Complex Tertiary Structure Governs RNA Editing of the Drosophila Paralytic Transcript by Leila E. Rieder B.A., Reed College, Portland, OR, 2006 M.A., Brown University, Providence, RI, 2011 A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in The Division of Biology and Medicine, The Department of Molecular Biology, Biochemistry, and Cellular Biology at Brown University PROVIDENCE, RHODE ISLAND May 2013 c Copyright 2013 by Leila E. Rieder ! This dissertation by Leila E. Rieder is accepted in its present form by The Division of Biology and Medicine, The Department of Molecular Biology, Biochemistry, and Cellular Biology as satisfying the dissertation requirement for the degree of Doctor of Philosophy. Date Dr. Robert Reenan, Advisor Recommended to the Graduate Council Date Dr. Michael McKeown, Reader (Chairman) Date Dr. Erica Larschan, Reader Date Dr. Eric Morrow, Reader Date Dr. Ronald Emeson, Outside Reader Approved by the Graduate Council Date Dr. Peter Weber, Dean of the Graduate School iii Vitae Leila Elizabeth Rieder Brown University Providence, RI 02912 518-428-2877 or leila.rieder@gmail.com Education Expected Doctor of Philosophy, Dr. Robert Reenan, PI May, 2013 Cellular Biology, Molecular Biology, and Biochemistry Brown University, Providence, Rhode Island 02912. May, 2011 Master of Arts, Dr. Robert Reenan, PI Cellular Biology, Molecular Biology, and Biochemistry Brown University, Providence, Rhode Island 02912. 2007-2008 Post-graduate classes Harvard University Extension, Cambridge, Massachusetts 02138. May, 2006 Bachelor of Arts, Biology Reed College, Portland, Oregon 97202. Additional Research Experience 2008-2009 PhD research, Dr. Michael Sussman, PI, University of Wisconsin-Madison, 53706. 2006-2008 Associate Scientist I, Molecular Profiling, Biogen Idec, Inc., Cambridge, Massachusetts 02142. summers Research Assistant, Marine Biological Laboratory, 2000-03, ’05 Woods Hole, Massachusetts 02543. Publications LE Rieder, CJ Staber, B Hoopengardener, and RA Reenan. 2013. Tertiary structural elements determine extent and specificity of messenger RNA editing. In press, Nature Communications. iv Savva, YA*, LE Rieder*, and RA Reenan. 2012. The ADAR Protein Family. Genome Biology 13(12):252-32. (*Equal contributors) Rieder, LE and RA Reenan. 2012. The Intricate Relationship between RNA Structure, Editing, and Splicing. Seminars in Cell and Developmental Biology 23(3): 281-8. Allaire NE, Rieder LE, Bienkowska J, Carulli JP. 2008. Experimental Compari- son and Cross-Validation of Affymetrix HT Plate and Cartridge Array Gene Expression Platforms. Genomics 92(5): 359-65. Rieder, LE and AF Mensinger. 2001. Strategies for Increasing Growth of Juvenile Toadfish. Biological Bulletin 201(2): 283-5. Grants, Awards, and Honors Sept., 2011 Poster prize, Brown University Cellular Biology, Molecular Biology and Biochemistry Department annual retreat. April, 2011 Poster prize, EMBO Workshop on chromatin structure, organization, and dynamics, Prague, CZ. 2009-2012 National Science Foundation Graduate Research Fellowship Award. 2008-2009 Molecular Biosciences Training Grant, University of Wisconsin-Madison Cellular Biology, Molecular Biology, and Biochemistry Department. 2005-2006 Howard Hughes Undergraduate Research Grants, Reed College. 2003-2006 Commendation for excellence in scholarship, Reed College. Teaching Experience 2009-2013 Mentor to various Brown undergraduate students. summer, Teaching Assistant, Genome Explorers, 2012 HHMI undergraduate summer course, Brown University. 2011 Teaching certificate recipient, Brown University Sheridan Center for Teaching and Learning. 2010 Teaching Assistant, Plant Organism, Brown University. 2006-2008 Teaching Assistant, Community Lab, Biogen Idec Inc., (middle and high school students). 2004-2006 Science Center Tutor, Organic and Introductory Chemistry, Reed College. v 2004-2006 Personal Tutor, Organic and Introductory Chemistry, Reed College. 2003-2006 Teaching Assistant, Introductory Chemistry Lab, Reed College. Professional Presentations Jan., 2013 Poster presenter, Gordon Conference on RNA Editing, Galveston TX. Jan., 2013 Poster presenter, Gordon-Kenan Research Seminar on RNA Editing, Galveston TX. Sept., 2012 Poster presenter, EMBO conference, Nice France. May, 2012 Selected speaker, RNA Society Annual Meeting, Ann Arbor IL. Sept., 2012 Selected speaker, Brown University Molecular Biology, Cellular Biology and Biochemistry Department annual retreat, Providence RI. April, 2011 Poster presenter, EMBO Workshop on chromatin structure, organization, and dynamics, Prague CZ. *Poster prize recipient. April, 2011 Co-chair, RNA Processing session, EMBO Workshop on chromatin structure, organization, and dynamics, Prague CZ. Feb., 2011 Speaker, Brown University Molecular Biology, Cellular Biology, and Biochemistry Department recruitment weekend, Providence RI. Jan., 2011 Poster presenter, Gordon Conference on RNA Editing, Galveston TX. Jan., 2011 Poster presenter, Gordon-Kenan Research Seminar on RNA Editing, Galveston TX. vi Dedication Both by Mark Nowogrodzki RNA editing The Fruit Fly In order to keep its cells in line, Sing praise for the humble fruit fly: Healthy, without aberrations, though with a tiny brain, A body constantly keeps a fine it teaches us the ”what” and ”why” Check on its genes’ equation. of nature’s vast domain. h h It monitors its DNA Whether Darwin or creation, And finds where corrections are due, whether great design or chance, And edits its genes in its own way it’s many generations Their balance to pursue. our knowledge help advance. h h That’s what my granddaughter’s quests Oh, what a humbling feeling embrace: that reason does defy, What are the body’s controls? that nature’s highest being Where and how does the edit take place, learns wisdom from a fly... And what are nature’s goals? h She found substitutions in the RNA chain Take place at particular spots, In which the pattern of the genes’ domain Is changed into loops and knots. h And, though it may earn her a degree, And maybe fame and glory, The puzzle still appears to be The same eternal story: h Are these peculiarities, strange and stark, Genetic imperfections, Or just a signal, to embark On editing corrections? h This is the problem that ever imbued The human intellect, And that through history was pursued: What’s cause and what’s effect? vii Acknowledgements I owe so many thanks to the people who have helped me become a proud, confident scientist, and who have supported me through the dark times as well as the bright. Thanks to: ... Dr. Allen Mensinger at the MBL and Dr. Keith Karoly at Reed College. The former shaped me into a young scientist and the latter turned me into a molecular biologist and geneticist. ... my unwavering support system in Madison, Wisconsin: Kelly April, Patricia Lindquist, Jessica Ciomperlik, and Rachel Nelson-Rodrigues. You kept me safe through the darkest time of my life, so that I could arrive whole and new on the other side. To Dr. Michael Sussman, a selfless PI who supported my transition to Brown without question. ... the University of Wisconsin-Madison CMB department, who wished me good luck without judgment, and the Brown University MCB department who welcomed me home. ... the friends I found at Brown when I arrived here lost and unsure. Together we suffered the ups and downs of graduate school. At each new challenge I drew strength from Rebecca Helm, Asli Sahin, Marcela Soruco, Selena Gell, and so many others. viii ... my lab mates throughout the years, without whom I would be lost. Dr. James Jepson, Dr. Yiannis Savva, Dr. Selena Gell, soon-to-be-Drs. Asli Sahin and Yao-Jen Chang. To the Brown undergraduate students, with whom I converse as colleagues and of whom I could not be more proud, Abby Kearson and Kasia Sierzputowska. To the Larschan lab members who provided my home away from home. ... Dr. Rob Reenan, who took a chance on me. I have never met a scientist more excited about research. To my academic committee and the professors, students, and staff in the Brown MCB department who collectively nurtured my professional scientific confidence. ... my family. My parents, Conly and Susan, who saw me through it all with just the right mix of caution, concern, and support. To my sister Ray, my personal inspiration and the bigger half of my heart. To Matt, whose strength of character never ceases to amaze me, and whose mere presence in my life is enough to shape me into a better person. To my grandpa, Mark Nowogrodzki, who catches all my typos and who effortlessly imbues meaning into poems that remind me to remember the beauty of science. ix Contents Vitae iv Dedication vii Acknowledgments viii 1 Introduction to RNA Editing 1 1.1 ADARs and A-to-I RNA editing . . . . . . . . . . . . . . . . . . . . . 3 1.1.1 ADARs: characteristic structural features . . . . . . . . . . . 6 1.1.2 ADARs: localization and function . . . . . . . . . . . . . . . . 8 1.1.3 Advantages of RNA editing . . . . . . . . . . . . . . . . . . . 9 1.1.4 Organismal impacts of individual RNA editing events . . . . . 12 1.2 RNA editing substrates . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.2.1 Promiscuous editing of perfectly duplex RNA substrates . . . 14 1.2.2 Imperfectly duplex editing substrates direct specific editing events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 1.2.3 Regulation of siRNA/miRNA biogenesis and function through RNA editing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 1.3 ADAR-substrate recognition . . . . . . . . . . . . . . . . . . . . . . . 21 1.4 The complex relationship between RNA splicing and editing . . . . . 25 1.4.1 Coupling of editing and splicing . . . . . . . . . . . . . . . . . 25 1.4.2 Evidence that editing precedes splicing . . . . . . . . . . . . . 27 1.4.3 Rare post-splicing editing events are possible . . . . . . . . . . 30 1.5 Future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 1.5.1 Frontiers in ADAR biology . . . . . . . . . . . . . . . . . . . . 32 1.5.2 Directions in RNA post-transcriptional processing . . . . . . . 34 1.6 Voltage-gated sodium channels . . . . . . . . . . . . . . . . . . . . . . 36 1.6.1 Sodium channel structure and function . . . . . . . . . . . . . 37 1.6.2 Sodium channelopathies . . . . . . . . . . . . . . . . . . . . . 38 1.6.3 Paralytic is the only sodium channel gene in Drosophila . . . . 39 1.6.4 Mechanisms of temperature-sensitive paralysis . . . . . . . . . 42 x 1.7 Splicing in paralytic is affected by a mutation in the Maleless RNA helicase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 1.8 Using Homologous recombination to engineer the Drosophila genome at the paralytic locus . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 2 Tertiary structural elements determine extent and specificity of messenger RNA editing 50 2.1 RNA editing of paralytic . . . . . . . . . . . . . . . . . . . . . . . . . 52 2.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 2.2.1 Evolutionary conservation and RNA structural predictions for an RNA editing site . . . . . . . . . . . . . . . . . . . . . . . 55 2.2.2 Knock-in mutagenesis demonstrates the requirement for dsRNA directing RNA editing in vivo . . . . . . . . . . . . . . . . . . 57 2.2.3 Structural occlusion of splicing signals modulates RNA editing 62 2.2.4 The hairpin structure is required for selective editing of one adenosine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 2.2.5 A long-range tertiary pseudoknot is mediated by the hairpin . 71 2.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 2.3.1 Complex and novel structural requirements for RNA editing . 72 2.3.2 Modulation of RNA editing by a structure separate from the central duplex . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 2.3.3 RNA editing specificity conferred by a tertiary pseudoknot . . 80 2.3.4 Broader implications for RNA editing . . . . . . . . . . . . . . 82 2.4 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 3 Temperature-dependent modulation of editing via RNA “thermis- tors” 89 3.1 RNA as cis-acting “molecular sensor” . . . . . . . . . . . . . . . . . . 91 3.1.1 Riboswitches . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 3.1.2 RNA thermometers: temperature, RNA structure, and gene regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 3.1.3 RNA “thermistors,” putative thermo-sensitive effectors of RNA editing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 3.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 3.2.1 Temperature changes affect editing of specific transcripts . . . 99 3.2.2 Conservation of editing responsiveness to temperature in Drosophil- idae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 3.2.3 The effect of temperature on global RNA editing trends . . . . 103 3.2.4 HR-generated RNA structural mutants . . . . . . . . . . . . . 104 3.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 3.4 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 4 Mlenap−ts , an RNA helicase, interacts with the paralytic transcript 112 4.1 The DExD/H box RNA helicase family . . . . . . . . . . . . . . . . . 114 4.1.1 Dual functionality of the Drosophila Maleless helicase . . . . . 115 xi 4.1.2 A model for Mlenap−ts -para interaction . . . . . . . . . . . . . 115 4.2 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 117 4.2.1 The Mlenap−ts helicase is rescued by the presence of the loxP in the paralytic pre-mRNA . . . . . . . . . . . . . . . . . . . . 117 4.2.2 Transcription Activator-Like Effector Nucleases (TALENs) . . 118 4.2.3 TALEN design . . . . . . . . . . . . . . . . . . . . . . . . . . 120 4.2.4 TALEN vector construction . . . . . . . . . . . . . . . . . . . 122 4.3 Future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 4.4 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 5 Conclusions and Future Directions 128 5.0.1 RNA editing by ADARs . . . . . . . . . . . . . . . . . . . . . 129 5.0.2 RNA editing of the paralytic transcript . . . . . . . . . . . . . 129 5.0.3 The complex relationship between editing and splicing . . . . 130 5.0.4 Temperature effects on RNA editing . . . . . . . . . . . . . . 131 5.0.5 Sodium channel epilepsy mutations . . . . . . . . . . . . . . . 132 A A new hammer in the homologous recombination toolbox: Cre- mediated trans recombination 134 B Temperature-responsiveness of individual editing sites 138 xii List of Tables 2.1 Specific paralytic primer sequences. . . . . . . . . . . . . . . . . . . . 87 4.1 Primers used in the synthesis of the loxP-TALEN EGFP vector. . . . 124 4.2 Primers used in the mlenap−ts splicing analysis. . . . . . . . . . . . . . 126 xiii List of Figures 1.1 Hydrolytic deamination. . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.2 ADAR architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.3 ADAR catalytic domain . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.4 ADAR localization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.5 Adenosine-to-inosine RNA editing may recode transcripts . . . . . . . 10 1.6 Promiscuous and specific RNA editing . . . . . . . . . . . . . . . . . 14 1.7 Diverse RNA structures that direct editing. . . . . . . . . . . . . . . 17 1.8 Editing of small RNAs. . . . . . . . . . . . . . . . . . . . . . . . . . . 21 1.9 Proximity of editing sites to intron-exon junctions. . . . . . . . . . . . 23 1.10 Editing affects splicing. . . . . . . . . . . . . . . . . . . . . . . . . . . 27 1.11 Consequences of specific versus non-specific editing . . . . . . . . . . 33 1.12 Voltage-gated sodium channel structure. . . . . . . . . . . . . . . . . 38 1.13 The Drosophila paralytic sodium channel locus. . . . . . . . . . . . . 41 1.14 Sodium current in response to temperature and channel abundance. . 42 1.15 Third instar larvae polytene chromosome staining of the wild type Mle vs. the Mlenap−ts mutant helicase. . . . . . . . . . . . . . . . . . 44 1.16 A model for the interaction of the Mlenap−ts helicase with the para transcript. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 1.17 An overview of the homologous recombination process . . . . . . . . . 47 2.1 Organization of para exon 19 editing sites and ADAR dependence. . . 54 2.2 Paralytic exon 19 editing throughout Drosophilidae . . . . . . . . . . 56 2.3 Genomic sequence conservation of paralytic intron 19 throughout Drosophil- idae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 2.4 Nucleotide resolution of paralytic intronic cis element conservation . . 59 2.5 Editing mutations introduced by homologous recombination (HR) demon- strate necessity of central duplex . . . . . . . . . . . . . . . . . . . . 61 2.6 Mutations in the donor site complementary sequence (DCS) effect editing and splicing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 2.7 Consequences of aberrant splicing from the DCS zip allele . . . . . . 65 2.8 Phenotypic consequences of aberrant splicing from the DCS zip allele 67 2.9 Editing at site 1, within the central duplex, requires the hairpin element 69 2.10 Hairpin conservation and editing consequences across Drosophilidae . 70 2.11 The group II self-splicing intron α/α’ kissing loop interaction . . . . . 71 2.12 The hairpin is involved in a tertiary pseudoknot that directs selective editing in the central duplex, and can be functionally replaced . . . . 73 xiv 2.13 Representative electropherograms from all HR-generated mutant Drosophila lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 2.14 Editing and structure in outgroup species . . . . . . . . . . . . . . . . 77 2.15 Diurnal and mating behavior in editing extremes. . . . . . . . . . . . 79 2.16 A model for paralytic pre-mRNA editing and splicing . . . . . . . . . 81 2.17 Specific P values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 2.18 Drosophilidae species abbreviations and P values . . . . . . . . . . . 85 3.1 RNA sensor families. . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 3.2 Progression of the Mixed Peak Image Analysis program. . . . . . . . 97 3.3 Accuracy of the Mixed Peak Image Analysis program. . . . . . . . . . 98 3.4 Temperature affects editing in specific transcripts. . . . . . . . . . . . 100 3.5 Phylogeny of Drosophila species used in this study. . . . . . . . . . . 101 3.6 A comparison of temperature responsiveness of select editing sites across Drosophilidae. . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 3.7 Global effects of temperature on editing levels, independently ranked. 103 3.8 Global effects of temperature on editing levels, ranked at 20◦ C. . . . . 104 3.9 Temperature effects of HR-generated mutations in paralytic. . . . . . 105 4.1 Model for Mlenap−ts interaction with para structural mutants. . . . . 116 4.2 The loxP suppresses the skipping phenotype in a mlenap−ts background.118 4.3 The loxP remnant may form a perfect duplex in the paralytic transcript.119 4.4 Structure of a Transcription Activator-Like Effector Nuclease. . . . . 120 4.5 Targeting the loxP for disruption using TALENs. . . . . . . . . . . . 121 4.6 The loxP remnant and TALEN targeting sequence. . . . . . . . . . . 122 4.7 Composition of the TALEN plasmid. . . . . . . . . . . . . . . . . . . 123 4.8 Proposed TALEN crossing scheme. . . . . . . . . . . . . . . . . . . . 125 4.9 TALEN targeting of the loxP will generate animals with various de- grees of loxP integrity. . . . . . . . . . . . . . . . . . . . . . . . . . . 125 5.1 Steric relationship between residues affected by RNA editing and epilepsy mutations in the sodium channel. . . . . . . . . . . . . . . . 132 A.1 Cre-mediated trans recombination. . . . . . . . . . . . . . . . . . . . 136 B.1 Editing response to temperature of 55 individual Drosophila editing sites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 xv Chapter One Introduction to RNA Editing 2 Parts of this chapter have been published as: Rieder, LE, and RA Reenan. (2011) The intricate relationship between RNA struc- ture, editing, and splicing. Seminars in Cell and Developmental Biology. May; 23(3):28-8. I wrote the manuscript and crafted the figures. Dr. Robert Reenan provided editorial feedback. and Savva, YA*, Rieder, LE*, and RA Reenan. (2012) The ADAR protein family. Genome Biology. Dec; 13(12):252-8. (*Denotes equal contributors) Dr. Yiannis Savva and I share first authorship and jointly wrote the manuscript. I crafted the figures with help from Dr. Robert Reenan and Dr. James Jepson. Dr. Robert Reenan also provided editorial feedback. 3 1.1 ADARs and A-to-I RNA editing The human genome is predicted to encode well over a million distinct, functional proteins [Nielsen et al., 2006], despite containing a mere 20,000-25,000 protein-coding genes [Nekrutenko, 2004]. Highly sophisticated post-transcriptional modifications of messenger RNAs lead to increased proteome diversity while limiting a genome’s size and buffering it from mutation [Jin et al., 2007, Tian et al., 2008, Yang et al., 2008]. These enhancing processes include alternative splicing [Schmucker et al., 2000], alternative polyadenylation [Xing and Li, 2009], and RNA editing [Reenan et al., 2000, Pullirsch and Jantsch, 2010]. Unlike splicing, which can facilitate inclusion or exclusion of large segments of coding sequence, adenosine-to-inosine (A-to-I) RNA editing generally chemically alters only a few specific nucleotides within the coding regions of a transcript, providing a mechanism to subtly reshape the properties of resulting proteins. While editing causes the modification of transcripts at the level of single nucleotides, collectively these events are essential for viability in some organisms and correct neurological function in others [Higuchi et al., 2000, Nishikura et al., 2000, Tonkin and Bass, 2003, Palladino et al., 2000]. Although the global effect of loss of editing has been demonstrated in several model organisms, we have yet to fully uncover the functional importance of such seemingly small individual recoding events. RNA editing was first discovered 25 years ago in Xenopus oocytes: injected antisense RNAs failed to form stable duplexes in vivo, suggesting an as-yet then uncharacterized unwinding activity [Rebagliati and Melton, 1987, Bass and Weintraub, 1987]. This activity was later attributed to the enzyme eventually named ADAR (Adenosine Deaminase acting on RNA) [Bass, 1997], and the unwinding was due to the conversion of select adenosines (A) into inosines (I) via hydrolytic deamination (Figure 1.1), destabilizing the duplexes as A-U Watson-Crick base pairs convert to I-U wobble pairs [Polson et al., 1991]. Because the properties of inosine mimic those of 4 guanosine (inosine will form two hydrogen bonds with cytosine, Figure 1.1), inosine is recognized as guanosine by the translational cellular machinery [Basillo et al., 1962]. Adenosine-to-inosine (A-to-I) RNA editing, therefore, effectively changes the primary sequence of RNA targets. "#! !"# "#+ + ! ! ! ! ! !" 123()*'& + ! ! ! ! !"# !", $%&'()*'& *'()*'& -./$'()*'&0 Figure 1.1: Hydrolytic deamination. A-to-I RNA editing involves the hydrolytic deamination of an adenosine into an inosine molecule. Inosine forms two hydrogen bonds with cytosine, resembling guanosine to the cellular machinery. ADAR enzymes are highly conserved in metazoa [Bass, 2002], although the number of genes and isoforms varies between species. Mammalian genomes encode three ADARs: ADAR1, and ADAR2, which are catalytically active [Nishikura, 2010], and ADAR3, which is thought to be catalytically inactive [Chen et al., 2000]. The C. elegans genome encodes two genes, CeADR1 and CeADR2 [Tonkin et al., 2002], while only a single adar locus is present in the Drosophila genome [Palladino et al., 2000] (Figure 1.2A). The squid [Palavicini et al., 2009] and hydra (personal observation, R. A. Reenan) genomes each encode a single adar locus. Furthermore, ADAR genes are also present in the genomes of both sea urchin and sea anemone, suggesting an early origin of RNA editing enzymes in metazoan evolution [Jin et al., 2009]. In contrast, ADAR genes do not appear to be present in fungal, plant and yeast genomes [Jin et al., 2009]. Interestingly, although prokaryotic genomes do not contain ADAR genes, 5 they do encode a transfer RNA (tRNA) adenosine deaminase (TadA), which modifies specific tRNAs [Wolf et al., 2002]. Eukaryotic orthologs of this RNA editing enzyme, adenosine deaminases acting on tRNAs (ADATs), are also conserved in metazoa and catalyze the deamination of specific adenosines to inosines at or adjacent to the tRNA anticodon [Gerber and Keller, 2001]. Sequence homology between the catalytic domains of ADARs and ADATs suggests a model in which tRNA modifying enzymes are ancestral to ADARs (Figure 1.2B). In this model, a duplicate ADAT gene acquired one or more dsRNA binding domains that allowed the protein to recognize and bind dsRNAs. This novel gene is thought to have conferred selective advantage due to repair of detrimental genomic mutations by A-to-I modifications at the mRNA level [Bass, 2002]. Z Z dsRBM dsRBM dsRBM Deaminase hADAR1 hADAR2 100 aa hADAR2 hADAR3 sqADAR2 sqADAR2a dADAR sqADAR2b hADAR3 ceADAR1 hADAR1 ceADAR2 hmADAR dADAR ceADAR2 hmADAR hADAT hADAT dADAT (a) dADAT (b) Figure 1.2: (a) Domain architecture of metazoan ADARs. The deaminase domain is depicted in blue, while the dsRBMs are shown in orange and Z-DNA binding domains, unique to human ADAR1, are presented in green. The human genome contains three ADAR genes (hADAR1-3). That of the squid, Loligo pealeii, contains an ADAR2-like gene (sqADAR2) that produces variants (a and b) through alternative splicing. C. elegans contains two genes (ceADAR1-2), while the genome of D. melanogaster encodes for only one (dADAR), an enzyme homologous to hADAR2. Although the dsRBM found in the Hydra magnapapillata genome are highly divergent, five such motifs are recognizable in hmADAR, the only identified gene in this family. Human and Drosophila ADAT architectures are included (red), as these enzymes are believe to be ancestral to present-day ADARs. (b) Cladogram based on ADAR catalytic domain sequences. MacVector was used to generate a relatedness tree based on the protein sequences of ADAR catalytic domains from different species. C. elegans ADAR2 is absent due to difficulty aligning the catalytic domain. Note that human and Drosophila ADATs (red) cluster as the outgroup. Reproduced from Savva et al. [2012]. In the decades of research following the initial discovery, RNA editing has 6 been proven to exert significant impact on proteomes and organisms. Mice deficient for ADAR1 have a heterozygous embryonic lethal phenotype [Nishikura et al., 2000]; those with complete lack of ADAR2 are seizure-prone and die soon after birth [Higuchi et al., 2000]. Drosophila null for dADAR, exhibit signs of severe neurological impairment, including locomotor uncoordination, temperature-sensitive paralysis and seizures [Palladino et al., 2000], while in C. elegans, both functional ADAR genes are required for normal chemotaxis [Tonkin et al., 2002]. Clearly the conserved process of A-to-I RNA editing is crucial for proper neurological function. Interestingly, genetically engineered Drosophila dADAR hypomorphs, which produce only 20% of dADAR protein compared to wild type animals, display alterations in complex behavior such as courtship, but not in general locomotor coordination [Jepson et al., 2011]. While RNA editing may serve different purposes in various organisms, ADARs are conserved among metazoa and share certain characteristics: between one and three tandem double-stranded RNA binding motifs (dsRBMs) and a C-terminal catalytic deamination domain. 1.1.1 ADARs: characteristic structural features ADAR enzymes share a common domain architecture consisting of a variable number of N-terminal dsRNA binding domains (dsRBD) and a C-terminal catalytic deaminase domain [Bass, 2002] (Figure 1.2 A). Human ADARs (hADARs) possess two or three dsRBDs, while the C. elegans enzymes (ceADARs) contain one or two. The single Drosophila ADAR (dADAR) contains two dsRBDs, similar to the mammalian ADAR2. In squid two ADAR enzymes (sqADARs) are generated via splicing from a single adar locus: while isoform 2b contains two dsRBDs, the inclusion of an alternative exon leads to the generation of an ADAR enzyme (sqADAR2a) containing an additional dsRBD, which confers higher enzymatic activity in vitro [Palavicini et al., 2009]. Interestingly, recent evidence suggests that this extra dsRBD of sqADAR2a is required 7 for appropriate RNA editing in the high salt conditions of a marine environment [Palavicini et al., 2012], suggesting a link between the regulation of the RNA editing process and changes in the physical environment. In agreement with this, additional evidence suggests that RNA editing may be regulated by temperature [Garrett and Rosenthal, 2012, Savva et al., 2012] (please see Chapter 3). Finally, the Hydra magnapapillata ADAR enzyme (hmADAR) contains five identifiable dsRBDs (Figure 1.2 A) (personal observation, R. A. Reenan). The crystal structure of the human ADAR2 deaminase domain (Figure 1.3) suggests that a catalytic center forms in the enzyme active site: a glutamic acid residue (E396) forms hydrogen bonds with a water molecule, while a histidine (H394) and two cysteine residues (C451 and C516) coordinate a zinc ion that activates the water molecule for nucleophilic attack (1.1) [Macbeth et al., 2005]. Buried in the catalytic core, an inositol hexakisphosphate (IP6 ) molecule stabilizes multiple arginine and lysine residues and is also required for catalytic activity [Macbeth et al., 2005]. (a) Top (b) Side Figure 1.3: Crystal structure of the human ADAR2 deaminase domain shown fro the top (a) and side (b). The catalytic core is formed between H394 (red), E396 (blue), C451 (orange), and C516 (gray). The core also includes a water molecule, zinc ion, and IP6 molecule (not shown). Reproduced from Savva et al. [2012]. 8 1.1.2 ADARs: localization and function The three mammalian adar genes give rise to four known isoforms: ADAR1p150, ADAR1p110, ADAR2, and ADAR3. ADAR1 variants and ADAR2 are expressed in many tissues, whereas the ADAR3 protein is only present in the brain [Chen et al., 2000, Melcher et al., 1996]. ADAR3 is thought to be catalytically inactive [Chen et al., 2000], but it is possible that it competes with ADARs 1 and 2 for RNA binding substrates, altering the editing profile through that mechanism. Alternative promoter usage within the ADAR1 transcript generates a full length (ADAR1p150) and a N-terminally truncated (ADAR1p110) isoform [Patterson and Samuel, 1995], both of which contain three dsRNA-binding domains and the deaminase domain. The ADAR1 isoforms differ in their cellular distributions: ADAR1p110 exclusively localizes to the nucleus [Desterro et al., 2003, Sansam et al., 2003], while ADAR1p150 shuttles in and out of the nucleus, although it accumulates in the cytoplasm [Patterson and Samuel, 1995, Poulsen et al., 2001]. The cytoplasmic localization of ADAR1p150 suggests that a select class of dsRNAs could be targeted outside the nucleus as mature mRNAs. However, cytoplasmic localization of other ADARs is unusual. Mammalian ADAR1p110 and ADAR2 are also present in the nucleolus and this localization is dependent on functional dsRBDs [Desterro et al., 2003, Sansam et al., 2003]. ADAR nucleolar localization may be mediated by dsRNAs generated from ribosomal RNAs (rRNAs) and small nucleolar RNAs (snoRNAs), both of which are concentrated within the nucleolus. However, to date there is no evidence that ADAR modification of rRNAs or snoRNAs occurs, thus, the significance of ADAR nucleolar localization remains enigmatic. Yet translocation of ADAR2 from the nucleolus to the nucleoplasm upon expression induction of a known editing target resulted broadly in higher editing levels in pre-mRNA substrates [Sansam et al., 2003]. Therefore, the nucleolar localization of ADARs may represent a mechanism 9 for regulating editing of pre-mRNAs through enzyme sequestration. In Drosophila, Jepson et al. [2011] examined the complete expression pattern and cellular distribution of dADAR using a genetically engineered allele containing an HA epitope tag inserted into the endogenous locus. Visualization of dADAR-HA revealed that Drosophila ADAR is predominately expressed in the nervous system [Jepson et al., 2011]. Furthermore, both transgenic dADAR specifically expressed in the salivary glands of 3rd instar larvae, as well as endogenous dADAR FIB tagged dADAR, dADAR localize FIB within the nucleus and accumulate within the nucleolus [Savva et al.,DAPI 2012] (Figure 1.4), consistent with observations regarding mammalian ADARs. (a) (b) (b) dADAR LAMIN dADAR dADAR LAMIN dADAR LAMIN LAMIN DAPI dADAR FIB dADAR dADAR FIB dADAR FIB FIB DAPI Figure 1.4: ADAR localization. (a) Transgenic HA-tagged ADAR (green) localizes within the nuclear envelope (lamin, red) and more specifically to the nucleolus (fibrillarin, red) in Drosophila salivary gland cells. (b) Endogenous HA-tagged ADAR (green) localizes to the Drosophila neuronal nucleus and colocalizes with the nucleolus, distinguished by the red fibrillarin signal (arrows). Reproduced from Savva et al. [2012]. 1.1.3 Advantages of RNA editing Adenosine-to-inosine RNA editing results in alteration of base pairing properties; inosine forms two hydrogen bonds with cytosine (Figure 1.1) and is recognized as guanosine by cellular machineries tasked with interpreting nucleic acid information, such as the translational system. Editing, therefore, effectively recodes RNA through A-to-G substitutions. These changes may have obvious impact in that editing has the 10 potential to recode stop codons (nonsense) into that of tryptophan (sense), resulting in read-through. Alternatively, editing can change key amino acid side chains (Figure 1.5) in conserved and functionally critical portions of the protein products. 2nd base U C A G UUU Phe UCU Ser UAU Tyr UGU Cys U UUC Phe UUA Leu UCC Ser UCA Ser UAC Tyr UAA STOP UGC Cys UGA STOP UUG Leu UCG Ser UAG STOP UGG Trp 1st base CUU Leu CCU Pro CAU His CGU Arg C CUC Leu CUA Leu CCC Pro CCA Pro CAC His CAA Gln CGC Arg CGA Arg CUG Leu CCG Pro CAG Gln CGG Arg AUU Ile ACU Thr AAU Asn AGU Ser nonpolar A AUC Ile AUA Ile ACC Thr ACA Thr AAC Asn AAA Lys AGC Ser AGA Arg AUG Met ACG Thr AAG Lys AGG Arg polar GUU Val GCU Ala GAU Asp GGU Gly basic G GUC Val GUA Val GCC Ala GCA Ala GAC Asp GAA Glu GGC Gly GGA Gly GUG Val GCG Ala GAG Glu GGG Gly acidic Figure 1.5: Adenosine-to-inosine RNA editing may recode transcripts. Recoding events may be silent (synonymous) or change amino acid residues (arrows; non-synonymous). Stop codons (black) may be edited to the tryptophan codon, resulting in read-through. Reproduced from Rieder and Reenan [2011]. What is the purpose of a system such as RNA editing versus simply fixing mutations at the DNA level? In fact, the effects of editing are more akin to genetic mutation than to alternative splicing, since the nature of editing alters the genetic code at the level of a single nucleic acid. Editing sites are sometimes hardwired as guanosine in closely related species, a phenomenon shown in both vertebrates [Li et al., 2009] and invertebrates [Yang et al., 2008, Tian et al., 2008], suggesting mutation at the DNA level is a viable alternative in select cases. Similar to alternative splicing, editing enables an organism (or even individual cells) to achieve genetic variation while safeguarding the genome from deleterious mutations [Gommans et al., 2009]. 11 Additionally, post-transcriptional events such as splicing and editing diversify the protein repertoire [Hanrahan et al., 2000] allowing for flexibility in protein properties and possible adaptation of novel protein functions. It is possible that cells such as neurons are particularly individualized through transcriptome diversification, at the levels of both transcription and transcript modifi- cation. ADARs are most highly expressed in the nervous system [Jepson et al., 2011] throughout development [Jacobs et al., 2009], although the presence of the protein does not necessarily cause high levels of editing. Rather, editing efficiency appears low in the embryonic mouse brain and gradually increases during development, imply- ing that a regulatory network carefully controls editing during development despite ADAR concentration [Wahlstedt et al., 2009]. The observation that ADARs modify transcripts implicated in fast chemical and electrical neurotransmission [Hoopen- gardner et al., 2003] has led to suggestions that editing contributed to the evolution of complex metazoa [Jin et al., 2009] and to higher cognitive function in primates [Paz-Yaacov et al., 2010, Eisenberg et al., 2005]. Transcripts for proteins involved in synaptic signaling, including those of ion channels and synaptic release proteins, are heavily edited [Hoopengardner et al., 2003, Hanrahan et al., 2000], and even singular editing events can substantially impact channel properties [Bhalla et al., 2004]. For example, the human potassium channel Kv 1.1 transcript is edited at a single adenosine. This one editing event causes a valine substitution for an isoleucine in the highly conserved ion-conducting pore, resulting in the effective removal of a single methyl group from the edited residue. Despite this subtle structural alteration, recoding of the transcript by RNA editing profoundly affects channel inactivation [Bhalla et al., 2004, Decher et al., 2010]. Similarly, the Drosophila Shaker potassium channel is edited at four locations, resulting in 16 possible channel isoforms. Each variant possesses subtly different biophysical properties. Furthermore, examination of multiply edited combinations revealed an intriguing degree of functional epistasis 12 between even distant editing sites [Ingleby et al., 2009]. It is evident that the complex interaction of multiple editing sites within a transcript confers the potential for functional fine-tuning. Editing can serve to increase proteomic diversity and produce variation within a cell population. While the significance of this process at the molecular level may appear subtle, the effect of single events at the organismal level can be quite striking. 