.

McGaraughty, S., Chu, K. L., Wismer, C. T., Mikusa, J., Zhu, C. Z., Cowart, M., Kowaluk, E. A., and Jarvis, M. F. (2001). Effects of A-134974, a novel adenosine kinase inhibitor, on carrageenan-induced inflammatory hyperalgesia and locomotor activity in rats: evaluation of the sites of action. J Pharmacol Exp Ther 296(2), 501-9.

McGaraughty, S., Cowart, M., and Jarvis, M. F. (2001). Recent developments in the discovery of novel adenosine kinase inhibitors: mechanism of action and therapeutic potential. CNS Drug Rev 7(4), 415-32.

Mitchelhill, K. I., Stapleton, D., Gao, G., House, C., Michell, B., Katsis, F., Witters, L. A., and Kemp, B. E. (1994). Mammalian AMP-activated protein kinase shares structural and functional homology with the catalytic domain of yeast Snf1 protein kinase. J Biol Chem 269(4), 2361-4.

Moffatt, B. A., Stevens, Y. Y., Allen, M. S., Snider, J. D., Periera, L. A., Todorova, M. I., Summers, P. S., Weretilnyk, E. A., Martin-Caffrey, L., and Wagner, C. (2002). Adenosine kinase deficiency is associated with developmental abnormalities and reduced transmethylation. Plant Physiol. 128, 812-821.

Moffatt, B. A., Wang, L., Allen, M. S., Stevens, Y. Y., Qin, W., Snider, J., and von Schwartzenberg, K. (2000). Adenosine kinase of Arabidopsis. Kinetic properties and gene expression. Plant Physiol 124(4), 1775-85.

Moore, F., Weekes, J., and Hardie, D. G. (1991). Evidence that AMP triggers phosphorylation as well as direct allosteric activation of rat liver AMP-activated protein kinase. A sensitive mechanism to protect the cell against ATP depletion. Eur J Biochem 199(3), 691-7.

188

Moorhead, G., Douglas, P., Morrice, N., Scarabel, M., Aitken, A., and MacKintosh, C. (1996). Phosphorylated nitrate reductase from spinach leaves is inhibited by 14-3- 3 proteins and activated by fusicoccin. Curr Biol 6(9), 1104-13.

Morel, J. B., Mourrain, P., Beclin, C., and Vaucheret, H. (2000). DNA methylation and chromatin structure affect transcriptional and post-transcriptional transgene silencing in Arabidopsis. Curr Biol 10(24), 1591-4.

Morris, B. A. M., Richardson, K. A., Haley, A., Zhan, X., and Thomas, J. E. (1992). The nucleotide sequence of the infectious cloned DNA component of tobacco yellow dwarf virus reveals features of geminiviruses infecting monocotyledonous plants. Virology 187, 633-642.

Mourelatos, Z., Dostie, J., Paushkin, S., Sharma, A., Charroux, B., Abel, L., Rappsilber, J., Mann, M., and Dreyfuss, G. (2002). miRNPs: a novel class of ribonucleoproteins containing numerous microRNAs. Genes Dev 16(6), 720-8.

Mourrain, P., Beclin, C., Elmayan, T., Feuerbach, F., Godon, C., Morel, J. B., Jouette, D., Lacombe, A. M., Nikic, S., Picault, N., Remoue, K., Sanial, M., Vo, T. A., and Vaucheret, H. (2000). Arabidopsis SGS2 and SGS3 genes are required for posttranscriptional gene silencing and natural virus resistance. Cell 101(5), 533- 42.

Nagar, S., Pedersen, T. J., Carrick, K. M., Hanley-Bowdoin, L., and Robertson, D. (1995). A geminivirus induces expression of a host DNA synthesis protein in terminally differentiated plant cells. Plant Cell 7(6), 705-19.

Napoli, C., Lemieux, C., and Jorgensen, R. (1990). Introduction of a Chimeric Chalcone Synthase Gene into Petunia Results in Reversible Co-Suppression of Homologous Genes in trans. Plant Cell 2(4), 279-289.

Noris, E., Jupin, I., Accotto, G. P., and Gronenborn, B. (1996). DNA-binding activity of the C2 protein of tomato yellow leaf curl geminivirus. Virology 217(2), 607-12.

Novick, R. P. (1998). Contrasting lifestyles of rolling-circle phages and plasmids. Trends Biochem Sci 23(11), 434-8.

Nurnberger, T., and Scheel, D. (2001). Signal transmission in the plant immune response. Trends Plant Sci 6(8), 372-9.

