Iowa State University Capstones, Theses and Retrospective Theses and Dissertations Dissertations

1993 In vitro analysis of the self-cleaving RNA of barley yellow dwarf Stanley Livingstone Silver Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/rtd Part of the Biochemistry Commons, Molecular Biology Commons, and the Pathology Commons

Recommended Citation Silver, Stanley Livingstone, "In vitro analysis of the self-cleaving satellite RNA of barley yellow dwarf virus " (1993). Retrospective Theses and Dissertations. 10274. https://lib.dr.iastate.edu/rtd/10274

This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Retrospective Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. _UMI

MICROFILMED 1993 | INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, begiiming at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 313/761-4700 800/521-0600

Order Number 9335023

In vitro analysis of the self-cleaving satellite RNA of barley yellow dwarf virus

Silver, Stanley Livingstone, Ph.D.

Iowa State University, 1993

UMI 300 N. ZeebRd. Ann Arbor, MI 48106

In vitro analysis of the self-cleaving satellite RNA of barley yellow dwarf virus

by

Stanley Livingstone Silver

A Dissertation Submitted to the Graduate Faculty in Partial Fulfillment of the Requirements for the Degree of DOCTOR OF PHILOSOPHY

Department: Plant Pathology Interdepartmental Major; Molecular, Cellular, and Developmental Biology

Approved:

Signature was redacted for privacy.

In Charge of Major Work Signature was redacted for privacy. or/the InterdeHartmental Major

Signature was redacted for privacy. De^rtapiept Signature was redacted for privacy. For the dM^^Co^ege

Iowa State University Ames, Iowa 1993 ii

DEDICATION

This dissertation is dedicated to my parents, Stewart and Jean Silver iii

TABLE OF CONTENTS

Page GENERAL INTRODUCTION 1

Group I Intron Splicing 1 Ribonuclease P; The First True RNA Enzyme 4 Group II Intron Splicing 6 Nuclear, Pre-mRNA Splicing 6 Small Catalytic 8 Reaction Mechanism 15 Target-specific 16 Barley Yellow Dwarf Virus 19 Satellite RNAs 20 Satellite RNA of Barley Yellow Dwarf Virus 2 6 Project Goals 27 Explanation of Dissertation Format 27

PAPER I. ALTERNATIVE TERTIARY STRUCTURE ATTENUATES SELF-CLEAVAGE OF THE IN THE SATELLITE RNA OF BARLEY YELLOW DWARF VIRUS 3 0 ABSTRACT 32 INTRODUCTION 33 MATERIALS AND METHODS 38 Synthesis of wildtype and mutant self-cleavage structures 38 Self-cleavage assays 39 Synthesis of end-labeled RNA 41 Nuclease digestion 42 RESULTS 43 Self-cleavage of wildtype and mutant RNAs 43 Nuclease sensitivity 46 iv

Page DISCUSSION 53 Effects of mutations on self-cleavage rate 53 Nuclease sensitivity 54 Comparison with other hammerheads 57 REFERENCES 61

PAPER II. NONCONSENSUS BASES IN THE sBYDV (+) RNA RNA AFFECT THE RATE OF HAMMERHEAD FORMATION AND SELF-CLEAVAGE EFFICIENCY 64 INTRODUCTION 66 MATERIALS AND METHODS 70 Synthesis of wildtype and mutant self-cleavage structures 70 Self-cleavage assays 72 Synthesis of end-labeled RNA and nuclease digestion 73

RESULTS 74 Self-cleavage of mutant RNAs 74 Nuclease sensitivity 79 DISCUSSION 84 Effects of mutations on self-cleavage rate 84 Nuclease sensitivity 86 Other hammerheads 88 REFERENCES 91

PAPER III. REPLICATION OF TRANSCRIPTS FROM A cDNA CLONE OF BARLEY YELLOW DWARF VIRUS SATELLITE RNA IN OAT PROTOPLASTS 94 ABSTRACT 96 INTRODUCTION 97 V

Page Self-cleavage of dimeric sBYDV RNA 100 Dependence of sBYDV RNA replication on BYDV genomic RNA 100 Replication of both strands of sBYDV RNA by a rolling circle mechanism 105 REFERENCES 109

PAPER IV. BIMOLECULAR CLEAVAGE REACTION AND THE DESIGN OF TARGET-SPECIFIC RIBOZYMES 113 INTRODUCTION 115 MATERIALS AND METHODS 123 Construction of ribozymes 123 In vitro transcriptions and cleavage assays 126 RESULTS AND DISCUSSION 129 Bimolecular cleavage 129 targeted ribozymes 132 REFERENCES 138

GENERAL SUMMARY 140 LITERATURE CITED 143 1

GENERAL INTRODUCTION

A ribozyme is defined as an RNA molecule that has the inherent ability to catalyze cleavage or ligation of covalent bonds (Kruger et al., 1982). The term ribozyme was coined initially to describe the mechanism by which the intervening sequence (IVS) of Tetrahymena thermophilia processes the

pre-rRNA to form mature rRNA. However, the IVS RNA is not truly an enzyme since it does not exhibit turnover ability or remain unchanged during the splicing reaction. Since the initial discovery of ribozymes in 1982, many other RNAs have been observed to exhibit autocatalysis. Such varied RNAs as Group II mitochondrial introns, pre- tRNAs, nuclear mRNA introns, several small, pathogenic RNAs, transcripts of Neurospora DNA, and transcript II from newt all demonstrate the capability of autocatalysis.

Group I Intron Splicing Group I introns were the first RNA molecules demonstrated to have at least quasi-catalytic activity (Kruger et al., 1982). The splicing mechanism involves two transesterifications (Cech et al., 1981; Fig. 1). The first involves a free guanosine or a 5'- phosphorylated form of guanosine serving as the nucleophile that attacks the phosphorous atom at the 5' splice site, resulting in the 2

A. Group I intron splicing pQqh

i Ujp —Gpi ? < UIOH + pQp—Qp[ Exon Intron Exon

i V|P| k + pQp—QQH B. RNase P cleavage Excised intron

- c{^ 5—J 1—3" 5- _0H J L_3' iWtRNA

C. Group II intron self-splicing, nuclear pre-mRNA splicing

r~ IpQU A PI J I I OH C3,^A^p[

A^p Lariat structure D. Self-cleaving, small pathogenic RNAs —p §• SL 3' A+ OH 2'-3'-cyclic phosphate

Figure l. Four major types of RNA-catalyzed RNA cleavage. The rectangles with one jagged end represents a less-than- full- length exon. The pre-mRNA and Group II intron splicing mechanisms are shown together; even though both RNAs most likely use a different mechanism to form an efficient cleavage structure (Symons, 1991b). 3

formation of a 3' 5' phosphodiester bond between the cofactor and the 5' end of the intron. The free 3' OH of exon 1 then attacks the 3' end splice site displacing the intron and ligating exon 1 and 2 (Cech, 1990). Efficient Group I intron splicing, in addition to the nucleotide cofactor, requires specific secondary and tertiary interactions within the RNA. The secondary structure consists of a complex of nine stem-loops collectively called the Michel-Davies model (Davies et al., 1982; Michel et al., 1982). Molecular modelling analysis, sequence comparisons, mutational analysis, and enzymatic probing have verified several of the proposed secondary features of the IVS RNA (Been & Cech, 1986; Cech, 1988; Ehrenmann et al., 1989). The tertiary structure is believed to form a cavity which facilitates GTP binding, whereas the primary structure of the IVS RNA confers RNA specificity via an internal guide sequence located near the splice site (Davies et al., 1982; Waring et al., 1986). In addition to RNA secondary and tertiary interactions, called mRNA maturases enhance RNA splicing in Neurospora crassi (Garriga & Lambowitz, 1986) and yeast mitochondria (Lazowaska et al., 1980). Finally, the IVS RNA also undergoes a secondary reaction that involves the cleavage of 19 from the

5' end to form the molecule L-19 IVS (Zaug et al., 1986). The 4

L-19 IVS has recently spurred great interest in the concept of using RNA molecules as target-specific endonucleases.

Ribonuclease F: The First True RNA Enzyme Ribonuclease P (RNase P) was the first and only RNA molecule shown thus far to have true catalytic activity without prior modification (Zaug et al., 1986). In prokaryotes a single gene may code for several copies of a tRNA; therefore, excision of a single, mature tRNA requires an endoribonucleolytic cleavage on both the 5' and 3' sides of each pre-tRNA. This discussion will be limited only to the events involved with the 5' pre-tRNA cleavage since the enzymes(s) involved in 3' processing have not been identified clearly. The 5' pre-tRNA sequence that undergoes cleavage is not conserved and thus, RNase P must recognize a structure common to all tRNA molecules. After recognition, the enzyme specifically cleaves the RNA by base hydrolysis to produce the

5' terminus (Fig. IB). The RNase P holoenzyme consists of one basic with a molecular mass of 14 kilodaltons (kD) and one single-stranded RNA of 377 nts in Escherichia coli (Baer et al., 1990) or 401 nts in Bacillus subtilis (Pace & Smith,

1990). Absence of either the protein or RNA component of RNase P inhibits pre-tRNA cleavage under physiological conditions (Baer et al., 1990). At high ionic strength, however, the RNA component alone can efficiently cleave the 5

pre-tRNA. This observation led to the suggestion that the RNA acts as the catalytic subunit and the protein serves to stabilize RNA/tRNA interactions (Baer et al., 1990). Using phylogenetic comparison of RNase P from B.

megaterium and E. coli, the putative secondary structure was

determined (Altmann, 1987). Structures within the RNA component of RNase P include a pseudoknot (Pace & Smith, 1990) and several nucleotides dispersed throughout the RNA that act collectively to provide the RNA secondary structure required for efficient pre-tRNA cleavage. Mutagenic analysis identified specific regions within the RNA that associate with the protein component (Shiraishi & Shimura, 1988). Finally, UV-cross-linking experiments have revealed two widely separated regions within RNase P that interact during the tRNA splicing reaction (Burgin & Pace, 1990). RNase P specificity for the pre-tRNA comes from an external guide sequence (E6S) composed of the 3' sequence of the tRNA acceptor stem (Fig. IB). The EGS of the pre-tRNA binds RNase P adjacent to the cleavage site (Forster & Altmann, 1990). For efficient cleavage, the EGS and a second conserved sequence within the pre-tRNA -NCCA- must reside on opposite strands. It has been proposed that a ribonuclease that will cleave RNA in a site-specific manner could be designed by modifying the EGS (Forster & Altmann, 1990). 6

Group II Intron Splicing Group II intron splicing has been observed only in the organelles of fungi and . To date, 70 Group II introns have been characterized (for review see Michel et al., 1989). Similar to pre-mRNA splicing. Group II intron splicing involves a two-step transesterification (see discussion below). Group I and Group II intron splicing differ with respect to the chemical species that initiates the first transesterification. A 3' OH from a free guanosine initiates group I intron splicing, whereas, Group II intron splicing results from an attack by a 2' OH from an adenosine located within the intron (Fig. IC; Jacquier, 1990). The result of the first nucleophilic attack in Group II intron processing is a lariat structure consisting of the intron and exon 2. The 3' OH of exon 1 then attacks the 5' end of exon 2 resulting in release of the intron and ligation of exon 1 and 2. Currently, Group II intron splicing is believed the result of specific folding of the RNA without a requirement for proteins, trans-acting factors or adenosine (for review see Michel et al., 1989).

Nuclear, Pre-mRNA Splicing To translate active proteins from an mRNA template, cells must have a highly specific mechanism to excise introns and ligate the resulting exons. The mechanism evolved to perform 7

this function is both the most complex and least understood of all the known RNA splicing reactions. Pre-mRNA splicing involves formation of a lariat intermediate and proceeds via a two-step transesterification reaction similar to Group II intron splicing (Cech, 1990; Jacquier, 1990; Fig. IC). In the pre-mRNA reaction, sequences adjacent to the 5' and 3' splice sites and the branch point are required but not sufficient for splicing (Ruby & Abelson, 1991). In addition to sequence requirements, several small, nuclear RNAs (snRNAs) called Ul, U2, U4, US, and U6 associate with the pre-mRNA along with trans-acting factors and other proteins in a temporally and spatially specific manner (for review see Ruby & Abelson, 1991; Zieve & Sauterer, 1990). ATP is also a required component for the protein/ interaction. Presently, many other as yet unidentified proteins and splicing factors are believed to be required for efficient pre-mRNA processing (Ruby & Abelson, 1991). It has been proposed that pre-mRNA splicing evolved from the less complex Group II intron splicing mechanism (Jacquier, 1990). In pre-mRNA processing, the snRNAs are believed to facilitate the folding of the pre-mRNA into a structure capable of splicing. Group II introns, on the other hand, do not require proteins for correct folding of the RNA (Jacquier, 1990). Determination of whether or not pre-mRNA processing evolved from the Group II intron splicing reaction, will 8

require an answer to the question: is pre-mRNA splicing an RNA- or protein- catalyzed reaction.

Small Catalytic RNAs Another type of ribozyme, first observed in a small pathogenic RNA of tobacco (Prody et al., 1986) results in the

production of RNA fragments with 2'-2' cyclic phosphate and 5' OH termini (Fig. ID). The self-cleavage reaction occurs most efficiently at pH 7.5 in the presence of MgClg and is

independent of a protein requirement (Prody et al., 1986). There are four distinct classes of this small ribozyme based on the primary and secondary structure at the cleavage site. The first class of ribozyme was first observed in the satellite RNA of tobacco ringspot virus (sToBRV RNA) (for review of satRNAs see below). Terminal analysis of sToBRV RNA cleavage fragments revealed the presence of a 5' hydroxyl and

a 2'-2' cyclic phosphate instead of the 5' phosphate and 3' OH termini observed in Group I intron splicing. Since different termini result from the small catalytic ribozyme, the cleavage mechanism responsible for self-cleavage most likely uses a different reaction pathway than either Group I intron or pre-tRNA splicing. Furthermore, Mg'^^, which is specifically

required for Group I intron cleavage, can be replaced with Mn^^, Pb®^, Zn^^, or even spermidine in the sToBRV cleavage reaction. Finally, unlike Group I intron splicing, sToBRV RNA 9

cleavage occurs efficiently without GTP (Prody et al., 1986). The mechanism proposed for sToBRV RNA cleavage involves a phospho-transfer reaction (Fig. 2) using the ribosyl 2' OH on the 5' side of the cleavage site as the nucleophilic species that attacks the partially positive phosphate center (Prody et al., 1986).

t o o

Figure 2. Phospho-transfer mechanism proposed for the small, catalytic ribozyme cleavage reaction (Symons, 1991b).

Comparison of six virusoid sequences and one satellite RNA sequence revealed a highly conserved region that folded into a hammerhead-shaped structure (Forster & Symons, 1987). Table 1 lists all the known RNAs that form the hammerhead motif. The hammerhead contains three helical domains (I, II and III) and a central core of eleven highly conserved, mainly single-stranded nucleotides (Forster & Symons, 1987) (Fig. 3A). Mutational analysis of the eleven conserved nucleotides Table 1 Small RNAs capable of self-cleavage in vitro

RNA and cleavage structure Size SNA. strand Reference

I. Cleavage bj the hammerhead A. #

Avocado sunblotch viroid (ASBVd) 246--251 1 Hutchins et al., 1986

Carnation stunt associated viroid (CarSAVd) 275 1 Hernandez et al., 1992

Peach latent mosaic viroid (PLHVd) 337 1 Hernandez, 1992 B. Encapsidated linear satellite RNAs of: Barley yellow dwarf virus (sBYDV) 322 (+)/{-) Miller et al., 1991 Tobacco ringspot virus (sToBRV) 359 -360 (+) Prody et al., 1986 Arabis mosaic virus (sArBV) 300 {+) Bruening, 1989 C. Encapsidated circular satellite RNAs of: Lucerne transient streak virus,(vLTSV) 324 (+)/(-) Forster & Symons, 1987 Solanum nodlflorum mottle virus (vSNHV) 378 (+) Symons, 1989 Subterranean clover mottle virus (vSCMoV) 3325388 (+) Davies et al., 1990 Velvet tobacco mottle virus (vVTMoV) 366 ( + ) Symons, 1989 D. RNA transcript of newt satellite II DNA 300 (+) Epstein & Gall, 1987 II. cleavage by hairpin/bent paperclip Encapsidated linear satellite RNA of ToBRV 359-360 (-) Breuning, 1989 III. cleavage not fully defined Hepatitis delta virus RNA (HDV RNA) 1700 (+)/{-) Perotta S Been, 1990 IV. Cleavage by completely undefined structure Neurospora, linear and circular mitochondrial 800 (+) Saville & Collins, 1990 RNAs 11

OlMVIM A. ConMHMN hamtiMrtiMd 5NNNNN- Œi'

B. Htlrp(ne«ttlytlomocM

Catalytic RNA ciMvae* Substrats RNA \

u AIM AC A m AOUC 40 O / QUO du uu cuoau JOAC euauu C U 1 111 II II nut 1 III 1*1111 CAO CA AO OACW IV III II

c. HDV Moondary Mnielur*

S'— u

Figure 3. Structures of the three partially-characterized self-cleavage sites of small, pathogenic RNAs. (A) Hammerhead motif employed by the majority of small, self-cleaving RNAs. Nucleotides conserved among the known hammerheads are in boxes. (B) The hairpin structure at the cleavage site of sToBRV (-) strand RNA. (C) The proposed structure of the catalytic site of HDV. 12

demonstrated that all are critical for efficient ribozyme activity (Ruffner et al., 1990; Sheldon & Symons, 1989). From these observations the conserved nucleotides are believed to be directly involved in the formation of the correct secondary and tertiary structure required for cleavage. Mutations involving the non-conserved bases of the hammerhead decreased the cleavage rate by inducing hammerhead instability (Ruffner et al., 1990). Mutation analysis of the hammerhead corroborates the prediction that as the stability of the hammerhead increases, so does the transition state stability with the end result being a faster cleavage rate (Ruffner et al., 1989; Ruffner et al., 1990). However, recent studies of an intermolecular-type ribozyme reaction suggested the possibility that the transition state, i.e. the hammerhead, should be sufficiently unstable to decrease the activation energy needed for cleavage (Fedor & Uhlenbeck, 1992). Therefore, the fastest cleaving RNAs will form a hammerhead with the least amount of energy. A second type of ribozyme was observed in the sToBRV RNA (-) strand (Table 1). This class utilizes a "hairpin" structure to achieve efficient site-specific cleavage (Hampe1 et al., 1990; Fig. 3B). Site-directed mutagenesis identified the nucleotides instrumental in hairpin structure (Hampe1 et al., 1990; Haseloff & Gerlach, 1989). Two discrete regions of RNA are required for efficient sToBRV (-) strand cleavage. 13

One region maps to a 12 nucleotide sequence in the proximity of the cleavage site, whereas the second maps to a position approximately 124 bases downstream of the cleavage site (Haseloff & Gerlach, 1989). The hairpin motif contains four helical domains that, when disrupted, result in loss of activity (Hampel et al., 1990). Sequences within the sToBRV RNA basepair to form the four helices that permit efficient self-cleavage. In addition to the sToBRV (-), the nucleotides of the pre-mRNA splicing component U6 snRNA can be envisioned to fold into the hairpin structure. Whether or not U6 snRNA catalyzes the cleavage step in pre-mRNA splicing will require further study (Tani et al., 1992). Hepatitis delta virus (HDV) (+) and (-) sense RNA represents a third class of ribozyme associated with a small, pathogenic RNA (Table 1). The HDV RNA is approximately 1700 nts in length and by far larger than any of the self-cleaving RNAs that infect plants. The only requirements for cleavage are a neutral pH and Mg^^ (Kuo et al., 1989; Sharmeen et al.,

1988). Computer modelling and mutation analysis of the full-length HDV RNA indicated four helical domains form during the cleavage reaction (Been et al., 1992; Fig. 3C). Mutations that partially disrupt the helical domains result in decreased cleavage rates. A putative tertiary interaction involves the formation of a pseudoknot between stem I and stem 14

II that may facilitate HDV RNA cleavage (Perrotta & Been, 1991). Deletions of the HDV RNA demonstrated that a minimum of 88 nts was sufficient for cleavage (Wu et al., 1990). Further deletions of the HDV RNA made using an alkaline degradation method showed an even smaller molecule could self-cleave in vitro. The minimum sequence required for cleavage includes one nucleotide upstream of the cleavage site and the 84 nts downstream of the cleavage site (Perrotta & Been, 1991) (Fig. 3C). This shortened version of the HDV RNA still has the potential to fold into a structure with two stem-loops. Using HDV RNA molecules of varying lengths, site-directed mutagenesis and computer modelling experiments provide equivocal data on the relative importance of the helical structures and single-stranded regions (Perrotta & Been, 1991; Thill et al., 1991; Wu et al., 1989; Wu et al., 1992). The helices possibly serve indirectly in the formation and stability of the cleavage structure under various in vivo conditions (Perrotta & Been, 1991). The last known type of small catalytic RNA was isolated from mitochondria of the Varkud-lc strain of Neurospora

(VSRNA). VSRNA is approximately 800 nts in length and exists in either a circular and/or linear form (Saville & Collins, 1990; Table 1). Both forms exist as long multimers in vivo. RNA transcribed from a VSDNA clone and incubated in the presence of Mg'^^ at neutral pH underwent a site-specific 15

cleavage with the formation of 5' OH and a 2'-3' cyclic phosphate fragments (Saville & Collins, 1990). A search of the VSRNA did not uncover any conserved sequences present in the other small, self-cleaving RNAs motifs. Presently, the secondary and tertiary interactions required for efficient Neurospora VSRNA self-cleavage are unknown.