1.1.4 Organismal impacts of individual RNA editing events Recoding through RNA editing can produce variation in a protein population if some transcripts are edited while others reflect the literal genome. Yet certain transcripts are edited at a very high level, similarly impacting every resulting protein product. For example, the mouse GluR-B transcript, which encodes an AMPA receptor subunit, is edited to nearly 100% at one adenosine, replacing the genomically encoded glutamine with an arginine (Q/R site). Mice deficient for ADAR2 exhibit post-natal mortality, usually due to fatal seizure events, but substitution of the arginine form of GluR-B for the unedited form completely rescues this phenotype [Higuchi et al., 2000]. Editing of this transcript, therefore, is absolutely necessary for the organism’s viability. In addition to changing key residues (Figure 1.5) in essential proteins, RNA editing is known to contribute to cellular immunity. The presence of dsRNA in the cytoplasm, as with viral infection, triggers host defense responses including editing by mammalian cytoplasmic ADAR1. In many cases editing of a viral RNA genome results in loss or retardation of pathogenicity [Samuel, 2011]. Yet this process has also been co-opted by viruses as a method of regulating complex replication cycles. The hepatitis delta virus possesses a single-stranded RNA genome encoding a lone protein, HDAg. However the viral anti-genome folds into a duplex structure and directs the cellular RNA editing machinery to edit an amber stop codon in the HDAg transcript, resulting in read-through and production of a longer isoform of the viral 13 protein. The presence of HDAg-Long is necessary to switch the viral replication cycle from genome replication to packaging [Casey, 2011]. The replication of other viruses including HIV-1 [Doria et al., 2009] and the measles virus [Ward et al., 2011] is also enhanced by ADAR1 editing [Gelinas et al., 2011]. While the impact of editing events involved in temporal control of viral replication or recoding of nearly all GluR-B transcripts is apparent, other editing events may have subtler effects, such as those localized to introns or to the third nucleotides of codons (Figure 1.5). Considerable evidence exists, however, that even seemingly insignificant editing events may have substantial impact through effect on RNA structural stability, splicing, and even global cellular responses such as RNA interference (RNAi). It is clear that editing is not a purely random force in the transcriptome. As with splicing, the RNA itself encodes the highly conserved informational signals that direct ADARs. 1.2 RNA editing substrates Early observations suggested that ADARs recognize only RNA substrates possessing duplex character with no activity detected on single-stranded RNA molecules [Bass and Weintraub, 1987]. This observation provoked two related questions: how do RNA substrates direct ADAR activity and how does the enzyme recognize target substrates? In vitro data revealed that specific deamination of single adenosines usually occurs on short imperfect duplex RNA, while long, highly stable, perfectly complementary duplexes are non-specifically, or promiscuously, modified at up to 50% of adenosine residues [Polson et al., 1991, Nishikura et al., 1991] (Figure 1.6 A). 14 A. Alu elements B. GluR-2 Q/R I A A I A I A A A I A I A I A I A I I C. Synaptotagmin-I I I I Figure 1.6: Promiscuous and specific RNA editing. Perfect duplex RNA structures are edited non-specifically (promiscuously), while short imperfect duplexes, with bulges, mismatches, and loops, are edited specifically. (A) Primate-specific Alu elements (orange) in reverse orientation may form long (approximately 300 basepairs) nearly perfect duplex editing substrates, which are edited by ADARs promiscuously. (B) The mammalian GluR-2 Q/R site is formed between an exon (blue) and downstream intron (gray). This imperfectly duplex substrate directs editing around a single nucleotide that results in a glutamine to arginine substitution in the AMPA receptor protein. (C) Part of the Drosophila synaptotagmin-1 transcript forms a complex pseudoknot between an exon (blue) and following intron (gray) that directs precise deamination at two sites. Reproduced from Rieder and Reenan [2011]. 1.2.1 Promiscuous editing of perfectly duplex RNA substrates Numerous efforts have recently attempted to identify the complete inosinomes [Wulff et al., 2011] of various organisms through deep sequence identification of the edited transcript population [Rosenberg et al., 2011, Bahn et al., 2011] or via other novel techniques [Li et al., 2009, Xia et al., 2005, Eisenberg, 2011]. Such endeavors reveal that a significant number of editing events occur in non-coding regions of the transcriptome–a phenomenon that spans metazoa, including primates, C. elegans [Wu et al., 2011, Morse et al., 2002] and Drosophila [Graveley et al., 2011]. The most 15 well studied example of non-coding ADAR activity is the promiscuous editing of primate-specific Alu repetitive elements, short interspersed elements of approximately 300 nucleotides, which inhabit up to 75% of human genes and comprise more than 10% of the human genome [Kim et al., 2004]. Given the prolificacy of Alu elements and their recent expansion in the human genome lineage, new insertions have frequently occurred in regions already occupied by other local Alu elements. The insertion of one element into the genome in reverse orientation to another nearby Alu can result in the formation of a highly complementary, stable, duplex structure in the transcript that may serve as an ADAR substrate (Figure 1.6 A). These elements, often found in gene-rich, more highly transcribed regions, lead to nearly perfect double-stranded RNA substrates in vivo and are heavily edited by ADAR [Athanasiadis et al., 2004, Paz-Yaacov et al., 2010, Levanon et al., 2004] (Figure 1.6 2A). In fact, due to Alu substrates, RNA editing in humans is at least an order of magnitude higher than that in mouse, Drosophila, or C. elegans [Eisenberg et al., 2005]. The purpose of promiscuous editing is unknown, although edited Alu-contain- ing transcripts can be retained in the nucleus [Zhang and Carmichael, 2001]. Also, since inverted Alu elements so readily form double-stranded structures, they may also serve as Dicer substrates to generate Alu-specific small interfering RNAs that guide a subsequent RNA interference response. An Alu duplex in one transcript could mediate degradation of other Alu-containing transcripts via the RNA-Induced Silencing Complex (RISC). For example, a pair of inverted Alus in the 3’ UTR of a GFP reporter mRNA strongly repress GFP expression in vivo, though a single element has no effect on the fluorescence signal [Chen et al., 2008]. Promiscuous editing of the Alu duplex would result in unwinding of the structure and loss of Dicer recognition, followed by alleviation of silencing. In this manner RNA editing could impact RNAi-based silencing and transcriptional regulation of many Alu-containing 16 genes. Indeed, RNAi is inhibited in vitro when target transcripts are first deaminated by ADAR2 [Scadden and Smith, 2001]. It is also interesting to note that aberrant chemotaxis phenotypes caused by C. elegans ADAR mutations [Tonkin et al., 2002] are rescued by deficiencies in various components of the RNAi pathway [Tonkin and Bass, 2003], suggesting that the modification activity of ADARs may unwind in vivo perfect dsRNA substrates capable of initiating an RNAi response through Dicer activity [Wu et al., 2011]. Promiscuous editing may also play a role in viral defense via interfacing with the RNAi pathway. Viruses with dsRNA genomes, for example, provide a perfect duplex substrate for promiscuous editing. These edited dsRNA viral genomes may be targeted for destruction: dsRNA with multiple I-U mismatches is cleaved by Tudor Staphylococcal Nuclease [Scadden, 2005], a component of the RISC. This targeting function appears to be conserved: mammalian ADAR1, ADAR2, and Drosophila dADAR are all capable of generating the preferred cleavage site when acting on dsRNA in vitro [Scadden, 2005]. Genome cleavage has an direct impact on viral pathogenicity. 1.2.2 Imperfectly duplex editing substrates direct specific editing events Perfectly duplex substrates are edited promiscuously, while imperfect structures containing bulges, mismatches, and loops, are edited much more selectively. Spe- cific editing targets are often identified due to A/G disparities between the cDNA sequence and that of the reference genome. To date, all confirmed ADAR substrates possess secondary structure around the region of editing; that is, the early-identified requirement of double-strandedness for editing has thus far not been violated. The secondary and tertiary structures of RNA around the editing site(s) in an imperfect RNA duplex serve to specifically direct editing of select adenosines. Also, it is rare 17 for inosine to replace adenosine in 100% of transcripts at these sites. Rather, levels of editing appear to be precisely modulated, can be evolutionarily conserved, and are dictated by RNA structure [Bratt and Ohman, 2003] (Figure 1.6 B, C). Imperfectly duplex double-stranded editing substrates are sometimes gen- erated within exonic sequences when the nascent transcript folds, creating simple hairpins [Bhalla et al., 2004, Hanrahan et al., 2000, Keegan et al., 2005, Ohlson et al., 2007]. Yet, more often intronic sequences with extensive complementarity to upstream or downstream exons containing the targeted adenosine(s) can pair with the exonic sequence to form structurally diverse RNA duplexes. These intronic cis elements, also known as editing site complementary sequences (ECS), can form simple exon-intron hairpin structures [Burns et al., 1997, Hanrahan et al., 2000, Higuchi et al., 1993, Lomeli et al., 1994, Wang et al., 2000] or more complex RNA secondary structures such as pseudoknots [Reenan, 2005] (Figure 1.7). Drosophila synaptotagmin-I sites C and D 5’ 3’ Mammalian GluR-2 R/G site 5’ 3’ Figure 1.7: Diverse RNA structures that direct editing. The complex pseudoknot of Drosophila synaptotagmin-1 is presented in contrast with the simple hairpin of mammalian GluR-2, both specific editing targets. Exons are represented in blue, introns in black. Adenosines targeted by ADAR are red. Reproduced from Savva et al. [2012]. The majority of ECS-mediated editing substrates usually contain mismatches, bulges, and loops. It is thought that the presence of such structural imperfections directs ADARs to specific locations on the RNA duplex without the requirement of primary sequence recognition. However, upon binding, ADARs do exhibit sequence 18 preferences for modifying select adenosines over others. In vitro studies using synthetic editing substrates revealed that ADAR enzymes preferentially target adenosines neighbored by a 5’ uridine, while adenosines with a 5’ guanosine neighbor are exceedingly rare [Lehmann and Bass, 2000, Polson and Bass, 1994]. In addition, adenosines found in mismatches with cytosines are edited more often when compared to other adenosines [Wong et al., 2001]. Specific RNA editing often leads to transcript recoding. Because inosine shares the base pairing properties of guanosine, the translational machinery interprets edited inosine as guanosine (Figure 1.1), altering the triplet codon, which can result in amino acid substitutions in protein products. Over half the triplet codons in the genetic code may be reassigned through RNA editing [Rieder and Reenan, 2011] (Figure 1.5). Due to the degeneracy of the genetic code, RNA editing can cause both silent and non-synonymous amino acid substitutions. However, statistically, RNA editing predominantly results in non-synonymous changes [Graveley et al., 2011] and thus favors the diversification of protein products. While the functional consequences of most specific recoding events are currently unknown, myriad studies in diverse model organisms indicate that specific RNA editing of certain mRNAs can result in profound changes in protein function [Bhalla et al., 2004, Dingledine et al., 1999, Ingleby et al., 2009, Rosenthal and Bezanilla, 2002, Seeburg et al., 2001]. As an example of RNA structure, the exonic GluR-2 Q/R site forms an imperfect RNA duplex by pairing with sequences in the downstream intron (Figure 1.6 B). Single nucleotide changes in the structure in either exonic or intronic sequence alter output editing levels of a mini-gene reporter at the Q/R site. This suggests that it is the precise RNA structure that finely tunes editing levels at this location [Higuchi et al., 1993]. As with the GluR-2 Q/R site, many target adenosines are encompassed in secondary structures composed of an exon and cis-acting elements present in a nearby intron. While most intronic sequence is under little selective pressure, 19 cis elements that participate in editing site duplex formation exhibit substantially higher sequence conservation compared to the surrounding intron. These observations allow putative identification of elements that participate in editing [Ingleby et al., 2009, Reenan et al., 2000]. Similarly, comparative genomics provides a useful tool to identify unknown editing sites based on conservation of nearby cis elements [Sixsmith and Reenan, 2007]. Prior to widespread use of deep sequencing technology comparative genomic approaches more than doubled the number of confirmed editing sites in Drosophila [Hoopengardner et al., 2003]. Although many cis-acting elements are localized in introns proximal to the edited exon, these sequences may also regulate editing from substantial distances. Structures that direct editing may be relatively simple, such as an imper- fectly base-paired hairpin [Stefl et al., 2010], but they may also be considerably more elaborate, as exemplified by the Drosophila synaptotagmin-1 editing target (Figure 1.6 C). Synaptotagmin-1 mRNA, which encodes the calcium sensor for fast neurotransmitter release at the synapse, forms arguably the most complex editing- directing RNA structure described in the literature. Editing sites C and D of the synaptotagmin-1 transcript are encompassed in a long-range exon-intron structure of over 1500 nucleotides. This structure appears to form an intricate pseudoknot, the exact conformation of which serves to precisely regulate editing levels [Reenan, 2005]. 1.2.3 Regulation of siRNA/miRNA biogenesis and function through RNA editing Deamination of adenosines found in non-coding regions influences the biogenesis and target recognition of small interfering RNAs (siRNAs) involved in the RNA interference (RNAi) pathway [Nishikura, 2006]. The biogenesis of siRNAs requires processing of long dsRNA precursors into 21-23nt RNA duplexes, a task carried out 20 by the Dicer class of RNase III-like ribonuclease enzymes [Bernstein et al., 2001]. The mature siRNAs generated by Dicer are unwound for the incorporation of a single-stranded RNA (guide RNA) into the RNA-induced silencing complex (RISC). Ultimately, the targeting of RISC by the small guide RNA initiates transcriptional and post-transcriptional sequence-specific silencing [Hannon, 2002]. Growing evidence supports the interaction between the editing and RNAi pathways. Since the RNA editing and RNAi pathways both involve dsRNAs, editing may antagonize the RNAi pathway at two levels. First, adenosine deamination can alter the perfectly duplex structure of dsRNA substrates, leading to poor Dicer processing and thus decreased siRNA concentrations (Figure 1.8). Second, RNA editing of siRNAs may prevent base pairing of the RISC with the original mRNA targets, inhibiting the cleavage step of RNAi [Nishikura, 2006]. Similarly, editing may change the target of an siRNA by altering the primary sequence of the guide RNA. siRNA biogenesis is progressively inhibited with increased RNA editing in vitro [Scadden and Smith, 2001, Scadden and Smith, 2001]. However, the in vivo functional consequences of the antagonism between editing and RNAi pathways are still unclear. Similar to the siRNA pathway, micro RNAs (miRNAs) are highly conserved, genomically encoded 20-25nt RNAs that mediate gene silencing and regulate diverse cellular processes including development, differentiation, and apoptosis [Bartel, 2004, Meister et al., 2004]. This class of small interfering RNAs is encoded within non-coding regions of the genome and form hairpin primary miRNA structures (pri-miRNAs), which are extensively processed to generate the mature miRNA. In their double- stranded state, these pri-miRNAs can serve as ADAR substrates (Figure 1.8). Indeed, several miRNA precursors undergo A-to-I RNA editing at specific adenosines [Blow et al., 2006, Luciano et al., 2004]. Editing of these precursors can inhibit further cleavage and processing [Yang et al., 2006], as well as regulation of 21 siRNA precursor pri-miRNA A AA A A A AAAA I II I I I pre I I I I fewer functional mature miRNAs toward edited siRNAs siRNAs alternative targets Figure 1.8: Editing of small RNAs. Editing can interfere with siRNA and micro RNA production and targeting. Perfectly duplex siRNA precursors are targets for hyper-editing by ADARs. Editing may result in improper Dicer processing and fewer functional siRNAs, or edited siRNAs. Primary (pri) microRNAs, imperfect duplexes, may be targets for specific editing, resulting in mature miRNAs toward alternative mRNA targets. Reproduced from Savva et al. [2012]. gene expression. A single edited adenosine within the 6nt targeting seed region of the miRNA is sufficient to redirect silencing to a new target [Kawahara et al., 2007] (Figure 1.8). In addition, A-to-I RNA editing is highly enriched at 3’ UTRs within miRNA targets, suggesting a regulatory role for RNA editing in translation [Gu et al., 2012]. More importantly, RNA editing can inhibit or initiate miRNA/target interactions via modification at important positions where complementarity is required for appropriate miRNA targeting. 1.3 ADAR-substrate recognition As discussed above, particular aspects of RNA structure serve to direct ADAR to edit specific adenosines. But how does the ADAR enzyme recognize substrates that vary in sequence almost as widely as the genome itself? No sequential or higher- order structure has yet been identified that is shared by all editing targets, which confounds the search for new inosinome candidates. Although ADARs recognize certain structures and precisely target adenosines in diverse transcripts, the only 22 feature common to all editing sites is a double stranded character. The binding of an ADAR enzyme to an RNA substrate requires direct contact between the dsRBDs and dsRNA substrates [Valente and Nishikura, 2007]. Nuclear magnetic resonance and X-ray crystallography studies reveal that the binding of double stranded RNA binding proteins (dsRBPs) in general to dsRNA relies less on sequence than on structure [Bycroft et al., 1995, Nanduri et al., 1998, Ryter and Schultz, 1998]. Because dsRNA adopts an A-helical structure, which contains a deep and narrow major groove, dsRNA binding proteins are thought to recognize substrates through indirect readout and RNA structure [Draper, 1995]. Nuclear magnetic resonance data obtained from the dsRBMs of rat ADAR2 in solution with a simple hairpin substrate revealed unexpected structure-dependent discrimination by the individual motifs. In complex with the 71 nucleotide imperfect duplex encompassing the GluR-2 R/G editing site, dsRBM1 recognizes the conserved pentaloop at the apex of the hairpin structure, while dsRBM2 recognizes two bulged bases adjacent to the singular editing site [Stefl et al., 2006] (Figure 1.9 3A). However, these structural features are specific to the GluR-2 R/G site and are not common to all editing targets. Consistent with the binding properties of other dsRBPs, ADARs can bind to any dsRNA without sequence specificity [Nishikura et al., 1991]. It has been widely assumed in the field that the dsRBDs of ADAR enzymes confer structural specificity during specific editing, determining which substrates are targeted and which adenosines are edited. The same group that reported structural recognition of the GluR-2 R/G site investigated this issue further and proposed a solution structure of the ADAR2 dsRBMs in complex with the GluR-2 R/G site that indicated sequence- specific recognition. Their data suggested that the ADAR2 dsRBMs contact the RNA minor groove, rather than the ribose backbone, allowing reading of the primary sequence [Stefl et al., 2010]. Two amino acid side chains contributed from each 23 A. GluR-2 R/G B. Adar2 intron 4 I II I II C. Drosophila dAdar exon 7 D. Paralytic exon 19 I I I I E. Paralytic exon 9 I Figure 1.9: Proximity of editing sites to intron-exon junctions. Editing sites are often found near exon-intron (blue-gray) boundaries, but structures that direct editing can be formed from entirely exonic sequence (blue). (A) The mammalian GluR-2 transcript R/G site is located just one nucleotide from the intron-exon boundary. (B) The RNA structure that directs editing of mammalian ADAR2 is entirely intronic. Editing at one site (arrow) creates an alternative AG 3’ splice acceptor site, leading to exonization of part of intron 4 (green). (C) Drosophila dAdar is edited at one adenosine in exon 7, substituting a glycine for the genomically-encoded serine. The cis elements that form this structure are included within the edited exon. (D) The Drosophila sodium channel transcript, paralytic, is edited at three adenosines in exon 19, located within 70 nucleotides of the downstream splice donor site. (E) Paralytic exon 9 is edited at a single site based on a structure formed from exonic sequence. Reproduced from Rieder and Reenan [2011]. dsRBM make sequence-specific contacts: each motif contacts a guanosine and an adenosine 9 or 10 nucleotides apart. In vitro studies revealed that the footprint of each dsRBM encompass 11-16 nucleotides of the 71 nucleotide hairpin GluR-2 substrate [Ohman et al., 2000]. However, it is not simply the dsRBMs of ADAR enzymes that provide substrate recognition; the C-terminal catalytic domain also contributes to editing specificity. Surprisingly, swapping the deamination domains of mammalian ADAR1 and ADAR2 reveals that these domains play the dominant role in determining editing specificity of the resulting chimeric proteins [Wong et al., 2001]. It is possible that the major function of the dsRBMs is to dimerize ADARs rather than contribute 24 to substrate specificity [Cho et al., 2003, Gallo et al., 2003], a view supported by the observation that ADAR1 and ADAR2 form homodimers independent of RNA substrate binding [Valente and Nishikura, 2007]. Both the amino terminal and dsRBM1 are required for dimerization of Drosophila dADAR on RNA, and point mutations in dsRBM1 are capable of abolishing both dimerization and editing activity. Nevertheless, while the amino terminus is required for activity and dimerization, dADAR peptides lacking this domain still recognize RNA substrates. This suggests that dADAR monomers bind individually and gain catalytic activity post-dimerization [Gallo et al., 2003]. Coupled editing sites are found on the same side of the RNA helix, indicating that only a single catalytic center remains active following dimerization [Enstero et al., 2009]. Yet recent in vitro work by Eggington et al. suggest that mammalian ADAR1 and ADAR2 are both able to edit when lacking N-terminal dsRBDs. Not only do these truncated enzymes effectively edit in vitro, but they are also able to selectively deaminate the same adenosines targeted by the full-length enzyme, although to different levels. The dsRBDs may function as RNA anchors [Stephens et al., 2004], but these data suggest that ADAR specificity is predominantly conferred, not by dsRBDs, but by the deaminase domain itself [Eggington et al., 2011]. However, experiments in an in vivo setting are required for the biological relevance of this observation to be understood. Regardless of how ADARs recognize target substrates, editing seemingly must be accomplished before the splicing machinery recognizes local intron-exon junctions, removes introns, and abolishes the double-stranded structures that direct ADARs. 25 1.4 The complex relationship between RNA splic- ing and editing Alternative splicing and RNA editing are both post-transcriptional modifications that diversify the resulting protein populations. These processes do not occur in isolation from each other; rather, editing and splicing are believed to be coupled, at the very least because both exonic and intronic sequences logically comprise most editing substrates in nascent transcripts. It is also thought that editing and splicing compete for access to the transcript and that one process contributes to the fine-tuning of the other. Further, recent high-throughput sequencing experiments revealed that co-transcriptional editing is widespread in Drosophila [Rodriguez et al., 2012]. 1.4.1 Coupling of editing and splicing Splicing is believed to occur co-transcriptionally [Goldstrohm et al., 2001] and recent evidence corroborates that spliceosomal assembly is required for proper transcription termination [Martins et al., 2011]. The RNA Polymerase II (Pol II) C-Terminal Domain (CTD) is thought to coordinate RNA processing events, including editing and splicing [Munoz et al., 2010]. Perhaps the most well known example of the interaction between editing and splicing is the editing by mammalian ADAR2 of its own transcript in intron 4 (Figure 1.9 3B). Editing of the Adar2 RNA alters an AA into an AI (recognized as AG), creating a new 3’ splice site (Figure 1.10). Edited forms of the transcript may then be alternatively spliced, and use of the new 3’ acceptor causes a change to the open reading frame and a truncated protein [Rueter et al., 1999]. The Pol II CTD appears to coordinate editing and splicing of Adar2 [Laurencikiene et al., 2006], suggesting that these post-transcriptional events may be coupled as a more general mechanism. The regulation of adar2 transcript splicing via auto-editing represents a negative regulatory feedback mechanism to 26 modulate ADAR2 protein levels since inhibition of auto-editing in vivo increases ADAR2 expression and editing at many target adenosines in the transcriptome [Feng et al., 2006]. The transcript of the Drosophila homolog of mammalian ADAR2, dADAR, is also edited and spliced to create isoforms of different target specificity. Transcripts that contain alternative exon 3a due to recognition of a non-canonical 5’ splice site also lack editing at the single target site in exon 7 [Keegan et al., 2005] (Figure 1.9 3C). The expression of dADAR isoforms with different specificity and activity is developmentally regulated and tissue-dependent [Marcucci et al., 2009], suggesting global editing is controlled in part by splicing of the dAdar transcript. Adenosines targeted for editing are disproportionately localized near splice junctions in the pre-mRNA [Rieder and Reenan, 2011]. Therefore, during formation of a dsRNA ADAR substrate, intronic cis-acting sequences can form RNA duplexes encompassing splicing sites and potentially obscuring them from the splicing machinery (Figure 1.7). Furthermore, through modification of select adenosines, ADARs can create or eliminate splicing sites, broadly affecting later splicing of the transcript. Similar to the translational machinery, the spliceosome interprets inosine as guanosine, and therefore, a canonical GU 5’ donor site and AG 3’ acceptor site can be created via the deamination of AU (IU = GU) and AA (AI = AG), respectively [Valente and Nishikura, 2005]. Correspondingly, RNA editing can destroy a canonical AG 3’ acceptor site (IG = GG) [Valente and Nishikura, 2005]. Yet even editing at sites that do not encode splicing signals can affect splicing. The mammalian GluR-2 transcript, for example, is edited at two non-synonymous sites: the Q/R and R/G sites, named for the residue changes resulting from editing. When the Q/R site, located in exon 11, as well as a downstream intronic hot spot, is edited, splicing is enhanced between exons 11 and 12 (Figure 1.10). In the same transcript, the R/G site is located just one nucleotide from the boundary between 27 exon 13 and the downstream intron (Figure 1.7). When this site is edited, splicing favors inclusion of exon 15 over that of exon 14 (Figure 1.10). The mechanism of how these particular editing sites affect splicing of the GluR-2 transcript is still unknown [Schoft et al., 2007]. Mammalian ADAR2 TTACAA 47 nt TTACAG Mammalian GluR-2 Q>R R>G A A A 11 12 13 14 15 16 Figure 1.10: Editing affects splicing. Mammalian ADAR2 auto-edits its own transcript, creating a novel splice site (red), which results in the inclusion of 47 nucleotides (yellow) and a frameshift in the coding sequence. In the mammalian GluR-2 transcript, editing at both the Q/R site in exon 11 and an intronic hotspot (red) is required for efficient removal of the downstream intron. Editing of the R/G site (red) reduces efficacy of downstream splicing and favors an alternative final exon configuration (yellow). Reproduced from Savva et al. [2012]. Molecular data presented above indicates that editing and splicing are coor- dinated. Subcellular localization of ADARs also provides clues to the relationship between editing and splicing. Mouse ADAR2 is concentrated in the nucleolus [Sansam et al., 2003] and Drosophila dADAR is localized predominantly in the nucleus [Jepson et al., 2011]. Once a pre-mRNA has been spliced it is quickly shuttled from the nucleus to the cytoplasm for translation [Luo and Reed, 1999, Valencia et al., 2008]. Splicing and editing, therefore, are likely both compartmentalized in the nucleus. 1.4.2 Evidence that editing precedes splicing Support for the coupling of editing and splicing comes from the nature of the structures that direct editing. Often these duplexes are formed between the region of an exon that 28 encompasses the editing site(s) and a nearby intron [Reenan, 2005, Hanrahan et al., 2000, Bratt and Ohman, 2003] (Figure 1.6 B,C; Figure 1.9 A, D). These structures can only form if editing precedes splicing: splicing results in loss of double-strandedness, prohibiting ADAR substrate recognition. Editing sites and the cis elements that direct editing tend to occur near exon/intron borders [Reenan, 2005, Rueter et al., 1999, Higuchi et al., 1993] (Figure 1.9 A, B, D). Because RNA double-strandedness is required to direct ADAR, and editing sites are often located near splice junctions, splicing signals may be buried in secondary structures. However, these structures must be resolved in order for the intron-exon boundary to be accurately recognized by the splicing machinery. Conserved secondary structures are enriched at alternative splice sites [Shepard and Hertel, 2008], and encompassing the 3 or 5’ splice site in local structure prevents spliceosomal recognition [Buratti and Baralle, 2004]. What does this mean for the splicing signals of constitutive exons, for example those represented in Figure 1.6 B and C, and Figure 1.9 A and D, which are involved in secondary structures directing editing? Resolution of these structures may be key. Such a structure is predicted to direct editing at three sites in constitutive exon 19 of the Drosophila sodium channel gene paralytic. This structure also encompasses the downstream exon-intron boundary (Figure 1.9 D). A mutation in a dsRNA helicase, Maleless, appears to result in decreased recognition of the 5’ splice site (obscured due to inclusion in secondary structure) and skipping of the edited exon–a splicing catastrophe that results in fewer functional sodium channels and whole organism temperature-sensitive paralysis [Reenan et al., 2000]. This evidence suggests that the helicase mediates interplay between editing and splicing: after deamination by dADAR the Maleless helicase resolves the secondary structure to expose the splicing signals. The mutant version of Maleless is either unable to resolve the structure or fails to dissociate from the region, resulting in exon skipping (see Figure 1.16). In 29 the mutant Maleless background correctly spliced transcripts are edited half as often as in a wild type helicase background. This finding suggests that the transcripts that do not fold into the structure recognized by dADAR are also those that are more easily identified by the splicing machinery. This hypothesis also suggests that the processes of splicing and editing compete for access to the nascent transcript. While in some cases editing and splicing may be in direct competition, in others editing is known to direct splicing and therefore must occur first. The GluR-2 transcript is edited at two major recoding sites, R/G in exon 13 and Q/R in exon 11, and both are controlled by cis elements near the exon/intron borders (Figure 1.9 A and Figure 1.6 B, respectively). The Pol II CTD enhances editing at the R/G site by preventing premature splicing; yet editing of this site reduces splicing efficiency of the downstream intron and favors alternative splicing events [Schoft et al., 2007]. When the R/G region encompassing the 5’ splice site was fused to the 3’ splice site of an adenovirus reporter, editing and splicing were each found to interfere with the other in vitro [Bratt and Ohman, 2003]. Similar to the R/G site, editing at the Q/R site, which occurs at nearly 100%, is also facilitated by the CTD [Ryman et al., 2007]. Editing at both the Q/R site and an intronic hot spot is required for efficient splicing of the following intron (Figure 1.10). Further, when these sites are not edited the transcript is improperly spliced. Perhaps this serves as a quality control mechanism by ensuring that only edited transcripts mature and are exported to the cytoplasm for translation [Schoft et al., 2007]. Similar to the GluR-2 transcript, editing of Alu elements, though non-specific (Figure 1.6 A), directs future splicing events. Rather than favoring one discrete alternative exon over another, however, Alu editing can result in exonization of a previously non-coding sequence [Athanasiadis et al., 2004]. For example, the human nuclear prelamin A recognition factor transcript includes two Alus in reverse orientation separated by 25 nucleotides, which may form a nearly perfect duplex. 30 Editing in the downstream Alu results in generation of a new 3’ splice signal and alteration of splicing enhancers. In addition, a premature stop codon within the Alu is edited to that for tryptophan. Collectively these editing events can result in exonization of previously intronic sequence, a process that is regulated in a tissue- dependent manner [Lev-Maor et al., 2007]. While the significance of exonization by edited Alu elements is unmistakable, the relationship between splicing and editing may be considerably more opaque. Two Drosophila transcripts demonstrate correlation between splicing and editing across some distance. One encodes a calcium channel and the other a nicotinic acetylcholine receptor, and editing in each is directed by intronic cis signals. No correlation between editing and splicing is observed within the channel transcript, in which the edited exon is considerably upstream from the alternative splicing event. However, the alternate forms of the receptor transcript, in which the editing site is downstream from the alternative exons, display different editing efficiencies [Agrawal and Stormo, 2005]. These observations further suggest that editing precedes splicing and that this relationship may occur over considerable distances within the transcript. Editing, therefore, must occur prior to splicing in most transcripts, either because the intron contains necessary cis-acting elements or because editing regulates subsequent splicing. Yet not every ADAR substrate contains an intronic component (Figure 1.9 C, E), and it is possible that select transcripts are deaminated post-splicing or even in isolation from other post-transcriptional processes. 