Order Number 0S07T65

RNA splicing inNeuroapora mitochondria: The interaction of tyrosyl-tRNA synthetase and group I

Guo, Qingbin, Ph.D.

The Ohio State University, 1992

U MI 300 N.ZeebRd. Ann Arbor, MI 48106 RNA SPLICING IN NEUROSPORA MITOCHONDRIA: THE INTERACTION OF TYROSYL-tRNA SYNTHETASE AND GROUP I INTRONS

Dissertation

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

By

Qingbin Guo, B.S.

* * * *

The Ohio State University 1992

Dissertation Committee: Approved by A.M. Lambowitz G.A. Marzluf AdvisaCj- C.A. Breitenberger Ohio State Biochemistry Program D.E . Schumm To my friend, Ching Dai, who was a Ph.D. candidate at The Ohio State University and passed away on January 15, 1990,

To my mother, my wife, and my son. ACKNOWLEDGEMENTS

This dissertation was made possible through the support, encouragement, and discussion of many people, who cannot possibly be mentioned here individually. I owe special gratitude to my dissertation advisor, Dr. Alan M. Lambowitz, for his advice, support, and friendship throughout this dissertation. I would also like to thank Dr. Roland Saldanha for his intelligence, suggestions, and generosity, both in science and in life, which had an influence in my life. I am also indebted to all of the members of the Lambowitz laboratory, whose talents, encouragements, discussions, and friendship made this dissertation work so joyous. I thank Drs. Caroline A. Breitenberger, George A. Marzluf, and Dorothy E. Schumm for serving on my dissertation committee. Finally, I wish to thank my mother, Guilian Zhao, for her help in the last year of this dissertation work. VITA

January 30, 1963...... Born - Harbin, P.R. China. July, 1985 ...... B.S. in Biology, Beijing University, Beijing, P.R. China. 1986 - Present...... Graduate student in Biochemistry Program, The Ohio State University, Columbus, Ohio.

PUBLICATIONS Guo, Q., Akins, R.A., Garriga, G. and Lambowitz, A.M. (1991). Structural analysis of the Neurospora mitochondrial large rRNA and construction of a mini-intron that shows -dependent splicing. J. Biol. Chem. 266, 1809-1819. Guo, Q. and Lambowitz, A.M. (1992). A tyrosyl-tRNA synthetase binds specifically to the catalytic core of a group I intron. Dev. 6 , 1357-1372.

FIELD OF STUDY Major Field: Biochemistry

iv TABLE OF CONTENTS

DEDICATIONS...... ii ACKNOWLEDGEMENTS...... iii VITA...... iv ABBREVIATIONS...... viii LIST OF TABLES...... ix LIST OF FIGURES...... x CHAPTER PAGE I. INTRODUCTION...... 1

Introns ...... 1

Group I introns...... 6 The Structure of Group I Introns...... 7 Sequence and Structural Requirements for the Splicing Activity of Group I Introns...... 9 Protein-dependent Group I Introns...... 10 Experimental System...... 13 and its Mitochondria...... 13 Neurospora Nuclear Mutants Defective in Splicing of the Mt Group I Introns...... 15 The cyt-18 and the CYT-18 Protein...... 17 II. MATERIALS AND METHODS...... 30 Strains of Neurospora and Growth Conditions...... 30 Isolation of Mt RNA and RNP Particles...... 31 Construction of mutant Neurospora Mt LSU Introns Used for Structural Analysis...... 31 Construction of Mutant Introns Used for the Localization of the CYT-18 Protein Binding Site on the Neurospora Mt Large rRNA Intron 35 Construction of Other Plasmids...... 37 Preparation of the CYT-18 Protein from Neurospora and E . coli...... 39 v In Vitro Synthesis of Neurospora Mt tRNATyr and Other tRNAs...... 39 Synthesis of in vitro Transcripts Containing Group I Introns...... 41 Splicing Assays...... 42 TyrRS Assays...... 43 Nitrocellulose Filter Binding Assays...... 45 UV-Crosslinking...... 4 6 III. STRUCTURAL ANALYSIS 07 THE NEUROSPORA MITOCHONDRIAL LARGE rRNA INTRON AND CONSTRUCTION 07 A MINI-INTRON THAT SHOWS PROTEIN-DEPENDENT SPLICING...... 53

introduction...... 53 Results...... 55 Secondary Structure Model of the Mt Large rRNA Intron...... 55 Deletion of the ORF...... 56 Deletions in the P2 Stem...... 57 Deletions in L8 ...... 58 Deletions of P4 and P6 and Nucleotide Substitutions in P9 ...... 58 Truncations at the 3' End of the Intron...... 59 Charaterization of in vitro Splicing of the Mini-Intron...... 60 Modification of the IGS Sequence and P9.0 Pairing...... 62 Comparison of the Splicing and TyrRS Activity of the CYT-18 Protein Isolated from 74A, cyt-4-1, and cyt-19-1...... 64 Discussion...... 65 IV. THE CYT-18 PROTEIN BINDS SPECIFICALLY TO THE CATALYTIC CORE OF Neurospora MITOCHONDRIAL LARGE rRNA INTRON...... 95 Introduction...... 95 Results...... 96 Direct Bindings of the CYT-18 Protein and Neurospora Mt Large rRNA Intron Assayed by UV-cros si inking Method...... 96 Quantitive Analysis of the Binding Using Nitrocellulose Filter Binding Method...... 98 The CYT-18 Binds to the Group I Intron Catalytic Core...... 100 Construction of Small RNAs Containing the CYT-18 Protein Binding Site...... 102

vi Discussion...... 104 V. THE CATALYTIC CORE OF NEUROSPORA MT LARGE rRNA INTRON MAY RESEMBLE THE STRUCTURE OFMT TYROSYL-tRNA...... 131 introduction...... 131 Results...... 132 Binding of the CYT-18 Protein to Other Group 11ntrons...... 132 Competition Between the Neurospora Mt tRNA1*' and Intron RNA for the CYT-18Protein ...... 133 Comparison of Group I Introns That Bind and Are Spliced by the CYT-18 Protein...... 135 Discussion...... 136 VI.REPLACEMENT OF THE LARGE EXTENSION OF P5 ELEMENT IN THE TETRAHYMENA LARGE rRNA INTRON BY NEUROSPORA MT TYROS YL-TRNA SYNTHETASE...... 160 Introduction...... 160 Results...... 161 Nitrocellulose Filter Binding of the CYT-18 Protein and Tetrahymena Large rRNA Intron and Its Derivative...... 161 CYT-18 Protein Functions in Splicing of the AP5abc Mutant of the Tetrahymena Large rRNA Intron...... 162 Construction of A Small RNA Containing the CYT-18-Binding Site of the Tetrahymena Intron...... 165 Discussion...... 165 VIII. CONCLUSIONS...... 179 LIST OF REFERENCES...... 183 ABBREVIATIONS bp base pair nt nucleotide kb kilobase kDa kiloDalton Kd dissociation constant IGS internal guide sequence LSU large ribosomal RNA mt mitochondrial Nc Neurospora crassa ORF open reading frame RNP ribonucleoprotein TyrRS tyrosyl-tRNA synthetase

viii LI8T OF TABLES Table 1. Plasmids containing the Neurospora mt large rRNA intron and its derivatives 4 8 2. Comparison of the splicing and TyrRS activity of the CYT-18 protein isolated from 74A, cyt-4-1, and cyt-19-1...... 94 3. The efficiency of the nitrocellulose filter binding between the CYT-18 protein and the Neurospora mt large rRNA intron...... 110 4. Binding of the CYT-18 protein to different introns...... 145 5. The Neurospora rat large rRNA intron inhibits TyrRS activity of the CYT-18 protein...... 150

ix LIST OP FIGURES Figure 1. Splicing mechanisms of the four groups of introns....20 2. The hypothetic secondary and tertiary structures of the group I introns...... 22 3. The hypothetic secondary structure of the Tetrahymena large rRNA intron...... 24 4. The map of the Neurospora crassa mi tochondr ia 1 genome...... 26 5. The comparison of the CYT-18 protein with other tyrosyl-tRNA synthetases...... 28

6 . The map of the plasmid pTYR...... 51 7. Hypothetical secondary structure of the Neurospora mt large rRNA intron...... 71

8 . Effect of deleting the ORF and all or part of the central hairpin on in vitro splicing of the mt large rRNA intron...... 73

9. Effect of deletions in L8 and P8 on in vitro splicing of the mt large rRNA intron...... 77

10. Effect of deletions in P4/P5 and P6 and nucleotide substitutions in P9 on in vitro splicing of the mt large rRNA intron...... 79 11. Effect of deletions from the 3' exon and 3 ' end of the intron on activity of the mt large rRNA intron...... 81 12. In vitro splicing of 388 nt mini-intron RNA under different conditions...... 83 13. Effect of high GTP or guanosine concentrations on in vitro splicing of 388 nt mini-intron...... 85 14. Effect of changing the IGS sequence on in vitro splicing of the mt large rRNA intron...... 87 x 15. Effect of modifying the P9.0 pairing on in vitro splicing of the mt large rRNA intron...... 89 16. Comparison of the TyrRS and splicing activities of the CYT-18 protein isolated from 74A, cyt-4-1, and cyt-19-1...... 92 17. Predicted secondary structure of the 388 nt derivative of the Neurospora mt large rRNA intron...106 18. UV-crosslinking experiment showing that the CYT-18 binds the Neurospora mt large rRNA intron directly...... 108 19. The efficiency of the nitrocellulose filter binding between the CYT-18 protein and the Neurospora mt large rRNA intron...... 112 20. Fraction of the active CYT-18 protein for the nitrocellulose filter binding experiments...... 114 21. Nitrocellulose filter binding of the CYT-18 protein to the Neurospora mt large rRNA intron...... 116 22. Stoichiometry of the complex between the CYT-18 protein and pBD5A/BanI intron RNA...... 118 23. Nitrocellulose filter binding of the CYT-18 protein to derivatives of the Neurospora mt large rRNA intron...... 120 24. Summary of the nitrocellulose filter binding of the CYT-18 protein to derivatives of the Neurospora mt large rRNA intron...... 122 25. Binding of the Neurospora mt large rRNA intron competed by a series of RNAs having 3' truncations of intron sequences...... 125 26. Inhibition of splicing activity of CYT-18 protein by a series of RNA derivatives of the Neurospora mt large rRNA intron truncated from the 3' end...... 127 27. Binding of CYT-18 protein to small RNAs containing different regions of the Neurospora mt large rRNA intron catalytic core...... 129 28. Splicing activity of CYT-18 protein is inhibited by the Neurospora mt tRNATyr...... 148

xi 29. Competitive inhibition of TyrRS activity of the CYT-18 protein by the Neurospora mt large rRNA intron...... 152 30. Comparison of the different introns that bind and are spliced by the CYT-18 protein...... 154 31. Comparison of the three-dimensional structures of group I introns and tRNA...... 156 32. Comparison of the Neurospora mt tRNA1^" and the Neurospora mt large rRNA intron core...... 158 33. The predicted secondary structure of the AP5abc derivative of the Tetrahymena large rRNA intron.... 172 34. The nitrocellulose filter binding of the CYT-18 protein with the Tetrahymena large rRNA intron or its P5abc deletion mutant (AP5abc)...... 174 35. The CYT-18 protein can rescue the splicing of the P5abc deletion mutant of the Tetrahymena large rRNA intron (AP5abc)...... 176

xii CHAPTER I INTRODUCTION

The discovery of introns is an important landmark in modern biology. This discovery has given rise to questions about the origin of introns and the mechanisms by which they are removed from the precursor RNAs so that mature RNAs can be produced for their biological function. In the 15 years since the discovery of the intron (Chow et al., 1977; Berget et al., 1977), considerable progress has been made toward understanding the mechanisms of the splicing reactions. In this chapter, I will first briefly review our current knowledge about all introns. Then I will describe some important aspects of group I introns on which my dissertation research has been focused. Finally, I will introduce our experimental system and provide a rationale for my dissertation work.

Introns Introns are common in eukaryotic genomes, but they have also been found in some genes of prokaryotes. They exist in genes encoding ribosomal RNA, transfer RNA, and messenger RNA in nuclear, mitochondrial, chloroplast, and viral genomes. The transcripts of such genes undergo precise removal of 1 2 introns to produce functional rRNA, tRNA and mRNA. Based on the sequence and RNA secondary structures, introns were originally classified into four major groups, i.e., group X, group II, nuclear mRNA, and tRNA introns (Cech, 1990). However, recent studies have found some introns that cannot be categorized into any of these four groups of introns. For example, Christopher and Hallick (1989) have found four novel introns in a chloroplast ribosomal protein operon in Euglena gracilis and categorized them as group III introns. The introns in the large rRNA gene and some tRNA genes of the archaebacteria are also different from the above four classes of introns and use a different mechanism for splicing (Kjems and Garrett, 1988; cf. Thompson and Daniels, 1990). Most of our knowledge about intron splicing mechanisms has been obtained through biochemical studies using in vitro splicing systems. These studies have shown that the original four groups of introns not only have different conserved sequences and secondary structure, but also use different splicing mechanisms. These splicing mechanisms are illustrated schematically in Figure 1 and briefly described as follows. (1) Many eukaryotic nuclear tRNAs have an intron in the anticodon arm. The introns in the precursor tRNAs are spliced out by two enzymes through a series of consecutive steps (Peebles et al., 1983; Greer et al., 1983). First, an endonuclease cleaves at the 5' and 3' splice sites. This 3 step produces a 2', 3'-cyclic phosphate at the 3/ end of the 5' exon and a hydroxyl group at 5' end of the 3' exon. Second, an RNA ligase with several activities adds a phosphate to the hydroxyl group of the 5' end of the 3' exon, then adenylates it. Third, the same RNA ligase with the phosphodiesterase activity breaks the 2'-3/ cyclic phosphate bond to yield a 2' phosphate at the 3' end of the 5' exon. Fourth, the RNA ligase ligates the two exons by forming a 3'- 5' phosphodiester bond. The phosphate in the 3'-S' phosphodiester linkage is derived from the 7-Phosphate of ATP in , but from the precursor tRNA in Hela cells (Winicov and Button, 1982; Filipowicz and Shatkin, 1983). Finally, the 2 ' phosphate in the ligated product is removed by a phosphatase to form mature tRNA. (2) The splicing of nuclear mRNA introns proceeds by two transesterification reactions (Padgett et al., 1986). First, the 2'-hydroxyl group of "A" at the branch site of the intron attacks the phosphate at the 5' splice junction and forms a

2'-51 phosphodiester bond. This step produces a free 3' hydroxyl group at the 3' end of the 5' exon and a lariat RNA containing the intron and 3' exon. Second, the hydroxyl group at the 3 ' end of the 5' exon attacks the phosphate group at the 3' splice junction. This results in cleavage at the 3 ’ junction and ligation of the two exons. ATP is required for the splicing of nuclear mRNA introns. (3) The splicing of group II introns is mechanistically 4 similar to that of the nuclear mRNA introns, as shown in Figure 1. First, the 2' hydroxyl group of a nucleotide in the intron attacks the phosphate group at the 5' splice site. This results in cleavage at the 5' splice site and formation of a lariat RNA containing the intron and 3' exon. Then, the 3' hydroxyl group of the 5/ exon attacks the 3# splice site, leading to cleavage at 3' splice site and ligation of the two exons. The in vitro splicing of group II introns does not require ATP. (4) Like the splicing of the nuclear mRNA and group II introns, the splicing of group I introns also proceeds through two consecutive transesterification reactions. The difference in the pathways between the nuclear mRNA introns, group II introns and group I introns is that the first transesterification reaction of group I introns is initiated by guanosine or its 5'-phosphorylated forms (GMP, GDP or

GTP), instead of a 2' hydroxyl group of a nucleotide in the nuclear mRNA and group II introns. Therefore, no lariat intermediate of the intron is formed during the splicing of group I introns. The second step is the same as nuclear mRNA and group II introns. The 3' hydroxyl group attacks the phosphate at the 3' splice site and results in the excision of the intron and ligation of the two exons. After excision from the precursor RNA, some group I introns, such as the Tetrahymena large rRNA intron, subsequently cyclize themselves to form circular molecules (Zaug et al., 1983). 5 During the cyclization, the excised linear intron is cleaved at a site 15 nucleotides from the 5' end and a phosphodiester linkage is formed between the 3'-hydroxyl group and the new 5' phosphate of the intron. This cyclization reaction is also intron-catalyzed. In contrast, some other group I introns do not form circular RNA molecules (e.g., Neurospora mt large rRNA intron; Garriga and Lambowitz, 1986). The sequence or structural requirements in the group I introns for the cyclization reaction are still not clear. The similarity of tranesterification reactions in the splicing of the nuclear mRNA, group II and group I introns suggests that these introns may be evolutionarily related (Cech, 1986; Cech and Bass, 1986; Padgett et al., 1986). It has been proposed that present nuclear mRNA introns are derived from self-splicing introns (Padgett et al., 1986; see Chapter VI for further discussion). Despite the mechanistic similarities, many differences exist in these introns. For example, some introns can self-splice in vitro, while others require to splice. All of the nuclear tRNA introns and mRNA introns require proteins to splice. As mentioned above, the splicing of tRNA introns involves two distinct enzymes, endonuclease and RNA ligase. Nuclear mRNA introns require many proteins and small nuclear RNA molecules (snRNA) to splice both in vitro and in vivo. These proteins and RNA molecules form a complex molecular apparatus termed the "spliceosome". The process of splice-site recognition and catalysis takes place in these catalytically active spliceosomes. Although most group I and group II introns also require protein factors to splice efficiently in vivo (Lambowitz and Perlman, 1990), many group I and group II introns have been shown to splice without any protein factors under in vitro conditions (Peebles et al., 1986; Cech, 1990). The discovery of the self-splicing Tetrahymena large rRNA intron by Cech and co-workers has changed our traditional concept of the nature of "enzymes" (Cech et al., 1981; Kruger et al., 1982). Provided with only Mg++ and guanosine, the Tetrahymena large rRNA intron can be spliced in vitro in an appropriate buffer without any protein catalyst or energy source. Since then, it has been realized that RNA could be an enzyme, named ribozyme. The studies of the ribozymes have brought a profound insight into the fundamental mechanisms of gene regulation.

Group I Introns

Group I introns are found in many fungal mitochondrial genes, some chloroplast and bacteriophage genes, the large rRNA genes of Tetrahymena, Physarum, and small or large rRNA genes of Pneumocystis and some algae (Cech, 1988) . They have also been found in bacterial tRNA genes (Kuhsel et al., 1990; Xu et al., 1990; Reinhold-Hurek and Shub, 1992). In this section, I will focus on three aspects of group I introns which are directly related to my dissertation work. 7 (1) The structure of group I introns. All group I introns share a common secondary structure, although different group I introns have minimal sequence (< 10%; Cech, 1990; Michel and Westhof, 1990). This conserved secondary structure consists of four major conserved sequence elements P, Q, R, S and a number of base paired stems, named PI to Pll, which fold the intron into a conserved structure required for splicing (see Figure 2) . In addition to P, Q, R, and S elements, other highly conserved nucleotides of the group I introns include the U residue preceding the 5' splice site and the G residue to which it pairs, a G-C base pair in P3, and the G which is the last nucleotide of the intron (Cech, 1988). Group I introns also have an internal guide sequence (IGS) which pairs with exon sequences adjacent to the 5' and 3' splice-sites and helps to bring the two splice-sites together (Cech, 1988). The hypothetical secondary structure of the best-studied group I intron, the Tetrahymena large rRNA intron, is shown in Figure 3. Based on sequences and ancillary structures, group I introns can be further categorized into several subgroups. Cech divided group I introns into two subgroups, group IA and group IB (Cech, 1988). Group IA introns have a stem or a pair of stems between P7[5'] and P3[3'], designated P7.1 or P7.1/P7.2. Group IB introns do not have an extra stem at this position, and the P7[5'] and P3[3'] are separated typically by a single nucleotide. In addition, the 8 bulged nucleotide of P7 in most group IA introns is a "C", but it is an "A" or "C" in the group IB introns. Almost all group IB introns have a large extension of the P5 stem, which is absent in most if not all group IA introns. Recently, Michel and Westhof (1990) have classified group I introns into 4 major subgroups and 10 minor subgroups based on their core alignments and a distance matrix by computer analysis. The group IA in the latter classification is equivalent to the group IA in Cech's classification. The group IB and IC are equivalent to group IB in Cech's classification. Group ID introns in the classification of Michel and Westhof have neither the extra stem between P7[5'] and P3[3], nor large extension of P5 stem; these introns were classified into group IB by Cech (1988) . The tertiary structure of the group I introns has not been experimentally determined, but a three-dimensional model of the conserved core of group I introns has been proposed recently by Michel and Westhof (1990). In this model, the core of group I introns consists of two major helices formed by coaxial stacking of the base-paired stems. One helix contains P5, P4, P6 and P6a, while the other contains P9, P7,

P3, and P8 . These two helices form a cleft containing the catalytically active site and guanosine-binding site. The 5' and 3' splice sites of the intron are brought together to the catalytically active site in the cleft and spliced. Two important features in this model are that it is compact and 9 that the most evolutionarily conserved sequence elements are within or around the putative active site at the cleft formed by the two helices. Although the tertiary structure has been modeled only for the Tetrahymena large rRNA intron and yeast mt large rRNA intron (Michel and Westhof, 1990; Jaeger et al., 1991), it is believed that the general features of this modeled tertiary structure are also shared by other group I introns. (2) Sequence and structural requirements for the splicing activity of group I introns. The sequence and structural requirements for the splicing activity of group I introns have been recently reviewed by Cech (1988; 1990) and Burke (1988). Here I will only focus on some general aspects. All group I introns share a conserved secondary structure, and presumably also a tertiary structure, which are clearly important for their catalytic activity. However, some sequences are located outside of the conserved regions of the conserved structure and can be deleted without losing splicing activity. Some examples of these sequences include the open reading frames (ORFs) in many group I introns, the P2.1 and P6b stems of the Tetrahymena large rRNA intron (see Figure 3; Burke, 1988). The smallest group I intron found in nature has only 204 nucleotides (Kuhsel et al., 1991). Most group I introns can probably be reduced to this size without losing their splicing activities. 10 The sequences and structures which form the catalytic core of group I Introns are generally required for the splicing of these introns. Deficiencies in the catalytic core result in the loss of the splicing activity of group I introns, but these introns may be spliced by splicing factors, such as the CYT-18 protein (Mohr et al., 1992). Some of the sequences outside of the catalytic core are also important for the splicing of group I introns. For example, the PI, P10 and IGS sequences serve to bring the 5' and 3' splice-sites together to the catalytic core. The P5abc stem of the Tetrahymena large rRNA intron is required to stabilize the catalytic core at low Mg2+ concentrations (van der Horst et al., 1991). Some sequences and structures in the group I introns may be required for interaction with protein factors, which help the intron to form the correct core structure required for splicing. (3) Protein-dependent group I introns. The proteins involved in the splicing of group I introns can be classified into two groups. They are intron-encoded proteins, which are also termed maturases, and host-encoded proteins. The intron-encoded proteins, i.e. maturases, which are required for the splicing of group I introns, were identified through genetic studies in the mitochondria of . The maturases encoded in the introns in the yeast mt genes, cob-12, cob-13 and cob-14, are required for 11 splicing of the introns encoding them (Lazowska et al., 1980; Lazowska et al., 1989). Because there is no in vitro splicing assay system available for these introns, whether these maturases function directly on group I introns remains unknown. Intron-encoded proteins are not all maturases. Some proteins encoded in group I introns are not related to their splicing, such as mt ribosomal protein S-5 encoded in the Neurospora mt large rRNA intron (Lambowitz et al., 1985). The host-encoded splicing factors for group I introns are often multifunctional proteins. Some of them function in splicing a number of different group I introns, while the others function in splicing of a single intron (Lambowitz and Perlman, 1990). These host-encoded splicing proteins for the group I introns will be described as follows according to their functions in addition to splicing.