Reaction Mechanism The one underlying theme of all RNAs that undergo self-cleavage is the importance of secondary and tertiary interactions that permit a particular phosphodiester bond to become sufficiently labile to cleave in the absence of a protein catalyst. Thus far, proteins involved in RNA-catalyzed RNA cleavage reactions seem to serve as stabilizers for the RNA secondary structure required for cleavage. RNAs that self-cleave using the hammerhead motif do so by a phospho-transfer mechanism similar to base hydrolysis (Fig. 2). Computer modelling studies of the hammerhead suggest: (i) the base at the cleavage site does not interact with other bases, (ii) the ribose-phosphate backbone at the cleavage site makes a sharp turn and (iii) the conformation of the ribose at the cleavage site becomes 2' endo (Mei et al., 1989). Collectively, these changes ensure that a specific phosphodiester bond is cleaved via an "in-line" attack by the 16

2' OH of the ribose on the 3' P to form the products with 2', 3' P, and 5' OH termini (Mei et al., 1989). The role plays in RNA-catalyzed RNA cleavage is still

unclear. In Group I introns and the hammerhead ribozymes, Mg^*

coordinates with one specific Pro-R oxygen (Dahm & Uhlenbeck, 1991; Slim & Gait, 1991; Uchimaru et al., 1993). By changing which Pro-R oxygen Mg^"*" binds, Uchimaru

et al. (1993) observed cleavage of normally resistant phosphodiester bonds. Therefore, it seems RNA secondary and tertiary structure determines which Pro-R oxygen Mg^^ binds and

ultimately specifies which phosphodiester bond is cleaved. Whether Mg^"*" serves a similar role in the hammerhead-type

ribozyme cleavage will require further study.

Target-specific Ribozymes Soon after RNA molecules were shown to self-cleave in vitro, the concept of targeting the catalytic portion of these self-cleaving molecules against a given substrate RNA was proposed (Haseloff & Gerlach, 1988; Uhlenbeck, 1987; Zaug et al., 1986). Of the known RNA ribozymes, the hammerhead type is by far the smallest and best characterized. Subsequently, the currently is most widely used for targeting cleavage of a specific RNA. The first step in constructing a target-specific ribozyme was to determine whether or not the catalytic domain of the hammerhead would 17

function in an intermolecular fashion. The first intermolecular reaction involved the sequences from the avocado sunblotch viroid (ASBV) hammerhead (Uhlenbeck, 1987). The ribozyme (Rz) was composed of all the conserved hammerhead nts with the exception of the sequence GAAA and the U at the cleavage site which were made part of the substrate RNA (S) (Uhlenbeck, 1987; Fig. 4A). When the S and Rz were combined at neutral pH in the presence of Mg^"*", cleavage of the

substrate RNA was observed. This result proved conclusively that the hammerhead ribozyme could effectively cleave RNA in a bimolecular manner if a minimum sequence (in this case only 43 nts) which contained all 11 conserved nts of the hammerhead motif were present (Uhlenbeck, 1987). Haseloff and Gerlach (1988) constructed a similar bimolecular system to test the Rz sequence from the sToBRV (+) strand hammerhead. In their system, the substrate sequence had only one conserved base from the hammerhead and the CUGANGA and GAAA sequences were part of the Rz (Fig. 4B). Target arms complementary to the substrate RNA were then combined with the catalytic domain of sToBRV to produce the complete ribozyme. When targeted against the chloramphenicol acetyltransferase (CAT) mRNA, they observed specific and efficient cleavage (Haseloff & Gerlach, 1988). Since this initial account, many ribozymes have been successfully targeted against a host of RNAs (for review see Symons, 18

A. C C G

Substrate pppGÇGÇÇ® hoCQCGG^ gAGCUCGGppp

U4® Ribozyme

B.

5> Substrate I f 3' xxxxxx lyyyyOxxxxxx xxxxx xxxxxx 3' II

C-G U AU G'C Ribozyme G'C A G G U

Figure 4. (A) Division of the ASBV hammerhead into substrate and catalytic domains (adapted from Uhlenbeck, 1987). Conserved bases of hammerhead are in bold italics. Arrow indicates cleavage site. (B) Gene-targeted ribozyme as proposed by Haseloff and Gerlach (1988). (I) indicates the one conserved nt in the substrate RNA. (II) identifies the target-arms which give specificity to the ribozyme. (Ill) is the catalytic domain of the ribozyme. Arrow indicates cleavage site, (adapted from (Haseloff & Gerlach, 1988). 19

1991b). RNAs successfully cleaved in vivo by a gene-targeted ribozyme include U7 snRNA, c-fos mRNA, HIV-1 RNA and the a-sarcin domain of 28S RNA (Gotten et al., 1989; Sarver et al., 1990; Saxena & Ackerman, 1990; Scanlon et al., 1991). Unfortunately, many of the gene-specific ribozymes, when tested in vivo, do not cleave their substrate RNA effectively. Parameters such as in vivo reaction conditions, length and composition of the ribozyme target arms and alternative conformations within the target arms all can dramatically affect the ribozyme's ability to cleave in vivo (see Section IV for further details). Further work will be required in order to design a rational approach to ribozyme design.

Barley Yellow Dwarf Virus Barley yellow dwarf virus (BYDV), the type member of the luteovirus group of plant , was classified as a new virus in 1951 (Oswald & Houston, 1951). From the historical record, however, symptoms of the virus were observed as early as the late 1800s. BYDV is now recognized as the most destructive virus capable of infecting small grains including oats, barley, and wheat (Plumb, 1983). In fact, grain yield losses observed from intentional BYDV infections of some wheat varieties have been greater than 60%. Even naturally occurring infections can result in a greater than 25% grain yield loss (Yamani & Hill, 1991). Symptoms of BYDV infection 20 include yellowing of plant tissue, moderate to severe stunting of the plant and, in severely affected plants, heads with many blasted or sterile spiklets. BYDV is phloem-limited and aphid-transmitted in a circulative and persistent manner. The virus particle consists of approximately 180 subunits of the 22kd coat protein (CP) arranged as an icosahedron with a diameter of 24 to 30 (nm). Five different serotypes of BYDV have been classified according to aphid specificity and serological characteristics (Rochow, 1970). These serotypes can be grouped into two subgroups based on their organizations, cytopathological effects, and serological relatededness. BYDV serotypes MAV, PAV and SGV belong to subgroup I and RMV and RPV belong to subgroup II. The complete nucleotide sequence of an Australian isolate of BYDV-PAV was determined and the genome organization deduced (Fig. 5A; Miller et al., 1988). The BYDV genome consists of a single (+) sense RNA approximately 6kb long. Gene expression strategies used by BYDV include a -1 ribosomal frameshift (Brault & Miller, 1992), subgenomic RNA synthesis, and an internal initiation and readthrough event (Dinesh-Kumar et al., 1992).

Satellite RNAS Several types of RNA exist that either by themselves or with the aid of a , can infect plants and cause disease. A short summary of each type is given below. 21

Readthrough Frameshift CP 22KK^ mm# I7K 39K 6.7K Leaky Scanning 5* 3' •sgRNM sgRNA2

B 71K j 22K m 72mmmm 17K 5' 3'

Figure 5. Genome organization of BYDV. Panel A is the genome organization of BYDV-PAV of subgroup I. Panel B depicts the genome organization of BYDV-RPV of subgroup II. Open reading frames are depicted by rectangles with the molecular mass of the putative product inside, abbreviation; sgRNA, subgenomic RNA (Miller et al., 1988; Vincent et al., 1991). 22

1. Satellite viruses are defective viruses that encode the genetic information for their own coat protein. They require a helper virus for replication. 2. Defective interfering RNAs (DI RNA) are viral RNAs that have lost a portion of the virus genome. Replication requires the presence of a full-length copy of the virus. DI RNAs, by definition, interfere with normal virus replication. 3. are unencapsidated RNA molecules ranging from 246 to 375 nts in length that do not require a helper virus for replication. Viroids exist as covalently closed, circular molecules that have a high degree of basepairing and thus are quite stable. 4. Satellite RNAs (satRNA) are defined as (i) completely dependent on a helper virus for replication, encapsidation and spread, (ii) having no known sequence homology with the plant or viral , and (iii) not being required by the helper virus. SatRNAs range from 300 to over 1400 nts in length. 5. Virusoids are a type of satRNA that vary in size from 324 to 388 nts and are encapsidated primarily as circular molecules. Thus far only the sobemovirus group of plant viruses demonstrate the ability to support virusoid replication. 6. Satellite-like RNAs are molecules that have qualities of a satRNA but do not fit completely the classical definition of a satRNA. 23

For the purpose of this discussion, only the satRNAs will be discussed in detail.

SatRNAs can be divided into two groups based on their length and coding capacity (Table 2). Large satRNAs associated with the nepovirus group of plant viruses have the ability to direct protein translation (Roossinck et al., 1992). At present, the role if any these proteins play in satRNA biology is unknown. The second group of smaller molecular weight RNA molecules does not direct translation of proteins in vivo. The first small satRNA was found associated with tobacco ringspot virus (sToBRV RNA) (Schneider, 1969). Further study of sToBRV RNA revealed the presence of both multimeric and circular forms in vivo (Schneider, 1977). Since the initial discovery of sToBRV RNA, small satRNAs of the sobemovirus (virusoids) and luteovirus groups have been found which also form multimeric and circular molecules in vivo (Francki, 1985; Keese & Symons, 1987; Miller, et al., 1991). These observations led to the proposal that these RNAs replicate via a rolling circle pathway (Branch & Robertson, 1984; Hutchins et al., 1985). A general outline of the rolling circle mechanism is depicted in Figure 6. After the virion enters the plant, the satRNA is uncoated. The (+) strand then circularizes followed by replication via the virally-encoded RNA-dependent RNA polymerase to produce long multimers of (-) 24

Table 2 Summary of satellite RNAs. This list contains known RNAs that fit the classical definition of a satellite RNA. Abbreviations; NR. no report; <-) no translation products detected; helper virus abbreviation in parentheses; kP. kilodaltons ^adapted from Roossinck. et al.. 19921.

Mol Mass of Helper Virus # nts encoded proteins (kD)

Large satRNAs Tomato blackring virus (TBRV) 1375 48 Chickory yellow mottle (CYMV) 1145 40 Arabis mosaic virus (ArMV) 1104 39 Strawberry latent ringspot (SLRV) 200 38 Myrobalan latent (MLRV) 1400 45 Grapevine Bulgarian latent (GBLV) 1500 NR Grapevine fanleaf (GFLV) 1114 37 Pea enation (PEMV) 900 NR

Small saRMAs Cucumber mosaic virus (CMV) 333-342 2-3.9 Peanut stunt virus (PSV) 369-393 Tobacco mosaic virus (ToBRV) 359 Arabis mosaic virus (ArBV) 300 NR Barley yellow dwarf (BYDV) 322 NR

Virusolds Velvet tobacco mottle (VTMoV) 365 Solanum nodiflorum mottle (SNMV) 377 Lucerne transient streak (LTSV) 324 Subterranean clover mottle (SCMoV) 332, 388 NR 25

strand satRNA. In sBYDV RNA, the (-) strand multimer cleaves into monomers that act as templates for (+) strand synthesis. Similar to the (-) strand, the (+) strand multimers undergo cleavage to form the encapsidated monomer.

(•)

(•) • • ##IW*av#g#

Virus (+) partcle \ (•) - © (+)

• •

Figure 6. Rolling circle replication model. Arrows indicate cleavage sites (adapted from Symons, 1991a).

Demonstration of the autocatalytic capabilities of Group I intron RNA (Kruger et al., 1982) and the ability of RNase P to cleave pre-tRNA into a mature tRNA (Guerrier-Takada et al., 1983) led Hutchins et al. (1985) to propose that these small, pathogenic RNAs cleaved in a manner analogous to the 26

RNA-catalyzed RNA cleavage of Tetrahymena Group I introns. Evidence for the autocatalytic cleavage of the multimeric satRNA into monomers was provided by Prody et al. (1986), who demonstrated the ability of sToBRV to cleave in vitro (discussed previously).

Satellite RNA of Barley Yellow Dwarf Virus Some isolates of BYDV-RPV contain a satellite RNA (Miller et al., 1991). Labeled cDNA of sBYDV RNA hybridized to a multimeric series of bands from purified virion RNA but not to genomic BYDV-RPV RNA. The smallest and most abundant band corresponded to an approximately 320 nt long RNA believed to be the monomer. Detection of a multimeric series as well as circular forms of RNA suggest the RNA replicates by a rolling circle mechanism (Miller et al., 1991). The RNA sequence was first determined by sequencing several overlapping cDNA clones. Most clones contained a 321 base pair repeat, consistent with the size estimated for the monomer by electrophoresis. Direct RNA sequencing revealed an A residue at the 3' end of the monomer that was absent from these clones. Thus the satRNA is actually 322 nts long. Several of these same clones were tested for cleavage ability by in vitro . The clones that contained the extra adenosine cleaved efficiently, while those that lacked the A cleaved poorly. Both the (+) and (-) strands of 27

the RNA have the conserved hammerhead self-cleavage motif at the site of cleavage. Closer examination of the (+) strand cleavage site revealed several sequence differences relative to the other known hammerheads (see Fig. 7A). In addition, the sequence at the (+) strand cleavage site predicted a novel secondary structure which included a putative pseudoknot due to basepairing between nts 6-10 and 34-30 (Fig. 7B).

Project Goals This project addressed three major goals. The first goal was to demonstrate the presence or absence of the proposed alternative structure at the sBYDV RNA cleavage site, and determine the significance of the nonconsensus bases present in the sBYDV RNA hammerhead. The second goal was to prove that the small molecular weight RNA associated with BYDV-RPV is in fact a satRNA. The third goal of the project was to design, construct, and test gene-targeted ribozymes predicted to cleave BYDV genomic RNA.

Explanation of Dissertation Format This dissertation contains three papers either published or in preparation for publication and one manuscript not expected to be submitted for publication at the present time. Paper I contains work performed by myself and W. A. Miller. 28

dMvagt •ito F 310UAUUUC QUQQAB^S20_ Tl ACAQ *0AO 3' MJOMa^CAOSilj^> I I I .i.W iiniir>.I I I I a

y-AXg

daavagi •ity 110 »0_ Tl 6* UWUUÇI i I I i I QuaoMill, ACAQ *g:§ 3' AUOAAOqC^CUI ÙàÙCi Aj®-0,^-Cao

.>3

stacked helices B (pseudoknot) Hammerhead

Figure 7. Alternative structures for the self-cleaving ribozyme in sBYDV (+) strand RNA. Panel A is sBYDV (+) RNA folded in the hammerhead motif. Panel B is sBYDV folded as the alternative structure when basepairing occurs between nts 6-10 and 30-34. Boxed bases are conserved among all hammerheads. Bases that differ from consensus are shown in outlined text. Numbering of nucleotides is based on the full-length satellite RNA (Miller & Silver, 1991). 29

Even though Miller is the principal author of Paper I, it was included in my dissertation since the work is so closely related to Paper II. Specifically, Miller constructed mutants Ml, M19, and M20, and I constructed mutants M2, Ml/2, M/19/20, and M5. Miller performed cleavage assays on all the mutant transcripts with the exception of MS. All nuclease sensitivity analysis was done by Miller. Papers I and II are formatted for publication in Nucleic Acids Research. Paper III is a paper submitted to the journal . Finally, Paper IV reports on research performed by myself but not ready for publication. References cited in the INTRODUCTION and the GENERAL SUMMARY are listed in the LITERATURE CITED section that follows the GENERAL SUMMARY. 30

ALTERNATIVE TERTIARY STRUCTURE ATTENUATES SELF-CLEAVAGE OF THE RIBOZYME IN THE SATELLITE RNA OF BARLEY YELLOW DWARF VIRUS 31

Alternative tertiary structure attenuates self-cleavage of the ribozyme in the satellite RNA of barley yellow dwarf virus

Miller W.A., Ph.D. Silver, S.L., M.S.

From the Department of Plant Pathology, Iowa State University, Ames, lA 50011 32

ABSTRACT

A self-cleaving satellite RNA associated with barley- yellow dwarf virus (sBYDV) contains a sequence predicted to form a secondary structure similar to catalytic RNA molecules (ribozymes) of the "hammerhead" class (17). However, this RNA differs from other naturally occurring hammerheads both in its very slow cleavage rate, and in some aspects of its structure. One striking structural difference is that an additional helix is predicted that may be part of an unusual pseudoknot containing three stacked helices. Nucleotide substitutions that prevent formation of the additional helix and favor the hammerhead increased the self-cleavage rate up to 400-fold. Compensatory substitutions, predicted to restore the additional helix, reduced the self-cleavage rate by an extent proportional to the calculated stability of the helix. Partial digestion of the RNA with structure-sensitive nucleases supported the existence of the proposed alternative structure in the wildtype sequence, and formation of the hammerhead in the rapidly-cleaving mutants. This tertiary interaction may serve as a molecular switch that controls the rate of self-cleavage and possibly other functions of the satellite RNA. 33

INTRODUCTION

Some viruses in the nepo-, luteo-, and sobemovirus groups contain small satellite RNAs that replicate via a rolling

circle mechanism (reviewed in 1,2). Circular and multimeric forms of these RNAs undergo self-cleavage at a specific site to generate the monomeric form that predominates in the virus particle. This self-cleavage is catalyzed by a subset of the satellite RNA sequence (ribozyme) flanking the cleavage site that is predicted to fold into a hammerhead-shaped structure

(3, 4; reviewed in 5). One viroid, avocado sunblotch viroid (ASBV, 6), and the transcripts of satellite 2 DNA of newt also self-cleave at a hammerhead structure (7). The consensus hammerhead structure consists of three short helices, whose sequences are not conserved, joined together by short single- stranded regions, most of which are conserved at the primary structure (sequence) level. In contrast, the loops connecting the distal portions of the helices can be varied in sequence (8), or removed altogether (9, 10), without loss of ribozyme activity. Much effort has been devoted to elucidation of the structure-function relationships of hammerhead ribozymes, including mutagenesis of the helices and the conserved single- stranded bases (11-16). 34

A self-cleaving satellite RNA was recently reported associated with the RPV serotype of barley yellow dwarf virus (sBYDV, [17] GenBank Accession number M63666). This is the first known satellite of a member of the luteovirus group. Encapsidated sBYDV RNA is a predominantly linear monomer, 322 nucleotides (nt) long, and it encodes no long open reading frames. In being a linear molecule it more closely resembles the satellite RNA of tobacco ringspot virus (sTobRV, 2) than the satellites of the sobemoviruses (also known as virusoids) in which the circular monomer is the most abundant form. The (-) strands of some but not all of the self-cleaving satellite RNAs also self-cleave, indicating variations in the rolling circle mechanisms (1, 2). Both the encapsidated (+) strand and the (-) strand of sBYDV RNA self-cleave (17). The (-) strand self-cleaves extremely rapidly and contains a perfect consensus hammerhead sequence flanking the self- cleavage site (5, 14). The sequence flanking the cleavage site in the encapsidated (+) strand of sBYDV fits most of the consensus rules for formation of a hammerhead structure. However, the (+) polarity sBYDV hammerhead differs from other naturally occurring hammerheads in at least three features (Fig. 1). First, the trinucleotide at the 5' side of the cleavage site is AUA, instead of GUC, which is found in all naturally occurring hammerheads except that of lucerne transient streak virus satellite (sLTSV) that cleaves at GUA 35

dMvage 310 320, f UAUUucouaaA %-sli L2c 9' AÛàÀAàoCÂCCW^11 1111 II III I'lni'ie. Q

c-a A-U rou-A ik dMvac A-U 1-1® 5* 310UAUuucauaaA^ACAa^c- ILSa S20_ tllidU^'Q-C 3' ÀÙÔÀMtC^ ®- -CSG A"

s::%ac-a CigHli '^1 ""QQ" Stacked helices B (pseudoknot) Hammerhead

Figure l. Proposed alternative structures for the self- cleaving ribozyme in sBYDV (+) strand RNA. Boxed bases are conserved among all hammerheads (5). Bases that differ from consensus are shown in outlined text. Numbering of nucleotides is based on the full-length satellite RNA (17). Numbering of the three major helices (HI, H2, H3) of the hammerhead is as in (4, 5). Single-stranded regions (loops) are prefixed with L. The individual helices and loops within compound stem-loop 2 are labeled as lettered subsets. Bases in loops LI and L2a expected to form the additional helix are in italics. Cleavage structure is drawn as a hammerhead (A) without Ll-L2a base pairing, and in the alternative conformation (B), in which LI and L2a base pair without disruption of other helices to form a pseudoknot with three stacked helices. Because of difficulties in two dimensional representation, lines indicating phosphodiester bonds have been added as necessary. Note the coaxial alignment of three helices in B with the helix derived from LI and L2a in the center. 36