1.4.3 Rare post-splicing editing events are possible Several entirely exonic structures that direct editing have been identified. The human potassium channel gene Kv 1.1 is intronless, yet edited via a 114 bp imperfectly duplex hairpin entirely contained within coding sequence [Bhalla et al., 2004]. One of the many editing sites in the Drosophila sodium channel transcript, paralytic, is directed 31 by an exonic hairpin [Hanrahan et al., 2000] (Figure 1.9 E). The Drosophila dAdar transcript itself undergoes editing based on a structure formed within exon 7 [Keegan et al., 2005] (Figure 1.9 C), which changes the subnuclear organization and specificity of the resulting enzyme for editing targets [Savva et al., 2012]. Auto-editing of the dAdar transcript decreases overall deaminase activity of the enzyme [Keegan et al., 2005], suggesting a feedback mechanism to control global editing levels. In theory, intermolecular structures formed between two mature mRNAs could form double-stranded ADAR substrates. In such an example, one exonic sequence would act in trans as a regulatory element to direct editing of the other. This concept has even been explored in vitro and abstracted to the therapeutic realm: targeted ribo-oligonucleotides could direct ADAR to correct disease-related genomic mutations at the post-transcriptional level [Woolf et al., 1995], though controlling deamination of precise adenosines would be incredibly challenging. In addition, there is as yet no in vivo evidence of post-spliced intermolecular structures attracting ADAR, and no editing sites have been identified that lack a cis regulatory sequence. Although it is possible that entirely exonic structures directing editing endure in cytoplasmic post-spliced mature mRNA, ADARs are most often found localized to the nucleus, as discussed above. However, some observations suggest ADARs may be present, at least transiently, in the cytoplasm. Mammalian ADAR1, for example, which has overlapping specificity with ADAR2 [Lehmann and Bass, 2000], shuttles between the nucleus and cytoplasm [Eckmann et al., 2001]. Inflammatory responses can cause increased ADAR1 expression in the cytoplasm [Liu et al., 1997] and the presence of extranuclear mammalian ADARs appears to be regulated by interferon responses [Yang et al., 2008, Patterson and Samuel, 1995]. Mammalian ADAR3, though not known to have catalytic activity, contains a nuclear localization signal recognized by nuclear importins. Certain isoforms of ADAR2 contain an ADAR3-like signal, suggesting ADAR3 may regulate sub-cellular localization of 32 more enzymatically-active ADARs by competing for importins [Maas and Gommans, 2009]. Sub-cellular localization of ADAR2 is also dependent on the phosphorylation- dependent peptidyl-prolyl cis/trans isomerase Pin1. In the absence of Pin1, ADAR2 mislocalizes to the cytoplasm. Cytoplasmic ADAR2 is targeted for degradation by WWP2, an E3 ubiquitin ligase [Marcucci et al., 2011], and so is unlikely to participate in editing of extranuclear targets. Therefore, rare post-splicing or extranuclear editing may occur. If present, RNA editing activity in the cytoplasm could serve a distinct function. Splicing, while usually confined to the nucleus, also occurs in the cytoplasm of neuronal cell dendrites [Glanzer et al., 2005]. This suggests that pre-mRNA processing on-site may allow more rapid responses to synaptic signals. Cytoplasmic editing could serve a similar purpose, as editing targets are involved in chemical and electrical synaptic signaling. Post-transcriptional modification at the synapse would allow selection of the most appropriate protein isoforms and near-instant responses to changes in the dendritic environment. The relationship between cytoplasmic editing and splicing may occur much like nuclear events or be transcript-dependent. 1.5 Future directions 1.5.1 Frontiers in ADAR biology Inappropriate RNA editing is associated with suicidal depression [Dracheva et al., 2008], cancer [Galeano et al., 2011], and neurodegenerative diseases [Blow et al., 2004] including amyotrophic lateral sclerosis [Kawahara et al., 2004, Takuma et al., 1999]. Thus a better understanding of the ADAR protein family functions will shed light into therapeutic strategies regarding these and other nervous system disorders. ADARs modify specific adenosines to inosines in short imperfect dsRNA templates, a well-characterized role that usually leads to recoding of transcripts and 33 peptides (Figure 1.11 a). In contrast, ADAR modification of long, perfect duplexes results in the deamination of up to 50% of the adenosines [Cattaneo, 1994, Fischer et al., 2008, Peters et al., 2003] affecting siRNA biogenesis and activity (Figure 1.11 b). Although much progress has been achieved in ADAR biology over the last two decades, the consequences of such promiscuous RNA editing, especially in vivo, are not well understood.                           (b)                 long, perfect duplexes            (a)       short, imperfect duplexes { { I transposons I I I  siRNAs I I A A T N C AA RISC ? ? I Ile Met transcript recoding regulation of  regulation of chromatin gene expression Figure 1.11: Consequences of specific versus non-specific editing. (a) Short, imperfect duplexes, such as Drosophila synaptotagmin-1, are specifically edited leading to transcript and peptide recoding. (b) Long, perfect dsRNA substrates, including those formed by nearby transposons in opposite orientation (green), are hyper-edited, leading to fewer or edited siRNAs. This may alter gene expression through RISC (yellow) targeting, but evidence also links the RNAi pathway to chromatin regulation. Reproduced from Savva et al. [2012]. Several lines of evidence indicate that promiscuous editing by ADARs an- tagonizes RNAi-mediated gene silencing [Nishikura, 2006, Wu et al., 2011]. Editing of endogenous siRNAs in Drosophila [Kawamura et al., 2008] suggests that ADARs interact with endogenous RNAi pathways, which are involved in somatic defense against transposable elements [Chung et al., 2008, Czech et al., 2008, Ghildiyal et al., 2008, Kawamura et al., 2008] and in the regulation of chromatin states [Fagegaltier 34 et al., 2009] (Figure 1.11 b). The consequences, however, of editing/RNAi interactions and the resulting effects on both gene expression and chromatin regulation remain enigmatic. Future studies should aim in deciphering the in vivo consequences of such interactions, as this avenue highlights a more global role for ADARs regarding the broad regulation of neuronal transcriptomes. 1.5.2 Directions in RNA post-transcriptional processing The relationship between RNA editing and splicing is far from resolved and necessitates future study. Single nucleotide polymorphisms at the genetic level can result in differential splicing between individual humans, and may even impact gene expression [Lalonde et al., 2011]. Nucleotide changes in the transcript, for example, those caused by RNA editing, may have similar impacts on splicing and overall gene expression. In this way aberrant editing might lead to disease states. Over 50% of known disease- causing mutations affect splicing [Hammond and Wood, 2011], and the role of editing in these diseases has not been addressed. The most widely used therapy to correct mis-splicing diseases is targeting antisense oligonucleotides to the pre-mRNA [Hammond and Wood, 2011], creating secondary structure around the splicing area of interest. Although these therapeutic oligonucleotides are chemically altered to resist nucleases and improve functionality and are not chemically equivalent to ribo-oligonucleotides, such therapies could also affect RNA editing by changing transcript structure, especially as editing sites tend to occur near intron-exon junctions (Figures 1.6 and 1.9). Off-target effects will have to be considered when designing antisense oligonucleotide therapies. To more fully understand the relationship between splicing and editing, and effectively interpret the roles of such modifications in disease, the kinetic association between these processes should be explored to probe the model in which editing precedes splicing. For example, sequestering constitutive splice sites in synthetic RNA 35 secondary structure, as is observed for endogenous alternative splice sites [Shepard and Hertel, 2008], should allow more time for editing of nearby adenosines, resulting in increased editing or even editing of new target nucleotides. Decreasing RNA structure by reducing or altering intronic cis sequences predicted to obstruct splice sites should allow less time for ADAR to deaminate nearby, and so editing would be predicted to ablate or decrease. It would be most informative to conduct such experiments in vivo where the editing and splicing machinery exist in the correct stoichiometry and whole organismal effects may be observed. However, altering secondary structure around constitutive splice signals may also result in aberrant splicing [Goguel et al., 1993] and effects on behavior or viability of the animal. Performing these experiments in vivo requires precision and accuracy to preserve the pre-mRNA sequence except where mutation is intentionally introduced. Although most often employed in yeast, homologous recombination techniques may be used to introduce genetic mutations into Drosophila [Staber et al., 2011] and have already been successfully employed to alter RNA editing in this organism [Jepson et al., 2011, Savva et al., 2012]. The homologous recombination approach has numerous benefits over traditional reporter transgenic constructs. For example, it allows manipulation of the endogenous gene, transcribed at the biologically relevant level rather than an exogenous driver generating overexpression of a reporter or minigene construct. This is especially important considering editing targets are most often involved in electrical and chemical neurotransmission, and altering gene dosage could have dramatic viability effects. In addition, manipulation of the endogenous gene allows for behavioral studies not possible using overexpression or transgene systems. Investigating the interaction of editing and splicing in vivo, therefore, is disease-relevant and necessary for modeling the coordination between post-trans- criptional processes. In addition, new editing sites identified through transcriptome 36 deep sequencing and other novel techniques continue to expand known inosinomes, and each new adenosine brings insight into possible structural or sequential patterns shared by ADAR target sites. Further, ADAR itself may provide a therapeutic tool for genetic diseases. For example, fusing the ADAR catalytic domain to the bacteriophage λ N-peptide is sufficient to target the chimeric protein to dsRNA tagged with the boxB RNA recognition element. Swapping the bulged nucleotides in the target RNA structure alters editing specificity of the substrate (J. Rosenthal, personal communication). These results suggest that ADAR may provide a useful tool to combat genetic diseases at the RNA level. 1.6 Voltage-gated sodium channels Complex neurological functions require precise control of neuronal properties such as membrane permeability to sodium ions. Voltage-gated sodium channels, responsible for the rising phase of the action potential in the membranes of neurons and other electrically excitable cells, must be flexible and sensitive enough to account for changing cellular and environmental conditions. Yet how sodium channel flexibility is achieved differs between organisms. For example, the human genome encodes at least nine major sodium channel genes (SCN1A-SCN9A), which may have subtly different properties [Mantegazza et al., 2010]. The Drosophila genome, however, contains just one sodium channel gene, paralytic (para), but the para transcript may be altered post-transcriptionally through alternative splicing and editing to produce over two million different channel isoforms [Graveley, 2001]. 37 1.6.1 Sodium channel structure and function It appears that vertebrate and invertebrate sodium channels are derived from a single ancestor, possibly a calcium channel, and have evolved independently in these two branches [Goldin, 2002]. Voltage-gated sodium channels consist of a highly processed α subunit of 260 kDa that is sufficient for functional sodium current, and auxiliary β subunits of approximately 36 kDa that are required for gating and normal kinetics [Goldin, 2002]. The β subunits may also be required for α subcellular targeting [Mantegazza et al., 2010]. The human genome contains nine sodium channel α genes and four β genes (SCN1B -SCN4B ), while that of Drosophila contains only the single para functional α subunit and a potential β subunit, TipE [Feng et al., 1995, Hodges et al., 2002]. TipE is coexpressed with para in most cells and coimmunoprecipitates from Xenopus oocytes, suggesting physical association [Hodges et al., 2002]. Sodium channels are comprised of four homology domains (I-IV), each with six trans-membrane α helices (S1-S6; Figure 1.12). The channel is composed of a voltage-sensing domain, formed by S1-S4 and a pore-forming domain, consisting of S5 and S6. Voltage-gated sodium channels play a key role in neuronal excitability: in response to initial membrane depolarization, sodium channels open within a few hundred microseconds (activation), resulting in an inward Na+ current and further membrane depolarization. Channels subsequently inactivate in milliseconds, preventing additional ion flow. Inactivation, however, is often incomplete, leading to persistent Na+ current, which is known to be important to neuronal function [Mantegazza et al., 2010]. During channel gating, positively-charged residues on S4 move across the membrane in a “screw-helical” fashion, stabilized by negative charges in the S1 helix [Paldi and Gurevitz, 2010]. 38 DI D II D III D IV +++++ +++++ +++++ +++++ N C Figure 1.12: Voltage-gated sodium channel structure. The α sodium channel subunit, embedded in the cell membrane (blue), consist of four homology domains (DI-IV), each with six membrane- spanning helices (S1-S6). The fourth helices of each domain form the voltage sensor (+). The editing sites discussed in this dissertation (Chapter 2) are located on S1 of DIII (red star). α subunits are able to function autonomously, but their gating and localization may be affected by β subunits (not shown). 1.6.2 Sodium channelopathies Perturbations in sodium channel structure or dosage can have substantial effects on channel function, neuronal activity, and resulting behavioral phenotypes. Mutations in sodium channel genes are implicated in the pathophysiology of human disorders as diverse as epilepsy, ataxia, migraine, neuropathic pain, autism, and neurodegenerative diseases including multiple sclerosis [Mantegazza et al., 2010, Meisler and Kearney, 2005]. Identified epilepsy mutations, such as those involved in generalized epilepsy with febrile seizures plus (GEFS+) and severe myoclonic epilepsy of infancy (SMEI), are found primarily in SCN1A and localize throughout the protein without any clear relationship to functional channel domains [Meisler and Kearney, 2005]. SMEI mutations, which mostly arise de novo, may be truncation or missense, while all identified GEFS+ mutations are missense in highly-conserved amino acids. Many studies suggest that most epilepsy-generating mutations lead to loss of channel function and haploinsufficiency [Mantegazza et al., 2010]. Many of the most widely-used antiepileptic drugs, such as phenytoin, car- bamezepine, and lamotrigine, inhibit the function of voltage-gated sodium channels [Mantegazza et al., 2010]. While it appears curious that epilepsy, a disorder charac- 39 terized by brain hyperexcitability, should be caused by loss-of-function mutations in sodium channel genes, it appears that the main gene affected in epileptic patients, SCN1A, is most highly expressed in inhibitory neurons. This gene is also often affected in migraine patients, although mutations range from loss of function to gain of function [Mantegazza et al., 2010]. Sodium channel disorders may be aggravated or made apparent by increases in temperature. For example, epileptic attacks can occur in children due to increases in ambient or body temperature resulting from the immune response to infection or vaccination. Indeed, discovery of paralytic in 1971 was due to a temperature-sensitive mutation causing adult Drosophila paralysis [Suzuki et al., 1971], which suggested a nervous system defect. It was not until 1989 that para was determined to be the Drosopihila sodium channel homolog [Loughney et al., 1989, Ramaswami and Tanouye, 1989]. By then, napts , another mutation believed to affect sodium channel abundance had been discovered [Wu et al., 1978]. It took until 2000 to discover that napts is an allele of maleless [Reenan et al., 2000] (mle), an RNA helicase involved in Drosophila dosage compensation, that affects paralytic splicing and therefore channel abundance. This mutation is discussed below and is specifically addressed in Chapter 4. Additionally, human GEFS+ SCN1A epilepsy mutations knocked into Drosophila paralytic result in a temperature-induced seizure phenotype due to reduced inhibitory activity in the central nervous system [Sun et al., 2012]. 1.6.3 Paralytic is the only sodium channel gene in Drosophila Sodium channel mutations can lead to a variety of neurological syndromes and widespread behavioral phenotypes, but due to redundancy in the human genome molec- ular and phenotypic effects of isolated mutations are difficult to study. Drosophila paralytic presents an ideal model in which to investigate sodium channel processing and function, and is the subject of Chapter 2. Localization of the editing sites 40 studied in this chapter are represented in Figure 1.12 (protein) and Figure 1.13 (locus). While mammalian genomes encode nine functional α sodium channel subunits, most invertebrate genomes, including that of Drosophila, contain only one, suggesting that invertebrates and vertebrates use different means of generating diversity in sodium channel function [Goldin, 2002]. Additionally, although some electrophysiological differences have been attributed to alternate mammalian channel spliceoforms, it is not clear how much functional diversity is achieved through alternative splicing, and no RNA editing has been identified in any mammalian sodium channel transcript. This is in sharp contrast to invertebrate sodium channels, which are heavily edited and alternatively spliced to generate protein diversity. Additionally, although C. elegans contains only 302 neurons, its genome encodes many more ion channel genes than the genome of Drosophila, an organism with about 250,000 neurons, again suggesting that Drosophila must achieve channel diversity through non-genomice molecular strategies [Littleton and Ganetzky, 2000]. Insect sodium channel spliceoforms generate a broad range of voltage-dependent activation and inactivation properties [Tan et al., 2002], and two A-to-I RNA editing sites in the German cockroach BgNav sodium channel transcript lead to amino acid substitutions and altered gating properties [Song et al., 2004]. The paralytic locus, located on the X chromosome, is over 65 Kb long, and contains 26 constitutive exons [Loughney et al., 1989]. In addition to a myriad of mutually exclusive exons and alternative exons [Thackeray and Ganetzky, 1995], the transcript contains a dozen sites of RNA editing (Figure 1.13). Taken independently, these post-transcriptional modifications could generate over two million sodium channel isoforms [Hanrahan et al., 2000]. Evidence suggests only a fraction of these post-transcriptional modification combinations are utilized, and that the range and ratios of transcript isoforms change throughout development, among tissues, 41 and between sexes [Thackeray and Ganetzky, 1994, Ingleby et al., 2009]. Sodium channel spliceoforms have different characteristics in the membrane, such as varying voltage-dependent activation or inactivation [Olson et al., 2008, Lin et al., 2009], yet even minor sequence changes in the sodium channel transcript may translate to structural/functional changes, and provide flexibility in local or global neuronal membrane properties. A AAA A AAA A AA Figure 1.13: The Drosophila paralytic sodium channel locus. The entire paralytic locus is over 65 Kb. Constitutive exons are represented in blue, alternative exons in yellow, and introns as dotted lines. Editing sites are denoted by red “A”s. If all of the post-transcriptional modifications are considered separately, the single paralytic locus is able to give rise to over two million protein isoforms. The editing sites within the red box are the subject of Chapter 2. 42 1.6.4 Mechanisms of temperature-sensitive paralysis As discussed above, sodium channel mutant phenotypes are exacerbated by in- creased temperature, which often leads to temperature-sensitive paralysis in affected Drosophila. Nelson and Wyman used Drosophila paralytic to propose a synergistic paralysis mechanism, suggesting that temperature-sensitive mutations in this locus uncover the inherent temperature sensitivity in all neurons. As temperature increases, neuronal repolarization occurs earlier due to an increase in the rates of activation and inactivation for both sodium and potassium currents [Huxley, 1959] (Figure 1.14 A). This “normal” effect of temperature on the action potential is augmented when there are fewer functional sodium channels, as with napts (see below) and temperature-sensitive para mutations (parats , Figure 1.14 B), leading to paralysis when the sodium current fails to reach spiking threshold [Nelson and Wyman, 1990] (Figure 1.14 C). A. Wild-type sodium channels B. Constant temperature C. Temperature and channel abundance 39ºC para-ts 39ºC Na+ current into cell 25ºC para-ts spike threshold WT WT 25ºC Time Time Time Figure 1.14: Sodium current in response to temperature and channel abundance. A. At elevated temperatures (red), sodium channels more rapidly activate and inactivate compared to at cooler temperatures (blue). B. Parats and other mutations that affect channel abundance, reduce overall sodium current (dashed blue line). C. When Drosophila mutants with reduced sodium channel abundance are subjected to elevated ambient temperatures, a paralysis phenotype ensues due to the synergistic interaction between the normal effects of temperature on sodium channel current and reduced number of functional sodium channels because the sodium current (dashed red line) fails to reach spiking threshold (dashed gray line). 43 1.7 Splicing in paralytic is affected by a mutation in the Maleless RNA helicase The napts mutation in Drosophila was originally believed to be a mutation in an α or β sodium channel gene, but was finally shown in 2000 to be a mutation in the maleless gene [Reenan et al., 2000] encoding a member of the DEAH RNA helicase family (Mle) involved in dosage compensation. In Mlenap−ts mutants, approximately 80% of paralytic transcripts are aberrantly spliced, always lacking constitutive exon 19 (harboring the three editing sites discussed in Chapter 2), and creating a widespread splicing catastrophe [Reenan et al., 2000]. At permissive temperatures, 20% of correctly spliced para transcript suffices for neuronal activity in mlenap−ts mutants. However, at elevated temperatures (39◦ C) more sodium channel function is required for proper neuronal activity and 20% of the normal concentration of transcripts is insufficient. As even wild type flies paralyze at 43◦ C, mlenap−ts and parats mutations uncover normally temperature-sensitive traits in neurons [Nelson and Wyman, 1990](Figure 1.14). Para null mutations, which are recessive lethal in a wild type background, become dominant lethal in a mlenap−ts background, while parats mutations, normally homozygous viable, become homozygous lethal. A duplication of wild type para in the genome rescues these effects, indicating that the observed phenotypes are due to para dosage [Ganetzky, 1984]. The mutation responsible for the mlenap−ts “splicing catastrophe” phenotype is unknown. Candidates include a 7 bp intronic insertion and a single amino acid mu- tation (T415S) in the ATP hydrolysis domain of the protein [Smith, 2004]. Mlenap−ts homozygous males are viable, while mle−/− males are inviable, indicating that the role of Mle in the Male-Specific Lethal complex required for dosage compensation is unperturbed by the nap-ts mutation. Polytene staining of napts and wild type 44 Canton-S (wild type Mle) Mle nap-ts Figure 1.15: Third instar larvae polytene chromosome staining of the wild type Mle vs. the Mlenap−ts mutant helicase. Female Canton-S polytenes are on the left, while female mlenap−ts chromosomes are on the right. Chromosomes are stained with DAPI (blue) while the helicase is stained with an anti-Mle antibody (green), courtesy of Mitzi Kuroda. Mle reveal similar binding patterns in females, suggesting that the Mlenap−ts helicase still localizes to active sites of transcription on polytenes (Figure 1.15). Curiously, homozygous mle−/− females are viable and do not have behavioral or molecular defects; a complete absence of the Mle protein does not result in exon skipping in the para transcript [Reenan et al., 2000]. ADAR 5’ 3’ Wild type Mle helicase Mle nap-ts helicase 5’ 5’ 3’ 3’ Correct splicing Aberrant splicing 18 19 20 18 20 Figure 1.16: A model for the interaction of the Mlenap−ts helicase with the para transcript. After dADAR (purple) edits three adenosines in para exon 19 (blue), a process that is dependent on the formation of RNA secondary structures, the wild type Mle helicase (yellow) resolves these structures prior to editing. The mlenap−ts mutation results in a recessive gain of function allele, giving rise to a mutant helicase (orange) that fails to unwind the RNA structure and/or obstructs the splicing machinery from removing the intron (black) leading to aberrant spliceoforms and temperature-senstive paralysis. Figure adapted from Reenan et al. [2000] 45 Therefore, the mlenap−ts mutation results in a recessive gain of function allele. The current model for Mlenap−ts -para interaction involves the wild type Mle helicase resolving para pre-mRNA secondary and tertiary structures after editing occurs but before splicing [Reenan et al., 2000] (Figure 1.16). In the complete absence of Mle, other RNA helicases participate in para unwinding, generating properly-spliced transcript isoforms. In an mlenap−ts background, the mutant helicase not only fails to resolve the RNA structures surrounding para exon 19, but also blocks the splicing machinery in some way, leading to aberrant splicing and a temperature-sensitive paralysis phenotype. Interestingly, the 20% of full-length para transcripts present in the mlenap−ts mutant are edited at a significantly lower level than the full-length transcripts in mle+/+ animals (37% vs. 73%) [Reenan et al., 2000]. This suggests that when the transcript folds into a structure that is not conducive to editing, the transcript is more likely to be correctly spliced even in a mlenap−ts background. In other words, it may be the RNA structures responsible for directing editing that interfere with the Mlenap−ts helicase. Chapter 4 discusses this theory in greater detail. 1.8 Using Homologous recombination to engineer the Drosophila genome at the paralytic locus Previously, a mini-gene reporter construct, encompassing para exon 18-20, including all intervening intronic sequences, was used to investigate the interaction between editing and splicing in the mlenap−ts background [Smith, 2004]. While the wild type version of this construct is edited at the same three locations as endogenous para exon 19, it is to a much lower level. Additionally, while the mini-gene recapitulated the “splicing catastrophe” due to Mlenap−ts , it was in the context of the endogenous wild type para gene. Finally, the mini-gene reporter system did not allow extrapolation 46 from genotype to phenotype. Therefore, to introduce precise genetic mutations into the Drosophila para locus, we used homologous recombination (HR). HR is a technique widely used in mouse and yeast genetics, in which precisely engineered mutations are introduced into an endogenous genetic locus. The context of promoter, enhancer elements, chromatin state, and chromosomal location is preserved. Such a technique is especially important for a gene such as para, as sodium channel dosage affects organismal behavior. HR was recently developed [Rong and Golic, 2000] and streamlined [Staber et al., 2011] in Drosophila, and research application for this versatile technique in the fly continues to expand. Once the transgenic animal has been created, the process (Figure 1.17) is predicted to take approximately three months. Briefly, a construct, contained on a pW25 vector with P-element insertion ends, is designed containing two “arms” homologous to the desired mutation site. The arms are flanked by restriction sites used for sub-cloning, flippase recognition targets (FRT), and I-SceI recognition sites for future genomic excision. Between the arms is a mini-white gene flanked by loxP sites, 6-frame stop codon sequences, and restriction cloning sites. The construct is introduced into syncytial Drosophila embryos where it randomly integrates into the genome. Transgenic flies are mapped and stocked. Flies carrying the transgenic construct are crossed to animals expressing heat-shock flippase and heat-shock I-SceI endonuclease. When the progeny are heat shocked at 39◦ C, expression of the flippase and I-SceI enzymes excises the construct from the genomic integration location (Figure 1.17). In a small percentage of cases the linearized fragment will recombine into the genome at the endogenous location via cellular DNA repair mechanisms. Loss of the transgene is scored by white ommatidia, while re-integration is scored by red ommatidia, together generating a mosaic-eyed animal. To ensure that the mini-white expression is due to integration after recom- 47 P-element transgene: homology arm 1 mini-white homology arm 2 FRT I-SceI mutation loxP loxP mutation I-SceI FRT hs-Flp, hs-I-SceI Excised fragment: homology arm 1 mini-white homology arm 2 Recombination Endogenous locus: Recombination in endogenous locus: homology arm 1 mini-white homology arm 2 Cre recombinase homology arm 1 homology arm 2 mutation mutation Figure 1.17: An overview of the homologous recombination process. A P-element transgene is engineered that contains two homology arms (light blue) flanked by restriction sites (not shown). Between the arms is a “mini-white” phenotypic marker (red) flanked by loxP sites (yellow). Arms are bracketed by FRT (flippase recognition target, purple) and I-SceI recognition sites (green). The desired mutations are introduced into the homology arms. After expression of heat-shock Flippase (Flp) and I-SceI recombinases, the construct is excised from the random genomic integration point and linearized. In a small percentage of cases, the fragment will recombine in the genome at the endogenous locus (blue). The addition of Cre recombinase, which specifically recognizes the loxP sequences, removes the mini-white, leaving behind a single loxP remnant of 76 bp, which is targeted to an intronic region of low sequence conservation. 48 bination and not because the construct failed to excise from the original genomic location, a flippase under control of an eyeless promoter (expressed only in the eye) is introduced. Progeny whose red eyes persist through the ey-FLP cross likely have successfully undergone homologous recombination at the endogenous locus. Finally, a Cre recombinase enzyme is introduced, which specifically recognizes the loxP se- quences and excises the mini-white reporter (Figure 1.17), leaving behind a single loxP “remnant” of 76 basepairs in the genome. Loss of the red eye indicates successful Cre-mediated excision, and mutants are verified via PCR, restriction digestion, and Sanger sequencing. The goal of HR is the integration of a desired mutation or combination of mutations into the endogenous locus in the genome. Control animals go through the same series of crosses as mutant lines and contain the wild type gene with the loxP insertion. For complete methodology see Staber et al. [2011]. Previously, assaying two mutations at once meant the design of a new HR construct, followed by injection and the complete HR procedure (Figure 1.17). However, I discovered that Cre recombinase swaps genomic material around loxP sites when two HR-generated alleles are placed in trans. Therefore, mutations that are on opposite sides of the loxP site (in different homology arms, Figure 1.17) may be easily combined through simple crossing methods. This technique, “Cre-mediated trans recombination,” is outlined in Appendix A. I used HR to introduce a series of precisely-designed mutations into the endogenous Drosophila para locus. These mutations were designed to probe the RNA structure(s) required to direct specific editing at three adenosines in exon 19 (Chapter 2). These structural mutations were also informative when I investigated the effect of temperature on RNA editing (Chapter 3), as the RNA structures (“thermistors”) that control editing may be temperature-responsive. Finally, I assayed the effect of various engineered mutant alleles of paralytic, which are predicted to alter RNA 49 structure, on the interaction between the transcript, the Mlenap−ts helicase, and RNA splicing (Chapter 4). Chapter Two Tertiary structural elements determine extent and specificity of messenger RNA editing 51 Parts of this chapter have been published as: Rieder, LE, CJ Staber, B Hoopengardner, and RA Reenan. (2013) Tertiary struc- tural elements determine extent and specificity of messenger RNA editing. Nature Communications, in press. Dr. Rob Reenan and I conceived and designed experiments and I performed all experiments. Dr. Rob Reenan designed, constructed, and performed transgenesis of HR constructs. Ms. Cindi Staber assisted in HR construct transgenesis and mapping. Dr. Barry Hoopengardner cloned and sequenced species introns. Dr. Robert Reenan and I wrote the manuscript and I crafted the figures. All authors provided editorial feedback. I would like to thank Dr. Yiannis Savva and Dr. James Jepson for the dADARhyp and dADAR5g1 Drosophila lines and for manuscript input, Mr. Matthew A. Reyna for assistance with figures, Ms. Rachel Whitaker for qRT-PCR guidance, Ms. Kasia Sierzputowska for Figure 2.6e and Dr. Erica Larschan for manuscript input. Dr. Barry Ganetzky supplied the paralk5 ;;;para dup/Ci stock. This material is based upon work supported by a National Science Foundation Graduate Research Fellowship awarded to LER (No. DGE 0228243). 52 2.1 RNA editing of paralytic The spatial folding of a given RNA in vivo poses an experimentally difficult but vital issue for key biological processes. Recognition of RNA substrates by binding proteins is believed to be largely dependent on secondary structure [Stefl et al., 2005], bestowed on these proteins by recognition motifs seeking RNA shapes rather than primary nucleotide sequence. Although recent evidence suggests that double- stranded (ds) RNA binding domains (RBDs) may be capable of primary sequence recognition via the shallow minor groove of duplex RNA [Masliah et al., 2012], three- dimensional RNA structure likely plays a major role in highly-specific protein-RNA recognition, supporting diverse cellular functions such as splicing, RNA interference, and translation. ADAR (adenosine deaminases that act on RNA) family members possess dsRBDs, recognize diverse RNA substrates, and catalyze the deamination of adenosines into inosines in a post transcriptional process known as A-to-I RNA editing [Savva et al., 2012]. Because the chemical properties of inosine mimic those of guanosine, cellular machines, including the ribosome, interpret inosine as guanosine. Therefore, RNA editing effectively causes an adenosine to guanosine change in the primary sequence of mRNAs, potentially recoding the resulting peptides. A-to-I editing is thought to exist as a mechanism to expand protein repertoires within the nervous system [Stapleton et al., 2006, Hoopengardner et al., 2003, Rosen- thal and Seeburg, 2012]. ADARs are strikingly conserved among metazoa [Nishikura, 2010, Savva et al., 2012] and engineered model organisms that lack ADAR activity display severe neurological phenotypes, including seizures and neurodegeneration [Higuchi et al., 2000, Palladino et al., 2000, Tonkin and Bass, 2003, Wang et al., 2000]. Furthermore, ADARs are localized to the neuronal nucleus [Savva et al., 2012] and nucleolus [Desterro et al., 2003], supporting the role of editing in nascent, immature RNA molecules. One exemplar Drosophila neuronal editing substrate is paralytic (para), the lone action-potential sodium channel gene in the fly genome [Ganetzky, 53 1984]. In contrast, nine voltage-gated sodium channel genes (Nav 1.1-Nav 1.9) are encoded in the mammalian genome [Meisler and Kearney, 2005]. Despite the fact that the Drosophila genome possesses only the para locus, para transcripts are extensively alternatively spliced and edited (Figure 2.1a). These post-transcriptional modifi- cations, taken independently, result in over two million potential sodium channel isoforms [Hanrahan et al., 2000]. ADAR proteins typically contain at least one dsRBD, correlating well with the observation that edited adenosines are associated with dsRNA structures [Bass and Weintraub, 1987]. Nevertheless, very few structures that direct editing have been probed to reveal the extent to which three-dimensional properties contribute to editing specificity. Complicating this picture is the observation that the isolated ADAR catalytic domain, lacking dsRBDs, has significant activity on certain specific target dsRNAs [Eggington et al., 2011], and that mutations in the catalytic domain of human ADAR2 can result in altered editing site specificity [Kuttan and Bass, 2012]. At the RNA level, ADAR substrates range from perfectly complementary long RNA hairpins, which are promiscuously hyper-edited at up to 40% of adenosines [Wahlstedt and Ohman, 2011], to shorter ( 25 bp) imperfectly paired substrates with bulges and loops, which are specifically edited at one or a few adenosines [Nishikura et al., 1991]. The menagerie of specifically edited substrates includes simple short hairpins, such as the 71-nt stem-loop directing editing at the mammalian GluR-2 R/G site [Aruscavage and Bass, 2000], as well as comparatively enormous pseudoknotted structures, such as that found in the Drosophila synaptotagmin-1 transcript, whose elements span thousands of nucleotides [Reenan, 2005]. Most RNA structures that recruit ADAR proteins involve exonic editing sites that interact via base pairing with complementary and conserved cis non-coding sequences from a nearby intron. Such imperfectly paired duplex structures direct modification to particular adenosines. 