(i) Aminoacyl-tRNA synthetases. The Neurospora mt tyrosyl- tRNA synthetase encoded by the cyt-18 gene functions in splicing of a number of introns, including introns in the genes of mt large rRNA, cob, coxl and ATPase 6. The detailed description of the cyt-18 gene will be given in the Experimental System section of this chapter. Another tRNA synthetase implicated in splicing of group I introns is the yeast mt leucyl-tRNA synthetase encoded by the nuclear NAM2 gene. Some in the NAM2 gene can suppress mutations in the cob-14 maturase for its inability to splice two closely related introns {cob-14 and cox-I4a) in Saccharomyces 12 cerevisiae (Labouesse et al., 1987). It is still not clear whether the mt leucyl-tRNA synthetase can function directly in the splicing of these two introns. (ii) Helicases. The yeast MSS116 gene encodes a protein with an RNA helicase domain. It is required for splicing of some group I and group II introns in the cob and coxl genes and may also function as a translation initiation factor (see Lambowitz and Perlman, 1990). (iii) Other multifunctional proteins. The yeast MSS18 and PET54 genes are both required for the splicing of the coxl- 153 intron and also function in mitochondrial (mt) protein synthesis (Seraphin et al., 1988; 1989; Valencik et al., 1989). Some other examples of multifunctional splicing factors include the Neurospora nuclear cyt-4 gene, yeast NAM1 and suv3 genes (see Lambowitz and Perlman, 1990). (iv) Single-functional proteins. Two proteins identified so far function only in splicing of group I introns (Lambowitz and Perlman, 1990). One is the CBP2 protein of Saccharomyces cerevisiae, which functions in the splicing of the cob-15 intron (McGraw and Tzagoloff, 1983; Gampel et al., 1989). Jn vitro splicing studies have shown that the conserved catalytic core structure is required for the function of the CBP2 protein on the intron (Gampel and Cech, 1991). The other identified protein is yeast MRS1, which cooperates with a maturase to function in the splicing of cob-13 intron (Kreike et al,1986; Kreike et al., 1987). 13 Experimental System (1) Neurospora crassa and its mitochondria. Neurospora crassa, commonly named pink bread mold, is an eukaryotic filamentous fungus. Like other fungi, it has asexual and sexual cycles. During the asexual cycle, the branched multinucleate filaments (termed hyphae) grow to form two kind of spore forms, microconidia and macroconidia (Davis and de Serres, 1970). These two kinds of conidia can grow into the branched hyphae again. During the sexual cycle, the ascogonia (female gametes) are first formed from the branched hyphae under conditions that are not suitable for the asexual cycle. Then the ascogonia fuse with conidia of the opposite mating type. After meiosis, fertilization and mitosis, the ascus containing eight ascospores is formed. The ascospores can grow into branched hyphae, which undergo the next sexual life cycle or enter the asexual cycle (Davis and de Serres,

1970). Neurospora crassa can grow in simple and inexpensive medium in the laboratory. Since its life cycle is relatively short, large amounts of materials for biochemical studies of Neurospora can be obtained in a short period of time. The genetics of Neurospora crassa are well-established. Therefore, Neurospora crassa is an ideal for biochemical and genetic studies. Neurospora crassa contains mitochondria, in which the energy required for metabolism is produced. The Neurospora 14 has its own genome and all the enzymatic machinery required for transcription and translation of the genetic information into functional proteins. The mt genome of Neurospora crassa is a circular double-stranded DNA molecule of 62 kb (Figure 4; Bernard et al., 1976; Terpstra et al., 1977). The mt DNA encodes components of the respiratory chain, subunits of an ATPase that couples the electron transfer to ATP synthesis, a full set of mt tRNAs, and the small and large rRNAs (Figure 4). Many of the genes in the Neurospora mt genome are interrupted by introns. Except for some natural isolates of wild-type strains in which group II introns have been found, all of the introns found in the mitochondria of laboratory strains of Neurospora crassa are group I introns (Collins, 1990). In addition to the 62 kb mt genome, the mitochondria of several strains of Neurospora crassa also have plasmids, which are small circular or linear DNAs (Lambowitz et al., 1985). Three noteworthy features of these plasmids are: (i) Some of mt plasmids may be related to the mt group I introns (Lambowitz et al., 1985); (ii) some of the plasmids have been shown to encode reverse transcriptase (Kuiper and Lambowitz, 1988); (iii) A single-stranded RNA transcribed from VS plasmid DNA of several strains of Neurospora mitochondria has been shown to have the RNA-catalyzed self-cleavage activity (Saville and Collins, 1990). 15 (2) Neurospora nuclear mutants defective in the splicing of mt group I introns. Although some group I introns can self-splice, most require protein factors for efficient splicing in vivo (Lambowitz and Perlman, 1990). To study the splicing of the protein-dependent group I introns, our laboratory has been focusing on the intron in the Neurospora mt large rRNA gene. The Neurospora mt large rRNA gene contains a 2.3 kb intron at a site near the 3' end of the large rRNA and has the characteristic sequences and secondary structure of group I introns, but it cannot self-splice in vitro under any conditions examined (Garriga and Lambowitz, 1983). An in vitro system has been established using Neurospora mt ribonucleoprotein (RNP) and soluble fractions to assay the splicing of this intron (Garriga and Lambowitz, 1986). Studies using this system have shown that the splicing of this intron uses the same mechanism as that first elucidated for the Tetrahymena large rRNA intron. Previous genetic studies in our laboratory have shown that the splicing of the Neurospora mt large rRNA intron is defective in eight Neurospora nuclear mutants. These eight nuclear mutants were mapped to three linkage groups, and thus defined three genes required for the splicing of mt large rRNA intron (Mannella et al., 1979; Bertrand et al., 1982). These three genes are named cyt-4, cyt-18, and cyt-19 because of their cytochrome deficiency. 16 The phenotype of the cyt-4 mutants is very complex. In addition to the deficiency in the splicing of mt large rRNA intron, the cyt-4 mutants are also defective in the 3' end synthesis of the mt large rRNA and accumulate large amounts of several aberrant RNAs derived from the large rRNA precursor (Garriga et al., 1984; Dobinson et al, 1989). The cyt-4 mutants are also defective in processing a number of other mt RNAs, which include cob, coJ, coll, and ATPase6 mRNAs and some tRNAs (Dobinson et al., 1989). Sequencing of the cyt-4 gene has shown that the protein encoded by cyt-4 has significant similarity to the SSD1/SRK1 protein of Saccharomyces cerevisiae and DIS3 protein of Schizosaccharomyces pombe, which may function as protein phosphatases that regulate cell cycle and mitotic chromosome segregation (Turcq et al., 1992). This suggests that RNA splicing and processing may be regulated by protein phosphorylation. However, how the CYT-4 protein promotes the splicing of group I introns is still unknown. The cyt-18 and cyt-19 mutants have been shown to be defective in splicing several group I introns in Neurospora mitochondria and may encode proteins that function generally in the splicing of group I introns (Collins and Lambowitz, 1985) . The cyt-19 gene has not been cloned so far and its biochemical function remains unknown. The cyt-18 gene has been cloned and extensively studied in our laboratory. Our current knowledge about the cyt-18 gene and its product CYT- 17 18 protein will be described in the following section. (3) The cyt-18 gene and the CYT-18 protein. The sequencing of the cyt-18 gene indicated that it has an ORF of 669 amino acids, which is homologous to the E. coli tyrosyl-tRNA synthetase. The cyt-18 mutants are deficient in both mt tyrosyl-tRNA synthetase (TyrRS) and splicing of the mt large rRNA intron in vivo (Akins and Lambowitz, 1987). Antibodies raised against TrpE-CYT-18 fusion proteins recognized an approximately 70 kDa protein encoded by the cyt-18 gene. In a variety of ion-exchange chromatographic columns, the highly purified 70 kDa protein has both mt TyrRS and splicing activities when assayed with the established in vitro aminoacylation and splicing methods (Majumder et al., 1989). Gel-filtration indicated that the complex having the tyrosyl-tRNA synthetase and splicing activities is 150 kDa or smaller. This suggested that CYT-18 protein may function in splicing in vitro by itself, in the form of monomer or dimer (Majumder et al., 1989). This assertion was verified by showing that the highly purified CYT-18 protein synthesized in E. coli can splice the Neurospora mt large rRNA intron in vitro by itself (Kittle et al., 1991). Structure-function relationships in the CYT-18 protein have been extensively studied in our laboratory by assaying the in vitro mutants of the CYT-18 protein purified from E, coli for their splicing and aminoacylation activities. Figure 5 compares five different tyrosyl-tRNA synthetases, including two bacterial TyrRSs from Bacillus stearothermophilus and E. coli and three fungal mt TyrRSs from Neurospora, Podospora, and Saccharomyces cerevisiae. These TyrRSs have some similarity in the nucleotide binding fold domain and tRNA binding domain (black box in Figure 5). Compared with bacterial TyrRSs, the mt TyrRSs have a mt targeting domain at the amino-terminus (striped box). Some domains are conserved only between the mt TyrRSs in Neurospora and Podospora, which also have splicing functions (gray boxes). These conserved domains include 65 amino acids at the N-terminus of the fungal mt TyrRSs. This domain has been shown to be required for the splicing activity of the CYT-18 protein, but not for its TyrRS activity when assayed with E. coli tyrosyl-tRNA (Cherniack et al., 1990). Some small linker-insertion mutations within or near the predicted ATP-binding domain abolished TyrRS activity but left substantial splicing activity. This suggested that the normal TyrRS activity of the CYT-18 protein is not required for its splicing activity (Kittle et al., 1991). A number of mutations in the C-terminal domain of the CYT-18 protein result in parallel inhibition of splicing and TyrRS activity. This suggests that these regions may be involved in the binding of both the intron and tRNA (Kittle et al., 1991). The finding that different regions of the CYT-18 protein required for splicing may also be involved in binding the tRNA suggests that similar interactions may be used for the 19 binding of both the intron and tRNA substrates. To understand the mechanism of the CYT-18 protein- dependent splicing reactions of group 1 introns, I have been focusing on how the CYT-18 protein interacts with these group I introns. First, I characterized the structure of the Neurospora mt large rRNA intron to understand the differences between the self-splicing and CYT-18-dependent group I introns and to facilitate the identification of the CYT-18 binding site in the intron. Second, I localized the CYT-18 protein binding site to a small region of the Neurospora mt large rRNA intron. Third, I attempted to understand how the CYT-18 protein promotes splicing by binding to these regions of a group I intron using a variety of approaches. Finally, I extended my studies from CYT-18-dependent introns to the self-splicing Tetrahymena large rRNA intron and studied the effect of the CYT-18 on the splicing of wild-type and P5abc deletion derivative of the Tetrahymena large rRNA intron. These studies have shown that the CYT-18 protein binds to the group I intron core. The CYT-18 protein may promote splicing by stabilizing the catalytically active structure of group I introns. The similarity between the group I introns and tRNA raises the possibility that tRNAs may have evolved from ancestors of group I introns. The studies of the effect of the CYT-18 protein on the self-splicing Tetrahymena large rRNA intron suggest that the protein-dependent introns may be derived from self-splicing introns. Figure 1. splicing mechanisms of the four groups of introns. The figure is from Cech (1990).

20 21

a ) Group I b) Group II C) Nuclear mRNA d ) Nuclear tRNA Self-Splicing Self-Splicing Spliceosomal Enzymatic

o I ENDONUCLEASE -OH

■ k in a s e T LIG A S E PHOSPHATASE

-0™“

+ Q/ MO

Figure 1. Figure 2. The hypothetical secondary and tertiary structures of the group I introns. Left is the secondary structure of group I introns (from Burke et al., 1987) . Right is a tertiary structural model of Tetrahymena large rRNA intron proposed by Michel and Westhof (1990). 5'SS and 3'SS: 5' and 3' splice sites. IGS: internal guide sequence. P, Q, R, and S indicate conserved sequence elements. Pi to P10 and LI to L9 indicate base-paired stems or loops of group I introns. G1 and G414 indicate the first and last guanine nucleotides of Tetrahymena large rRNA intron.

22 3-D Model of Group I Intron

Secondary Structure of Group I Introns P10 P5 P90

P4 G414

5'SS' P7 P4 5 ' - J mini mm I iirri/ P7 PI 013') 3'SS P6a It P3

P8

Figure 2. Figure 3. The hypothetical secondary structure of the Tetrahymena large rRNA intron. The figure is from Burke et al. (1987). The arrows indicate 5' and 3' splice sites. PI to P9 and LI to L9 indicate base-paired stems and loops.

24 erhmn Lre RA Intron rRNA Large Tetrahymena S'- PI 5003 n u , ■c u-G u-ft c-G c-G u-G ft-U ft-U ft-U I G G LI G G - ft • ft- ft - ftft-U— - G-CftHGACCGUC A ft ft U O u u c A u ZUG U-ft PZ U-G ,CG I p, C-G 8 I I I U IL8 111 C 0 C-G C-G 0 C n - UG P3 U-G G-8 U-ft U-H U-ft ft-U fl-U H-U ft-U fl-U G-C C G-C ft G U 2 u“« L2 lR tG RUCAGflC f, u UCUUCUC e u flGRftGGG c t C ft P8 ft—U L2.1 ft n b 5 1 P7 G P2.1 ftUHHGftUH CUGGCUGU CUGGCUGU I I I I II I n. GRCUCU PSD . iflGfl Mill P5a G-CGH IIIII I U lU-GC- CG —R-UWJGGAGUflCUCGuaog G— C-G— UflGUC-GGHCC-G iue 3. Figure

- -O-^0^ U-G P9 h O00"** c 5 P u GcUlG yflGUUCUC ,uUGHcoUflGGu y u L9 ft U „ U U I U GC C f W G U c c u W 1 c U C f t B C A G H U-H U-H 6 6 P6b P6a P6 C-G C-G C-G C-G II II 11 I III C IL6b I ■ I P9.ia u-ft G G G R G U ft U n n t G ft t G ft L5c L9.ia G-C G-C glgP9.1G-C ft-U U—G o-c o-c fl-U G C G

U U U U ft G G U G U U G L9.2 C-G C-G ulft P9.2a U-ft U-ft gljj P9 2 U-ft C-G H-U G-U G-U ft ft u

Figure 4. The map of the Neurospora crassa mt genome. The figure is from Collins (1990).

26 27

URFL/M S-rRNA c o m

ATP6 A ATP8

exon L-rRNA

URFk intron URFj it intron ORF unassigned ORF

tRNA

^ promoter

i.2.3.^c. EcoRl fragments

URFN COI URFu ND4L

Figure 4. Figure 5. The comparison of the CYT-18 protein with other TyrRSs. B.s.: TyrRS of Bacillus stearothermophilus (Winter et al., 1983); E.c.: TyrRS of Escherichia coli (Barker et al., 1982); S.c.: TyrRS of Saccharomyces capensis (Hill and Tzagoloff, unpubl.); P.a.: TyrRS of Podospora anserina (KMmper et al., 1992); N.c.: TyrRS of Neurospora crassa (Akins and Lambowitz, 1987). The black boxes indicate the homologous regions among all of these TyrRSs. The striped boxes indicate the homologous regions of TyrRSs in Saccharomyces, Podospora, and Neurospora. The gray boxes indicate the homologous regions between Neurospora and Podospora. The figure was drawn and kindly provided by Dr. G. Mohr.

28 Comparison of TyrRS proteins

B.s. 1 34 83 161 207 217 308 399 416 419

E.c.

1 37 89 167 213 222 313 404 421424

S.c.

1 62 89 141 231 277 290 381 472489492

P. a.

240 286 309 399 442 4BB 607 624 541 586 640

N.c.

1 33 99 161 239 285 311 402 452 497 515 532 549 594 669

sequence Nucleotide binding fold tRNA binding

Figure 5. toVO CHAPTER II MATERIALS AMD METHODS

In this chapter, the major procedures used for the dissertation research are described. Other generally used procedures such as DNA preparations, cloning, and DNA treatments with some modifying enzymes will not be described here and can be found in "Molecular Cloning11 by Sambrook et al. (1989). The boiling lysis method was used for "minipreps" of plasmid DNA (pp 1.29; Sambrook et al., 1989). The alkaline lysis method was used for large-scale preparations of plasmid DNA (pp 1.33; Sambrook et al., 1989). The strategies and methods of cloning and transformation used in Chapters III, IV, and VI were as described (pp 1.53-pp 1.89; Sambrook et al., 1989). The treatment of DNA with nuclease Bal31 used in Chapters III and IV was described by Sambrook et al. (pp 5.73; 1989).

Strains of Neurospora and Growth Conditions The Neurospora crassa strains used in this study were wild-type 74-OR23-1A (designated 74A), and mutants cyt-4-1 (AEG-193a), cyt-19-1 pan-2 (GK-13a; Bertrand et al., 1982). Mycelia were grown in liquid culture containing Vogel's minimal medium plus 2% sucrose. Growth times and temperatures 30 31 were: wild-type 74A, 14 h, 25°C; cyt-4-1, 24 h, 25°C; cyt-19- 1, 24 h, 25°C (Akins and Lambowitz, 1987). Procedures for maintaining strains, preparing conidia, and growing conidia in liquid culture were as described (Davis and de Serres, 1970; Lambowitz et al. 1979).

Isolation of Mt RNA and RNP Particles Mitochondria were purified by the modified flotation gradient method (Lambowitz, 1979). RNA was extracted from mitochondria by the SDS-urea plus phenol-chloroform-isoamyl alcohol procedure (Garriga et al., 1984). RNP particles were isolated from mitochondria by centrifugation through 1.85 M sucrose containing 0.5 M KC1 buffer, as described (Garriga and Lambowitz, 1986).

Construction of the Plasmids Containing the Neurospora Mt Large rRNA intron Derivatives Used for the construction of the Mini-intron The plasmids containing the Neurospora mt large rRNA intron and its derivatives are summarized in Table 1. Recombinant plasmid pHXll contains the entire 2.3 kb mt large rRNA intron and flanking exon sequences cloned behind the bacteriophage T3 promoter in Bluescribe vector (pBS(+); Stratagene, La Jolla CA.; Akins and Lambowitz, 1987). Derivatives of pHXll lacking the ORF were originally constructed by Dr. R. A. Akins using Bal31 (Fast) nuclease (International , Inc., New Haven, CT) from the 32 BsmI site in the ORF. The resulting deletion plasmids were ligated with bacteriophage T4 DNA ligase and then sequenced by the dideoxy method (Sanger et al. , 1977) . One of the resulting plasmids, pHXHGl, has a deletion beginning at position 544 and ending at position 2060. This plasmid, linearized with EcoRI and transcribed with bacteriophage T3 RNA polymerase, gives an in vitro transcript having a 5' exon of 844 nts, an intron of 779 nts, which lacks the ORF, and a 3 ' exon of 525 nts. Plasmid pHXHG2 is the same as pHXHGl, except that a 779 bp Sall-Bglll fragment was deleted from the 5' exon of pHXHGl by digestion with Sail and Bglll, fill-in with the Klenow fragment of DNA polymerase I and ligation with T4 DNA ligase. pHXHG2 gives an in vitro transcript with a 5' exon of 65 nts. Recombinant plasmids pHX9422, pHX7, pHX46, and pHX57 were originally constructed by Dr. R. A. Akins. They were derived from pHXHG2 by site-directed mutagenesis (Kunkel, 1985; Kunkel et al., 1987) to remove all or part of the central hairpin (see Fig. 7) . The oligonucleotides used were: pHX94 22: 5'-ATTTAAGGATGGCATGCTGAGGCTTCACC-3' pHX7: 5'- ATTTAAGGATGGCACACCCTTAAATAGAA-3' pHX46: 5 r-ATTTAAGGATGGCATATTGAAGGACAAACA-3' pHX5 7: 5'-CCCAATATATTATCGAAGGACAAACATCCC-3' Plasmid pHX993 is identical to pHX9422, but lacks 18 bp of the polylinker between the Hindlll and Bglll sites in the 33 5' exon. To delete sequences in L8, pHX993 was cut at the remaining Hindlll and PstI sites in L8 and digested with Bal31 (Fast) nuclease (International Biotechnologies, Inc.) for times ranging from 15 sec to 3 min. The Bal31 treated plasmids were ligated with T4 DNA ligase in the presence of Ncol linker (CCCATGGG). Plasmids having inserts of the appropriate size were identified by gel electrophoresis, and deletion end points were determined by dideoxy sequencing (Sanger et al., 1977), using synthetic oligonucleotide primers complementary to sequences in the intron. The positions of some deletions are shown in Figure 9 of Chapter III. pBDl, pBD5, pBD6, pBD8, and pBD9 have the Ncol linker at the deletional junction. pBD2, pBD3, pBD4, and pBD7 do not contain the linker. pBD5A is the same as pBD5, except that the 5' exon is identical to that in pHX9422, with a length of 65 nts.

Plasmid pBD541 was derived from pBD5A by deleting P4 and P5 by site-directed mutagenesis (Kunkel, 1985; Kunkel et al., 1987), using the oligonucleotide: 5'-GCTAAGATAAATTTCGCTTAAATT AACCCAATATATTATC-3'. The deletion starts at position 303 and ends at position 334 (see Figure 10 in Chapter III). Starting with pBD5A, sequences in P6 were deleted by Bal31 digestion from the Nhel site (see Figure 10 in Chapter III). The plasmid was cut with Nhel, treated with Bal31 (Slow) nuclease (International Biotechnologies, Inc.) and ligated with T4 DNA ligase. The resulting plasmids were 34 sequenced by the dideoxy method (Sanger et al., 1977), using synthetic oligonucleotide primers complementary to sequences in the intron. Sequences deleted in the plasmids are shown in Figure 10 of Chapter III. Plasmid pBD591 is a derivative of pBD5A in which an

EcoRI site was inserted in P9 by changing G2266 and A 2267 to T and C, respectively (see Figure 11 of Chapter III). These changes were made by site-directed mutagenesis, using the oligonucleotide 5'-GATTTATTTCCAGAATTCACATAATGG-3'. Starting with pBD5A, sequences from the 3' exon and 3• end of the intron were deleted by Bal31 digestion. pBD5A was digested with Hindlll and Fspl. The 0.5 kb Hindlll-FspI fragment containing the intron and exons was purified on a 1.4% agarose gel and cloned into the Hindlll/HincII site of Bluescribe vector (pBS(+)) to produce plasmid pBD5A-Fsp. This plasmid was linearized with Xbal, which cuts in the polylinker, treated with Bal31 (Slow) nuclease (International Biotechnologies, Inc.) for 2 to 70 min and ligated in the presence of BamHI linker (5'-CCGGATCCGG-3'). The deletional junctions were sequenced by the dideoxy method (Sanger et al., 1977), using M13 forward primer (Bethesda Research Laboratories). The positions of some deletions are shown in Figure 11. All the constructs shown have the BamHI linker at the deletional junction. Plasmid pBD5A-IGS is related to pBD5A, but has the normal IGS sequence replaced with that of the yeast mt large 35 rRNA intron. This replacement was made by site-directed mutagenesis using the oligonucleotide: 5'-GAATATTGAGGGACAAGGG GGTAAATTATCCCTAGCGTAGCT-3'. Plasmid pBD5/UC is a derivative of pBD5A-IGS in which the P9.0 interaction was changed from GAA/U229SU to GAA/C2295U. This was accomplished by changing U2295 to C by site-directed mutagenesis using the oligonucleotide: 5'-GCCTGTTCGACTTGTTATC -3'.

Construction of the Plasmids Containing the Neurospora Mt Large rRNA Intron Derivatives Used for the Localization of the CYT-18 Protein Binding site on the Intron. Plasmids pHXll, pHX9422 and pBD5A contain different forms of the Neurospora mt large rRNA intron cloned behind the phage T3 promoter in pBS(+) vector (see above). All three plasmids were linearized with BanI for in vitro transcription. The transcript from pHXll consists of an 844 nt 5 ' exon, the full length 2296 nt intron, and a 50 nt 3' exon. The transcripts from pHX9422 and pBD5A consist of a 65 nt 5' exon, 583 or 388 nt-derivatives of the intron, respectively, and a 50 nt of 3' exon. The mutant introns used in this study are related to pBD5A. To construct plasmids having 5' truncations of intron sequences, plasmid pBD5, which is the same as pBD5A except for the polylinker region (see above), was linearized at the SphI site in the P2 stem and digested with Bal31 (Slow) nuclease at 30°C for 10 to 60 min. The Bal31-digested 36 plasmids were ligated with T4 DNA ligase in the presence of a HindiII-linker (CCAAGCTTGG). Plasmids containing both the Hindlll-linker and appropriate inserts were isolated, and the deletion end points were sequenced (Sanger et al., 1977), using the M13 reverse primer. Hindlll/EcoRI-fragments containing selected truncated introns were recloned between the Hindlll- and EcoRI- sites of pBS(-) to give plasmids pL95, pL97, pL107, and pL129, with the number in the plasmid name indicating the last nt of the truncation. These plasmids were linearized with BanI and transcribed using phage T3 RNA polymerase (Stratagene). Plasmids pBD5910, pBD591 and pBD5908 used to obtain RNAs having 3' truncations of intron sequences were described (Guo et al., 1991), but have been renamed pR342, pR359, and pR367 to indicate the last nt of the truncation. Plasmids with 3/ truncations upstream of L8 were constructed by Bal31- digestion of pBD5A linearized at the Ncol site at the end of the P8. The Bal31-treated plasmids were ligated with T4 DNA ligase in the presence of BamHI linker (CCGGATCCGG). Plasmids containing both the BamHI linker and inserts of appropriate size were isolated, and the deletion end points were sequenced (Sanger et al., 1977), using a primer complementary to the 3' exon (CCGCGCACTCATTTTGTACAC). The resulting plasmids are named pR239, pR243, pR244, and pR258. These plasmids were linearized with BamHI and transcribed using T3 RNA polymerase. RNAs R175 and R312 were obtained 37 from pBD5A linearized with Nhel or Ncol, respectively. Plasmid pBD541, which has a deletion of P4 and P5, was described (Guo et al., 1991), but has been renamed pA108-139 to indicate the extent of the deletion. pA258-323 was constructed by combining pR258 and pR244 at the BamHI sites in the linkers. Other plasmids having internal deletions of intron sequences were constructed from pBD5A by site-directed mutagenesis (Kunkel et al., 1987), using the following synthetic oligonucleotides: pA115-133: 5'-GATAAATTTCGCTTCAAATTATTTTTGAAATTAACCC-3' pAl49-233: 5'-CCTTATAGTCGTTGAACGATTATTTCGCTTCAAATT-3' pA155-224: 5'-GAACGATTTTATCTTNTAGATAAATTTCGCTTCAAATT-3' pA252-282: 5'-AAAAAGAAAGTTGGGTGTTTATAGTCGTTGAACG-3' Plasmids having internal deletions were linearized with BanI and transcribed using T3 RNA polymerase. pRL96-242 and pRL98-242 were constructed by combining 5' truncations pL95 and pL97, respectively, with 3' truncation pR243 via the Nhel site in P6b. Starting from these plasmids, pI75 and pI77 were constructed by deleting sequences distal to P6a by site-directed mutagenesis, using the oligonucleotide for pAl55-224. These plasmids were linearized with BamHI and transcribed using T3 RNA polymerase.

Construction of Other Plasmids. Plasmid pDJD173c was provided by Gerald F. Joyce (The Scripps Research Institute, La Jolla, CA) . This plasmid 38 contains the AP5abc deletion in the pT7L-21 plasmid which has the L-21 version of the Tetrahymena large rRNA intron cloned in the pUC18 vector (Been and Cech, 1988). Plasmid pT7TTlA3, which contains 48 nt of 5' exon, 414 nt of the wild-type Tetrahymena large rRNA intron, and 45 nt intron, was provided by T.R. Cech (University of Colorado, Boulder, CO; Zaug et al., 1986). I replaced the Bglll-SphI fragment in the insert of the plasmid pT7TTlA3 with the Bglll-SphI fragment in the insert of the plasmid pDJD173c to make the plasmid pT7TTlA3AP5abc, which was used for the study in Chapter VI. Plasmid pP5abcX was provided by Tan Inoue (Salk Institute, La Jolla, CA) . The plasmid pP5abcX contains the P5a, b, c stems of the Tetrahymena large rRNA intron cloned behind the phage T7 promoter of the pTZ18U vector (van der Horst et al., 1991). The plasmids pT7TTlA3AP5abc and pP5abcX were linearized with EcoRI and Xbal and transcribed with T7 RNA polymerase to make the AP5abc precursor RNA and P5abc RNA, respectively. Plasmid pTHIGHX, which was used for the study in Chapter VI, was constructed by PCR cloning a sequence of the Tetrahymena large rRNA intron from P3[5'] to J6/7 of the plasmid pDJD173c using the following oligonucleotides: ThighS: 5'-CGGAAGCTTAATACGACTCACTATAGGTCAAATTGCGGGAAAG-3' Th igh 3: 5'-CGGGATCCTGAACTGCATCCATATC-3' The Thigh5 oligonucleotide contains a phage T7 promoter and a Hindlll site at the 5' end. The Thigh3 oligonucleotide 39 contains a BamHI site at the 3' end. PCR product cut with both BamHI and Hindlll was isolated from a 2% agarose gel, and then cloned into the Hindlll/BamHI site of the pUC18 vector. Nonspecific control transcripts were synthesized from pHXll/PvuII (Neurospora mt large rRNA gene, 171 nt RNA containing sequences in the 5' exon) or pBTC20/KpnI (Neurospora cyt-20 gene, 869 nt RNA corresponding to anti­ sense strand; Kubelik et al., 1991), using T3 RNA polymerase.

Preparation of CYT-18 Protein from Neurospora and E. coli. The CYT-18 protein was obtained from mt RNPs of Neurospora wild-type 74A by heparin-Sepharose chromatography (Majumder et al., 1989) or synthesized in E. coli from the expression plasmid pEX560 (Kittle et al., 1991) . The E. coli- synthesized CYT-18 protein used for binding assays was extensively purified using a recently developed procedure involving polyethyleneimine-precipitation of nucleic acids, ammonium sulfate fractionation and chromatography in a CM- Sepharose column (Saldanha and Lambowitz, unpublished data). The final protein fractions were about 95% pure, with the major impurities being slightly smaller proteins that cross­ reacted with anti-CYT-18 antibodies.