(4). However, functional hammerheads have been constructed with many variations from GUC (7, 14). Second, an unpaired cytosine (C-24) is present immediately 3' to the conserved single-stranded CUGANGA sequence, and an unpaired adenine (A-73) occurs immediately 5' of the conserved GAAAN sequence. In almost all other cases, bases in these positions are hydrogen bonded to form the first pair of the vertical helix 2 (5, 14). No others have an unpaired base analogous to A-73. The third and most striking, unusual feature is the subject of this paper: the "loop" (LI) connecting strands of helix 1 (HI) would be expected to pair with a sequence (L2a), which is drawn as a single-stranded loop in the hammerhead conformation in compound helix 2 (Fig. 1). The calculated stability of the helix formed by Ll-L2a base pairing

(AG = -10.1 kcal/mol) is the strongest in the entire satellite

RNA. Thus, structures containing the Ll-L2a helix may be energetically preferred, and loops LI and L2a would not be expected to remain single-stranded as shown in the hammerhead conformation in Fig. lA. Because the formation of the Ll-L2a helix does not replace any of the base pairing involved in hammerhead formation, it may be possible for all helices to remain intact, giving the unusual pseudoknot-like structure shown in Fig. IB. This structure contains an interesting set of three coaxially stacked helices, the central one being the 37 Ll-L2a helix. Although such a structure may not be possible due to torsional constraints, we sought to determine if it could form by observing the effects of mutagenesis of the putative Ll-L2a helix on the self-cleavage rate, and on overall secondary and tertiary structure. 38

MATERIALS AND METHODS

Synthesis of Wildtype and Mutant

Self-cleavage Structures RNA molecules were synthesized by in vitro transcription of sequences in pGEM3Zf(-) (Promega, Madison, WI). Mutations were introduced by the method of Kunkel (18), using a kit from Biorad (Richmond, CA). Plasmid pSSl contained the wildtype (+) strand sBYDV hammerhead sequence: bases 310-322 and 1-89 of the satellite sequence (Fig.s 1 and 2). Cleavage occurs between bases 322 and 1. pSSl was derived from the slightly larger plasmid pPCSl (17) by site-directed mutagenesis with the primer: 5' GTTATCCACGAAATAGGA'ICCTATAGTGAGTCGTATTAC 3'. Bases 5' of the apostrophe are complementary to satellite RNA sequence. The complement of the T7 RNA polymerase promoter is shown in italics. This resulted in deletion of 25 vector-derived bases and 11 satellite RNA bases not involved in formation of the hammerhead structure. Transcripts of Xjbal-linearized pSSl and mutants contained vector bases G6A and CUCUAG at the 5' and 3' ends, respectively. The following containing mutant sBYDV self-cleavage structures were constructed with 39

the indicated primers. (See maps and sequence alterations in Fig. 2.) pMl: 5' CGTCGTCAGACAGTAGGGCCTCTGTTATCCACGAA 3';

pM2: 5' CCAGCCTTCTAGTCCGGCCCGATACGTCGTCAGAC 2'; pM19; 5' AGACAGTACGAGCTCTGTTATC 3'; pM20; 5' TTCTAGTCCGCTCGGATACGTCG 3'; pM5: 5' GGATTTCTGTATCTATT'GATACGTCGTCAGAC 3'. Underlining indicates base changes, and the apostrophe indicates site of deletion. All plasmids were sequenced in the transcribed regions by the dideoxy method with Taq polymerase (Promega). In our nomenclature, a plasmid and its transcript are named similarly except that the "p" of the plasmid name is omitted in designating the transcript.

Self-cleavage Assays [a-^'pjUTP-labeled RNAs were transcribed in vitro as described previously (17) from XJbal-linearized plasmids.

Uncleaved transcripts were purified by elution (19) after electrophoresis of transcription products on an 8% polyacrylamide, 7M urea gel. Self-cleavage reactions contained approx. 10,000 cpm gel-purified, uncleaved transcription product, 1 mg/ml tRNA, 50 mM Tris, pH 7.5, and 10 mM MgClg. Reactions were started by addition of the MgCl^.

To stop the reaction, 5 /il aliquots were removed at designated time points, added to an equal volume of 7 M urea, 30 mM EDTA 40

pSSI and mutant leti I and II

aawaga product!:' le ; j 101 PM5 , f — 77 aio T LI M Xbm\ OMwaga nzzzzzzzzzB^, products; is 07

B 881 (wlldtyp*) M19/20 Au/y_yoii!i LC ?•

NA kcalAnol ~2000 11 mln

Figure 2. A. Maps of inserts in transcription plasmids. Large bold box represents sBYDV RNA sequence; shaded regions (flanking bases 310 and 89) are not required for hammerhead formation. Horizontal arrow marks the transcription start site; vertical arrowhead indicates self-cleavage site. Expected cleavage products of transcripts generated by T7 polymerase-catalyzed transcription of XJbal-linearized DNA are shown as bold lines below map. Sizes of transcripts (in nt) are indicated below each one. Specific changes in LI and L2a (shaded with diagonal lines) in mutant sets I and II are shown in panel B. Dashed lines indicate region (bases 30-63) of pSSl that was deleted in pM5. B. Schematic representations of mutants with base changes (outlined text) in the putative Ll-L2a helix. Mutants are grouped into sets I and II as described in the text. Only the sequences of LI and L2a are shown. Their positions in the hammerhead are shown in SSI (wildtype) at left. Empty box in mutant M5 represents deletion of L2a sequence. The name of each mutant (above) and free energies (below) of potential Ll-L2a helices are indicated. Helices were identified and free energies calculated by using the RNA Structure Editor (RNASE) computer program (23) which used the parameters in (40). NA = not applicable. The bottom line shows the half- (tj/a) of uncleaved RNAs calculated from graphs in Fig. 3. 41

and immediately frozen at -SO^C. Samples were thawed and

denatured by boiling for one min immediately prior to electrophoresis on 8% polyacrylamide, 7 M urea sequencing-type gels. Bands were cut out and radioactivity determined by liquid scintillation counting.

Synthesis of End-labeled RNA Each of the steps below was terminated by making solutions 50 mM in EDTA, extracting once with phenoliCHClg

(1:1), and ethanol precipitation in 2M ammonium acetate pH 6.2. Step 1: Unlabeled transcripts were synthesized as described above except that the reactions contained: 5 - 15 mg of linearized template, 0.5 mM of all four NTPs, 100 units RNasin, in a 100 ml final reaction volume. Step 2:

Transcription products were treated with 5 units calf intestinal phosphatase for 10 min in the absence of magnesium as in (20). Step 3: End-labeling reactions (50 ml) contained dephosphorylated transcript, 40 units RNasin, 100 mCi [a-®^P]ATP (3000 Ci/mmole), 10 units polynucleotide kinase, 50 mM Tris, pH 8.0, 10 mM MgClg. To minimize self-cleavage during labeling, MgClg was added last and reactions were incubated only 5 to 10 min before addition of EDTA to stop the reaction. Step 4: Uncleaved transcripts were purified by gel electrophoresis as already described. 42

Nuclease Digestion All nucleases except VI (Pharmacia, Milwaukee, WI) were from BRL (Gaithersberg, MD). Nuclease digestions were under the same conditions of temperature and solution used in the self-cleavage assays. Reactions (5 fxl) contained approx. 10,000 cpm per reaction of gel-purified, end-labeled uncleaved RNA, 1 mg/ml tRNA, 50 mM Tris, pH 8.0, 10 mM MgClg, and either

0.2 units VI nuclease, 0.1 units T2 nuclease, 0.2 units T1 nuclease, or water. Nuclease treatment was initiated 20 to 50 s after addition of MgClg. After 5 min at 37°C, reactions were terminated by adding an equal volume of 7M urea, 30 mM EDTA and freezing at -SO^C. Products were separated on 12% polyacrylamide gels containing 7M urea. Sequencing reactions (T1 and 0 M nucleases) under denaturing conditions (21, 22) and ladder formation by partial alkaline hydrolysis were performed by using a kit from BRL. 43

RESULTS

Self-cleavage of Wildtype and Mutant RMAs Transcripts from a plasmid (pPCSl) spanning the (+) strand cleavage site were shown previously to self-cleave (17). However, the rate of cleavage and the effects of a large number of extraneous bases were unknown. To allow transcription of an RNA that contained the hammerhead sequence with minimal vector-derived bases, 36 bases were deleted from pPCSl to create pSSl (Fig 2). The RNA transcript (SSI) from Xjbal-linearized pSSl contained only 3 vector-derived bases at the 5' end and at most 6 at the 3' end (Figs. 2A, 5). Optimal conditions for self-cleavage of SSI were 50 mM Tris-HCl, pH 7.5, 10 mM MgClg, at 37°C. The reaction showed a broad magnesium optimum, and little sensitivity to temperature or denaturation treatments such as boiling and snap cooling (4) prior to addition of Mg'^^ (data not shown). The cleavage products of all the transcripts were as expected: a 120 nt full-length fragment, a 101 nt 3' fragment, and a 19 nt 5' fragment (Fig. 3). The intensity of the full-length band decreased as the intensities of the fragments increased with time. Transcript SSI cleaved with an approximate half- of 1500 to 2500 min as calculated by extrapolation of several 44

SS1 M1 MS M1/M2 SSI

0 30 60 90 120 150 IBO 210 140 270 jOO

SSI I\/I19 M20 M19/20 min (•—•) 0 60 120 180 240 jOO J60 420 480 540 600 ^ MM M V 1.00 a-" ' SSl

0.50 va MÎ9/20

il-a M 20 0.10

(1

10 15 20 25 30 35 min.(

SS1 MS

3'F

• • ••• 5*

Figure 3. Self-cleavage of wildtype and mutant RNAs. Left: Autoradiographs of transcripts after incubation in self- cleavage conditions for times shown in graphs at right, and denaturing polyacrylamide gel electrophoresis. Mobilities of full-length (F) transcript, and 3' and 5' fragments are indicated. Right: Semilog plots of fraction of full-length transcript that remains uncleaved were calculated from radioactivity in bands of autoradiographs at left. In the middle graph, note that M19 and M20 are plotted against a different scale (bottom) than are SSI and M19/20 (top scale). SSI was included as a standard in all assays. 45

separate experiments (Figs. 2, 3). To determine if the slow rate of cleavage was due to formation of the alternative base pairing (Ll-L2a; Fig. 1), the potential to form this helix was disrupted by mutagenesis. Two three-membered sets of mutants were constructed (Fig. 2). Each set included transcripts with alterations to LI, L2a, or to LI and L2a on the same molecule such that potential base pairing, but not wildtype sequence was restored in the double mutant. In an additional mutant (M5), bases 30-63, including the L2a region, were deleted (Fig. 2). This prevented formation of the Ll-L2a helix, and resulted in a hammerhead similar in size to those of the sobemovirus satellites (virusoids), sTobRV (5), and sBYDV RNA (-) strand (17).

Mutant set I (mutants Ml, M2, and Ml/2), was designed to conserve the 100% G+C content of LI and L2a. Three bases were changed in either LI or L2a, or both LI and L2a. The half- lives of mutant RNAs Ml and M2, in which the proposed Ll-L2a base pairing was disrupted, were 310 and 48 min, respectively (Figs. 2, 3). The double mutant Ml/2 did not self-cleave detectably. In contrast to mutant set I, set II was chosen after exhaustive computer searches (23) of many sequences containing different mutations predicted that this set of mutations should form no significant unintended base pairing.

Mutant set II (M19, M20 and M19/20) was constructed by introducing only a single base change in either LI or L2a, or 46

both LI and L2a. Each of the single mutants, M19 and M20, cleaved hundreds-fold more efficiently than wildtype (Figs. 2, 3). Seventeen percent of M19 RNA remained uncleaved after 10 hr, presumably because this proportion of the molecules folded into an inactive conformation. The double mutant M19/20 cleaved 24- to 36-fold more slowly than either of the single mutants, but at least 10-fold faster than wildtype. Mutant MS also cleaved quite rapidly (Figs. 2, 3). The full-length

transcript (86 nt) and the 2' (67 nt) and 5' (19 nt) fragments migrated as expected.

Nuclease Sensitivity The next set of experiments employed structure-sensitive nucleases to determine whether the overall tertiary structure exists as predicted in Fig. IB. 5' end-labeled transcripts were partially digested with nucleases T2 (cuts all single- stranded bases [24]), T1 (cuts single-stranded guanosine nucleotides), or VI (cuts double-stranded and some base- stacked regions [25,26]), under conditions identical to those used in the cleavage assays. To avoid misleading secondary cleavages, enzyme concentrations were adjusted so that the RNA was cleaved less than once per molecule, on average. The digestion times were short enough (5 min) to allow analysis of the RNA structure before it self-cleaved. The possibility that the nuclease digestion reveals structures of the cleavage 47

products is eliminated in all bands greater than 19 nt long, because self-cleavage of the 5' end-labeled RNA would remove

label from the 2' product. Products of nucleolytic digestions of the wildtype and all the mutant transcripts were separated on denaturing gels alongside cleavage products obtained with nucleases T1 and ^ M (cuts after A and U residues) under fully denaturing conditions (Fig. 4). A wide variation in nuclease susceptibility of phosphodiester bonds under self-cleavage conditions relative to denaturing conditions indicated regions of strong structure. Nuclease T2 cut single-stranded G residues only weakly, compared to its activity on the other three bases. The prominent 19 base band in all samples (including those lacking nuclease) except Ml/2, that were digested in self-cleavage conditions is the 5' self-cleavage fragment. It migrated exactly as expected on the basis of the

T1 and

^ccnoccxa>c»o>o>ooooo^ooDcoco>OD>oc>c:>caoooonoc»ooD^Co

' tn , • ; I i : i(i Hfîjs * I » »•• • M U) DoeoD fioocacî < I Hf I'.'. { u III11 J «•I « (; M e #*#####«**##### 00000 00000 % Hi CO ' m a »«UUttllll|lUBUKI •' *$% * It ooooc) 60^00 »« . I • 1 III 111 t 1 It • I I I I I II II t nn(IIiir.'!(iiumr.ît!i!Ri ### M'W# « til I : lilt I I tli llllf #*«' M # il o@gSc^ • I I I # 4 vo • If II I (III Ujl ' I II I » \ Hit l« if I XVVVVWVV VW'AWVWVW.m.VA vv V V, ooeoo 00^00 *« .' ##* # I It IV: ti I 11 n mt m . 0 •••III ••• •» / t I t I I : M : i i : - Oj.rccoocoo>c»o>o>ooQoooncococo>ao>ooc>co»^oo>o»:^'' Co util tê II t « 4 $ I II I ' lilt II lltlftli th ^ ^000 ^o@^

ft- /• / têfi / // t Ê t ! t I I : • • f I i ll Hi iiitimitiiiiiiaiiiii • «•«•.1111% • « tttt •• tit lit I s 00000 0@S2)C^ @0g)5)0 Figure 5. Locations and relative susceptibilities of nuclease cleavage sites derived from Fig. 4. V = VI site, T = T1 or T2 site. The size of the V or T is proportional to the intensity of the band at that site. Because the digestion pattern of helix 3 was virtually identical in all constructs, it is shown only for wildtype (SSI) RNA. Mutant sets I and II are indicated. Vector-derived bases are in lower case. Mutations are in outlined font. For ease of comparison, the sequences are drawn as hammerheads regardless of whether nuclease sensitivity supported the existence of this structure. 51

s /uaWI? Ï1.VV ...ua pcuagaaaugaa<^caccua a -ê "a "a a a CpA c ua u-%, :s- c râ Q. % Ag JÂ" °- ; JÂ" @ f i °*%s-^ Ac-Q^c-q y: ^U-AÔÏ c-g c-q' c-q c-q tc-q, c-g o & V SSI M1 M2 Ml/2 B JQ. ..aua ..MA ..auaW'''^ ...UÀ ..aja juf - ..ua%# m a c% i V •s.,.s- L a-u \-u: /\ iê:k % ^c : i- t M5 %C.Q° '^AC.G° 8:g 8:2 ~8# G-% _C c-q c-qU-AÛO c-gc-q c-q c-q V <& Mig M20 Mig/20 52

more susceptible to single-strand-specific nucleases in most of the single mutants in which Ll-L2a base pairing was expected to be disrupted (M2, M19, M20, M5), than in those in which Ll-L2a base pairing was predicted (SSI, Ml/2, M19/20). This is especially noticeable when the T1 digests under cleavage and denaturing conditions are compared across all the RNAs (Fig. 4). Concomitantly, in the double mutants, nucleotides in L2a were digested somewhat more readily by nuclease VI than were the L2a bases in the single mutants. In all RNAs, loops L2b, L2c, and L2d were highly susceptible to single-strand-specific nucleases. Predicted helix H2a was generally more susceptible to VI, with the exception of Ml in which it appeared highly single-stranded. H2b and H2c were not cleaved well by any nucleases, except for H2b in M2 and Ml/2 which were cleaved unexpectedly by single strand-specific nucleases. The conserved CUGANGA loop was sensitive to single-stranded nucleases in all but Ml. Thus, the nuclease cleavage pattern of Ml RNA was radically different from that expected, unlike the patterns generated by the other transcripts. 53

DISCUSSION

Effects of Mutations on Self-cleavage Rate By showing that disruption of the potential Ll-L2a helix dramatically increased self-cleavage activity and that restoration of the potential helix with a new sequence reduced activity, we have provided strong evidence for the existence of the helix in the self-cleavage structure of sBYDV (+) strand RNA. Because disruption of this base pairing is predicted to favor formation of a hammerhead RNA, the results also support the proposal (3,4) that the hammerhead is indeed the self-cleavage structure. That different disruptions to the Ll-L2a helix have different cleavage rates implies that other unpredicted folding may be taking place. It is noteworthy that the rates of cleavage of SSI and mutants Ml/2 and M19/20 are inversely proportional to the calculated stability of helix Ll-L2a (Fig. 2). Thus, we propose that an equilibrium exists between the self-cleaving hammerhead and the noncleaving stacked helix structure. Even if the latter is favored, any brief formation of the hammerhead would allow irreversible cleavage. This explains the slow but steady rates of cleavage of SSI and M19/20. In Ml/2, apparently the helix is so strong that the hammerhead 54

never forms. The situation is somewhat analogous to the self- cleavage structure of virus (HDV) RNA. Although HDV RNA shows no evident relationship to a hammerhead, it seems to form a structure in which one helix inhibits self- cleavage (27). Denaturing conditions such as <1 mM Mg^"*", high

temperature, or presence of urea or formamide accelerate the cleavage rate (28). However, unlike HDV RNA, the rate of sBYDV RNA cleavage does not increase at elevated temperatures such as 50°, nor does it increase at low magnesium

concentrations, or after boiling and snap-cooling. Perhaps this is because the Ll-L2a helix is the most stable in the molecule, and helices required for hammerhead formation (e.g., HI) would denature before Ll-L2a. Of the known hammerheads (5), only the (+) and (-) strands of sLTSV have potential base pairing between the distal loops of helices 1 and 2 like (+) sBYDV. However, in that case, such base pairing is not predicted to be as stable as that involved in hammerhead formation.

Nuclease Sensitivity The self-cleavage assays of wildtype and mutant sequences do not distinguish between the following two possibilities: (i) the three stacked helices form as proposed in the wildtype

ribozyme with no disruption of the hammerhead (HI, H2, H3) helices (Fig. lb), and (ii) the Ll-L2a base pairing causes 55

major disruption of the hammerhead by preventing one or more of the hammerhead helices from forming and perhaps allowing many new helices to form. To distinguish between these possibilities, the RNAs were probed with structure-sensitive nucleases. Nuclease sensitivity of the wildtype self-cleavage structure (SSI, Fig. 5) is, for the most part, consistent with the proposed structure. However, other than H3, predicted helices are not clearly shown by VI digestion. This may be due to the wide variation in ability of nuclease VI to act on certain helical regions (25, 26, 29). Certain regions (e.g.,

putative helices H2b, H2c) were not cleaved by VI nuclease, presumably due to steric hindrance in which the enzyme could not interact with nucleotides in the interior of the RNA molecule (30). In addition, some regions predicted to be single stranded were partially cut by VI, perhaps due to the ability of VI to cut at nonpaired but stacked bases (26, 29).

Helix HI was cleaved by single- and double-strand-specific nucleases, perhaps because it "breathes" during the nuclease treatment owing to its low stability (AG = -5.3 kcal/mol).

Thus the structural information derived from the VI digests is somewhat limited. Single-strand-specific nuclease digestions gave more clear-cut results than the VI digestions. The observation that loops L2b, L2c, L2d, and the conserved CUGANGA loop (with 56

one exception) remained sensitive to nucleases T1 and T2 provides strong evidence that the entire vertical stem-loop 2 region forms as in the proposed structures (Fig. 1). Unexpected alterations to structures of RNAs in mutant set I were observed (Fig. 5). The most unexpected changes were in Ml which explains why it cleaves so much more poorly than the other mutants that disrupt Ll-L2a base pairing. Indeed, a computer search (23) revealed that many new helices may form as a result of the three base changes in both Ml and M2 (not shown), perhaps explaining why neither mutant cleaved as well as M19 or M20. Although the nuclease sensitivity results do not support the structure of Ml as drawn, the functional assay shows that disruption of Ll-L2a still allowed the RNA to form a functional cleavage structure at least a portion of the time, causing it to cleave more readily than wildtype. Unlike Ml, the nuclease sensitivity of M2 supports the existence of a functional hammerhead. Note the striking increase in sensitivity to single-stranded nucleases of loop LI in M2 versus SSI. The disrupted helix H2b in M2 is not predicted to be essential for a functional hammerhead. Thus, it is not surprising that this RNA cleaves much more rapidly than Ml.