Nykanen, P. (2001). On conceptualization of a decision support system in health informatics. Medinfo 10(Pt 1), 503-7.

Orozco, B. M., and Hanley-Bowdoin, L. (1996). A DNA structure is required for geminivirus replication origin function. J Virol 70(1), 148-58.

189

Palatnik, J. F., Allen, E., Wu, X., Schommer, C., Schwab, R., Carrington, J. C., and Weigel, D. (2003). Control of leaf morphogenesis by microRNAs. Nature 425(6955), 257-63.

Palauqui, J. C., Elmayan, T., Pollien, J. M., and Vaucheret, H. (1997). Systemic acquired silencing: transgene-specific post-transcriptional silencing is transmitted by grafting from silenced stocks to non-silenced scions. Embo J 16(15), 4738-45.

Palmer, K. E., and Rybicki, E. P. (1998). The molecular biology of mastreviruses. Adv Virus Res 50, 183-234.

Park, S. H., Hur, J., Park, J., Lee, S., Lee, T. K., Chang, M., Davi, K. R., and Kim, J. (2002). Identification of a tolerant locus on Arabidopsis thaliana to hypervirulent beet curly top virus CFH strain. Mol Cells 13(2), 252-8.

Paszkowski, J., and Whitham, S. A. (2001). Gene silencing and DNA methylation processes. Curr. Opin. Plant Biol. 4, 123-129.

Pedersen, T. J., and Hanley-Bowdoin, L. (1994). Molecular characterization of the AL3 protein encoded by a bipartite geminivirus. Virology 202(2), 1070-5.

Peele, C., Jordan, C. V., Muangsan, N., Turnage, M., Egelkrout, E., Eagle, P., Hanley- Bowdoin, L., and Robertson, D. (2001). Silencing of a meristematic gene using geminivirus-derived vectors. Plant J 27(4), 357-66.

Pelissier, T., Thalmair, S., Kempe, D., Sanger, H. L., and Wasseneger, M. (1999). Heavy de novo methylation at symmetrical and non-symmetrical sites is a hallmark of RNA-directed DNA methylation. Nucleic Acids Res. 27, 1625-1643.

Phillips, G. J., Park, S.-K., and Huber, D. (2000). High copy number plasmids compatible with commonly used cloning vectors. BioTechniques 28, 400-408.

Pickford, A. S., Catalanotto, C., Cogoni, C., and Macino, G. (2002). Quelling in Neurospora crassa. Adv Genet 46, 277-303.

Pilartz, M., and Jeske, H. (1992). Abutilon mosaic geminivirus double-stranded DNA is packed into minichromosomes. Virology 189(2), 800-2.

Plasterk, R. H. A. (2002). RNA silencing: The genome's immune system. Science 296, 1263-1265.

Pringle, C. R. (1999a). Virus nomenclature. Arch Virol 144(7), 1463-6.

190

Pringle, C. R. (1999b). Virus taxonomy--1999. The universal system of virus taxonomy, updated to include the new proposals ratified by the International Committee on Taxonomy of Viruses during 1998. Arch Virol 144(2), 421-9.

Pruss, G., Shi, X. M., Carrington, J. C., and Vance, V. B. (1997). Plant viral synergism: The potyviral genome encodes a broad-range pathogenicity enhancer that transactivates replication of heterologous viruses. Plant Cell 9, 859-868.

Purcell, P. C., Smith A, M, Halford N. G. (1998). Antisense expression of a sucrose nonfermenting-1-related protein kinase sequence in potato results in decreased expression of sucrose synthase in tubers and loss of sucrose-inducibility of sucrose synthase transcripts in leaves. The Plant Journal 14, 195-202.

Qu, F., Ren, T., and Morris, T. J. (2003). The coat protein of turnip crinkle virus suppresses posttranscriptional silencing at an early initiation step. J. Virol. 77, 511-522.

Ratcliff, F., Harrison, B. D., and Baulcombe, D. C. (1997). A similarity between viral defense and gene silencing in plants. Science 276, 1558-1560.

Razin, A. (1998). CpG methylation, chromatin structure and gene silencing-a three-way connection. Embo J 17(17), 4905-8.

Reed, J. C., Kasschau, K. D., Prokhnevsky, A. I., Gopinath, K., Pogue, G. P., Carrington, J. C., and Dolja, V. V. (2003). Suppressor of RNA silencing encoded by Beet yellows virus. Virology 306(2), 203-9.