In Vitro Synthesis of Neurospora Nt tRNATyr and other tRNAs

Plasmid pTYR, which was used to synthesize the Neurospora mt tRNATyr, was constructed by J. Gianelos in our 40 laboratory by PCR cloning the Neurospora mt tRNA1*' gene from plasmid pHP2, which contains the Pstl-5a fragment of wild- type 74A mtDNA in pBR322 (Heckman and RajBhandary, 1979). The primers for PCR were: 5'-AGCTAGCAATTAACCCTCACTAAA(A/G)GGA GGGTTCCGTTTGTTGG-3' and 5'-CGGTACCTGGTAGGAGAGAAAGGAA-3'. The 5' primer contains a terminal Nhel site for cloning and introduces a bacteriophage T3 promoter at the 5' end of the tRNA1*' gene. The 3f primer contains a BstNI site, which is used to linearize the plasmid for synthesis of an RNA ending in CCA, and a terminal Kpnl site for cloning. After PCR amplification, the 121 nt PCR product was purified by electrophoresis in a 2% agarose gel, digested with Nhel and Kpnl, and inserted into the Xbal/Kpnl site of pUC19 (Yanisch- Perron et al., 1985; Figure 6). To synthesize the Neurospora mt tRNATyr, pTYR was linearized with BstNI and transcribed using bacteriophage T3 RNA polymerase. Following transcription, the DNA template was digested with DNase I (Pharmacia LKB , Inc., Piscataway, NJ; 5 units, 5 min, 37°C) . After two cycles of extraction with phenol-CIA (phenol-chloroform-isoamyl alcohol, 25:24:1) and ethanol-precipitation, the RNA pellet was dissolved in 300 /il 10 mM Tris-HCl (pH 7.5), 1 mM EDTA, and the in vitro-synthesized tRNA was purified by DEAE- Sephacel chromatography at room temperature. For this purification step, about 1 mg of dissolved RNA was loaded on a DEAE-Sephacel column (0.7 x 4 cm column; 0.6 ml bed 41 volume), which had been equilibrated with 50 mM NaCl, 10 mM Tris-HCl (pH 7.5) , 1 mM EDTA. The column was washed with 250 mM NaCl, 10 mM Tris-HCl (pH 7.5), 1 mM EDTA, and the tRNA was eluted with 800 mM NaCl, 10 mM Tris-HCl (pH 7.5), 1 mM EDTA. The eluted tRNA was precipitated with ethanol and redissolved in 10 mM Tris-HCl (pH 7.5), 1 mM EDTA. The in vitro tRNATyr transcript is recognized efficiently by CYT-18 protein. In parallel experiments with the same CYT-18 protein preparation, the K,,, value for the in vitro- synthesized Neurospora mt tRNA1^ was 7 ± 3 jxM compared to 25 ± 6 /xM for purified E. coli tRNA1*', and the was about half that for the E. coli tRNA (Cherniack et al., 1990; Cherniack, 1991; Kittle et al., 1991). E. coli tRNA”*' and yeast tRNA”1' were transcribed from plasmids p67CF0 and p67YF0 (provided by Dr. Olke Uhlenbeck, U. Colorado). The plasmids were digested with BstNI and transcribed using T7 RNA polymerase (Sampson and Uhlenbeck, 1988). The resulting tRNA transcripts were purified as described above for the Neurospora mt tRNA7^ transcript.

Synthesis of In Vitro Transcripts Containing Group I Introns. In vitro transcripts were synthesized from recombinant plasmids linearized with appropriate restriction enzymes, using bacteriophage T3 or T7 RNA polymerase (Stratagene), essentially as described (Yisraeli and Melton, 1989). Following transcription, the DNA template was digested with 42 DNase I (Pharmacia LKB Biotechnology, Inc; 5 units, 5 min, 37°C), and the transcript was phenol-CIA-extracted and ethanol-precipitated twice. Standard transcription reactions were carried out in a volume of 200 Ml with 0.5 mM rNTPs for 60 min at 37 °C. To synthesize large amounts of RNAs for competition experiments, transcription was carried out in a volume of 400 Ml with 2 mM rNTPs. To synthesize high specific activity 32P-labeled transcripts for nitrocellulose filter binding assays, transcription was carried out in a volume of 50 nl with 0.4 mM ATP, GTP, CTP, 0.04 mM UTP and 150 /iCi a-[32P]-UTP (3,000 Ci/mmole; NEN-DuPont, Boston, MA) . To synthesize 32P-labeled transcripts for splicing assays, transcription was carried out in a volume of 50 m1 with 0.4 mM ATP, GTP, CTP and UTP plus 150 /xCi a-[32P]-UTP (3,000 Ci/mmole; NEN-DuPont).

Splicing Assays

The RNAs tested for splicing were transcribed from linearized plasmids described above, using bacteriophage T3 or T7 RNA polymerase. Splicing activity was isolated from mt RNP preparations from wild-type 74A by chromatography in heparin-Sepharose columns, as described by Majumder et al. (1989). Splicing reactions using unlabeled precursor RNA were carried out as described (Majumder et al., 1989) in 20 Ml reaction medium containing 0.5 nq of in vitro transcript substrate, 4 Ml CYT-18 protein preparation (0.8 M9 CYT-18 protein preparation from Neurospora 74A RNPs or 6 to 30 nM purified E. coli-synthesized CYT-18 protein), 100 mM KC1, 5 mM MgCl2, 20 mM Tris-HCl (pH 7.5), 5 mM DTT, and 40 tiCi a-

[32P]-GTP (3000 Ci/mmole; NEN-DuPont) . Reactions were initiated by addition of the CYT-18 protein, incubated at 37°C

for 10 to 60 min as indicated in the individual experiments, and stopped by addition of 180 jul of 10 mM Tris-HCl (pH 7.5) , 5 mM EDTA, 10 /xg total E.coli tRNA (Sigma Chemical Co.) and 200 til of phenol-CIA. The samples were extracted twice with phenol-ClA and precipitated with ethanol. Splicing assays using the 32P-labeled in vitro transcripts (approximately

20,000 cpm; 2ng) were the same except that the GTP concentration was 250 nM and reactions were incubated at 37°C for the period of time indicated. After phenol-CIA- extraction and ethanol-precipitation, samples were analyzed by electrophoresis in a 5% polyacrylamide gel (20:1 acrylamide:bisacrylamide) containing 8 M urea in IX TBE buffer (90 mM Tris, 90 mM boric acid, 2.5 mM EDTA) or in 1.5% agarose gels containing 0.1% SDS in 0.5X TBE buffer. The gels were dried and autoradiographed overnight on X-ray film. For competition study of the CYT-18 activity, the intron precursor RNA (pHX9422) was mixed with the competitor RNAs first. The reaction was initiated by addition of the CYT-18 protein. The reactions and analysis were the same as above. 44

TyrRS Assays TyrRS activity of the CYT-18 protein was assayed essentially as described (Akins and Lambowitz, 1987). For experiments involving large numbers of samples, TyrRS assays were carried out with CYT-18 protein in micrococcal nuclease- digested mt RNP preparations from wild-type 74A (Akins and Lambowitz, 1987). The use of these preparations circumvents problems due to instability of TyrRS activity associated with purified mitochondrial CYT-18 protein (Majumder et al., 1989) . TyrRS reactions were carried out in 20 /ul of reaction medium containing 4 ul CYT-18 protein preparation (28 /ug RNP protein) , 1 ng tRNA substrate, 100 mM KCl, 5 mM MgCl2, 20 mM Tris-HCl (pH 8.8), and 1 fzCi of L-[3,5-3H]-tyrosine (50 Ci/mmol; Amersham Co., Arlington Heights, IL) . tRNA substrates were E. coli tRNATyr (type 2; Subriden RNA, Rolling

Bay, WA) or Neurospora mt tRNATyr synthesized from pTYR (see

above). The reaction was initiated by addition of enzyme preparation, incubated at 30°C for 10 min, and terminated by addition of 0.5 ml of 10% TCA-0.1% SDS. TCA-precipitable radioactivity was collected on Whatman GF/C filter (Whatman LabSales, Hillsboro, OR) . The filters were washed nine times with 3 ml of 2.5% TCA-0.1% SDS and counted using Ready Protein Liquid Scintillation Cocktail (Beckman Instruments, Inc., Fullerton, CA). For kinetics study of the TyrRS activity of the CYT-18 protein, the tRNATyr was mixed with the competitor intron 45 first. The reaction was initiated by addition of CYT-18 protein from 74A mt RNP.

Nitrocellulose Filter Binding Assays Approx. 5 pM 32P-labeled RNA (18 fiCi/pmole) was mixed with different amounts of CYT-18 protein in 500 fil of TMK buffer (10 mM Tris-HCl (pH 7.5), 5 mM MgCl2, 100 mM KC1) containing 50 fig acetylated bovine serum albumin (New England Biolabs, Inc., Beverly, MA) plus non-specific competitors 40 fig poly(rC) (Pharmacia LKB Biotechnology, Inc.) and 10 fig of yeast total RNA (Sigma Chemical Co. , St. Louis, MO) . The samples were incubated for 2 min at 37°C, followed by 15 min at room temperature, then placed on ice to await filtration. Samples were filtered under gentle vacuum through a nitrocellulose filter (HAWP02500; Millipore Co., Bedford, MA), which had been pre-wetted with the TMK buffer plus 40 /xg/ml denatured salmon sperm DNA (Sigma Chemical Co.). After filtration, the filters were washed three times with 0.9 ml

TMK buffer plus 40 /Ltg/ml denatured salmon sperm DNA and then dried at 80°C. Radioactivity retained on the filter was measured by liquid scintillation counting, using Ready Protein Liquid Scintillation Cocktail (Beckman Instruments, Inc.). Binding data were analyzed as described by Yarus and Berg (1967), with the proportion of active CYT-18 protein (approx. 60%) determined by the method of Riggs et al. (1970). The binding of the CYT-18 protein to the intron 46 containing transcript was saturated after 1 min at 37°C and was stable for at least 20 min. In eleven nitrocellulose filter binding experiments, the average Kd for the pBD5A/BanI transcript containing the Neurospora mt large rRNA intron was 6.2 nM, the range of the Kds was 3 to 11 nM, and the standard deviation was 2.1 nM.

UV-crosslinking For UV crosslinking experiments, 32P-labeled in vitro transcript was synthesized using T3 RNA polymerase in 50 jul of standard reaction medium containing 1 mM ATP, CTP and GTP, 0.08 mM 5' bromodeoxyuridine and 4 fiCi 32P-UTP (3,000 Ci/mmole, NEN-DuPont). The 32P-labeled in vitro transcript (40,000 to 50,000 cpm) was incubated with CYT-18 protein under the standard splicing conditions for ten min at 37°C. The reaction mixtures were then placed in a plastic petri dish on a Fotodyne (Model 3-3000) transilluminator and irradiated with 300 nm UV light for six min, unless otherwise indicated. After irradiation, unprotected RNA was digested with 2 mg/ml of RNase A at 30°C for 30 min. Proteins were analyzed by SDS-polyacrylamide gel electrophoresis, using the gel system of Laemlli with a 10% separating gel and a 3% stacking gel. Gels were run in a Biorad mini-gel apparatus at 200 V for 30 min. The proteins on the gel were transferred to a nitrocellulose filter using an electroblot apparatus and analyzed by immunoblotting with the CYT-18 47 antibody preparation C18-2 and autoradiography to detect 32P- labeled RNA fragments that had been crosslinked to the protein. Table 1. Plasmids containing Neurospora mt large rRNA intron and its derivatives. The construction of these plasmids is described in the text of this chapter. "Enzymes indicate the restriction enzymes used to linearize the plasmids for transcription. All of the plasmids were transcribed with phage T3 RNA polymerase. b(+linker) indicates that half of the BamHI linker at the 3' end of the linearized plasmids was transcribed and attached to the 3' end of the transcript.

48 Table l

Plasmid Vector Enzyme" in vitro transcript (5’E-I-3’E) pH X ll pBS(+) EcoRI 844-2296-525 nt pH X llG l pBS(+) EcoRI 844-779-525 nt pHXUG2 pBS(+) EcoRI 65-779-525 nt pHX9422 pBS(+) BaNI 65-583-50 nt pHX7 pBS(+) BanI 65-551-50 nt pHX46 pBS(+) BanI 65-536-50 nt pHX57 pBS(+) BanI 65-509-50 nt pHX993 pBS(+) BanI 47-583-50 nt pBDl pBS(+) BanI 47-347-50 nt pBD2 pBS(+) BanI 47-373-50 nt pBD3 pBS(+) BanI 47-415-50 nt PBD4 pBS(+) BanI 47-430-50 nt pBD5 pBS(+) BanI 47-388-50 nt pBD6 pBS(+) BanI 47-413-50 nt pBD7 pBS(+) BanI 47-433-50 nt pBD8 pBS(+) BanI 47-445-50 nt PBD9 pBS(+) BanI 47-478-50 nt pBDSA pBS(+) BanI 65-388-50 nt pBD5A-IGS pBS(+) BanI pBD5A: IGS replaced with that of yeast a* intron. pBD5/UC pBS(+) BanI pBD5A-IGS: GAA/U^jU of P9.0 pairing changed to G A A /C^U . pL95 pBS(-) BanI pBD5A: 5’ truncation to J2/3 (nt 95). pL97 pBS(-) BanI pBD5A: 5’ truncation to P3 (nt 97). pL107 pBS(-) BanI pBD5A: 5’ truncation to P4 (nt 107). Table l. (continued)

pL129 pBS(-) BanI pBD5A: 5’ truncation to P5 (nt 129). pBD5901 pBS(+) BamHI 65-388-15 nt pBD5902 pBS(+) BamHI 65-388-11 nt pBD5903 pBS(+) BamHI 65-388-9 nt pBD5904 PBS(+) BamHI 65-388(+linker)b pBD5905 pBS(+) BamHI 65-382(+linker) pBD5906 pBS(+) BamHI 65-377(+linker) pBD5907 pBS(-t-) BamHI 65-372(+linker) pBD5908 pBS(-t-) BamHI 65-366(+linker) pBD5909 pBS(+) BamHI 65-246(+linker) pBD5910 pBS(+) BamHI 65-241(+linker) pR367 pBS(+) BamHI pBD5A: 3’ truncation to P9.1 (nt 367). pR359 pBS(+) BamHI pBD5A: 3’ truncation to P9 (nt 359). pR342 pBS(+) BamHI pBD5A: 3’ truncation to J7/9 (nt 342). pR312 pBS(+) BamHI pBDSA: 3’ truncation to L8 (nt 312). pR243 PBS(+) BamHI pBD5A: 3’ truncation to J6/7 (nt 243). pR239 pBS(+) BamHI pBD5A: 3’ truncation to P6 (nt 239). pR175 pBS(+) BamHI pBDSA: 3’ truncation to P6b (nt 175). pA108-139 pBS(+) BanI pBD5A: AP4-P5. pAl 15-133 pBS(+) BanI pBD5A: AP5. pA 149-233 pBS(+) BanI pBDSA: AP6a-P6b. pA155-224 pBS<+) BanI pBD5A: AL6a-P6b. pA252-282 pBS(+) BanI pBDSA: AP7.1-P7.2. pA258-323 pBS(+) BanI pBDSA: AP8. pA258-323 pBS(+) BamHI pBD5A: 3’ truncation to P7.1 (nt 258). pI75 pBS(-) BamHI RNA containing 75 nt of P4-P5-P6 with sequences distal to P6a deleted. Figure 6. The map of the plasmid pTYR. The Neurospora mt tyrosyl-tRNA gene is linked to the bacteriophage T3 promoter and cloned into the Xbal/Kpnl site of pUC19 vector. The plasmid was constructed by J. Gianelos.

51 52

Xbal BstNI Kpnl Nc m t tRNAty, geneI T3 Promoter Nc mt tRNAty, geneI pTYR ICTAGCAATTAACCCTCACTAAAK G g 5C «3G T A d

Sapl lacZ'

Afil EcoO109I Ori pUC19 Aatll

AlwnI SspI

Amp Seal Bsal

Figure 6 . CHAPTER III STRUCTURAL ANALYSIS OF THE Neurospora MITOCHONDRIAL LARGE rRNA INTRON AND CONSTRUCTION 07 A MINI-INTRON THAT SHOWS PROTEIN-DEPENDENT SPLICING

Introduction

The Neurospora mitochondrial (mt) large rRNA intron is an example of a group I intron that is not self-splicing in vitro. This intron has a length of 2.3 kb and contains an open reading frame (ORF) of 426 amino acids that encodes the mt small subunit ribosomal protein S-5 (Burke and RajBhandary, 1982; Lambowitz et al., 1985). Experiments using a protein-dependent in vitro splicing system showed that the Neurospora mt large rRNA intron is spliced by the same guanosine-initiated transesterification reactions used by self-splicing group I introns, but in this case, splicing is dependent upon proteins, which are presumably required for correct folding of the precursor RNA (Garriga and Lambowitz, 1986). Genetic and biochemical studies have shown that a key protein required for splicing the mt large rRNA intron is the mt tyrosyl-tRNA synthetase (mt tyrRS), which is encoded by nuclear gene cyt-18 of Neurospora (Akins and Lambowitz,

53 54 1987). Although additional proteins, such as CYT-4 and CYT- 19 proteins, may be required for efficient splicing in vivo, highly purified preparations of the CYT-18 protein isolated from Neurospora mitochondria or synthesized via expression plasmids in E . coli are by themselves sufficient to splice the mt large rRNA intron under in vitro conditions (Majumder et al., 1989; Kittle et al., 1991). An experimental advantage of the Neurospora mt system is that unspliced 35S pre-rRNA accumulates in the mitochondria of cyt-18 mutants and can be isolated in mt ribonucleoprotein (RNP) particles. These RNP particles contain large subunit ribosomal proteins and other components required for splicing (LaPolla and Lambowitz, 1979; Majumder et al., 1989). The large amounts of unspliced 35S pre-rRNA that accumulate in the cyt-18 mutants made it possible to analyze its structure by direct experimental methods (Grimm et al., 1981; Wollenzien et al., 1983). When 35S pre-rRNA was psoralen- crosslinked in RNPs, the predominant crosslinked structure detected by electron microscopy was a 2.2 kb intron loop. This intron loop appears to reflect a base pairing interaction between sequences at or near opposite splice sites, very possibly the interaction between the IGS and the 5' and 3' exons. By contrast, the predominant structure in deproteinized RNA was a relatively large (100 bp) hairpin structure located in the center of the molecule, near the 5 ' splice site. The psoralen-crosslinking experiments indicated 55 that the intron loop and "central hairpin" structure are alternate conformations of 35S pre-rRNA and that proteins preferentially stabilize the intron loop conformation (Wollenzien et al., 1983). In the work presented in this chapter, I analyzed the structural requirements for splicing the mt large rRNA intron by in vitro mutagenesis, using the protein-dependent in vitro splicing system developed for this intron. I constructed a mini-intron of 388 nts that retains the IGS and core structure, but lacks regions that are not essential for splicing, including the ORF, all of L8 and most of the central hairpin. I found that this mini-intron is spliced in a protein-dependent manner, but still cannot self-splice under the conditions tested. Strengthening the PI, P10 or P9.0 pairings did not enable the mini-intron to self-splice. My results suggest that the inability of the Neurospora mt large rRNA intron to self-splice reflects the deficiency of a structure or activity required for the 5' splice site cleavage reaction, either in the intron core itself or in the interaction between the core and the PI stem.

Results

Secondary structure Model of the Mt Large rRNA Intron Figure 7 shows a hypothetical secondary structure of the Neurospora mt large rRNA intron. This structure is typical of the subclass of group I introns, designated group IA, 56 which are distinguished from other group I introns by several idiosyncracies in the core structure, including an extra stem-loop or pair of stem-loops (P7.1 and P7.2) between sequence elements R and P3[3'] (Cech, 1988). Features that may affect the ability of the Neurospora mt large rRNA intron to self-splice include the unusual location of the 5' splice site near the top of the PI stem, rather than within the stem as in most group I introns (cf., Waring and Davies, 1984; Doudna et al., 1989) and the long, stable hairpin structure (P2 stem; 272 nt), immediately downstream from the 5' splice site. The ORF encoding mt ribosomal protein S-5 is contained

within loop L-8 . The Neurospora mt large rRNA intron also contains PstI- and HindiII-palindromes, repetitive sequence elements in Neurospora mtDNA (Burke and RajBhandary, 1982). The left most PstI- and Hindlll-palindromes are located in extensions of the P2 stem, with the PstI-palindrome forming

the apex of the stem (see Figure 8C). The right most Pstl-

and Hindlll-palindromes are located in L8 , downstream of the ORF (see Figure 9, right). Base pairing between opposite PstI- and Hindlll-palindromes could also, in principle, impede the ability of the intron to self-splice.

Deletion of the ORF To define minimum sequences required for splicing, we constructed mutant introns containing relatively large deletions. First, starting with pHXll, which contains the full length intron, the ORF encoding mt ribosomal protein S-5 57 was deleted by Bal31 digestion from the Bsml site in the ORF. The resulting plasmid, pHXHGl, gives an in vitro transcript containing an intron of 779 nt. Splicing of introns from in vitro transcript substrates was assayed by addition of 32P-GTP to the 5' end of the intron excised during splicing. After gel electrophoresis and autoradiography, splicing reactions show a 32P-labeled band corresponding to the excised intron and a slightly larger band corresponding to the 1-3'E intermediate, resulting from

G-dependent cleavage at the 5' splice site. Figure 8A shows that the intron from pHXHGl is spliced by preparations of the CYT-18 protein, indicating that the ORF sequence is not essential for splicing. Comparison of the amount of excised intron obtained with pHXll and pHXHGl shows that the smaller intron with the ORF deleted is spliced more efficiently than the full length intron, presumably reflecting in part its smaller size.

Deletions in the P2 Stem The large, central hairpin structure, which had been identified in 35S pre-rRNA, corresponds to an extended P2 stem, located near the 5' splice site (Figure 8C). Following the rules of nomenclature for group I introns, extensions at the P2 stem are divided into segments designated P2a, P2b, etc. (Figure 8C) . Starting with plasmid pHXHG2, which contains the ORF deletion and shorter 5' exon, different 58 segments of the central hairpin were deleted by site-directed mutagenesis. As shown in Figure 8B, most of the central hairpin (228 of 272 nt in pHX7) could be deleted without impairing splicing by the CYT-18 protein. However, complete deletion of the P2 stem (pHX57) or deletions that remove one side of the bottom of the stem (pHX46) inactivate splicing. These findings are similar to those for the Tetrahymena nuclear rRNA intron where deletion of the P2 stem inhibits self-splicing in vitro (Price et al., 1985). The smaller introns generated by progressive deletions in the central hairpin are spliced more efficiently than the larger introns.

Deletions in L8 Starting from construct pHX993, from which the ORF and the extensions of the P2 stem have been deleted, I deleted additional sequences in L8 by Bal31 digestion. As shown in

Figure 9, all of L8 and the last base pair of P8 could be deleted without impairing splicing (deletion A5). Further deletions, A2 and A4, which extend into P8 , impair splicing, but do not inactivate it entirely. Al and A3, which extend into P7, are totally inactive. The A5 deletion removes the remaining PstI- and Hindlll-palindromes, indicating these are not required for splicing in vitro.

Deletion of P4 and P6 and Nucleotide Substitutions in P9 Figure 10 shows analysis of constructs from which P4/P5 or P6 have been deleted or which have nucleotide substitutions in P9. Plasmid pBD541 was derived from pBD5A by site-directed mutagenesis to delete P4/P5 (nucleotides 303-334). The resulting intron is inactive in protein- dependent splicing. Plasmids pBD561 to pBD564 have deletions in P 6 made by Bal31 digestion from the Nhel site. Deletions

Al (pBD561) and A2 (pBD562), which affect only P6b, have no effect on protein-dependent splicing. A larger deletion A3, which extends one nucleotide into P6a, impairs protein- dependent splicing, and a deletion A4, which extends into P6 and P4, inhibits protein-dependent splicing completely. Plasmid pBD59l which contains two nucleotide substitutions, introducing an EcoRI-site at the base of P9 (G2266 and A2267 replaced by T and C, respectively), gives an intron that is fully active in splicing. The two nucleotide substitutions disrupt base pairs in P9, but it is conceivable that the stem could still form, anchored by the GC base pairs at its base (Figure 11 bottom).

Truncations at the 3' End of the Intron Starting with plasmid pBDSA, sequences from the 3' exon and 3' end of the intron were progressively deleted by Bal31 digestion from the FspI site to generate a series of plasmids numbered sequentially pBD5901 to pBD5910. These plasmids were linearized, using a BamHI-site introduced at the end of the deleted sequence (see Chapter II), and transcribed with T3 RNA polymerase to produce a series of in vitro transcripts 60 lacking different segments of the intron and 3' exon (Figure 11). In addition, pBD591 (see above) was used to obtain an in vitro transcript truncated at the EcoRI site introduced into P9. Figure 11 shows that truncated introns lacking the 3 ' splice site and P9.1 are still active, as judged by ability to splice or carry out the 5' cleavage reaction.

The smallest active intron, A 8 , comes within 4 nts of the base of P9. However, in pBD591/EcoRI, deletion of an additional 8 nts, which truncates the intron in the P9 stem, abolishes the 5' splice site cleavage activity, indicating that this part of P9 is critically required. We note that the A4 intron gives a band corresponding to 1-3'E due to the BamHI-linker at the 3' end of the transcript, and the A5 intron gives a band which is slightly smaller than the intron and may reflect activation of a cryptic splice site at

The results for the truncations show that the 3' splice site and intron sequences downstream of P9 are not required for functional binding of the CYT-18 protein.

Characterization of In Vitro Splicing of the Mini-intron The preceeding experiments led to construction of a mini-intron of 388 nts that still shows protein-dependent splicing. The mini-intron lacks the ORF and incorporates both the A5 deletion, which removes all of L8 (see Figure 9), and the pHX9422 deletion, which removes most of the extended

P2 stem (see Figure 8C). The intron was tested for ability 61 to self-splice under a variety of different conditions, including systematic variation of KC1, Mg+2, GTP, guanosine and temperature. As shown in Figures 12 and 13, the mini- intron was not self-splicing under any condition tested. Protein-dependent splicing was optimal at 75 to 100 mM KC1,

5 mM Mg+2, and 45°C. The KC1 and Mg+2 optima are essentially the same as those for the full length intron, whereas the temperature optimum is somewhat higher (45°C compared to 37°C to 40°C) and splicing occurs at a lower temperature than for the full length intron (25°C compared to 30°C; cf. , Garriga and Lambowitz, 1986). As reported previously for the full length intron, increasing the Mg+2 concentration above 5 mM resulted in a progressive decrease in the ratio of excised intron to the partial product 1-3'E, resulting from G- dependent cleavage at the 5' splice site (cf., Garriga and Lambowitz, 1986). The mini-intron remained active in protein-dependent splicing when Mn+2 was substituted for Mg+2, up to M n +2 concentrations of 20 mM (not shown), or when (NH4)2

S04 was substituted for KC1, up to (NH4)2S04 concentrations of

50 mM (not shown). However, neither substitution of Mn+2 for

Mg+2 nor (NH4)2S04 for KCl enabled the intron to self-splice.