The base changes in mutant set II resulted in no significant changes in nuclease sensitivity outside the predicted regions, with the exception of slight unexpected cuts in LI by VI in M19 and M20. Finally, M5, in which most 57

of the entire vertical stem including L2a was deleted, behaved as predicted. This is the first use of structure-sensitive nucleases to show that hammerhead ribozymes indeed fold as predicted by the model of Forster and Symons (4).

Comparison with Other Hammerheads Even the fastest-cleaving mutants do not cleave as rapidly as conventional hammerhead ribozymes such as sBYDV (-) strand (17) or sTobRV (+) strand (8). This may be due to the other deviations from consensus (Fig. 1). The AUA instead of a GUC at the 5' side of the cleavage site is not likely to significantly reduce the rate of cleavage relative to other hammerheads. Ruffner et al. (14) showed that replacement of the G or the C of the GUC with As at either position had only minimal effect on cleavage by an ASBV-derived ribozyme. However, when the bases of the proximal base pair of helix 2 were changed to C and A, analogous to C-24 and A-73 in sBYDV RNA, cleavage by the ASBV-derived ribozyme was reduced over one hundred-fold. Given this and similar observations with sLTSV mutagenesis (15), it is perhaps surprising that mutants M19, M20, and M5, all of which contain unpaired bases C-24 and A-73, cleave as well as they do. The results do not prove the existence of the three stacked helices, but they support this model (Fig. IB). This structure fits the definition of a pseudoknot (31); however. 58

it has three instead of the usual two coaxially stacked helices. It is possible that helix HI would be pulled apart due to torsional constraints when the Ll-L2a helix forms, but the nuclease sensitivity does not correlate with this. It is possible that yet another conformation exists within the context of the full-length satellite RNA. Other examples of alternative conformations for hammerhead domains have been reported. The self-cleavage sequences in the (+) and (-) strands of sLTSV may exist in equilibrium between the hammerhead and as part of the rod-shaped full-length molecule (4). The self-cleavage sites in newt satellite 2 transcripts (32) and ASBV RNA (33) may fold differently in the context of a dimer of the full satellite RNA, compared to the sequence context of the isolated hammerhead. These RNAs have an extremely short (2 or 3 base pair) helix 3. In the minimal hammerhead context, cleavage can occur by a bimolecular double-hammerhead structure (34), whereas dimeric satellite RNA self-cleaves mostly via single-hammerheads (32, 33). Thus, it will be interesting to determine the cleavage rate of dimeric sBYDV RNA. However, the sBYDV situation is quite different from the newt and ASBV in that we observed intra- hammerhead conformational changes rather than inter-hammerhead base pairing. Due to the length of helix 3, we have no reason to invoke a bimolecular double-hammerhead mechanism (35). The sLTSV alternative conformation is also unlike sBYDV in that it 59

involves base pairing to regions outside the hammerhead domain (4). Finally, alternative conformations have also been observed when the ribozyme and substrate are located on separate molecules (13, 16). Because the Ll-L2a helix is the strongest uninterrupted helix in the entire satellite, and the strands are in such close proximity, we predict that the helix exists even in the context of full-length and multimeric RNAs. The biological role of the stacked-helix structure is unknown. An intriguing possibility is that the Ll-L2a helix may form left-handed Z-RNA (reviewed in 36). It contains the alternating G-C sequence required for Z-RNA. Also, torque may be imposed by the strands of the distal end of putative helix HI on helix Ll-L2a because these strands would be expected to join LI at opposite sides of the one-half turn, 5-base Ll-L2a helix. This may confer negative supercoiling that is known to favor Z-DNA (37). Z-RNA has been shown to exist in living cells (38, 39), but its identity and biological role are unknown. Regardless of whether Z-RNA forms, it is possible that the stacked helix structure performs a function different from self-cleavage. Nearly half (153 nt) of the 322 nt sBYDV RNA is involved in formation of either the (+) strand or complement of the (-) strand cleavage structure. This leaves only 169 nucleotides divided into two tracts of 104 and 65 bases (17) to perform all the other functions of the satellite RNA, including an , an origin of assembly 60 and probably other functions. Thus, it could be envisioned that sequences within the self-cleavage structure contribute to one of these other functions, compromising the ability to self-cleave in the process. The stacked-helix may serve as a molecular switch to modulate the transition between these two functions. 61

REFERENCES

1. Symons, R. H. (1991) Molec. Plant-Microbe Interact. A, 111-120. 2. Bruening, G., Passmore, B. K., van Toi, H., Buzayan, J. M., & Feldstein, P. A. (1991) Molec. Plant-Microbe Interact. A, 219-225. 3. Forster, A. C., & Symons, R. H. (1987) 50, 9-16.

4. Forster, A. C., & Symons, R. H. (1987) Cell 49, 211-220.

5. Bruening, G. (1989) Methods Enzvmol. 180, 547-558. 6. Hutchins, C. J., Rathjen, P. D., Forster, A. C., & Symons, R. H. (1986) Nucleic Acids Res. 14, 3627-3640.

7. Epstein, L. M., & Gall, J. G. (1987) Cell 48, 535-543. 8. Haseloff & Gerlach (1989) Gene 82, 43-52.

9. Uhlenbeck, O. C. (1987) Nature 328, 596-600. 10. Haseloff, J., & Gerlach, W. L. (1988) Nature 334, 585-591. 11. Koizumi, M., Iwai, S., & Ohtsuka, E. (1988) FEES Letters 228, 228-230. 12. Ruffner, D. E., Dahm, S. C., & Uhlenbeck, O. C. (1989) Gene 82, 31-41. 13. Fedor, M. J., & Uhlenbeck, O. C. (1990) Proc. Natl. Acad. Sci. USA 87, 1668-1672. 14. Ruffner, D. E., Stormo, G. D., & Uhlenbeck, 0. C. (1990) Biochemistry 29, 10695-10702. 15. Sheldon, C. C., & Symons, R. H. (1989) Nucleic Acids Res. 14, 5679-5685. 16. Heus, H. A., Uhlenbeck, O. C., & Pardi, A. (1990) Nucleic Acids Res. 18, 1103-1108. 62

Miller, W. A., Hercus, T., Waterhouse, P. M., & Gerlach, W. L. (1991) Virology 183, 711-720.

18 Kunkel, T. A., Roberts, J. D., & Zakour, R. A. (1987) Methods Enzvmol. 154, 367-382. 19 Sambrook, J., Fritsch, E. F., & Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual. Second edition. Cold Spring Harbor Laboratory Press, Cold Spring Harbor.

20 Conway, L., & Wickens, M. (1989) Methods Enzvmol. 180, 369-379.

21 Donis-Keller, H., Maxaiti, A. M., & Gilbert, W. (1977) Nucleic Acids Res. 8, 3133-3142. 22 Simoncsits, A., Brownlee, G. G., Brown, R. S., Rubin, J. R., & Guilley, H. (1977) Nature 269, 833-836. 23 Cedergren, R., Gautheret, D., Lapalme, G., & Major, F. (1988) CABIOS 4, 143-146.

24. Vary, C. P. H., & Vournakis, J. N. (1984) Nucleic Acids Res. 12, 6763-6778.

25. Lockard, R. E., & Kumar, A. (1981) Nucleic Acids Res. 9, 5125-5140. 26. Lowman, H. B., & Draper, D. E. (1986) J. Biol. Chem. 261, 5396-5403.

27. Perotta, A. T, & Been, M. D. (1991) Nature 350, 434-436.

28. Rosenstein, S. P., & Been, M. D. (1990) Biochemistry 29, 8011-8016. 29. Puglisi, J. D., Wyatt, J. R., & Tinoco Jr, I. (1988) Nature 331, 283-286.

30. Knapp, G. (1989) Methods Enzymol. 180, 192-213. 31. Pleij, C. W. A., Rietveld, K., & Bosch, L. (1985) Nucleic Acids Res. 13, 1717-1731. 32. Epstein, L. M., & Pabon-Pena, L. M. (1991) Nucleic Acids Res. 19, 1699-1705. 33. Davies, C., Sheldon, C. C., & Symons, R. H. (1991) Nucleic Acids Res. 19, 1893-1898. 63

34. Forster, A. C., Davies, C., Sheldon, C. C., Jeffries, A. C., & Symons, R. H. (1988) Nature 334, 265-267. 35. Sheldon, C. C., & Symons, R. H. (1989) Nucleic Acids Res. 14, 5665-5677. 36. Tinoco Jr., I., Davis, P. W., Hardin, C. C., Puglisi, J. D., Walker, G.T., & Wyatt, J. (1987) Cold Spring Harbor Svmp. Quant. Biol. 52, 135-146. 37. Singleton, K., Klysik, J., Stirdivant, S. M., & Wells, R. D. (1982) Nature 299, 312-316. 38. Zarling, D. A., Calhoun C. J., Hardin, C. C., & Zarling, A. H. (1987) Proc. Natl. Acad. Sci. USA 84, 6117-6121. 39. Zarling, D. A., Calhoun C. J., Feuerstein, B. G., & Sena, E. P. (1990) J. Mol. Biol. 211, 147-160. 40. Freier, S. M., Kierzek, R., Jaeger, J. A., Sugimoto, N., Caruthers, M. H., Neilson, T., & Turner, D. H. (1986) Proc. Natl. Acad. Sci. USA 83, 9373-9377. 41. Schulz, V. P., & Reznikoff, W. S. (1990) J. Mol. Biol. 211, 427-445. 64

PAPER II. NONCONSENSUS BASES IN THE sBYDV (+ ) RNA AFFECT THE RATE OF HAMMERHEAD FORMATION AND SELF- CLEAVAGE EFFICIENCY 65

Nonconsensus bases in the sBYDV (+) RNA affect the rate of hammerhead formation and self-cleavage efficiency

Silver, S.L., M.S. Miller, W.A., Ph.D.

From the Department of Plant Pathology, Iowa State University, Ames, lA 50011 66

INTRODUCTION

Self-cleavage of a small molecular weight RNA was first observed in the satellite of tobacco ringspot virus (sToBRV) (+) strand (Prody et al., 1986). The self-cleavage reaction has a requirement for Mg^"^, and unlike cleavage by RNase P

(Baer et al., 1990) and pre-mRNA splicing (Ruby & Abelson, 1991), does not require protein at physiological conditions. Demonstration of self-cleavage in the satellite RNAs of luteo- and sobemoviruses (Bruening, et al., 1991; Symons, 1991b), in avocado sunblotch viroid (ASBVd; Hutchins et al., 1986), peach latent mosaic viroid (Hernandez & Flores, 1992; PLMVd) and carnation stunt associated viroid or satellite RNA (CarSAV; Hernandez et al., 1992) and in the transcripts of satellite II DNA of newt (Epstein & Gall, 1987) indicate the broad occurrence of this phenomenon. These self-cleaving RNAs are believed to replicate by a rolling circle mechanism which results in long multimers that cleave into circular and/or linear monomers (Bruening, et al., 1991; Symons, 1991a). The nucleotides spanning the cleavage site of the above RNAs fold into a structure referred to as a hammerhead. The hammerhead consists of three helices, whose sequences are not conserved and a central, predominately single-stranded region composed 67

of highly conserved bases (Forster & Symons, 1987b; Forster & Symons, 1987c). Extensive mutagenic analysis of the hammerhead confirmed the importance of the conserved nucleotides and the structure during cleavage (Miller & Silver, 1991; Ruffner et al., 1989; Ruffner et al., 1990; Sheldon & Symons, 1989). The first satellite RNA of a luteovirus was recently discovered associated with barley yellow dwarf virus, serotype RPV (sBYDV RNA) (Miller et al., 1991). sBYDV RNA is a 322 nucleotide long RNA that is encapsidated primarily as a linear molecule. Both the (+) and (-) strands of sBYDV RNA contain the hammerhead motif (Miller et al., 1991). The (-) strand cleavage site closely resembles the hammerhead of the sToBRV (+) strand cleavage site (Bruening, 1989); whereas, the sBYDV (+) strand hammerhead has several additional structural features not present in most other hammerheads (Fig. 1). First, AUA is the trinucleotide on the 5' side of the cleavage site in sBYDV (+) RNA, while all but one other known hammerhead structure has the sequence GUC. The one exception is the virusoid of lucerne transient streak virus (sLTSV), which has the trinucleotide GUA in this position. Mutagenic analysis, however, revealed that the trinucleotide need not be totally conserved for efficient cleavage (Ruffner et al., 1990; Sheldon & Symons, 1989). Second, mutagenic analysis and nuclease sensitivity probing of sBYDV (+) RNA 68

demonstrated that stems-loops I and II of the hammerhead have the potential to basepair and form an alternative structure containing three coaxially stacked helices (Miller & Silver, 1991).

dMvags 5* 310uauuucquaqapâacaa^c 9' I I I I I I I I I I t TC lloiie.I I I I P% K C_Q|G=SI

rou-A dMvaaa il® 310 iLia aaa e UAUUUCaUQQA Wo-Ç 3' AUdAAO^CACCU iM\ lAZPU-AlUQal & "UqqU Stacked helices B (pseudoknot) Hammerhead

Figure 1. Proposed alternative structure for the self- cleaving ribozyme in sBYDV (+) strand RNA. Boxed bases are conserved among all hammerheads (Bruening, 1989). Bases that differ from consensus are shown in outlined text. Numbering of nucleotides is based on the full-length satellite RNA. The three major helices of hammerhead are labeled HI, H2, and H3. Loops present in the hammerhead are prefixed with L. Bases in loops LI and L2a expected to form the additional helix are in italics. Panel A is the hammerhead of sBYDV (+) RNA drawn without the putative basepairing between LI and L2a. Panel B is the alternative structure that forms when LI and L2a interact. 69

The third structural feature of sBYDV (+) RNA not generally found in hammerhead ribozymes, and the one addressed by the current study, is the occurrence of one unpaired cytosine (C-24) immediately 3' to the conserved single-stranded CUGANGA sequence and an unpaired adenine (A-73) directly 5' of the conserved GAAAN sequence. Only two other self-cleaving RNAs have unpaired nucleotides corresponding to bases 24 or 73 of sBYDV RNA (+) hammerhead. In sLTSV the (+) strand hammerhead has a uracil corresponding to C-24 (Forster & Symons, 1987b) and recently the CarSAVd (-) strand was found to have an unpaired cytosine and adenine located at exactly the same sites within the hammerhead as the sBYDV (+) RNA (Hernandez et al., 1992). This study determined the effects of deleting nucleotides C-24 and A-73 on the rate of cleavage and hammerhead formation. The data are consistent with the hypothesis that formation of alternative structures at the cleavage site disrupt the integrity of the hammerhead, thus regulating the rate of cleavage. 70

MATERIALS AMD METHODS

Synthesis of Hildtype and Mutant

Self-cleavage Structures RNA molecules were synthesized by in vitro transcription of sequences in pGEM3Zf(-) (Promega, Madison, WI). Plasmid pSSl contained the wildtype (+) strand sBYDV hammerhead sequence: bases 310 to 322 and 1 thru 89 of the satellite sequence (Figs. 1 & 2). Cleavage occurs between bases 322 and 1. Vector bases GGA and CUCUAG were present at the 5' and 3' ends, respectively of wildtype and mutant transcripts from Xbal-linearized pSSl. Each mutation was introduced by the two-step PGR method of Herlitze and Koenen (1990). The following derivatives of pSSl were made with the indicated primers that are complementary to the satellite RNA (+) sequence. pM3; 5' CGCGGATAC_*TCGTCAGACAG 3'; pM4:5' GTGGATTTC_*GTATCTATTTG 3' and pM20: 5' TTCTAGTCCGCTCGGATACTCG 3'. Underlining indicates base changes and * indicates base deletion. The first PGR involved 25 cycles of amplification (Amplitaq, Perkin-Elmer-Cetus) using the M13 reverse primer and the mutagenic oligomer. The purified double-stranded PGR product and the M13 forward primer were used for the second PGR using the same target DNA (pSSl). Products from the 71

310 i LI L2a T7 80 Xbal % Cleavage -fl products: .3' 19 101

Figure 2. Maps of inserts in transcription plasmids. Large bold box represents sBYDV RNA sequence; shaded regions (flanking bases 310 and 89) are not required for hammerhead formation. Horizontal arrow marks the transcription start site; vertical arrowhead indicates self-cleavage site. Expected cleavage products of transcripts generated by T7 polymerase-catalyzed transcription of Xbal-linearized DNA are shown as bold lines below the map. Sizes of transcripts (in nt) are indicated below each one. LI and L2a are indicated by hatched regions. Sites of mutations are indicated. second PGR were digested with BamHI/Hindlll, gel-purified and ligated into BamHI/Hindlll-cut pSSl. Synthesis of the double mutant pM3/4 proceeded by first constructing pM3 followed by a second PGR mutagenesis that used pM3 instead of pSSl as the target DNA. The triple mutant, pM3/4/20, was synthesized using pM3/4 as the target DNA. All base deletions and alterations were confirmed by double-stranded DNA sequencing (Promega). During the sequencing of the mutations, we discovered a non-templated addition of an adenosine at the 5' side of the site to which M20 annealed. The PCR-generated conversion was circumvented by using "Hot Tub" DNA polymerase (Amersham) for the first PGR amplification and Taq polymerase 72

for the second amplification. Generally, 50-75% of resulting clones had the correct mutation without the base conversion. In our nomenclature, a plasmid and its transcript are named similarly except that the "p" of the plasmid name is omitted in the transcript.

Self-cleavage Assays [a-^^P] GTP-labeled RNA was transcribed from Xbal- linearized plasmid DNA as described by Milligan and Uhlenbeck (1989). Approximately 0.1 mg/ml of T7 RNA polymerase isolated as described previously was used per standard transcription (Grodberg & Dunn, 1988; Zawadski & Gross, 1991). Full-length transcripts were purified from an 8% polyacrylamide, 7M urea gel (Sambrook et al., 1989). Self-cleavage assays (80 ^ll) contained approximately 20,000 counts per minute (cpm) of gel-purified, uncleaved transcript, SOmM Tris, pH 7.5 and lOmM MgClg. The assay was initiated by addition of the MgClg. The reaction was incubated at 37° C and at indicated time points, 8

/il aliguots were taken and added to an equal volume of "solution D" (10 M urea and 30 mM EDTA; BRL). Samples were denatured by boiling for one minute immediately prior to electrophoresis. Bands corresponding to the uncleaved transcript and cleavage products were quantified using Phosphorimage analysis (Molecular Dynamics). The half-life (T.) of the uncleaved transcript was determined by computing 73

the slope of the best-fit line of the cleavage data and solving for the X-intercept using the software CoPlot and CoStat.