54 The diversity of ADAR targets provides a unique source for studying RNA-protein and RNA-RNA interactions. a 100 bp 1 23 A A AA b 19 20 AAA DCS HP ECS c A d 5’ 22 nt AAA ECS 112 nt DCS 276 nt 3’ 255 nt A AA HP e Canton-S f dADARhyp g dADAR5g1 C G A T TA A T TGA A G A CA A T TA A T TGA A A A CA A T T A A T TGA A A A 1 2 3 1 2 3 1 2 3 Figure 2.1: Organization of para exon 19 editing sites and ADAR dependence. (a) The para locus contains 26 constitutive exons (blue), 13 alternative exons (yellow) and 11 sites of RNA editing (red). The boxed region in (a) is enlarged in (b), showing the three nearby editing sites in constitutive exon 19. (c) Sequence conservation of the region in (b) from 13 species of Drosophilidae found in the UCSC Genome Browser. Three regions of intronic homology, the ECS (editing site complementary sequence), DCS (donor site complementary sequence), and HP (hairpin) are putative cis sequences. (d) The indicated region of the para transcript from exon 19 (blue) and the downstream intron (black), folded in Sfold, show the conserved cis sequences denoted in (c) involved in double-stranded structures. Edited adenosines are shown in red. (e) Wild type (Canton-S) editing at the three sites in exon 19 is apparent as mixed A/G peaks in electropherograms from reverse-transcribed para mRNA. (f ) Editing at these sites is sensitive to dADAR levels, as editing in a hypomorphic dADAR mutant (dADARhyp ) are greatly reduced. (g) Editing is completely abolished in the dADAR5g1 null mutant. Containing eleven highly specific RNA editing sites (Figure 2.1a), paralytic mRNA provides an exceptional opportunity to study the enzyme-substrate inter- actions that direct RNA editing. However, previous in vivo studies of paralytic have only demonstrated an implied necessity for secondary structure [Reenan et al., 2000]. Here, we describe the in vivo molecular consequences of RNA structural perturbations, introduced via knock-in into the endogenous para locus in exon 19 and the downstream intron, using homologous recombination (HR) [Gong and Golic, 55 2003]. These mutations, designed to probe conserved cis-sequences in predicted RNA structures, establish the predicted necessity of imperfect double-strandedness around three nearby editing sites. We also demonstrate the presence of an accessory duplex, formed around the splicing donor site, which modulates an interaction between RNA splicing and the extent of editing. Finally, our data demonstrate the essential nature of a highly conserved tertiary pseudoknot that directs ADAR to only one of three nearby adenosines in the central duplex. Replacing the pseudoknot with a distinct yet functional natural example indicates the generality of the mechanism. Our data suggest that proximity of editing sites within a central duplex does not imply functional coupling. We show that extent and selectivity of editing can be determined by structural components assembled from non-local sequence elements. 2.2 Results 2.2.1 Evolutionary conservation and RNA structural predic- tions for an RNA editing site Exon 19 in the Drosophila paralytic transcript possesses three edited adenosines within 12 nucleotides (Figure 2.1b), and is predicted to form dsRNA secondary structure (Figure 2.1d). An assay of paralytic RNA editing levels in wild type and mutant Drosophila can accurately be performed via RT-PCR and Sanger sequencing. Editing at the three adenosines (sites 1-3) in exon 19 is apparent as mixed A/G peaks in wild type electropherograms (Figure 2.1e). Editing at these adenosines is sensitive to levels of Drosophila ADAR (dADAR), as flies carrying a hypomorphic dAdar allele (approximately 80% dADAR reduction) [Jepson et al., 2011] show reduced editing levels (Figure 2.1f ) and dAdar null males [Palladino et al., 2000] do not edit these sites (Figure 2.1g). In addition, the presence of editing at these three adenosines is conserved across Drosophila species [Reenan et al., 2000] (Figure 2.2). 56 a Zind Dwil Dequ Dtro Dpse Dana Dbip Dkik Dfic Deug Dtak Dyak Dere Dsec Dsim Dmel Dmau Drho Dele Dbia Camo Cpro Ddim Spat Dbus Dsil Dgri Dmen Dvir Dpol Dimm Dsig Dpic Dmer Dmoj Dhyd Zsep b Dwil Dpse Dyak Dsec Dmel Dvir Dhyd CN A T TA A T TGA A G A CGA T TA A T TGA A G A CGA T TA A T TGA N GA C G A T T A A T TG A A N A C G A T TA A T T GA A G A CG A T TA A T TGA N GA CG A T TA A T TGA G GA Dequ Dana Dere Dsim Dmer CGA T TA A T TGA A G A C GA T TA A T TGA N GA C G A T TA A T T G A N G A CG A T TA A T TGA G G A CG A T TA A T TGA A GA Dtro CG A T TA A T TGA N GA Figure 2.2: Paralytic exon 19 editing throughout Drosophilidae. (a) Cladogram of Drosophilidae. (b) Electropherograms of para exon 19 editing sites from species indicated in (a). For full species names please see Figure 2.18. 57 While the previous study of these sites experimentally revealed the minimal region of primary sequence encompassing all intronic cis-elements (including an editing site complementary sequence or ECS) necessary for RNA editing, only two species were compared when making these predictions [Reenan et al., 2000]. We suspected that our analysis was far too limited to reveal more elusive conserved elements. Therefore, we obtained the genomic regions encompassing exon 19, intron 19, and exon 20 from 37 species of the genus Drosophilidae (Figure 2.3). Sequence alignments revealed several areas of extreme intron conservation in addition to the expected ECS element (Figure 2.1c, Figure 2.3). When we used the RNA folding algorithm Sfold [Ding et al., 2004] to predict dsRNA features of the paralytic transcript in this region, three striking features were predicted in the ensemble of all structures, and corresponded to regions of high sequence conservation (Figure 2.1c, d). The expected central ECS-directed duplex feature places the three exonic editing sites within a 77 bp-long imperfect duplex formed between the exon and the region of the downstream intron with homology to the edited region (ECS; Figure 2.4a). In addition, a highly invariant region of the intron, which we call the donor-site complementary sequence (DCS; Figure 2.4b), is predicted to form a secondary duplex RNA structure encompassing the 5’ splice donor site. Finally, a predicted intronic local hairpin (HP) of unknown function is highly conserved in all species surveyed (Figure 2.4c). 2.2.2 Knock-in mutagenesis demonstrates the requirement for dsRNA directing RNA editing in vivo Traditional experimental manipulation of RNA editing structures are usually con- ducted in vitro [Daniel et al., 2012] with purified components or through the use of synthetic mini-gene reporters in tissue culture [Tian et al., 2011, Bhalla et al., 2004]. Many of these studies are capable of generally recapitulating editing and 58 Exon 19 Intron 19 (DCS) (HP) (ECS) Figure 2.3: Genomic sequence conservation of paralytic intron 19 throughout Drosophilidae. Red indicates complete base conservation while pink represents near conservation, orange some conservation, and gray no conservation. White dashes indicate inserted nucleotides. Putative cis elements are denoted in parentheses. For complete ordered species list please see Figure 2.4a. Exonic sequence is denoted in blue while intronic is in black. 59 a Editing site complementary sequence (ECS) Dtro Dtak Dmen Dhyd Dgri Dfic Deug Dere Dsim Dequ Dele Dbus Dbip Dbia Dana Cpro Camo Dpse Drho Dpol Dmau Zsep Spat Dpic Dyak Ddim Dwil Dvir Dsil Dsig Zind Dsec Dmoj Dmer Dmel Dkik Dimm b Donor site complementary sequence (DCS) c Hairpin (HP) Figure 2.4: Nucleotide resolution of paralytic intronic cis element conservation. Color-coding as in Figure 2.3. (a) Editing site complementary sequence (ECS) and Drosophilidae species abbreviations. For full species names please see Figure 2.18. (b) Donor site complementary sequence (DCS). (c) Hairpin (HP). 60 delineating the most important cis elements. Model genetic systems have explored the effects of altered ADAR levels in vivo, but provide little detail about the fine structure of ADAR targets. We chose to use ends-out homologous recombination (HR) to study the in vivo details of RNA editing, where precise introduction of various specific mutations into the endogenous Drosophila paralytic locus [Staber et al., 2011] preserves endogenous gene regulation. Briefly, the process of HR introduces mutations into the genome, as well as a 76 bp LoxP remnant directed to a region of the intron with low sequence conservation [Staber et al., 2011]. Specifically, we targeted the LoxP remnant to the non-conserved region of the intron between DCS and HP elements (Figure 2.5a). The initial targeting construct encompasses a selectable mini-white marker gene flanked by LoxP sites, unique restriction sites, and two homology arms into which the desired mutations are engineered (Figure 2.5b). After targeting, introduction of a Cre recombinase removes the mini-white gene between the LoxP sites, leaving a single LoxP remnant. A wild type construct is also targeted containing no mutations in the para locus, providing the appropriate control for all other mutants created through HR (Figure 2.5c, Figure 2.6a, Figure 2.9a). We first assessed the role of the predicted ECS by precisely excising the 39 intronic nucleotides predicted to pair with the edited region. As expected, when the para ECS is deleted from intron 19, editing at exon 19 is completely abolished (Figure 2.5d) compared to the LoxP control (Figure 2.5c). This suggests that the ECS directs the formation of a central duplex RNA structure necessary for editing at all three adenosines. Although all three edited adenosines require the ECS, editing at only the first and third sites result in amino acid substitutions in the sodium channel peptide (Glu to Arg, Asn to Asp, respectively). We next mutationally hardwired the first and third edited adenosines to guanosines, mimicking complete ADAR modification at these sites (Figure 2.5e). This mutation (GAG) restores predicted base pairing at site 61 a A AA 18 19 20 h DCS HP ECS b BsiWI (arm1) AscI Acc65I (arm 2) NotI LoxP mw+ LoxP c LoxP control ECS e GAG CA A T TA A T TGA A N A C G A T TA A T TGA G G A 1 2 3 1 2 3 f 2.5 d ECS delete Normalized Editing (G/G+A) 2.0 *** C A A T TA A T TGA A A A 1.5 1.0 0.5 LoxP Control GAG 0.0 1 2 3 2 Site Figure 2.5: Editing mutations introduced by homologous recombination (HR) demonstrate necessity of central duplex. (a) The region of the para locus targeted by HR encompasses exon 19 and the putative cis elements (black). (b) The targeting construct for HR involves two arms (white), homologous to the endogenous para locus and bordered by unique restriction sites. A mini-white reporter gene flanked by LoxP sites (red) is situated between the arms and targeted to a region in the intron of low sequence conservation. (c) As a proper control, a construct containing wild type arms is targeted to create the LoxP control line, which is wild type except for the LoxP remnant in the intron. Editing at the three sites occurs in this control. (d) When the 39 nt ECS intronic cis sequence is deleted through HR (ECS delete), editing is completely abolished at all three adenosines. (e) The GAG mutation replaces the non-silent first (Glu > Arg) and third (Asn > Asp) adenosines with guanosines (green), resulting in sodium channel peptides representing only the completely edited form. (f ) The GAG mutation (red) causes editing at the silent site 2 to increase compared to the LoxP control (gray). Bars represent standard error: ∗ ∗ ∗ P < 0.0001, please see Figure 2.17 for specific P values. 62 3 (versus an A-C mismatch, Figure 2.5c), resulting in a dramatic 2.5 fold increase in editing at synonymous site 2 (Figure 2.5f ). These data are consistent with previous observations [Reenan et al., 2000], which suggest that editing of site 2 is coupled to editing at site 3. In addition, ADAR enzymes prefer editing adenosines neighbored by a 3 guanosine and a 5’ adenosine [Lehmann and Bass, 2000]. Adenosines mismatched to a cytosine are also favored [Wong et al., 2001], suggesting that editing at site 3, an ideal dADAR target, precedes editing at site 2, which may occur as a sequential byproduct under little or no selective pressure. These results suggest that the GAG mutation improves the context of site 2 within the ECS/editing site duplex, resulting in increased editing levels. 2.2.3 Structural occlusion of splicing signals modulates RNA editing A highly conserved intronic sequence (DCS) downstream of the 5’ splice donor of intron 19 was predicted to generate a separate dsRNA structure encompassing the 5’ splice donor in wild type Drosophila (Figure 2.6a). We therefore used HR to precisely delete the DCS (24 nt) (Figure 2.6b). In “DCS delete” alleles of para, editing is reduced uniformly but not eliminated at all three sites compared to the LoxP control (Figure 2.6d, Figure 2.13b). We considered the possibility that the wild-type DCS duplex retards splicing, thereby enhancing editing, since the ECS is within the inron and more efficient splicing may disfavor editing by removing this crucial cis element. Thus, we designed a mutation (DCS zip) to extend the DCS duplex by 9 bp, which we hypothesized would energetically stabilize the DCS duplex. The insertion of the DCS zip mutation into the para intron (Figure 2.6c) resulted in substantially increased editing at all three sites (Figure 2.6d, Figure 2.13c). Our data support in vitro experiments documenting the interplay between editing and splicing [Rieder and Reenan, 2011]. 63 a LoxP control DCS d 3.0 276 nt Genotype *** DCS delete LoxP control b DCS delete DCS zip *** Normalized Editing (G/G+A) *** 276 nt 2.0 c DCS zip *** *** *** 265 nt 1.0 e 1 2 3 4 1000 0.0 1 2 3 800 Site 600 f 591 bp 500 (FL) 546 bp (a) 400 GU 426 bp h (19) A AA (20) 300 GU GU GC (d) (c) (b) 200 228 bp (a) 591 bp (b) 546 bp (c) 426 bp 100 bp (d) 228 bp Figure 2.6: Mutations in the donor site complementary sequence (DCS) effect editing and splicing. (a) In the LoxP control, the wild type DCS intronic sequence is predicted to involve the donor site in secondary structure. (b) The “DCS delete” mutation excises the conserved 24 nt DCS cis sequence from the intron. (c) Seven intronic nucleotides (green) are added to the wild type DCS (“DCS zip”), and are predicted to result in a more stable secondary structure around the splice donor. (d) Data for each site are normalized to editing in the LoxP control. Editing at all three sites significantly decreases in the DCS delete Drosophila lines compared to the control, while editing significantly increases in animals homozygous for the DCS zip allele. Bars represent standard error (∗ ∗ ∗ P < 0.0001, please see Figure 2.17 for specific P values). (e) PCR was performed on cDNA reverse-transcribed from para mRNA. Not all para transcripts from the DCS zip allele are full-length (FL). Lane 1, 100 bp ladder; lane 2, LoxP control; lane 3, DCS delete; lane 4, DCS zip. Sizes of bands are indicated on the 1.5% agarose gel. (f ) Aberrant bands from (e) were isolated and sequenced to determine the activated cryptic donor sites within constitutive exon 19 (blue) and alternative exon h (yellow). Primers (green) in exons h and 20 produce the full-length PCR product (a) from the wild type splice donor. Aberrant shorter products (b-d) are formed due to activation of cryptic alternative donor sites in exons h and 19. 64 When the splicing machinery recognizes the donor site, the intron is removed and further processing, maturation, and export should decrease the likelihood that dADAR recognizes the transcript due to loss of double-strandedness. Further, we observed that para transcripts with a wild type intron or deleted for the DCS were properly spliced, while those from the DCS zip allele exhibited improper splicing, generating mRNA size variants (Figure 2.6e). Sequencing of aberrant products from the DCS zip animals revealed that they result from cryptic splice donor site activation from within constitutive exon 19 and the upstream alternative exon h (Figure 2.6f ). All of the sequenced variants are predicted to result in nonsense mutations and, if translated, in substantially truncated non-functional sodium channel proteins. These observations strongly support the notion that the DCS cis element is involved in a dsRNA sequestering the splice donor, and that this structure modulates RNA editing levels in the central dsRNA duplex containing sites 1-3 by orchestrating the pace of intron removal by the splicing machinery. We performed quantitative RT-PCR on RNA extracted from DCS zip and LoxP control animals using primers specific to the para spliceoforms detectable by conventional PCR (Figure 2.7a). This analysis revealed a significant decrease in the full length spliceoform in the DCS zip mutants compared to the LoxP control, while the presence of several aberrant spliceoforms was detectable and significantly increased in the mutant animals (Figure 2.7b). In addition, the DCS zip allele also gives rise to significantly more total para transcript than the LoxP control allele. These data suggest that transcription from the para locus is upregulated in DCS zip animals, perhaps in attempt to compensate for low levels of properly spliced para transcript and Para sodium channel protein. The para locus encodes the only action potential-generating sodium channel gene in Drosophila [Ganetzky, 1984], and excitability of the fly nervous system is highly sensitive to para dosage [Stern et al., 1990]. Animals hemi- or homozygous 65 a b h (19) A AA GU (20) Genotype 14 DCS zip GU GU GC LoxP control 80 288bp 426bp FL Fold change (normalized to GAPDH and LoxP) 12 c Female paralysis at 39°C 100 * 10 60 DCS Zip * Percentage of flies paralyzed 75 8 40 6 50 * LoxP Control 4 20 25 * DCS Delete 2 0 0 0 0 100 200 300 Total transcript Full length splicing 288 spliceoform 426 spliceoform Time (sec) Figure 2.7: Consequences of aberrant splicing from the DCS zip allele (a) Location of qRT-PCR primers. Orange primers pick up full-length spliceoforms, while the red and purple primer pairs capture the 288bp and 426bp spliceoforms, respectively. (b) Quantitative RT-PCR shows total para transcript, detected by a primer pair spanning the 3 most exon-exon junction of the mature mRNA, is significantly increased from the DCS zip allele (purple) compared to the LoxP control (gray) (P =0.0228), while full- length transcript is significantly decreased (P =0.0493). The 288bp and 426bp aberrant spliceoforms are both significantly increased in the DCS zip animals (P =0.0404 and P =0.0037, respectively). Primers are as in (a). The 546bp spliceoform detectable by conventional PCR (Figure 2.6e) is not detectable via qRT-PCR. Each bar represents three biological replicates and three technical replicates of each biological replicate. (c) Female flies were treated as males in Figure 2.8a. Flies homozygous for the DCS zip allele paralyze significantly faster than those of the LoxP control genotype (P < 0.0001). N = 105 flies per genotype. Ribbons represent standard deviations. 66 for temperature-sensitive (ts) para alleles, which reduce para expression, display ts paralytic phenotypes, while null alleles of para result in unconditional lethality. Therefore, we assessed the phenotypic consequences of altered para splicing in our DCS zip mutant. Drosophila hemi- or homozygous for the DCS zip mutation show a significant ts-paralytic phenotype (Figure 2.8a, Figure 2.7c), similar to other parats alleles. To investigate the dosage of full-length functional para transcript originating from the DCS zip allele (Figure 2.6c), we crossed females homozygous for the DCS zip allele to males null for endogenous para on the X chromosome, but with a rescuing copy of para in a duplication translocated to the fourth chromosome. The lethality of female progeny from this cross demonstrates that the DCS zip allele is haploinsufficient. However, female siblings with an extra dose of the wild type para locus carried on the fourth chromosome duplication were rescued from this inviability (Figure 2.8b). Interestingly, although males contain only one X chromosome, males hem- izygous for the DCS zip allele are viable, likely due to the upregulation of genes on the X chromosome through the process of dosage compensation [Larschan et al., 2011]. In summary, we show that the extent of the DCS/donor site duplex strongly modulates the degree of RNA editing in a generic manner. The level of editing of all sites in the central duplex can be tuned up or down, and we propose that this is an effect modulated by the pace of splicing in vivo. While even subtle changes to intronic sequences predicted to strengthen this duplex enhance editing at the molecular level, they do so at the cost of incurring a substantial and presumably deleterious phenotypic consequence at the organismal level. 67 paralk5 para dup a b ;;; Male paralysis at 39°C Y Ci 100 Y ;;; Ci Y ;;; para dup LK5 ;;; Ci LK5 ;;; para dup DCS Zip LoxP ;;; + Percentage of flies paralyzed LoxP 75 LoxP N=446 N=378 N=578 N=612 DCS zip ;;; + 50 DCS zip DCS zip N=412 N=529 N=0 N=630 25 DCS Delete LoxP Control 0 0 100 200 300 Time (sec) Figure 2.8: Phenotypic consequences of aberrant splicing from the DCS zip allele (a) Male flies of the indicated genotypes were placed in room temperature glass vials and then submerged in a circulating 39◦ C water bath. Time to first paralysis event was recorded over 5 minutes. Flies hemizygous for the DCS zip allele paralyze significantly faster than those with the LoxP control allele (P < 0.0001). N = 105 flies per genotype. Ribbons represent standard deviations. (b) Virgin females homozygous for the LoxP control or DCS zip alleles were crossed to males with a para deletion on the X chromosome (paralk5 ) rescued by a wild type chromosomal duplication on the fourth chromosome (para dup). Allele contributions from each parent are shaded. No female progeny with only one copy of the DCS zip para allele were found (red shaded box), although this lethality is rescued by the wild type para duplication on the fourth chromosome. Lethality is not observed in male progeny hemizygous for the DCS zip allele, a phenomenon that is likely due to dosage compensation of the male X chromosome. 68 2.2.4 The hairpin structure is required for selective editing of one adenosine The highly conserved intronic hairpin (HP) structure (Figure 2.9a) has no obvious relationship to either the ECS or DCS duplexes, and is without precedence in other known ADAR targets. Surprisingly, when the hairpin is deleted through HR editing is preferentially abolished at site 1, while editing at sites two and three remain near that of the LoxP control (Figure 2.9g, Figure 2.13d). Although the precise nucleotide sequence of the hairpin stem differs between species, comparative sequence analyses revealed extensive co-variation of stem sequences that maintain base pairing (Figure 2.10a, Figure 2.10b), suggesting that the structure (Figure 2.9b) rather than primary sequence is necessary for editing. In particular, two unpaired bulged regions of the HP stem are almost invariably pyrimidines, and the HP stem resembles a barbell. We therefore introduced the “HP zip” mutation, in which some of the bulged pyrimidines were paired, converting the HP stem into a more perfect duplex (Figure 2.9c). In this case, editing was severely reduced at the first adenosine, while editing at sites 2 and 3 occurred at control levels (Figure 2.9g, Figure 2.13e). The HP stem is predicted to adopt a barbell structure due to the symmetric pyrimidine bulges (Figure 2.9b). While this structure is preserved in most species of Drosophilidae, as divergent as D. melanogaster (Figure 2.10a) and D. melanica (Figure 2.10b), D. willistoni subgroup members lack the predicted HP barbell due to stem mutations that pair the second group of pyrmidines (Figure 2.10c). Species within this subgroup therefore may provide an example of natural variation of the HP that resembles the HP zip synthetic allele (Figure 2.9c). We therefore compared editing at site 1 to editing at site 3 in species across Drosophilidae (Figure 2.2). Editing at site 1 compared to site 3 was significantly decreased in two closely related species, D. willistoni and D. equinoxialis, which lack the canonical HP barbell shape (Figure 2.10c), compared to D. melanogaster (Figure 2.10d). These data further 69 a LoxP control HP c HP zip e HP stem 2 b HP conservation d HP stem 1 f HP stem double 8 2 2 2 7 12-15 g 2.0 Genotype LoxP control HP delete Normalized Editing (G/G+A) HP zip * 1.5 HP stem 1 *** HP stem 2 ** ** HP stem double *** 1.0 0.5 0.0 1 2 3 Site Figure 2.9: Editing at site 1, within the central duplex, requires the hairpin element. (a) The highly conserved predicted wild type hairpin (HP) element. (b) Conservation of the HP across Drosophilidae. Black regions of the structure are conserved, while red are variable. Blue represents conserved pyrimidines bulges. Numbers refer to the nucleotides in the loops or basepairs in the stem. (c) The HP zip mutation pairs some of the bulged, conserved pyrimidines (green) into the HP stem. (d) The HP stem 1 mutation introduces a destabilizing PacI site into one side of the stem, while (e) the HP stem 2 mutation introduces the PacI mutation from the opposite side. (f ) Combining the HP stem 1 and 2 mutations (HP stem double) restores predicted base pairing in the HP stem. (g) Data for each site are normalized to editing in the LoxP control (gray). Deleting the HP (black) results in a selective loss of editing at site 1, while pairing the pyrimidine bulges (HP zip; yellow) results in a reduction of editing at site 1 compared to the LoxP control. Both the HP stem 1 and 2 mutations (orange and red, respectively) selectively abolish editing at site 1, while the restorative double mutation (HP stem double; brown) rescues editing at site 1. Bars represent standard error (∗ P < 0.05, ∗∗ P < 0.001, ∗ ∗ ∗ P < 0.0001, please see Figure 2.17 for specific P values). 70 suggest that the HP stem shape serves to modulate RNA editing at site 1. a D. melanogaster d 1.5 *** * Normalized to Site 3 Editing Site 1 Editing (G/G+A) 1.0 b D. melanica 0.5 c D. willistoni 0.0 Dmer Dequ Dana Dpse Dhyd Dmel Dyak Dsec Dsim Dere Dtro Dwil Dvir Figure 2.10: Hairpin conservation and editing consequences across Drosophilidae. (a) The predicted HP structure in D. melanogaster, compared to (b) that of D. melanica, which also contains the “barbell” shape due to bulged pyrimidines. In contrast, the HP of D. willistoni (c) does not have the predicted barbell. Nucleotides that differ from (a) are shown in red in (b) and (c). (d) Editing of site 1 normalized to site 3 across species. The site 1 to site 3 ratio in D. melanogaster is represented by the black dotted line. This editing ratio is significantly decreased in two closely related species, D. willistoni and D. equinoxialis, compared to D. melanogaster. For species abbreviations please see Figure 2.18. Bars represent standard error (∗ P < 0.05, ∗ ∗ ∗ P < 0.0001, for exact P values please see Figure 2.18). To test the notion that the duplex HP stem is necessary for editing at site 1, we introduced helix-disrupting mutations in the 5’ or 3’ side of the stem (Figure 2.9d, e). These mutations, “HP stem 1” and “HP stem 2,” both selectively abolish editing at site 1, while restoring the base pairing by combining these two compensatory mutations (“HP stem double,” Figure 2.9f ) rescues editing at the first adenosine (Figure 2.9g, Figure 2.13f -h). Thus, editing at site 1, residing in a separate duplex distant from the HP, requires both the presence of the HP as well as the specific paired and unpaired bases within that structure. This is particularly interesting considering that two other adenosines only ten nucleotides away in the central duplex appear to be unaffected by the presence or integrity of the HP. 71 2.2.5 A long-range tertiary pseudoknot is mediated by the hairpin A closer inspection of both the HP primary sequence as well as its potential to contribute to other structural elements was revealing. First, while the HP stem covaries between Drosophila species, the HP loop is completely invariant (Figure 2.9b, Figure 2.4c). Secondly, a short sequence just 3’ to the ECS cis element is also completely conserved, was not predicted to base pair in standard folding predictions, and creates an exact docking site for the hairpin loop sequence–a potential tertiary pseudoknot (PK) interaction (Figure 2.12a). Such interactions are very difficult to predict with current RNA folding software programs except for in the smallest input sequences. Furthermore, there is no precedence for the absolute requirement of a long distance tertiary structural element orchestrating specific RNA editing in a different dsRNA helix. Side: Top: Figure 2.11: The group II self-splicing intron α/α’ kissing loop interaction. Figures created with MacPyMOL freeware. Loop and dock nucleotides are colored yellow and red, respectively. We were struck by the similarity of the predicted para PK to a structural ele- 72 ment of the group II self-splicing intron from the extremophilic bacterium Oceanobacil- lus iheyensis [Toor et al., 2008]. In this autocatalytic intron, the α/α kissing loop interaction is required to scaffold the formation of a five-way helical junction, ensuring the proper formation of a three-dimensional structure necessary for self-catalysis (Figure 2.11) [Harris-Kerr et al., 1993]. The primary sequence of the α/α kissing loop interaction differed from that of the para hairpin loop/dock interaction by only three nucleotides. We used HR to interconvert the hairpin loop sequence to the α sequence from the kissing loop interaction (Figure 2.12b) and the docking site into that of the α (Figure 2.12c). We found that these mutations, designed to disrupt the putative tertiary para PK, independently, selectively, and completely abolish editing at para site 1 (Figure 2.12e, Figure 2.13i-j). Combining the mutations in the double mutant, which recapitulates the naturally occurring α/α kissing loop interaction of the group II self-splicing intron within the paralytic intron (Figure 2.12d) restores editing at the site one adenosine (Figure 2.12e, Figure 2.13k). Thus, our data strongly suggest that not only does the tertiary pseudoknot exist in vivo, but it appears to use a structural variant similar to a natural example and serves to selectively direct the RNA editing enzyme to specifically deaminate only one of three nearby edited adenosines in the central duplex. 2.3 Discussion 2.3.1 Complex and novel structural requirements for RNA editing The RNA informational complement of only four building blocks, amplified by simple base pairing rules, nevertheless allows for an almost infinite variety of secondary and tertiary structures, surfaces, and chemistries. In silico approaches such as comparative genomics and structural predictions are very useful in identifying important base 73 a Predicted pseudoknot 5’ 3’ b Loop > ! c Dock > !’ d Loop/dock > !/!’ e 2.0 Genotype LoxP control *** loop > ! *** *** dock > !’ Normalized Editing (G/G+A) 1.5 loop/dock > !/!’ *** *** *** *** 1.0 0.5 0.0 1 2 3 Site Figure 2.12: The hairpin is involved in a tertiary pseudoknot that directs selective editing in the central duplex, and can be functionally replaced. (a) The predicted intronic tertiary pseudoknot is composed of seven base pairs. Exon, blue; intron, black; editing sites, red. (b) The “Loop > α” mutation alters three nucleotides (green) in the HP loop so that the sequence recapitulates the α sequence from the group II self-splicing intron. (c) The “Dock > α’ ” mutation introduces three nucleotide changes (green) in the docking site of the predicted para pseudoknot, altering the sequence into that of the self-splicing intron α’ sequence. (d) The double “Loop/dock > α/α’ ” mutation restores the predicted pseudoknot interaction using the α and α’ sequences (green) from the group II self-splicing intron kissing loop interaction. (e) Editing is normalized to that in the LoxP control (gray). Both the “Loop > α” and the “Dock > α’ ” mutation (light and dark purple, respectively) abolish editing at site 1, while the double mutation, “Loop/dock > α/α’ ” (blue) rescues editing at this site. Bars represent standard error (∗ ∗ ∗ P < .0001, please see Figure 2.17 for specific P values). 74 a LoxP: b DCS delete: c DCS zip: CA A T TA A T TGA A N A C A A T TA A T TGA A A A C G A T TA A T TGA G G A 1 2 3 1 2 3 1 2 3 d HP delete: e HP zip: C A A T T A A T TG A A N A C A A T TA A T TGA A A A 1 2 3 1 2 3 f HP stem 1: g HP stem 2: h HP double: C A A T T A A T TGA A A A C A A T T A A T TGA A A A C A A T T A A T TG A A A A 1 2 3 1 2 3 1 2 3 i Loop > !: j Dock > !’: k Loop/dock > !/!’: C A A T T A A T TG A A N A C A A T T A A T TG A A N A C N A T T A A T T GA N G A 1 2 3 1 2 3 1 2 3 Figure 2.13: Representative electropherograms from all HR-generated mutant Drosophila lines. (a) LoxP control, reproduced from Figure 2.5c. (b-k) Mutations are as indicated. 75 pairing interactions in secondary structures. However, higher order RNA structures or subsidiary structural components and their spatial relationships may be difficult to discern. The enzyme-substrate requirement for RNA editing by ADARs has been pursued using both in vitro and in vivo systems. In general, most studies have focused on the necessity for relatively simple local secondary structural features of a more or less perfect central RNA duplex. In contrast, our studies reveal structural elements that are distant from the central duplex containing the edited adenosines, and play roles in both modulating and specifying sites of RNA editing in vivo. 2.3.2 Modulation of RNA editing by a structure separate from the central duplex Many A-to-I editing sites are found proximal to intron-exon boundaries, and edited adenosine(s) and splicing signals are encompassed in the same dsRNA region [Rieder and Reenan, 2011]. In these cases, proximity of splicing boundaries and tempo of splicing are regulators of editing. The highly conserved subsidiary DCS/donor site RNA duplex is suggestive of a balance between splicing and editing; postponing splicing by occlusion of splicing signals would extend the time allotted for RNA folding, ADAR binding, or both, resulting in increased editing. Yet, efficient donor site recognition by the splicing machinery must eventually occur. We previously reported that the “no action potential” (nap) mutation in the Maleless dsRNA helicase results in a splicing catastrophe around para exon 19 [Reenan et al., 2000], causing exon skipping due to avoidance of the 5’ donor site for the edited exon. Therefore, precedence suggests that this constitutive splicing event requires unique processing. The deletion of the DCS decreases editing overall at sites in the central duplex and appears to have no deleterious effect on splicing. Thus, the main effect of the wild type DCS appears to be as an upward modulator of editing. In contrast, the DCS zip mutation adds 7 intronic nucleotides 271 nucleotides 76 away from the splice donor (Figure 2.6c), designed to extend the DCS/donor site RNA helix. Even this modest non-coding change resulted in dramatic increase in RNA editing, at the cost of cryptic donor site activation within exon 19 and the upstream alternative exon h (Figure 2.6f ). Due to this aberrant splicing, we hypothesize that fewer full-length sodium channel transcripts are produced (Figure 2.7b) resulting in a sodium channel hypomorphic allele. A distal intronic duplex in the mammalian Gabra-3 transcript was recently reported to stimulate specific editing, but this element does not include splicing signals [Daniel et al., 2012]. The authors hypothesize that the Gabra-3 element serves to increase ADAR concentration locally, which is quite different from our proposed mechanism. Interestingly, the predicted DCS of Musca domestica is extended in both directions and a fourth adenosine is edited within this structure (Figure 2.14a-b, e). The predicted M. domestica DCS, while more extensive than that in Drosophilidae, is predicted to contain a large central nucleotide bulge, which perhaps aides the splicing machinery in donor site recognition by destabilizing the DCS structure. Sodium channel abundance affects Drosophila behavior, development, repro- duction, and aging [Garber et al., 2012]. In both sexes, flies hemi or homozygous for the DCS zip allele are sensitive to temperature (Figure 2.8a, Figure 2.7c), displaying a paralysis phenotype consistent with that of other temperature-sensitive para alleles that are due to loss of para activity [Ganetzky, 1984]. Thus, while substantial in- creases in editing occur in the DCS zip mutant, this increase comes at the expense of splicing defects and likely explains why the DCS is not larger in size and effect than that observed. Splicing and editing occur co-transcriptionally [Rodriguez et al., 2012], and recent evidence points toward a role for RNA structure in splicing [McManus and Graveley, 2011, Zhang et al., 2011]. Our data suggest that an RNA duplex sequestering splicing signals, distant from the central duplex, is finely-tuned in size to increase RNA editing levels by effecting splicing, while simultaneously not exceeding 77 a b Musca domestica 5’ 1 2 3 4 ECS DCS ? d CA AT T A A T T G A A N A e A A T G AG T A G C HP 3’ c Sarcophaga bullata 1 2 3 4 Figure 2.14: Editing and structure in outgroup species. (a) Genomic sequence comparison between Musca domestica (top) and Sarcophaga bullata (middle). Consensus sequence is on the bottom. Red nucleotides represent conservation, while gray denotes disparity. The relevant cis elements and four editing sites are noted. (b) Predicted para structure of Musca domestica. Four editing sites are shown in red, while intronic nucleotides that differ from the sequence of D. melanogaster are in green. Although the HP structure resembles that in Drosophilidae, the HP sequence is completely different. (c) The predicted HP from Sarcophaga bullata. Nucleotides different from Musca domestica (b) are show in orange. (d) Electropherograms representing editing of adenosines 1-3 and (e) 4 in M. domestica, which is not edited in Drosophilidae. 