As shown in Figure 13, experiments using uniformly labeled transcript showed that increasing the concentration of GTP or guanosine to 800 pH and 400 nVL, respectively, well above their Km's in the protein-dependent reaction, did not enable the mini-intron to self-splice, as judged by lack of products 62 corresponding in size to the excised intron or ligated exons. Other experiments showed that further increase of the GTP concentration to 10 mM had a similar lack of effect (not shown).

Modification of the IGS Sequence and P9.0 Pairing The experiments above show that wholesale reduction in size of the intron, deletion of the ORF, or reduction of the large hairpin near the 5' splice site do not enable the Neurospora mt large rRNA intron to self-splice. As indicated previously, the Neurospora mt large rRNA intron is unusual in that the conserved UG base pair, which defines the 5' splice site, is located at the top of the PI stem rather than within the stem as in most group I introns (Waring and Davies, 1984; Doudna et al., 1989). To test whether an appropriate modification of the PI pairing would enable the intron to self-splice, we changed the IGS sequence so that it is the same as that in the yeast mt large rRNA intron, which is inserted at the same location in the mt large rRNA gene and is self-splicing in vitro. As shown in Figure 14A, the modified PI interaction is identical to that for the yeast intron, with the conserved UG base pair now located two nucleotides from the top of the stem. At the same time, the P10 interaction is also strengthened relative to the Neurospora intron, but differs from the yeast P10 interaction by one GO base pair, reflecting a difference in the exon sequence. As shown in Figure 14, the construct containing 63 the yeast IGS was still not self-splicing under any conditions examined, including systematic variation in temperature (Figure 14A), MgCl2 and KC1 concentrations

(Figures 14B, C) or (NH4)2S04 concentration (not shown) . The modified intron remains active in protein-dependent splicing, but has somewhat different temperature, Mg+2 and KC1 optima (Figure 14). Interestingly, the change of the IGS increases the ratio of excised intron to 1-3'E intermediate, presumably reflecting increased efficiency of the second step of the splicing reaction as a result of the strengthened P10 pairing. The P9.0 pairing involves base pairing between the two nucleotides preceding the intron's 3' terminal G-residue and nucleotides in J7/9 of the intron core (see Figure 14; Burke, 1989; Burke et al., 1990; Michel et al., 1989; 1990). The P9.0 interaction, along with P10, plays a role in binding the 3' splice site to the intron and brings the conserved G- residue at the 3' end of the intron into proximity with the guanosine binding site, which has been shown to include the conserved GC base pair adjacent to the puckered base in P7 (Michel et al., 1989). Recently, Michel et al. (1990) noted that P9.0 in the Tetrahymena intron and other group IB introns involves base pairing of the 1st and 2nd nucleotides of J7/9 with the 2nd and 3rd nucleotides upstream of the 3' splice site, whereas in group IA introns, the potential pairing generally involves the 2nd and 3rd nucleotides of 64 J7/9 (see Figure 15, bottom) . In the Neurospora intron, which belongs to group IA, the P9.0 can be drawn in either way, beginning at either the first or second nucleotide of J7/9 (pBD5; Figure 15) . Starting with the modified IGS- construct, I changed the P9.0 pairing so that it should be identical to that of the self-splicing Tetrahymena intron, which is a group IB intron. As shown by the assays in Figure 15, the resulting modified intron (pBD5/UC) is still spliced in a protein-dependent manner, but is not self-splicing under any conditions examined. The products of the protein- dependent splicing reaction show an increased ratio of partial product 1-3'E to full length intron, suggesting the second step of the splicing reaction has been inhibited, presumably as a result of forcing the intron toward the group IB configuation of P9.0 at the expense of the normal group IA configuration. This result implies that the three- dimensional architecture of this region is somewhat different between group IA and group IB introns, so that only the appropriate P9.0 pairing can be accomodated with optimal efficiency.

Comparison of the Splicing and TyrRS Activity of the CYT-18 Protein Isolated from 74A, cyt-4-1, and cyt-19-1 by Heparin Sepharose Chromatography The Neurospora mt large rRNA intron is deficient in splicing in cyt-4-1 and cyt-19-1 mutants. To understand the function of the cyt-4 and cyt-19 genes, I examined whether the CYT-18 protein isolated from these mutants is deficient in splicing or TyrRS activity. The CYT-18 protein was isolated from mt RNPs of 74A, cyt-4-1 (AEG-193a) and cyt-19-l strains by heparin-Sepharose (HS) columns as described by Majumder et al. (1989). The amount of the CYT-18 proteins purified through the HS column was measured by Western blotting as shown in Figure 16B. The intensities of the CYT- 18 protein bands on the Western blot filter was determined by scanning the bands with a densitometer. After the calibration, the volume of the fraction of the CYT-18 preparation, which gave approximately the same intensity of the immunostaining, were determined. Approximately the same amount of the CYT-18 proteins from the HS column fractions were used for the comparison of splicing activity (Figure 16A). The results are summarized in Table 2. The TyrRS and splicing activities for the HS column-purified CYT-18 protein from cyt-4-1 mutant are close to those from wild-type 74A. The TyrRS activity of the CYT-18 protein from cyt-19-l mutant is about the same as that of 74A, but its splicing activity is slightly higher than the CYT-18 protein from 74A, with a ratio of 1:1.8 (Table 2).

Discussion

I have constructed mutant derivatives of the Neurospora mt large rRNA intron in order to identify structural features that are required for splicing or impede its ability to self- 66 splice. A mini-intron of 388 nt, which lacks the long ORF, all of L8 and most of the central hairpin (P2 stem) , but which retains the core structure and IGS, was spliced in the same protein-dependent manner as the full length intron. The complete deletion of P2 or further deletions that affect the core structure, such as deletions in P4, P6 , P7 or P9, inactivate splicing. The latter results indicate that splicing of the mt large rRNA intron requires the conserved group I intron core structure that presumably retains catalytic activities required for splicing. Group I introns have been found at the same location as the Neurospora mt large rRNA intron in other fungi, including Saccharomyces, Kluyveromyces, Aspergillus, and Podospora. The Aspergillus and Podospora introns are clearly cognates of the Neurospora intron. Burke (1983) found that most of the size difference between the Neurospora mt large rRNA intron and the smaller Aspergillus mt large rRNA intron (1678 nt) could be accounted for by four relatively GC-rich insertions in the Neurospora intron. Insertion I (nucleotides 7 to 267) includes Pstl-L and Hindlll-L and corresponds to most of the extended central hairpin structure. Insertion II

(nucleotides 376 to 414) is in P6b, insertion III (nucleotides 1559 to 1783) is in the ORF encoding S-5, and

insertion IV (nucleotides 2074 to 2203) is in L8 and includes Pstl-R and Hindlll-R. We find that these extra sequences can be deleted without impairing splicing of the intron in vitro. 67 However, in vivo, they may be adapted to serve functions in RNA splicing or other processes, e.g., by providing protein binding sites or by helping to fold the intron, bringing opposite splice sites into proximity. Since some of the insertions include repetitive sequence elements of Neurospora mtDNA, it is likely they were acquired by the Neurospora intron after the divergence of Neurospora and Aspergillus. Like the Neurospora intron, the Aspergillus and Podospora

introns contain ORFs in L8 that are homologous to the S-5 ribosomal protein (Netzker et al., 1982; Burke, 1983; Cummings et al., 1989). Interestingly, the Podospora intron contains an additional downstream ORF of 111 amino acids, which is homologous to the ORF encoded in the Neurospora

ATPase 6 intron (Cummings et al., 1989). Comparison of the Neurospora mt large rRNA intron with the yeast mt large rRNA intron shows that sequences in P3,

P6 , J6/7, P7, P7.1, P7.2, and between P9.1 and the 3' splice site are very similar between the two introns (74 out of 96 nucleotides identical), whereas other regions, including P2,

P5 and P8 , have little similarity. Although the Neurospora and yeast mt large rRNA introns have not been considered cognates (Dujon, 1989) , in our opinion, it is not possible to distinguish whether the two introns resulted from a single insertion that predated the divergence of Neurospora and yeast or whether they are two different introns that inserted at the same location in different . Both introns 68 contain a long ORF in L8 . However, the ORF in the Neurospora intron encodes the mt ribosomal protein S-5, whereas the ORF in the yeast intron encodes a site-specific endonuclease that promotes transfer of the intron to intronless alleles of the gene (see Lambowitz, 1989). Since the endonuclease is presumably required for insertion of the intron, the most likely hypothesis is that it was originally present in both introns and the ORF encoding S-5 is a more recent acquisition. It remains possible, however, that S-5 was originally an intron-encoded protein, perhaps even an endonuclease, that became adapted to function as a ribosomal protein. Mota and Collins (1988) have previously found that the Neurospora ND1 intron has two different ORF#s in different strains of the same species. After deletion of extraneous sequences, the functional core of the Neurospora mt large rRNA intron is similar in size and structure to that of other group I introns. The finding that the mini-intron shows the same protein-dependent splicing as the full length intron suggests that the CYT-18 protein binds to conserved sequences or structures in the intron, rather than any of the optional insertions. A binding site that includes conserved sequences or structures in group I introns is expected from the finding that the CYT- 18 protein is required for splicing a number of different group I introns in vivo (Collins and Lambowitz, 1985; Lambowitz et al., 1985). 69 Although splicing of the Neurospora mt large rRNA intron is protein-assisted, mutations in the intron core have essentially the same effect as in self-splicing group I introns. Thus, in the Neurospora intron, as in the

Tetrahymena intron, deletions affecting P2, P4/5, P6 or P9 completely inhibit splicing or 5' splice site cleavage activity, whereas deletions in L8 , peripheral to P6 , or downstream of P9 have little if any effect on activity of the intron (Burke 1988; Cech, 1988, 1990). Likewise, changes in the IGS and P9.0 interactions affect the second step of the splicing reaction in a manner that is readily accounted for based on previous results with self-splicing group I introns (Michel et al., 1989, 1990; Burke et al., 1990). As discussed above, our results support the previously noted difference in P9.0 pairings between group IA and group IB introns (Michel et al., 1990) by showing that changing the pairing to favor the group IB configuration decreases the efficiency of the second step in splicing (Figure 15). This implies that a critical region of the intron core, which includes the guanosine binding site, has a somewhat different structure in group IA and group IB introns. The lack of self-splicing of the rat large rRNA intron indicates either that the core structure of the intron is deficient in a catalytic activity that must be provided by a protein or that some feature of the intron structure must be stabilized by a protein. At a minimum, this activity or 70 structure is required for the 5' splice site cleavage reaction, since this does not occur under in vitro conditions in the absence of proteins. Our results show that lack of self-splicing of the Neurospora mt large rRNA intron is not due to the large size of the intron, nor the extended P2 stem, which can be greatly reduced without enabling the intron to self-splice. Further, the lack of self-splicing does not appear to be due to unusual features or instability of the PI, P10 or P9.0 pairings, which can be made similar to those in self-splicing introns without enabling the intron to self-splice. The remaining candidates include a presently unknown structure in the core that is required to activate the 5' splice site, the binding site for the guanosine cofactor, which could in principle be located in the intron or the protein, or the interaction between the core and the PI stem. Figure 7. Hypothetical secondary structure of the Neurospora mt large rRNA intron. The intron has a length of 2296 nts with an ORF of 1278 nts, which encodes mt ribosomal protein S-5. The sequence of the intron was determined by Burke and RajBhandary (1982), but has three corrections based on our sequencing: has been added; C350 (previously C349) has been changed to U; and A225l (previously A2250) has been deleted. The latter change affects the J7/9 region shown in Figure 15. Structural conventions are based on Burke et al. (1987). The core structure is typical of group IA introns.

71 72

p-231nt— |

'-c C' u A 320 G u G c U A A G c A u A u u u U A 0 A U A- A U A P5C C 20— A u A U U A 310— U A c A U A \ u A A A— 330 A 354 u u A t 5’SS— G„ U A I 1 A A A U 376 A U 1 A 1PZ P a u O C 390 Plc : u ' c A ti­ 280 N cc G A GU S c 1 A ll P3 u A PS AA P6j a G PSb A A C U C U 5'-AAAAGCUACGCUA |----- C GAI UUUAUCUU GUGGGUCCUAGC CAC UAGC CGCUA G AACCCA OGCU-AAAAUACAA CAUCCGUCAUCG-CUG----AUCC GUGAU C 498 C C U ! A U i i C AAU-36nt-GUA C 430 A A 410 A00 J U r3 R A ! 2296 ’UAU uuuuuucuuu IAUCAGCA 420 Ii AUA AAAAAAGAAAUGAAGAA- r*'1AUAGUC-UCAAC CA---AA UGAUAACAACUUdAACACGCUAAUUU-3' ; u P8 s P7 G ORF(127Bnt) r I r — 2221 A U Spw I 3'SS I A P9 A A A G G A--2280 UAA-391nl-OA A-- C U u c A U; p7.i A P7.2 U C A A A U G A C I U U U— 454 2262 C C A

470

Figure 7. Figure 8. Effect of deleting the ORF and all or part of the central hairpin on in vitro splicing of the mt large rRNA intron. The CYT-18 protein was isolated from mt RNP preparations of wild-type 74 A by heparin-Sepharose chromatography, as described by Majumder et al. (1989).

Splicing was assayed by addition of 32P-GTP to the 5' end of the excised intron under the standard splicing conditions at 37°C for 60 min. (A) Deletion of ORF. Starting from plasmid pHXll, which contains the full length intron, the ORF and adjoining sequences in L8 were deleted by Bal31 digestion, as described in Chapter II. pHXHGl has a 1517 nt deletion including the ORF. pHXll and pHXHGl were linearized with FcoRI and transcribed with T3 RNA polymerase. The samples were analyzed by electrophoresis in a 1.5% agarose gel. Lane 1: pHXHGl; Lane 2: pHXll. Molecular weight markers (M) are

Hindlll/EcoRI fragments of bacteriophage X DNA. (B) Deletion of the central hairpin. Starting from plasmid pHXHG2, derivatives of the intron lacking all or part of the central hairpin (P2 stem) were constructed by oligonucleotide- directed mutagenesis, as described in Chapter II. These plasmids were linearized with Banl and transcribed with bacteriophage T3 RNA polymerase. The samples were analyzed by electrophoresis in a 5% polyacrylamide/8 M urea gel. Lane

73 74 1: pHXHG2; Lane 2: pHX9422; Lane 3: pHX7; Lane 4: pHX46;

Lane 5: pHX57; Lane 6 : pHX9422. (C) shows the nucleotides sequence and predicted secondary structure of the central hairpin (P2 stem). The boundaries of the deletions in pHX7, pHX9422, pHX46, and pHX57 are indicated. PstI- and Hindlll- palindromes (PstIL and HindlllL) are highlighted. Short, solid arrows near the base of the hairpin indicate locations of RNase III sites (Guo et al., 1991). 75

O X X Xa. X a.

2822 nt 1-3 E 2297 nt I

1305 nt 1-3'E

780 nt I

i

ORF pHX11 H ------L . n ■ ■ 3665 nt

PHX11G1 H ■■ 2148 nt

Figure 8. 76 Figure 8 (continued)

CM (M CSI •'■I 1 1 1 5 \»-U' O SI c-o c-a B X S N S!l)r X a tQ. l Q. l Q. P Q. Q, c-a c-u c-n c -a Hi

ucck'U m

/IN 5-c P2« , fc uBBBBu-i>'U '«— 780 nt Llc uggg \ ± / COI-A <— 552 nt U*»wJJ0 uu‘* •» ifk PI*

Hi

65 779 50 nt c I c-c PHX11G2 »*IMU

•196 nt p8X9M 2 1 * pKX»42; I prtX7 pHX9422 ■ V ecu n r r r uc»— ;-c pJUUb •228 nt * tec ixc-t tv'* pHX7 - V / c-u pKX?®-c_K o-c •243 nt pHX46 ■ V

-270 nt S‘8 - , . pHX57 B1..V- Figure 9. Effect of deletions in L8 and P8 on in vitro splicing of the mt large rRNA intron. Starting with plasmid pHX993 (identical to pHX9422, except for deletion of 18 bp from 5' exon) , deletions in L8 and P8 were generated by Bal31 digestion, as described in Chapter II. pHX9422 (control) and plasmids containing deletions were linearized with BanI and transcribed with bacteriophage T3 RNA polymerase. The CYT-18 protein was isolated from mt RNP preparations of wild-type 74A by heparin-Sepharose chromatography, as described by Majumder et al. (1989). Splicing was assayed by addition of 32P-GTP to the 5' end of the excised intron under the standard conditions at 37°C for 60 min. The samples were analyzed by electrophoresis in a 5% polyacrylamide/8 M urea gel. Lane l and 11: pHX9422; Lane 2: A9, pBD9; Lane 3: A8, pBD8,* Lane 4: A7, pBD7; Lane 5: A6, pBD6; Lane 6; A5, pBD5; Lane 7: A4, pBD4; Lane 8: A3, pBD3; Lane 9: A2, pBD2; Lane 10: Al, pBDl. Right shows sequence and predicted structure of L8 and P8 and boundaries of deletions that were tested. PstI- and Hindlll- palindromes (PstIR and HindlllR) are highlighted. P and H indicate PstI and Hindlll sites used for Bal31 deletions.

77 Figure Figure 10. Effect of deletions in P4/P5 and P6 and nucleotide substitutions in P9 on in vitro splicing of the mt large rRNA intron. Plasmid pBD541 was derived from pBD5A by site-directed mutagenesis to delete P4/P5. Plasmids having deletions in P6 were derived from pBD5A by Bal31 digestion from Nhel site. Plasmid pBD591 has an EcoRI-site inserted at the base of P9 (see Figure 11) . All of the plasmids were linearized with BanI and transcribed with bacteriophage T3 RNA polymerase. Splicing of 0.5 fig of pBD5A/BanI RNA and 1 fig of other RNAs was assayed by addition of 32p-GTP to the 5' end of the excised intron under the standard conditions for 60 min. The samples were analyzed by electrophoresis in a 5% polyacrylamide/8 M urea gel. Lane 1: pBD5A; Lane 2: pBD591; Lane 3: pBD541; Lane 4: Al, pBD561; Lane 5: A2, pBD562; Lane

6 : A3, pBD563; Lane 7: A4, pBD564. Bottom shows sequence and predicted structure of P4/P5 and F6 region and boundaries of deletions that were tested.

79 80

S 53 3 (&9 Q9° CD •- a a a. < 9! 3 P4/P5 "IS1 \ 7 50.'? t pBD5Al ■I---“------Ml 550 m

389 nl » il"J> 6£jlt \ / 50 nt pBD561-- Ml}______m p B D S 6 4 p 6

♦EcoRI 65 n t \ / 50 nt PBD591 ■ m 501 nr P9

320 i U A A A U U u o I) U A rSM Uc AC A V 310--U A U A A A-330 A Al A U M A U All - ! 376 390 ■mu * A3C a dl 1A1 fcGU 1 ■“ "'"i: f l*i AA«« \ A c * I, a A G U C U / uaauu agcga^ uuuauSuu GUGGGUGCUAGC CAC UACC CGCUA G UCCU-AAAAUAGAA C Al) CCGUGAUCG*GUG--- AUCG GUGAU G ore i rti * u rib | : t c ' 410 A A A3 400 A4

Figure 10. Figure 11. Effect of deletions from the 3' exon and 3' end of the intron on activity of the mt large rRNA intron. Starting with pBD5A, sequences from the 3' exon and 3' end of the intron were deleted by Bal31 digestion from the FspI site in the 3' exon followed by religation in the presence of a BamHI linker, as described in Chapter II. Plasmids having 3' truncated sequences were linearized with BamHI. Plasmid pBD591, in which an EcoRI site was introduced at the base of P9, was linearized with EcoRI. pBD5A (control) was linearized with Banl. The linearized plasmids were transcribed with bacteriophage T3 RNA polymerase. Activity

was assayed by addition of 32P-GTP to the 5' end of the intron under standard conditions for 30 min. The samples were

analyzed by electrophoresis in a 5% polyacrylamide/8 M urea

gel. Lane 1: pBD5A; Lane 2: Al, pBD5901; Lane 3: A 2 ,

pBD5902; Lane 4: A3, pBD5903; Lane 5: A4, pBD5904; Lane 6 :

A5, pBD5905; Lane 7: A 6 , pBD5906; Lane 8 : A7, pBD5907; Lane

9: A 8 , pBD5908; Lane 10: pBD591; Lane 1 1 : A9, pBD5909; Lane 12: A10, pBD5910; Lane 13: pBD5A. Bottom shows the sequence and predicted structure of this region and the positions of truncations that were tested.

81 82

■389 nt * H

65 nt P9 15 nt pRD5A/?spl 1 *68 nt

pBD5901 — P&D5910

KcoBI pBD591 /EcoRI M h

--U

P7 2272 Al AS h$i -UAUCACCA/ . 1 . 1 I -AUACUC--U6AiAAC CA...... AA UGAUAACAAGUtXjAA£AlXCUAAl1 . .11UUUCCGCA-3' S I uP9.1 nU A_„cA— C *A * A 1 U A A'—a? i s P9A u U C G : A pBD'591/EcoRI U G

2262

Figure 11. Figure 12. In vitro splicing of 388 nt mini-intron RNA under different conditions. pBD5A was linearized with BanI and transcribed with bacteriophage T3 RNA polymerase. Splicing assays were under standard conditions for 60 min, except for varying (A) temperature, (B) MgCl2, or (C) KCl.

83 B c

_*grf-l> -cyt-ll -cyi-ll -cyt-ia ^ Tamp “ n--- *— ------1 y- Q — ■ ,------, iiSlMiiTrrtT-" (C> = 4 2111Zaa6 s’e* iXS 0sas8?l8 §S° *s 85 f? S28fi S

Figure 1 2 . 00 it* Figure 13. Effect of high GTP or guanosine concentrations on in vitro splicing of 388 nt mini-intron. In vitro transcript from pBD5A digested with BanI was uniformly labeled with [a-

32P]UTP. Transcripts were incubated under splicing conditions with or without preparations of CYT-18 protein for 60 min, and products were analyzed by electrophoresis in a 6% polyacryalmide/8 M urea gel. Lane 1: control, pBDSA/Banl transcript incubated with neither GTP nor CYT-18 protein; Lane 2: pBD5A/BanI transcript without incubation; Lane 3: pBD5A/BanI transcript incubated with CYT-18 protein preparation plus 100 fiM GTP; Lanes 4 to 9: pBDSA incubated with indicated concentration of GTP or G in the absence of CYT-18 protein preparation. Most lanes show two very light, high molecular weight bands, which are present in the original transcript preparation and which could reflect incomplete digestion of pBDSA with BanI.

85 86

£ -cyMB

503 nt 439 nt 389 nt

115M % El E2

Figure 13. Figure 1 4 . Effect of changing the IGS sequence on in vitro splicing of the mt large rRNA intron. Top shows splicing activity of pBD5A-IGS, which contains the yeast IGS sequence. Splicing assays were under standard conditions for 60 min, except for varying: (A) temperature, (B) MgCl2, or (C) KC1. Bottom shows structures of the PI and P10 pairings in the Neurospora (Nc) mt large rRNA intron, the yeast (Sc) mt large rRNA intron, and the Neurospora intron with the yeast IGS (pBD5A-IGS). Shaded areas show nucleotides that were changed.

87 B

-r^r«yt-1»-cyt-H . ...______* c y :______» -H ______- g r t - n j _ * cy t- 1>|ww^i -cyl-ilprairti " KCl (iq«j - ** (mM) (mM)

l-Tt- -—atm i-3t— * - _ | _

Me L$U rflN A M ra ft S e LSU rflNA In trtn pfiOSA-KSS

5SS 5 SS Pt j - acgcua gggau- 5 - c u a ggga uifc AQ- - acgcua g g g a u&AU------, i I I I l I i M l I l l J M I | M .UUCClLQUii-UB;-□ CCCUGUUCC CCCAUIJU- .MGiuauu tc c c caw y J / Mill / M l i M M I / ill M i l l U Ga a cag gcua a - UG aaca g g gu a auau -3 UGaacag g cua auuu- U ) Pio U t Pto M I P10 3 SS X 3 SS 3 SS

Figure 14. 00 00 Figure 15. Effect of modifying the P9.0 pairing on in vitro splicing of the mt large rRNA intron. Top shows splicing activity of the intron containing the modified P9.0 pairing. Plasmid pBD5/UC was linearized with BanI and transcribed with bacteriophage T3 RNA polymerase. In vitro transcripts were incubated with or without preparations of CYT-18 protein under standard conditions for 60 min, except for varying (A) MgCl2, or (B) temperature. In (B), the lane for transcript incubated at 70°C in the absence of CYT-18 protein shows a very light band at approximately the position expected for unspliced transcript, which may have bound some of the 32P-GTP label. (C) shows structures of the two possible P9.0 pairing of the Neurospora mt large rRNA intron (pBD5) , the P9.0 pairing of the Tetrahymena nuclear rRNA intron and the modified P9.0 pairing of Neurospora rRNA intron (pBD5/UC). Box indicates nucleotide that was changed.

89 B >eyt-1* -cyt-l* + cyt-18

,n288SSS? o 3*888888 §' MgCmnHI) 37 pBDSA

■ 1-rE 389M 389nt-

Figure 15. VO o 91 Figure 15 (continued)

Neurospora Intron (pBD5)

P7 P7 U C A G C A — U U C C A I I 1 I I P9.0 P9.0 A Q U C — U G A I I I (P9.P9.1) — A U A G U C — UGAA IUlU i (P9, P9.1) UlU 3 ' SS |y| a 3" SS-> a c I

Tetrahymena intron PBD5/UC

/ u

P7 A° P7 A A U C A G A C — U A U C A G C A I I I I I I P9.0 I I I I I I I P9.0 U A G U C — GGA — A UAGUC — UGA I I (P9, P9.1, P9.2) (P9, P9.1) CU 3' SS-»G u 3’ SS- a a Figure 16. Comparison of the TyrRS and splicing activities of the CYT-18 protein isolated from 74A, cyt-4-l, and cyt-19- 1 strains by heparin-Bepharose chromatography. (A) Comparison of the splicing activity. The splicing was assayed using the unlabeled intron RNA (pHX9422/BanI) and 40

juCi of a-32P-GTP under the standard conditions at 37°C for 60 min. The CYT-18 proteins were purified by heparin-Sepharose columns. The amount of CYT-18 proteins used in the splicing experiments are: 74A, 0.67 jug; cyt-4-1, 0.13 fig; cyt-19-1, 0.458 jug. Three sets of samples were run in parallel in the gel for better accuracy. The splicing bands are cut from dried gel and counted in Ready Protein Scintillation liquid (Beckman). (B) Western blot. Same preparations of the CYT- 18 proteins as those in (A) were used. The western blot was carried out using the C18-1 antibody as described by Majumder et al.(1989). The amount of CYT-18 proteins used in the experiment are: 74A, 2.7 /ig; cyt-4-1, 0.52 jug; cyt-19-1, 1.83 jug. Left shows a shorter exposure and right shows a longer exposure of photography. The intensities of the CYT-18 band were scanned with a densitometer and the peaks on the printed papers were cut and weighed to determine the ratio of the amount of CYT-18 protein.