Synthesis of End-labeled RNA

and Nuclease Digestion End-labeled uncleaved transcripts were prepared and gel-purified as described previously (Conway & Wickens, 1989; Miller & Silver, 1991) with the following modifications. Unlabeled RNA was transcribed according to Milligan (1989) with T7 RNA polymerase isolated in our laboratory and products treated with 5 units of calf intestinal phosphotase for 20 minutes. Nuclease digestion was performed essentially as described previously (Miller & Silver, 1991) with the following changes. Nuclease digestion assays contained 20,000 cpm per reaction of gel-purified, end-labeled uncleaved RNA, 1 mg/ml tRNA, 50 mM Tris, pH8.0, 10 mM MgClg and either 0.1 U VI

nuclease, 0.1 U T1 nuclease, 0.1 U T2 nuclease or water. The nuclease reaction was initiated by MgClg, incubated for 10

minutes and quenched by addition of solution D. For each reaction, the RNA was denatured by boiling for 1 min and then separated by electrophoresis on a 12% polyacrylamide, 7 M urea gel. Cleavage products were visualized by autoradiography. 74

RESULTS

Self-cleavage of Mutant RNAs RNA transcribed from Xbal-linearized pSSl (Miller & Silver, 1991) produced a 120 nt long molecule that in the presence of 10 mM MgClg and 50 mM Tris, pH 7.5, self-cleaves

to form a 101 nt 2' fragment and 19 nt 5' end fragment. The Tx/2 of the uncleaved transcript determined from three separate

assays ranged from 1400 to 2100 min (Fig. 3). Similar rates of cleavage were observed previously (Miller & Silver, 1991). The unpaired cytosine (C-24) or the unpaired adenosine (A-73) were deleted from pSSl resulting in pM3 and pM4, respectively. The Ti/2 of M3 and M4 calculated from three separate assays were not significantly different from the T1/2 of SSI (Fig. 3 &

4). The double mutant M3/4, however, had a much shorter T^/g

(39 min) than either of the single mutants or the wildtype sequence (Fig. 3 & 4). In all transcripts with the M4 mutation, an unexpected C-83 to A-83 transversion was induced by the PGR mutagenesis. The location of the transversion is in stem III which is the most stable and least likely affected of the three helices. Exhaustive computer searches revealed little potential for alternative basepairing due to the transversion. Figure 3. Self-cleavage of wildtype and mutant RNAs. Left: Autoradiographs of transcripts after incubation in self- cleavage conditions for times shown in graphs at right, and after denaturing polyacrylamide gel electrophoresis. Mobilities of full-length (F) transcript, and 3' and 5' fragments are indicated. Right; Plots of fraction of full- length transcript that remained uncleaved were calculated from radioactivity in bands of autoradiographs at left. Ê

• •" Wiif f 5

fraction uncleaved Figure 4. Autoradiographs after incubation of wildtype and mutant RNAs with the indicated nucleases. Products were analyzed on 12% polyacrylamide, 7 M urea gels. Lanes are labeled for the enzyme treatment, except that lanes marked N had no enzyme added and L indicates partial alkaline hydrolysis ladder. ^ indicates ^ M digest which cut A residues slightly more than U's. SEQ indicates nuclease digestions under denaturing (sequencing ) conditions. Other digests were in self-cleavage conditions. Readable sequences of SSI are indicated alongside the autoradiographs. Left side of gel contains position of the CUGANGA region; whereas, the right side of gel, positions of bases in LI (lower set) and L2a (upper set) are indicated. uouoy. çuoup mmninnnn i ti n I I I llir orà k. •riHT'rT""!'!'!!'!' V • r I 1 I II l ! Spp.'l ) i 'M ' I I I I M I III 9 1 V;i # ## #*#' III i Ill III 2Ë| I •• I## « ' I #*## ' • '' • M II l»l I I # «Il I II M III ' I

ypui^ (juDi^ RtMiMiililillinini I II •« SJD<(^ DUOUO m tiinii! iiini i î ; ! "t CO H . • / III r" 1 • • i—MmimBinwmmmi ii w# f t-1 Hi m um iflONH I I'll 0 It t III! II II I## I I %\\\ I*## III I t £11)) n i^J) )j J J 1111 III • > n M 191 II E f ' # I i'tÊ t: f' I it mi >^i , ». 1,1 ,, # a 00 ^jdC^y jjuoi^ youoy ououo w wwwv \ \ un » « ' *•' 1 • ; 1 I III * •••••Mfinmn• M «»• » • M9 * f• MmU' I I II I Mill I# I I nil I II II# # t #•« '( • ( III lillU III I III I _ II I ' I ! - Il II III M • T 'i#y 11 « z M •« ' t*-< «41 'I# *## ' »«' Mill â » • 0

To look further at how the mutants M20 and M3/4 act to enhance hammerhead formation by disrupting inactive, alternative structures, the triple mutant M3/4/20 was constructed. The first attempt at constructing the mutant by PGR using pM3/4 DNA resulted in an unintended G to A transition at G-44. The half-life of the resulting M3/4/20a transcript was approximately 570 min which is significantly faster than SSI but not nearly as fast as would be expected from earlier M20 and M3/4 cleavage data. A modified PGR mutagenesis protocol which uses "Hot Tub" polymerase resulted in the M20 mutation without the PGR-generated transition. The M3/4/20 transcript had a half-life of approximately 6 min which is not significantly different from the 5 min half-life of the M20 transcript determined previously (Miller & Silver, 1991).

Nuclease Sensitivity To determine the tertiary structure of each RNA during cleavage, structure-sensitive nucleases were employed. 5' end-labeled transcripts were partially digested with nucleases T2 (cuts all single-stranded nucleotides; Vary & Vournakis, 1984), T1 (cuts single-stranded guanosine nucleotides), or VI (cuts double-stranded and some base-stacked regions; Lockard & Kumar, 1981). For each enzyme, the reaction was performed in standard cleavage conditions (10 mM MgGlg and 50 mM Tris, 80

pH 7.5 at 37° C). The digestion times were sufficiently short

(10-15 min) to study the RNA structure before complete cleavage. Products of nucleolytic digestions of the wildtype and all the mutant transcripts were separated on denaturing polyacrylamide gels along with the products obtained from nucleases T1 and ^ M (cuts after A and U bases) performed in denaturing conditions. Products were visualized by autoradiography (Fig. 4). The prominent 19 base band in all samples carried out in self-cleavage assay conditions (including the no enzyme control) with the exception of SSI and M4, is the 5'-end cleavage product. It migrated as expected on the basis of the T1 and 0 M sequencing ladders. Minor, nonspecific RNA hydrolysis was detected in the no enzyme control lane; none of which was as significant as the nuclease-specific products. A compilation of all sites and frequency of nuclease digestion was constructed for each RNA (Fig. 5). In all transcripts, including ones containing the PCR-generated base conversion at position 88, helix 3 has a strong degree of double-stranded character as shown by the frequency of VI cutting and the absence of either T1 or T2 digestion. Helix I, on the other hand, showed a high degree of single-stranded character as evidenced by the frequency of T1 and T2 digestions. When all transcripts are compared at the sites corresponding to LI and L2a, mutations expected to directly 81

disrupt basepairlng (M3/4/20 and M3/4/20a) show a higher degree of single-strandedness than transcripts expected to maintain this helix (SSI, M3, M4, and M3/4). Furthermore, M3 and M3/4 have an intermediate level of susceptibility to Tl and T2 digestion in these loops (Fig 4, 5). In all the mutants, loop L2d was highly susceptible to single strand- specific nucleases. Loop L2c was generally observed as single-stranded in the mutants with the exception of M3/4/20 and M3/4/20a which showed insensitivity to all enzymes. Loop L2b was highly susceptible to single-strand-specific nucleases in all the mutants but M3. Helices H2a and H2c were generally not susceptible to nuclease Tl or T2; whereas this region was highly susceptible to these enzymes in both M3/4/20 and M3/4/20a. The conserved CUGANGA single-stranded region was cut as expected, with the exception of M3/4/20 and M3/4/20a which became highly resistant to nuclease digestion. Clearly, the most significant change in cleavage pattern when all mutants were compared occurred in the CUGANGA region (Fig. 4 & 5). Figure 5. Locations and relative susceptibilities of nuclease cleavage sites derived from Fig. 4. V = VI site, T = T1 or T2 site. The size of the T or V marker is proportional to the intensity of the band at that site. Helix III digestion pattern is nearly the same for all transcripts; therefore, it is shown only with the SSI map. The sequences are drawn as hammerheads regardless of whether nuclease sensitivity supported the existence of the structure. Base conversions are in italics. Base deletions are depicted by empty rectangles. ve/tftn at/*lt\/i n!°n

gW-n D-OL )

ovo-0 -w'OOO Ow nao Jf ''iTi y? l^vXS . .oyoyvnv S.Vl' '( >/' 't ' 'f

m BM ISS M& "tfga 8-|< îV'^-n es 4 Q-0_ J)-0_ /% i I -n O', ff A 4^5 , YOPqYP^YYPOY^YW w [/i^SîjA^^WAnPAOT/STo®"'".» m 1

E8 84

DISCUSSION

Effects Of Mutations on Self-cleavage Rate Fedor and Uhlenbeck (1992) outlined a kinetic pathway for an intermolecular reaction involving a ribozyme and a substrate RNA (Fig. 6). According to this model, cleavage is limited by either formation of an active structure, i.e., a hammerhead or by the rate at which products are released from the ribozyme. kj describes the rate at which the two RNAs

interact and fold into a hammerhead, kg is the actual clevage rate and kscharacterizes the rate of product release. In general, the kinetic pathway proposed for the bimolecular reaction can also describe what occurs in the intramolecular cleavage reaction, except that the latter is zero order and the bimolecular reaction is first order. Therefore, any mutation that affects the cleavage rate of the sBYDV (+) RNA hammerhead can be described by the reaction rates in the Fedor and Uhlenbeck model. Any alternative structures within the hammerhead should be discussed relative to how they affect the formation of the hammerhead. Since actual phosphodiester bond cleavage (kg) is believed to be instantaneous and virtually irreversible, most if not all alternative structures that form within the hammerhead will affect either ki or kg but not kg. 85

The basepairing between LI and L2a which disrupts hammerhead formation in the sBYDV (+) RNA self-cleavage domain is predicted to be the most stable helix (AG= -10.1 kcal/mol) in the satellite RNA (Miller & Silver, 1991). The unpaired nucleotides (C-24 and A-73) situated at the top of stem II of the hammerhead may be crucial for the formation of the alternative basepairing (Fig. 1). Deletion of both bases results in a significant increase in the cleavage rate, suggesting that C-24 and A-73 indeed play a role in making the alternative, non-cleaving structure sterically possible.

E+S E-S E-P1-P2 E+P1+P2

E-P1+P2 ^3

Figure 6. Proposed reaction pathway for bimolecular cleavage (adapted from Fedor & Uhlenbeck, 1992).

Upon comparison, the mutant transcript that directly inhibits basepairing between LI and L2a, M3/4/20, cleaves much faster than M3/4. This is consistent with the idea that M3/4 acts indirectly to disrupt Ll/L2a helix formation and inhibits but does not completely prevent helix formation. Finally, 86

M3/4/20 cleaved as fast as M20 (Miller & Silver, 1991); while, the double mutant M3/4 cleaves significantly slower. We concluded that deleting C-24 and A-73 in a transcript with M20, has little additional positive effect on the cleavage rate. Therefore, these data indicate the major alternative structure in the SSI transcript that inhibits cleavage is the putative basepairing between LI and L2a. In general, there seems to be two ways in which the rate of SSI cleavage can be increased: directly by changing nts so that the two regions (LI and L2a) are no longer complementary, or by deleting C-24 and A-73. In either case, the kj of the cleavage reaction is increased since alternative structure(s) other than the hammerhead are less likely to form.

Nuclease Sensitivity The phosphodiester bond cleavage of small molecular weight RNAs is: (i) highly dependent on the formation of the hammerhead structure (Forster & Symons, 1987b; Sheldon & Symons, 1989) and (ii) instantaneous and essentially irreversible (Fedor & Uhlenbeck, 1992). Furthermore, Haldane (1930) proposed that the transition state intermediate of any enzyme-catalyzed reaction must be unstable to successfully lower the activation energy. Therefore, when using structure- sensitive nucleases to look at a cleaving RNA, care must be taken in the interpretation of the data. The slow-cleaving 87

transcripts SSI, M3, and M4 will exist, generally, as inactive molecules and thus sensitivity probing will most likely reflect the presence of these inactive forms. However, with the fast-cleaving transcripts, the probing may reveal structure of the minority of molecules that form inactive structures, but would not detect the majority which will have self-cleaved. For example, the very fast-cleaving transcript (M3/4/20) and the significantly less active RNA (M3/4/20a) have nearly the same nuclease probing map (Fig. 5). This result indicates that nuclease probing can not always discriminate between active and inactive molecules. Secondly, in the case of transcripts that have the M20 mutation (M20, M3/4/20 and M3/4/20a), the CUGANGA region becomes significantly less accessible to single strand analysis. This change in enzyme susceptibility was the most significant in the entire RNA molecule. Because formation of the hammerhead depends on this region to be single-stranded (Forster & Symons, 1987a), and nuclease probing indicated some of the fast cleaving transcripts to be double-stranded in this region, we conclude that this is the structure of the fraction of transcripts that do not cleave even with extended times of incubation. Even though the LI and L2a basepairing is disrupted (Fig 4 & 5), other alternative structures are also possible. In fact, GCG software using the RNA secondary structure analysis applications, MFOLD and PLOTFOLD, indicates 88

additional basepairing is possible in both M3/4/20 and M3/4/20a between nts 20-29 and 48-57 that is not present in any of the other mutant transcripts. Finally, in nearly all the transcripts, helix I has a high degree of single-stranded character (Fig. 4 & 5) and thus the cleavage site has an "open" structure. For ASBVd self-cleavage, a double hammerhead model was proposed that involved the binding of a second ASBVd molecule to the first in a manner that extends stem III thus in effect stabilizing the very short and unstable helix III. It seems unreasonable to believe stem I can be so unstable and still cleave efficiently. Therefore, our data are consistent with the proposal that structures observed with nuclease sensitivity probing are ones that occur in RNAs that self-cleave more slowly than the nucleases cut. These structures may reflect those parts of the RNA that inhibit the formation of the active hammerhead structure. These data might help explain the observation that purified transcripts of sLTSV require a boiling/snapcooling step before efficient cleavage is realized (Forster & Symons, 1987).

Other Hammerheads Generally speaking, the cleavage rate observed for sBYDV (+) RNA is slower than for other self-cleaving RNAs. Even the (-) strand hammerhead of sBYDV RNA cleaves at a much faster rate than its counterpart in the (+) sense (Miller et al.. 89

1991). It was believed earlier that the alternative bases found in the (+) strand hammerhead were responsible for the slow cleavage rate. However, mutagenic analysis of the trinucleotide 5' to the cleavage site (Ruffner, et al., 1990) and the discovery of the self-cleaving RNA of CarSV (Hernandez et al., 1992) seem to indicate that the decrease in cleavage efficiency of sBYDV (+) RNA is not due to the alternative bases. More likely the slow cleavage rate of sBYDV is the result of all the non-conserved nts interacting to form stable alternative conformations with themselves and with the conserved nts that make up the hammerhead. This would be analogous to the global effects seen in proteins where one amino acid change not in the active site per se still interferes with catalytic activity. On the other hand, bases that are a long distance from the cleavage site may in effect inhibit formation of alternative structures {Ll-L2a helix). In this case, the cleavage rate would increase. For example, a permuted dimeric clone of sBYDV RNA that is composed of a greater number of wildtype nts on either side of the cleavage site cleaves much more efficiently than the less-than full- length monomers (Miller & Silver, 1991; Silver et al., in preparation). In summary, the rate of hammerhead-mediated self-cleavage is likely affected by not only the conserved nucleotides but also by the non-conserved bases. It is the 90 global interaction of all the nucleolids within a self- cleaving RNA that governs the actual rate of cleavage. 91

REFERENCES

Baer, M. F., Arnez, J. G., Guerrier-Takada, C., Vioque, A., & Altmann, S. (1990). Preparation and characterization of RNase P from E. coli. Methods in Enzvmoloav. 181. 569-582. Bruening, G. (1989). Compilation of self-cleaving sequences from plant virus satellite RNAs and other sources. Methods in Enzvmoloav. 1980. 546-558.

Bruening, G., Passmore, B. K., Toi, H. v., Buzayan, J. M., & Feldstein, P. A. (1991). Replication of a plant virus satellite RNA: Evidence favors transcription of circular templates of both polarities. Molecular Plant-Microbe Interactions. 4, 219-225. Conway, L., & Wickens, M. (1989). Modification interference analysis of reactions using RNA substrates. Methods in enzvmoloav. 180. 369-379. Epstein, L. M., & Gall, J. G. (1987). Self-cleaving transcripts of satellite DNA from the newt. Cell. 48. 535-543. Fedor, M. J., & Uhlenbeck, 0. C. (1992). Kinetics of intermolecular cleavage by hammerhead ribozymes. Biochemistry. 31. 12042-12054. Forster, A. C., & Symons, R. H. (1987a). Self-cleavage of plus and minus RNAs of a virusoid and a structural model for the active sites. Cell. 49. 211-220. Forster, A. C., & Symons, R. H. (1987b). Self-cleavage of plus and minus RNAs of a virusoid and a structural model for the active sites. Cell. 49, 211-220. Forster, A. C., & Symons, R. H. (1987c). Self-cleavage of virusoid RNA is performed by the proposed 55-nucleotide active site. Cell. 50. 9-16. Grodberg, J., & Dunn, J. J. (1988). omp T encodes Escherichia coli outer membrane protease that cleaves T7 RNA polymerase during purification. J. Bacterid. 170. 1245-1253. 92 Haldane, J. B. S. (1930). Enzymes. London: Longmans. Herlitze, S., & Koenen, M. (1990). A general and rapid mutagenesis method using polymerase chain reaction. Gene. 91, 143-147. Hernandez, C., Daros, J. A., Elena, S. F., Moya, A., & Flores, R. (1992). The strands of both polarities of a small circular RNA from carnation self-cleave in vitro through alternative double-and single-hammerhead structures. Nucleic Acids Research. 20.6323-6329. Hernandez, C., & Flores, R. (1992). Plus and minus RNAs of peach latent mosaic viroid self-cleave in vitro via hammerhead structures. Proceedings of the National Academv of Sciences. USA. 89. 3711-3715. Hutchins, C. J., Rathjen, P. D., Forster, A. C., & Symons, R. H. (1986). Self-cleavage of plus and minus RNA transcripts of avocado sunblotch viroid. Nucleic Acids Research. 14. 3627-3641. Lockard, R. E., & Kumar, A. (1981). Mapping tRNA structure in solution using double-strand-specific ribonuclease Vi from cobra venom. Nucleic Acids Research. 9, 5125-5140. Miller, W. A., Hercus, T., Waterhouse, P. M., & Gerlach, W. L. (1991). A satellite RNA of barley yellow dwarf virus contains novel hammerhead structure in the self-cleavage domain. Virology. 183. 711-720. Miller, W. A., & Silver, S. L. (1991). Alternative tertiary structure attenuates self-cleavage of the ribozyme in the satellite RNA of barley yellow dwarf virus. Nucleic Acids Research. 19, 513-520. Milligan, J. F., & Uhlenbeck, O. C. (1989). Synthesis of small RNAs using T7 RNA polymerase. Methods in Enzvmology. 180. 51-62. Prody, G. A., Bakos, J. T., Buzayan, J. M., Schneider, I. R., & Bruening, G. (1986). Autolytic processing of dimeric plant virus satellite RNA. Science. 231. 1577-1580. Ruffner, D. E., Dahm, S. C., & Uhlenbeck, O. C. (1989). Studies on the hammerhead RNA self-cleaving domain. Gene. 82/ 31-41. Ruffner, D. E., Stormo, G. D., & Uhlenbeck, O. C. (1990). Sequence requirements of the hammerhead RNA self-cleavage reaction. Biochemistry. 29, 10695-10702. 93

Sambrook, J., Fritsch, E. F., & Maniatis, T. (1989). Molecular cloning: A laboratory manual (2nd ed.)« Cold Spring Harbor: Cold Spring Harbor Laboratory Press. Sheldon, C. C., & Symons, R. H. (1989). Mutagenesis analysis of a self-cleaving RNA. Nucleic Acids Research. 17, 5679-5685. Symons, R. H. (1991a). The intriguing viroids and virusoids: What is their information content and how did they evolve? Molecular Plant-Microbe Interactions. 4, 111-121. Symons, R. H. (1991b). Ribozymes. Critical Reviews in Plant Sciences. 10. 189-234. Vary, C. P. H., & Vournakis, J. N. (1984). RNA structure analysis using T2 ribonuclease:detection of pH and metal ion induced conformational changes in yeast tRNA ". Nucleic Acids Research. 12, 6763-6778. Zawadski, V., & Gross, H. J. (1991). Rapid and simple purification of T7 RNA polymerase. Nucleic Acids Research. 19, 1948. 94

REPLICATION OF TRANSCRIPTS FROM A CLONE OF BARLEY YELLOW DWARF VIRUS SATELLITE RNA IN OAT PROTOPLASTS 95

Replication of transcripts from a cDNA clone of barley yellow dwarf virus satellite RNA in oat protoplasts

Silver, S.L., M.S.

Miller, VI.A., Ph.D.

From the Department of Plant Pathology, Iowa State University, Ames, lA 50011 96

ABSTRACT

A small RNA associated with an isolate of barley yellow dwarf virus (BYDV, RPV serotype) has been described which has the physical properties of a satellite RNA (Miller et al., 1991). Here, we demonstrate that this RNA has the biological properties of a satellite RNA; it depends on helper virus (BYDV genomic) RNA for replication and the helper RNA does not depend on the satellite. To separate helper from satellite RNA, a permuted dimeric clone was constructed from which infectious satellite RNA could be transcribed in vitro. The dimeric transcript self-cleaved to produce monomeric satellite RNA. When this RNA was co-electroporated with BYDV genomic RNA into oat protoplasts, replication of both RNAs was detected by Northern hybridization. Satellite RNA was incapable of independent replication. The presence of discrete oligomeric forms of (+) and (-) sense satellite RNA in infected protoplasts suggests that both strands replicate by a rolling circle mechanism. 97

INTRODUCTION

Satellite RNAs of plant viruses are single-stranded molecules that are completely dependent on a helper virus for replication, encapsidation and spread (for review see Roossinck et al., 1992). The helper virus is able to replicate and spread in the absence of the satellite RNA. Satellite RNAs have little sequence similarity with either the helper virus RNA or host RNA. Virus disease symptoms can be altered by the presence of the satellite RNA. In the majority of cases, satellite RNAs attenuate symptoms with a concomitant decrease in helper virus titer (Gerlach et al., 1986; Roossinck et al., 1992; Waterworth et al., 1979). One class that we call small satellite RNAs can form circles and multimers that undergo a self-cleavage reaction during replication. These include satellite RNAs of the nepovirus, sobemovirus (virusoids) and luteovirus groups (Francki, 1985; Keese & Symons, 1987; Miller et al., 1991; Prody et al., 1986). The termini at the cleavage site include a 5'-hydroxyl and a 2'-3'-cyclic phosphodiester (Buzayan et al., 1986; Miller et al., 1991). It has been proposed that the small satellite RNAs replicate by a rolling circle mechanism in which large multimers are generated and then 98

cleaved into monomer-sized molecules (Branch & Robertson, 1984; Hutchins et al., 1985). This self-cleavage occurs with a conserved hammerhead-shaped domain or less frequently a hairpin structure (Forster & Symons, 1987a; Hampel et al., 1990). The hammerhead is composed of three helical domains (I, II, and III) and a central core of eleven conserved, predominantly single-stranded nucleotides (Forster & Symons, 1987a) that are critical for self-cleavage (Ruffner et al., 1990; Sheldon & Symons, 1989). We described previously a satellite-like RNA associated with barley yellow dwarf virus (sBYDV RNA) which has the above physical properties of a small satellite RNA (Miller et al., 1991). However, we had not demonstrated that this RNA is a true satellite RNA by showing that it requires BYDV for propagation and that BYDV does not require the satellite. To pursue this question we first constructed a cDNA clone from which full-length sBYDV RNA could be transcribed. This allowed certain separation of helper and satellite RNA. The in vitro-transcribed sBYDV RNA was then co-electroporated along with BYDV virion RNA into oat protoplasts. The isolates of BYDV we used for virion RNA isolation have been shown previously not to have an associated satellite RNA. This is similar to the approach of Gerlach et al.,(1986) who co- inoculated tobacco plants with tobacco ringspot virus (ToBRV) and the sToBRV RNA. Because BYDV is not mechanically 99

transmissible to whole plants (Oswald, 1951) we used a protoplast system instead. To construct an infectious sBYDV RNA transcript with correct termini, we started with a permuted dimeric clone (pBWS30) described previously (Miller et al., 1991). Both copies of the dimer in this clone contain a deletion of the 3/-terminal base (nt 322) of monomeric sBYDV RNA (+) strand. This deletion rendered (+) sense transcripts of pBWS30 extremely slow at self-cleavage. This is not surprising since this resulted in deletion of the non-basepaired residue at the

5' side of the self-cleavage site required by the hammerhead (Miller et al., 1991). Base 322 was restored to the pBWS30 sequence by site-directed mutagenesis at both cleavage sites. The first mutagenesis reaction was according to Kunkel (1987) using the oligomer 5'- TACGCGCTCTGT (ACGT)ATCCACGAAATAG-3' which was degenerate at the site of base insertion. The oligomer hybridizes on the dimeric insert in plasmid pBWS30 across each cleavage site (Figure lA). The resulting clones were screened by double-stranded DNA sequencing (Promega).