Reinhart, B. J., and Bartel, D. P. (2002). Small RNAs correspond to centromere heterochromatic repeats. Science 297(5588), 1831.

Restrepo, M. A., Freed, D. D., and Carrington, J. C. (1990). Nuclear transport of plant potyviral proteins. Plant Cell 2(10), 987-98.

Rice, J. C., and Allis, C. D. (2001). Histone methylation versus histone acetylation: new insights into epigenetic regulation. Curr Opin Cell Biol 13(3), 263-73.

Richards, E. J., and Elgin, S. C. (2002). Epigenetic codes for heterochromatin formation and silencing: rounding up the usual suspects. Cell 108(4), 489-500.

Rogers, S. G., Bisaro, D. M., Horsch, R. B., Fraley, R. T., Hoffmann, N. L., Brand, L., Elmer, J. S., and Lloyd, A. M. (1986). Tomato golden mosaic virus A component DNA replicates autonomously in transgenic plants. Cell 45, 593-600.

Roitsch, T. (1999). Source-sink regulation by sugar and stress. Curr Opin Plant Biol 2(3), 198-206.

191

Roitsch, T., Balibrea, M. E., Hofmann, M., Proels, R., and Sinha, A. K. (2003). Extracellular invertase: key metabolic enzyme and PR protein. J Exp Bot 54(382), 513-24.

Ronne, H. (1995). Glucose repression in fungi. Trends Genet 11(1), 12-7.

Roth, S., Kumme, J., and Schuller, H. J. (2003). Transcriptional activators Cat8 and Sip4 discriminate between sequence variants of the carbon source-responsive promoter element in the yeast Saccharomyces cerevisiae. Curr Genet.

Ruiz, M. T., Voinnet, O., and Baulcombe, D. C. (1998). Initiation and maintenance of virus-induced gene silencing. Plant Cell 10, 937-946.

Ryals, J., Lawton, K. A., Delaney, T. P., Friedrich, L., Kessmann, H., Neuenschwander, U., Uknes, S., Vernooij, B., and Weymann, K. (1995). Signal transduction in systemic acquired resistance. Proc Natl Acad Sci U S A 92(10), 4202-5.

Ryals, J. A., Neunschwander, U. H., Willits, M. G., Molina, A., Steiner, H.-Y., and Hunt, M. D. (1996). Systemic acquired resistance. Plant Cell 8, 1809-1819.

Salt, I., Celler, J. W., Hawley, S. A., Prescott, A., Woods, A., Carling, D., and Hardie, D. G. (1998). AMP-activated protein kinase: greater AMP dependence, and preferential nuclear localization, of complexes containing the alpha2 isoform. Biochem J 334 ( Pt 1), 177-87.

Sanz, A. I., Fraile, A., Garcia-Arenal, F., Zhou, X., Robinson, D. J., Khalid, S., Butt, T., and Harrison, B. D. (2000). Multiple infection, recombination and genome relationships among begomovirus isolates found in cotton and other plants in Pakistan. J Gen Virol 81(Pt 7), 1839-49.

Sato, R., Goldstein, J. L., and Brown, M. S. (1993). Replacement of serine-871 of hamster 3-hydroxy-3-methylglutaryl-CoA reductase prevents phosphorylation by AMP-activated kinase and blocks inhibition of sterol synthesis induced by ATP depletion. Proc Natl Acad Sci U S A 90(20), 9261-5.

Saunders, K., Bedford, I. D., Yahara, T., and Stanley, J. (2003). Aetiology: The earliest recorded plant virus disease. Nature 422(6934), 831.

Saunders, K., Lucy, A., and Stanley, J. (1991). DNA forms of the geminivirus African cassava mosaic virus consistent with a rolling circle mechanism of replication. Nucleic Acids Res 19(9), 2325-30.

Scholes, J. D., Lee, P.J., Horton, P. and Lewis, D.H (1994). Invertase: understanding changes in the photosynthetic and carbohydrate metabolism of barley leaves infected with powdery mildew. New Phytologist 126, 216-222.

192

Schwarz, D. S., Hutvagner, G., Du, T., Xu, Z., Aronin, N., and Zamore, P. D. (2003). Asymmetry in the assembly of the RNAi enzyme complex. Cell 115(2), 199-208.

Schweizer, P., Pokorny, J., Schulze-Lefert, P., and Dudler, R. (2000). Technical advance. Double-stranded RNA interferes with gene function at the single-cell level in cereals. Plant J 24(6), 895-903.