92 93

c ut ‘C 'TlS. U I I I11 5 1 I* & S' e- & & & S

• a * a *

shorter longer exposure exposure

Figure 16. Table 2. Comparison of the TyrRS and Splicing Activities of the CYT-18 Protein isolated from 74A, cyt-4-1, and cyt-19-1. The CYT-18 protein was isolate from mt RNP by heparin- Sepharose chromatography. The ratio of the amount of the CYT-18 protein used in the subsequent splicing and TyrRS assayes was determined by scanning the CYT-18 protein band on the Western blot with a densitometer. The TyrRS activity was assayed with E. coli tRNATyr using the standard procesure described in Chapter II. The splicing activity was assayed by addition of 32P-GTP to the excised intron RNA (pHX9422/BanI) using the standard procesure described in Chapter II.

Ratios 74/4 cyt-19-1 cyt-4-1 Amount of CYT-18 Protein 1 1.06 1.2 TyrRS Activity 1 0.83 1.4 Splicing Activity 1 1.9 1.6 TyrRS Activity/CYT-18 Protein 1 0.8 1.2 Splicing Activity/CYT-18 Protein 1 1.8 1.3

94 CHAPTER IV THE CYT-18 PROTEIN BINDS SPECIFICALLY TO THE CATALYTIC CORE OF Neurospora MITOCHONDRIAL LARGE rRNA INTRON

Introduction

To facilitate studies on the interaction between the CYT-18 protein and the intron RNA, we constructed a mini- intron of 388 nt, compared to 2.3 kb for the original intron, that is still spliced in a protein-dependent manner (see Chapter III; Guo et al., 1991). Experiments with mutant derivatives of this mini-intron suggested that the inability of the Neurospora intron to self-splice reflects instability of a structure in the intron core or in the interaction between the core and the PI stem containing the 5' splice site (see Chapter III; Guo et al., 1991). In reverse splicing catalyzed by the CYT-18 protein, the protein appears to bind to the free intron to form an active RNP complex, which then binds ligated exon RNA (Mohr and Lambowitz, 1991). Finally, recent experiments have shown that the CYT-18 protein can suppress structural defects in the catalytic core of mutant bacteriophage T4 td or yeast u+ introns (Mohr et al., 1992). Together, these findings suggest that the CYT-18

95 96 protein promotes splicing by binding to the intron and facilitating formation of the catalytically active structure of the intron core. Similarly, the yeast CBP2 protein, which is a group I intron splicing factor specific for cob intron 5, has been shown to require an intact catalytic core of the intron for binding, indicating either that the protein binds to the core or that the core is required for proper display of sequences that bind the CBP2 protein (Gampel and Cech, 1991). The precise identification of the CYT-18 protein binding site in the intron RNA is the key to understanding how the protein functions in splicing group I introns and its evolutionary significance. In this chapter, I localized the CYT-18 protein binding site to a conserved catalytic core region of the intron by testing the binding of the CYT-18 protein to a variety of truncation and deletion mutants of the Neurospora mt large rRNA intron using the nitrocellulose filter binding technique. Considered together with recent functional analysis (Mohr et al., 1992), our results indicate that the CYT-18 protein promotes splicing by binding to the intron core and stabilizing it in the catalytically active conformation.

Results

UV-cross linking Experiments Show that CYT-18 Protein Binds to the Neurospora Mt Large rRNA Intron Directly 97 Figure 17 shows the 388 nt-derivative of the Neurospora mt large rRNA intron (pBD5A) used as the starting point for this study. This intron was constructed by deleting sequences corresponding to the ORF, which is ordinarily encoded in the intron, as well as non-essential parts of P2 and L8 . We showed that splicing of this 388 nt-intron is completely dependent on the CYT-18 protein (see Chapter III) . To confirm that the CYT-18 protein binds directly to the large rRNA intron, I carried out the UV-crosslinking experiments shown in Figure 18. In these experiments, 32P- labeled in vitro transcripts, which had incorporated BrdU at concentrations below those that interfere with splicing activity, were incubated with purified E. coli-synthesized CYT-18 protein under splicing conditions. After 10 min incubation at 37°C, RNA-protein crosslinking was carried out with 300 nm UV light, RNA-protein complexes were digested with RNase A, and proteins analyzed by SDS-PAGE. Gel patterns were electroblotted to nitrocellulose filter and analyzed by autoradiography to detect proteins containing crosslinked 32P-labeled RNA (Figure 18A) and by immunoblotting with anti-CYT-18 antibody to detect and quantitate CYT-18 protein in the samples (Figure 18B). As shown in Figure 18, the 70 kDa CYT-18 protein was crosslinked to the pBD5A/BanI transcript containing the 388 nt intron, but not to a control transcript containing exon sequences (pHXll cut with PvuII) or 32P-GTP by itself (not 98 shown here). The specific crosslinking to in vitro transcripts containing the intron was competed by the homologous intron-containing RNA, but not by similar concentrations of exon RNA, total yeast RNA, or poly(rC) (Figure 18) or by 1 mg/ml heparin (not shown) . This indicates that the CYT-18 protein binds directly to the Neurospora mt large rRNA intron in a specific manner.

Quantitive Analysis of the Interaction Between the CYT-18 Protein and the Neurospora Mt Large rRNA Intron Using Nitrocellulose Filter Binding Method Nitrocellulose filter binding method now has been widely used to characterize the interaction between purified protein and nucleic acids, either DNA or RNA, since it was first used to study the interaction between the ribosome and rRNA by Nirenberg and Leder (1964). This method is ideally suitable for kinetic studies of the interaction since the free DNA or RNA can go through the nitrocellulose filter and be separated easily from the protein-bound DNA.

In initial experiments, I used the nitrocellulose filter binding method to assay the binding of the purified CYT-18 protein to a 503 nt in vitro transcript containing the 388 nt intron derivative. In these experiments, low concentrations of 32P-labeled in vitro transcript were incubated with increasing concentrations of purified CYT-18 protein and binding was assayed by retention of the 32P-labeled RNA on a 99 nitrocellulose filter. Different amounts of intron RNA were incubated with excess CYT-18 protein under the standard conditions, as described in Chapter II. The total amount of RNA added in each reaction and the amount of RNA retained on the filter, which indicates the amount of RNA bound to the CYT-18 protein, were determined, as shown in Table 3. Since the CYT-18 protein is in excess, the ratio of the amount of RNA retained on the filter and the amount of input RNA is defined as binding efficiency. The amount of RNA retained on the filters is plotted against the total amount of input RNA, as shown in Figure 19. The slope of this plot represents the filter binding efficiency. The filter binding efficiency of the CYT-18 protein and the intron indicated by the slope of the plot is 64%. That the efficiency is less than 100% is observed generally with this technique. This could be either due to the disruption of the complexes by filtration process or due to the inactive RNA in the reaction (Yarus and Berg, 1967; Carey et al., 1983). Figure 20 shows that approximately 60 % of the CYT-18 protein is active in the binding to the pBD5A/BanI intron. The inactive protein could be due to (i) mis-folding during expression in E.coli; (ii) the protein purification procedure; (iii) storage; (iv) handling during the experiments.

As shown in Figure 21A, the CYT-18 protein bound strongly to the transcript containing the intron (pBD5A/BanI), whereas control transcripts (pHXll/PvuII and 100 pBD5A/NheI) showed only weaker binding at higher protein concentrations. Figure 2IB shows that specific binding of the CYT-18 protein to the transcript containing the intron was competed efficiently by the same transcript (pBD5A/BanI), but not by control transcripts (pHXll/PvuII or pBTC20/KpnI). Based on eleven nitrocellulose filter binding experiments, like that in Figure 21A, we calculated that the CYT-18 protein binds to the intron with a Kd of 6 ± 2 nM. The Kd value was not affected by the presence of 0.2 or 0.5 mM GTP (not shown) and is essentially the same as that for excised intron RNA (5.5 nM corrected for the proportion of active CYT-18 protein; Mohr and Lambowitz, 1991). The stoichiometry of the intron-protein complex was determined by using a protein concentration near the Kd and varying the RNA concentration by two orders of magnitude above and below the Kd. As shown in Figure 22, a double­ reciprocal plot of the binding data gave a straight line with a y-intercept close to one (1.16), indicating that there is one RNA binding site per CYT-18 protein dimer. From the slope of the line in Figure 22, we calculated a Kd of 7.8 ±

0.7 nM, in good agreement with the Kd measured by varying the protein concentration.

The CYT-18 Protein Binds to the Group I Intron Catalytic Core To localize the CYT-18 protein binding site in the intron RNA, we tested derivatives of the 388 nt mt large rRNA 101 intron containing truncations or internal deletions, for their ability to bind the CYT-18 protein by the nitrocellulose filter binding method. Binding curves for 5'- and 3'- truncations are shown in Figure 23, and Kds for the mutant introns are summarized in Figure 24. From the 5' end, introns having truncations that remove 97 nt and extend up to the second nucleotide of P3[5'] retained essentially full binding activity for the CYT-18 protein (#'s l and 2, L95 and L97; Kds=7 and 6 nM, respectively), indicating that the 5' splice site, PI, P2 and part of P3 are not required for binding. However, truncations that extend as few as ten nucleotides further to the second nucleotide of P4[5'] resulted in sharply decreased binding (#'s 3 and 4, L107 and L129). From the 3' end, the largest truncation that retained full binding activity was R3 67, which removes 22 nt from the 3' end of the intron (#5, Kd=5 nM) . More extensive 3' truncations that delete sequences between P9 and P6 resulted in a progressive increase in Kd up to 13 nM for a truncation in J6/7 (#9, R243 in Figure 23). Although this corresponds to a relatively small decrease in binding affinity, it is easily detectable in the binding assays and appears to be significant. The same effect was observed in two independent repeats of the experiment in which transcripts were assayed in parallel and was confirmed by additional experiments in which the 3' truncations were tested for ability to compete 102 binding of the wild-type intron (pBD5A/BanI, Figure 25) and by experiments testing the ability of mutant introns to compete with wild-type intron for splicing activity {Figure

26). Truncations that extend beyond J6/7 into P6 resulted in strongly decreased binding, as judged by loss of ability to compete with the wild-type intron for splicing activity (pR175 in Figure 26) and by nitrocellulose filter binding assays (e.g., R175, #10 in Figure 23B; or R239, which deletes only one nt into P6 , in Figure 25). In addition to the regions identified by 5'- and 3'- truncation, we found that several internal regions of the intron could be removed without affecting binding (Figure

24B) . These include P6b and part of L6a, (#14, Kd=6 nM) , P7.1 and P7.2 (#15, Kd=6 nM) and part of P8 (#16, Kd=7 nM) .

Deletion of P5 or P6a resulted in moderate decreases in binding (#12, Kd=43 nM; #13, Kd=52 nM), whereas deletion of P4

and P5 near the base of P4 resulted in strongly decreased binding (#11; Figure 24B) .

Construction of Small RNAs Containing the CYT-18 Protein Binding site Together, the above results indicate that the intron contains a high affinity binding site for the CYT-18 protein in the P4-P5-P6-P6a region, and there may be an additional contribution from the P7-P9 region. To confirm the location of the high affinity binding site, we synthesized an RNA that 103 contains a 75 nt region of the intron encompassing P4[5#] to

P6[3'], excluding sequences distal to L6a (RNA 1; Figure 27). The CYT-18 protein bound this RNA with a Kd=l3 nM, in good agreement with the results expected from 5' and 3' truncations. Starting with this small RNA, we tested derivatives that add additional 3' sequences. Addition of P7[5'] had no effect on Kd (RNA2; Figure 27), whereas addition of the 3' end of the intron including P7 and P9 gave a Kd of 6 nM, similar to that for the full length intron (RNA 3; Figure 27). Since the experiments in Figure 24 show that sequences distal to P9 have no effect on binding, we infer that the tighter binding of the latter RNA is due to sequences in P7-P9. In control experiments, addition of similar-length sequences that did not restore P7 and P9 either had no effect on binding or resulted in an increased Kd, presumably due to nonspecific interference with the binding of CYT-18 (not shown). The contribution of P7-P9 could indicate that P7-P9 is required for correct folding of P4-P5-P6-P6a or that the intron contains at least two binding sites for CYT-18, one in

P4—P5—P6—P6a and the other in P7-P9. Based on the measured Kds, we calculate that binding of CYT-18 to P4-P5-P6-P6a contributes -10.7 kcal/mole to AG°, whereas its potential binding to P7-P9 would contribute only an additional -0.5 kcal/mole. By itself, P7-P9 is expected to have a relatively high Kd, above the range readily measured by the 104 nitrocellulose filter binding assay. We attempted to measure the binding of the P7-P9 region to CYT-18 by inhibition of TyrRS activity, but in this experiment an RNA beginning in

P6b (Nhel site) and extending through the 3' end of intron had a K,=2.4 fxM, only slightly better than inhibition by non­ specific RNAs (K,s=4 to 12 /iM; not shown).

Discussion

I localized the CYT-18 protein binding site in the Neurospora mt large rRNA intron to a highly conserved region of the catalytic core of the Neurospora mt large rRNA intron. This indicates that the CYT-18 protein may function in splicing by binding to highly conserved structural features of the group I intron core and stabilizing the core in a conformation required for catalytic activity. My results show that the CYT-18 protein binds strongly to the P4-P5-P6-P6a region of the intron core. By itself, an RNA containing 75 nt from this region binds the CYT-18 protein with a Kd=13 nM (-10.7 kcal/mole). Intron mutants lacking P5 and P6a still bind CYT-18 relatively strongly (Kds=43 and 52 nM, respectively; Figure 24B), suggesting that key features are probably in P4-P6. Comparison of six different group I introns that both bind and can be spliced in vitro by the CYT-18 protein showed relatively little sequence similarity in P4-P6, raising the possibility that the protein recognizes conserved secondary and/or tertiary 105 structure features of this region. The previous finding that regions required for splicing are distributed throughout the CYT-18 protein and include both N- and C-terminal domains suggests that the protein contacts the intron at more than one site (see Introduction). In addition to P4-P5-P6-P6a, we find that the P7-P9 region is required for maximal binding of the CYT-18 protein, either because it is necessary for optimal folding of P4-P5-P6-P6a or because it contains a second binding site for the protein. The possibility that CYT-18 binds to these two regions is consistent with findings with the phage td intron that CYT-18 exerts its effect over a wide area of the catalytic core and can efficiently suppress mutations in P7 and P9, as well as

P4 and P6 (Mohr et al., 1992). If there is a second binding site in the P7/P9 region, the higher Kd for this site could reflect either weaker interaction with the protein or the coupling of binding to an energetically unfavorable conformational change. Figure 17. Predicted secondary structure of the 388 nt- derivative of the Neurospora mt large rRNA intron. The intron is cloned in pBD5A. Upper case letters indicate intron sequences, and lower case letters indicate exon sequences. 5/SS and 3'SS indicate 5'- and 3'- splice sites, respectively.

106 107

r4 4nts-j U A A "G C" A U A U u u A U 120— U U U A U A U A U A— 130 U A ..CO 20— A U PS A U U A U A C A U A u U U A — 80 A 5'SS— § U U A A A 160 a G U A U 1t a U a a PAOQ A U Pig C u A P2 A U C A 9 C A U 90 P4C G G A GU 9 U A U j P3 U A P6 AA P6a A G P6b A A G U G 5'-uacgcua U— ---- C GAUAAUAUAUUGGGUUAAUU AGCGA UUUAOCUU GUGGGUGCUAGC CAC UAGC CGCUA G AACCCA UGCU~AAAAUAGAA CAUCCGUGAUCG-GUG AUCG-GUGAU G C C U [ A « I 300 C 230 A A 210 U I I R A CC-AU uuuuuucouu -UAUCAGCA GGAUA AAAAAAGAAAUGAA6AA- —AUAGUC-UGAAC GA---AA (JGAUAACAAGOUGaacaggcuaauuu-3' U { P8 "F S P7 C G U A 320 r P9 A U A V P9.1 T A U A A A 3 S S A U G A A U 350— A U u c U U A I 280— A G A 370 U U G P7.2 u P7.1 u c A— 260

Figure 17. Figure 18. UV-crosslinking experiment showing that the CYT- 18 protein binds the Neurospora mt large rRNA intron directly. About 50 nM of the E. coli-synthesized CYT-18 protein was incubated with 100,000 cpm of MP-UTP labeled intron RNA at 37°C for 10 min. The samples were irradiated with 300 nM UV light (Fortodyne 3-3000) through the petri dish lid for 2 min. After UV-irradiation, the samples were treated with 1 mg/ml of RNAse A at 30°C for 30 min, and then loaded onto the SDS protein gel. The complex of the CYT-18 protein and the intron fragments were electro-blotted to a nitrocellulose filter, and the filter was immuno-stained with the C18-1 antibody, as shown in (B). After the immuno- staining, the filter was autoradiographed, as shown in (A). pBD5A/BanI is the transcript containing the 388 nt intron derivative of the Neurospora mt large rRNA intron. pHXll/PvuII is the transcript containing the 171 nt 5/ exon of the mt large rRNA. The numbers on the right are the molecular weight markers. The numbers on the top show the amount of competitors used in the reaction mixture for the UV-crosslinking experiments.

108 1 0 9

pHX11/Pvu n pHX11/Pvu ♦ pBD5A/BanI + yoast RNA + poly (rC) pB D 5A /B anI pHXII/Pvu H pBD5A/BonI

(0 I f) ■>< > > 0 (0 N CD

o o a cn I 1 <5 i otto ro otto (0 i» o ) A 0 W o

Figure 18 Table 3. The efficiency of the nitrocellulose filter binding between the CYT-18 protein and the Neurospora mt large rRNA Intron. Nitrocellulose filter binding of varying amounts of intron RNA (pBD5A/BanI) and 377 nM CYT-18 protein was assayed in 250 nl volume using the standard procedure as described in Chapter II. The CYT-18 protein was provided by Dr. Roland Saldanha and purified from the overexpressed CYT-18 protein in E. coli. The input RNA (cpm) indicates the total amount of radioactive counts per min added in the reaction. The RNA on filter indicates the radioactive counts per min on the filters.

110 Table 3

Input RNA (cpm) RNA on Filter (cpm) 3,424,699 2,193,580 1,693,169 1,076,633 354,324 230,905 266,857 161,996 175,512 113,014 87,169 55,829 36,341 24,982 27,462 17,596

18,137 11,298 8,947 5,743 3,924 2,525 2,884 1,773 1,842 1,256 931 698 Figure 19. Nitrocellulose filter binding efficiency of the CYT-18 protein and Neurospora mt large rRNA intron. The varying amounts of 32P-intron RNA (pBD5A/BanI; from 0.09X104 cpm to 3 4 2X104 cpm or from 8.5X10'8 /xmoles to 2.25X10'11 /umoles) was incubated in 250 nl volume with an excess of the CYT-18 protein (377 nM). The amounts of RNA remained on the filters (cpm) are plotted relative to the amount of total input RNA (cpm) . The slope of the plot indicates the ratio of the amount of the intron RNA and the total amount of RNA added in the reaction. This ratio was defined as filter binding efficiency (Yarus and Berg, 1967). The amounts of input RNA and RNA bound (cpm) are shown in Table 3. The method is according to Yarus and Berg (1967).

112 RNA Retained on Filter (X104 cpm) 50 -5 200 250 100 150 50 0 100 iue 19. Figure Input RNA (X104 cpm) (X104 RNA Input 200 300 400 113 Figure 20. Fraction of the active CYT-18 protein for nitrocellulose filter binding experiments. Fixed amount of intron RNA (810 nM) was incubated with varying amounts of CYT-18 protein (from 50 to 3000 nM) . At the half maximum binding of the intron RNA, the CYT-18 protein concentration is about 710 nM, as indicated by the dashed lines in the figure. At the half-maximum binding, the RNA bound is assumed to be half of the total RNA, i.e. 405 nM. The proportion of active CYT-18 protein is 57% (= 405/710 nM) . The method is according to Riggs et al. (1970).

114 RNA Bound (percent) 20 30 40 50 70 60 0 0 0 500 1000 Y-8 rti (nM) Protein CYT-18 iue 20. Figure 1500

2000

2500

3000

3500 115 Figure 21. Nitrocellulose filter binding of the CYT-18 protein to the Neurospora mt large rRNA intron. (A) Binding of CYT-18 protein to the pBD5A/BanI transcript containing the 388 nt-derivative of the Neurospora mt large rRNA intron and to nonspecific RNAs (pHXll/PvuII, 171 nt RNA containing 5' exon sequences; and pBD5A/NheI, 2 39 nt RNA containing 65 nt of 5' exon plus 174 nt of intron truncated in P6b). The percentage of J2P-labeled RNA retained on the filter is plotted as a function of concentration of CYT-18 protein dimer (nM). (B) Binding of the CYT-18 protein to the intron RNA (pBD5A/BanI) competed by the same RNA or nonspecific RNAs (pHXll/PvuII, see above, and pBTC20/KpnI, 869 nt transcript corresponding to the anti-sense strand of Neurospora nuclear gene cyt-20). The amount of RNA bound in the presence of competitor divided by the amount bound in the absence of the competitor (8) is plotted as a function of the competitor RNA concentration (nM).

116 117 A

60 —

5 0 -

3 40 — TJ c g 3 0 - pBD5A/NheI I % 2 0 -i

1 0 -?

40 80 120 160 CYT-18 Protein (nM)

B 1.0

0.8 -

0.6 --'

B 0.4 —

0.2 --

pBD5A/BanI

40 120 160 200 Competitor RNA (nM)

Figure 21. Figure 22. stoichiometry of the complex between the CYT-18 protein and pBD5A/Banl intron RNA. The reciprocal of the fraction of CYT-18 protein in the complex (1/v) is plotted against the reciprocal of the concentration of free RNA (1/ [RNAfree], nM"1) . The amount of bound RNA was corrected for the efficiency of binding (0.64, see Figure 19). The concentration of free RNA [RNAfree] at each point was calculated by subtracting the concentration of bound RNA corrected for efficiency from the total RNA concentration (Yarus and Berg, 1967). The data are averages for duplicate samples.

118 1 1 fv 0.5 — 0.5 3.0 -r 3.0 2.0 - - 2.5 . - - 1.5 — 0.05 Figure 22. Figure 0.15 0.2 119 Figure 23. Nitrocellulose filter assay of binding of tbe CYT-18 protein to derivatives of the Neurospora mt large rRNA intron. (A) Binding of CYT-18 protein to derivatives of the mt large rRNA intron having 5'- truncations. (B) Binding of CYT-18 protein to derivatives of the mt large rRNA intron having 3'- truncations. The percentage of 32P-labeled RNA retained on the filter is plotted as a function of concentration of CYT-18 protein dimer (nM). All experiments were repeated at least once with essentially the same results. Circled numbers identify RNAs depicted schematically in Figure 24.

120 _RNA Sound (percent) 0 7 RNA Bound (percent)

CD

ro

i . c© „ I \ \ cn n ® © 0 © * ©

ro M Figure 24. Summary of tbs binding of the CYT-18 protein to derivatives of tbe Neurospora mt large rRNA intron. (A) 5'- or 3'- truncations and (B) internal deletions. RNAs were:

(1) L95, Kd=7 nM; (2) L97, Kd=6 nM; (3) and (4) L107 and L129, binding not saturated at 168 nM CYT-18 protein; (5) R367, Kd=5 nM; (6) R359, Kd=9 nM; (7) R342, Kd=ll nM; (8) R312, Kd=ll nM; (9) R243, Kd=13 nM; (10) R175, binding not saturated at 168 nM CYT-18 protein; (11) A108-139, binding not saturated at 188 nM CYT-18 protein; (12) A115-133, Kd=43 nM; (13) A149-233,

Kd=52 nM; (14) A155-224, Kd=6 nM; (15) A252-282, Kd=6 nM; (16) A258-323, Kd=7 nM. Kd values in figure are averages from two independent nitrocellulose filter binding experiments, except for 14, 15 and 16 which are from one experiment. The intact 388 nt mt large rRNA intron (pBD5A/BanI) was run in parallel with mutant RNAs in all experiments.

122 123

A. 5'- and 3'- Truncations

U A A A U U U 120— U U u A u A @ -> 1 6 8 nM c G A U I U A 1P5 UA @ ->168 nM i A k A @ - 6 nM V AU AU >168 nM © 7 nM A u 1P4 c G I \ U A P61 aa P6a a- 5 ' — (9 5nt)--UUGGGUUAAUU AGCGA UUUAUCUU 68nt P3 AACCCA UGCU-AAAAUAGAA e c u AA- 300 J

c u | A A CC-AU UUUUUUCUUU uaucagca* @ -13nM U GGAUA AAAAAAGAAAUGAAGAA- AUAGUC-UGAAC GAAAU-— (2 2 nt)— 3' G 4 U P8 P7 AF U (5) - 5 nM 7 A A G ; ® - 9 nM ® -11 nM A ® -11 nM U A- G U A U ; 280--A U G 3 A — 355 U A U A U G P7.2 U A c G P7.1 U U G U c U A--260

Figure 24. 124 Figure 24 (continued)

B. Internal Deletions

I------@ ->188 nM-—,

— (12)-43nM U A A A U U u u u u P5 c A _u_&___ l U A A A A--135 A U \llO- A U p4 A U \ ^ Q j* __ P3 ' U-A- P6 AA I P6a [a— - (95nt)— UOGGGUUAAUU AGCGA UUUAUCtKJ AACCCA UGCU-AA|AAUAGAA "J 68nt(14)-6nM -52 nM C C U | AAj--- 1 i 300 J J \ c u i I A \______A CC-AU UUUUUUCUUU f-1-UAUCAGCA U GGAUA AAAAAAGAAAUGAAGAA- — I-AUAGUC—UGAAC GAAAU— (22nt)— 3 ' G U P8 f1P7 c G A_ AU a I A U / A P91 -L bJ G U G 1 U A- G A U P7.2 u A U P 7.1 UA A U G A -355 U A A u G U A G 1 u U U I c A I A I I____ @ -7nM — J

(Hi) - 6nM ------Figure 25. Binding of the Neurospora mt large rRNA intron competed by a series of RNAs having 3' truncations of intron

sequences. 32P-labeled RNA containing the intron (pBD5A/BanI) was mixed with different concentrations of unlabeled competitor RNAs indicated in the figure. CYT-18 protein was added and nitrocellulose filter binding assays were carried out under standard conditions. The amount of RNA bound in the presence of competitor divided by the amount bound in the

absence of the competitor (0 ) is plotted as the function of the competitor RNA concentration. All experiments were repeated at least once with essentially the same results.