For unknown reasons, only the 3' cleavage site was mutagenized by this method. The 5' cleavage site was mutagenized separately with the same primer by the PGR method of Herlitze and Koenen (1990). The Hindlll-Bglll fragment of pBWS30 was cloned into Bglll/Hindlll-cut pSLllSO (Pharmacia) to create pHB-FL. The first PGR involved 25 cycles of amplification 100

(Amplitaq, Perkin-Elmer-Cetus) using the M13 reverse primer and the above mutagenic oligomer. The purified PGR product and the M13 forward primer were used for the second PGR using the same target DNA (pHB-FL). Products from the second PGR were digested with Bglll/Hindlll, gel-purified and ligated into Bglll/Hindlll-cut pBWS30A that had been mutated previously at the 3' cleavage site. Insertion of the deoxyadenylate at both cleavage sites was confirmed by sequencing. Other bases inserted at the cleavage site were confirmed by sequencing but were not tested for cleavage.

Self-cleavage of Dimeric sBYDV RNA The resulting permuted, dimeric wild type (pFL-WT) clone was linearized with EcoRI and transcribed using SP6 RNA polymerase (Promega) and [a-®^P] GTP (NEN/DuPont). An aliquot

of the transcription reaction was electrophoresed on a 6% polyacrylamide gel and radioactive bands detected by autoradiography.

Dependence of sBYDV RNA Replication

on BYDV Genomic RNA A mixture of satellite-free isolates consisting of PAV-IL and RPV-NY was propagated in oats (Avena sativa cv. Clintland

64) from which virus and viral RNA were extracted as described by Waterhouse et al. (1986). Non-radiolabeled, permuted 101

dimeric sBYDV RNA was synthesized by in vitro transcription of EcoJîJ-linearized pFL-WT using SP6 RNA polymerase (Ambion

Megascript kit, Austin, TX). The monomer obtained by self- cleavage (as in Fig. IB), was gel-purified (Sambrook et al., 1989). One hundred nanograms of this monomer and/or 100 ng of viral RNA (17-fold molar excess of sBYDV RNA) were used as inoculum. Protoplasts were prepared and electroporated as described by Dinesh-Kumar et al. (1992). Total RNA was isolated from protoplasts at various times after

A.

11 ^ 1 1 SP6 — u Self-cleaved transcript:

nts: 269 322 122 B. 4' FL-WT 1 1 M . .-.A#

Figure l. Map of permuted dimeric clone and transcripts of sBYDV RNA. Panel A shows the map of sBYDV RNA in the transcription vector, pBWS3 0A. Shaded regions indicate sequences involved in (+) strand cleavage. Arrows indicate cleavage sites. Expected cleavage product sizes for (+) strand are indicated. Abbreviations: R, EcoRI; S, Sau3AI; B, Bglll;, H, Hindlll. Panel B shows an autoradiograph of transcripts of in vitro transcribed (+) strand sBYDV RNA cleavage products after denaturing polyacrylamide gel electrophoresis. 102

electroporation, and analyzed by Northern hybridization (Fig. 2). In RNA from protoplasts inoculated with BYDV genomic RNA, a (-) sense viral RNA probe complementary to the 5' end of the BYDV-RPV genome, detected an approximately 5.5 kb band that co-migrated with purified virion RNA (Fig. 2A). This was not simply detection of inoculum RNA, because no viral RNA was detected 5 min after inoculation (Fig. 2A, lane 1). Because it spanned only the 5'-terminal 250 bases, this probe was not expected to detect 3'-terminal subgenomic RNAs. A similar pattern was detected when the blot was stripped and re-probed with (-) sense BYDV-PAV genomic RNA (data not shown). Thus, RNAs of both serotypes replicated in oat protoplasts. The mixture of serotypes was used because such a mixture is known to support sBYDV RNA replication in plants (Miller et al., 1991) and the serotypes act synergistically (Baltenberger et al., 1987) resulting in exacerbation of symptoms. Probes specific for both the (+) and (-) strands of sBYDV RNA hybridized to low molecular weight RNAs (Fig. 2B and C) that accumulated by 30 hrs. post-inoculation. Neither probe hybridized to RNA from protoplasts inoculated with sBYDV RNA alone, indicating that it could not replicate in the absence of BYDV genomic RNA. A slight decrease in BYDV-RPV genomic RNA was observed consistently when sBYDV was co-electroporated (Fig. 2A, compare lanes 2 and 3 with 5 and 6). This is somewhat surprising because BYDV isolates that contain sBYDV Figure 2. Northern hybridization of replication products in oat protoplasts. Five micrograms of total RNA isolated from protoplasts was electrophoretically separated on each lane of a 1% agarose gel and blotted to Genescreen (duPont/NEN) nylon membrane for hybridization with the probes indicated at right. After autoradiography, radioactive probe was removed from the blot by boiling for 15 min in O.lxSSC, 0.1% SDS, to allow for hybridization with a different probe. Protoplasts were inoculated with BYDV genomic RNA only (lanes 1-3), both BYDV genomic RNA and sBYDV RNA transcript (lanes 4-6), and sBYDV RNA transcript only (lanes 7-9). Protoplasts were harvested at 5 min (lanes 1, 4, 7), 30 hr (lanes 2, 5, 8) or 72 hr (lanes 3, 6, 9) after electroporation. Lane 10: 0.1 pg BYDV genomic RNA. Lane 11: 0.1 nq (+) sense monomeric sBYDV RNA transcript. Lane 12: 0.1 nq (-) sense sBYDV RNA transcript (including partial cleavage products) obtained by T7 RNA polymerase transcription from Hindlll-linearized pFL-WT. Mobilities of BYDV genomic RNA (G) and monomeric, dimeric and trimeric forms of sBYDV RNA (1, 2, 3) are indicated at left. A. Autoradiograph after probing with an a-[^^P]-labeled (-) sense transcript obtained by transcription of Apal-linearized pML5 with SP6 polymerase. pML5 contains an EcoRV fragment (bases 809-1191, R. Beckett, unpublished data) of the BYDV-RPV genome inserted in the EcRV site of pGEMSZ (Promega). B. After removal of RPV genomic RNA-specific probe, the blot was reprobed with a-[ P]-labeled (-) sense sBYDV RNA transcribed as described above. C. After removal of (-) sense sBYDV RNA probe, the blot was reprobed with a-[ P]-labeled (+) sense sBYDV RNA obtained by transcription as in Fig. 1. 104

Inoculum RNA: BYDV BYDV& sBYDV only sBYDV only 1 2 3 4 5 6 7 8 9 10 11 12 . Probe: (-) BYDV- I RPV RNA

(-) sBYDV transcript I

(+)sBYDV transcript 105

RNA in infected plants give high yields of virus (Miller et al., 1991). Additional controls must be performed to verify that the decrease in BYDV genomic RNA accumulation in protoplasts is caused specifically by sBYDV RNA.

Replication of Both Strands of sBYDV RNA by a Rolling Circle Mechanism An oligomeric series of monomer through pentamer is visible for (+) strand sBYDV RNA, and monomer through tetramer for (-) strands. The higher molecular weight smear of (+) sense satellite RNA may represent higher order multimers that cannot be resolved on this gel system which was used to allow detection of both genomic and satellite RNAs. In both strands, the amount of multimer decreased with increasing size. There was no cross-hybridization between satellite and genomic RNA. (-) sense transcript hybridized weakly with itself (Fig. 2B, lane 12), but (+) sense transcript did not self-hybridize (Fig. 2C). Thus, the similarity of the (-) sense pattern (Fig. 2C) to the (+) sense pattern (Fig. 2C) is not due to cross-hybridization. The (-) strand pattern differs from previous studies of satellite and viroid replicative intermediates in infected plants (Branch et al., 1982; Hutchins et al., 1985) by the relative ease of detection and the distinct multimeric series of bands. The difficulty in detection in previous cases was probably due to the excess 106

of (+) sense satellite RNA in infected tissue and dissociation of bound, unlabeled RNA from the nylon membrane (Hutchins et al., 1985). At least two factors may have contributed to the ease of minus strand detection here. We bound the extracted RNA to the nylon membrane by the more efficient method of UV cross-linking rather than baking. Secondly, the RNA was extracted from protoplasts which give synchronous infection and vastly higher yields of RNA per weight of tissue than do whole plants. Finally, the ratio of (+) to (-) strands of sBYDV RNA may not be as great as with other satellite RNAs. The multimeric series of bands suggests that both strands replicate by a rolling circle mechanism (Branch & Robertson, 1984; Hutchins et al., 1985; Kiefer et al., 1982). By this symmetrical model of replication, the monomeric (+) sense RNA circularizes to act as a template for production of multimeric complementary strands which undergo self-cleavage to form monomers which then circularize and the cycle is repeated to produce (+) sense monomers. Although other models are possible, the evidence with similar satellites such as sTobRV (Bruening et al., 1991) strongly support this model. Thus sBYDV RNA resembles sTobRV RNA, sLTSV RNA (Forster & Symons, 1987a), avocado sunblotch viroid (Hutchins et al., 1986) and carnation stunt associated RNA (Hernandez & Flores, 1992) in that both strands self-cleave during replication. This 107

contrasts with satellites of VTMoV, SCMoV and SNMV in which only the (+) strand circularizes and self-cleaves. The (-) strand forms a linear, multimeric (-)-stranded template for (+) strand synthesis (Chu, 1983; Davies et al., 1990; Hutchins et al., 1985). This construction of an infectious sBYDV RNA transcript will allow us to determine its serotype-specificity and effect on disease symptoms. sBYDV RNA was isolated from a mixture of Australian PAY and RPV isolates of BYDV and from pure RPV derived from this mixture (Miller et al., 1991). To determine which BYDV serotypes are capable of supporting BYDV will require rigorous separation of viral genomic RNAs, preferably by construction of full-length clones. It is noteworthy that the RNA-dependent RNA polymerase of the RPV serotype, but not the PAV serotype, is related to those of the sobemoviruses which are known to support similar satellite RNAs (Habili & Symons, 1989; Vincent, 1991). Thus, helpers capable of supporting sBYDV RNA replication may be limited to RPV-like serotypes of BYDV. To study the effect of sBYDV RNA on disease symptoms will require inoculation of plants with aphids that acquired satellite-containing virus from protoplasts (Young, 1991) or by agroinfection (Leiser et al., 1992). Other satellite-like RNAs are associated with luteoviruses. The ST9-associated RNA of BWYV may be a large gene-encoding satellite (Chin et al, 1993), and a satellite 108

RNA has been shown to be associated with ground nut rosette virus, a non-luteovirus that depends on ground nut rosette assister luteovirus for transmission and disease (Murant, 1990). However, sBYDV RNA is the first luteovirus-associated RNA shown to fulfill completely the definition of a satellite. To understand how a satellite RNA affects symptom development and how the helper virus and satellite RNA co-exist in vivo demands defining more clearly the three-way relationship between the satellite RNA, helper virus and host plant. 109

REFERENCES

Baltenberger, D. E., Ohm, H. W., & Foster, J. E. (1987). Reactions of oat, barley and wheat to infection with barley yellow dwarf virus isolates. Crop Sciences. 27, 195-198. Branch, A. D., Robertson, H. D. (1984). A replication cycle for viroids and other small infectious RNAs. Science. 223. 450—455. Branch, A. D., Robertson, H. D., Greer, C., Gegenheimer, P., Peebles, C., Abelson, J. (1982). Cell-free circularization of viroid progeny RNA by an RNA ligase from wheat germ. Science. 217. 1147-1149. Bruening, G. (1989). Compilation of self-cleaving sequences from plant virus satellite RNAs and other sources. Methods in Enzvmoloav. 180. 546-558. Bruening, G., Passmore, B. K., Toi, H. v., Buzayan, J. M., & Feldstein, P. A. (1991). Replication of a plant virus satellite RNA: Evidence favors transcription of circular templates of both polarities. Molecular Plant- Microbe Interactions. 4, 219-225. Buzayan, J. M., Gerlach, W. L., & Bruening, G. (1986). Satellite tobacco ringspot virus RNA: A subset of the RNA sequence is sufficient for autolytic processing. Proceedings of the National Academy of Sciences. USA. 83. 8859-8862. Chin, L., Foster, J. L., Falk, B. W. (1993). The Beet Western Yellows Virus ST9- Associated RNA shares structural and Nucleotide sequence homology with Carmo-like viruses. Virology. 192. 473-482. Chu, P. W. G., Francki, R. I. B., Randies, J. W. (1983). Detection, isolation and characterization of high molecular weight double-stranded RNAs in plants infected with velvet tobacco mottle virus. Virology. 126. 480-492. 110

Davles, C., Haseloff, J., & Symons, R. H. (1990). Structure, self-cleavage, and replication of two viroid-like satellite RNAs(virusoids) of subterranean clover mottle virus. Viroloav. 177. 216-224. Dinesh-Kumar, S. P., Brault, V., & Miller, W. A. (1992). Precise mapping and in vitro translation of a trifunctional subgenomic RNA of barley yellow dwarf virus. Viroloav. 187. 711-722. Forster, A. C., & Symons, R. H. (1987a). Self-cleavage of plus and minus RNAs of a virusoid and a structural model for the active sites. Cell. 49. 211-220. Forster, A. C., & Symons, R. H. (1987b). Self-cleavage of virusoid RNA is performed by the proposed 55-nucleotide active site. Cell. 50. 9-16. Francki, R. I. B. (1985). Plant Virus Satellites. Annual Review of Microbioloav. 39. 151-174. Gerlach, W. L., Buzayan, J. M., Schneider, I. R., & Bruening, G. (1986). Satellite tobacco ringspot virus RNA: Biological activity of DNA clones and their in vitro transcripts. Viroloav. 151. 172-185. Habili, N., & Symons, R. H. (1989). Evolutionary relationship between luteoviruses and other RNA plant viruses based on sequence motifs in their putative RNA polymerases and nucleic acid helicases. Nucleic Acids Research. 17, 9543-9555. Hampel, A., Tritz, R., Hicks, M., Cruz, P. (1990). Hairpin catalytic RNA model: Evidence for helices and sequence requirement for substrate RNA. Nucleic Acids Research. 18, 299-304. Herlitze, S., Koenen, M. (1990). A general and rapid mutagenesis method using polymerase chain reaction. Gene. 91/ 143-147. Hernandez, C., & Flores, R. (1992). Plus and minus RNAs of peach latent mosaic viroid self-cleave in vitro via hammerhead structures. Proceedings of the National Academy of Sciences. USA. 89. 3711-3715. Ill

Hutchins, C. J., Keese, P., Visvader, J. E., Rathjen, P. D., Mclnnes, J. L., & Symons, R. H. (1985). Comparison of multimeric plus and minus forms of viroids and virusoids. Plant Molecular Biology. 4, 293-304. Hutchins, C. J., Rathjen, P. D., Forster, A. C., & Symons, R. H. (1986). Self-cleavage of plus and minus RNA transcripts of avocado sunblotch viroid. Nucleic Acids Research. 14. 3627-3641. Keese, P., Symons, R. H. (1987). The structure of viroids and virusoids. Boca Raton: CRC Press. Kiefer, M. C., Daubert, S. D., Schneider, I. R., Bruening, G. (1982). Multimeric forms of satellite tobacco ringspot virus RNA. Virology. 137. 371-381. Kunkel, T. A., Roberts, J. D., Zakour, R. A. (1987). Rapid and efficient site-specific mutagenesis without phenotypic selection. Methods In Enzvmology. 154. 367-382. Leiser, R. M., Ziegler-Graff, V., Reutenauer, A., Herrbach, E., Lemaire, O., Guilley, H., Richards, K., Jonard, G. (1992). Agroinfection as an alternative to insects for infecting plants with beet western yellows luteovirus. Proceedings of the National Academy of Sciences. 89. 9136-9140. Miller, W. A., Hercus, T., Waterhouse, P. M., & Gerlach, W. L. (1991). A satellite RNA of barley yellow dwarf virus contains a novel hammerhead structure in the self-cleavage domain. Virology. 183. 711-720. Murant, A. F. (1990). Dependence of groundnut rosette virus on its satellite RNA as well as on groundnut rosette assister luteovirus for transmission by Aphis craccivora. Journal of General Virology. 71, 2163-2166. Oswald, J. W., Houston, B. R. (1951). A new virus disease of cereals, transmissible by aphids. Plant Disease Reporter. 35, 471-475. Prody, G. A., Bakos, J. M., Buzayan, J. M., & Schneider, I. R., Bruening, G. (1986). Autolytic processing of dimeric plant virus satellite RNA. Science. 231. 1577-1580. Roossinck, M. J., Sleat, D., & Palukitis, P. (1992). Satellite RNAs of plant viruses; Structures and biological effects. Microbiological Reviews. 56. 265-279. 112 Ruffner, D. E., Stormo, G. D., & Uhlenbeck, O. C. (1990). Sequence requirements of the hammerhead RNA self- cleavage reaction. Biochemistry. 29. 10695-10702. Sambrook, J., Fritsch, E. F., Maniatis, T. (1989). Molecular cloning: A laboratory manual (2nd ed.). Cold Spring Harbor: Cold Spring Laboratory Press. Sheldon, C. C., & Symons, R. H. (1989). Mutagenesis analysis of a self-cleaving RNA. Nucleic Acids Research. 17. 5679-5685. Vincent, J. R., Lister, R. M., Larkins, B. A. (1991). Nucleotide sequence analysis and genomic organization of the NY-RPV isolate of barley yellow dwarf virus. Journal of General Virology. 72. 2347-2355. Waterhouse, P. M., Gerlach, W. L., Miller, W. A. (1986). Serotype-specific and general luteovirus probes from cloned cDNA sequences of barley yellow dwarf virus. Journal of Virology. §!_, 1273-1281. Waterworth, H. E., Kaper, J. M., & Tousignant, M. E. (1979). CARNA5, the small cucumber mosaic virus-dependent replicating RNA, regulates disease expression. Science. 204. 845-847. Young, M. J., Kelly, L., Larkin, P. J., Waterhouse, P. M., Gerlach, W. L. (1991). Infectious in vitro transcripts from a cloned cDNA of barley yellow dwarf virus. Virology. 180. 372-379. 113

BIMOLECULAR CLEAVAGE REACTION AND THE DESIGN OF TARGET-SPECIFIC RIBOZYMES 114

Biitiolecular cleavage reaction and the design of target specific ribozymes

Silver, S.L., M.S.

Miller, VI.A., Ph.D.