78 a size beyond which splicing is compromised (Figure 2.16a). Suggestively, editing levels are increased in Drosophila carrying a mutant strain of “slow” RNA polymerase II (pol II) compared to Canton-S wild type animals with a “normal” pol II (Figure 2.16b-c). The C4 point mutation in the largest polymerase subunit confers on the enzyme a slower elongation rate, and is known to cause alternative splicing on endogenous genes in both human cells and Drosophila [de la Mata et al., 2003]. Because splicing is believed to occur co-transcriptionally, it is possible (but speculative) that splicing is also slowed when transcription rate is retarded. With this speculation, the data presented in Figure 2.16b is consistent with our model (Figure 2.16a) in which slower splicing (due to slowed pol II) results in an increased level of editing at all three para sites. However, the slow pol II and Canton-S lines do not share a genetic background and therefore these data are simply suggestive. Additionally, we assayed behavioral phenotypes of the editing extremes: the GAG mutant, in which all para transcripts are 100% edited at sites 1 and 3, resulting in Para channels with Arg and Asp residues from the edited codons, and the ECS delete mutant, in which editing is abolished at sites 1-3 leading to 100% Glu and Asn residues in the peptide. The sodium channel α subunit consists of four similar domains, each consisting of six transmembrane helices, and the exon 19 editing sites map to the first transmembrane helix of domain III. The editing extreme mutations conferred no detectable activity or mating phenotype (Figure 2.15), suggesting, as is consistent with theories of mRNA editing, that these changes result in subtle alterations in channel function. 79 a LoxP control activity b ECS delete activity N = 117 N = 121 70 70 60 60 average beam breaks average beam breaks 50 50 40 40 30 30 20 20 10 10 0 0 00:30 06:00 12:00 18:00 24:00 00:30 06:00 12:00 18:00 24:00 time (half hour bins), averaged over 3 days time (half hour bins), averaged over 3 days c GAG activity d Activity comparison N = 98 70 70 LoxP control ECS delete 60 average beam breaks 60 GAG average beam breaks 50 50 40 40 30 30 20 20 10 10 0 0 00:30 06:00 12:00 18:00 24:00 00:30 06:00 12:00 18:00 24:00 time (half hour bins), averaged over 3 days time (half hour bins), averaged over 3 days e Latency f Latency g Latency (all) (successful mating) (UNsuccessful mating) 60 60 60 Latency (s) Latency (s) Latency (s) 40 40 40 20 20 20 N = 27 N = 38 N = 34 N = 33 N = 17 N = 31 N = 21 N = 25 N = 10 N=7 N = 13 N=8 0 0 0 CS LoxP control GAG ECS delete CS LoxP control GAG ECS delete CS LoxP control GAG ECS delete Genotype Genotype Genotype Proportion of time spent courting Proportion of time spent courting Proportion of time spent courting h Time spent courting i Time spent courting j Time spent courting by all males by successful males by UNsuccessful males 1.2 1.2 1.2 0.8 0.8 0.8 0.4 0.4 0.4 N = 27 N = 38 N = 34 N = 33 N = 17 N = 31 N = 21 N = 25 N = 10 N=7 N = 13 N=8 0.0 0.0 0.0 CS LoxP control GAG ECS delete CS LoxP control GAG ECS delete CS LoxP control GAG ECS delete Genotype Genotype Genotype Number of males assayed k Total number of assays l Courting time before copulation by successful males 60 Successful 400 Time (s) Unsuccessful 40 200 20 N = 27 N = 38 N = 34 N = 33 N = 17 N = 31 N = 21 N = 25 0 0 CS LoxP control GAG ECS delete CS LoxP control GAG ECS delete Genotype Genotype Figure 2.15: Diurnal and mating behavior in editing extremes. (a) Circadian activity (beam breaks) of LoxP control males (N = 117), averaged over 3 days into half-hour bins. White bars represent lights-on, while gray bars represent lights-off. Black shadows represent standard deviation. (b) Activity of ECS delete males (N = 121). (c) Activity of GAG males (N = 98). (d) Comparison of activity between LoxP control males (black), ECS delete males (blue) and GAG males (red). Gray lines represent lights-on and lights-off, as in (a-c). (e-g) Mating latency (time before courtship initiation) of all males (e), successful males (f ), and unsuccessful males (g). Numbers are indicated within bars and vertical lines represent standard deviations. The red line aligns with LoxP control average in each graph. (h-j) Total time spent courting by all males (h), successful males (i), and unsuccessful males (j). (k) Total number of assays by male genotype. Green represents successful males, while red represents unsuccessful males. (l) Courting time before copulation of successful males. 80 2.3.3 RNA editing specificity conferred by a tertiary pseu- doknot Complex tertiary interactions that extend base pairing to extrahelical and non- canonical base pairs in dsRNA structures are exemplified by the pseudoknot. PK structures are essential to diverse biological processes. For example, in protein syn- thesis, the central PK of the 30S ribosomal subunit serves to scaffold the proper arrangement of three major structural domains [Brink et al., 1993]. Likewise, telom- erase function is dependent on a conserved core PK within the telomerase RNA linked to the RNA template strand for DNA synthesis [Blackburn and Collins, 2011]. In contrast to these structural PKs, there are fewer and less highly conserved PKs reported in mRNA. However, PKs figure prominently in viral mRNAs in the process of ribosomal frameshifting, a process vital to viral life cycles [Brierley et al., 2007]. ECS elements underlie the central duplex containing sites of ADAR action and, while they can be found at a significant distance from editing sites, are generally thought to contain all of the information for ADAR specificity. For example, the para ECS appears to be necessary and sufficient for editing at sites 2 and 3 (Figure 2.16d). However, the PK interaction we report here is unique in its characteristics as a tertiary structure directing specific RNA editing. Length and helix discontinuities (loops and bulges) of RNA central duplexes limits ADAR activity to particular adenosines [Ohman et al., 2000]. We propose that the exon/ECS duplex is inconsistent in either size or structure with a standard binding of dADAR to the duplex for editing at site 1, and necessitated the evolution of a more complex structure to orchestrate binding of dADAR and engage the catalytic domain on the site 1 adenosine (Figure 2.16e). The paralytic pseudoknot interaction reported herein encompasses seven Watson-Crick base-pairing interactions. Although the hairpin loop and docking site are absolutely conserved in all the Drosophilidae species we investigated (Figure 2.9b, Figure 2.4c), replacing the pseudoknot sequences with a naturally occurring motif 81 a wt Splicing Editing DCS zip DCS delete b Slow RNA Polymerase II c Canton-S C G A T T A A T TG A A G A C G A T T A A T TG A A G A 1 2 3 1 2 3 d II e I 5’ 5’ 3’ 3’ Figure 2.16: A model for paralytic pre-mRNA editing and splicing. (a) Editing (red) and splicing (blue) are inversely correlated. When splicing is very slow or inefficient, editing is high, for example, in the para “DCS zip” allele. If splicing is rapid or highly efficient, the paralytic structure is quickly resolved, resulting in lower levels of editing, for example, transcripts from the para “DCS delete” allele. (b) Editing in “slow” RNA Polymerase II mutant Drosophila, compared to editing in (c) Canton-S with a “normal” Polymerase. (d) dADAR dsRNA binding domains (orange) bind to only the ECS to edit sites 2 and 3 into inosine (red). This occurs before the splicing machinery (blue) recognizes the splice donor. (e) dADAR binds to the ECS and HP duplexes to correctly position the catalytic domain (red) to edit para site 1. 82 (the kissing loop sequences from the group II self-splicing intron) not only completely supports editing in vivo, but increases editing at all three edited adenosines (Figure 2.12e). While the Drosophila PK sequences are invariant, the α/α’ sequences from group II introns do co-vary, preserving structure while altering primary sequence [Michel et al., 1989]. Indeed, seven base-pairs is thought to be a minimum required for rapid nucleotide annealing [Cisse et al., 2012], although certain kissing loops comprising as little as two base pairs can have surprising stability [Chen and Garcia, 2012]. Musca domestica and Sarcophaga bullata show RNA editing at sites 1-3 (Figure 2.14a, d), and comparative genomics reveal a conserved hairpin, located between the orthologous DCS and ECS elements, with an invariant 8-nucleotide loop (Figure 2.14a-c). Curiously, the loop sequence differs from that seen in Drosophilidae, and no obvious conserved docking site was observed. Nevertheless, we believe that the HP in these species directs site 1 editing, probably through a long-range tertiary interaction that remains to be determined. 2.3.4 Broader implications for RNA editing Recent effort in the editing field has focused on identification of new editing sites in multiple organisms via RNAseq, thereby compiling comprehensive inosinomes [Wulff et al., 2011]. Yet these studies are often controversial [Li et al., 2011], and pose substantial analytical challenges due to false positives [Bass et al., 2012]. More broadly, our observations suggest that proximity of editing sites does not imply functional coupling. Future efforts to find polymorphisms affecting editing levels in human genes should not assume that standard ECS elements are sufficient to determine editing extent and specificity. Editing is implicated in an expanding list of diseases, including breast cancer [Shah et al., 2009], suicidal depression [Lyddon et al., 2012] and ALS. Substantial 83 evidence points toward ADAR2 mis-regulation in motor neurons of ALS patients [Hideyama et al., 2012], and recent identification of a hexanucleotide repeat associated with ALS [Renton et al., 2011] has strengthened the hypothesis that aberrant RNA processing is involved in ALS, possibly through the seclusion of RNA binding proteins such as ADAR, with irregular transcripts in inclusion bodies [Baloh, 2012]. Additional insights into the secondary and tertiary RNA structures that direct editing may allow for prediction of potential editing substrates, and help to rule out false positives identified via deep sequencing methods. In addition, artificial editing substrates may be designed, enabling the co-option of endogenous ADAR enzymes as tools in specific RNA therapies. For example, a 937 bp antisense sequence designed to trigger RNA editing and degradation of the HIV env transcript, is already in clinical trials [Burnett et al., 2012]. Further, the adenosine preference of the human ADAR protein is mutable [Kuttan and Bass, 2012], and the enzyme itself therefore presents a possible target for genetic disease therapy. 2.4 Methods Sequence alignments Sequence alignments were performed using the ClustalW multiple sequence alignment package included with MacVectortm 7.2 (Accelrys Inc.). Open and extend gap penalties were set to 1.0. The 37 species alignment was then used for the cladogram in the supplemental material. The cladogram was generated using the Neighbor- joining method, bootstrap (1000 reps). RNA structural predictions Structural predictions for regions encompassing all elements and the HP element alone were performed using the SFOLD web-based algorithm (http://sfold.wadsworth.org/cgi- bin/index.pl) using the SRNA package to generate general features and output for 84 Raw editing P value Significance Genotype Site level (SE) vs. LoxP code LoxP control 1 0.314 (0.0157) NA GAG 1 1 NA DCS delete 1 0.102 (0.0115) < 0.0001 *** DCS zip 1 0.623 (0.0073) < 0.0001 *** HP delete 1 0 < 0.0001 *** HP zip 1 0.103 (0.0080) < 0.0001 *** HP stem 1 1 0 < 0.0001 *** HP stem 2 1 0 < 0.0001 *** HP stem dbl 1 0.171 (0.0039) < 0.0001 *** Loop > a 1 0 < 0.0001 *** Dock > a' 1 0.008 (0.0066) < 0.0001 *** Loop/Dock > a/a' 1 0.497 (0.0240) < 0.0001 *** LoxP control 2 0.234 (0.0168) NA GAG 2 0.566 (0.0097) < 0.0001 *** DCS delete 2 0.076 (0.0123) < 0.0001 *** DCS zip 2 0.684 (0.0058) < 0.0001 *** HP delete 2 0.167 (0.0084) 0.0003 ** HP zip 2 0.194 (0.0058) 0.0644 HP stem 1 2 0.145 (0.0037) < 0.0001 *** HP stem 2 2 0.136 (0.0078) < 0.0001 *** HP stem dbl 2 0.181 (0.0037) 0.0056 * Loop > a 2 0.214 (0.0074) 0.8318 Dock > a' 2 0.353 (0.0081) < 0.0001 *** Loop/Dock > a/a' 2 0.410 (0.0191) < 0.0001 *** LoxP control 3 0.387 (0.0170) NA GAG 3 1 NA DCS delete 3 0.205 (0.0141) < 0.0001 *** DCS zip 3 0.809 (0.0058) < 0.0001 *** HP delete 3 0.362 (0.0120) 0.6976 HP zip 3 0.382 (0.0075) 1 HP stem 1 3 0.339 (0.0064) 0.0728 HP stem 2 3 0.311 (0.0143) 0.0009 ** HP stem dbl 3 0.383 (0.0043) 1 Loop > a 3 0.385 (0.0079) 1 Dock > a' 3 0.514 (0.0080) < 0.0001 *** Loop/Dock > a/a' 3 0.586 (0.0277) < 0.0001 *** Figure 2.17: Specific P values. P values are the result of one-way ANOVAs with Dunnett post-hoc tests (alpha = 0.05). Testing is always against the LoxP control. Significance codes: ∗ P < 0.05, ∗∗ P < 0.001, ∗ ∗ ∗ P <0.0001. 85 Flybase Drosophila P value Significance abbreviation species vs. Dmel code Zsep Zaprionus sepsoides Dhyd Drosophila hydei 0.0568 Dmoj Drosophila mojavensis Dmer Drosophila mercatorum mercatorum 0.8212 Dpic Drosophila picticornis Dsig Drosophila similis grenadensis Dimm Drosophila immigrans Dpol Drosophila polychaeta Dvir Drosophila virilis 0.986 Dmen Drosophila melanica Dgri Drosophila grimshawi Dsil Drosophila silvestris Dbus Drosophila busckii Spat Scaptodrosophila pattersoni Ddim Drosophila dimorpha Cpro Chymomyza procnemis Camo Chymomyza amoena Dbia Drosophila biauraria Dele Drosophila elegans Drho Drosophila rhopaloa Dmau Drosophila mauritiana Dmel Drosophila melanogaster NA Dsim Drosophila simulans 0.9973 Dsec Drosophila sechellia 0.2119 Dere Drosophila erecta 1 Dyak Drosophila yakuba 1 Dtak Drosophila takahashii Deug Drosophila eugracilis Dfic Drosophila ficusphila Dkik Drosophila kikkawai Dbip Drosophila bipectinata Dana Drosophila ananassae 0.8583 Dpse Drosophila pseudoobscura pseudoobscura 0.8429 Dtro Drosophila tropicalis 0.6376 Dequ Drosophila equinoxialis 0.0131 * Dwil Drosophila willistoni <0.0001 *** Zind Zaprionus indianus Figure 2.18: Drosophilidae species abbreviations and P values. Drosophilidae species abbreviations and species, as given in Flybase. P values for the 13 species represented in Figure 2.2b and Figure 2.10d (bold) are the result of one-way ANOVAs with Dunnett post-hoc tests (alpha = 0.05). Comparison is to D. melanogaster. Significance codes: ∗ P < 0.05, ∗∗ P < 0.001, ∗ ∗ ∗ P < 0.0001. 86 statistical RNA folding. Drosophila stocks Stocks were maintained at 25◦ C under 12-h light/dark cycles on standard cornmeal molasses food. The slow RNA polymerase mutant Drosophila line “RpII215C4 ” was obtained from Bloomington Stock Center (No. 3663). Ends-out homologous recombination of the paralytic locus We performed ends-out homologous recombination after Staber et al. [2011]. This technique involves two 2.5 Kb arms homologous to the para locus and flanked by recognition sites for the Flp and I-Sce I endonucleases. We cloned and sequenced homology arms in pTOPO (Life Technologies) and then ligated them into a p[w25.2] targeting vector. We achieved mutations in the desired region using the Quik Change II XL Site-directed Mutagenesis Kit (Agilent Technologies). The mini-gene white+ , situated between the arms and flanked by LoxP sites, acts as a selectable eye color marker and is removed later with Cre-recombinase. The vector was introduced into the Drosophila genome by injection (Genetic Services, Inc.). We targeted the endogenous para locus so as to generate multiple independent events. We removed the white+ marker from each line by crossing in Cre-recombinase and isolating individuals carrying targeted alleles with a single LoxP remnant. We then validated targeted alleles via Sanger sequencing (University of Wisconsin Biotech- nology Center). RNA editing analyses For each mutation analyzed we performed six analyses, two PCRs each from three independent Drosophila HR lines. Species-specific editing data were obtained from 3 PCR replicates derived from a single RNA sample. For each sample we extracted RNA from male heads (N = 15 − 20) using Tri Reagent (Molecular Research Center, Inc.). We amplified cDNAs via RT-PCR using a para-specific primer (Table 2.1), performed PCR and electrophoresed samples on agarose gels. Species-specific editing 87 data were obtained by designing PCR and RT primers to regions of homologous sequence (Table 2.1). We cleaned the products using Wizard Gel and PCR Cleanup Kits (Promega) and detected editing by Sanger sequencing (University of Wisconsin Biotechnology Center). We determined editing ratios by measuring the area under select A and G nucleotide traces in Adobe Photoshop. Table 2.1: Specific paralytic primer sequences. Step Forward (F) and Reverse (R) Primers Reverse transcription CCTGTTCGTTGGTTAAGCACTTTGC PCR (D. melanogaster ) F: CAGATGATTGGCAACTCAATTAACCACC R: CGCGACCATGTTCGATATGGCAGGTAG PCR (species) F: CCATAAGAATCGACCATTCAAGGACG R: CTTTGAAGCCGAGCGCCAACCACTTG Sanger sequencing CTTTGAAGCCGAGCGCCAACCACTTG qRT-PCR (total transcript) F: CAACGACGTCTACAACTTCAAGACC R: CTATTACCGCAGGTCATGTGGTAG qRT-PCR (full-length transcript) F: GGATGGGGCAATTTACGACTGAAAAC R: GTGGCAGATGTACATCTTCTAATGC qRT-PCR (426 bp spliceoform) F: GGTCTCGACGAGGAACTGGAC R: CCTGCAGTATGGGTCTTTGTGGCAG qRT-PCR (288 bp spliceoform) F: CTCAATTAACCACCAAGACAATAG R: CTTTGAAGCCGAGCGCCAACCACTTG qRT-PCR(GAPDH) F: GACGAAATCAAGGCTAAGGTCG R: AATGGGTGTCGCTGAAGAAGTC Identification of cryptic splice site activation We cleaned PCR products, as above, and sequenced individual products. Please see Table 2.1 for specific primer sequences. Temperature-sensitivity paralysis assay We raised Drosophila at room temperature (18◦ C) and tested all animals within 24 hours after ecclosion. We mouth pipetted single flies into room temperature glass vials, which we then submerged in a circulating 39◦ C water bath. We measured time from submersion until first detectable paralytic event, defined as 5 seconds lying on the back or (rarely) side. We tested 105 flies per genotype (35 each from 3 independent HR lines) per sex. 88 Quantitative RT-PCR assay We isolated mRNA from 40 male Drosophila heads per biological replicate using oligo d(T)25 magnetic beads (New England BioLabs). We normalized and reverse- transcribed the mRNA using the iScript cDNA Synthesis Kit (Bio-Rad). We then performed quantitative RT-PCR using the SYBR Green PCR Master Mix (Applied Biosystems) and ran plates on the 7500 Fast Real-Time PCR System machine (Applied Biosystems). We performed three technical replicates per biological replicate, and three biological replicates for both the LoxP control and DCS zip lines. We used primers specific to the mature GAPDH transcript to normalize the data from replicates (Table 2.1). To detect total para transcript, we designed primers specific to the 3 most exon-exon junction of the mature para mRNA. To detect spliceoforms we designed primer pairs within exons 19 and 20 (see Figure 2.7a, Table 2.1). We performed melting curves on each PCR product to ensure primer specificity. Statistical analyses For editing analyses, we performed one-way ANOVAs (alpha = .05) followed by Dunnett post hoc tests. ∗ P < 0.05, ∗∗ P < 0.001, ∗ ∗ ∗ P < 0.0001. All experimental lines were compared to LoxP control lines. For specific P values please see Figure 2.17. For paralysis analyses we performed logrank tests on the raw data. We performed one-way ANOVAs on qRT-PCR data after Rieu and Powers [Rieu and Powers, 2009]. Chapter Three Temperature-dependent modulation of editing via RNA “thermistors” 90 The work for this chapter was carried out in collaboration with Dr. Yiannis Savva (Brown), Mr. Matthew Reyna (RPI). I wrote the manuscript, crafted the figures, and carried out data analysis. Dr. Yiannis Savva and I shared data collection responsibilities. Mr. Matthew Reyna scripted a program to compute editing ratios. Editorial feedback was contributed by all authors. Thanks to Mr. Yao-jen Chang and Ms. Patricia Santos for western help, the laboratory of Dr. Stephen Helfand for incubator use, and Dr. James E.C. Jepson for tagged endogenous dADAR mutants. 91 3.1 RNA as cis-acting “molecular sensor” The “RNA World hypothesis,” first suggested by Walter Gilbert in 1986, proposes that RNA is the ancient molecule [Gilbert, 1986] capable of self-replication, chemical reaction catalysis (e.g. “ribozymes”), and environmental sensing. While DNA is limited to double-stranded helical shapes, the repertoire of possible RNA secondary and tertiary structures appears limitlesss. Wan et al. recently measured RNA folding energies genome-wide at single-nucleotide resolution, revealing that classes of RNAs fall into categories of thermodynamic stability [Wan et al., 2012]. These findings reveal, unsurprisingly, that RNA secondary structure is correlated to its function, and that both of these properties–structure and stability–contribute to transcriptional regulation [Bonetti and Carninci, 2012]. RNA structures are dynamic, and, as detailed below, they are incredibly responsive to input in the form of molecular and environmental factors. It is the mutable property of RNA structure that allows RNA to act as sensor and elicit rapid responses on the cellular level [Wan et al., 2011]. RNA structures are involved in transcription, splicing, cellular localization, silencing, translation, and RNA turnover [Wan et al., 2011]. Therefore, it is not surprising that RNAs also play a role in transcriptional regulation in response to environmental signals. Although the process of environmental sensing is often carried out by proteins, a range of RNA molecules have recently been discovered that recognize ligands or other signals and control post-transcriptional gene expression. Amazingly, these RNA sensors use only the four ribonucleotide building blocks, in contrast to the twenty amino acids available to proteins. In addition, RNA sensors are often highly specific, a property that has garnered much interest from the medical field. For example, RNA aptamers, single-stranded nucleic acids that bind to molecular ligands and inhibit their functions (Figure 3.1 A), are easily selected in an in vitro process called SELEX (Systematic Evolution of Ligands by EXponential enrichment) [Tuerk and Gold, 1990]. The first aptamer therapeutic, Pfizer’s Macugen 92 to treat age-related macular degeneration, was approved by the FDA in 2005. However, it remains the only aptamer therapeutic on the market [Sundaram et al., 2013]. 3.1.1 Riboswitches A sub-class of RNA aptamers,“riboswitches,” were discovered in 2002 in bacteria [Mironov et al., 2002, Nahvi et al., 2002, Winkler et al., 2002]. A riboswitch, which typically consists of an aptamer domain and an expression domain (Figure 3.1 B), senses cellular metabolites to modulate gene expression through changes in RNA structure. For example, the vitamins B1 , B2 , and B12 directly interact with mRNAs to control operons involved in metabolite biosynthesis and transport [Serganov and Nudler, 2013]. The btuB gene in E. coli and the cob operon in S. typhimurium both bind coenzyme B12 , precisely regulating internal cellular concentrations of this metabolite. More than 200 similar variants of this riboswitch have been found in diverse bacteria [Mandal and Breaker, 2004]. Although widespread in prokaryotes, only a single family of riboswitch has been found in eukaryotes [Wan et al., 2011]. Instead of exerting transcriptional control, eukaryotic thymidine pyrophosphate (TPP) riboswitches regulate alternative splicing. Riboswitch structures form in some eukaryotic mRNAs, involving an intergenic region or 3’ UTR sequence that base pairs around a splice junction. In the absence of the ligand, TPP, normal splicing occurs. However, in the presence of TPP, the RNA structure is altered such that the alternative exon is included, leading to translation termination or aberrant peptide synthesis [Serganov and Nudler, 2013]. TPP-regulated riboswitches are found in bacteria, plants, and fungi [Sudarsan et al., 2003], suggesting they represent an ancestral form of genetic control. Therefore, the ability of RNAs to sense environmental signals and effect change at the level of post-transcriptional gene expression is found in diverse life forms. 93 A. Aptamer B. Riboswitch C. RNA thermometer D. RNA “thermistor” I I RBS AUG aptamer domain ? ribosome A A RBS AUG expression domain Figure 3.1: RNA sensor families. A. Aptamers, which may be DNA or RNA, bind small molecule ligands (purple) and inhibit their fuction. B. Bacterial RNA riboswitches contain an aptamer domain and an expression domain. When the aptamer domain binds a ligand (purple, often a cellular metabolite), the RNA changes shape and affects subsequent translation. Certain sequences of the RNA (orange) have alternative cis binding partners. C. RNA thermometers, most often found in bacteria, structurally sequester the ribosome binding site (RBS) and start codon (red). When the structure melts, the ribosome (green) may bind and translate the RNA. RNA thermometers most often regulate heat- and cold-shock genes and virulence factors. D. We suggest that there exist RNA “thermistors” structures, sometimes complex, required for ADAR (pink) recognition and A-to-I editing, that are sensitive to temperature. Thus a thermistor could affect RNA editing levels in response to temperature fluctuations. 94 3.1.2 RNA thermometers: temperature, RNA structure, and gene regulation Yet RNA sensing is not limited to molecular ligands; RNA structure is fundamentally sensitive to temperature. The differential stability of RNA molecules in the transcrip- tome correlates to the diverse cellular roles for RNA structures. For example, Wan et al. detected specific “signatures” of yeast RNA melting temperatures that detected RNA properties, including the polarity of mRNA open reading frames, regulatory motifs in 3’ UTRs and noncoding versus coding RNAs [Wan et al., 2012]. Computa- tional programs are reasonably accurate when predicting secondary structure of short (<700 bp) input RNA sequences [Wan et al., 2011]. Yet the probability of finding this, the most thermodynamically-stable (minimum free energy) structure, drops as temperature increases [Huynen et al., 1997]. At higher temperatures RNA molecules dynamically adopt various probable secondary and tertiary structures. RNA thermometers, as their name suggests, are riboswitches that are sensitive to temperature (Figure 3.1 C). RNA thermometers are found in bacteria, which employ these temperature-senstive regulatory elements to control transcription of heat-shock, cold-shock, and some virulence genes (thought to be triggered by entry into a warm-blooded host) [Narberhaus et al., 2006]. All known bacterial RNA thermometers sequester the ribosome binding site (Shine-Dalgarno sequence) and start codon at certain temperatures to inhibit translation. Simple synthetic RNA thermometers have been created by structural occlusion of the ribosome binding site, indicating that they most likely operate though a basic melting mechanism [Neupert et al., 2008]. Although there is no direct evidence of RNA thermometers in eukaryotes, several avenues of evidence suggest that some eukaryotic RNA structures respond to temperature to control translation. For example, many eukaryotic mRNAs contain complex IRES (internal ribosome entry sites) sequences that are believed to form 95 secondary or tertiary structures [Perard et al., 2013, Filbin et al., 2013] that may be responsive to temperature. In addition, eukaryotic cells also contain a heat- shock response pathway, suggesting a mechanism of genetic control by temperature change [Narberhaus et al., 2006]. With the addition of intronic sequences and the compartmental separation of transcription and translation, eukaryotic RNA thermometers could be considerably less simple and conserved than those found in bacteria, confounding detection. Temperature-sensitive structures found in eukaryotic mRNA could theoretically act anywhere in the transcript to alter splicing, translation, transport, degradation, or protein binding. 3.1.3 RNA “thermistors,” putative thermo-sensitive effec- tors of RNA editing A post-transcriptional modification known to involve diverse secondary and tertiary RNA structures is adenosine-to-inosine RNA editing (please see Chapters 1 and 2). We reasoned that there might be a class of eukaryotic RNA thermometer- like structures, “thermistors,” that, instead of controlling post-transcriptional gene expression, rather exert subtle pressure on RNA editing levels (Figure 3.1 D). In this way the transcriptome and proteome could be responsive to temperature via RNA structure. Certain evidence exists in the literature suggesting that RNA editing is tem- perature-sensitive. Savva et al. demonstrated that dADAR auto-editing of the dAdar transcript decreases as temperature increases. As the ratio of edited to unedited dADAR isoforms fine-tunes the global inosinome as well as complex organismal behavior, these data point toward an intriguing role of thermal control in RNA editing [Savva et al., 2012]. Additionally, not only is RNA editing in octopus delayed rectifier K+ channel transcripts responsive to ambient water temperature, but this response appears to be adaptive to the organism [Garrett and Rosenthal, 2012]. 96 To search for RNA thermistors we chose to use Drosophila, a poikilotherm with known robust RNA editing machinery and a well-characterized inosinome [Savva et al., 2012]. Although poikilotherms are at the mercy of their environment as they cannot regulate their own body temperatures, they have evolved physiological processes that allow them to function within the natural temperature range of their habitats [Economos and Lints, 1986]. It is possible that differential RNA editing is one process that allows poikilothermic animals such as Drosophila to function at varying environmental temperature. The above observations led us to undertake an exhaustive survey of 55 Drosophila editing sites across a 20◦ C biologically-relevant temperature range (10◦ C- 30◦ C). Global editing patterns suggest that editing decreases at higher temperatures, but examination of individual editing sites reveal a broad range of temperature- responsive editing patterns. To test the involvement of RNA structures, we repeated the experiment using Drosophila with HR-generated structural mutations in the paralytic editing substrate, revealing slightly altered temperature sensitivity due to changing structure. Our results suggest that RNA editing is sensitive to temperature, and that this response is effected by RNA “thermistors,” temperature-sensitive secondary and tertiary structures that direct editing. 3.2 Results In order to undertake this ambitious experiment, we first needed a better method of RNA editing analysis. Previous methods involved capturing a screen shot of a mixed peak chromatogram, importing it into Adobe Photoshop, and tracing each peak in order to calculate the area under each curve (in pixels). These numbers were then recorded and stored in a spreadsheet for further analysis. This method, while accurate [Rinkevich et al., 2012], is incredibly laborious and time consuming. We 97 therefore partnered with a mathematician, Matthew Reyna (Rensselaer Polytechnic Institute), who designed a program to automate the process. This program, Mixed Peak Image Analysis, requires the user to identify the mixed peak and returns the numerical editing ratio as well as intermediate data from each stage of the procedure (Figure 3.2). The program allows the simultaneous processing of many samples, is incredibly time-efficient, and is an accurate substitution for by-hand data processing (Figure 3.3), which may introduce user bias. A. B. Original C. G, isolated D. G, completed E. G, shaded F. A, isolated G. A, completed H. A, shaded Figure 3.2: Progression of the Mixed Peak Image Analysis program. The program requires the user to select mixed peaks from a chromatogram (e.g. position 890 in A.), and save the original (B.) in an input folder. The program then performs the isolation (C., F.), completion (D., G.), and area calculation (E., H.) steps for the guanosine (green) and adenosine (black) traces. It then returns a .txt file containing calculated editing ratios. We grew flies at 25◦ C before transferring newly-ecclosed animals to incubators held at 10, 20, and 30◦ C. At 20◦ C Drosophila eggs are the most viable [Economos and Lints, 1986], therefore the 20◦ C temperature range chosen is biologically-relevant. After 72 hours of acclimation the animals were frozen at -80◦ C and then processed as described below. The Mixed Peak Image Analysis program was used for all samples. 98 0.6 Editing Ratio (calculated by user) 0.5 0.4 0.3 0.3 0.4 0.5 0.6 Editing Ratio (calculated by program) Figure 3.3: Accuracy of the Mixed Peak Image Analysis program. The program was compared to the standard method of editing calculation, which requires the user to import .tiff files into Adobe Photoshop, trace curves, and record and calculate peak area ratios by hand. The same peaks were analyzed by the program and by hand. Editing ratios are G/G+A. The red line represents best fit, y = 0.9996x, R2 = 0.9987. 99 3.2.1 Temperature changes affect editing of specific tran- scripts To investigate the effect of temperature on editing in specific transcripts we limited our scope to Drosophila melanogaster, in which the inosonome has been well-characterized [Savva et al., 2012]. We investigated 55 editing sites in sixteen different transcripts. The results from each individual site are compiled in Appendix B. Temperature-sensitive editing at individual sites illustrates the full range of possible patterns. For example, editing in the Dopamine/Ecdysteroid receptor transcript sites 1-3 are all very sensitive to temperature: as temperature increases, editing decreases substantially (Figure 3.4 A). However, editing at synaptotagmin- 1 sites 2-4 is largely temperature resistant (Figure 3.4 B). Paralytic sites 1-3 are resistant to temperature between 10 and 20◦ C, but sensitive at 30◦ C (Figure 3.4 C), while editing of site 6 in shab is sensitive to temperature, but increases at higher temperatures (Figure 3.4 D). Notably, shab site 7, located adjacent to site 6, is edited at 100% and is temperature-insensitive within the 20◦ C range studied. 3.2.2 Conservation of editing responsiveness to temperature in Drosophilidae In order to determine whether the observed temperature-dependent editing responses are conserved in other species of Drosophila, we studied editing of several transcripts in five closely-related Drosophilidae species (Figure 3.5). Interestingly, we observed that certain sites, such as complexin site 2 and synaptotagmin-1 site 4, are very tightly regulated not only across temperatures, but also between the species investigated (Figure 3.6). Generally, if an editing site is responsive to changes in temperature in one species, it is also responsive in other species (for example, dAdar auto-editing), although the temperature-dependent 100 A. The Dopamine/Ecdysteroid receptor transcript (dop) sites 1­3 A AN T T T A T CG A AC C C T AG T A A AT T T AT C G A AC C C T AG T A A AT T T AT C G A AC C C T AG T       10!C                   20!C                      30!C    B. synaptotagmin­1 (syt) sites 2­4 A A G A A G AC A A G TN T TG C AAG AAG ACAAG TGT TG C A A G A AG AC A A G T N T TGC       10!C                   20!C                      30!C    C. paralytic (para) sites 1­3 C G A T T A A T TG A N G A C GA T T A A T TGA A G A C A A T T A A T TGA A A A       10!C                   20!C                      30!C    D. shab sites 6 and 7 TG G T AG T C G T G G TG G T C G TG G T G G T C G       10!C                   20!C                      30!C    Figure 3.4: Temperature affects editing in specific transcripts. A. Editing at sites 1-3 in the Dopamine/Ecdysteroid receptor transcript decreases with increasing temperature. B. Editing of synaptotagmin-1 sites 2-4 are temperature-insensitive between 10 and 30◦ C. C. Editing at paralytic sites 1-3 is stable between 10 and 20◦ C, yet decreases dramatically at 30◦ C. D. Editing at shab site 6 is one of the few sites surveyed that increases with rising temperature. In contrast, adjacent site 7 is edited at 100% regardless of temperature. In all figures, black carrots denote editing sites. 