A Dissertation Entitled Isolation of An ARGONAUTE Gene in Pelargonium and Identification Of Candidate Genes Regulated Through ARGONAUTE4-Dependent RNA-Dependent DNA Methylation In Arabidopsis By Jie He Submitted as partial fulfillment of the requirements for the Doctor of Philosophy degree in Biology Advisor: Scott Leisner, Ph.D College of Graduate Studies The University of Toledo December 2009 An Abstract of Isolation of An ARGONAUTE Gene in Pelargonium and Identification Of Candidate Genes Regulated Through ARGONAUTE4-Dependent RNA-Dependent DNA Methylation In Arabidopsis Jie He Submitted as partial fulfillment of the requirements for the Doctor of Philosophy degree in Biology The University of Toledo December 2009 RNAi-induced gene silencing plays a role in plant DNA methylation and defense. While most gene silencing studies have been performed on annuals, little is known about the expression of key components of this process (like ARGONAUTE proteins) in ornamentals. Using a combination of polymerase chain reaction techniques, an ARGONAUTE4 gene, PhAGO4, was isolated from Pelargonium. PhAGO4 encodes a predicted product of 934 amino acids that contains the PAZ and PIWI domains typical of ARGONAUTE (AGO) proteins. Phylogenetic analyses indicate that PhAGO4 clusters with other plant AGO4 proteins. Organ expression patterns of the AGO4 genes in Pelargonium and Arabidopsis show intriguing differences. AGO4 RNA levels decline with leaf age in both Arabidopsis and Pelargonium. In contrast AGO4 RNA levels in roots relative to leaves are higher in ii Pelargonium than in Arabidopsis. Both Arabidopsis and Pelargonium AGO4 showed higher RNA levels in flowers than leaves or roots. Even though flowers show higher levels of PhAGO4 RNA when compared to leaves and roots, protein gel blot analysis shows that at the protein level, the reverse is true. RNA interference (RNAi) is a regulatory mechanism found in all eukaryotes that occurs at either the transcriptional or post-transcriptional level. Two key players in Arabidopsis transcriptional gene silencing are small RNAs (sRNAs) and ARGONAUTE (AGO) proteins, specifically AGO4. Therefore, a combination of microarray-based genome-wide transcript profiling, a potential sRNA binding profile, and methylome datasets was used to identify candidate AGO4-regulated protein coding target genes. The highest number of AGO4-associated small RNAs with the potential for binding protein- coding genes showed a preference for either the upstream 1000 bp portion (promoter) or intronic regions. Microarray studies identified 243 up-regulated genes in ago4-1 mutant plants when compared to wild type. Interestingly, 33 up-regulated genes correlated with loci containing potential AGO4-associated sRNA binding sites. AGO4 is thought to recruit enzymes catalyzing non-CG methylation. The majority of potential sRNA binding sites did show a correlation with the non-CG methylation pattern. Taken together, these 33 protein-coding genes are good candidates for direct AGO4 regulation. iii Table Of Contents Abstract ii Table of Contents iv List of Figures vi List of Tables vii Abbreviations viii I. Introduction 1 II. Materials and Methods 25 III. A Pelargonium ARGONAUTE4 Gene Shows Organ-Specific Expression And Differences In RNA And Protein Levels 37 Abstract 39 Introduction 40 Materials And Methods 42 Results 47 Discussion 50 References 54 IV. Identification Of Candidate Genes Regulated Through ARGONAUTE4- Dependent RNA-Dependent DNA Methylation In Arabidopsis 68 Abstract 69 Introduction 70 Materials And Methods 73 Results 76 Discussion 84 iv References 91 V. AGO4’s role in Anti-viral defense 105 Abstract 105 Introduction 106 Materials And Methods 111 Results 112 Discussion 115 References 118 VI. Discussion / Future work 126 References 133 v List of Figures Fig. 1. RNA silencing and RISC complex 22 Fig. 2. TGS and PTGS pathways 23 Fig. 3. Ten Argonaute proteins in Arabidopsis 24 Fig. 4. Arabidopsis Actin2 gene Melt Curve Chart for SYBR-490 36 Fig. 5. Structure of the predicted Pelargonium AGO4 protein 62 Fig. 6. Conservation of the DDH catalytic motif in AGO proteins 62 Fig. 7. Evolutionary relationships among AGO homologs 64 Fig. 8. Expression of AGO4 in Arabidopsis and Pelargonium organs 66 Fig. 9. The Arabidopsis ago4-1 mutant shows accelerated development 99 Fig. 10. Mapping AGO4-associated small RNAs to protein coding gene regions within the Arabidopsis genome 100 Fig. 11. Gene Ontology analysis of potential AGO4-regulated genes 101 Fig. 12. The potential sRNA binding sites present within the SDC gene correlate with sites of reduced non-CG methylation and expression of this gene is up-regulated in the ago4 mutant 102 Fig. 13. Average CG content for the potential AGO4-regulated targets 104 Fig. 14. Photoperiod influences AGO4 expression in Pelargonium 121 Fig. 15. Photoperiod influences PFBV RNA and protein levels 122 Fig. 16. PFBV RNA and protein levels correlate differently under different photoperiods 123 Fig. 17. CaMV infection affects expression of candidate AGO4 target genes 124 vi List of Tables Table 1. Components of RNA directed