Department of Plant Pathology, Iowa State University, Ames 50011 115

INTRODUCTION

The first successful demonstration of a bimolecular reaction involving an autocatalytic RNA was the ribozyme L-19 IVS derived from the pre-rRNA of T. thermophilia (Zaug et al.,

1986). As an endoribonuclease the L-19 IVS was directed against a specific site within the a-globin mRNA (Zaug et al., 1986). With a length of nearly 400 nts, the L-19 IVS RNA may undergo changes in conformation resulting in formation of inactive ribozyme structures in vivo. Additionally, the kinetics of L-19 IVS RNA is characterized by Briggs-Haldane kinetics, in which high affinity binding to the substrate and the subsequent slow release of product is observed. The end result is the propensity of the ribozyme to bind both incorrect and correct substrate molecules equally well; thereby cleaving a higher percentage of incorrect RNA molecules (Herschlag, 1991). For these reasons, the potential usefulness of the L-19 IVS ribozyme in vivo may be limited. Uhlenbeck (1987) has suggested using the hammerhead catalytic domain as a ribozyme since it is composed of only 50 nucleotides (Forster & Symons, 1987) which reduces the risk of inactive ribozyme formation in vivo. In addition, the 116

hammerhead ribozyme is expected to cleave with Michaelis- Menten kinetics; unlike the L-19 TVS ribozyme. Uhlenbeck (1987), using the hammerhead of ASBV, dissected the enzyme and substrate components into two separate molecules (Fig lA). The substrate RNA contained a three-base loop that connected the conserved nucleotides GAAA and U of the hammerhead. The ribozyme component contained the remaining conserved hammerhead nucleotides. After the substrate and ribozyme RNAs were transcribed and annealed in the presence of Mg^^, significant cleavage of the substrate RNA

into two fragments was observed (Uhlenbeck, 1987). The RNA enzyme demonstrated turnover capability without undergoing modification and thus was truly catalytic. This observation was a significant development in the use of gene-targeted ribozymes; since, now an RNA could potentially inactivate, not only one copy of an RNA as in the case of an antisense molecule, but several. Many examples of hammerheads now exist that demonstrate bimolecular cleavage capability (Gerlach et al., 1987; Jeffries & Symons, 1989; Koizumi et al., 1988). In 1988 a more general acting hammerhead-derived ribozyme was proposed which required the substrate RNA to contain only the conserved U located one base upstream of the cleavage site (Haseloff & Gerlach, 1988) (Fig IB). Since the catalytic domain contained the majority of conserved nucleotides, almost any substrate RNA could theoretically be targeted for 117

A. C C G A-U C-G / Substrate^yi^-f^^ pppQCQÇÇ® yÇÇAQÇoH hoCGCGG^ çAGCUCQGppp Ribozyme

B.

Substrate 5' f xxxxxxyyyyyyGlgcjyyyyy^xxxxxx xxxxxxD TMXXXXXX II II

lAA-U III G-C Ribozyme

G U

Figure 1. (A) Division of the ASBV hammerhead into substrate and catalytic domains. Conserved bases of hammerhead are in bold and italics. Arrow indicates cleavage site (Uhlenbeck, 1987). (B) Gene-targeted ribozyme as proposed by Haseloff and Gerlach (1988). (I) indicates the one conserved nucleotide in the substrate RNA. (II) identifies the target arms which give specificity to the ribozyme. (Ill) is the catalytic domain of the ribozyme. Arrow indicates cleavage site (Haseloff & Gerlach, 1988). 118

cleavage. This ribozyme was constructed using 10 conserved nucleotides of sToBRV (+) RNA connected by a 4 basepair helix and a tetraloop. It was designed with "arms" that would base- pair to the bacterial chloramphenical acetyl-transferase (CAT) mRNA which included the U of the target. The CAT transcript and ribozyme were incubated and significant cleavage of the RNA was observed (Haseloff & Gerlach, 1988). Since the first demonstration of a hammerhead-type ribozyme cleaving an unrelated substrate RNA, other target-specific ribozymes have been constructed (for review see Symons, 1991). When designing gene-targeted ribozymes it is imperative to consider the reaction pathway proposed for the intermolecular cleavage reaction (Fedor & Uhlenbeck, 1992; Fig. 2). The formation and dissociation of the enzyme- substrate complex (E-S) is defined by the rate constants

and k.i respectively. The actual cleavage reaction is defined

by kgand dissociation of products from the ribozyme is defined

by kg and k^. Three main parameters must be

E-P2+P1

E+8 E^ ^ E-P1-P2 E+P1+P2

E-P1+P2

Figure 2. Proposed reaction pathway for bimolecular cleavage (Fedor & Uhlenbeck, 1992) 119

considered during ribozyme design. The first involves selecting a ribozyme that has minimal sequence requirements. As mentioned previously, the hammerhead catalytic domain is excellent in this regard. The ribozyme described by Haseloff and Gerlach (1988) has only 22 nts that are involved directly in catalysis. Further deletions within the ribozyme, i.e., the short stem-loop connecting the two regions of conserved bases in the ribozyme only decreases the cleavage rate (McCall et al., 1992). Therefore, the 22 nt long sequence is likely the shortest catalytic domain still capable of cleavage. Secondly, alternative conformations within the target sequence can pose significant difficulties for E-S formation. For instance, some hammerhead ribozymes with a slow rate of product formation have a low binding affinity for the substrate RNA (Fedor & Uhlenbeck, 1992; Xing & Whitton, 1993).

It is believed this observation is the result of an alternative secondary structure at the cleavage site. Identification of target sites with a significant likelihood of forming secondary structures can be accomplished using RNA structure analysis software. A more time-consuming method by which to find RNA secondary structures involves structure- sensitive nucleases and chemical modification analysis. The last consideration when designing a ribozyme is the length and composition of the target arms. Changes in either parameter result in dramatic effects on the rates of substrate 120

dissociation (k.j) and product release (kg and k^) (Fedor &

Uhlenbeck, 1992). By increasing the length of the target arms, a corresponding increase in the stability of the E-S complex occurs. A more stable E-S leads to a higher number of incorrect RNAs binding to the ribozyme (lowered discrimination index), because a few mismatches between ribozyme and target will be tolerated and subsequently the cleavage of more incorrect RNA molecules (Herschlag, 1991). Since the rate of dissociation of small RNA duplexes is dependent on the length and composition of the RNA (Tinoco, 1982), the rate of product dissociation of the hammerhead ribozyme also depends directly on the length and sequence of the target arms. This is a major dilemma since both E-S complex formation and product dissociation are dependent on the same parameters. This dilemma is evident from mutagenic analysis of the target arms. Base changes that increase kg and k^ also leads to an increase in k.i. Haldane (1930) argued that instability of the E-S complex is in fact the result of a lowering of the activation energy between the E-S and E-P during enzyme catalysis. Instability of the E-S complex is beneficial when attempting to lower the activation energy. However, the E-S complex must be sufficiently stable to allow cleavage. When designing a ribozyme, therefore, the target arms must be long enough to form a stable E-S complex but not so long as to lower the 121

discrimination index of the ribozyme or inhibit the release of products from the ribozyme. Presently, the optimal length and composition of the target arms that results in the most efficient cleavage of a substrate RNA in vivo is not known. However, many ribozymes that cleave their substrate RNA efficiently have been constructed. One of the best characterized ribozymes cleaves c-fos mRNA (Scanlon et al., 1991). The ribozyme was shown to

significantly decrease the level of several proteins whose translation is regulated by the c-fos mRNA product, whereas an

inactive form of the ribozyme had no influence on c-fos

regulated proteins. The accumulation of c-fos mRNA cleavage

products in vivo, however, was not determined. Therefore, as is typical for most in vivo ribozyme studies, it is unclear whether the ribozyme acts as a catalytic entity or simply as a better antisense molecule than the inactive, mutated ribozyme. To ascertain the efficacy of target-specific ribozymes, it is imperative that the experimental approach allows one to distinguish between cleavage of the target and mere antisense inhibition. In the present work, we dissect the sBYDV RNA hammerhead into separate ribozyme and substrate transcripts and demonstrate that a wildtype and a mutant catalytic domain have the ability to associate with the substrate RNA in a manner consistent with hammerhead formation. We also show that a 122 ribozyme targeted against the 5' end of BYDV-PAV is able to specifically cleave a portion of the viral genomic RNA in vitro. 123

MATERIAL AND METHODS

construction of Ribozymes

B19 and B5/19 ribozymes To test whether the sBYDV RNA hammerhead can function in an intermolecular fashion, the catalytic and substrate domains of the (+) hammerhead were separated by placing bases 310 to 10 on one transcript (substrate) and bases 10 through 89 on a second transcript (ribozyme). The first bimolecular enzyme (B19) was constructed by cutting pSSM19 (Miller & Silver, 1991) and pGEM3Z (Promega, Madison, WI) with Sad and Drall. The 2357 basepair (bp) vector fragment from the pSSM19 digestion was ligated to the 898 bp fragment from the pGEM3Z digestion. The second bimolecular enzyme, 35/19 was constructed from the clone pSSM5 (Miller & Silver, 1991). SSM5 contains a 34 base deletion of nucleotides 30 to 63. The M19 mutation containing the Sad site was introduced by site- directed mutagenesis of pSSMS with the primer 5' AGACAGTACGAGCTCTGTTATC 3' using the method of Kunkel et al., (1987). The resulting mutant pSSM5/19 (Fig. 3B) was divided into catalytic and substrate domain as described above. 124

A. Substrate

.310 MO AUUUC aiOQA r BiBiBi^ùàÀÀàcËÀMÙ •••uj.MJiiii.w.ijaa tr

Ribozyme

"uqqu

B. Substrate 3' **^^UUUC (HJCKÏA^ACAQ'P®" 3' UauCAUGCU CGûGCt M lAAGCGG

TOU-A Ribozyme

Figure 3. Bimolecular hammerhead formation of sBYDV (+) strand RNA. Boxed bases are conserved among all hammerheads.Numbering of nucleotides is based on the full- length satellite RNA. Cleavage sites indicated by an arrow. (A) B 19 and substrate RNA cleavage structure. (B) B 5/19 and substrate RNA cleavage structure. 125

Gene-targeted ribozymes Sites within the BYDV genomic RNA containing the sequence AUA were targeted for cleavage using the catalytic domain of the sBYDV RNA (+) strand hammerhead. The ribozymes were constructed by annealing complementary oligomers encoding the ribozyme sequence to obtain a double stranded fragment with sticky ends. The fragment was cloned into the transcription vector pGEM3Z. Oligomers were synthesized using /3- cyanoethylphosphoramidite chemistry (Nucleic Acids Facility, Iowa State University). The subgenomic (SUBI) and upstream (UPl) ribozymes (see below for specific target sites) were cloned using the oligomers in Figure 4. Prior to cloning, the oligomers were phosphorylated with T4 kinase. The oligomers were annealed using a standard annealing buffer (40 mM Tris- HCl, pH 7.4, and 4 mM MgClg) and approximately 1 /xg of each oligomer (Ausubel et al., 1989). The annealing reaction was heated to 95° C for 1 min followed by slow cooling to 22° C.

The annealed oligomers were ligated into BamHI-PstI digested pGEM3Z. Successful transformation of E. coli (strain DH5aF') with these ligation mixes required serial dilutions of the ligation reaction; since high levels of DNA have been implicated in decreased transformation efficiency. 126

Subgenomic Ribozyme Oligomers BarnH Prt 5' QArcCATACTAACCTQ(TA)CQACQTATCAATAQATACAaAAATAAACCCTATCCrGC/* 3' 3' 0TAT

Upstream Ribozyme Oligomers BwnH Pit) 5' QArCCTTGCTTTAQACTQ(TA)CQACQTATCAATAQATACAQAAATQTnTACTCTQCA 3' 3' GAACQAAATCTQAC(A'T7QCTQCATAGtrTATCTATQTCTTTACAAAATCWa 5'

Ribozyme 175 oligomer 3' ATATCACTCABCATAAT 5' 5' TAAAQCQQTTTCQTCCTCACQQACTCATCAQAAAQACTTCCTATAQTQAQTCQTATTA 3'

Figure 4. Oligonucleotides used in cloning the gene-targeted ribozymes or in the oligo-directed transcription of Rz 175. Bases involved in restriction enzyme sites are in italics. Bases in bold are the ribozyme target arms. Nucleotides in bold and underlined make up the T7 RNA polymerase promoter. Sequences involved in the catalytic domain of the ribozyme are boxed.

In Vitro Transcription and Cleavage Assays

B19 and B5/19 ribozymes Transcripts of the bimolecular enzymes were synthesized by digesting pB19 and pB5/19 with Hind III, extracting with phenol/chloroform and ethanol precipitating for 3 0 min at -80° C. The linearized DNA was resuspended in 0.1% diethylpyrocarbonate (DEPC)-treated water and quantified by determining the O.D.260nm* A typical ICQ nl transcription 127 reaction contained transcription buffer (40 mM Tris-HCl, pH 8.0, 6 mM MgClg, 5 mM DTT, 1 mM spermidine), 0.5-1.0 mM NTP,

100 units of RNasin, 10 jug of linearized template and 150 units of T7 RNA polymerase. The reaction was incubated at 37°

C for 1-3 hours. For a-[®^P]-labeled transcripts, 50 nCi of a-

[®^P]GTP (NEN/duPont) was added to the reaction. Substrate RNA

was transcribed from Sacl-linearized pSSM19 (Miller & Silver, 1991). Each RNA was run on a 6% polyacrylamide gel containing 7 M urea and full-length substrate and ribozyme molecules were purified by excising the radioactive bands and eluting the RNA from the gel by constant shaking in elution buffer (0.5 M ammonium acetate, 1 mM EDTA, pH 8.0, and 0.1% SDS; Sambrook et al., 1989). Radioactivity was determined by liquid scintillation counting. Cleavage assays were performed as described previously (see Material and Methods, Section I).

Gene-targeted ribozymes Ribozymes UPl and SUBI were transcribed from Pst I- linearized duplex DNA using T7 RNA polymerase. Full-length ribozyme transcripts were gel-purified as above. Ribozyme 175 (Rz 175) was transcribed directly from an oligomer containing the complement of the T7 RNA polymerase promoter (Fig. 4) according to the method of Milligan and Uhlenbeck (1989). The 17-mer containing the T7 RNA polymerase promoter was combined in an equal molar ratio with the synthetic template containing the ribozyme and the complement of the T7 RNA polymerase 128 promoter (Fig. 4). Annealing of the two oligomers was performed by heating to 95° C in Tris, pH 7.5, and 10 mM MgClg followed by quick cooling to room temperature. The resulting partially double-stranded oligomer was directly used for transcription of Rzl75. UPl substrate RNA was transcribed from EcoRI-linearized pPA142 (Miller et al., 1988) and SUBI substrate was transcribed from EcoRl-digested pSP-17 (Dinesh-Kumar et al., 1992). Substrate RNA for Rzl75 was transcribed from EcoRV- digested pPAV-6 (Dinesh-Kumar et al., 1992). All substrates were internally labeled using a-[®^P]-GTP and gel-purified from a 6% polyacrylamide gel as described previously. 129

RESULTS AND DISCUSSION

Bimolecular Cleavage The efficiency of the sBYDV (+) RNA hammerhead in a bimolecular cleavage reaction was determined by dividing the catalytic and substrate domains of pSS19 and pSSM5 into separate transcripts (Fig. 3A & B). The cleavage assay contained either B5/19 or B19 and a 5-fold molar excess of substrate RNA. After incubating the ribozyme and substrate for 1 hour at 3 0° C in 10 mM MgClg, the cleavage products were separated on a 10% polyacrylamide gel containing 7 M urea followed by quantification of fragments by phosphoryimage analysis (Molecular Dynamics). B19 converted 4% of the substrate into the expected 18 base and 8 base products (Fig. 5). Incubation of B5/19 and S19 resulted in cleavage of nearly 38% of the substrate RNA into products. Incubation of the substrate and B5/19 transcripts at either 37° C or 45° C resulted in approximately 65% of the substrate RNA being converted into products (Fig. 6). The B5/19 cleavage rate is similar to other bimolecular reactions that use the hammerhead catalytic domain of the satellite RNA of tobacco ringspot virus (sToBRV RNA) (Haseloff & Gerlach, 1988) and avocado sunblotch viroid (ASBVd) (Uhlenbeck, 1987). The only sequence difference between B19 130

30°C 37t 45°C

z A z £ z o i + o i + o (n A (0 (0 Ë CO (0 i!

#

3' PRODUCT

Figure 5. Bimolecular cleavage by ribozyme B19. Autoradiograph of transcripts subjected to denaturing polyacrylamide gel electrophoresis after incubation in cleavage conditions for 1 hour at 30| C. Mobilities of full-length substrate (S) and 5' and 3' products are indicated. Abbreviations: Rz, ribozyme; S, substrate. 131

m #

^ 3' PRODUCT

5' PRODUCT

Figure 6. Bimolecular cleavage by ribozyme B5/19. Autoradiograph of transcripts subjected to denaturing polyacrylamide gel electrophoresis after incubation in cleavage conditions for 1 hour at temperatures indicated. Mobilities of full-length (S) and 5' and 3' products are indicated. Abbreviations: Rz, ribozyme; S, substrate. 132

and B5/19 is the 34 base deletion in B5/19 (Fig. 3). Secondary structure comparison of each ribozyme/substrate pair using the software MFOLD and PLOTFOLD in the GCG package (Zuker, 1989) revealed no significant alternative conformation that could account for the poor cleavage efficiency of B19. A significant difference, however, was observed in the calculated -AG values of the two ribozyme/substrate complexes. B19 forms a more stable hammerhead than does B5/19. According to Fedor and Uhlenbeck (1992), increasing the stability of a hammerhead above an optimal point could potentially cause an actual decline in the rate of product release. The -AG of the B19-S19 complex is -26.9 whereas the -AG of the B 5/19-S19 complex is significantly more positive with a value of -20.4. Further deletions and base changes in the catalytic domain along with nuclease probing experiments will be required to fully characterize the chemistry of enzyme association and product release.

Gene Targeted Ribozymes Ribozymes designed to down-regulate a particular gene have met with some success in vivo (for review see Gotten & Birnstiel, 1989; Scanlon et al., 1991/ Symons, 1991; Xing & Whitton, 1993). Thus we attempted to define ribozyme target sites within the genome of BYDV-PAV. The first ribozyme tested, UPl, was directed at a sequence in BYDV-PAV spanning 133

the nucleotides 1018 to 1039 of the 39K open reading frame (ORF) (Fig. 7). UPl resulted in no detectable cleavage after 1 hour at 37° C (data not shown). The second ribozyme tested against the BYDV genome was SUBI. This ribozyme anneals at a site extending over nucleotides 2769 to 2789. 2769 is the exact start of subgenomic RNA 1 formed during BYDV infection (Fig. 7) (Dinesh-Kumar et al., 1992). Similar to the UPl ribozyme, SUBI resulted in no significant cleavage of BYDV genomic RNA. A survey of the UPl target sequence using the program, RNASE did not reveal any obvious secondary structures within the target site. Upon further analysis using GCG-MFOLD (Zuker, 1989), the 3' end of the target sequence demonstrated the potential to form duplex DNA that may interfere with association of the ribozyme to the cleavage site sufficiently to inhibit cleavage of the target sequence. A search of the SUBI target sequence using nuclease probing analysis revealed a significant conformation within the subgenomic promoter sequence (Dinesh-Kumar et al., 1992) (Fig. 8). The target sequence is bounded by the two arrows (Fig. 8) and forms a stem-loop structure that clearly would inhibit the annealing of ribozyme and substrate. After the failure of two ribozyme constructs to cleave genomic BYDV RNA in vitro, the third ribozyme we tested used the catalytically-important nts from the sToBRV hammerhead. 134

Readthrough Frameshift CP 22K 60K(POL) 1 (t7K • mm f 6.7K Leaky Scanning 5' 3'

Subitrat*

®A g®Aa® y^A RJbozynw qIC A A UA ButwtrMi i s'aAUfOGajuuAmauuMUAi s-

M Rixsyini Q-C A A UA •

RbozytiM A a Q U

Figure 7. Sequences targeted for ribozyme-mediated cleavage. (A) Genome organization of BYDV-PAV. (B) For each ribozyme, the catalytic domain and the target site are given. Arrows indicate the cleavage site. 135

G G®° A A U-G C-G A-U C-G C-G A-U

u% ^ « G A90 CP ORF start

U-A^O y y C^

'°UU.A a'=AAU4°AGUU-AU

\ U-G G-C

;....G^"'-'AC»UUAGCUc'ty CAAAUCAAAAUGGC A... 3' ' 130 17K ORF Start

Figure 8. Proposed secondary structure of the subgenomic promoter sequence of BYDV-PAV. Arrows indicate start and end points of Rz-175 target site (Dinesh-Kumar et al., 1992). 136

With the sToBRV-derived ribozyme, the 5' side of the cleavage site must now have a GUC trinucleotide instead of the AUA trinucleotide. The target site in BYDV genome was chosen after extensively searching the entire BYDV genome for a cleavage site that had little opportunity for secondary structure formation. A 5' end sequence was selected because cleavage here should inhibit an early step in virus replication. The cleavage assay contained 3 0 nM ribozyme and 100 nM substrate RNA. Approximately 25% of the substrate RNA was converted into product after 180 min incubation in standard cleavage conditions at 37° C (Fig. 9). Future attempts to designing target-specific ribozymes against BYDV genomic RNA will involve the careful selection of numerous sites within the genomic RNA that do not form predicted stable secondary structures. Nuclease probing experiments along with determining the cleavage efficiency of ribozymes at many sites within the BYDV RNA will give insight into avoiding ribozyme inactivity due to alternative conformations. In addition, the smallest, most efficient ribozyme will need to be designed. This will entail determining what role(s) the conserved and non-conserved bases serve in the cleavage reaction and how changes within these bases change the dynamics of hammerhead formation and product dissociation. 137

SUBSTRATE

5' PRODUCT

3' PRODUCT

Figure 9. Rz-175 cleavage assay. Autoradiograph of transcripts after incubation for indicated times. Mobilities of full-length substrate (5' 258 bases of BYDV-PAV) and 5' and 3' products are given. 138