Seemanpillai, M., Dry, I., Randles, J., and Rezaian, A. (2003). Transcriptional silencing of geminiviral promoter-driven transgenes following homologous virus infection. Mol Plant Microbe Interact 16(5), 429-38.

Settlage, S. B., Miller, A. B., and Hanley-Bowdoin, L. (1996). Interactions between geminivirus replication proteins. J Virol 70(10), 6790-5.

Shahbazian, M., Young, J., Yuva-Paylor, L., Spencer, C., Antalffy, B., Noebels, J., Armstrong, D., Paylor, R., and Zoghbi, H. (2002). Mice with truncated MeCP2 recapitulate many Rett syndrome features and display hyperacetylation of histone H3. Neuron 35(2), 243-54.

Shalitin, D., and Wolf, S. (2000). Cucumber mosaic virus infection affects sugar transport in melon plants. Plant Physiol 123(2), 597-604.

Sharp, P. A. (2001). RNA interference-2001. Genes Dev. 15, 485-490.

Silhavy, D., Molnar, A., Lucioli, A., Szittya, G., Hornyik, C., Tavazza, M., and Burgyan, J. (2002). A viral protein suppresses RNA silencing and binds silencing- generated, 21- to 25-nucleotide double-stranded RNAs. EMBO J. 21, 3070-3080.

Singh, B., Hao, W., Wu, Z., Eigl, B., and Gupta, R. S. (1996). Cloning and characterization of cDNA for adenosine kinase from mammalian (Chinese hamster, mouse, human and rat) species. High frequency mutants of Chinese hamster ovary cells involve structural alterations in the gene. Eur J Biochem 241(2), 564-71.

Slack, F. J., Basson, M., Liu, Z., Ambros, V., Horvitz, H. R., and Ruvkun, G. (2000). The lin-41 RBCC gene acts in the C. elegans heterochronic pathway between the let-7 regulatory RNA and the LIN-29 transcription factor. Mol Cell 5(4), 659-69.

Slocombe, S. P., Laurie, S., Bertini, L., Beaudoin, F., Dickinson, J. R., and Halford, N. G. (2002). Identification of SnIP1, a novel protein that interacts with SNF1-related protein kinase (SnRK1). Plant Mol Biol 49(1), 31-44.

Smith, N. A., Singh, S. P., Wang, M. B., Stoutjesdijk, P. A., Green, A. G., and Waterhouse, P. M. (2000). Total silencing by intron-spliced hairpin RNAs. Nature 407(6802), 319-20.

193

Spychala, J., Datta, N. S., Takabayashi, K., Datta, M., Fox, I. H., Gribbin, T., and Mitchell, B. S. (1996). Cloning of human adenosine kinase cDNA: Sequence similarity to microbial ribokinases and fructokinases. Proc. Natl. Acad. Sci. USA 93, 1232-1237.

Stanley, J. (1983). Infectivity of the cloned geminivirus genome requires sequences from both DNAs. Nature 305, 643-645.

Stanley, J., Latham, J., Pinner, M. S., Bedford, I., and Markham, P. G. (1992). Mutational analysis of the monopartite geminivirus beet curly top virus. Virology 191, 396- 405.

Stanley, J., Markham, P. G., Callis, R. J., and Pinner, M. S. (1986). The nucleotide sequence of an infectious clone of the geminivirus beet curly top virus. EMBO J. 5, 1761-1767.

Steimer, A., Amedeo, P., Afsar, K., Fransz, P., Scheid, O. M., and Paszkowski, J. (2000). Endogenous targets of transcriptional gene silencing in Arabidopsis. Plant Cell 12(7), 1165-78.

Strahl, B. D., and Allis, C. D. (2000). The language of covalent histone modifications. Nature 403(6765), 41-5.

Sturm, A., and Chrispeels, M. J. (1990). cDNA cloning of carrot extracellular β- fructosidase and its expression in response to wounding and bacterial infection. Plant Cell 2, 1107-1119.

Sugden, C., Crawford, R. M., Halford, N. G., and Hardie, D. G. (1999a). Regulation of spinach SNF1-related (SnRK1) kinases by protein kinases and phosphatases is associated with phosphorylation of the T loop and is regulated by 5'-AMP. Plant J. 19, 433-439.

Sugden, C., Donaghy, P. G., Halford, N., and Hardie, D. G. (1999b). Two SNF1-related protein kinases from spinach leaf phosphorylate and inactivate 3-hydroxyl-3- methylglutaryl-coenzyme A reductase, nitrate reductase, and sucrose phosphate synthase in vitro. Plant Physiol. 120, 257-274.