125 126

1 .2 -r

1.0

9 0.6 j

-R 243

0.2 + R359 R367 pBDSA 200 400 600 800 1000 Competitor RNA (nM)

Figure 25. Figure 26. Inhibition of splicing activity of the CYT-18 protein by a series of RNA derivatives of the N&urospora mt large rRNA intron truncated from the 3' end. A series of 3'- truncated intron derivatives were tested for their abilities to compete with the intron RNA for the splicing activity of the CYT-18 protein. The plasmids were linearized with the enzyme indicated and transcribed with bacteriophage T3 RNA polymerase. The intron RNA (pHX9422/BanI in vitro transcript) was mixed with different amounts of truncated intron derivatives as indicated in the figure. The reactions were initiated by adding 42 nM of CYT-18 protein prepared from E. coli and incubated at 37°C under standard splicing conditions for 6 min. "I" represents intron; "i-3'E" represents intron plus 3'-exon. The transcript lane is the pHX9422/BanI in vitro transcript only. The "no competitor" lane is the splicing reaction without any competitors.

127 a 1 -1 KJ J5. S sj J fo _ Ki SI? - sJ? no competitor no

Figure 26. 8ZT Figure 27. Binding of CYT-18 protoin to small RNAs containing different regions of the Neurospora mt large rRNA intron catalytic core. (1) pI75/BamHI, RNA containing 75 nt of the intron encompassing J3/4 to J6/7 with sequences distal to P6a deleted. (2) pR258/BamHI, RNA containing the same 75 nt intron sequences as RNA (#1) plus an additional 15 nt adding P7[5'j. (3) pA258-323/BanI, RNA containing the same intron sequences as RNA (#2) plus an additional segment restoring P7 and P9 and extending into the 3' exon. In addition to intron sequences, all three RNAs have the vector sequence between the T3 promoter and HindiII site (GGGAACAAAAGCTTGG) attached to the 5' end of the RNA, and RNAs (#1) and (#2) have part of a BamHI linker (CCGGATC) attached to the 3' end. RNA (#3) has a BamHI linker sequence (CCGGATCCGG) between P7[5'J and P7[3']- Kd values for (#1) and (#3) are averages from two separate experiments, whereas the Kd value for (#2) is from one experiment.

129 130

U A A A u U u U u U A P5 U A c G AU U A u A A A A AU AU P4 A u C G u A P6 P6a a 5 '— GGGUUAAUU AGCGA UUUAUCU UGCU-AAAAUAGA A U A -AGUGGA C I® -13 nM (3)-6nM ( 2) -1 3 nM BamHI UAUCAGCA <~| linker AUAGUC-UGAAC GA- —AAUGAUAAC- | A P7 C G U...... T ' 3 ‘ GAAAUGAAGA A U A 5 P 9 -1 BanI P9 U A AA u G AA AU U C UA GA U G

Figure 27. CHAPTER V THE CATALYTIC CORE OF NEUROSPORA HT LARGE rRNA INTRON MAY RESEMBLE THE STRUCTURE OF MT TYROSYL-tRNA

Introduction

The cyt—18 gene encodes a bifunctional protein, which has both TyrRS and splicing activity (Akins and Lambowitz, 1987). That the CYT-18 protein can splice a number of different group I introns in Neurospora mitochondria suggests that the CYT-18 protein recoganizes conserved structural features in these introns (Collins and Lambowitz, 1985). The fact that the CYT-18 protein is an aminoacyl-tRNA synthetase raises the possibility that these structural features resemble those found in tRNAs (Akins and Lambowitz, 1987). Previous studies in our laboratory have identified some regions of the CYT-18 protein required for splicing or TyrRS activities. Parallel inhibition of splicing and TyrRS activity have been observed in a number of mutations in the CYT-18 proteins. First, a number of linker insertion mutations at the C-terminal domain inhibit both TyrRS and splicing activities. Second, deletions in the idiosyncratic N-terminal domain inhibit both splicing and TyrRS activities when assayed with the Neurospora mt tRNATyr (Cherniack, 1991) .

131 132 The finding that different regions of the CYT-18 protein required for splicing may also be involved in binding the tRNA suggests that similar interactions may be used for the binding of both the intron and tRNA substrates. In the last chapter, I presented data showing that the protein binding site or sites for CYT-18 are located in highly conserved regions of the catalytic core in the Neurospora mt large rRNA intron. The major binding site is in the regions between P3 and P6 and the potential low affinity binding site is in the P7-P9 regions. In this chapter, I show that the CYT-18 protein can also bind strongly to a number of other group I introns. The Neurospora mt large rRNA intron and tyrosyl-tRNA can compete with each other for the synthetase or splicing activity of the CYT-18 protein. My results suggest that the group I intron catalytic core, which is specifically recognized by the synthetase, has structural features that resemble those in tRNAs.

Results

Binding of the CYT-18 Protein to other Group I Introns Since the CYT-18 protein can function in the splicing of a number of group I introns in Neurospora mitochondria, I examined the binding of the CYT-18 protein to other group I introns using the nitrocellulose filter binding method. The results summarized in Table 4 showed that the CYT-18 protein 133 binds strongly to a number of other group I introns, Including mt introns from Neurospora, Podospora, and yeast, as well as the td, sunY and nrdB introns of bacteriophage T4. The introns that bind the CYT-18 protein strongly belong to different subclasses and include self-splicing, as well as non-self-splicing introns. An apparent exception is the Tetrahymena nuclear rRNA intron, but recent studies have shown that CYT-18 binds strongly to a derivative of this intron that lacks the large peripheral structure P5abc, which presumably obscures the CYT-18 binding site (Kd=9 nM; see Chapter VI) . The CYT-18 protein did not bind strongly to group II introns, as expected since it does not function in splicing group II introns (Collins and Lambowitz, 1985). The finding that the CYT-18 protein binds strongly to a number of different group I introns suggests that it interacts with conserved structural features of these introns.

Competition between the intron RNA and Neurospora Mt tRNA1*' for the CYT-18 Protein We wished to compare the ability of the CYT-18 protein to bind group I introns with its ability to bind tRNA75".

Since sufficient quantities of Neurospora mt tRNA75" could not be isolated from mitochondria, we constructed a plasmid, pTYR, which enabled us to synthesize large amounts of Neurospora mt tRNA75" in vitro from the phage T3 promoter. The in vitro transcript lacks modified nucleotides, but is still 134 recognized efficiently by the CYT-18 protein (see Figure 6 in Chapter II). To investigate whether the intron and tRNA bind to similar sites in the CYT-18 protein, I tested their ability to compete for synthetase and splicing activity, respectively. Figure 28 shows that the Neurospora mt tRNATyr

inhibited splicing of the mt large rRNA intron. Although inhibition required a relatively high concentration of the Neurospora mt tRNATyr (>25 /iM at 0.11 jtM intron RNA; see Figure Legend), this tRNA was nevertheless a better inhibitor of splicing than other tRNAs tested, including E. coli tRNA13", E. coli tRNAVtl, total E. coli or yeast tRNA and in vitro transcribed E. coli or yeast tRNA”1*, all of which inhibited

only at higher concentrations. Based on the amount of the Neurospora mt tRNATyr required for half-maximal inhibition and the measured Kd for the complex between the CYT-18 protein and

the intron RNA, we calculated that the Kd for the binding of

Neurospora mt tRNA1*' is approx. 2 in reasonable agreement with the Kn, value for the tRNA in the aminoacylation reaction

(7 ± 3 fMj Cherniack, 1991) . This K,,, has been a reasonable

approximation to Kd for synthetases (Schimmel and Soli, 1979;

Schimmel, 1989) . The finding that Neurospora mt tRNATyr is a

better inhibitor of splicing than is E. coli tRNATyr is consistent with their relative K„s in the aminoacylation reaction (7 and 25 /liM, respectively; see Chapter II) and suggests that the more efficient inhibition may depend on 135 structural features that are present in the Neurospora mt tRNA73'1, but not in the E. coli tRNA1*'. In reciprocal experiments, transcripts containing the Neurospora mt large rRNA intron (pHX9422/BanI and pBD5A/BanI) strongly inhibited TyrRS activity of the CYT-18 protein, using either E. coli or Neurospora mt tRNA73", whereas similar amounts of control RNAs (total yeast RNA and pHXll/PvuII containing 5' exon) had a relatively small effect on synthetase activity (Table 5) . To determine the nature of the inhibition, TyrRS activity was assayed with different amounts of Neurospora mt tRNATyr substrate and intron inhibitor. A Hanes-Woolf plot of the data gave a series of parallel lines, indicating that the intron RNA is a competitive inhibitor (Figure 29) . The calculated K,, approx. 39 nM, is in reasonable agreement with the Kd value for the intron at the pH of synthetase reaction medium (Kd=26 nM atpH

8 .8 ) and shows again that the intron binds more strongly than the tRNA (Km=5.7 /xM in this experiment) . The finding that the inhibition is competitive provides evidence that the binding site for the intron overlaps that for the tRNA.

Comparison of Group I Introns That Bind and Are Spliced by the CYT-18 Protein In general, different group I introns have minimal sequence homology (<10%), but all share conserved secondary and tertiary structures that are required for catalytic 136 activity (Cech, 1990; Michel and Westhof, 1990; Heuer et al., 1991). Figure 30 is a schematic comparing six different introns (Nc LSU, Nc ND1, Sc LSU, Pa LSU1, T4 td, and Tt LSU with P5abc deleted) that both bind and can be spliced in vitro by the CYT-18 protein. The region containing the high affinity binding site, P4-P5-P6-P6a, has little sequence similarity in these introns, except for the A-residues in

J3/4, L4/5 and L6/6a, which are conserved in most group 1 introns. Outside of this region, the only extensive conserved sequence is P7, which is highly conserved in all group I introns.

Discussion

I show here that the CYT-18 protein, the Neurospora mt TyrRS, binds strongly to a number of different group I introns from several organisms. Previous work showed that the CYT-18 protein functions in splicing a number of different group I introns in Neurospora mitochondria (Collins and Lambowitz, 1985) and could also promote splicing of mutants of the phage T4 td or yeast w+ introns that have structural defects in different regions of the catalytic core

(e.g., P4, P5, P6 , J6/7, P7, P8 and P9; Mohr et al., 1992). Together, these results indicate that the CYT-18 protein functions in splicing by binding to highly conserved structural features of the group I intron core and presumably stabilizing the core in a conformation required for catalytic 137 activity. The findings that the Neurospora mt large rRNA intron is a competitive inhibitor of aminoacylation and that the Neurospora mt tRNA1*' inhibits splicing indicate that the intron binding site in the CYT-18 protein overlaps the binding site for the tRNA. This conclusion is consistent with findings that regions of the CYT-18 protein required for splicing, including part of the idiosyncratic N-terminal domain, may also be involved in binding tRNA75" (see

Introduction) and that small linker insertions in the putative C-terminal tRNA binding domain have parallel effects on splicing and synthetase activity (Kittle et al., 1991; Mohr et al., unpublished data).

The binding of CYT-18 protein to the intron RNA (Kd=6 nM) appears to be substantially stronger than its binding to tRNA1*' (estimated Kd= 2 to 7 /iM; see Results) . A caveat here is that for experimental reasons the tRNAs used for assays were either E. coli tRNATyr or an in vitro transcript of

Neurospora crassa mt tRNA15", which lacks modified nucleotides and for that reason may not be optimally recognized by the synthetase. Nevertheless, the observed Kd in the jiM range is similar to that for other tRNA/synthetase complexes, which generally have relatively high Kds to facilitate rapid turnover for protein synthesis (Schimmel, 1987, 1989). The tighter binding of the intron indicates either that additional interactions are involved and/or that binding of 138 the intron requires less distortion of RNA structure than does binding of the tRNA (cf., Rould et al., 1989). Since the difference in Kds for the intron and tRNA corresponds to a relatively small difference in AG° (-11.2 kcal/mole for the intron compared to -8.5 kcal/mole for the tRNA), a relatively small number of additional interactions may be involved. Although different group I introns have only minimal sequence homology and can be divided into subclasses based on ancillary structures, all have similar secondary and tertiary structures that include a highly conserved core region required for catalytic activity (Cech 1990; Michel and Westhof, 1990; Heuer et al., 1991). The CYT-18 protein binds directly to this catalytic core and presumably recognizes structural features that are conserved in different group I introns. Since most group I introns have not coevolved with the CYT-18 protein and are not normally dependent on it for splicing, we infer that these structural features are conserved for some other reason, presumably because they are required for catalytic activity. Figure 31A shows a three-dimensional model for the Tetrahymena rRNA intron catalytic core interacting with PI and P10 (Michel and Westhof, 1990). Although the Tetrahymena intron belongs to a different structural class than the Neurospora mt large rRNA intron, it nevertheless binds CYT-18 strongly (Kd= 9 nM; see Chapter VI) and should contain all structural features essential for this binding. According to 139 the structural model, the catalytic core of the intron consists of two extended helices, one formed by stacking of P6-P4-P5 and the other P8-P3-P7, with the active site formed by a cleft between the two helices. The high affinity binding site for CYT-18 is located on one helix and the possible low affinity site is on the other helix. If there are binding sites on both helices, a major function of the CYT-18 protein may be to stabilize the two helices in the correct relative orientation to form the active site. This possibility is consistent with the finding that the CYT-18 protein suppresses mutations in the phage td or yeast w* introns' J6/7 region, which is believed to be involved in tertiary contacts that play a key role in establishing the correct orientation of the two major helices (Michel and Westhof, 1990; Mohr et al., 1992). Further, it may also explain why the CYT-18 protein can suppress mutations in many different regions of the intron core, all of which potentially interfere with establishing the correct orientation of the two helices. The findings that the CYT-18 protein is a synthetase and uses the same or overlapping regions to bind both the intron and tRNA have three explanations, which are not mutually exclusive: (i) the protein may bind to different RNA motifs in the intron core and the tRNA, (ii) the protein may bind to small sequences or structural motifs that are present in both the intron core and the tRNA, or (iii) the protein may 140 recognize similar overall three-dimensional structures of the intron core and tRNA. Figures 3 IB and C compare the model for the part of the group I intron core involved in binding CYT-18 protein with the three-dimensional structure of yeast tRNAPbe. Michel and

Westhof (1990) noted that the stacked helix formed by P6 and P4 is structurally analogous to the tRNA helix formed by stacking of the anticodon and D-arms in being stabilized by base triples with both the incoming and outgoing single­ stranded regions (J3/4 and J6/7 in the case of the intron). The potential correspondence between the P4/P6 helix and the D-arm/anticodon arm helix is supported by the finding that the C-terminal domain of the protein, which in bacterial TyrRSs is believed to interact with the anticodon arm of tRNATyr (Bedouelle, 1990) , is required for high affinity binding of the intron and may therefore interact with the P4/P6 region (Mohr et al., unpublished data). Figure 32 compares the sequences of the Neurospora mt tRNA1^ and the Neurospora mt large rRNA intron. For the purpose of comparison, the intron is drawn unconventionally in the form of a clover leaf, using the potential correspondence between the D-arm/anticodon arm and the P4/P6 region of the intron for orientation. The degree of sequence similarity is surprisingly high (circled nucleotides), but only a subset of the matching nucleotides correspond to those conserved in other group I introns shown to bind CYT-18 141 functionally (shaded nucleotides). Of the features that are conserved, the most striking is the similarity in both sequence and secondary structure between the P7 stem and the variable arm of the tRNA. In addition, P9 has a length of 7 bp, the same as that of an acceptor stem in five of the six introns to which CYT-18 binds functionally (Figure 30), the exception being the Tetrahymena intron where P9 consists of 5 bp with a bulged nucleotide and there is one extra nucleotide in L9. The relatively high degree of sequence similarities between the Neurospora mt large rRNA and mt tRNA1*'’ may reflect co-evolution with the CYT-18 protein and could contribute to the somewhat tighter binding of the Neurospora mt large rRNA intron relative to other group I introns (Table 4). The potential correspondence between P7 and the variable arm is intriguing, since the Neurospora mt tRNATyf belongs to a subclass of tRNAs (type II) characterized by long variable arms (Rich and RajBhandary, 1976) . The long variable arm has been implicated in the recognition of E. coll tRNA1*' (Himeno et al., 1990) and is an idiosyncratic feature that could interact with the idiosyncratic N-terminal domain of the CYT- 18 protein and contribute to the Neurospora mt tRNA1*' being a better inhibitor of splicing than is E. coll tRNA1*'. In the case of yeast tRNA8", which also has a long variable arm, footprinting experiments have shown that SerRS binds to both the variable arm and the base of the anticodon stem (Dock- 142 Bregeon et al., 1990). If the situation is similar for the Neurospora mt tyrRS, a key structure involved in recognition may be the cognate of the variable arm, P7, hinged off the cognate of the anticodon arm, P4/P6 (cf. Dock-Bregeon et al., 1989, 1990). The finding that splicing activity of the Neurospora mt TyrRS requires an idiosyncratic N-terminal domain not found in bacterial or yeast mt TyrRSs suggests that the Neurospora mt TyrRS may have adapted to function in splicing relatively recently in evolution, after the divergence of Neurospora and yeast (Cherniack et al., 1990; KSmper et al., 1992). However, the present results raise the possibility that the TyrRS may have been predisposed to function in splicing by virtue of its ability to recognize conserved structural features of group I introns that resemble those in tRNAs. The possible structural similarities between group I introns and tRNAs could reflect convergent evolution leading to the presence of similar RNA sequence or structural motifs. The alternate possibility is that group I introns and tRNAs are evolutionarily related. In principle, either RNA species could have evolved into the other. However, if we assume that group I introns are ancient, we can speculate that catalytically active RNAs, which were the ancestors of group I introns, also gave rise to tRNAs. The possibility that group I RNAs evolved into tRNAs is consistent with the hypothesis that life began in an "RNA world", in which case all the machinery for protein synthesis, including tRNAs, would have had to evolve from catalytic RNAs. The possible involvement of catalytic RNAs in the evolution of protein synthesis was suggested early on (reviewed in Darnell and Doolittle, 1986) and has been reinforced by recent findings that the Tetrahymena ribozyme binds arginine (Yarus, 1988) and that antibiotics that interact with the decoding region of E. coli 16S rRNA specifically inhibit splicing of group I introns (von Ahsen et al., 1991). Weiner and Maizels (1987) proposed that tRNAs originated as "genomic tags" at the 3' ends of self- replicating, catalytically active RNAs and that such RNAs gave rise to both tRNAs and synthetases, which functioned in their own aminoacylation by RNA catalysis. Recently, Piccirilli et al. (1992) provided evidence for the feasibility of such a reaction by demonstrating that the Tetrahymena ribozyme could catalyze the reverse of an

aminoacylation reaction, hydrolysis of an aminoacy1- bond attached to an oligonucleotide ending in CCA. Although present day group I introns do not contain a CCA terminus, the Mauriceville plasmid, an autonomous element, which may be related to group I introns, has a CCA terminus and a 3' terminal tRNA-like structure that may be a "genomic tag" for a reverse transcriptase encoded by the element (Kuiper and Lambowitz, 1988). The objective now will be to solve the structure of the intron and its complex with the TyrRS to 144 determine the extent of the structural similarities. Table 4. Binding of the CYT-18 protein to different introns.

The binding of the CYT-18 protein to 32P-labeled RNAs containing different introns was measured by the nitrocellulose filter method. Group I introns are classified according to Michel and Westhof (1990). References: This paper; bProvided by Dr. Peter Zassenhaus (St. Louis University); the transcript was synthesized with T7 RNA polymerase and consists of 483 nt 5' exon, 1143 nt intron, 61 nt 3'exon; cHeinen, 1991; dBelfort et al., 1987; eGott et al.,

1986; fShub et al., 1987; *Wallweber and Lambowitz, unpublished data. For pNDl/PvuII, the transcript was synthesized with T3 RNA polymerase and consists of 539 nt 5' exon, 1118 nt intron, and 738 nt 3' exon. For pND4/PvuI, the transcript was synthesized with T3 RNA polymerase and consists of 708 nt 5' exon, 1449 nt intron, and 984 nt 3' exon. For pND4L/EcoRI, the transcript was synthesized with T3 RNA polymerase and consists of 696 nt 5'exon, 1492 nt intron, and 178 nt 3' exon; hZaug et al., 1986; ‘Guo and

Lambowitz, unpublished data. The transcript was synthesized with T7 RNA polymerase and consists 130 nt 5' exon, 1449 nt intron, and 285 nt 3' exon; jGarriga and Lambowitz, 1984; kDib- Hajj, 1990; 'Jarrell et al., 1988. * indicates that binding was not saturated at the indicated concentration of CYT-18

145 146 protein. Data are mean ± standard deviation for at least two independent experiments for each intron. The E. coli- synthesized CYT-18 protein used in most of these experiments was kindly provided by Dr. Roland Saldanha. 147

Table 4

Intron Plasmid Group Kj (nM) Nc. mt LSU pBDSA/Banl* IA1 6 ± 2 (mini-intron)

Nc. mt LSU pHX 11/Ban I* IA1 10 ± 2 (full length)

Sc. mt LSU pBS 188’ Nc. ND4L pND4L/EcoRI* IC1 58 ± 9 Nc. ND3 pND3/XbaT IC2 46 ± 8 Nc. ND4 pND4/PvuI* IC2 51 ± 1 Nc. cob-11 pSP64-H2a/XmnIj ID 38 ± 8

Sc. cob-11 pSDHl/BamHP II > 188’

Sc. coxl-157 pJD20/HindIII‘ II >188* Figure 28. Splicing activity of CYT-18 protein is inhibited by the Neurospora mt tRNA1*'. Different competitor RNAs in amounts (/ug) indicated in the Figure were premixed with 0.5 pg (0.11 /xM) of pHX9422/BanI transcript, which contains a 583 nt-derivative of the Neurospora mt large rRNA intron. The molar concentrations of the tRNA species at 15 fig ranged from 25 to 29 fM. Splicing was initiated by addition of CYT-18 protein purified from wild-type 74A mt RNPs by heparin-

Sepharose chromatography and assayed by addition of a-[32PJ-

GTP to the 5' end of the excised intron. Neurospora crassa (N.c.) mt tRNA75" was transcribed from plasmid pTYR/BstNI, and

E. coli tRNA”16 and yeast tRNAPhe were transcribed from plasmids p67CF0/BstNI and p67YF0/BstNI, respectively. E . coli tRNATyr was obtained from Subriden RNA (Rolling Bay, WA) . E. coli tRNAVd and mixed E. coli and yeast tRNAs were obtained from

Sigma. The somewhat different concentrations of Neurospora mt tRNA15" required for inhibition in the top and bottom panels reflect the use of different CYT-18 protein preparations. The experiments shown in both panels were repeated at least once with essentially the same results.

148 149

o CD x CM CM E. coli N.c. mt E. coll E. coll yeast fl" ,Val CT> tRNA tRNATyr tRNATyr tRNA' tRNA

a. 15 30 60 15 30 60 15 30 60 'l5~30 60 'lS~~30 60 pHX9422/BanI

5 8 4 nt

N.c. mt E.coli yeost tRNATyr tRNAPh* tRNAph* Z 0 ^ 0 4 0 “ 20 30 4(J '20 30 4 0

Figure 28. Table 5. The Neurospora mt large rRNA Intron Inhibits TyrRS activity of the CYT-18 protein. The tRNATyr substrate (l fig) was pre-mixed with competitor RNA at concentrations indicated and TyrRS reactions were initiated by adding CYT-18 protein preparation (28 fig micrococcal nuclease-digested mt RNP preparation from N. crassa wild-type 74A ). The experiment was repeated three times using different CYT-18 protein preparations with similar results.

150 151

Table 5

Substrate Competitor Amount (/ig) cpm Activity (percent) E. coli None 0 17,841 100 tRNA1*1 Nc mt large rRNA 5 50 0.3 intron: pHX9422/BanI 10 74 0.4 20 132 0.7 Nc mt large rRNA 5 0 0 intron: pBD5A/Banl 10 6 0 20 1,070 3 5’ exon: pHXll/PvuII 5 15,712 88 10 14,992 84 20 11,196 63 yeast total RNA 5 25,427 142 10 25,357 142 20 22,759 127 Nc mt None 0 16,561 100 tRNA1* Nc mt large rRNA 5 407 2.5 intron: pHX9422/BanI 10 381 2.3 20 401 2.4 Nc mt large rRNA 5 479 2.9 intron: pBD5A/BanI 10 451 2.7 20 494 3.0 5’ exon: pHXll/PvuII 5 16,566 100 10 15,869 96 20 13,632 82 Figure 29. Competitive inhibition of TyrRS activity of the CYT-18 protein by the Neurospora mt large rRNA intron. Different amounts of Neurospora mt tRNA75" substrate were pre­ mixed with intron RNA (pBD5A/BanI) at the concentrations indicated in the Figure. TyrRS reactions were initiated by addition of CYT-18 protein preparation (28 fig micrococcal nuclease-digested mt RNPs from N. crassa wild-type 74A ). The substrate concentration (S, /xM Neurospora mt tRNATyr) divided by the rate of product formation (v, /uM/min) is plotted as a function of the substrate concentration (S, fiK) (Dixon and Webb, 1979) . [I] indicates the concentration of the inhibitor intron RNA (pBDSA/Banl). The data are averages for duplicate samples.

152 ° 0 , Figure 30. Comparison of the differant introns that bind and are spliced by the CYT-18 protein. The introns compared are the Neurospora crassa (Nc) mt large rRNA intron, the Saccharomyces cerevisiae (Sc) mt large rRNA intron (Dujon, 1980; in vitro splicing, G. Mohr, G. Wallweber and A. Lambowitz, unpublished data), the Podospora anserina (Pa) mt large rRNA intron 1 (Heinen, 1991), the bacteriophage T4 td intron (Chu et al., 1984; Mohr et al., 1992), the Neurospora crassa ND1 intron (Burger and Werner, 1985; in vitro splicing, G. Wallweber and A. Lambowitz, unpublished data), and the Tetrahymena (Tt) large rRNA intron derivative lacking P5a,b,c (Joyce et al., 1989; in vitro splicing, see Chapter VI). Nucleotides conserved in all introns are indicated in the structure. "R" and "YM indicate conserved purine or pyrimidine, respectively. "N1' indicates any nucleotide found at a position. For some introns, the published structures of the P6 /P6a junction region has been modified to maximize similarity.