DNA methylation 21 Table 2. Gene accession data for sequence analysis 57 Table 3. Primers used for cloning, sequencing, and real time RT-PCR amplification of AGO4s and Actins 60 Table 4. List of candidate AGO4-regulated targets 96 Table 5. Comparison of common targets of different effectors of the RdDM pathway 97 Table 6. List of primers used for valiation of microarray results 97 Table 7. The Arabidopsis ago4-1 mutant of is susceptible to CaMV 120 vii Abbreviations RNAi RNA interference PIWI P-element induced wimpy testis RISC RNA induced silencing complex AGO argonaute PTGS posttranscriptional gene silencing TGS transcriptional gene silencing RITS RNA-induced transcriptional silencing RdRP RNA dependent RNA polymerase PAZ Domain found in PIWI, ARGONAUTE, ZWILLE proteins DDH Asp Asp His SUP SUPERMAN (SUP) CMT3 CHROMOMETHYLASE 3 SUVH4 SU(VAR)3-9-RELATED4 KYP KRYPTONITE NRPD1a subunit of the RNA Polymerase IV and V complexes NRPD1b subunit of the RNA Polymerase IV and V complexes RDR2 RNA Dependent RNA Polymerase 2 DCL3 DICER3 DRD1 Defective in RNA-Directed DNA Methylation1 DRM2 Domains Rearranged Methyltransferase2 RdDM RNA-directed DNA methylation DNMT1 DNA (cytosine-5-)-methyltransferase 1 viii DNMT3a/b DNA (cytosine-5-)-methyltransferase 3a/b MET1 Methyltransferase 1 HEN1 Hua enhancer1 CLSY1 CLASSY1 SSH suppression subtractive hybridization AFLP cDNA-amplified fragment length polymorphisms TUs transcriptional units AGI Arabidopsis Genome Initiative PDR pathogen-derived resistance H3K9 histone H3 lysine 9 PFBV Pelargonium flower break virus CaMV Cauliflower mosaic virus ix Chapter One Introduction RNAi RNA interference (RNAi) is a system within living cells that helps to control which genes are active and how active they are. Based on parsimony-based phylogenetic analysis, the most recent common ancestor of all eukaryotes most likely already possessed an early RNA interference pathway; the absence of the pathway in certain eukaryotes is thought to be a derived characteristic (Cerutti, H 2006). This ancestral RNAi system probably contained at least one Dicer-like protein, one Argonaute, one PIWI protein, and an RNA-dependent RNA polymerase that may have also played other cellular roles. The ancestral function of the RNAi system is generally agreed to have been immune defense against genetic elements such as transposable elements and viral genomes (Cerutti, H 2006). Related functions such as histone modification may have already been present in the ancestor of modern eukaryotes, although other functions such as regulation of development by miRNAs are thought to have evolved later. In 2006, Andrew Fire and Craig C. Mello shared the Nobel Prize in Physiology or Medicine for their work on RNA interference in the nematode worm C. elegans (Fire, A 1998). From a cellular mechanism perspective, RNAi is an RNA-dependent gene silencing process that is controlled by two RNA cleavage enzymes, DICER and the RNAi-induced silencing complex (RISC) as well as small RNAs which provide specificity to RISC activities (Tolia, NH 2007). This pathway is initiated by the enzyme 1 Dicer, which cleaves double-stranded RNA (dsRNA) molecules into short fragments of ~20 nucleotides. One of the two strands of each fragment, known as the guide strand, is then incorporated into the RISC, while the other is degraded. The most well-studied outcome of post-transcriptional gene silencing, is that when the guide strand base pairs with a complementary sequence of a messenger RNA molecule, it induces cleavage by Argonaute, the catalytic component of the RISC complex. Interestingly, this process is known to spread systemically throughout the organism (Klahre, U 2002). However, the interaction of sRNAs with Argonaute proteins need not always result in cleavage of the target RNAs but may also result in translational arrest or lead to inhibition of transcription (see below). Binding of the short double-stranded RNA molecules the catalytic RISC component Argonaute is a key step in RNAi (Hutvagner, G 2008) When the dsRNA is exogenous (coming from infection by a virus with an RNA genome or via laboratory manipulations, e.g., transgenes), the RNA is either transported into the cytoplasm (in the case of transgenes), or introduced there (in the case of most viral infections) and cleaved to short fragments by the enzyme dicer. The initiating dsRNA can also be endogenous (naturally originating within the cell), as in pri-microRNAs expressed from RNA-coding genes in the genome. The primary transcripts from such genes are first processed to form the characteristic stem-loop structure of pre-miRNA in the nucleus, then exported to the cytoplasm to be cleaved by dicer. Thus, the two dsRNA pathways, exogenous and endogenous, converge at the RISC complex. (Fig. 1). RNAi is also called RNA silencing and can be categorized in at least two different ways (Brodersen, P 2006). Based on their small RNA origin (sRNA), RNAi pathways 2 are divided into either the siRNA- or the miRNA-silencing pathway. siRNA and miRNA are incorporated into related RISCs, termed siRISC and miRISC, respectively (Tang, G 2005).