Sung, Y. K., and Coutts, R. H. (1995). Pseudorecombination and complementation between potato yellow mosaic geminivirus and tomato golden mosaic geminivirus. J Gen Virol 76 ( Pt 11), 2809-15.

Sung, Y. K., and Coutts, R. H. (1996). Potato yellow mosaic geminivirus AC2 protein is a sequence non-specific DNA binding protein. FEBS Lett 383(1-2), 51-4.

194

Sunter, G., and Bisaro, D. M. (1989). Transcription map of the B genome component of tomato golden mosaic virus and comparison with A component transcripts. Virology 173(2), 647-55.

Sunter, G., and Bisaro, D. M. (1991). Transactivation in a geminivirus: AL2 gene product is needed for coat protein expression. Virology 180(1), 416-9.

Sunter, G., and Bisaro, D. M. (1992). Transactivation of geminivirus AR1 and BR1 gene expression by the viral AL2 gene product occurs at the level of transcription. Plant Cell 4(10), 1321-31.

Sunter, G., and Bisaro, D. M. (1997). Regulation of a geminivirus coat protein promoter by AL2 protein (TrAP): evidence for activation and derepression mechanisms. Virology 232(2), 269-80.

Sunter, G., and Bisaro, D. M. (2003). Identification of a minimal sequence required for activation of the tomato golden mosaic virus coat protein promoter in protoplasts. Virology 305(2), 452-62.

Sunter, G., Hartitz, M. D., and Bisaro, D. M. (1993). Tomato golden mosaic virus leftward gene expression: autoregulation of geminivirus replication protein. Virology 195(1), 275-80.

Sunter, G., Hartitz, M. D., Hormuzdi, S. G., Brough, C. L., and Bisaro, D. M. (1990). Genetic analysis of tomato golden mosaic virus: ORF AL2 is required for coat protein accumulation while ORF AL3 is necessary for efficient DNA replication. Virology 179(1), 69-77.

Sunter, G., Stenger, D. C., and Bisaro, D. M. (1994). Heterologous complementation by geminivirus AL2 and AL3 genes. Virology 203(2), 203-10.

Sunter, G., Sunter, J. L., and Bisaro, D. M. (2001). Plants expressing tomato golden mosaic virus AL2 or beet curly top virus L2 transgenes show enhanced susceptibility to infection by DNA and RNA viruses. Virology 285(1), 59-70.

Sutherland, C. M., Hawley, S. A., McCartney, R. R., Leech, A., Stark, M. J., Schmidt, M. C., and Hardie, D. G. (2003). Elm1p is one of three upstream kinases for the Saccharomyces cerevisiae SNF1 complex. Curr Biol 13(15), 1299-305.

Tamaru, H., and Selker, E. U. (2001). A histone H3 methyltransferase controls DNA methylation in Neurospora crassa. Nature 414, 277-283.

Tanaka, H., Masuta, C., Uehara, K., Kataoka, J., Koiwai, A., and Noma, M. (1997). Morphological changes and hypomethylation of DNA in transgenic tobacco

195

expressing antisense RNA of the S-adenosyl-L-homocysteine hydrolase gene. Plant Mol Biol 35(6), 981-6.

Tang, F. C., Yang, H. B., Meng, G. L., Li, C. J., Shang, K. G., Zhang, B., and Xue, Y. F. (2003). RNA interference directed by small hairpin RNA expressed in COS-7 cells. Yi Chuan Xue Bao 30(4), 295-300.

Tatlisumak, T., Takano, K., Carano, R. A., Miller, L. P., Foster, A. C., and Fisher, M. (1998). Delayed treatment with an adenosine kinase inhibitor, GP683, attenuates infarct size in rats with temporary middle cerebral artery occlusion. Stroke 29(9), 1952-8.

Tecsi, L. I., A. J. Maule (1994). Metabolic alterations in cotyledons of Cucurbita pepo infected by cucumber mosaic virus. Journal of Experimental Botany 45(280), 1541-1551.

Tecsi, L. I., Smith, A. M., Maule, A. J., and Leegood, R. C. (1996). A Spatial Analysis of Physiological Changes Associated with Infection of Cotyledons of Marrow Plants with Cucumber Mosaic Virus. Plant Physiol 111(4), 975-985.

Tenllado, F., and Diaz-Ruiz, J. R. (2001). Double-stranded RNA-mediated interference with plant virus infection. J Virol 75(24), 12288-97.