154 155

Kd (nM)

Nc. LSU 6 o Nc. ND1 16 P5 Sc. LSU 11 Pa. LSU-1 36 A T4 . td 16 R A Tt. LSU-AP5abc 9 R-Y

P4 P3 P6 AR pga 5 /_. -A— R— A YY y r r ------Y Y A AUCAGNN R-GR-AUAGUC NR— Y-R----- 3 ' P7 P9 (

Figure 30. Figure 31. comparison of the three-dimensional structures of group I introns and tRNA. (A) Three-dimensional model of the Tetrahymena large rRNA intron from Michel and Westhof (1990). G1 and G414 refer to the 5'- and 3'-nts of the intron. (B) Same as (A) , except that regions not required for binding CYT-18 protein have been deleted and the segment corresponding to P6a has been closed by a loop. (C). Three- dimensional structure of yeast tRNAPhe adapted from Freifelder (1987).

156 P9 TwC loop P 5 — 3 P90 P 9.0 CCA G414 terminus

P4 P7

P6

P6a P6a

Anticodon loop

B

Figure 31. 157 Figure 32. Comparison of the Neurospora mt tRNA1*1, and the Neurospora mt large rRNA intron core. The structure of Neurospora mt tRNATyr is from Heckman et al. (1979) . The intron structure is drawn as a cloverleaf to facilitate comparison to the tRNA1*'. Circled nucleotides are identical

in the two structures. Shaded nucleotides are those conserved in the six group I introns that bind and are spliced by CYT-18 (see Figure 30).

158 Neurospora mt tRNA^ Neurospora mt LSU Intron

A @ G@ ™ <8? °VV“9? ® A ° ® ( U ) U © A U l T A A a A © ® ® I I I I I I I I I I GG ^%>G©^ I I I I ©®A)G-LU" 6 t ^ C ‘ ^®A{g)U A A | U U U © A A ... [J G _ ® C @ © ° ® U® P5 P4 D a ® m?G / ^ 4» — (A) u u -® A — U A — U G — C ® U C ® (A) A U ® u — a P6a G u ®

Figure 32. CHAPTER VI REPLACEMENT OF THE LARGE EXTENSION OF P5 ELEMENT IN TETRAHYMENA LARGE rRNA INTRON BY THE NEUROSPORA MT TYROSYL tRNA SYNTHETASE

Introduction

In previous chapters, I showed that the CYT-18 protein can bind to a number of group I introns from several different organisms. These introns belong to different subgroups and include self-splicing introns as well as non­ self-splicing introns. The major binding site of the CYT-18 protein on the Neurospora mt large rRNA intron is in the P4- P5-P6 regions of the group I intron core. The potential low affinity binding site is in the P7-P9 region. Mohr et al. (1992) have reported recently that the CYT-18 protein can function in splicing defective introns in the bacteriophage T4 td gene and yeast a>+ gene. These results suggest that the

CYT-18 protein may promote splicing by stabilizing the catalytic core structure required for catalysis. It has been speculated that the protein-dependent introns evolved from self-splicing introns by a process in which the structurally and/or catalytically important RNA

160 161 elements were gradually lost and replaced by trans-acting RNA elements and protein factors (Padgett et al., 1986). To understand the connections between the self-splicing introns and protein-dependent introns, I studied the function of the CYT-18 protein in the splicing of a derivative of the self­ splicing Tetrahymena large rRNA intron and found a functional similarity between the CYT-18 protein and a structurally important RNA element of the Tetrahymena large rRNA intron.

Results

Nitrocellulose Filter Binding of the CYT-18 Protein to the Tetrahymena Large rRNA Intron or its PSabc Deletion Mutant Although the CYT-18 protein can bind to a number of group I introns from Neurospora, Podospora, yeast, and bacteriophage, it is not able to bind to the Tetrahymena large rRNA intron at relative high concentrations (up to 188 nM; see chapter V) . Compared to the six introns that can both bind and be spliced by the CYT-18 protein (see Chapter V), the Tetrahymena large rRNA intron has an extra large extension of P5 stem, i.e., PSabc stems (Figure 33). Our previous studies have shown that the major binding site of the CYT-18 protein is in the P4-P5-P6 region of the catalytic core of the Neurospora mt large rRNA intron, which is structurally conserved in all group I introns. Accordingly, we speculate that the PSabc stems in the Tetrahymena large rRNA intron may obscure the binding of the CYT-18 protein to 162 the P4-P5-P6 regions of the intron. To test the possibility that the PSabc element obscures the binding of the CYT-18 protein, I examined the nitrocellulose filter binding of the CYT-18 protein to a P5abc deletion mutant of the Tetrahymena intron (APSabc), as shown in Figure 34. The wild-type Tetrahymena large rRNA intron cannot bind to the CYT-18 protein at concentrations up to 188 nM. However, the P5abc deletion mutant of the Tetrahymena intron (APSabc) can bind to the CYT-18 protein strongly, with a Kd of 9 nM, which is about the same as that of Neurospora mt large rRNA intron. This finding indicates that the CYT-18 protein binding site exists in the Tetrahymena large rRNA intron but is obscured by the PSabc stems.

The CYT-18 Protein Functions in Splicing of the APSabc Mutant of the Tetrahymena Large rRNA Intron It has been reported that the PSabc deletion mutant of the Tetrahymena large rRNA intron is defective in splicing under standard conditions for the intact intron (5 mM

MgCl2/200 mM NH4Cl, pH7.5), but it can self-splice at higher magnesium concentrations (15 mM MgCl2/200 mM NH4Cl/2 mM spermidine, pH 7.5; Joyce et al., 1989). Recent studies have shown that this PSabc deletion mutant intron can be spliced under standard conditions when the PSabc RNA is added in trans (van der Horst et al., 1991). Since the CYT-18 protein 163 can bind strongly to AP5abc mutant of the Tetrahymena large rRNA intron, I tested whether this mutant intron derivative can also be spliced by the CYT-18 protein in the absence of the P5abc RNA, as shown in Figure 35. The autoradiogram confirms that the AP5abc intron does not splice under the standard conditions (5 mM MgCl2/200 mM

NH4C1, pH 7.5), in the absence of the P5abc RNA or the CYT-18 protein (Figure 35A, lanes 2 and 5, controls) . In the presence of P5abc RNA, the splicing of AP5abc occurs and produces bands corresponding to the intron and intron-3'exon (Figure 35A, lanes 4 and 7: L-IVS and IVS-3'exon). In the splicing reaction using the 32P-UTP labeled transcripts, a band corresponding to ligated exons is also observed (lane 4, 5'exon-3'exon) . The splicing of the AP5abc RNA can also occur in the presence of the CYT-18 protein (lanes 3 and 6 ). However, the mutant CYT-18 protein A22 (aa 38-59 deletion) or 604 (3' truncation to aa 452) cannot splice the AP5abc intron (data not shown; proteins provided kindly by Dr. G. Mohr). The heat-treated CYT-18 protein (100°C, 15 min) also fails to function in splicing of AP5abc intron (not shown). Despite the similarity between the CYT-18-rescued and P5abc-rescued splicing reactions of AP5abc intron, the CYT-18 protein and P5abc have some differences in their abilities to catalyze some additional reactions instead of splicing. In the P5abc-mediated reaction as shown in lane 4 of Figure 35, three major extra bands can be observed. The band at the top 164 of the gel (C-IVS) is the product of a cyclization reaction of the excised linear intron at the site 15 nucleotides from the 5' end of the intron (Zaug et al., 1983). A second band (5'exon-IVS), below the precursor band, is the product of a group I intron-mediated specific hydrolysis at the 3' splice- site Zaug et al., 1985). The third band, L15IVS, is below the intron band (L-15IVS) and results from the reopening of the circular intron RNA (Sullivan and Cech, 1985). The pattern and ratios of different bands are similar to those of wild-type Tetrahymena intron with the P5abc RNA present in trans (Figure 35B). However, these products, resulting from the group I intron-specific side-reactions, were not seen in the CYT-18-rescued reactions. This indicates that the CYT-18 can rescue the activity of 5' splice site cleavage, 3' splice site cleavage, and ligation of the two exons, but not any other activities of the intron. I also examined the splicing of AP5abc intron in the presence of either P5abc RNA or CYT-18 for increasing periods of time (not shown). In both cases, the splicing of AP5abc became saturated in 60 min. The other products, which are due to the cyclization, hydrolysis at 3' splice site, and reopening of the circular form, were not observed in the CYT- 18-promoted splicing of AP5abc intron for up to 3 hours. The splicing of the wild-type Tetrahymena large rRNA intron was not affected by the CYT-18 protein, but the formation of the circular intron was inhibited by the P5abc 165 RNA present in trans (Figure 35B). This effect of the P5abc RNA on the formation of the circular intron was observed in two repeated experiments. This result is different from that of van der Horst et al. (1991), in which they demonstrated that the P5abc RNA has no effect to the splicing of the wild- type Tetrahymena large rRNA intron. When the P5abc RNA was provided in trans in the splicing reactions of the CYT-18-dependent Neurospora mt large rRNA intron and ND1 intron, neither intron can splice under the standard conditions (5 mM MgCl2/20Q mM NH4C1, pH 7.5). These results suggest that either the deficiencies in the Neurospora mt large rRNA intron and ND1 intron for their lack of the self-splicing activity are not simply the same as those in the Tetrahymena AP5abc intron or the P5abc RNA cannot interact with the Neurospora mt large rRNA intron and ND1 intron. Since the Tetrahymena AP5abc intron can still splice at high magnesium concentration while the Neurospora mt large rRNA intron and ND1 intron cannot self-splice under any conditions tested, the Tetrahymena APSabc intron may have a more easily repairable structure.

Construction of Small RNA Containing the CYT-18 Binding Site of the Tetrahymena Intron The major binding site of the CYT-18 protein on the Neurospora mt large rRNA intron is in the regions from P3 to J6/7. To identify the CYT-18 protein-binding site on the 166 Teyrahymena large rRNA intron, I constructed a plasmid (pTHIGHX) which contains the AP5abc intron sequence from P3 to J6/7 (see Chapter II for the construction of the plasmid). The RNA transcribed from this plasmid was tested for its binding to the CYT-18 protein. Under the standard filter binding conditions (5 mM MgCl2/100 mM KC1/20 mM Tris-HCl, pH7.5), this RNA cannot bind to the CYT-18 protein (up to 165 nM) . When the MgCl2 concentration was increased to 10 mM, 20 mM, and 50 mM, the RNA still fails to bind the CYT-18 protein (data not shown).

Discussion

Here I showed that the CYT-18 protein can bind to the P5abc deletion mutant of the Tetrahymena large rRNA intron and rescue its splicing activity. First, these results support our finding that CYT-18 protein bind to the conserved catalytic core structure of the group I introns. Second, these results suggest that one function of the CYT-18 protein may be the same as the P5abc RNA. Finally, these findings show that a protein can substitute functionally for an RNA structure in a self-splicing intron. Our previous studies have shown that the major binding site of the CYT-18 protein on the Neurospora mt large rRNA intron is in the P4-P5-P6 regions and the potential low affinity binding site is in the P7-P9 regions (see chapter IV). In addition to the Neurospora mt large rRNA intron, a 167 number of other group I introns from several organisms can also bind to the CYT-18 protein. Comparison of these introns shows little sequence homology and suggests that the CYT-18 protein binds to the conserved structural features of these group I introns. Accordingly, it is very likely that the major binding site of the CYT-18 protein in the Tetrahymena large rRNA intron is also in the P4-P5-P6 regions of the intron. We found, however, that the wild-type Tetrahymena large rRNA intron cannot bind to the CYT-18 protein. Several observations suggest that this binding site of the CYT-18 protein is obscured by P5abc stems in the wild-type intron: (1) The six group I introns, which can bind and be spliced by the CYT-18 protein, have no P5abc stems or other kind of large extensions. (2) In the structural studies of Heuer et al. (1991), 8/13 nucleotides in P4 , 9/12 nucleotides in P5, and 4/4 nucleotides in P6 are protected from Fe(II) -EDTA cleavage in the wild-type Tetrahymena large rRNA intron. However, in phage T4 introns of sunY and td genes which are both group IA introns and have no extended P5 stem, only 1/12 nucleotides in P4, 2/6 (td) or 3/8 (sunY) nucleotides in P5, and 2/4 nucleotides in P6 are protected from Fe(II) -EDTA cleavage. The extra protections in the wild-type Tetrahymena intron are probably due to the Existence of the P5abc element. This possibility is also supported by the finding that A183 and adjacent portions of P5a and the core regions are 168 all located inside the tertiary structure of the intron. (3)

Recent studies suggest that A 183 of P5a may interact with G212 of P4 (Pyle et al., 1990). In this chapter, I showed that the AP5abc mutant of the Tetrahymena large rRNA intron can bind strongly to the CYT-18 protein, although the wild-type intron cannot bind. This indicates that the CYT-18 binding site is hindered by the P5abc element and suggests that the CYT-18 protein and P5abc RNA probably bind to the same region in the intron core. However, the small RNA containing this region of the Tetrahymena large rRNA intron cannot bind to the CYT-18 protein when tested by nitrocellulose filter binding assays. A possible interpretation is that the correct structure in this region cannot be formed by itself and has to be stablized by the elements outside this region in the Tetrahymena large rRNA intron. The P 6 and P6a stems have 2 and 3 bp in the Tetrahymena intron, but they have 4 and 8 bp in the Neurospora mt large rRNA intron. This instability of the P6 and P6a in the Tetrahymena intron may be the cause of decreased binding to the CYT-18 protein. Answering how the CYT-18 protein functions in the splicing of group I introns is central toward understanding of CYT-18-dependent and other protein-dependent splicing mechanisms. The results presented in chapter V showed that the CYT-18 protein binds to the core regions of the group I intron and suggest that the CYT-18 protein probably promotes splicing of group I introns by stabilizing its catalytically 169 active structure. The results shown in this chapter support this assertion. Since the P5abc RNA and the CYT-18 protein can replace each other for promoting the splicing of the Tetrahymena intron, one of the functions of the CYT-18 protein may be the same as the P5abc RNA, which binds to the Tetrahymena intron and may stabilize its catalytic RNA structure. For those introns which are able to self-splice but do not have the PSabc stems, the catalytic structure might be stabilized in some other ways. In fact, the interaction between the P7.2 and P9.2 in the phage sunY intron has been confirmed recently (Michel et al., 1992). This interaction may serve to stablize the catalytic core of the sunY intron. Here, I showed that the essential RNA element P5abc of the Tetrahymena intron for the self-splicing activity under low magnesium condition can be replaced by the CYT-18 protein. Theoretically, this kind of replacement could also occur during evolution. If self-splicing introns are ancient (see discussion of Chapter V) , it is likely that the RNA elements in the self-splicing group I and group II introns are gradually replaced by some protein factors to evolve into the protein-dependent introns, and finally lead to present- day nuclear mRNA introns. During this evolutionary process, the essential RNA elements in the introns may develop to function in the following three ways. (1) Some may be kept for certain advantages of being in the intron. For example. 170 the core structure of the group I introns survived billions of years of natural selection and is still kept in the intron. The conserved branch site and junction sequences in the nuclear mRNA introns may also be such examples. (2) Some may develop to function in trans or could function in trans. The small nuclear RNAs in the spliceosome of the nuclear mRNA intron may be the examples of this group. The P5abc element is a good example that can function both in cis and in trans. (3) Some may be replaced by protein factors. Although the CYT-18 protein and P5abc RNA can replace each other for rescuing the splicing activity of the Tetrahymena large rRNA intron, they have some differences in their abilities to catalyze reactions other than splicing. In addition to the normal splicing activity, P5abc RNA also promotes additional group I intron-specific reactions (cyclization of excised intron, hydrolysis at 3' splice-site, and hydrolysis of the circular intron). However, these reactions are suppressed by the CYT-18 protein during the splicing of AP5abc intron. The products resulting from the group I intron-specific side-reactions are not found in the CYT-18-mediated splicing reactions of the Neurospora mt large rRNA intron, ND1 intron (Guo et al., 1991; Wallweber and Lambowitz, unpubl.). There are three possible explanations for the suppression of the side-reactions by the CYT-18 protein, (i) The sites required for these side-reactions are obscured by the binding of the CYT-18 protein, or the CYT-18 171 protein increases the ability of the intron to discriminate against the mismatched substrates in the cyclization reaction. (ii) The structures required for the normal splicing activity and side-reactions are different and only the former is stabilized by CYT-18. (iii) The PSabc RNA itself is involved directly in the catalysis of the side- reactions. Except that the cyclization side-reaction may serve to drive the splicing reaction forward, the circular form of the intron or the other products of the side-reactions seem to have no biological functions in Tetrahymena (Zaug et al., 1983). If these side-reactions do not have any functions, they may not need to be kept in the intron. This may be the reason that the CYT-18 or other protein factors are adapted to replace only those necessary functions of the RNA elements in the self-splicing group I introns during evolution. The structure of the intron/CYT-18 complex may be better than the original self-splicing group I introns since the unwanted side-reactions are minimized in the CYT-18-mediated splicings. The only possible function of the side-reactions, i.e. driving the forward splicing reaction, could be carried out in some other ways. For example, the splicing products may be degraded faster. Figure 33. The predicted secondary structure of the APSabc derivative of the Tetrahymena large rRNA intron. The structure is modified from Burke et al. (1987). Uppercase letters indicate intron sequences; lowercase letters indicate exon sequences. Arrows indicate 5' and 3' splice-sites, respectively. The P5abc structure including P5a, P5b, P5c, L5b and L5c is indicated in the figure.

172 173

Tt large rRNA intron PSabc

L1 ft c , u n „ u - a a " uu u ft-U c U U G„ n c r. ft ft C C-G 2 ? c a c-G u c o-c n j k j k p2 t ™ -#-13c-G ft-U U-ft P’ IH M pj g P6 P6a P6b ®* c-G-ft-A-A'A-U— ft- t^CAftGflCCTUCA A AU U - A ^ CUGGduGU^ ^ ^ f tc g lin c iu ^ y n C U U G U C y U C n jh o u HGflftGCO LS “ni.LoS^ftUftftOftUftSSS^GftCCnuUCUUClXAUAAOAUAUAGUC-GGACC-G C-G— O— ft-UJUGGAGUACUCGuoog 3’ n u __ U-ft U-0 ft-U A PS c-G ft-U C-G T U I G-C U-ft P9 gig g^P91 £ 3 P9 2 u U G-C U-ft U n ft A G U 0 u A U ft U U (a G G G U *•* ft G U-fl t£c £3 P9.24 P9.1a u-fl G-U G-C C-G A A C ft U ft ft A G C L9.1* L9 .2

Figure 33. Figure 34. The nitrocellulose filter binding of the CYT-18 protein to the Tetrsthymena large rRNA intron or its PSabc deletion mutant (APSabe). The percentage of 32P-labeled RNA retained on the filter is plotted as a function of concentration of the CYT-18 protein dimer (nM). The Tetrahymena large rRNA intron RNA and AP5abc RNA were transcribed from plasmids pT7TTlA3 and pT7TTlA3AP5abc linearized with EcoRI using phage T7 RNA polymerase. The Neurospora mt large rRNA intron RNA was transcribed from pBD5A plasmid linearized with BanI using phage T3 RNA polymerase. The nitrocellulose filter binding experiments were carried out using the standard procedure as described in Chapter II. The data are averages of duplicate samples. The initial nitrocellulose binding experiments using plasmid pDJD173c, which contains AP5abc deletion in Tetrahymena L21 ribozyme (see chapter II) , were carried out together with Lingyun Zhao (Kd = 9 nM; data not shown) .

174 RNA Bound (percent) 20 40 60 SO 0 20 Y-8 rti (nM) Protein CYT-18 Figure 34. Figure 60 8040 0 120 100 Tt large rRNA intron rRNA large Tt cm ag RA intron rRNA large mt Nc Tt A intron Tt PSabc 140 160 175 Figure 35. The CYT-18 protein can splice the PSabc deletion mutant of the Tetrahymena large rRNA intron (APSabe). (A) The splicing of the AP5abc intron in the presence of the P5abc RNA or the CYT-18 protein. Lanes 1-4: splicing assayed using the 32P-UTP labeled precursor RNA. Lanes 5-7: splicing assayed using the precursor intron RNA and 32P-GTP. Lane 1:

32P-UTP labeled AP5abc precursor RNA only; lane 2 and 5: AP5abc precursor RNA incubated under the standard splicing conditions; lanes 3 and 6 : AP5abc precursor RNA under the standard splicing conditions in the presence of 6.6 fiM E . coli expressed CYT-18 protein; lanes 4 and 7: AP5abc precursor RNA under the standard splicing conditions in the presence of 5 /aM P5abc RNA. (B) The splicing of the wild- type Tetrahymena large rRNA intron in the presence of the PSabc RNA or the CYT-18 protein. The splicing reactions were carried out in the buffer containing 5 mM MgCl2, 200 mM NH4C1,

50 mM Tris-HCl (pH 7.5) at 30°C for 2 hours. For splicing using 32P-UTP labeled transcript, 81,000 cpm of AP5abc precursor RNA (specific activity: 1 x 106 cpm//Ltg) and 0.2 mM GTP were used in each reaction. For splicing assayed using intron transcript and 32P-GTP, l fig of AP5abc precursor RNA and 40 fiCi [a-32]GTP (3,000 Ci/mmole) were used in a 40 /ul reaction. The APSabe precursor RNA is the same as that

176 177 described by van der Horst et al. (1991), but transcribed from a different plasmid pT7TTlA3AP5abc (see chapter II for description). The P5abc RNA was transcribed using T7 RNA polymerase from plasmid pP5abcX linearized with Xbal. The plasmid pP5abc is from Tan Inoue (The Salk Institute, La Jolla, CA) and has been described by van der Horst et al. (1991). Abbreviations: L-IVS, linear intron; C-IVS, circular intron without the first 15 nucleotides; 5'exon-IVS, 5' exon plus intron; IVS-3'exon, intron plus 3' exon; pre-RNA, precursor transcript of plasmid pT7TTlA3AP5abc/EcoRI. -»C o « ® 2 r i ® i 9 < < < O z < CO CO CO 3 > CO Mill I Iran script control + CYT-18 I + PSabc control 0 I + CYT-18 0 I +P5abc i.

0 ~ transcript s ■ . 0 control ? • 0p + PSabc f 0 I 0 + CYT-18 > ^1 K k ! T T «! < -o o

- CO O 0 9 2 5 < “ it > CO 3 178 CHAPTER VII CONCLUSIONS

Some group I introns have been shown to self-splice in vitro, but most require proteins to splice efficiently in vivo. The Neurospora mt large rRNA intron requires the CYT- 18, a mitochondrial tyrosyl-tRNA synthetase, to splice both in vitro and in vivo. My dissertation research focused on the interaction between the group I introns and the CYT-18 protein and aimed to understand the mechanism of this protein- dependent splicing reaction. To facilitate the identification of the protein binding site on the Neurospora mt large rRNA intron, I analyzed in vitro mutants of the intron and constructed a 388 bp mini- intron from the 2.3 kb of original intron. This mini-intron retains conserved sequences and secondary structure of group I intron and is still spliced in a protein-dependent manner, as the wild-type intron, but further modifications that affect different regions of the intron core abolish splicing, suggesting that the conserved RNA structure is required for catalysis and cannot be substituted by the CYT-18 protein. Using the nitrocellulose filter binding, UV-crosslinking, and in vitro splicing methods, I found that the CYT-18 protein

179 180 can bind to the Neurospora mt large rRNA intron. By testing truncation or deletion mutants of the intron for their ability to bind to the CYT-18 protein and to compete for filter binding and splicing of the wild-type intron, the major binding site of the CYT-18 protein was localized to a 75 nt region in P4 and P6 stems of the intron. By itself, this RNA binds to the CYT-18 protein with Kd of 13 nM, compared to 6 nM for the wild-type intron. The rest of the binding ability may be contributed by another 40 nt region in P7 and P9 stems of the intron. The CYT-18 protein can also bind to a number of other group I introns from Neurospora, Podospora, yeast, and bacteriophage. These introns have little overall sequence homology, suggesting that the CYT-18 protein interacts with highly conserved structural features of the group I intron core. My results suggest that the Neurospora mt large rRNA intron and tRNATyr may resemble each other. First, they compete for the same or overlapping binding sites, as judged by the findings that the mt large rRNA intron is a competitive inhibitor of aminoacylation and Neurospora mt tRNATyr inhibits splicing. Second, the stoichiometry for the binding of the CYT-18 to the rRNA intron is the same as that found in the bacterial system, where Bacillus tyrosyl-tRNA is believed to bind across the surface of the two subunits of the synthetase. Third, the comparison of the binding regions of the large rRNA intron with the Neurospora mt tRNATyr revealed some 181 similarities in primary sequence, secondary and tertiary structures. These results suggest that the CYT-18 protein binds to the intron in a similar way as it binds to the tRNA1*1. The structural similarity between the group I intron core and tRNA raises the possibility that tRNA may have structural features that resemble and could have evolved from a catalytic RNA, which was the ancestor of group I introns. Although the CYT-18 protein cannot bind to the wild-type Tetrahymena large rRNA intron, it can bind to a derivative of this intron in which the PSabc stems of the intron have been deleted. The APSabe derivative of the Tetrahymena large rRNA intron is defective in self-splicing at low MgCl2 concentration, but its splicing activity can be rescued by PSabc RNA in trans. Like PSabc, the CYT-18 protein can also promote the splicing of APSabe derivative of the Tetrahymena intron. This indicates that an important RNA sequence element in a self-splicing intron can be replaced by a protein. The finding supports the hypothesis that self-splicing introns may evolve into protein-dependent introns, such as comtemporatory nuclear mRNA introns. In addition, this finding suggests that the CYT-18 protein may function similarly to PSabc in binding to and stabilizing the catalytic structure of the group I intron. The above studies have lead to a better understanding of the splicing mechanism of group I introns promoted by the CYT- 18 protein. However, many aspects of how the CYT-18 functions 182 in splicing of group I introns remain to be investigated. Further studies, including the X-ray structure of the complex formed by CYT-18 and group I introns or CYT-18 and tyrosyl- tRNA, should provide further insight into the action of the CYT-18 protein on the splicing of group I introns. References

Akins, R.A. and Lambowitz, A.M. (1987). A protein required for splicing group I introns in Neurospora mitochondria is mitochondrial tyrosyl-tRNA synthetase or a derivative thereof. Cell 50, 331-345. Barker, D.G., Bruton, C.J. and Winter, G. (1982). The tyrosyl-tRNA synthetase from Escherichia coli. Complete nucleotide sequence of the structural gene. FEBS Lett. 150, 419-423. Bedouelle, H. (1990). Recognition of tRNA7^ by tyrosyl-tRNA synthetase. Biochimie 72, 589-598. Been, M.D. and Cech, T.R. (1988). RNA as an RNA polymerase: net elongation of an RNA primer catalyzed by the Tetrahymena ribozyme. Science 239, 1412-1416. Belfort, M. , Chandry, P.S., and Pedersen-Lane, J. (1987). Genetic delineation of functional components of the group I intron in the phage T4 td gene. Cold Spring Harbor Symp. Quant. Biol. 52, 181-192. Berget, S.M., Moore, C. and Sharp, P.A. (1977). Spliced segments at the 5' terminus of adenovirus 2 late mRNA. Proc. Natl. Acad. Sci. USA 74, 3171-3175. Bernard, U., Goldthwaite, C. and Ktintzel, H. (1976). Physical map of Neurospora crassa mitochondrial DNA and its transcription. Nucleic. Acids Res. 3, 3101-3108. Bertrand, H., Bridge, P., Collins, R.A., Garriga, G. and Lambowitz, A.M. (1982). RNA splicing in Neurospora mitochondria. Characterization of new nuclear mutants with defects in splicing the mitochondrial large rRNA. Cell 29, 517-526. Bordonn&, R., Bandlow, W., Dirheimer, G., and Martin, R.P. (1987) . A single base change in the extra-arm of yeast mitochondrial tyrosine tRNA affects its conformational stability and impairs aminoacylation. Mol. Gen. Genet. 206, 498-504.