101 D. melanogaster D. simulans D. sechellia D. yakuba D. erecta D. ananassae Figure 3.5: Phylogeny of Drosophila species used in this study. D. ananassae represents an outgroup and was not included in the investigation. Based on the Drosophilidae phylogeny from the UCSC Genome Browser [Meyer et al., 2012]. editing trend is not always in the same direction (for example, uncoordinated-13 ). It is possible that the observed temperature-responsiveness of editing is a result of altered RNA structures, perhaps due to single nucleotide changes or polymorphisms within the genes of the Drosophila species studied here. As most structures that direct editing are formed between exonic and intronic sequences, changes in primary sequence often occur in intronic cis elements, leading to alterations in RNA structure and a corresponding change in editing. However, it is notable that both dAdar [Keegan et al., 2005] and unc-13 are editing sites directed by entirely exonic secondary structure, whereas complexin and synaptotagmin-1 [Reenan, 2005] sites are directed by structures comprised of paired exon and intron sequences. One would therefore expect the structures of adar and unc-13 to be under higher sequence, and therefore structural, conservation, leading to highly similar editing at these sites across Drosophilidae. While editing in the dAdar transcript satisfies this prediction, editing in unc-13 is highly variable (Figure 3.6), suggesting that either the unc-13 RNA structure or the specificity of editing machinery has evolved in the species studied, leading to the observed variability. These observations suggest broadly that when editing is tightly controlled across temperature it is also conserved between species. This further suggests that 102 Adar Unc13 1.0 1.0 species 0.8 0.8 yak Editing level (G/G+A) Editing level (G/G+A) sim 0.6 0.6 sec 0.4 0.4 ere 0.2 0.2 mel 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (! C) Temperature (! C) Cpx1 Cpx2 Cpx3 1.0 1.0 1.0 0.8 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 0.0 0.0 0.0 10 15 20 25 30 10 15 20 25 30 10 15 20 25 30 Temperature (! C) Temperature (! C) Temperature (! C) Syt2 Syt3 Syt4 1.0 1.0 1.0 0.8 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 0.0 0.0 0.0 10 15 20 25 30 10 15 20 25 30 10 15 20 25 30 Temperature (! C) Temperature (! C) Temperature (! C) Figure 3.6: A comparison of temperature responsiveness of select editing sites across Drosophilidae. Species color-coding is constant in all graphs. Included are dAdar (Adar), uncoordinated-13 (Unc13), complexin (Cpx) sites 1-3, and synaptotagmin-1 (Syt) sites 2-4. Bars represent standard error. Pink: D. yakuba (yak), green: D. simulans (sim), blue: D. sechellia (sec), orange: D. erecta (ere), black: D. melanogaster (mel). 103 the RNA structures that direct editing at certain sites are highly thermodynamically stable and are also preserved during evolution. Yet structures that are inherently more temperature labile, allowing editing levels to fluctuate with temperature, are less conserved. 3.2.3 The effect of temperature on global RNA editing trends We investigated the global effect of temperature on RNA editing in D. melanogaster. Figure 3.7 independently plots all sites studied by rank at different temperatures. When plotted thusly, it is apparent that while global editing profiles at 10 and 20◦ C appear very similar, global editing at 30◦ C is decreased, suggesting that, at least on a global scale, editing decreases as temperature increases. 1.0 Temperature (! C) 30 20 0.8 10 average editing (G/G+A) 0.6 0.4 0.2 0.0 10 20 30 40 50 rank (temperature­dependent) Figure 3.7: Global effects of temperature on editing levels, independently ranked. Each site is ranked separately at each temperature. Bars represent standard error. 104 When the same data is instead presented as in Figure 3.8, in which all sites are ranked only at 20◦ C, the independent axis then represents individual editing sites. From this presentation of the data it is apparent that, while the global trend may suggest decreased editing at 30◦ C, each individual site behaves very differently with respect to its temperature profile. 1.0 Temperature (! C) 30 20 shab6 0.8 10 average editing (G/G+A) 0.6 dop1 0.4 caID4 caID3 0.2 para3 para1 para2 0.0 10 20 30 40 50 all sites ranked at 20!C  Figure 3.8: Global effects of temperature on editing levels, ranked at 20◦ C. Sites are ranked by editing level at 20◦ C and this rank is preserved for editing at 10◦ C and 30◦ C. The temperature range at each site is represented by vertical gray dotted lines. Notable sites are identified. Bars represent standard error. 3.2.4 HR-generated RNA structural mutants Temperature, therefore, impacts editing at individual sites as well as globally. We hypothesized that these editing changes are brought about by the melting of RNA structures that direct editing. At higher temperatures these specific structures occupy a smaller proportion of the structural space, leading to a decrease in editing. Similarly, at lower temperatures editing would be predicted to increase as these structures become more stable. 105 a LoxP control b DCS delete c DCS zip Site 1 Site 2 276 nt 265 nt Site 3 1.0 1.0 1.0 0.8 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 0.0 0.0 0.0 10 15 20 25 30 10 15 20 25 30 10 15 20 25 30 Temperature (!!C) Temperature (!!C) Temperature (!!C) d Loop > " e Dock > "’ f Loop/dock > "/"’ 1.0 1.0 1.0 0.8 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 0.0 0.0 0.0 10 15 20 25 30 10 15 20 25 30 10 15 20 25 30 Temperature (!!C) Temperature (!!C) Temperature (!!C) Figure 3.9: Temperature effects of HR-generated mutations in paralytic. The wild type paralytic transcript folds into a complex pseudoknotted structure to direct editing at three sites (see Chapter 2). (a) In the loxP control animal, all three sites are decreased at 30◦ C, similar to the temperature- responsiveness of wild type paralytic (Figure 3.4 C). (b) The “DCS delete” mutation maintains a similar pattern, although all three sites are decreased overall, as is seen with this mutation in Figure 2.6 d. (c) Similarly, the “DCS zip” mutation, which increases secondary structure around the splice donor and is known to universally increase editing at all three sites (Figure 2.6 d) displays a temperature-dependent pattern similar to the loxP control, although curiously site 1 appears temperature-resistant in this mutant. The “Loop > α” (d) and “Dock > α’ ” mutations, which selectively abolish editing at site 1 (Figure 2.12 e), are temperature-responsive at sites 2 and 3 similar to the loxP control (a). Similarly, the “Loop/Dock > α/α’,” which rescues editing at site 1, appears to have a similar temperature profile to that seen in the loxP control. In all graphs, site 1 is red, site 2 is blue, and site 3 is orange. Bars represent standard error. Relevant HR-introduced mutations are shown above graphs. Deleted nucleotides are noted by dashed gray lines, while altered nucleotides are shown in green. Blue represents exon, while black denotes intronic sequence. 106 We tested the editing response of several HR-generated RNA structural mutations in the para transcript (see Chapter 2) at varying temperatures, compared to the loxP control (Figure 3.9 a). For example, the “DCS delete” mutation, which is predicted to decrease secondary structure, decreases editing at all three para editing sites, while the “DCS zip” mutation, which increases secondary structure, universally increases editing (Figure 2.6 d). The temperature-responsive editing profile of these structural mutations appears to be a product of both temperature and structure: the DCS zip mutation displays decreased editing at all three sites in response to elevated temperature (Figure 3.9 b), yet editing is still decreased from the loxP control (Figure 3.9 a). Similarly, the DCS zip mutation also shows decreased editing at 30◦ C, yet editing at all three sites is higher than in the loxP control (Figure 3.9 c). In addition, we tested the temperature-responsiveness of the paralytic pseu- doknot mutations (Figure 2.12). The individual pseudoknot mutations, “Loop > α” and “Dock > α’,” both selectively abolish editing at site 1, while preserving editing at sites 2 and 3. The double mutation, “Loop/Dock > α/α’,” which is predicted to restore the tertiary psedudoknot structure, rescues editing at site 1. These mutations reveal temperature sensitivity similar to that seen in the loxP control (Figure 3.9 d-f ), suggesting again that for these particular editing sites, editing levels are affected by both RNA structure and temperature. In this case the effect appears additive. 3.3 Discussion RNA structures possess the ability to bind molecular ligands with high specificity, reg- ulating diverse cellular tasks such as transcriptional regulation and protein synthesis. Structures are inexorably linked to RNA splicing. For example, the Drosophila Down Syndrome Cell Adhesion Molecule (Dscam) gene contains 48 mutually exclusive ex- ons in cluster 6. Exon selection is regulated by competing RNA secondary structures, 107 and the thermodynamic strength of the structure governing exon selection is directly correlated with frequency of exon inclusion [May et al., 2011]. RNA structure also mediates splicing in Saccharomyces cerevisiae, and alternative 3’ splice site selection, governed by RNA secondary structure, is at least partially regulated by temperature [Plass et al., 2012]. RNA thermometers, involved in transcriptional regulation in prokaryotes, are now not the only RNA structures that are temperature-responsive, though they may easily be the best characterized. RNA editing, another post-transcriptional process dictated by RNA secondary and tertiary structures (please see Chapter 2), may also be sensitive to temperature changes, especially in a poikilothermic animal such as Drosophila. We therefore investigated the temperature-responsiveness of 55 well- characterized Drosophila editing sites [Savva et al., 2012] across a biologically-relevant 20◦ C range [Dillon et al., 2009]. Our analysis revealed that editing is globally decreased at 30◦ C (Figure 3.7), suggesting that this temperature may affect the RNA structures involved in editing. As D. melanogaster show fairly constant viability and fertility between approximately 12 and 30◦ C [Hoffman, 2010], we therefore suggest that this trend may represent an adaptive cellular response rather than a breakdown in editing efficiency. Indeed, the sensitivity of certain editing sites to elevated temperature is in the opposite direction of this global trend (Figure 3.4 D). Figure 3.8 suggests that each individual site behaves differently in response to temperature changes, an observation consistent with the data presented in Figure 3.4. Also congruent is the previous observation that certain sites are sensitive to dADAR isoform and concentration of dADAR protein [Jepson et al., 2011]. In addition, all known editing sites are encompassed in duplex RNA structure, although other than double-strandedness, these sites have no identified common structural property. This suggests that sites may behave differently because variable RNA structures act as 108 mechanistically distinct “thermistors,” depending on the function of the editing site residue. This hypothesis is supported by the observation that neighboring sites within a transcript, which are predicted to be involved in the same governing RNA struc- ture, often show similar temperature profiles (Figure 3.4 A-C). However, there are other sites (Figure 3.4 D) in which adjacent sites behave differently with respect to temperature. This could indicate that these sites are “more important” than labile sites. Indeed, the top 25% of sites ranked in Figure 3.8 are largely insensitive to temperature (with a few noted exceptions). This suggests that sites that are edited close to 100% are edited at this level regardless of temperature, perhaps because these structures are the most thermodynamically stable and possible alternative structures are much less likely to form within the investigated temperature range. In support, Tian et al. observed that while editing of constructs representing the wild type Gabra3 I/M site is insensitive to temperature, synthetic constructs containing structural mutations are temperature-sensitive, suggesting an evolutionary advantage to temperature-insensitivity at this particular site [Tian et al., 2011]. The species represented in the above analysis, although closely related (Figure 3.5), represent diverse geographical (and temperature) ranges. While D. melanogaster and D. simulans are cosmopolitan species, D. yakuba is found in savannah climes, D. erecta is found in west central Africa, and D. sechellia was previously confined to the Seychelles islands [Lachaise et al., 1988]. Our investigation into temperature sensitivity of editing across species was brief. Nevertheless, it appears that when editing level is tightly regulated across temperature, it is similarly regulated across species, suggesting conservation of stable RNA structures. However, editing sites that are more temperature-sensitive are also more variable across species (Figure 3.6), suggesting that labile RNA structures are under less selective pressure. However, drawing broad conclusions based on only eight editing sites in four transcripts is 109 unwise. To test our hypothesis that the responsiveness of RNA editing to tempera- ture is mediated by RNA structure, we tested several HR-generated paralytic RNA structural mutations at different temperatures. These mutations are described in detail in Chapter 2 and represent bidirectional changes to both a secondary (DCS) and tertiary (pseudoknot) structural element. The paralytic sites investigated in this study are responsive to temperature primarily at 30◦ C, at which they drastically decrease (Figure 3.4 C). Interestingly, in this particular transcript, editing level appears to be responsive to both structure and temperature in an additive manner (Figure 3.9), suggesting levels of regulation at both the molecular and abiotic levels. However, this observation may be specific to these particular editing sites in paralytic and is not necessarily representative of global trends. 3.4 Methods Drosophila stocks We raised wild type (Canton-S) Drosophila at 25◦ C in humidity-controlled incubators with 12-hour light/dark cycles. We collected animals daily as they ecclosed and shifted newly emerged animals to one of three temperatures: 10, 20 and 30◦ C, using temperature, humidity, and light-controlled (12-hour light/dark) incubators. From experience we know that flies can live within this temperature range for 72 hours, although they are most often subjected to temperatures between 20 and 30◦ C. We allowed these animals to remain in the incubators for 72 hours, long enough for the RNA editing machinery to adjust to the temperature, and then immediately froze flies from this population for long-term storage at -80◦ C. Additionally, we raised and collected D. erecta, D. sechellia, D. simulans, and D. yakuba in the same manner as D. melanogaster above. We did the same with 110 mutant D. melanogaster in which the endogenous dAdar gene had been HA-tagged using homologous recombination. Drosophila with engineered mutations in paralytic were generated through homologous recombination (please see Chapter 1). The paralytic mutations are extensively described in Chapter 2. RNA editing analyses For each unique species/genotype/temperature sample, we extracted RNA from male fly heads (N = 15 − 20) using Tri Reagent (Molecular Research Center, Inc.). We chose only males to control for sex-related editing changes. For each sample, three separately-extracted RNA biological replicates were used. We amplified cDNAs via RT-PCR using random primers, performed PCR using target-specific primers (2 per technical PCR replicates per RNA sample) and electrophoresed samples on agarose gels. We cleaned the PCR products using Wizard Gel and PCR Cleanup Kits (Promega) and detected editing by Sanger sequencing (University of Wisconsin Biotechnology Center). For species-specific editing, we designed primers to regions of high sequence conservation. Analysis of editing using Mixed Peak Imaging program Our collaborator, Matthew Reyna (RPI) wrote a program to calculate the ratio of mixed peak areas. This program represents a rapid method of editing calculation. It requires the user to identify mixed peaks, which the program analyzes to give editing ratios. The program also returns data on each stage of the calculation (Figure 3.2). Briefly, the algorithm can be broken down into five main tasks, where only a single task is provided by the user. The computer performs the final four tasks and provides feedback for each task and warnings when a task may provide suspect results. After the user identifies a mixed peak for analysis, the program isolates each curve by discarding information corresponding to or influenced by adjacent peaks. The program then completes each curve by performing a linear interpolation and extrapolation to recover missing parts of each curve. The program finally computes 111 the area under each curve and calculates the mixed peak ratio of the curves, a number between 0 and 1, which corresponds to the editing level (G/G+A). Chapter Four Mlenap−ts, an RNA helicase, interacts with the paralytic transcript 113 The TALEN work for this chapter was carried out in collaboration with Ms. Kasia Sierzputowska, an undergraduate student at Brown University. Ms. Sierzputowska and I crafted the figures with help from Ms. Marianna Neubauer. I wrote the manuscript and Ms. Sierzputowska provided editorial feedback. Thanks to Dr. James Gagnon (Harvard University) for TALEN design advice. Thanks to Dr. Koen Venken (Baylor College of Medicine, Department of Biochemistry and Molecular Biology) for the gift of a 3X-P3- and EGFP-containing plasmid. This research was funded in part through a Brown Research Grant to Ms. Kasia Sierzputowska. 114 In theory, RNA editing, which often relies on connectivity between an exonic and intronic sequence, is linked to splicing (see Chapter 1). However, until 2000 this relationship was only hypothesized [Seeburg, 2000]. Reenan et al. discovered that the napts (no action potential-temperature sensitive) mutation in the Maleless (Mle) RNA helicase causes a “splicing catastrophe” in the paralytic (para) transcript [Reenan et al., 2000]. This evidence suggests that the double-stranded RNA structures required for RNA editing must be resolved prior to efficient splicing [Seeburg, 2000]. Rent high-throughput sequencing efforts reveal that co-transcriptional RNA editing is widespread in Drosophila [Rodriguez et al., 2012], suggesting that defining the relationship between editing and splicing in para may reveal trends of a broader post-translational relationship. 4.1 The DExD/H box RNA helicase family Maleless, along with its mammalian homolog, RNA helicase A (RHA), is a member of the DExD/H box RNA helicase family. Members of this enzyme family share a conserved core containing motifs involved in ATP hydrolysis and RNA binding, yet their N- and C-terminal domains vary widely, suggesting diverse functionality. DExD/H helicases are associated with almost all aspects of RNA metabolism, including synthesis, pre-mRNA processing, turnover, and folding [Fuller-Pace, 2000]. Evidence links RHA directly to RNA polymerase II [Nakajima et al., 1997, Anderson et al., 1998], suggesting that RHA specifically is involved in transcription. Interest- ingly, the ATPase/helicase activity of these enzymes does not seem to be involved in their function in transcription, and so they may play multiple roles in the cell [Fuller-Pace, 2000]. A dual role for Mle was uncovered due to the napts mutation. 115 4.1.1 Dual functionality of the Drosophila Maleless helicase While Mle appears to be a general dsRNA helicase, it is best understood as an important player in the Drosophila dosage compensation pathway. Mle is one of five integral protein components of the Male Specific Lethal (MSL) complex, which is present only in males and targets the X-chromosome for transcriptional upregulation. The MSL complex in males lacking this helicase cannot properly target the X chromosome for upregulation, leading to lethality [Kuroda et al., 1991]. Mle is also expressed in females, hinting at functionality outside of dosage compensation, yet females lacking Mle are viable and without discernible phenotype. Further evidence for the dual functionality of Mle was provided when Reenan et al. discovered that the Mlenap−ts mutant form of the helicase causes aberrant para transcript splicing: constitutive exon 19 and often additional upstream exons are spliced out of the mature transcript (Figure 1.16 a), leading to a temperature-sensitive paralytic phenotype. While the helicase/ATPase activity of Mle is required for its function in dosage compensation [Lee et al., 1997, Morra et al., 2008], mutations in these domains do not appear to affect para splicing. Conversely, the napts phenotype, which is likely due to a transversion C to G mutation in the Walker A NTP-binding motif (GxGKTT to GxGKTS) [Zhang and Grosse, 2004], does not result in male- specific lethality or affect Mle targeting in females (Figure 1.15) or MSL complex targeting in males. 4.1.2 A model for Mlenap−ts -para interaction The current hypothesis for Mlenap−ts -para interaction involves nap-ts as a recessive gain of function allele. It was hypothesized that the mutant helicase either cannot dissociate from the secondary structure around para exon 19 or structurally blocks the splicing machinery from recognizing the splice donor [Reenan et al., 2000] (Figure 1.16). We reasoned that mutations in the para pre-mRNA secondary and tertiary 116 structures could rescue or antagonize the splicing and behavioral defects resulting from Mlenap−ts . nap-ts a WT para 5’ aberrant splicing 18 20 3’ b HP delete 5’ 3’ ? OR !" c DCS delete 5’ ? #" correct splicing 18 19 20 3’ OR 5’ d HP loop > ! ? 3’ e DCS zip 5’ ? Synthetic Lethality 18 20 3’ Figure 4.1: Model for Mlenap−ts interaction with para structural mutants. (a) Mlenap−ts (orange) causes a “splicing catastrophe” in wild type para background, causing exon skipping. (b-d) It is possible that para structural mutants that decrease secondary structure, such as the “HP delete” (b) or “DCS delete” (c), or those that alter tertiary structure, including the HP “Loop > α” (d), will rescue the napts splicing phenotype because the mutant helicase can now resolve the RNA secondary/tertiary structures prior to splicing. (e) The mlenap−ts mutation may be functionally lethal in the context of “DCS zip” allele of para, as increased secondary structure will further interfere with proper splicing, leading to even fewer full-length Para sodium channels. For example, structural mutations that decrease secondary structure, such as the para HP, DCS, or ECS delete mutations, or those that alter tertiary structure, 117 including the HP “Loop > α” and “Dock > α’ ” mutations, might not require the Mlenap−ts helicase for structural resolution before splicing (Figure 4.1 b-d). Alternatively, the structural changes around para exon 19 might not attract the mutant helicase, leading to structural resolution by other mechanisms and a rescue of the molecular and phenotypic defects. In contrast, the para DCS zip allele increases secondary structure around the splice donor and even in a wild type Mle background this allele generates aberrant splicing and a temperature-senstive paralysis phenotype (Figure 2.8). We reasoned that the homo- or hemizygous DCS zip allele in combination with homozygous mlenap−ts might result in synthetic lethality (Figure 4.1 e). Animals homozygous for both the DCS zip and mlenap−ts alleles are hypothesized to produce even fewer full-length sodium channel transcripts compared to either single mutant. 4.2 Results and Discussion 4.2.1 The Mlenap−ts helicase is rescued by the presence of the loxP in the paralytic pre-mRNA We therefore combined our para structural mutations generated through homologous recombination (Chapter 2) with the mlenap−ts mutation in flies. Surprisingly, we discovered that every HR-generated para mutation, including the otherwise wild type loxP control, resulted in a rescue of the napts splicing phenotype (Figure 4.2). Surprised by this observation, we turned to the loxP remnant, a 76 basepair intronic insertion that remains in the genome after the HR process [Staber et al., 2011]. The remnant includes not only the loxP sequence (34 bp), but also flanking 6-frame stop codons (14 bp each), and AscI (8 bp) and Acc65I (6 bp) restriction digestion sites required for arm subcloning (Figure 4.6). The core of the remnant is therefore palindromic and is predicted to form a perfect hairpin structure in the 118 1 2 3 4 5 6 7 8 9 10 1000bp Figure 4.2: The loxP suppresses the skipping phenotype in a mlenap−ts background. Lane 1: 1 Kb ladder; Lane 2: mlenap−ts ; Lane 3: para-loxP ; mlenap−ts ; Lane 4: para-loxP; Lane 5: para-ECS delete ; mlenap−ts ; Lane 6: para-ECS delete; Lane 7: para-DCS delete ; mlenap−ts ; Lane 8: para-DCS delete; Lane 9: Canton-S; Lane 10: 100 bp ladder. Primers are given in Table 4.2 para intron between the DCS and Hairpin (Figure 4.3). We hypothesized that the addition of the loxP hairpin results in the rescue of the splicing catastrophe in the presence of Mlenap−ts . In an effort to disrupt the loxP structure (Figure 4.3), we first repeatedly targeted the para loxP with constitutively-expressed Cre recombinase. However, it became apparent that Cre preserves the integrity of the loxP to a striking degree. This observation led us to search for alternative molecular strategies for targeted loxP disruption. 4.2.2 Transcription Activator-Like Effector Nucleases (TAL- ENs) TALENs (Transcription Activator-Like Effector Nucleases) are a novel genome editing technology that presents an ideal tool for loxP disruption. Recently, two back-to-back Science papers broke the cipher that governs DNA sequence recognition by TAL 119 5’ CTA GA GA CTA TCT GTC GA TAG TCA ATA GA AC TCT T TC AT T GTA GA TA G AG CAT CAT ATACAT 3’ Figure 4.3: The loxP remnant may form a perfect duplex in the paralytic transcript. Exon 19 is depicted in blue with three edited adenosines (red). The downstream intron with conserved cis elements is shown in black. Structural elements are consistent with those discussed in Chapter 2. The loxP remnant, including two 6-frame stop codons (orange) and the loxP site (red) may form a perfect duplex in the pre-mRNA structure. (Transcription Activator-Like) Effector proteins [Moscou and Bogdanove, 2009, Boch et al., 2009]. These transcription factors, which are injected into plant cells by bacterial pathogens in order to manipulate cellular processes, contain highly-conserved repeat arrays. Two hyper-variable residues in each repeat, repeat variable diresidues (RVDs), are responsible for DNA sequence recognition (Figure 4.4). Soon after this discovery, TAL DNA recognition domains were fused to the Fok1 nuclease, creating TALENs (TALE Nucleases, Figure 4.4) that could be employed for genome cleavage [Joung and Sander, 2013]. Unlike zinc-finger nucleases, TALENs are highly sequence specific, have few off-target effects, and are very inexpensive to design and synthesize. Further, TALENs are significantly more mutagenic than zinc- finger nucleases [Chen et al., 2013], predicting their widespread use in biotechnology. Relatively few rules govern the choice of target DNA sequence [Cermak et al., 2011], and kits are widely available to facilitate assembly. TALENs have been employed in most model systems, including mouse [Sung et al., 2013], human cell lines [Kim et al., 2013, Ding et al., 2013], Xenopus laevis [Sakuma et al., 2013], and Danio rerio [Bedell et al., 2012, Zu et al., 2013], as well as in non-model animals such as cow and pig [Carlson et al., 2012] and medaka 120 Figure 4.4: Structure of a Transcription Activator-Like Effector Nuclease. A TALEN consists of an N-terminal nuclear localization signal (NLS), a variable number of repeats, and a C-terminal Fok1 nuclease (pink). Residues 12 and 13 from each repeat represent the repeat-variable di-residue (RVD), and via the RVD each 34 amino acid repeat targets a single nucleotide. The four most common RVDs each target a separate nucleotide in the DNA. Image courtesy of Marianna Neubauer and Kasia Sierzputowska. [Ansai et al., 2013], suggesting that TALENs are useful tools in a variety of biological model systems. The Jiao group efficiently targeted the Drosophila yellow gene for disruption using transiently-expressed TALENs [Liu et al., 2012], suggesting TALENs are also operative and useful in the fly. However, to date, this study remains the sole experimental publication employing TALENs in Drosophila, although a second study evaluated TALEN use in Drosophila as a proof of utility [Sakuma et al., 2013]. We chose to synthesize TALENs targeted to the loxP remnant. Because the remnant is palindromic, we were able to design a single TALEN target sequence that provides correct target and spacer lengths, allowing Fok1 homodimerization and targeted cleavage within the loxP (Figure 4.5). 4.2.3 TALEN design To date, all studies using TALENs have injected TALEN-encoding mRNA into cells (or syncytial embryos, in the case of Drosophila), allowing for transient TALEN 121 Figure 4.5: Targeting the loxP for disruption using TALENs. Because the loxP remnant (yel- low/green) is palindromic, we targeted a single TALEN to this region, creating double-stranded breaks in the loxP when the Fok1 dimerizes. Image courtesy of Marianna Neubauer and Kasia Sierzputowska. expression. For several reasons we choose to express the loxP-targeting TALEN (“loxP-TALEN”) directly from the Drosophila genome. First, traditional elegant Drosophila genetics often involves crossing flies to generate appropriate in vivo combinations of alleles. In keeping with the litany of genetic tools used in the fly, a genomically-expressed TALEN could easily be crossed in and out of a population. Secondly, stable genomic expression of the loxP-TALEN would allow researchers using homologous recombination in the fly to cross in the TALEN, if loxP disruption is desired. In this way, the loxP-TALEN would become a tool extending the use of HR in Drosophila. This would also mean that laboratories unequipped for embryo injection could keep a stable stock of loxP-TALEN animals without the need to constantly inject loxP-TALEN mRNA into different Drosophila lines. Finally, the stable genomic expression of a TALEN from a genome has yet to be characterized. The loxP, as a foreign DNA remnant, presents a unique opportunity to safely express a TALEN from the Drosophila genome, allowing us to characterize any unintended consequences from genomic expression without targeting endogenous sequence. When designing our TALEN strategy we decided to take advantage of the HR process. To use the loxP-TALEN in “post-Cre” individuals, whose genomes contain a single loxP, would require genotyping. Instead, we chose to use “pre-Cre” animals, whose genomes still contain the mini-white phenotypic marker flanked by two identical loxP sites (Figure 1.17). In this experimental design the loxP-TALEN is predicted to target both loxP sites, removing the mini-white marker and resulting in an easily-visible mosaic-eyed animal while simultaneously disrupting the loxP 122 integrity. In addition, TALENs have yet to be used to target an entire gene for removal rather than disruption. 4.2.4 TALEN vector construction Because of our above experimental design, we constructed a vector using the pUAST backbone. The vector contains the TALEN gene, assembled using the AddGene Golden Gate TALEN Assembly Kit [Cermak et al., 2011], an EGFP phenotypic marker, and P-element ends for genomic insertion (Figure 4.7). First the pUAST vector was modified from containing the mini-white phenotypic marker to containing the EGFP marker under control of an eye-specific 3X-P3 promoter. The 3X-P3 and EGFP sequences were kindly donated by Dr. Koen Venken (Baylor College of Medicine, Department of Biochemistry and Molecular Biology). These sequences were amplified from the donated plasmid using the primers indicated in Table 4.1 and inserted into the pUAST backbone in place of the mini-white using PciI. Start  (0) End   (132) P 5 C C GAT GGAAT T T C T T GC T C C GC T C T GGC GC GC C C T AGAC T AGT C T AGAT AAC T T C GT AT AGC AT AC AT T AT AC GAAGT T AT C T AGAC T AGT C T AGGGT AC C GAGAAGAT AC T AT GT AT T T T GGT AGC T T AAA 3 3 GGC T AC C T T AAAGAAC GAGGC GAGAC C GC GC GGGAT C T GAT C AGAT C T AT T GAAGC AT AT C GT AT GT AAT AT GC T T C AAT AGAT C T GAT C AGAT C C C AT GGC T C T T C T AT GAT AC AT AAAAC C AT C GAAT T T 5 P Genomic AscI 6 Frame Stop loxP 6 Frame Stop Acc65I Start of mini­white TALEN spacer TALEN Figure 4.6: The loxP remnant and TALEN targeting sequence. The loxP remnant consists of the loxP sequence (green, 34 bp), flanking 6-frame stop codons (orange, 14 bp each), and AscI (pink, 8 bp) and Acc65I (lavender, 6 bp) sites. The targeted TALEN sequence (purple) spans the 6-frame stop and loxP sequences. The remnant upstream of the mini-white gene (red) is depicted. The repetitive domain of the TALEN was designed to target the loxP remnant (Figure 4.6) The TAL repeats were assembled in the “pTAL2” backbone vector included in the AddGene Golden Gate TALEN Kit. The sequence encoding the Fok1 nuclease was captured from the provided TAL “pTAL3-His” plasmid using primers indicated in Table 4.1. The Fok1 was cloned into pTAL2 after the C-terminal TAL domain using BbvCI. The entire TALEN gene was then inserted into the EGFP pUAST vector using EcoRI. The TALEN gene was inserted under control of a UAS- 123 HSP 70 promoter and so will express constitutively in the presence of a Tublin-GAL4 driver. EcoRI  (1) TALEN_in_EGFP_pUAST BbvCI  (3025) 12,121 bp BbvCI  (3729) EcoRI  (3933) PciI  (5263) (6487)  PciI Figure 4.7: Composition of the TALEN plasmid. The pUAST backbone encodes P-element ends (peach) for genomic insertion, and an EGFP marker (green) under control of the 3X-P3 eye-specific promoter (blue). The complete TALEN sequence (orange, yellow, and purple) is under control of a UAS-HSP 70 promoter (green and white). The EGFP was cloned into the pUAST backbone using PciI, while the TALEN sequence was added using EcoRI after BbvCI insertion of the Fok1 sequence. 4.3 Future directions Once the loxP-TALEN EGFP-pUAST plasmid validation is complete it will be introduced into the Drosophila genome by injection (Genetic Services, Inc.). Progeny of injected flies will be scored for EGFP eyes, indicating successful random P-element 124 Table 4.1: Primers used in the synthesis of the loxP-TALEN EGFP vector. Underlines represent restriction sites added by primers. Forward (F) and Reverse (R) Primers R. Enzyme Fok1 F: CAAACACCGGATCAGGCGTCTTTG BbvCI R: GGCTGAGGATTCAATCTTAAGAAAC EGFP F: TACATGTCAAACATGAGAATTGGTCGAG PciI R: TACATGTTATTGATCATAATCAGCCATACC genomic insertion. These animals will be mapped and stocked and used in the TALEN crossing scheme proposed in Figure 4.8. This scheme will introduce the tubulin-GAL4 driver, the UAS-TALEN, and loxP-containing para alleles (from “pre-Cre” flies) into the same animal. These flies will express the EGFP phenotypic marker in all eye ommatidia, while the action of the loxP-TALEN will excise the mini-white marker from some ommatidia, resulting in eyes that are mosaic for mini-white expression. As loxP-TALEN expression will be driven by a ubiquitous tubulin driver, it will act on the genome not only in the eye, but also in the germline. F2 progeny from the crossing scheme are predicted to have either red eyes, indicating that the TALEN did not act on the dual loxP sequences to excise the mini-white marker, or white eyes, indicating the intended action of the TALEN (Figure 4.8). We hypothesize that TALENs may be effectively expressed from the Drosophila genome. This is the first study in which TALENs are not transiently expressed through mRNA injection, and so will provide an alternative method for TALEN introduction into an important model organism. Transient TALEN expression results in various imprecise indels and nucleotide additions to the targeted sequence [Joung and Sander, 2013]. It appears that TALENs continue to bind, cut, and rebind until all or most of their target sequence is deleted (Dr. James Gagnon, personal communication), resulting in somatic mosaic individuals. However, our strategy of driving TALEN expression from the genome followed by isolation of F2 progeny (Figure 4.8) will result in animals that each represent a single allele generated from TALEN targeting. Therefore, characterizing TALEN action will be as simple as genotyping F2 flies. 125 + ; UAS - TALEN ; + para - loxP ; ; tub - Gal4 Y CyO TM6 X para - loxP TM3 mini-white para - loxP ; UAS - TALEN ; tub - GAL4 Y + TM6 mini-white and para - ? Y Figure 4.8: Proposed TALEN crossing scheme. “Pre-Cre” flies will be used because the para locus (blue) contains a mini-white gene (red) flanked by target loxP sites (orange). We can then score for the loss of red pigment, indicative of TALEN action. The scheme depicts a TALEN transgene on chromosome 2, followed by the EGFP adult eye marker, but TALENs on the X or on the 3rd could be used in similar crosses. A ubiquitous driver, such as tubulin-GAL4 on chromosome 3, drives TALEN expression in all cells throughout development, resulting in animals mosaic for mini-white expression. X-chromosomes from the F1 generation are isolated by crossing to an Fm7 balancer line (not shown), and F2 progeny lacking the red eye color will then be balanced and stocked for analysis. Image adapted from Kasia Sierzputowska. 5’ CTA GA GA CTA CTA CTA TCT GTC GA GA GA TAG GA CTA GA CTA TCA ATA TCT GT TC G C GA AC GA CTA TG ATC TA TCT T TC TCA GA AT T GTA GA TAA GA GA TA G TCT C AG AT T TT CAT CATA CG ATCAT 3’ ? ? suppression of splicing phenotype nap-ts splicing phenotype 18 19 20 18 20 Figure 4.9: TALEN targeting of the loxP will generate animals with various degrees of loxP integrity. The intact loxP suppresses the splicing phenotype seen in Mlenap−ts animals, while various loxP remnant deletions will allow us to map how much of the loxP is required for phenotype suppression. 