Thomma, B. P., Penninckx, I. A., Broekaert, W. F., and Cammue, B. P. (2001). The complexity of disease signaling in Arabidopsis. Curr Opin Immunol 13(1), 63-8.

Toroser, D., Plaut, Z., and Huber, S. C. (2000). Regulation of a plant SNF1-related protein kinase by glucose-6-phosphate. Plant Physiol 123(1), 403-12.

Triezenberg, S. J., Kingsbury, R. C., and McKnight, S. L. (1988). Functional dissection of VP16, the trans-activator of herpes simplex virus immediate early gene expression. Genes Dev. 2, 718-729.

Triezenberg, S. J., LaMarco, K. L., and McKnight, S. L. (1988). Evidence of DNA:protein interactions that mediate HSV-1 immediate early gene activation by VP16. Genes Dev. 2, 730-742.

Tsukaya, H., Oshima, T., Naito, S., Chino, M., and Komeda, Y. (1991). Sugar-dependent expression of the CHS-A gene for chalcone synthase from petunia in transgenic Arabidopsis. Plant Physiol. 97, 1414-1421.

Uknes, S., Mauch-Mani, B., Moyer, M., Potter, S., Williams, S., Dincher, S., Chandler, D., Slusarenko, A., Ward, E., and Ryals, J. (1992). Acquired resistance in Arabidopsis. Plant Cell 4(6), 645-56.

196

van der Krol, A. R., Mur, L. A., Beld, M., Mol, J. N., and Stuitje, A. R. (1990). Flavonoid genes in petunia: addition of a limited number of gene copies may lead to a suppression of gene expression. Plant Cell 2(4), 291-9.

van Loon, L. C., and van Kammen, A. (1970). Polyacrylamide disc electrophoresis of the soluble leaf proteins from Nicotiana tabacum var. "Samsun" and "Samsun NN". II. Changes in protein constitution after infection with tobacco mosaic virus. Virology 40(2), 190-211.

van Wezel, R., Liu, H., Tien, P., Stanley, J., and Hong, Y. (2001). Gene C2 of the monopartite geminivirus tomato yellow leaf curl virus-China encodes a pathogenicity determinant that is localized in the nucleus. Mol Plant Microbe Interact 14(9), 1125-8.

van Wezel, R., Liu, H., Tien, P., Stanley, J., and Hong, Y. (2002). Mutation of three cysteine residues in tomato yellow leaf curl virus-China C2 protein causes dysfunction in pathogenesis and posttranscriptional gene silencing-suppression. Mol. Plant Microbe Interact. 15, 203-208.

Van Wezel, R., Liu, H., Wu, Z., Stanley, J., and Hong, Y. (2003). Contribution of the zinc finger to zinc and DNA binding by a suppressor of posttranscriptional gene silencing. J Virol 77(1), 696-700. van, W. R., Dong, X., Liu, H., Tien, P., Stanley, J., and Hong, Y. (2002). Mutation of three cysteine residues in Tomato yellow leaf curl virus-China C2 protein causes dysfunction in pathogenesis and posttranscriptional gene-silencing suppression. Mol Plant Microbe Interact 15(3), 203-8.

Vance, V., and Vaucheret, H. (2001). RNA silencing in plants: Defense and counterdefense. Science 292, 2277-2280.

Vance, V. B. (1991). Replication of potato virus X RNA is altered in coinfections with potato virus Y. Virology 182(2), 486-94.

Vanitharani, R., Chellappan, P., and Fauquet, C. M. (2003). Short interfering RNA- mediated interference of gene expression and viral DNA accumulation in cultured plant cells. Proc Natl Acad Sci U S A 100(16), 9632-6.

Vargason, J. M., Szittya, G., Burgyan, J., and Tanaka Hall, T. M. (2003). Size selective recognition of siRNA by an RNA silencing suppressor. Cell 115, 799-811.

Vijayan, P., Shockey, J., Levesque, C. A., Cook, R. J., and Browse, J. (1998). A role for jasmonate in pathogen defense of arabidopsis. Proc Natl Acad Sci U S A 95(12), 7209-14.

197

Vincent, O., and Carlson, M. (1999). Gal83 mediates the interaction of the Snf1 kinase complex with the transcription activator Sip4. Embo J 18(23), 6672-81.

Vincent, O., Townley, R., Kuchin, S., and Carlson, M. (2001). Subcellular localization of the Snf1 kinase is regulated by specific beta subunits and a novel glucose signaling mechanism. Genes Dev 15(9), 1104-14.