183 184 Burger, G. and Werner, S. (1985). The mitochondrial URF1 gene in Neurospora crassa has an intron that contains a novel type O f URF. J. Mol. Biol. 186, 231-242. Burke, J.M. (1983). Structure of intervening sequences within mitochondrial genes of Neurospora crassa and the identification of an intron sequence implicated in RNA splicing, Ph.D. thesis, Massachusetts Institute of Technology, Cambridge, MA. Burke, J.M. (1988). Molecular genetics of group I introns: RNA structure and protein factors required for splicing - a review. Gene 73, 273-294. Burke, J.M. (1989). Selection of the 3' splice site in group I introns. FEBS Lett. 250, 129-133. Burke, J.M. and RajBhandary, U.L. (1982). Introns within the large rRNA gene of N. crassa mitochondria: a large open reading frame and consensus sequence possibly important in splicing. Cell 31, 509-520. Burke, J.M., Esherick, J.S., Burfeind, W.R. and King, J.L. (1990). A 3' splice site-binding sequence in the catalytic core of a group I intron. Nature 334, 80-82. Burke, J.M., Belfort, M., Cech, T.R., Davies, R.W., Schweyen, R.J., Shub, D.A., Szostak, J.W. and Tabak, H.F. (1987). Structural conventions for group I introns. Nucl. Acids Res. 15, 7217-7221. Carey, J., Cameron, V., de Haseth, P.L. and Uhlenbeck, O.C. (1983). Sequence-Specific Interaction of R17 Coat Protein with Its Ribonucleic Acid Binding Site. Biochem. 22, 2601- 2610. Cech, T.R. (1986) . The generality of self-splicing RNA: Relationship to nuclear mRNA splicing. Cell 44, 207-210. Cech, T.R. (1988). Conserved sequences and structures of group I introns: building an active site for RNA catalysis - a review. Gene 73, 259-271. Cech, T.R. (1990). Self-splicing of group I introns. Ann. Rev. Biochem. 59, 543-568. Cech, T.R. and Bass, B.L. (1986). Biological catalysis by RNA. Ann. Rev. Biochem. 55, 599-629. Cech, T.R., Zaug, A.J. and Grabowski, P.J. (1981). In vitro splicing of the ribosomal RNA precursor of Tetrahymena: Involvement of a guanosine nucleotide in the excision of 185 the intervening sequence. Cell 27, 487-496. Cherniack, A.D. (1991). Involvement of mitochondrial tyrosyl tRNA synthetase in splicing: identification of an N- terminal domain that functions in splicing. Ph.D. Thesis, The Ohio State University. Cherniack, A.D., Garriga, G., Kittle, J.D. Jr., Akins, R.A. and Lambowitz, A.M. (1990). Function of Neurospora mitochondrial tyrosyl-tRNA synthetase in RNA splicing requires an idiosyncratic domain not found in other synthetases. Cell 62, 745-755. Chow, L.T., Gleinas, R.E., Broker, T.R. and Roberts, R.J. (1977). An amazing sequence arrangement at the 5' ends of adennovirus 2 messenger RNA. Cell 12, 1-8. Christopher, D.A. and Hallick, R.B. (1989) . Euglena gracillis chloroplast ribosomal protein operon: a new chloroplast gene for ribosomal protein L5 and description of a novel organelle intron categary designated group III. Nucleic Acids Res. 17, 7591-7608. Chu, F.K., Maley, G.F., Maley, F. and Belfort, M. (1984). An intervening sequence in the thymidylate synthase gene of bacteriophage T4. Proc. Natl. Acad. Sci. USA 81, 3049- 3053. Collins, R.A. and Lambowitz, A.M. (1985) . RNA splicing in Neurospora mitochondria. Defective splicing of mitochondrial mRNA precursors in the nuclear mutant cytl8- 1. J. Mol. Biol. 184, 413-428. Collins, R.A. (1990). Neurospora crassa laboratory strain 74- OR23-1A: mitochondrial genes, in Genetic Maps: locus maps od complex genomes. O'Brien, S.J. (ed). Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. Cummings, D.J., Domenico, J.M. and Nelson, J. (1989). DNA sequence and secondary structure of the large subunit rRNA coding regions and its two class I intron of mitochondrial DNA from Podospora anserina. J. Mol. Evol. 28, 242-255. Darnell, J.E., and Doolittle, W.F. (1986). Speculations on the early course of evolution. Proc. Natl. Acad. Sci. USA 83, 1271-1275. Davis, R.H. and de Serres, F.J. (1970). Genetic and microbiological research techniques for Neurospora crassa. Method Enzymol. 17A, 79-143. Dib-Hajj, S.D. (1990). In vitro studies of domains 1, 3 and 186 5 of self-splicing group II introns of yeast mtDNA. Ph.D. Thesis, The Ohio State University. Dixon, M . , and Webb, E. C. (1979). Enzymes, 3rd ed. , p346- 347, Academic Press, New York. Dobinson, K.F., Henderson, M., Kelly, R.L., Collins, R.A. and Lambowitz, A.M. (1989). Mutations in nuclear gene cyt-4 of Neurospora crassa result in pleiotropic defects in processing and splicing of mitochondrial RNAs. Genetics 123, 97-108. Dock-Bregeon, A.C., Westhof, E., Giege, R. and Moras, D. (1989). Solution structure of a tRNA with a large variable region: yeast tRNA8". J . Mol. Biol. 206, 707-722. Dock-Bregeon, A.C., Garcia, A., Giege, R. and Moras, D. (1990). The contacts of yeast tRNA8" with seryl-tRNA synthetase studied by footprinting experiments. Eur. J. Biochem. 188, 283-290. Doudna, J.A., Cormack, B.P. and Szostak, J.W. (1989). RNA structure, not sequence, determines the 5' splice-site specificity of a group I intron. Proc. Natl. Acad. Sci., U.S.A. 8 6 , 7402-7406. Dujon, B. (1980). Sequence of the intron and flanking exons of the mitochondrial 2IS rRNA gene of yeast strains having different alleles at the u> and rib-1 loci. Cell 20, 185- 197. Dujon, B. (1989). Group I introns as mobile genetic elements: facts and mechanistic speculations - a review. Gene 82, 91-114. Filipowicz, W. and Shatkin, A. J. (1983) . Origin of splice junction phosphate in tRNAs processed by Hela cell extract. Cell 32, 547-557. Freifelder, D. (1987). Molecular Biology. Second edition. Jones and Bartlett Publishers, Inc., Boston. Gampel, A. and Cech, T.R. (1991) . Binding of the CBP2 protein to a yeast mitochondrial group I intron requires the catalytic core of the RNA. Genes Dev. 5, 1870-1880. Gample, A., Nishikimi, M. and Tzagoloff, A. (1989). CBP2 protein promotes in vitro excision of a yeast mitochondrial group I intron. Mol. Cell. Biol., 9, 5424- 5433. 187 Garriga, G. and Lambowitz, A.M. (1983). RNA splicing in Neurospora mitochondria. The large rRNA intron contains a noncoded, 5'-terminal guanosine residue. J. Biol. Chem. 258, 14745-14748. Garriga, G., Bertrand, H. and Lambowitz, A.M. (1984). RNA splicing in Neurospora mitochondria: nuclear mutants defective in both splicing and 3' end synthesis of the large rRNA. Cell 36, 623-634. Garriga, G. and Lambowitz, A.M. (1984). RNA splicing in Neurospora mitochondria: self-splicing of a mitochondrial intron in vitro. Cell 39, 631-641. Garriga, G. and Lambowitz, A.M. (1986). Protein-dependent splicing of a group I intron in ribonucleoprotein particles and soluble fractions. Cell 46, 669-680. Gott, J.M., Shub, D.A., Belfort, M. (1986). Multiple self­ splicing introns in bacteriophage T4: evidence from auto- catalytic GTP-labeling of RNA in vitro. Cell 47, 81-87. Greer, C.L., Peebles, C.L., Gegenheimer, P. and Abelson, J. (1983). Mechanism of action of a yeast RNA ligase in tRNA splicing. Cell 32 537-546. Grimm, M.F., Cole, M.D. and Lambowitz, A.M. (1981). Ribonucleic acid splicing in Neurospora mitochondria: secondary structure of the 35S ribosomal precursor ribonucleic acid investigated by digestion with ribonuclease II and by electron microscopy. Biochemistry 20, 2836-2842.

Guo, Q., Akins, R.A., Garriga, G. and Lambowitz, A.M. (1991). Structural analysis of the Neurospora mitochondrial large rRNA intron and construction of a mini-intron that shows protein-dependent splicing. J . Biol. Chem. 266, 1809-1819. Heckman, J.E. and RajBhandary, U.L. (1979). Organization of tRNA and rRNA genes in N. crassa mitochondria: intervening sequence in the large rRNA gene and strand distribution of the RNA genes. Cell 17, 583-595. Heckman, J.E., Alzner-Deweerd, B. and RajBhandary, U.L. (1979). Interesting and unusual features in the sequence of Neurospora crassa mitochondrial tyrosine transfer RNA. Proc. Natl. Acad. Sci. USA 76, 717-721. Heinen, U. (1991). Mitochondriale genexpression bei pilzen: Molekulare analysen zur nukleo-zytoplasmatischen wechselwirkung. Dissertationes Botanicae 166, J. Cramer, Berlin. 188 Heuer, T.S., Chandry, P.S., Belfort, M., Celander, D.W. and Cech, T.R. (1991). Folding of group I introns from bacteriophage T4 involves internationalization of the catalytic core. Proc. Natl. Acad. Sci. USA 8 8 , 11105- 11109. Himeno, H., Hasegawa, T., Ueda, T., Watanabe, K. and Shimizu, M. (1990). Conversion of aminoacylation specificity from tRNA15" to tRNA8" in vitro. Nucleic Acids Res. 18, 6815- 6819. Jacobson, A.B., Good, L., Simonetti, J. and Zuker, M. (1984). Some simple computational methods to improve the folding of large RNAs. Nucleic Acids Res. 12, 45-52. Jaeger, L., Westhof, E. and Michel, F. (1991). Function of Pll, a tertiary base pairing in self-splicing introns of subgroup IA. J. Mol. Biol. 221, 1153-1164. Jarrell, K.A., Dietrich, R.C. and Perlman, P.S. (1988). Group II intron domain 5 facilitates a trans-splicing reaction. Mol. Cell. Biol. 8 , 2361-2366. Joyce, G.F., van der Horst, G. and Inoue, T. (1989). Catalytic activity is retained in the Tetrahymena group I intron despite removal of the large extension of element P5. Nucleic Acids Res. 17, 7879-7889. Kamper, U., Kiick, U., Cherniack, A.D. and Lambowitz, A.M. (1992). The mitochondrial tyrosyl-tRNA synthetase of Podospora anserina is a bifunctional enzyme active in protein synthesis and RNA splicing. Mol. Cell. Biol., 12, 499-511, 1992. Kittle, J.D. Jr., Mohr, G., Gianelos, J., Wang, H. and Lambowitz, A.M. (1991) . The Neurospora mitochondrial tyrosyl-tRNA synthetase is sufficient for group I intron splicing in vitro and uses the carboxy-terminal tRNA- binding domain along with other regions. Genes Dev. 5, 1009-1021. Kjems, J. and Garrett, R.A. (1988). Novel splicing mechanism for the ribosomal RNA intron in the archaebacterium Desulfurococcus mobilis. Cell 54, 693-703. Kreike, J., Schulze, M., Pillar, T., Korte, A. and Rodel, G. (1986). Cloning of a nuclear gene MRS1 involved in the excision of a single group I intron (bI3) from the mitochondrial COB transcript in S. cerevisiae. Curr. Genet. 11, 185-191. 189 Kreike, J., Schulze, M. , Ahne, F. and Lang, B.F. (1987). A yeast nuclear gene, MRS1, involved in mitochondrial RNA splicing: nucleotide sequence and mutational analysis of two overlapping reading frames on opposite strands. EMBO J. 6 , 2123-2129.

Kruger, K., Grabowski, P.J., Zaug, A.J., Sands, J., Gottschling, D.E. and Cech, T.R. (1982). Self-splicing RNA: Autoexcision and autocyclizationof the ribosomal RNA intervening sequence of Tetrahymena. Cell 31, 147-157. Kubelik, A.R., Turcq, B. and Lambowitz, A.M. (1991). The Neurospora crassa cyt-20 gene encodes cytosolic and mitochondrial valyl-tRNA synthetases and may have a second function in addition to protein synthesis. Mol. Cell. Biol. 11, 4022-4035. Kuhsel, M.G., Strickland, R. and Palmer, J.D. (1990). An ancient group I intron shared by eubacteria and chloroplasts. Science 250, 1570-1573. Kuiper, M.T.R. and Lambowitz, A.M. (1988). A novel reverse transcriptase activity associated with mitochondrial plasmids of Neurospora. Cell 55, 693-704. Kunkel, T.A. (1985). Rapid and efficient site-specific mutagenesis without phenotic selection. Proc. Natl. Acad. Sci. U.S.A. 82, 488-492. Kunkel, T.A., Roberts, J.D. and Zakour, R.A. (1987). Rapid and efficient site-specific mutagenesis without phenotypic selection. Methods Enzymol. 154, 367-382. Labouesse, M., Herbert, C.J., Dujardin, G. and Slonimski, P.P. (1987). Three suppressor mutations which cure a mitochondrial RNA maturase deficiency occur at the same codon in the open reading frame of NAM2 gene. EMBO J. 6 , 713-721. Lambowitz, A.M. (1979). Preparation and analysis of mitochondrial ribosomes. Methods Enzymol. 59, 421-433. Lambowitz, A.M., and Perlman, P.S. (1990). Involvement of aminoacyl-tRNA synthetases and other proteins in group I and group II intron splicing. Trends Biol. Sci. 15, 440- 444. Lambowitz, A.M., LaPolla, R.J. and Collins, R.A. (1979). Mitochondrial ribosome assembly in Neurospora: two- dimensional gel electrophoretic analysis of mitochondrial ribosomal proteins. J. Cell Biol. 82, 17-31. 190 Lambowitz, A.M., Akins, R.A., Garriga, G., Henderson, M. , Kubelik, A.R. and Maloney, K.A. (1985). in Achievements and Perspectives of Mitochondrial Research, Volume II: Biogenesis, (Quagliariello, E., Slater, E.C., Palmieri, R., Saccone, C., and Kroon, A.M., eds) pp 237-247, Elsevier Science Publishers, Amsterdam. Lambowitz, A.M. (1989). Infectious introns. Cell 56, 323-326. LaPolla, R.J. and Lambowitz, A.M. (1979). Binding of mitochondrial ribosomal proteins to a mitochondrial ribosomal precursor RNA containing a 2.3-kilobase intron J. Biol. Chem. 254, 11746-11750. Lazowska, J. , Jacq, C. and Slonimski, P.P. (1980). Sequence of introns and flanking exons in wild-type and box3 mutants of cytochrome b reveals an interlaced splicing protein coded by an intron. Cell 22, 3 33-348. Lazowska, J., Claisse, M., Gargouri, A., Kotylak, Z., Spyridakis, A. and Slonimski, P.P. (1989). Protein encoded by the third intron of cytochrome b gene in Saccharomyces cerevisiae is an mRNA maturase. J . Mol. Biol. 205, 275- 289. Majumder, A.L., Akins, R.A., Wilkinson, J.G., Kelley, R.L., Snook, A.J. and Lambowitz, A.M. (1989). Involvement of tyrosyl-tRNA synthetase in splicing of group I introns in Neurospora crassa mitochondria: biochemical and immunochemical analysis of splicing activity. Mol. Cell. Biol. 9, 2089-2104. Mannella, C.A., Collins, R.A., Green, M.R. and Lambowitz, A.M. (1979). Defective splicing of mitochondrial rRNA in cytochrome-deficient nuclear mutants of Neurospora crassa. Proc. Natl. Acad. Sci. USA 76, 2635-2639. McGraw, P. and Tzagoloff, A. (1983). Assembly of the mitochondrial membrane system. Characterization of a yeast nuclear gene involved in the processing of the cytochrome b pre-mRNA. J. Biol. Chem. 258, 9459-9468. Michel, F., Hanna, M., Green, R., Bartel, D.P. and Szostak, J.W. (1989). The guanosine binding site of the Tetrahymena ribosome. Nature 342, 391-395. Michel, F., Netter, P., Xu, M. and Shub, D.A. (1990). Mechanism of 3' splice site selection by the catalytic core of the sunY intron of bacteriophage T4: the role of a novel base-pairing interaction in group I introns. Genes Dev. 4, 777-788. 191 Michel, F., Jaeger, L., Westhof, E., Kuras, R., Tihy, F., Xu, M. and Shub, D.A. (1992) . Activation of the catalytic core of a group intron by a remote 3' splice junction. Genes Dev. 6, 1373-1385. Michel, F. and Westhof, E. (1990). Modelling of the three- dimensional architecture of group I catalytic introns based on comparative sequence analysis. J. Mol. Biol. 216, 585-610. Michel, F., Jaeger, L., Westhof, E., Kuras, R., Tihy, F., Xu, M. and Shub, D.A. (1992) . Activation of the catalytic core of a group I intron by a remote 3' splice junction. Genes Dev. 6 , 1373-1385. Mohr, G., and Lambowitz, A.M. (1991). Integration of a group I intron into a ribosomal RNA sequence promoted by a tyrosyl-tRNA synthetase. Nature 354, 164-167. Mohr, G., Zhang, A., Gianelos, J.A., Belfort, M. and Lambowitz, A.M. (1992). The Neurospora CYT-18 protein suppresses defects in the phage T4 td intron by stabilizing the catalytically active structure of the intron core. Cell 69, 483-494. Mota, E.M. and Collins, R.A. (1988). Independent evolution of structural and coding regions in a Neurospora mitochondrial intron. Nature 332, 654-656. Netzker, R., Kochel, H.G., Basak, N. and Kiintzel, H. (1982). Nucleotide sequence of Aspergillus nidulans mitochondrial genes coding for ATPase subunit 6 , cytochrome oxidase subunit 3, seven unidentified proteins, four tRNAs and L- rRNA. Nucleic Acids Res. 10, 4783-4794. Nirenberg, M. and Leder, P. (1964). RNA codewords and protein synthesis: the effect of trinucleotides upon the binding of rRNA to ribosomes. Science 145, 1399-1407. Noller, H.F. (1991). Drugs and the RNA world. Nature 353, 302-303. Padgett, R.A., Grabowski, P.J., Konarska, M.M., Seiler, S. and Sharp, P.A. (1986). Splicing of messenger RNA precursors. Ann. Rev. Biochem. 55, 1119-1150. Peebles C.L., Gegenheimer, P. and Abelson, J. (1983). Precise excision of intervening sequences from precursor tRNAs by a membrane-associated yeast endonuclease. Cell 32, 525- 536. Peebles, C.L., Perlman, P.S., Mecklenburg, K.., Petrilo, 192 M.L., Tabor, J.H., Jarrell, K.A. and Cheng, H.-L. (1986). A self-splicing RNA excises an intron lariat. Cell 44, 213-223. Piccirilli, J.A., McConnell, T.S, Zaug, A.J., Noller, H.F. and T.R. Cech (1992). Aminoacyl esterase activity of the Tetrahymena ribozyme. Science 256, 1420-1424.

Price, J.V., Kieft, 6 .L., Kent, J.R., Sievers, E.L. and Cech, T.R. (1985). Sequence requirements for self-splicing of the Tetrahymena thermophila pre-ribosomal RNA. Nucleic Acids Res. 13, 1871-1889. Pyle, A.M., McSwiggen, J.A. and Cech, T.R. (1990). Direct measurement of oligonucleotide substrate binding to wild- type and mutant ribozymes from Tetrahymena. Proc. Natl. Acad. Sci. U.S.A. 87, 8187-8191. Reinhold-Hurek, B. and Shub, D.A. (1992). Self-splicing introns in tRNA genes of widely divergent bacteria. Nature 357, 173-176. Rich, A. and RajBhandary, U.L. (1976). Transfer RNA: molecular structure, sequence, and properties. Ann. Rev. Biochem. 45, 805-860. Riggs, A.D., Suzuki, H. and Bourgeois, S. (1970). lac repressor-operator interaction. I. equilibrium studies. J. Mol. Biol. 48, 67-83. Rould, M.A., Perona, J. J., Soil, D. and Steitz, T. A. (1989). Structure of E . coli glutaminyl-tRNA synthetase complexed with tRNAab and ATP at 2.8 A resolution. Science 246, 1135-1142. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, New York. Sampson, J.R. and Uhlenbeck, 0. C. (1988). Biochemical and physical characterization of an unmodified yeast phenylalanine transfer RNA transcribed in vitro. Proc. Natl. Acad. Sci. USA 85, 1033-1037. Sanger, F., Nicklen, S. and Coulson, A.R. (1977). DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74, 5463-5467. Saville, B.J. and Collins, R.A. (1990). A site-specific self­ cleavage reaction performed by a novel RNA in Neurospora mitochondria. Cell 61, 685-696. 193 Schimmel, P.R. and S311, D. 1979. Aminoacyl-tRNA synthetases: General features and recognition of transfer RNAs. Ann. Rev. Biochem. 48, 601-648. Schimmel, P. (1987). Aminoacyl tRNA synthetases: general scheme of structure-function relationships in the polypeptides and recognition of transfer RNAs. Ann. Rev. Biochem. 56, 125-158. Schimmel, P. (1989). Parameters for the molecular recognition of transfer RNAs. Biochem. 28, 2747- 2759. Shub, D.A., Xu, M.-Q., Gott, J.M., Zeeh, A. and Wilson, L.D. (1987). A family of autocatalytic group I introns in bacteriophage T4. Cold Spring Harbor Symp. Quant. Biol. 52, 193-200. Seraphin, B., Simon, M . , Boulet, A. and Faye, G. (1989). Mitochondrial splicing requires a protein from a novel helicase family. Nature. 337, 84-87. Seraphin, B., Simon, M. and Faye, G. (1988). MSS18, a yeast nuclear gene involved in splicing of intron AI5B of the mitochondrial coxl transcript. EMBO J. 7, 1455-1464. Sullivan, F.X. and Cech, T.R. (1985). Reversibility of cyclization of the Tetrahymena rRNA intervening sequence: implication for the mechanism of splice site choice. Cell 42, 639-648. Terpstra, P., Holtrop, M. and Kroon, A.M. (1977). A complete cleavage map of Neurospora crassa mtDNA obtained with endonucleases EcoRI and BamHI. Biochim. Biophys. Acta 475, 571-588. Thompson, L.D. and Daniels, C.J. (1990). Recognition of exon- intron boundaries by the Halobacterium volcanii tRNA intron endonuclease. J . Biol. Chem. 265, 18104-18111. Turcq, B., Dobinson, K.F., Serizawa, N. and Lambowitz, A.M. (1992). A protein required for RNA processing and splicing in Neurospora mitochondria is related to gene products involved in cell cycle protein phosphatase functions. Proc. Natl. Acad. Sci. USA 89, 1676-1680. Valencik, M.L., Kloeckener-Gruissem, B., Poyton, R.O. and McEwen, J.E. (1989). Disruption of the yeast nuclear PET54 gene blocks excision of mitochondrial intron aI58 from pre-mRNA for cytochrome c oxidase subunit I. EMBO J. 8 , 3899-3904. van der Horst, G., Christian, A. and Inoue, T. (1991). 194 Reconstitution of a group I intron self-splice reaction with an activator RNA. Proc. Natl. Acad. Sci. USA 8 8 , 184- 188. von Ahsen, U., Davies, J. and Schroeder, R. (1991). Antibiotic inhibition of group I ribozyme function. Nature 353, 368-370. Waring, R.B. and Davies, R.W. (1984). Assessment of a model for intron RNA secondary structure relevant to RNA self­ splicing - a review. Gene 28, 277-291. Weiner, A.M. and Maizels, N. (1987). tRNA-like structures tag the 3' ends of genomic RNA molecules for replication: implications for the origin of protein synthesis. Proc. Natl. Acad. Sci. USA 84, 7383-7387. Winicov, I. and Button, J.D. (1982). Nuclear ligation of RNA 5’-OH kinase products in tRNA. Mol. Cell. Biol. 2, 241- 249. Winter, G., Koch, G.L.E., Hartley, B.S. and Barker, D.G. (1983). The aminoacid sequence of the tyrosyl-tRNA synthetase from Basillus stearothermophilus. Eur. J. Biochem. 132, 383-387. Wollenzien, P.L., Cantor, C.R., Grant, D.M. and Lambowitz, A. M. (1983). RNA splicing in Neurospora mitochondria: structure of the unspliced 35S precursor ribosomal RNA detected by psoralen cross-linking. Cell 32, 397-407. Xu, M . , Kathe, S.D., Goodrich-Blair, H., Nierzwicki-Bauer, S.A. and Shub, D.A. (1990). Bacterial origin of a chloroplast intron: conserved self-splicing group I introns in . Science 250, 1568-1570. Yanisch-Perron, C., Vieira, J. and Messing, J. (1985). Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mpl8 and pUC19 vectors. Gene 33, 103-119. Yarus, M. and Berg, P. (1967) . Recognition of tRNA by aminoacyl tRNA synthetases. J. Mol. Biol. 28, 479-490. Yarus, M. (1988). A specific binding site composed of RNA. Science 240, 1751-1758. Yisraeli, J.K. and Melton, D.A. (1989). Synthesis of long, capped transcripts in vitro by SP6 and T7 RNA polymerases. Methods Enzymol. 180, 42-50.

Zaug, A.J., Grabowski, P.J. and Cech, T.R. (1983). 195 Autocatalytic cyclization of an excised intervening sequence RNA is a cleavage-ligation reaction. Nature 301, 578-583. Zaug, A.J., Been, M.D. and Cech, T.R. (1986). The Tetrahymena ribozyme acts like an RNA restriction endonuclease. Nature 324, 429-433. Zaug, A.J., Kent, J.R. and Cech, T.R. (1985). Reactions of the intervening sequence of the Tetrahymena ribosomal ribonucleic acid precursor: pH dependence of cyclyzation and site-specific hydrolysis. Biochemistry 24, 6211-6218.