126 This strategy will also generate animals whose para loci represent various degrees of loxP integrity (Figure 4.9). This will provide two benefits. First, it will allow us to map the loxP location that interferes with the Mlenap−ts mutant helicase, possibly through direct binding. Secondly, we may target the loxP in animals containing para intronic mutations generated through HR. We may select loxP deletions that we know from control individuals do not suppress the splicing phenotype. We may then assay the interaction of the mutant helicase with the HR-generated structural mutants, without the interference of the intact loxP. In these experiments, we hypothesize that mutations that disrupt the wild type para structure (see Chapter 2) will rescue proper para splicing, while the DCS zip mutation, which results in an aberrant splicing pattern even in the presence of wild type Mle, will be synthetically lethal in the presence of Mlenap−ts (Figure 4.1). 4.4 Methods RNA splicing analysis RNA was isolated and reverse-transcribed as in Chapter 2. cDNA was amplified via PCR using primers specified in Table 4.2 Table 4.2: Primers used in the mlenap−ts splicing analysis. Location Forward (F) and Reverse (R) primers Constit. exon 16 F:CCTTCTTCTTGGCCACCGTTGTCATCGGC Alt. exon “N” R: CCTCATGCCCTGCATACGGGACATGGCAC TALEN construction The TALEN was constructed using the Golden Gate TALEN Kit available through AddGene. The repetitive region was designed to target the sequence AGTCTA- GATAACTTC, as this sequence satisfies targeting guidelines [Cermak et al., 2011], and is present as a palindrome in the loxP remnant (see Figure 4.5), allowing us to synthesize only a single TALEN (Figure 4.6). 127 Due to kit design, backbone plasmid pTAL3-His contained the required Fok1 sequence and no convenient subcloning restriction sites, while backbone plasmid pTAL2 contained convenient EcoRI subcloning sites but no Fok1 sequence [Cermak et al., 2011]. We therefore amplified the Fok1 sequence from pTAL3-His using the primers indicated in Table 4.1. After the TAL repeats were assembled in the pTAL2 vector, the Fok1 sequence was added using BbvCI and T4 DNA ligase (Promega). The successful addition of Fok1 was verified by PCR and Sanger sequencing. EGFP-pUAST construction We replaced the majority of the mini-white gene in the pUAST P-element vector with an EGFP gene under control of a 3X-P3 promoter. The plasmid supplying the 3X-P3 and EGFP sequences was kindly donated by Dr. Koen Venken (Baylor College of Medicine, Department of Biochemistry and Molecular Biology). The relevant fragment was amplified from the supplied vector using the primers indicated in Table 4.1. The 3X-P3 and EGFP sequences were subcloned into the pUAST backbone using PciI and T4 DNA ligase. The addition of the EGFP was verified by PCR and Sanger sequencing. Vector assembly The complete TALEN fragment was subcloned out of pTAL2 and into EGFP-pUAST using EcoRI and T4 DNA ligase. Bacterial strains and transformation Vectors were grown in One Shot TOP10 Competent Cells (Invitrogen). Cells were transformed according to the protocol supplied with the cells. Colonies were grown on LB agar plates containing the appropriate antibiotic at each stage of vector construction [Cermak et al., 2011]. Several colonies were picked and grown in 3 mL of liquid LB medium containing the appropriate antibiotic. Plasmids were isolated from cultures with the PureYield Plasmid Miniprep system (Promega). Chapter Five Conclusions and Future Directions 129 5.0.1 RNA editing by ADARs Post-transcriptional modifications, such as adenosine-to-inosine RNA editing and alternative splicing, diversify the proteome while limiting the necessary size of the genome. Although splicing globally rearranges existing information within the transcript, the conserved process of RNA editing recodes the message through single nucleotide changes, often at very specific transcript locations. Because inosine is interpreted as guanosine by the cellular machineries, editing effectively results in the substitution of a guanosine for an adenosine in the primary RNA sequence. A highly conserved group of enzymes, the adenosine deaminase acting on RNA (ADAR) family, mediates this reaction and is highly expressed in the metazoan nervous system [Savva et al., 2012]. Mutation or deletion of ADAR genes results in striking phenotypes in multiple model organisms, including seizure episodes, extreme uncoordination, and neurodegeneration. Neuronal transcripts are specifically targeted for editing. These include mRNAs of proteins involved in electrical and chemical neurotransmission, including pre-synaptic release machineries, and voltage- and ligand-gated ion channels. One such transcript is paralytic, the only functional voltage-gated sodium channel gene in the Drosophila genome [Ganetzky, 1984]. While mammalian genomes contain nine sodium channel genes, insects achieve channel diversity through post-transcriptional modification of the single channel transcript [Goldin, 2002]. 5.0.2 RNA editing of the paralytic transcript We therefore investigated RNA editing in Drosophila paralytic to probe the extent of the RNA structures directing editing. Using homologous recombination, we surgically altered conserved cis elements associated with a cluster of ADAR modification sites within the endogenous para gene. In addition to the local requirement for a central imperfect RNA duplex containing the modified adenosines, we demonstrated that 130 a secondary RNA duplex containing the splice donor significantly modulates RNA editing. When this structure is destroyed, editing decreases, perhaps due to more rapid intron removal, while a subtle non-coding mutation, extending base pairing of this accessory helix, is predicted to stabilize the secondary structure and leads to higher levels of editing. This mutation also confers significant temperature-sensitive phenotypic consequences via effects on splicing, suggesting a complex relationship between these post-transcriptional processes. Through mutation/counter-mutation, we also uncovered a highly conserved intronic long-range tertiary pseudoknot that is absolutely required for deamination of one particular adenosine in the central duplex. Replacement of this tertiary pseudoknot with sequences from a Group II self-splicing RNA kissing loop results in a functional structure that efficiently directs editing. Our results demonstrate that complex RNA tertiary structures, which may be difficult to predict computationally, form in vivo and can regulate RNA editing events. These structures may assemble at a considerable distance from the edited adenosines [Rieder et al., 2013]. 5.0.3 The complex relationship between editing and splicing Precise control of editing is dictated by duplex structures in the RNA, often formed between the exonic region surrounding the editing site and cis regulatory elements localized to a nearby intron, suggesting that editing must precede splicing [Rieder and Reenan, 2011], and additional recent evidence suggests widespread co-transcriptional editing [Rodriguez et al., 2012]. However, the precise relationship between the post- or co-transcriptional processes of splicing and editing remains unclear. Paralytic again presents an ideal system in which to investigate this complex relationship, as the napts mutation in the Maleless dsRNA helicase generates a “spicing catastrophe” around edited para intron 19 [Reenan et al., 2000]. It is possible that release of the complex secondary and tertiary RNA structures directing editing would 131 rescue the mlenap−ts molecular and temperature-senstive phenotypes. Surprisingly, we discovered that the loxP remnant alone, which is predicted to form a perfect duplex in the RNA, rescues this splicing phenotype. We are therefore currently targeting a TALEN to the remnant with the goals of both loxP disruption, as well as development of the TALEN technology in Drosophila. This technology will hopefully generate animals with degrees of loxP disruption, allowing us to determine the extent of loxP structural integrity that interacts with the Mlenap−ts mutant helicase and restores the splicing catastrophe. Moreover, the loxP-targeting TALEN may be used in combination with HR-generated alleles at any locus, providing an accessory tool to homologous recombination in Drosophila. 5.0.4 Temperature effects on RNA editing RNA structures fulfill numerous cellular roles at all levels of gene expression. RNA thermometers, for example, are prokaryotic structural elements that control trans- lation by sequestering the ribosome binding site and start codon within secondary structure [Narberhaus et al., 2006]. In response to temperature changes, these struc- tures rearrange to allow translation. The RNA structure therefore acts as a molecular sensor, eliciting cellular responses in response to abiotic factors. We propose that RNA structures involved in editing act as “thermistors,” modulating editing levels in response to temperature. Indeed, global editing is decreased at 30◦ C compared to 10 and 20◦ C, although individual editing sites behave differently in response to temperature shifts. This likely indicates a general melting of RNA structures at elevated temperature, but also reflects the range of RNA structures known to direct specific editing. While the response of editing to temperature is known to be adaptive in some cold blooded species [Garrett and Rosenthal, 2012], it is not known if the temperature-sensitivity of editing in poikilothermic Drosophila is adaptive or if it instead represents a breakdown of the editing system, either the 132 signals in the form of RNA structure or the dADAR editing machinery itself. 5.0.5 Sodium channel epilepsy mutations Curiously, epilepsy-causing mutations in the human sodium channel gene SCN1A occur throughout the channel and confer varying degrees of symptom severity [Meisler and Kearney, 2005], while the only Drosophila homolog, paralytic, is edited throughout the transcript to achieve channel diversity. A GEFS+ (Generalized Epilepsy with Febrile Seizures Plus) mutation found in human patients (K1270T), when introduced into Drosophila paralytic, confers a temperature-sensitive seizure phenotype [Sun et al., 2012]. This particular GEFS+ mutation occurs on S2 of DIII (red on yellow helix in Figure 5.1) in the sodium channel protein, and is predicted to be in close steric proximity to the residue altered by the unique Musca domestica editing site (M1230V, bottom green on white helix in Figure 5.1). Figure 5.1: Steric relationship between residues affected by RNA editing and epilepsy mutations in the sodium channel. Two views of the six trans-membrane helices of Para DIII are depicted, S1 in white, S2 in yellow, and S3-S4 in blue. Residues affected by RNA editing in para exon 19 are shown in green. The top sites are found in Drosophilidae (Q1213R, N1217D), while the bottom (M1230V) is unique to M. domestica (Figure 2.14). An amino acid substitution found in GEFS+ patients (K1270T, red) is localized to S2. In addition, an amino acid substitution (red) leading to a more severe form of epilepsy, SMEI (Severe Myoclonic Epilepsy of Infancy, or Dravet Syndrome), localizes to S1 close to the residues affected by editing in Drosophilidae. Figure courtesy of Dr. Robert Reenan. 133 The residues (Q1213R, N1217D) affected by the Drosophila paralytic editing sites studied in Chapter 2 localize to S1 of DIII (upper green on white helix in Figure 5.1). While an amino acid substitution (W1204R) leading to a more severe epilepsy presentation, SMEI (Severe Myoclonic Epilepsy of Infancy, or Dravet Syndrome), is present in S1 within fifteen amino acids of the edited residues (red on white helix in Figure 5.1). This observation raises the intriguing possibility of steric interaction between the edited residues and the residues involved in epilepsy. Further, the GEFS+ mutation was introduced into the endogenous locus via HR and is downstream of the loxP. This feature allows easy Cre-mediated trans recombination (Appendix A) with any cis element or editing mutation upstream of the loxP. Combining these mutations into a single para allele would reveal synergistic or antagonistic effects of changes conferred by editing on the epilepsy phenotype. Although behavioral assays were not sensitive enough to detect any subtle phenotype conferred by mutations in para editing, these changes might be revealed in the background of the temperature-sensitive phenotype. Appendix A A new hammer in the homologous recombination toolbox: Cre-mediated trans recombination 135 Homologous recombination (HR) is widely used in mouse and yeast genetics, but rarely in Drosophila engineering. Our lab has recently streamlined HR [Staber et al., 2011], and the use of this important technique is rapidly expanding. Often, as in this manuscript, we seek to introduce multiple mutations into a single genetic locus. These mutations are designed to fall on homology arms (please see Figure 1.17) bisected by a LoxP site. After HR the LoxP site remains in the genome as the single remnant of the HR process. Previously, the desire to combine mutations, which may reveal synergistic or antagonistic molecular interactions, necessitated creation of a completely new construct containing both mutations, followed by introduction of the construct into Drosophila embryos and the subsequent HR process. However, we reasoned that mutations on opposite targeting arms (within the same gene) might be recombined in vivo using Cre Recombinase (Figure A.1 a). ”Cre-mediated trans recombination” might easily allow for combination of mutations using simple Drosophila crossing, and would likely be useful in other model organisms in which HR is used. The final step of HR produces male flies expressing Cre Recombinase and containing the targeted mutation. These flies lack the mini-white marker due to the action of Cre in the zygote or early embryo [Staber et al., 2011]. These flies, therefore, may be immediately used in Cre-mediated trans recombination, if the addition of another mutation is desired. To test this concept we used “DCS delete” (arm 1) male flies produced during HR and crossed them to females with a GEFS+ epilepsy mutation (Figure A.1 b). GEFS+ (Generalized Epilepsy with Febrile Seizures +) is a human epilepsy disorder, often caused by a mutation in the SCN1A sodium channel gene. One known human GEFS+ mutation has been introduced into Drosophila paralytic (K1270T) to create a fly epilepsy disease model [Sun et al., 2012]. We chose this mutation because it is located in arm 2 and introduces a SexAI restriction site. The DCS delete mutation similarly introduces a PacI site, and therefore the presence of both mutations may be scored by double digest (Figure A.1 c). 136 a. Cre b. DCS delete ; CyoCre X GEFS+ mut1 loxP Y GEFS+ XloxP mut2 Fm7a X DCS delete ; CyoCre Y GEFS+ mut1 loxP mut2 DCS delete - GEFS+ loxP Y c. PacI SexAI DCS delete loxP GEFS+ d. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Figure A.1: Cre-mediated trans recombination. a. In the presence of Cre Recombinase (yellow), genomic material (blue) may be exchanged between chromosomes via the LoxP sites (yellow). This facilitates combining mutations (purple and green) introduced via HR within a single locus, provided they map to opposite arms. b. To test the theory of Cre-mediated trans recombination, we used the DCS delete mutation on para arm 1, which introduces a PacI restriction site, and the GEFS+ epilepsy mutation on para arm 2, which introduces a SexAI restriction site. Through crossing, we created transheterozygous females expressing Cre and assayed the genotypes of their sons via PCR and restriction digest (c). d. Proof of concept genotyping agarose gel for Cre-mediated trans recombination. Lane 1: 100 bp ladder. Lanes 2-13: progeny males from transheterozygous mothers. Lanes 14-16: control flies (14: wild type, 15: DCS delete single mutant, 16: GEFS+ single mutant). Lane 17: 1 Kb ladder 137 Results from this experiment are shown in Figure A.1 d, in which lanes 2-13 represent progeny males of unknown genotypes. Lanes 14-16 represent control animals (14: wild type X chromosome from a CyoCre individual, 15: DCS delete, 16: GEFS+). Lanes 3,4,10, and 12 represent animals in which the two mutations have been combined through Cre-mediated trans recombination. Lanes 2,5,7,8 and 9 represent males with the GEFS+ mutation, while lanes 6 and 11 show digest of the DCS delete mutation. Lane 13 represents a wild type chromosome with neither the GEFS+ or DCS delete mutation, which is also produced by Cre-mediated trans recombination (Figure A.1 a). To my knowledge this is the first example of Cre inducing interchromosomal recombination that preserves overall chromosome sequence and order, although a similar technique has been used in mouse embryonic stem cells to exchange nonhomol- ogous chromosome sequences [Collins et al., 2000, Smith et al., 1995, Van Deursen et al., 1995]. In fission yeast, the Cre/LoxP system has been used to demonstrate genomic organization, as interactions and LoxP-mediated recombinations between nonhomologous loci are dependent on proximity [Molnar and Kleckner, 2008, Burgess and Kleckner, 1999]. Cre-mediated trans recombination will be invaluable in Drosophila HR, as it may avoid many months of work to obtain interesting and informative double mutant alleles, which may reveal synergistic, antagonistic, or rescue interactions. Now that this technique has been demonstrated, HR targeting constructs may be specifically designed so that mutations thought to interact are contained on opposite targeting arms. Appendix B Temperature-responsiveness of individual editing sites 139 We investigated fifty-five editing sites in sixteen Drosophila transcripts that had been previously characterized [Savva et al., 2012]. The results of each site are presented below. adar Caps_1 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) SloA_1 StnB_1 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Unc 1.0 0.8 Editing level (G/G+A) 0.6 0.4 0.2 0.0 10 15 20 25 30 Temperature (° C) 140 Rdl1_1 Rdl1_2 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Rdl1_3 Rdl1_4 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Rdl2_5 Rdl2_6 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) 141 Eag1_2 Eag1_3 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Eag1_4 Eag1_5 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Eag2_6 1.0 0.8 Editing level (G/G+A) 0.6 0.4 0.2 0.0 10 15 20 25 30 Temperature (° C) 142 CaID_1 CaID_2 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) CaID_3 CaID_4 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) CaID_5 1.0 0.8 Editing level (G/G+A) 0.6 0.4 0.2 0.0 10 15 20 25 30 Temperature (° C) 143 Syt_2 Syt_3 1.0 1.0 Editing level (G/G+A) 0.8 0.8 Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Syt_4 1.0 0.8 Editing level (G/G+A) 0.6 0.4 0.2 0.0 10 15 20 25 30 Temperature (° C) 144 a2d_1 a2d_2 1.0 1.0 Editing level (G/G+A) 0.8 0.8 Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Sbd_1 Sbd_2 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) 145 Cpx_1 Cpx_2 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Cpx_3 Para_1 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Para_2 Para_3 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) 146 Shab_1 Shab_2 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Shab_3 Shab_4 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Shab_5 Shab_6 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Shab_7 1.0 0.8 Editing level (G/G+A) 0.6 0.4 0.2 0.0 10 15 20 25 30 Temperature (° C) 147 Da5_f2_1 Da5_f2_2 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Da5_f2_3 1.0 0.8 Editing level (G/G+A) 0.6 0.4 0.2 0.0 10 15 20 25 30 Temperature (° C) 148 Da5_f3_4 Da5_f3_5 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Da5_f3_6 Da5_f3_7 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Da5_f3_8 1.0 0.8 Editing level (G/G+A) 0.6 0.4 0.2 0.0 10 15 20 25 30 Temperature (° C) 149 Dop_1 Dop_2 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Dop_3 Dop_4 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Dop_5 Dop_6 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) 150 Dop_8 Dop_9 1.0 1.0 0.8 0.8 Editing level (G/G+A) Editing level (G/G+A) 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 10 15 20 25 30 10 15 20 25 30 Temperature (° C) Temperature (° C) Dop_10 1.0 0.8 Editing level (G/G+A) 0.6 0.4 0.2 0.0 10 15 20 25 30 Temperature (° C) Figure B.1: Editing response to temperature of 55 individual Drosophila editing sites. Adr, (dAdar, the RNA editing enzyme). Caps (calcium-activated protein for secretion, a regulator of exocytosis). SloA (slowpoke A, a calcium-activated potassium channel). StnB (stoned-B, a regulator of endocytosis). Unc (uncoordinated-13, a regulator of exocytosis). Rdl (resistant to dieldrin, a GABA receptor α subunit). Eag (ether a go-go, a voltage-gated potassium channel α subunit). CaID (Ca-alpha1D, a voltage-gated calcium channel α subunit). Syt (synaptogagmin-1, a regulator of exocytosis. Ard (ard, a nicotinic acetylcholine receptor α subunit). Sbd (Sbd, a nicotinic acetylcholine receptor β subunit. Cpx (complexin, a regulator of exocytosis). Para (paralytic, a voltage-gated sodium channel α subunit). Shab (shab, a voltage-gated potassium channel). Da5 (Dα5, a nicotinic acetylcholine receptor α subunit). Da6 (Dα6, a nicotinic acetylcholine receptor α subunit). Dop (DopEcR, a Dopamine/Ecdysteroid receptor). Editing sites as in Savva et al. 2011. Bibliography [Agrawal and Stormo, 2005] Agrawal, R. and Stormo, G. D. (2005). Editing effi- ciency of a Drosophila gene correlates with a distant splice site selection. RNA 11, 563–6. [Anderson et al., 1998] Anderson, S., Schlegel, B., Nakajima, T., Wolpin, E. and Parvin, J. (1998). BRCA1 protein is linked to the RNA polymerase II holoenzyme complex via RNA helicase A. Nature Genetics 19, 254–256. [Ansai et al., 2013] Ansai, S., Sakuma, T., Yamamoto, T., Ariga, H., Uemura, N., Takahashi, R. and Kinoshita, M. (2013). Efficient targeted mutagenesis in medaka using custom-designed transcription activator-like effector nucleases. Genetics 193, 739–749. [Aruscavage and Bass, 2000] Aruscavage, P. J. and Bass, B. L. (2000). A phyloge- netic analysis reveals an unusual sequence conservation within introns involved in RNA editing. RNA 6, 257–69. [Athanasiadis et al., 2004] Athanasiadis, A., Rich, A. and Maas, S. (2004). Widespread A-to-I RNA editing of Alu-containing mRNAs in the human tran- scriptome. PLoS Biol 2, e391. [Bahn et al., 2011] Bahn, J. H., Lee, J. H., Li, G., Greer, C., Peng, G. and Xiao, X. (2011). Accurate Identification of A-to-I RNA editing in human by transcriptome sequencing. Genome Res 22. [Baloh, 2012] Baloh, R. H. (2012). How do the RNA-binding proteins TDP-43 and FUS relate to amyotrophic lateral sclerosis and frontotemporal degeneration, and to each other? Curr Opin Neurol 25, 701–7. [Bartel, 2004] Bartel, D. P. (2004). MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116, 281–97. [Basillo et al., 1962] Basillo, C., Wahba, A., Lengyel, P., Speyer, J. and Ochoa, S. (1962). Synthetic polynucleotides and the amino acid code. V. Proc Natl Acad Sci U S A 48, 613–616. [Bass et al., 2012] Bass, B., Hundley, H., Li, J. B., Peng, Z., Pickrell, J., Xiao, X. G. and Yang, L. (2012). The difficult calls in RNA editing. Nat Biotechnol 30, 1207–9. 151 152 [Bass, 1997] Bass, B. L. (1997). RNA editing and hypermutation by adenosine deam- ination. Trends Biochem Sci 22, 157–62. [Bass, 2002] Bass, B. L. (2002). RNA editing by adenosine deaminases that act on RNA. Annu Rev Biochem 71, 817–46. [Bass and Weintraub, 1987] Bass, B. L. and Weintraub, H. (1987). A Developmen- tally Regulated Activity That Unwinds Rna Duplexes. Cell 48, 607–613. [Bedell et al., 2012] Bedell, V., Wang, Y., Campbell, J., Poshusta, T., Starker, C., Krug, R., Tan, W., Penheiter, S., Ma, A., Leung, A., Fahrenkrug, S., Carlson, D., Voytas, D., Clark, K., Essner, J. and Ekker, S. (2012). In vivo genome editing using a high-efficiency TALEN system. Nature 491, 114–118. [Bernstein et al., 2001] Bernstein, E., Caudy, A. A., Hammond, S. M. and Hannon, G. J. (2001). Role for a bidentate ribonuclease in the initiation step of RNA interference. Nature 409, 363–6. [Bhalla et al., 2004] Bhalla, T., Rosenthal, J. J., Holmgren, M. and Reenan, R. (2004). Control of human potassium channel inactivation by editing of a small mRNA hairpin. Nat Struct Mol Biol 11, 950–6. [Blackburn and Collins, 2011] Blackburn, E. H. and Collins, K. (2011). Telomerase: an RNP enzyme synthesizes DNA. Cold Spring Harb Perspect Biol 3. [Blow et al., 2004] Blow, M., Futreal, P. A., Wooster, R. and Stratton, M. R. (2004). A survey of RNA editing in human brain. Genome Res 14, 2379–87. [Blow et al., 2006] Blow, M. J., Grocock, R. J., van Dongen, S., Enright, A. J., Dicks, E., Futreal, P. A., Wooster, R. and Stratton, M. R. (2006). RNA editing of human microRNAs. Genome Biol 7, R27. [Boch et al., 2009] Boch, J., Scholze, H., Schornack, S., Landgraf, A., Hahn, S., Kay, S., Lahaye, T., Nickstadt, A. and Bonas, U. (2009). Breaking the code of DNA binding specificity of TAL-type III effectors. Science 326, 1509–1512. [Bonetti and Carninci, 2012] Bonetti, A. and Carninci, P. (2012). Structure by se- quencing: discovery of hidden messages in the noncoding RNA fraction. Mol Cell 48, 153–5. [Bratt and Ohman, 2003] Bratt, E. and Ohman, M. (2003). Coordination of editing and splicing of glutamate receptor pre-mRNA. RNA 9, 309–18. [Brierley et al., 2007] Brierley, I., Pennell, S. and Gilbert, R. J. (2007). Viral RNA pseudoknots: versatile motifs in gene expression and replication. Nat Rev Micro- biol 5, 598–610. [Brink et al., 1993] Brink, M. F., Verbeet, M. P. and de Boer, H. A. (1993). Forma- tion of the central pseudoknot in 16S rRNA is essential for initiation of translation. EMBO J 12, 3987–96. [Buratti and Baralle, 2004] Buratti, E. and Baralle, F. E. (2004). Influence of RNA secondary structure on the pre-mRNA splicing process. Mol Cell Biol 24, 10505– 14. 153 [Burgess and Kleckner, 1999] Burgess, S. and Kleckner, N. (1999). Collisions be- tween yeast chromosomal loci in vivo are governed by three layers of organization. Genes and Development 13, 1871–1883. [Burnett et al., 2012] Burnett, J. C., Zaia, J. A. and Rossi, J. J. (2012). Creating genetic resistance to HIV. Curr Opin Immunol 24, 625–32. [Burns et al., 1997] Burns, C. M., Chu, H., Rueter, S. M., Hutchinson, L. K., Can- ton, H., Sanders-Bush, E. and Emeson, R. B. (1997). Regulation of serotonin-2C receptor G-protein coupling by RNA editing. Nature 387, 303–8. [Bycroft et al., 1995] Bycroft, M., Grunert, S., Murzin, A. G., Proctor, M. and St Johnston, D. (1995). NMR solution structure of a dsRNA binding domain from Drosophila staufen protein reveals homology to the N-terminal domain of ribosomal protein S5. Embo J 14, 3563–71. [Carlson et al., 2012] Carlson, D., Tan, W., Lillico, S., Stverakova, D., Proudfoot, C., Christian, M., Voytas, D., Long, C., Whitelaw, B. and Fahrenkrug, S. (2012). Efficient TALEN-mediated gene knockout in livestock. Proc Natl Acad Sci U S A 109, 17382–17387. [Casey, 2011] Casey, J. L. (2011). Control of ADAR1 Editing of Hepatitis Delta Virus RNAs. Curr Top Microbiol Immunol 353. [Cattaneo, 1994] Cattaneo, R. (1994). Biased (A–¿I) hypermutation of animal RNA virus genomes. Curr Opin Genet Dev 4, 895–900. [Cermak et al., 2011] Cermak, T., Doyle, E., Christian, M., Wang, L., Zhang, Y., Schmidt, C., Baller, J., Somia, N., Bogdanove, A. and Voytas, D. (2011). Efficient design and assembly of custom TALEN and other TAL effector-based constructs for DNA targeting. Nucleic Acids Research 39, e82. [Chen and Garcia, 2012] Chen, A. A. and Garcia, A. E. (2012). Mechanism of en- hanced mechanical stability of a minimal RNA kissing complex elucidated by nonequilibrium molecular dynamics simulations. Proc Natl Acad Sci U S A 109, E1530–9. [Chen et al., 2000] Chen, C. X., Cho, D. S., Wang, Q., Lai, F., Carter, K. C. and Nishikura, K. (2000). A third member of the RNA-specific adenosine deaminase gene family, ADAR3, contains both single- and double-stranded RNA binding domains. Rna 6, 755–67. [Chen et al., 2008] Chen, L. L., DeCerbo, J. N. and Carmichael, G. G. (2008). Alu element-mediated gene silencing. EMBO J 27, 1694–705. [Chen et al., 2013] Chen, S., Oikonomou, G., Chin, C., Niles, B., Liu, J., Lee, D., Antosheckin, I. and Prober, D. (2013). A large-scale in vivo analysis reveals that TALENs are significantly more mutagenic than ZFNs generated using context- dependent assembly. Nucleic Acids Res 41, 2769–2778. [Cho et al., 2003] Cho, D. S., Yang, W., Lee, J. T., Shiekhattar, R., Murray, J. M. and Nishikura, K. (2003). Requirement of dimerization for RNA editing activity of adenosine deaminases acting on RNA. J Biol Chem 278, 17093–102. 154 [Chung et al., 2008] Chung, W. J., Okamura, K., Martin, R. and Lai, E. C. (2008). Endogenous RNA interference provides a somatic defense against Drosophila transposons. Curr Biol 18, 795–802. [Cisse et al., 2012] Cisse, I., Kim, H. and Ha, T. (2012). A rule of seven in Watson- Crick base-pairing of mismatched sequences. Nat Struct Mol Biol 19, 623–7. [Collins et al., 2000] Collins, E. C., Pannell, R., Simpson, E. M., Forster, A. and Rabbitts, T. H. (2000). Inter-chromosomal recombination of Mll and Af9 genes mediated by cre-loxP in mouse development. EMBO Rep 1, 127–32. [Czech et al., 2008] Czech, B., Malone, C. D., Zhou, R., Stark, A., Schlingeheyde, C., Dus, M., Perrimon, N., Kellis, M., Wohlschlegel, J. A., Sachidanandam, R., Hannon, G. J. and Brennecke, J. (2008). An endogenous small interfering RNA pathway in Drosophila. Nature 453, 798–802. [Daniel et al., 2012] Daniel, C., Veno, M. T., Ekdahl, Y., Kjems, J. and Ohman, M. (2012). A distant cis acting intronic element induces site-selective RNA editing. Nucleic Acids Res 40, 9876–86. [de la Mata et al., 2003] de la Mata, M., Alonso, C. R., Kadener, S., Fededa, J. P., Blaustein, M., Pelisch, F., Cramer, P., Bentley, D. and Kornblihtt, A. R. (2003). A slow RNA polymerase II affects alternative splicing in vivo. Mol Cell 12, 525–32. [Decher et al., 2010] Decher, N., Streit, A. K., Rapedius, M., Netter, M. F., Marzian, S., Ehling, P., Schlichthorl, G., Craan, T., Renigunta, V., Kohler, A., Dodel, R. C., Navarro-Polanco, R. A., Preisig-Muller, R., Klebe, G., Budde, T., Baukrowitz, T. and Daut, J. (2010). RNA editing modulates the binding of drugs and highly unsaturated fatty acids to the open pore of Kv potassium channels. EMBO J 29, 2101–13. [Desterro et al., 2003] Desterro, J. M., Keegan, L. P., Lafarga, M., Berciano, M. T., O’Connell, M. and Carmo-Fonseca, M. (2003). Dynamic association of RNA- editing enzymes with the nucleolus. J Cell Sci 116, 1805–18. [Dillon et al., 2009] Dillon, M., Wang, G., Garrity, P. and Huey, R. (2009). Thermal preference in Drosophila. Journal of Thermal Biology 34, 109–119. [Ding et al., 2013] Ding, Q., Lee, Y., Schaefer, E., Peters, D., Veres, A., Kim, K., Kuperwasser, N., Motola, D., Meissner, T., Hendriks, W., Trevisan, M., Gupta, R., Moisan, A., Banks, E., Friesen, M., Schinzel, R., Xia, F., Tang, A., Xia, Y., Figueroa, E., Wann, A., Ahfeldt, T., Daheron, L., Zhang, F., Rubin, L., Peng, L., Chung, R., Musunuru, K. and Cowan, C. (2013). A TALEN genome-editing system for generating human stem cell-based disease models. Cell Stem Cell 12, 238–251. [Ding et al., 2004] Ding, Y., Chan, C. Y. and Lawrence, C. E. (2004). Sfold web server for statistical folding and rational design of nucleic acids. Nucleic Acids Res 32, W135–41. [Dingledine et al., 1999] Dingledine, R., Borges, K., Bowie, D. and Traynelis, S. F. (1999). The glutamate receptor ion channels. Pharmacol Rev 51, 7–61. 155 [Doria et al., 2009] Doria, M., Neri, F., Gallo, A., Farace, M. G. and Michienzi, A. (2009). Editing of HIV-1 RNA by the double-stranded RNA deaminase ADAR1 stimulates viral infection. Nucleic Acids Res 37, 5848–58. [Dracheva et al., 2008] Dracheva, S., Patel, N., Woo, D. A., Marcus, S. M., Siever, L. J. and Haroutunian, V. (2008). Increased serotonin 2C receptor mRNA editing: a possible risk factor for suicide. Mol Psychiatry 13, 1001–10. [Draper, 1995] Draper, D. E. (1995). Protein-RNA recognition. Annu Rev Biochem 64, 593–620. [Eckmann et al., 2001] Eckmann, C. R., Neunteufl, A., Pfaffstetter, L. and Jantsch, M. F. (2001). The human but not the Xenopus RNA-editing enzyme ADAR1 has an atypical nuclear localization signal and displays the characteristics of a shuttling protein. Mol Biol Cell 12, 1911–24. [Economos and Lints, 1986] Economos, A. C. and Lints, F. A. (1986). Developmen- tal temperature and life span in Drosophila melanogaster. II. Oscillating temper- ature. Gerontology 32, 28–36. [Eggington et al., 2011] Eggington, J. M., Greene, T. and Bass, B. L. (2011). Pre- dicting sites of ADAR editing in double-stranded RNA. Nat Commun 2, 319. [Eisenberg, 2011] Eisenberg, E. (2011). Bioinformatic Approaches for Identification of A-to-I Editing Sites. Curr Top Microbiol Immunol 353. [Eisenberg et al., 2005] Eisenberg, E., Nemzer, S., Kinar, Y., Sorek, R., Rechavi, G. and Levanon, E. Y. (2005). Is abundant A-to-I RNA editing primate-specific? Trends in Genetics 21, 77–81. [Enstero et al., 2009] Enstero, M., Daniel, C., Wahlstedt, H., Major, F. and Ohman, M. (2009). Recognition and coupling of A-to-I edited sites are determined by the tertiary structure of the RNA. Nucleic Acids Res 37, 6916–26. [Fagegaltier et al., 2009] Fagegaltier, D., Bouge, A. L., Berry, B., Poisot, E., Sis- meiro, O., Coppee, J. Y., Theodore, L., Voinnet, O. and Antoniewski, C. (2009). The endogenous siRNA pathway is involved in heterochromatin formation in Drosophila. Proc Natl Acad Sci U S A 106, 21258–63. [Feng et al., 1995] Feng, G., Deak, P., Chopra, M. and Hall, L. M. (1995). Cloning and functional analysis of TipE, a novel membrane protein that enhances Drosophila para sodium channel function. Cell 82, 1001–11. [Feng et al., 2006] Feng, Y., Sansam, C. L., Singh, M. and Emeson, R. B. (2006). Altered RNA editing in mice lacking ADAR2 autoregulation. Mol Cell Biol 26, 480–8. [Filbin et al., 2013] Filbin, M., Vollmar, B., Shi, D., Gonen, T. and Kleft, J. (2013). HCV IRES manipulates the ribosome to promote the switch from translation initiation to elongation. Nat Struct Mol Biol 20, 150–158. [Fischer et al., 2008] Fischer, S. E., Butler, M. D., Pan, Q. and Ruvkun, G. (2008). Trans-splicing in C. elegans generates the negative RNAi regulator ERI-6/7. Na- ture 455, 491–6. 156 [Fuller-Pace, 2000] Fuller-Pace, F. (2000). DExD/H box RNA helicases: multifunc- tional proteins with important roles in transcriptional regulation. Nucleic Acids Research 34, 4206–4215. [Galeano et al., 2011] Galeano, F., Tomaselli, S., Locatelli, F. and Gallo, A. (2011). A-to-I RNA editing: The ”ADAR” side of human cancer. Semin Cell Dev Biol 23. [Gallo et al., 2003] Gallo, A., Keegan, L. P., Ring, G. M. and O’Connell, M. A. (2003). An ADAR that edits transcripts encoding ion channel subunits functions as a dimer. EMBO J 22, 3421–30. [Ganetzky, 1984] Ganetzky, B. (1984). Genetic studies of membrane excitability in Drosophila: lethal interaction between two temperature-sensitive paralytic muta- tions. Genetics 108, 897–911. [Garber et al., 2012] Garber, G., Smith, L. A., Reenan, R. A. and Rogina, B. (2012). Effect of sodium channel abundance on Drosophila development, reproductive capacity and aging. Fly (Austin) 6, 57–67. [Garrett and Rosenthal, 2012] Garrett, S. and Rosenthal, J. J. (2012). RNA editing underlies temperature adaptation in K+ channels from polar octopuses. Science 335, 848–51. [Gelinas et al., 2011] Gelinas, J. F., Clerzius, G., Shaw, E. and Gatignol, A. (2011). Enhancement of replication of RNA viruses by ADAR1 via RNA editing and inhibition of RNA-activated protein kinase. J Virol 85, 8460–6. [Gerber and Keller, 2001] Gerber, A. P. and Keller, W. (2001). RNA editing by base deamination: more enzymes, more targets, new mysteries. Trends Biochem Sci 26, 376–84. [Ghildiyal et al., 2008] Ghildiyal, M., Seitz, H., Horwich, M. D., Li, C., Du, T., Lee, S., Xu, J., Kittler, E. L., Zapp, M. L., Weng, Z. and Zamore, P. D. (2008). Endogenous siRNAs derived from transposons and mRNAs in Drosophila somatic cells. Science 320, 1077–81. [Gilbert, 1986] Gilbert, D. M. (1986). Temporal order of replication of Xenopus laevis 5S ribosomal RNA genes in somatic cells. Proc Natl Acad Sci U S A 83, 2924–8. [Glanzer et al., 2005] Glanzer, J., Miyashiro, K. Y., Sul, J. Y., Barrett, L., Belt, B., Haydon, P. and Eberwine, J. (2005). RNA splicing capability of live neuronal dendrites. Proc Natl Acad Sci U S A 102, 16859–64. [Goguel et al., 1993] Goguel, V., Wang, Y. and Rosbash, M. (1993). Short artificial hairpins sequester splicing signals and inhibit yeast pre-mRNA splicing. Mol Cell Biol 13, 6841–8. [Goldin, 2002] Goldin, A. L. (2002). Evolution of voltage-gated Na(+) channels. J Exp Biol 205, 575–84. 