Voinnet, O. (2001). RNA silencing as a plant immune system against viruses. Trends Genet 17(8), 449-59.

Voinnet, O. (2002). RNA silencing: small RNAs as ubiquitous regulators of gene expression. Curr. Opin. Plant Biol. 5, 444-451.

Voinnet, O., and Baulcombe, D. C. (1997). Systemic signalling in gene silencing. Nature 389(6651), 553.

Voinnet, O., Lederer, C., and Baulcombe, D. C. (2000). A viral movement protein prevents spread of the gene silencing signal in Nicotiana benthamiana. Cell 103(1), 157-67.

Voinnet, O., Pinto, Y. M., and Baulcombe, D. C. (1999). Suppression of gene silencing: A general strategy used by diverse DNA and RNA viruses of plants. Proc. Natl. Acad. Sci. USA 96, 14147-14152.

Voinnet, O., Rivas, S., Mestre, P., and Baulcombe, D. (2003). An enhanced transient expression system in plants based on suppression of gene silencing by the p19 protein of tomato bushy stunt virus. Plant J 33(5), 949-56.

Voinnet, O., Vain, P., Angell, S., and Baulcombe, D. C. (1998). Systemic spread of sequence-specific transgene RNA degradation in plants is initiated by localized introduction of ectopic promoterless DNA. Cell 95(2), 177-87.

Volpe, T. A., Kidner, C., Hall, I. M., Teng, G., Grewal, S. I., and Martienssen, R. A. (2002). Regulation of heterochromatic silencing and histone H3 lysine-9 methylation by RNAi. Science 297(5588), 1833-7. von Schwartzenberg, K., Kruse, S., Reski, R., Moffatt, B., and Laloue, M. (1998). Cloning and characterization of an adenosine kinase from Physcomitrella involved in cytokinin metabolism. Plant J. 13, 249-257.

Wang, D., Harper, J. F., and Gribskov, M. (2003). Systematic trans-genomic comparison of protein kinases between Arabidopsis and Saccharomyces cerevisiae. Plant Physiol 132(4), 2152-65.

198

Wang, H., Hao, L., Shung, C.-Y., Sunter, G., and Bisaro, D. M. (2003). Adenosine kinase is inactivated by geminivirus AL2 and L2 proteins. Plant Cell 15, 3020-3032.

Ward, E. R., Uknes, S. J., Williams, S. C., Dincher, S. S., Wiederhold, D. L., Alexander, D. C., Ahl-Goy, P., Metraux, J. P., and Ryals, J. A. (1991). Coordinate Gene Activity in Response to Agents That Induce Systemic Acquired Resistance. Plant Cell 3(10), 1085-1094.

Waseem, N. H., Labib, K., Nurse, P., and Lane, D. P. (1992). Isolation and analysis of the fission yeast gene encoding polymerase delta accessory protein PCNA. Embo J 11(13), 5111-20.

Wasseneger, M., Heimes, S., Reidel, L., and Sanger, H. (1994). RNA-directed de novo methylation of genomic sequences in plants. Cell 76, 567-576.

Wassenegger, M. (2000). RNA-directed DNA methylation. Plant Mol Biol 43(2-3), 203- 20.

Wassenegger, M., Heimes, S., Riedel, L., and Sanger, H. L. (1994). RNA-directed de novo methylation of genomic sequences in plants. Cell 76(3), 567-76.

Waterhouse, P. M., Graham, M. W., and Wang, M. B. (1998). Virus resistance and gene silencing in plants can be induced by simultaneous expression of sense and antisense RNA. Proc Natl Acad Sci U S A 95(23), 13959-64.

Waterhouse, P. M., Wang, M. B., and Lough, T. (2001). Gene silencing as an adaptive defense against viruses. Nature 411, 834-842.

Watson, M. A. (1951). The effect of infection with beet yellows and beet mosaic viruses on the carbohydrate content of sugar-beet leaves, and on translocation. Ann. Appl. Biology 38, 276-289.

Weekes, J., Ball, K. L., Caudwell, F. B., and Hardie, D. G. (1993). Specificity determinants for the AMP-activated protein kinase and its plant homologue analysed using synthetic peptides. FEBS Letters 334, 335-339.

Weinberg, R. A. (1995). The retinoblastoma protein and cell cycle control. Cell 81(3), 323-30.

Weretilnyk, E. A., Alexander, K. J., Drebenstedt, M., Snider, J. D., Summers, P. S., and Moffatt, B. A. (2001). Maintaining methylation activities during salt stress. The involvement of adenosine kinase. Plant Physiol. 125, 856-865.