REFERENCES

Gotten, M., & Birnstiel, M. L. (1989). Ribozyme mediated destruction of RNA in vivo. The EMBO Journal. 8, 3861-3866. Dinesh-Kumar, S. P., Brault, V., & Miller, W. A. (1992). Precise mapping and in vitro translation of a trifunctional subgenomic RNA of barley yellow dwarf virus. Virology. 187. 711-722. Fedor, M. J., & Uhlenbeck, O. C. (1992). Kinetics of intermolecular cleavage by hammerhead ribozymes. Biochemistrv. 31. 12042-12054. Forster, A. C., & Symons, R. H. (1987). Self-cleavage of virusoid RNA is performed by the proposed 55-nucleotide active site. Cell. 50. 9-16. Gerlach, W. L., Llewellyn, D., & Haseloff, J. (1987). Construction of a plant disease resistance gene from the satellite RNA of tobacco ringspot virus. Nature. 328. 802-805. Haldane, J. B. S. (1930). Enzymes. London: Longmans. Haseloff, J., & Gerlach, W. L. (1988). Simple RNA enzymes with new and highly specific endoribonuclease activities. Nature. 334. 585-591. Herschlag, D. (1991). Implications of ribozyme kinetics for targeting the cleavage of specific RNA molecules in vivo: more isn't always better. Proceedings of the National Academy of Sciences. USA. 88. 6921-6925. Jeffries, A. J., & Symons, R. H. (1989). A catalytic 13-mer ribozyme. Nucleic Acids Research. 17. 1371-1377. Koizumi, M., Iwai, S., & Ohtsuka, E. (1988). Construction of a series of several self-cleaving RNA duplexes using synthetic 21-mers. FEB. 228. 228-230. 139

Kunkel, T. A., Roberts, J. D., & Zakour, R. A. (1987). Rapid and efficient site-specific mutagenesis without phenotypic selection. Methods In Enzvmoloav. 154. 367-382. McCall, M. J., Hendry, P., & Jennings, P. A. (1992). Minimal sequence requirements for ribozyme activity. Proceedings of the National Academy of Sciences. 89. 5710-5714. Miller, W. A., & Silver, S. L. (1991). Alternative tertiary structure attenuates self-cleavage of the ribozyme in the satellite RNA of barley yellow dwarf virus. Nucleic Acids Research. 19, 513-520. Miller, W. A., Waterhouse, P. M., & Gerlach, W. L. (1988). Sequence and organization of barley yellow dwarf virus genomic RNA. Nucleic Acids Research. 16, 6097-6111. Milligan, J. F., & Uhlenbeck, O. C. (1989). Synthesis of small RNAs using T7 RNA polymerase. Methods in Enzvmoloav. 180. 51-62. Sambrook, J., Fritsch, E. F., & Maniatis, T. (1989). Molecular cloning; A laboratory manual (2nd ed.). Cold Spring Harbor: Cold Spring Laboratory Press. Scanlon, K. J., Jiao, L., Funato, T., Wang, W., Tone, T., Rossi, J. J., & Kashani-Sabet, M. (1991). Ribozyme-mediated cleavage of c-fos mRNA reduces gene expression of DNA synthesis enzymes and metallothionein. Proceedings of the National Academy of Sciences. USA. 88, 10591-10595. Symons, R. H. (1991). Ribozymes. Critical Reviews in Plant Sciences. 10. 189-234. Uhlenbeck, O. C. (1987). A small catalytic oligoribonucleotide. Nature. 328. 596-600. Xing, Z., & Whitton, L. (1993). An anti-lymphocytic choriomeningitis virus ribozyme expressed in tissue culture cells diminishes viral RNA levels and leads to a reduction in infectious virus yield. Journal of Virology. 67. 1840-1847. Zaug, A. J., Been, M. D., & Cech, T. R. (1986). Tetrahymena ribozyme acts like an RNA restriction enzyme. Nature. 324. 429-433. Zuker, M. (1989). On finding all suboptimal foldings of an RNA molecule. Science. 244. 48-52. 140

GENERAL SUMMARY

Barley yellow dwarf virus (BYDV) is the most widespread and destructive plant virus disease of small grain cereals (Plumb, 1983). Significant grain yield losses due to BYDV have been observed in both deliberate and natural infections (Yamani & Hill, 1991). Several methods have been employed to control BYDV infestations. For example, (i) monitoring of aphid populations to determine the best time for planting or for application of insecticides, (ii) use of an aphid parasite to decrease the aphid population, and (iii) breeding for resistance cultivars using the Yd2 gene from Ethiopian barley, have all been used in the past with varying degrees of success. Less classical approaches to disease management have also been implemented to control virus infestations. Mechanisms such as coat protein-mediated resistance and the use of satellite RNAs have been used with other plant viruses. Since no one control strategy can be or has been proven completely effective in eliminating yield losses due to BYDV, it is necessary to further characterize BYDV and the recently discovered satellite of BYDV with respect to pathogenesis. This study was limited to the analysis of the self-cleavage site within sBYDV (+) RNA, synthesis, initial characterization 141

of a full-length infectious sBYDV clone, and the design and testing of target-specific ribozymes against BYDV genomic RNA. In PAPER I and II, we attempted to further characterize the hammerhead structure that is believed to be required for self-cleavage of sBYDV (+) RNA. In addition to having the 11 conserved nts found in all hammerheads, the (+) strand cleavage site of sBYDV has several novel features. First, the trinucleotide at the 5' side of the cleavage site is AUA, rather than the usual GUC. Second, two unpaired bases C-24 and A-73 are located at the top of stem II. Only one other self-cleaving RNA has bases analogous to C-24 and A-73. Finally, an additional helix not accounted for in the hammerhead motif may form within the cleavage domain of sBYDV (+) RNA. From site-directed mutagenesis and structure sensitive probing of a less-than-full-length sBYDV RNA transcript, we demonstrated that the alternative structure does form and results in a very inefficient cleavage reaction. These data are consistent with the proposal (Fedor & Uhlenbeck, 1992) that all bases in and near the cleavage site affect the rate of cleavage. Deletion of both of the unpaired, nonconserved nts C-24 and A-73 had a positive influence on the cleavage rate. This further underscores the importance of many bases in the hammerhead in cleavage. PAPER III involved the cloning of a full-length, infectious sBYDV. Initial characterization of the construct 142

proved the small RNA found associated with BYDV-RPV is indeed a true satellite RNA. Furthermore, preliminary results suggest that the presence of the sBYDV with BYDV inhibits the replication of the helper virus. More rigorous testing of this observation should indicate whether this inhibition is valid or due to experimental variation. PAPER IV involved the use of the hammerhead motif of sBYDV (+) RNA for the design of a target-specific ribozyme against BYDV genomic RNA. Although we observed cleavage of BYDV RNA in vitro, cleavage of BYDV RNA in vivo will be more difficult to obtain. Such inhibitory effects as inability of the ribozyme to find and anneal to the target RNA, slow rates of product dissociation and less than optimal concentrations within the cell will most probably decrease cleavage efficacy and thus several additional ribozyme constructs will need to be tested. To effectively control yield losses imposed by BYDV infestations, a better understanding of how BYDV and sBYDV affect pathogenesis is mandatory. With full-length infectious clones of the sBYDV and helper virus, questions concerning the relationships between satellite RNAs, different helper virus serotypes, host plant species, and environment can now be addressed. 143

LITERATURE CITED

Altmann, S. (1987). Ribonuclease P: An enzyme with a catalytic RNA subunit. Advances in Enzvmoloav. 62. 1-36. Baer, M. F., Arnez, J. G., Guerrier-Takada, C., Vioque, A., & Altmann, S. (1990). Preparation and characterization of RNase P from E. coli. Methods in Enzvmoloav. 181. 569-582. Been, M. D., & Cech, T. R. (1986). Sequence specificity of Tetrahymena pre-rRNA self-splicing, trans-splicing and RNA enzyme activity. Cell. 47. 207-216. Been, M. D., Prrotta, A. T., & Rosenstein, S. P. (1992). Secondary structure of the self-cleaving RNA of Hepatitis delta virus; Applications to catalytic RNA design. Biochemistry. 31. 11843-11852. Branch, A. D., & Robertson, H. D. (1984). A replication cycle for viroids and other small infectious RNAs. Science. 223. 450-455.

Brault, v., & Miller, W. A. (1992). Translational frameshifting mediated by a viral sequence in plant cells. Proceedings of the National Academy of Sciences. 89. 2262-2266. Bruening, G. (1989). Compilation of self-cleaving sequences from plant virus satellite RNAs and other sources. Methods in Enzvmoloav. 180. 546-558. Burgin, A. B., & Pace, N. R. (1990). Mapping of the active site of ribonuclease P RNA using a subunit containing a photoaffinity agent. EMBO J.. £, 4111-4115. Cech, T. R., Zaug, A. J., Grabowski, P. J. (1981). In vitro splicing of the ribosomal RNA precursor of Tetrahymena: Involvement of a guanosine nucleotide in the excision of the intervening sequence. Cell. 27. 487-490. Cech, T. R. (1990). Self-splicing of group I introns. Annual Review of Biochemistry. 59. 543-568. 144

Gotten, M., Schaffner, G., & Birnstiel, M. L. (1989). Ribozyme, antisense RNA, and antisense DNA inhibition of U7 small nuclear ribonucleoprotein-mediated histone pre-mRNA processing in vitro. Molecular and Cellular Bioloav. 9, 4479-4487. Dahm, S. C., & Uhlenbeck, O. C. (1991). Role of divalent metal ions in the hammerhead RNA cleavage reaction. Biochemistrv. 30. 9464-9469. Davies, C., Haseloff, J., & Symons, R. H. (1990). Structure, self-cleavage, and replication of two viroid-like satellite RNAs (virusoids) of subterranean clover mottle virus. Viroloav. 177. 216-224. Davies, R. W., Waring, R. B., Ray, J. A., Brown, T. A., & Scazzocchio, C. (1982). Making ends meet: A model for RNA splicing in fungal mitochondria. Nature. 300. 719-724. Dinesh-Kumar, S. P., Brault, V., & Miller, W. A. (1992). Precise mapping and in vitro translation of a trifunctional subgenomic RNA of barley yellow dwarf virus. Viroloav. 187. 711-722. Ehrenmann, K., Schroeder, R., Chandry, P. S., Hall, D. H., & Belfort, M. (1989). Sequence specificity of the P6 pairing for splicing of the group I td intron of phage T4. Nucleic Acids Research. 17, 9147-9163. Epstein, L. M., & Gall, J. G. (1987). Self-cleaving transcripts of satellite DNA from the newt. Cell. 48. 535-543. Fedor, M. J., & Uhlenbeck, 0. C. (1992). Kinetics of intermolecular cleavage by hammerhead ribozymes. Biochemistrv. 31. 12042-12054. Forster, A. C., & Altmann, S. (1990). External guide sequence for an RNA enzyme. Science. 249. 783-786. Forster, A. C., & Symons, R. H. (1987). Self-cleavage of plus and minus RNAs of a virusoid and a structural model for the active sites. Cell. 49, 211-220. Francki, R. I. B. (1985). Plant virus satellites. Annual Review of Microbioloav. 39, 151-174. Garriga, G., & Lambowitz, A. M. (1986). Protein dependent splicing of a group I intron in ribonucleoprotein particles and soluble fractions. Cell. 46. 669-680. 145

Guerrier-Takada, C., Gardiner, K., Marsh, T., Pace, N., & Altman, S. (1983). The RNA moiety of ribonuclease P is the catalytic subunit of the enzyme. Cell. 35, 849-857. Hampel, A., Tritz, R., Hicks, M., & Cruz, P. (1990). Hairpin catalytic RNA model: Evidence for helices and sequence requirement for substrate RNA. Nucleic Acids Research. 18, 299-304. Haseloff, J., & Gerlach, W. L. (1988). Simple RNA enzymes with new and highly specific endoribonuclease activities. Nature. 334. 585-591. Haseloff, J., & Gerlach, W. L. (1989). Sequences required for self-catalysed cleavage of the satellite RNA of tobacco ringspot virus. Gene. 82. 43-52. Hernandez, C., Daros, J. A., Elena, S. F., Moya, A., & Flores, R. (1992). The strands of both polarities of a small circular RNA from carnation self-cleave in vitro through alternative double- and single-hammerhead structures. Nucleic Acids Research. 20. 6323-6329. Hernandez, C., & Flores, R. (1992). Plus and minus RNAs of peach latent mosaic viroid self-cleave in vitro via hammerhead structures. Proceedings of the National Academy of Sciences. USA. 89. 3711-3715. Hutchins, C. J., Keese, P., Visvader, J. E., Rathjen, P. D., Mclnnes, J. L., & Symons, R. H. (1985). Comparison of multimeric plus and minus forms of viroids and virusoids. Plant Molecular Biology. 4, 293-304. Hutchins, C. J., Rathjen, P. D., Forster, A. C., & Symonds, R. H. (1986). Self-cleavage of plus and minus RNA transcripts of avocado sunblotch viroid. Nucleic Acids Research. 14, 3627-3641. Jacquier, A. (1990). Self-splicing group II and nuclear pre-mRNA introns: How similar are they? Trends in Biochemical Sciences. 15, 351-354. Keese, P., & symons, R. H. (1987). The structure of viroids and virusoids. Boca Raton; CRC Press. Kruger, K., Grabowski, P. J., Zaug, A. J., Sands, J., Gottschling, D. E., Cech, T. R. (1982). Self-splicing RNA: Autoexcision and autocyclization of the ribosomal RNA intervening sequence of Tetrahymena. Cell. 31. 147-157. 146

Kuo, M., Sharmeen, L., Dinter-Gottlieb, G., & Taylor, J. (1988). Characterization of self-cleaving RNA sequences on the genome and antigenome of human hepatitis delta virus. Journal of Viroloav. 62. 4439-4444. Lazowaska, J., Jacq, C., & Slonimski, P. P. (1980). Sequence of introns and flanking exons in wildtype and box3 mutants of cytochrome b reveals an interlaced splicing protein coded by an intron. Cell. 22. 333-348. Mei, H. Y., Kaaret, T. W., & Bruice, T. C. (1989). A computational approach to the mechanism of self-cleavage of hammerhead RNA. Proceedings of the National Academy of Sciences. USA. 86. 9727-9731. Michel, F., Jacquier, A., & Dujon, B. (1982). Comparison of fungal mitochondrial introns reveals extensive homologies in RNA secondary structure. Biochimie. 64. 867-881. Michel, F., Umesono, K., & Ozeki, H. (1989). Comparison and functional anatomy of group II catalytic introns—A review. Gene, 5-30. Miller, W. A., Hercus, T., Waterhouse, P. M., & Gerlach, W. L. (1991). A satellite RNA of barley yellow dwarf virus contains a novel hammerhead structure in the self-cleavage domain. Viroloav. 183. 711-720. Miller, W. A., & Silver, S. L. (1991). Alternative tertiary structure attenuates self-cleavage of the ribozyme in the satellite RNA of barley yellow dwarf virus. Nucleic Acids Research. 19, 513-520. Miller, W. A., Waterhouse, P. M., & Gerlach, W. L. (1988). Sequence and organization of barley yellow dwarf virus genomic RNA. Nucleic Acids Research. 16, 6097-6111. Oswald, J. W., & Houston, B. R. (1951). A new virus disease of cereals, transmissible by aphids. Plant Disease Reporter. 35. 471-475. Pace, N. R., & Smith, D. (1990). Ribonuclease P; Function and variation. Journal of Biological Chemistrv. 265. 3587-3590. Perrotta, A. T., & Been, M. D. (1991). A pseudoknot-like structure required for efficient self-cleavage of hepatitis delta virus RNA. Nature. 350. 434-436. 147

Plumb, R. T. (1983). Barley yellow dwarf virus—A global problem. London: Blackwell Sciences. Prody, G. A., Bakos, J. T., Buzayan, J. M., Schneider, I. R., & Bruening, G. (1986). Autolytic processing of dimeric plant virus satellite RNA. Science. 231. 1577-1580. Rochow, W. F. (1970). Barley yellow dwarf virus: Phenotypic mixing and vector specificity. Science. 167. 875-878. Roossinck, M. J., Sleat, D., & Palukitis, P. (1992). Satellite RNAs of plant viruses; Structures and biological effects. Microbiological Reviews. 56. 265-279. Ruby, S. W., & Abelson, J. (1991). Pre-mRNA splicing in yeast. Trends in Genetics. 2» 79-85. Ruffner, D. E., Dahm, S. C., & Uhlenbeck, O. C. (1989). Studies on the hammerhead RNA self-cleaving domain. Gene. 82, 31-41. Ruffner, D. E., Stormo, G. D., & Uhlenbeck, O. C. (1990). Sequence requirements of the hammerhead RNA self-cleavage reaction. Biochemistry. 29. 10695-10702. Sarver, N., M.Cantin, E., Chang, P. S., Zaia, J. A., Ladne, P.A., Stephens, D. A., & Rossi, J. J. (1990). Ribozymes as potential anti-HIV-1 therapeutic agents. Science. 247. 1222-1225. Saville, B. J., & Collins, R. A. (1990). A site-specific self-cleavage reaction performed by a novel RNA in Neurospora mitochondria. Cell. 61. 685. Saxena, S. K., & Ackerman, E. J. (1990). Ribozymes correctly cleave a model substrate and endogenous RNA in vivo. The Journal of Biological Chemistry. 28, 17106-17109. Scanlon, K. J., Jiao, L., Funato, T., Wang, W., Tone, T., Rossi, J. J., & Kashani-Sabet, M. (1991). Ribozyme- mediated cleavage of c-fos mRNA reduces gene expression of DNA synthesis enzymes and metallothionein. Proceedings of the National Academy of Sciences. USA. 88, 10591-10595. Schneider, I. R. (1969). Satellite-like particle of tobacco ringspot virus that resembles tobacco ringspot virus. Science. 166. 1627-1629. Schneider, I. R. (1977). Defective plant viruses. Montclair: Allanheld, Osmun. 148

Sharmeen, L., Kuo, M. Y. P., Dinter-Gottlieb, G., & Taylor, J. (1988). Antigenomic RNA of human hepatitis delta virus can undergo self-cleavage. Journal of Virology. 62. 2674. Sheldon, C. C., & Symons, R. H. (1989). Mutagenesis analysis of a self-cleaving RNA. Nucleic Acids Research. 17. 5679-5685. Shiraishi, H., & Shimura, Y. (1988). Functional domains of the RNA component of ribonuclease P revealed by chemical probing of mutant RNAs. EMBO J.. 7, 3817-3821. Slim, G., & Gait, M. J. (1991). Configurationally defined phosphorothioate-containing oligoribonucleotides in the study of the mechanism of cleavage of hammerhead ribozymes. Nucleic Acids Research. 19, 1183-1188. Symons, R. H. (1989). Self-cleavage of RNA in the replication of small of plants and . Trends in Biochemical Sciences. 14 November. 445-450. Symons, R. H. (1991a). The intriguing viroids and virusoids: What is their information content and how did they evolve? Molecular Plant-Microbe Interactions. 4, 111-121. Symons, R. H. (1991b). Ribozymes. Critical Reviews in Plant Sciences. 10, 189-234. Tani, T., Takahashi, Y., & Ohshima, Y. (1992). Activity of chimeric RNAs of U6 snRNA and (-) sTRSV in the cleavage of a substrate RNA. Nucleic Acids Research. 20, 2991-2996. Thill, G., Blumenfeld, M., Lescure, F., & Vasseur, M. (1991). Self-cleavage of a 71 nucleotide-long ribozyme derived from hepatitis delta virus genomic RNA. Nucleic Acids Research. 19, 6519-6525. Uchimaru, T., Uebayasi, M., Tanabe, K., & Taira, K. (1993). Theoretical analyses on the role of Mg2+ ions in ribozyme reactions. The FASEB Journal. 7, 137-142. Uhlenbeck, O. C. (1987). A small catalytic oligoribonucleotide. Nature. 328. 596-600. Vincent, J. R., Lister, R. M., & Larkins, B. A. (1991). Nucleotide sequence analysis and genomic organization of the NY-RPV isolate of barley yellow dwarf virus. Journal of General Virology. 72. 2347-2355. 149

Waring, R. B., Towner, P., Minter, S. J., & Davies, R. W. (1986). Splice-site selection by a self-splicing RNA of Tetrahymena. Nature. 321. 133-139. Wu, H. N., Lin, Y. J., Lin, F. P., Makino, S., Chang, M. F., & Lai, M. C. C. (1989). Human hepatitis delta RNA subfragments contain an autocleavage activity. Proceedings of the National Academy of Sciences. 86. 1831-1835. Wu, H. N., Wang, Y. J., Hung, C. F., Lee, H. J., & Lai, M. M. C. (1992). Sequence and structure of the catalytic RNA of hepatitis delta virus genomic RNA. Journal of Molecular Biology. 223. 233-245.

Yamani, M. E., & Hill, J. H. (1991). Crop loss assessment and germplasm screening for resistance to barley yellow dwarf virus in west-central Morocco. Phvtopath. Medit. 30, 93-102. Zaug, A. J., Been, M. D., & Cech, T. R. (1986). Tetrahymena ribozyme acts like an RNA restriction enzyme. Nature. 324. 429-433. Zieve, G. W., & Sauterer, R. A. (1990). Cell biology of the snRNA particles. Critical Reviews of Biochemistry and Molecular Biology. 25. 1-46.