157 [Goldstrohm et al., 2001] Goldstrohm, A. C., Greenleaf, A. L. and Garcia-Blanco, M. A. (2001). Co-transcriptional splicing of pre-messenger RNAs: considerations for the mechanism of alternative splicing. Gene 277, 31–47. [Gommans et al., 2009] Gommans, W. M., Mullen, S. P. and Maas, S. (2009). RNA editing: a driving force for adaptive evolution? BioEssays 31, 1137–45. [Gong and Golic, 2003] Gong, W. J. and Golic, K. G. (2003). Ends-out, or replace- ment, gene targeting in Drosophila. Proc Natl Acad Sci U S A 100, 2556–61. [Graveley, 2001] Graveley, B. R. (2001). Alternative splicing: increasing diversity in the proteomic world. Trends Genet 17, 100–7. [Graveley et al., 2011] Graveley, B. R., Brooks, A. N., Carlson, J. W., Duff, M. O., Landolin, J. M., Yang, L., Artieri, C. G., van Baren, M. J., Boley, N., Booth, B. W., Brown, J. B., Cherbas, L., Davis, C. A., Dobin, A., Li, R., Lin, W., Malone, J. H., Mattiuzzo, N. R., Miller, D., Sturgill, D., Tuch, B. B., Zaleski, C., Zhang, D., Blanchette, M., Dudoit, S., Eads, B., Green, R. E., Hammonds, A., Jiang, L., Kapranov, P., Langton, L., Perrimon, N., Sandler, J. E., Wan, K. H., Willingham, A., Zhang, Y., Zou, Y., Andrews, J., Bickel, P. J., Brenner, S. E., Brent, M. R., Cherbas, P., Gingeras, T. R., Hoskins, R. A., Kaufman, T. C., Oliver, B. and Celniker, S. E. (2011). The developmental transcriptome of Drosophila melanogaster. Nature 471, 473–9. [Gu et al., 2012] Gu, T., Buaas, F. W., Simons, A. K., Ackert-Bicknell, C. L., Braun, R. E. and Hibbs, M. A. (2012). Canonical A-to-I and C-to-U RNA editing is enriched at 3’UTRs and microRNA target sites in multiple mouse tissues. PLoS One 7, e33720. [Hammond and Wood, 2011] Hammond, S. M. and Wood, M. J. (2011). Genetic therapies for RNA mis-splicing diseases. Trends Genet 27, 196–205. [Hannon, 2002] Hannon, G. J. (2002). RNA interference. Nature 418, 244–51. [Hanrahan et al., 2000] Hanrahan, C. J., Palladino, M. J., Ganetzky, B. and Reenan, R. A. (2000). RNA editing of the Drosophila para Na(+) channel transcript. Evolutionary conservation and developmental regulation. Genetics 155, 1149–60. [Harris-Kerr et al., 1993] Harris-Kerr, C. L., Zhang, M. and Peebles, C. L. (1993). The phylogenetically predicted base-pairing interaction between alpha and alpha’ is required for group II splicing in vitro. Proc Natl Acad Sci U S A 90, 10658–62. [Hideyama et al., 2012] Hideyama, T., Yamashita, T., Aizawa, H., Tsuji, S., Kakita, A., Takahashi, H. and Kwak, S. (2012). Profound downregulation of the RNA editing enzyme ADAR2 in ALS spinal motor neurons. Neurobiol Dis 45, 1121–8. [Higuchi et al., 2000] Higuchi, M., Maas, S., Single, F. N., Hartner, J., Rozov, A., Burnashev, N., Feldmeyer, D., Sprengel, R. and Seeburg, P. H. (2000). Point mutation in an AMPA receptor gene rescues lethality in mice deficient in the RNA-editing enzyme ADAR2. Nature 406, 78–81. 158 [Higuchi et al., 1993] Higuchi, M., Single, F. N., Kohler, M., Sommer, B., Sprengel, R. and Seeburg, P. H. (1993). Rna Editing of Ampa Receptor Subunit Glur-B - a Base-Paired Intron-Exon Structure Determines Position and Efficiency. Cell 75, 1361–1370. [Hodges et al., 2002] Hodges, D., Lee, D., Preston, C., Boswell, K., Hall, L. and O’Dowd, D. (2002). tipE regulates Na+ -dependent repetitive firing in Drosophila neurons. Molecular and Cellular Neuroscience 19, 402–416. [Hoffman, 2010] Hoffman, A. (2010). Physiological climatic limits in Drosophila: patterns and implications. The Journal of Experimental Biology 213, 870–880. [Hoopengardner et al., 2003] Hoopengardner, B., Bhalla, T., Staber, C. and Reenan, R. (2003). Nervous system targets of RNA editing identified by comparative genomics. Science 301, 832–6. [Huxley, 1959] Huxley, A. F. (1959). Ion movements during nerve activity. Ann N Y Acad Sci 81, 221–46. [Huynen et al., 1997] Huynen, M., Gutell, R. and Konings, D. (1997). Assessing the reliability of RNA folding using statistical mechanics. J Mol Biol 267, 1104–12. [Ingleby et al., 2009] Ingleby, L., Maloney, R., Jepson, J., Horn, R. and Reenan, R. (2009). Regulated RNA editing and functional epistasis in Shaker potassium channels. J Gen Physiol 133, 17–27. [Jacobs et al., 2009] Jacobs, M. M., Fogg, R. L., Emeson, R. B. and Stanwood, G. D. (2009). ADAR1 and ADAR2 expression and editing activity during forebrain development. Dev Neurosci 31, 223–37. [Jepson et al., 2011] Jepson, J. E., Savva, Y. A., Yokose, C., Sugden, A. U., Sahin, A. and Reenan, R. A. (2011). Engineered alterations in RNA editing modulate complex behavior in Drosophila: regulatory diversity of adenosine deaminase act- ing on RNA (ADAR) targets. J Biol Chem 286, 8325–37. [Jin et al., 2007] Jin, Y., Tian, N., Cao, J., Liang, J., Yang, Z. and Lv, J. (2007). RNA editing and alternative splicing of the insect nAChR subunit alpha6 tran- script: evolutionary conservation, divergence and regulation. BMC Evol Biol 7, 98. [Jin et al., 2009] Jin, Y., Zhang, W. and Li, Q. (2009). Origins and evolution of ADAR-mediated RNA editing. IUBMB Life 61, 572–8. [Joung and Sander, 2013] Joung, J. and Sander, J. (2013). TALENs: a widely ap- plicable technology for targeted genome editing. Nature Reviews 14, 49–55. [Kawahara et al., 2004] Kawahara, Y., Ito, K., Sun, H., Aizawa, H., Kanazawa, I. and Kwak, S. (2004). Glutamate receptors: RNA editing and death of motor neurons. Nature 427, 801. [Kawahara et al., 2007] Kawahara, Y., Zinshteyn, B., Sethupathy, P., Iizasa, H., Hatzigeorgiou, A. G. and Nishikura, K. (2007). Redirection of silencing targets by adenosine-to-inosine editing of miRNAs. Science 315, 1137–40. 159 [Kawamura et al., 2008] Kawamura, Y., Saito, K., Kin, T., Ono, Y., Asai, K., Suno- hara, T., Okada, T. N., Siomi, M. C. and Siomi, H. (2008). Drosophila endogenous small RNAs bind to Argonaute 2 in somatic cells. Nature 453, 793–7. [Keegan et al., 2005] Keegan, L. P., Brindle, J., Gallo, A., Leroy, A., Reenan, R. A. and O’Connell, M. A. (2005). Tuning of RNA editing by ADAR is required in Drosophila. EMBO J 24, 2183–93. [Kim et al., 2004] Kim, D. D., Kim, T. T., Walsh, T., Kobayashi, Y., Matise, T. C., Buyske, S. and Gabriel, A. (2004). Widespread RNA editing of embedded alu elements in the human transcriptome. Genome Res 14, 1719–25. [Kim et al., 2013] Kim, Y., Kweon, J., Kim, A., Chon, J., Yoo, J., Kim, H., Kim, S., Lee, C., Jeong, E., Chung, E., Kim, D., Lee, M., Go, E., Song, H., Kim, H., Cho, N., Bang, D., Kim, S. and Kim, J. (2013). A library of TAL effector nucleases spanning the human genome. Nat Biotechnol 31, 251–258. [Kuroda et al., 1991] Kuroda, M., Kernan, M., Kreber, R., Ganetzky, B. and Baker, B. (1991). The maleless protein associates with the X chromosome to regulate dosage compensation in Drosophila. Cell 66, 935–947. [Kuttan and Bass, 2012] Kuttan, A. and Bass, B. L. (2012). Mechanistic insights into editing-site specificity of ADARs. Proc Natl Acad Sci U S A 109, E3295–304. [Lachaise et al., 1988] Lachaise, D., Cariou, M., David, J., Lemeunier, F., Tsacas, L. and Ashburner, M. (1988). Evolutionary Biology vol. 22, chapter Historical Biogeography of the Drosophila-melanogaster species subgroup, pp. 159–225. New York: Plenum Press. [Lalonde et al., 2011] Lalonde, E., Ha, K. C., Wang, Z., Bemmo, A., Kleinman, C. L., Kwan, T., Pastinen, T. and Majewski, J. (2011). RNA sequencing reveals the role of splicing polymorphisms in regulating human gene expression. Genome Res , 21, 545–54. [Larschan et al., 2011] Larschan, E., Bishop, E. P., Kharchenko, P. V., Core, L. J., Lis, J. T., Park, P. J. and Kuroda, M. I. (2011). X chromosome dosage com- pensation via enhanced transcriptional elongation in Drosophila. Nature , 471, 115–8. [Laurencikiene et al., 2006] Laurencikiene, J., Kallman, A. M., Fong, N., Bentley, D. L. and Ohman, M. (2006). RNA editing and alternative splicing: the impor- tance of co-transcriptional coordination. EMBO Rep , 7, 303–7. [Lee et al., 1997] Lee, C., Chang, K., Kuroda, M. and Hurwitz, J. (1997). The NTPase/helicase activities of Drosophila maleless, an essential factor in dosage compensation. EMBO , 16, 2671–2681. [Lehmann and Bass, 2000] Lehmann, K. A. and Bass, B. L. (2000). Double-stranded RNA adenosine deaminases ADAR1 and ADAR2 have overlapping specificities. Biochemistry , 39, 12875–84. 160 [Lev-Maor et al., 2007] Lev-Maor, G., Sorek, R., Levanon, E. Y., Paz, N., Eisenberg, E. and Ast, G. (2007). RNA-editing-mediated exon evolution. Genome Biol , 8, R29. [Levanon et al., 2004] Levanon, E. Y., Eisenberg, E., Yelin, R., Nemzer, S., Halleg- ger, M., Shemesh, R., Fligelman, Z. Y., Shoshan, A., Pollock, S. R., Sztybel, D., Olshansky, M., Rechavi, G. and Jantsch, M. F. (2004). Systematic identification of abundant A-to-I editing sites in the human transcriptome. Nature Biotechnology , 22, 1001–1005. [Li et al., 2009] Li, J. B., Levanon, E. Y., Yoon, J. K., Aach, J., Xie, B., Leproust, E., Zhang, K., Gao, Y. and Church, G. M. (2009). Genome-wide identification of human RNA editing sites by parallel DNA capturing and sequencing. Science , 324, 1210–3. [Li et al., 2011] Li, M., Wang, I. X., Li, Y., Bruzel, A., Richards, A. L., Toung, J. M. and Cheung, V. G. (2011). Widespread RNA and DNA sequence differences in the human transcriptome. Science , 333, 53–8. [Lin et al., 2009] Lin, W. H., Wright, D. E., Muraro, N. I. and Baines, R. A. (2009). Alternative splicing in the voltage-gated sodium channel DmNav regulates activa- tion, inactivation, and persistent current. J Neurophysiol , 102, 1994–2006. [Littleton and Ganetzky, 2000] Littleton, J. T. and Ganetzky, B. (2000). Ion chan- nels and synaptic organization: analysis of the Drosophila genome. Neuron , 26, 35–43. [Liu et al., 2012] Liu, J., Li, C., Yu, Z., Huang, P., Wu, H., Wei, C., Zhu, N., Shen, Y., Chen, Y., Zhang, B., Deng, W. and Jiao, R. (2012). Efficient and specific modifications of the Drosophila genome by means of an easy TALEN strategy. Journal of Genetics and Genomics , 39, 209–215. [Liu et al., 1997] Liu, Y., George, C. X., Patterson, J. B. and Samuel, C. E. (1997). Functionally distinct double-stranded RNA-binding domains associated with alter- native splice site variants of the interferon-inducible double-stranded RNA-specific adenosine deaminase. J Biol Chem , 272, 4419–28. [Lomeli et al., 1994] Lomeli, H., Mosbacher, J., Melcher, T., Hoger, T., Geiger, J. R., Kuner, T., Monyer, H., Higuchi, M., Bach, A. and Seeburg, P. H. (1994). Control of kinetic properties of AMPA receptor channels by nuclear RNA editing. Science , 266, 1709–13. [Loughney et al., 1989] Loughney, K., Kreber, R. and Ganetzky, B. (1989). Molec- ular analysis of the para locus, a sodium channel gene in Drosophila. Cell , 58, 1143–54. [Luciano et al., 2004] Luciano, D. J., Mirsky, H., Vendetti, N. J. and Maas, S. (2004). RNA editing of a miRNA precursor. Rna , 10, 1174–7. [Luo and Reed, 1999] Luo, M. J. and Reed, R. (1999). Splicing is required for rapid and efficient mRNA export in metazoans. Proc Natl Acad Sci U S A , 96, 14937–42. 161 [Lyddon et al., 2012] Lyddon, R., Dwork, A. J., Keddache, M., Siever, L. J. and Dracheva, S. (2012). Serotonin 2c receptor RNA editing in major depression and suicide. World J Biol Psychiatry , Epub. [Maas and Gommans, 2009] Maas, S. and Gommans, W. M. (2009). Identification of a selective nuclear import signal in adenosine deaminases acting on RNA. Nucleic Acids Res , 37, 5822–9. [Macbeth et al., 2005] Macbeth, M. R., Schubert, H. L., Vandemark, A. P., Lingam, A. T., Hill, C. P. and Bass, B. L. (2005). Inositol hexakisphosphate is bound in the ADAR2 core and required for RNA editing. Science , 309, 1534–9. [Mandal and Breaker, 2004] Mandal, M. and Breaker, R. (2004). Gene regulation by riboswitches. Molecular Cell Biology , 5, 451–463. [Mantegazza et al., 2010] Mantegazza, M., Curia, G., Biagini, G., Ragsdale, D. S. and Avoli, M. (2010). Voltage-gated sodium channels as therapeutic targets in epilepsy and other neurological disorders. Lancet Neurol , 9, 413–24. [Marcucci et al., 2011] Marcucci, R., Brindle, J., Paro, S., Casadio, A., Hempel, S., Morrice, N., Bisso, A., Keegan, L. P., Del Sal, G. and O’Connell, M. A. (2011). Pin1 and WWP2 regulate GluR2 Q/R site RNA editing by ADAR2 with opposing effects. EMBO J , 30. [Marcucci et al., 2009] Marcucci, R., Romano, M., Feiguin, F., O’Connell, M. A. and Baralle, F. E. (2009). Dissecting the splicing mechanism of the Drosophila editing enzyme; dADAR. Nucleic Acids Res , 37, 1663–71. [Martins et al., 2011] Martins, S. B., Rino, J., Carvalho, T., Carvalho, C., Yoshida, M., Klose, J. M., de Almeida, S. F. and Carmo-Fonseca, M. (2011). Spliceosome assembly is coupled to RNA polymerase II dynamics at the 3’ end of human genes. Nat Struct Mol Biol , 18. [Masliah et al., 2012] Masliah, G., Barraud, P. and Allain, F. H. (2012). RNA recog- nition by double-stranded RNA binding domains: a matter of shape and sequence. Cell Mol Life Sci , Epub. [May et al., 2011] May, G., Olson, S., McManus, C. and Graveley, B. (2011). Com- peting RNA secondary structures are required for mutually exclusive splicing of the Dscam exon 6 cluster. RNA , 18, 222–229. [McManus and Graveley, 2011] McManus, C. J. and Graveley, B. R. (2011). RNA structure and the mechanisms of alternative splicing. Curr Opin Genet Dev , 21, 373–9. [Meisler and Kearney, 2005] Meisler, M. H. and Kearney, J. A. (2005). Sodium chan- nel mutations in epilepsy and other neurological disorders. J Clin Invest , 115, 2010–7. [Meister et al., 2004] Meister, G., Landthaler, M., Dorsett, Y. and Tuschl, T. (2004). Sequence-specific inhibition of microRNA- and siRNA-induced RNA silencing. Rna , 10, 544–50. 162 [Melcher et al., 1996] Melcher, T., Maas, S., Herb, A., Sprengel, R., Higuchi, M. and Seeburg, P. H. (1996). RED2, a brain-specific member of the RNA-specific adenosine deaminase family. J Biol Chem , 271, 31795–8. [Meyer et al., 2012] Meyer, L., Zweig, A., Karolchik, D., Kuhn, R., Wong, M., Sloan, C., Rosenbloom, K., Roe, G., Rhead, B., Raney, B., Pohl, A., Malladi, V., Li, C., Lee, B., Learned, K., Kirkup, V., Hsu, F., Heitner, S., Harte, R., Haeussler, M., Guruvadoo, L., Goldman, M., Giardine, B., Fujita, P., Dreszer, T., Diekhans, M., Cline, M., Clawson, H., Barber, G., Haussler, D. and Kent, W. (2012). The UCSC Genome Browser database: extensions and updates 2013. Nucleic Acids Res , 41, D64–D69. [Michel et al., 1989] Michel, F., Umesono, K. and Ozeki, H. (1989). Comparative and functional anatomy of group II catalytic introns–a review. Gene , 82, 5–30. [Mironov et al., 2002] Mironov, A. S., Gusarov, I., Rafikov, R., Lopez, L. E., Shatalin, K., Kreneva, R. A., Perumov, D. A. and Nudler, E. (2002). Sensing small molecules by nascent RNA: a mechanism to control transcription in bacte- ria. Cell , 111, 747–56. [Molnar and Kleckner, 2008] Molnar, M. and Kleckner, N. (2008). Examination of interchromosomal interactions in vegetatively growing diploid Schizosaccha- romyces pmobe cells by Cre/loxP site-specific recombination. Genetics , 178, 99–112. [Morra et al., 2008] Morra, R., Smith, E., Yokoyama, R. and Lucchesi, J. (2008). The MLE subunit of the Drosophila MSL complex uses its ATPase activity for dosage compensation and its helicase activity for targeting. Molecular and Cellular Biology , 28, 958–966. [Morse et al., 2002] Morse, D. P., Aruscavage, P. J. and Bass, B. L. (2002). RNA hairpins in noncoding regions of human brain and Caenorhabditis elegans mRNA are edited by adenosine deaminases that act on RNA. Proc Natl Acad Sci U S A , 99, 7906–11. [Moscou and Bogdanove, 2009] Moscou, M. and Bogdanove, A. (2009). A simple cipher governs DNA recognition by TAL effectors. Science , 326, 1501. [Munoz et al., 2010] Munoz, M. J., de la Mata, M. and Kornblihtt, A. R. (2010). The carboxy terminal domain of RNA polymerase II and alternative splicing. Trends Biochem Sci , 35, 497–504. [Nahvi et al., 2002] Nahvi, A., Sudarsan, N., Ebert, M. S., Zou, X., Brown, K. L. and Breaker, R. R. (2002). Genetic control by a metabolite binding mRNA. Chem Biol , 9, 1043. [Nakajima et al., 1997] Nakajima, T., Uchida, C., Anderson, S., Lee, C., Hurwitz, J., Parvin, J. and Montminy, M. (1997). RNA helicase A mediates association of CBP with RNA polymerase II. Cell , 90, 1107–1112. [Nanduri et al., 1998] Nanduri, S., Carpick, B. W., Yang, Y., Williams, B. R. and Qin, J. (1998). Structure of the double-stranded RNA-binding domain of the 163 protein kinase PKR reveals the molecular basis of its dsRNA-mediated activation. Embo J , 17, 5458–65. [Narberhaus et al., 2006] Narberhaus, F., Waldminghaus, T. and Chowdhury, S. (2006). RNA thermometers. FEMS Microbiol Rev , 30, 3–16. [Nekrutenko, 2004] Nekrutenko, A. (2004). Reconciling the numbers: ESTs versus protein-coding genes. Mol Biol Evol , 21, 1278–82. [Nelson and Wyman, 1990] Nelson, J. C. and Wyman, R. J. (1990). Examination of paralysis in Drosophila temperature-sensitive paralytic mutations affecting sodium channels; a proposed mechanism of paralysis. J Neurobiol , 21, 453–69. [Neupert et al., 2008] Neupert, J., Karcher, D. and Bock, R. (2008). Design of sim- ple synthetic RNA thermometers for temperature-controlled gene expression in Escherichia coli. Nucleic Acids Res , 36, e124. [Nielsen et al., 2006] Nielsen, M. L., Savitski, M. M. and Zubarev, R. A. (2006). Extent of modifications in human proteome samples and their effect on dynamic range of analysis in shotgun proteomics. Mol Cell Proteomics , 5, 2384–91. [Nishikura, 2006] Nishikura, K. (2006). Editor meets silencer: crosstalk between RNA editing and RNA interference. Nat Rev Mol Cell Biol , 7, 919–31. [Nishikura, 2010] Nishikura, K. (2010). Functions and regulation of RNA editing by ADAR deaminases. Annu Rev Biochem , 79, 321–49. [Nishikura et al., 2000] Nishikura, K., Wang, Q., Khillan, J. and Gadue, P. (2000). Requirement of the RNA editing deaminase ADAR1 gene for embryonic erythro- poiesis. Science , 290, 1765–1768. [Nishikura et al., 1991] Nishikura, K., Yoo, C., Kim, U., Murray, J. M., Estes, P. A., Cash, F. E. and Liebhaber, S. A. (1991). Substrate specificity of the dsRNA unwinding/modifying activity. EMBO J , 10, 3523–32. [Ohlson et al., 2007] Ohlson, J., Pedersen, J. S., Haussler, D. and Ohman, M. (2007). Editing modifies the GABA(A) receptor subunit alpha3. Rna , 13, 698–703. [Ohman et al., 2000] Ohman, M., Kallman, A. M. and Bass, B. L. (2000). In vitro analysis of the binding of ADAR2 to the pre-mRNA encoding the GluR-B R/G site. RNA , 6, 687–97. [Olson et al., 2008] Olson, R. O., Liu, Z., Nomura, Y., Song, W. and Dong, K. (2008). Molecular and functional characterization of voltage-gated sodium channel variants from Drosophila melanogaster. Insect Biochem Mol Biol , 38, 604–10. [Palavicini et al., 2012] Palavicini, J. P., Correa-Rojas, R. A. and Rosenthal, J. J. (2012). Extra double-stranded RNA binding domain (dsRBD) in a squid RNA editing enzyme confers resistance to high salt environment. J Biol Chem , 287, 17754–64. [Palavicini et al., 2009] Palavicini, J. P., O’Connell, M. A. and Rosenthal, J. J. (2009). An extra double-stranded RNA binding domain confers high activity to a squid RNA editing enzyme. Rna , 15, 1208–18. 164 [Paldi and Gurevitz, 2010] Paldi, T. and Gurevitz, M. (2010). Coupling between residues on S4 and S1 defines the voltage-sensor resting conformation in NaChBac. Biophys J , 99, 456–63. [Palladino et al., 2000] Palladino, M. J., Keegan, L. P., O’Connell, M. A. and Reenan, R. A. (2000). A-to-I pre-mRNA editing in Drosophila is primarily in- volved in adult nervous system function and integrity. Cell , 102, 437–49. [Patterson and Samuel, 1995] Patterson, J. B. and Samuel, C. E. (1995). Expression and regulation by interferon of a double-stranded-RNA-specific adenosine deami- nase from human cells: evidence for two forms of the deaminase. Mol Cell Biol , 15, 5376–88. [Paz-Yaacov et al., 2010] Paz-Yaacov, N., Levanon, E. Y., Nevo, E., Kinar, Y., Harmelin, A., Jacob-Hirsch, J., Amariglio, N., Eisenberg, E. and Rechavi, G. (2010). Adenosine-to-inosine RNA editing shapes transcriptome diversity in pri- mates. Proc Natl Acad Sci U S A , 107, 12174–9. [Perard et al., 2013] Perard, J., Leyrat, C., Baudin, F., Drouet, E. and Jamin, M. (2013). Structure of the full-length HCV IRES in solution. Nat Commun , 19, 1612. [Peters et al., 2003] Peters, N. T., Rohrbach, J. A., Zalewski, B. A., Byrkett, C. M. and Vaughn, J. C. (2003). RNA editing and regulation of Drosophila 4f-rnp ex- pression by sas-10 antisense readthrough mRNA transcripts. Rna , 9, 698–710. [Plass et al., 2012] Plass, M., Codony-Servat, C., Ferreira, P., Vilardell, J. and Eyras, E. (2012). RNA secondary structure mediates alternative 3’ss selection in Saccharomyces cerevisiae. RNA , 18, 1103–1115. [Polson and Bass, 1994] Polson, A. G. and Bass, B. L. (1994). Preferential selec- tion of adenosines for modification by double-stranded RNA adenosine deaminase. Embo J , 13, 5701–11. [Polson et al., 1991] Polson, A. G., Crain, P. F., Pomerantz, S. C., McCloskey, J. A. and Bass, B. L. (1991). The mechanism of adenosine to inosine conversion by the double-stranded RNA unwinding/modifying activity: a high-performance liquid chromatography-mass spectrometry analysis. Biochemistry , 30, 11507–14. [Poulsen et al., 2001] Poulsen, H., Nilsson, J., Damgaard, C. K., Egebjerg, J. and Kjems, J. (2001). CRM1 mediates the export of ADAR1 through a nuclear export signal within the Z-DNA binding domain. Mol Cell Biol , 21, 7862–71. [Pullirsch and Jantsch, 2010] Pullirsch, D. and Jantsch, M. F. (2010). Proteome diversification by adenosine to inosine RNA editing. RNA Biol , 7, 205–12. [Ramaswami and Tanouye, 1989] Ramaswami, M. and Tanouye, M. A. (1989). Two sodium-channel genes in Drosophila: implications for channel diversity. Proc Natl Acad Sci U S A , 86, 2079–82. [Rebagliati and Melton, 1987] Rebagliati, M. R. and Melton, D. A. (1987). Anti- sense Rna Injections in Fertilized Frog Eggs Reveal an Rna Duplex Unwinding Activity. Cell , 48, 599–605. 165 [Reenan, 2005] Reenan, R. A. (2005). Molecular determinants and guided evolution of species-specific RNA editing. Nature , 434, 409–13. [Reenan et al., 2000] Reenan, R. A., Hanrahan, C. J. and Ganetzky, B. (2000). The mle(napts) RNA helicase mutation in drosophila results in a splicing catastrophe of the para Na+ channel transcript in a region of RNA editing. Neuron , 25, 139–49. [Renton et al., 2011] Renton, A. E., Majounie, E., Waite, A., Simon-Sanchez, J., Rollinson, S., Gibbs, J. R., Schymick, J. C., Laaksovirta, H., van Swieten, J. C., Myllykangas, L., Kalimo, H., Paetau, A., Abramzon, Y., Remes, A. M., Kaganovich, A., Scholz, S. W., Duckworth, J., Ding, J., Harmer, D. W., Her- nandez, D. G., Johnson, J. O., Mok, K., Ryten, M., Trabzuni, D., Guerreiro, R. J., Orrell, R. W., Neal, J., Murray, A., Pearson, J., Jansen, I. E., Sondervan, D., Seelaar, H., Blake, D., Young, K., Halliwell, N., Callister, J. B., Toulson, G., Richardson, A., Gerhard, A., Snowden, J., Mann, D., Neary, D., Nalls, M. A., Peu- ralinna, T., Jansson, L., Isoviita, V. M., Kaivorinne, A. L., Holtta-Vuori, M., Iko- nen, E., Sulkava, R., Benatar, M., Wuu, J., Chio, A., Restagno, G., Borghero, G., Sabatelli, M., Heckerman, D., Rogaeva, E., Zinman, L., Rothstein, J. D., Sendtner, M., Drepper, C., Eichler, E. E., Alkan, C., Abdullaev, Z., Pack, S. D., Dutra, A., Pak, E., Hardy, J., Singleton, A., Williams, N. M., Heutink, P., Pickering-Brown, S., Morris, H. R., Tienari, P. J. and Traynor, B. J. (2011). A hexanucleotide repeat expansion in C9ORF72 is the cause of chromosome 9p21-linked ALS-FTD. Neuron , 72, 257–68. [Rieder et al., 2013] Rieder, L., Staber, C., Hoopengardner, B. and Reenan, R. (2013). Tertiary structural elements determine extent and specificity of messenger RNA editing. Nature Communications , accepted. [Rieder and Reenan, 2011] Rieder, L. E. and Reenan, R. A. (2011). The intricate relationship between RNA structure, editing, and splicing. Semin Cell Dev Biol , 23, 281–8. [Rieu and Powers, 2009] Rieu, I. and Powers, S. J. (2009). Real-time quantitative RT-PCR: design, calculations, and statistics. Plant Cell , 21, 1031–3. [Rinkevich et al., 2012] Rinkevich, F., Schweitzer, P. and Scott, J. (2012). Antisense sequencing improves the accuracy and precision of A-to-I editing measurements using the peak height ratio method. BMC Res Notes , 5, 63. [Rodriguez et al., 2012] Rodriguez, J., Menet, J. S. and Rosbash, M. (2012). Nascent-seq indicates widespread cotranscriptional RNA editing in Drosophila. Mol Cell , 47, 27–37. [Rong and Golic, 2000] Rong, Y. and Golic, K. (2000). Gene targeting by homolo- gous recombination in Drosophila. Science , 288, 2013–2018. [Rosenberg et al., 2011] Rosenberg, B. R., Dewell, S. and Papavasiliou, F. N. (2011). Identifying mRNA editing deaminase targets by RNA-Seq. Methods Mol Biol , 718, 103–19. 166 [Rosenthal and Bezanilla, 2002] Rosenthal, J. J. and Bezanilla, F. (2002). Extensive editing of mRNAs for the squid delayed rectifier K+ channel regulates subunit tetramerization. Neuron , 34, 743–57. [Rosenthal and Seeburg, 2012] Rosenthal, J. J. and Seeburg, P. H. (2012). A-to-I RNA editing: effects on proteins key to neural excitability. Neuron , 74, 432–9. [Rueter et al., 1999] Rueter, S. M., Dawson, T. R. and Emeson, R. B. (1999). Reg- ulation of alternative splicing by RNA editing. Nature , 399, 75–80. [Ryman et al., 2007] Ryman, K., Fong, N., Bratt, E., Bentley, D. L. and Ohman, M. (2007). The C-terminal domain of RNA Pol II helps ensure that editing precedes splicing of the GluR-B transcript. RNA , 13, 1071–8. [Ryter and Schultz, 1998] Ryter, J. M. and Schultz, S. C. (1998). Molecular basis of double-stranded RNA-protein interactions: structure of a dsRNA-binding domain complexed with dsRNA. Embo J , 17, 7505–13. [Sakuma et al., 2013] Sakuma, T., Hosoi, S., Woltjen, K., Suzuki, K., Kashiwagi, K., Wada, H., Ochiai, H., Miyamoto, T., Kawai, N., Sasakura, Y., Matsuura, S., Okada, Y., Kawahara, A., Hayashi, S. and Yamamoto, T. (2013). Efficient TALEN construction and evaluation methods for human cell and animal popula- tions. Genes Cells , 18, 315–326. [Samuel, 2011] Samuel, C. E. (2011). ADARs: Viruses and Innate Immunity. Curr Top Microbiol Immunol , 353. [Sansam et al., 2003] Sansam, C. L., Wells, K. S. and Emeson, R. B. (2003). Modu- lation of RNA editing by functional nucleolar sequestration of ADAR2. Proc Natl Acad Sci U S A , 100, 14018–23. [Savva et al., 2012] Savva, Y. A., Jepson, J. E., Sahin, A., Sugden, A. U., Dorsky, J. S., Alpert, L., Lawrence, C. and Reenan, R. A. (2012). Auto-regulatory RNA editing fine-tunes mRNA re-coding and complex behaviour in Drosophila. Nat Commun , 3, 790. [Scadden, 2005] Scadden, A. D. (2005). The RISC subunit Tudor-SN binds to hyper- edited double-stranded RNA and promotes its cleavage. Nat Struct Mol Biol , 12, 489–96. [Scadden and Smith, 2001] Scadden, A. D. and Smith, C. W. (2001). RNAi is an- tagonized by A–¿I hyper-editing. EMBO Rep , 2, 1107–11. [Schmucker et al., 2000] Schmucker, D., Clemens, J. C., Shu, H., Worby, C. A., Xiao, J., Muda, M., Dixon, J. E. and Zipursky, S. L. (2000). Drosophila Dscam is an axon guidance receptor exhibiting extraordinary molecular diversity. Cell , 101, 671–84. [Schoft et al., 2007] Schoft, V. K., Schopoff, S. and Jantsch, M. F. (2007). Regula- tion of glutamate receptor B pre-mRNA splicing by RNA editing. Nucleic Acids Res , 35, 3723–32. 167 [Seeburg, 2000] Seeburg, P. (2000). RNA helicase participates in the RNA editing game. Neuron , 25, 261–263. [Seeburg et al., 2001] Seeburg, P. H., Single, F., Kuner, T., Higuchi, M. and Spren- gel, R. (2001). Genetic manipulation of key determinants of ion flow in glutamate receptor channels in the mouse. Brain Res , 907, 233–43. [Serganov and Nudler, 2013] Serganov, A. and Nudler, E. (2013). A decade of ri- boswitches. Cell , 152, 17–24. [Shah et al., 2009] Shah, S. P., Morin, R. D., Khattra, J., Prentice, L., Pugh, T., Burleigh, A., Delaney, A., Gelmon, K., Guliany, R., Senz, J., Steidl, C., Holt, R. A., Jones, S., Sun, M., Leung, G., Moore, R., Severson, T., Taylor, G. A., Teschendorff, A. E., Tse, K., Turashvili, G., Varhol, R., Warren, R. L., Watson, P., Zhao, Y., Caldas, C., Huntsman, D., Hirst, M., Marra, M. A. and Aparicio, S. (2009). Mutational evolution in a lobular breast tumour profiled at single nucleotide resolution. Nature , 461, 809–13. [Shepard and Hertel, 2008] Shepard, P. J. and Hertel, K. J. (2008). Conserved RNA secondary structures promote alternative splicing. RNA , 14, 1463–9. [Sixsmith and Reenan, 2007] Sixsmith, J. and Reenan, R. A. (2007). Comparative genomic and bioinformatic approaches for the identification of new adenosine-to- inosine substrates. Methods Enzymol , 424, 245–64. [Smith et al., 1995] Smith, A. J., De Sousa, M. A., Kwabi-Addo, B., Heppell-Parton, A., Impey, H. and Rabbitts, P. (1995). A site-directed chromosomal translocation induced in embryonic stem cells by Cre-loxP recombination. Nat Genet , 9, 376–85. [Smith, 2004] Smith, L. (2004). Genetic interactions between splicing and RNA editing in Drosophila. PhD thesis,, University of Connecticut. [Song et al., 2004] Song, W., Liu, Z., Tan, J., Nomura, Y. and Dong, K. (2004). RNA editing generates tissue-specific sodium channels with distinct gating properties. The Journal of Biological Chemistry , 279, 32554–32561. [Staber et al., 2011] Staber, C. J., Gell, S., Jepson, J. E. and Reenan, R. A. (2011). Perturbing A-to-I RNA editing using genetics and homologous recombination. Methods Mol Biol , 718, 41–73. [Stapleton et al., 2006] Stapleton, M., Carlson, J. W. and Celniker, S. E. (2006). RNA editing in Drosophila melanogaster: New targets and functional conse- quences. RNA , 12, 1922–32. [Stefl et al., 2010] Stefl, R., Oberstrass, F. C., Hood, J. L., Jourdan, M., Zimmer- mann, M., Skrisovska, L., Maris, C., Peng, L., Hofr, C., Emeson, R. B. and Allain, F. H. (2010). The solution structure of the ADAR2 dsRBM-RNA complex reveals a sequence-specific readout of the minor groove. Cell , 143, 225–37. [Stefl et al., 2005] Stefl, R., Skrisovska, L. and Allain, F. H. (2005). RNA sequence- and shape-dependent recognition by proteins in the ribonucleoprotein particle. EMBO Rep , 6, 33–8. 168 [Stefl et al., 2006] Stefl, R., Xu, M., Skrisovska, L., Emeson, R. B. and Allain, F. H. (2006). Structure and specific RNA binding of ADAR2 double-stranded RNA binding motifs. Structure , 14, 345–55. [Stephens et al., 2004] Stephens, O. M., Haudenschild, B. L. and Beal, P. A. (2004). The binding selectivity of ADAR2’s dsRBMs contributes to RNA-editing selectiv- ity. Chem Biol , 11, 1239–50. [Stern et al., 1990] Stern, M., Kreber, R. and Ganetzky, B. (1990). Dosage effects of a Drosophila sodium channel gene on behavior and axonal excitability. Genetics , 124, 133–43. [Sudarsan et al., 2003] Sudarsan, N., Barrick, J. E. and Breaker, R. R. (2003). Metabolite-binding RNA domains are present in the genes of eukaryotes. RNA , 9, 644–7. [Sun et al., 2012] Sun, L., Gilligan, J., Staber, C., Schutte, R. J., Nguyen, V., O’Dowd, D. K. and Reenan, R. (2012). A knock-in model of human epilepsy in Drosophila reveals a novel cellular mechanism associated with heat-induced seizure. J Neurosci , 32, 14145–55. [Sundaram et al., 2013] Sundaram, P., Kurniawan, H., Byrne, M. E. and Wower, J. (2013). Therapeutic RNA aptamers in clinical trials. Eur J Pharm Sci , 48, 259–71. [Sung et al., 2013] Sung, Y., Baek, I., Kim, D., Jeon, J., Lee, J., Lee, K., Jeong, D., Kim, J. and Lee, H. (2013). Knockout mice created by TALEN-mediated gene targeting. Nat Biotechnol , 31, 23–24. [Suzuki et al., 1971] Suzuki, D. T., Grigliatti, T. and Williamson, R. (1971). Temperature-sensitive mutations in Drosophila melanogaster. VII. A mutation (para-ts) causing reversible adult paralysis. Proc Natl Acad Sci U S A , 68, 890–3. [Takuma et al., 1999] Takuma, H., Kwak, S., Yoshizawa, T. and Kanazawa, I. (1999). Reduction of GluR2 RNA editing, a molecular change that increases calcium influx through AMPA receptors, selective in the spinal ventral gray of patients with amyotrophic lateral sclerosis. Ann Neurol , 46, 806–15. [Tan et al., 2002] Tan, J., Liu, Z., Nomura, Y., Goldin, A. and Dong, K. (2002). Alternative splicing of an insect sodium channel gene generates pharmacologically distinct sodium channels. The Journal of Neuroscience , 22, 5300–5309. [Thackeray and Ganetzky, 1994] Thackeray, J. R. and Ganetzky, B. (1994). Devel- opmentally regulated alternative splicing generates a complex array of Drosophila para sodium channel isoforms. J Neurosci , 14, 2569–78. [Thackeray and Ganetzky, 1995] Thackeray, J. R. and Ganetzky, B. (1995). Con- served alternative splicing patterns and splicing signals in the Drosophila sodium channel gene para. Genetics , 141, 203–14. [Tian et al., 2008] Tian, N., Wu, X., Zhang, Y. and Jin, Y. (2008). A-to-I editing sites are a genomically encoded G: implications for the evolutionary significance and identification of novel editing sites. RNA , 14, 211–6. 169 [Tian et al., 2011] Tian, N., Yang, Y., Sachsenmaier, N., Muggenhumer, D., Bi, J., Waldsich, C., Jantsch, M. F. and Jin, Y. (2011). A structural determinant required for RNA editing. Nucleic Acids Res , 39, 5669–81. [Tonkin and Bass, 2003] Tonkin, L. A. and Bass, B. L. (2003). Mutations in RNAi rescue aberrant chemotaxis of ADAR mutants. Science , 302, 1725. [Tonkin et al., 2002] Tonkin, L. A., Saccomanno, L., Morse, D. P., Brodigan, T., Krause, M. and Bass, B. L. (2002). RNA editing by ADARs is important for normal behavior in Caenorhabditis elegans. EMBO J , 21, 6025–35. [Toor et al., 2008] Toor, N., Keating, K. S., Taylor, S. D. and Pyle, A. M. (2008). Crystal structure of a self-spliced group II intron. Science , 320, 77–82. [Tuerk and Gold, 1990] Tuerk, C. and Gold, L. (1990). Systematic evolution of lig- ands by exponential enrichment: RNA ligands to bacteriophage T4 DNA poly- merase. Science , 249, 505–10. [Valencia et al., 2008] Valencia, P., Dias, A. P. and Reed, R. (2008). Splicing pro- motes rapid and efficient mRNA export in mammalian cells. Proc Natl Acad Sci U S A , 105, 3386–91. [Valente and Nishikura, 2005] Valente, L. and Nishikura, K. (2005). ADAR gene family and A-to-I RNA editing: diverse roles in posttranscriptional gene regula- tion. Prog Nucleic Acid Res Mol Biol , 79, 299–338. [Valente and Nishikura, 2007] Valente, L. and Nishikura, K. (2007). RNA binding- independent dimerization of adenosine deaminases acting on RNA and dominant negative effects of nonfunctional subunits on dimer functions. J Biol Chem , 282, 16054–61. [Van Deursen et al., 1995] Van Deursen, J., Fornerod, M., Van Rees, B. and Grosveld, G. (1995). Cre-mediated site-specific translocation between nonhomol- ogous mouse chromosomes. Proc Natl Acad Sci U S A , 92, 7376–80. [Wahlstedt et al., 2009] Wahlstedt, H., Daniel, C., Enstero, M. and Ohman, M. (2009). Large-scale mRNA sequencing determines global regulation of RNA edit- ing during brain development. Genome Res , 19, 978–86. [Wahlstedt and Ohman, 2011] Wahlstedt, H. and Ohman, M. (2011). Site-selective versus promiscuous A-to-I editing. Wiley Interdiscip Rev RNA , 2, 761–71. [Wan et al., 2011] Wan, Y., Kertesz, M., Spitale, R. C., Segal, E. and Chang, H. Y. (2011). Understanding the transcriptome through RNA structure. Nat Rev Genet , 12, 641–55. [Wan et al., 2012] Wan, Y., Qu, K., Ouyang, Z., Kertesz, M., Li, J., Tibshirani, R., Makino, D. L., Nutter, R. C., Segal, E. and Chang, H. Y. (2012). Genome-wide measurement of RNA folding energies. Mol Cell , 48, 169–81. [Wang et al., 2000] Wang, Q., O’Brien, P. J., Chen, C. X., Cho, D. S., Murray, J. M. and Nishikura, K. (2000). Altered G protein-coupling functions of RNA editing isoform and splicing variant serotonin2C receptors. J Neurochem , 74, 1290–300. 170 [Ward et al., 2011] Ward, S. V., George, C. X., Welch, M. J., Liou, L. Y., Hahm, B., Lewicki, H., de la Torre, J. C., Samuel, C. E. and Oldstone, M. B. (2011). RNA editing enzyme adenosine deaminase is a restriction factor for controlling measles virus replication that also is required for embryogenesis. Proc Natl Acad Sci U S A , 108, 331–6. [Winkler et al., 2002] Winkler, W., Nahvi, A. and Breaker, R. R. (2002). Thiamine derivatives bind messenger RNAs directly to regulate bacterial gene expression. Nature , 419, 952–6. [Wolf et al., 2002] Wolf, J., Gerber, A. P. and Keller, W. (2002). tadA, an essential tRNA-specific adenosine deaminase from Escherichia coli. Embo J , 21, 3841–51. [Wong et al., 2001] Wong, S. K., Sato, S. and Lazinski, D. W. (2001). Substrate recognition by ADAR1 and ADAR2. RNA , 7, 846–58. [Woolf et al., 1995] Woolf, T. M., Chase, J. M. and Stinchcomb, D. T. (1995). To- ward the therapeutic editing of mutated RNA sequences. Proc Natl Acad Sci U S A , 92, 8298–302. [Wu et al., 1978] Wu, C. F., Ganetzky, B., Jan, L. Y., Jan, Y. N. and Benzer, S. (1978). A Drosophila mutant with a temperature-sensitive block in nerve conduc- tion. Proc Natl Acad Sci U S A , 75, 4047–51. [Wu et al., 2011] Wu, D., Lamm, A. T. and Fire, A. Z. (2011). Competition between ADAR and RNAi pathways for an extensive class of RNA targets. Nat Struct Mol Biol , 18. [Wulff et al., 2011] Wulff, B. E., Sakurai, M. and Nishikura, K. (2011). Elucidating the inosinome: global approaches to adenosine-to-inosine RNA editing. Nat Rev Genet , 12, 81–5. [Xia et al., 2005] Xia, S., Yang, J., Su, Y., Qian, J., Ma, E. and Haddad, G. G. (2005). Identification of new targets of Drosophila pre-mRNA adenosine deami- nase. Physiol Genomics , 20, 195–202. [Xing and Li, 2009] Xing, D. and Li, Q. Q. (2009). Alternative polyadenylation: a mechanism maximizing transcriptome diversity in higher eukaryotes. Plant Signal Behav , 4, 440–2. [Yang et al., 2006] Yang, W., Chendrimada, T. P., Wang, Q., Higuchi, M., Seeburg, P. H., Shiekhattar, R. and Nishikura, K. (2006). Modulation of microRNA pro- cessing and expression through RNA editing by ADAR deaminases. Nat Struct Mol Biol , 13, 13–21. [Yang et al., 2008] Yang, Y., Lv, J., Gui, B., Yin, H., Wu, X., Zhang, Y. and Jin, Y. (2008). A-to-I RNA editing alters less-conserved residues of highly conserved coding regions: implications for dual functions in evolution. RNA , 14, 1516–25. [Zhang et al., 2011] Zhang, J., Kuo, C. C. and Chen, L. (2011). GC content around splice sites affects splicing through pre-mRNA secondary structures. BMC Ge- nomics , 12, 90. 171 [Zhang and Grosse, 2004] Zhang, S. and Grosse, F. (2004). Multiple functions of nuclear DNA helicase III (RNA helicase A) in nucleic acid metabolism. Acta Biochimica et Biophysica Sinica , 36, 177–183. [Zhang and Carmichael, 2001] Zhang, Z. and Carmichael, G. G. (2001). The fate of dsRNA in the nucleus: a p54(nrb)-containing complex mediates the nuclear retention of promiscuously A-to-I edited RNAs. Cell , 106, 465–75. [Zu et al., 2013] Zu, Y., Tong, X., Wang, Z., Liu, D., Pan, R., Li, Z., Hu, Y., Luo, Z., Huang, P., Wu, Q., Zhu, Z., Zhang, B. and Lin, S. (2013). TALEN-mediated precise genome modification by homologous recombination in zebrafish. Nature Methods , 10, 329–331.