199

White, J. L. B. (1983). Protein changes in wheat infected with wheat streak mosaic virus and in barley infected with barley stripe mosaic virus. Physiol. Plant Pathol. 22, 87-100.

Wiesner, J. B., Ugarkar, B. G., Castellino, A. J., Barankiewicz, J., Dumas, D. P., Gruber, H. E., Foster, A. C., and Erion, M. D. (1999). Adenosine kinase inhibitors as a novel approach to anticonvulsant therapy. J Pharmacol Exp Ther 289(3), 1669- 77.

Wilson, W. A., Hawley, S. A., and Hardie, D. G. (1996). Glucose repression/derepression in budding yeast: SNF1 protein kinase is activated by phosphorylation under derepressing conditions, and this correlates with a high AMP:ATP ratio. Curr Biol 6(11), 1426-34.

Winder, W. W., Wilson, H. A., Hardie, D. G., Rasmussen, B. B., Hutber, C. A., Call, G. B., Clayton, R. D., Conley, L. M., Yoon, S., and Zhou, B. (1997). Phosphorylation of rat muscle acetyl-CoA carboxylase by AMP-activated protein kinase and protein kinase A. J Appl Physiol 82(1), 219-25.

Witters, L. A., and Kemp, B. E. (1992). Insulin activation of acetyl-CoA carboxylase accompanied by inhibition of the 5'-AMP-activated protein kinase. J Biol Chem 267(5), 2864-7.

Woods, A., Munday, M. R., Scott, J., Yang, X., Carlson, M., and Carling, D. (1994). Yeast SNF1 is functionally related to mammalian AMP-activated protein kinase and regulates acetyl-CoA carboxylase in vivo. J Biol Chem 269(30), 19509-15.

Wu-Scharf, D., Jeong, B.-R., Zhang, C., and Cerutti, H. (2000). Transgene and transposon silencing in Chlamydomonas reinhardtii by a DEAH-box RNA helicase. Science 290, 1159-1162.

Xie, Q., Sanzburgos, P., Hannon, G. J., and Gutierrez, C. (1996). Plant cells contain a novel member of the retinoblastoma family of growth regulatory proteins. EMBO J. 15, 4900-4908.

Xie, Q., Suarez-Lopez, P., and Gutierrez, C. (1995). Identification and analysis of a retinoblastoma binding motif in the replication protein of a plant DNA virus: requirement for efficient viral DNA replication. Embo J 14(16), 4073-82.

Xie, Z., Kasschau, K. D., and Carrington, J. C. (2003). Negative feedback regulation of Dicer-Like1 in Arabidopsis by microRNA-guided mRNA degradation. Curr Biol 13(9), 784-9.

200

Yalpani, N., Silverman, P., Wilson, T. M., Kleier, D. A., and Raskin, I. (1991). Salicylic acid is a systemic signal and an inducer of pathogenesis-related proteins in virus- infected tobacco. Plant Cell 3(8), 809-18.

Ye, K., Malinina, L., and Patel, D. J. (2003). Recognition of small interfering RNA by a viral suppressor of RNA silencing. Nature 426, 874-878.

Zamore, P. D. (2002). Ancient pathways programmed by small RNAs. Science 296, 1265-1269.

Zamore, P. D., Tuschi, T., Sharp, P. A., and Bartel, D. P. (2000). RNAi: dsRNA directs the ATP-dependent cleavage of mRNA at 21-23 nt intervals. Cell 101, 25-33.

Zeng, Y., Wagner, E. J., and Cullen, B. R. (2002). Both natural and designed micro RNAs can inhibit the expression of cognate mRNAs when expressed in human cells. Mol Cell 9(6), 1327-33.

Zeng, Y., Yi, R., and Cullen, B. R. (2003). MicroRNAs and small interfering RNAs can inhibit mRNA expression by similar mechanisms. Proc Natl Acad Sci U S A 100(17), 9779-84.

Zhang, S. C., Ghosh, R., and Jeske, H. (2002). Subcellular targeting domains of Abutilon mosaic geminivirus movement protein BC1. Arch Virol 147(12), 2349-63.

Zhu, C. Z., Mikusa, J., Chu, K. L., Cowart, M., Kowaluk, E. A., Jarvis, M. F., and McGaraughty, S. (2001). A-134974: a novel adenosine kinase inhibitor, relieves tactile allodynia via spinal sites of action in peripheral nerve injured rats. Brain Res 905(1-2), 104-10.

201