Introduction

Since their discovery in the early 1980s, RNA enzymes or ribozymes have been the subject of much investigation. Numerous studies have been performed to elucidate the biochemistry of how certain RNA molecules can fold into complex tertiary structures to form active sites and perform catalysis. Other studies have focused on identifying the roles that RNA enzymes play in cell biology. More recently, even more attention has been focused on the study of ribozymes because it was recognized that these RNA enzymes can potentially be quite useful for a variety of gene therapy applications.

The first discovered ribozyme was the self-splicing group I intron from Tetrahymena thermophila. The reaction mediated by this RNA enzyme has now been extensively characterized and the mechanism by which it excises itself from precursor ribosomal RNAs (pre-rRNA) without the aid of proteins is well understood (1-3). The second ribozyme to be recognized was the RNA subunit of RNase P. RNase P catalyzes the removal of upstream sequences on precursor-tRNAs to produce mature 5' ends on tRNA molecules in a wide variety of cell types (3,4).

Several other catalytic RNA motifs have been discovered that are naturally associated with plant and human pathogens. The hammerhead and hairpin ribozymes are derived from satellite RNAs from plant viroid and virusoids, and the hepatitis delta virus (HDV) ribozyme is derived from a short single-stranded RNA virus found in some patients with hepatitis B virus. Each of these small RNA enzymes catalyzes a self-cleavage reaction that is believed to play a major role in the replication of these single-stranded RNA pathogens (5).

All these self-cleaving ribozymes have been reengineered so that they can cleave other target RNA molecules in trans in a sequence-specific manner.(6) This ability to specifically cleave targeted RNAs has led to much speculation about the potential utility of trans-cleaving ribozymes as inhibitors of gene expression (7-11). In addition, the group I self-splicing ribozyme from Tetrahymena can be reengineered to perform splicing on a targeted RNA molecule in trans. It has been argued that such trans-splicing ribozymes may prove to be effective at repairing mutant cellular transcripts by cleaving off mutant nucleotides and ligating on functional RNA sequences (12-14).

The purpose of this chapter is not to provide an extensive review of the enzymology of ribozymes or to catalog the published results demonstrating that ribozymes may become useful reagents for gene therapy applications. Both of these topics have been extensively reviewed elsewhere (2-14). Moreover, we do not discuss the use of synthetic ribozymes, and leave the description of various gene transfer and expression systems that can be employed to deploy ribozymes to the other chapters in this book. Rather, we attempt to present a focused account of the potential utility of catalytic RNAs for gene therapy by first presenting an overview of the basic biochemistry of well-characterized ribozymes and then discussing how trans-cleaving and trans-splicing ribozymes may be employed for a variety of gene therapy applications. Our hope is that this approach will enhance the reader's understanding of the potential utility of ribozymes for both gene inhibition and genetic repair.

II. CATALYTIC RNAs

Five classes of catalytic RNAs have been extensively characterized. Each class of ribozyme adopts a characteristic secondary and tertiary structure that is required to assemble a catalytic center and perform catalysis (Fig. 1) (6). In addition, these classes of ribozymes differ in size, and the mechanism

Figure 1 Secondary structures of five classes of ribozymes. All 5 ribozymes, shown in solid black lines, have been engineered to cleave specific target RNAs, shown as dashed lines. The base pairing formed between the ribozymes and their target RNAs are shown and the site of cleavage of the substrate RNAs is indicated by an arrow. (A) The group I ribozyme. P1 through P9 represent the conserved based paired regions found in group I introns. The internal guide sequence (IGS) is shown paired to a target RNA. The substrate is cleaved just 3' of the conserved G-U wobble base pair. (B) RNase P. RNase P holoenzyme is depicted as a cartoon and contains an RNA subunit as well as a protein cofactor. The substrate for RNase P cleavage is indicated bound to an external guide sequence (EGS) just 5' of a free 5'-NCCA-3' sequence. Cleavage of the targeted RNA occurs just across from the end of the EGS-target duplex. (C) The hammerhead ribozyme. The hammerhead ribozyme is shown bound to a target RNA through base pairing interactions formed by stems I and III. The-single-stranded regions encompass the catalytic core of the ribozyme. Cleavage of the substrate RNA occurs at an unpaired residue positioned between stems I and III. (D) The hairpin ribozyme. The hairpin binds its target RNA through 2 base pairing regions called helix 1 (H1) and helix 2 (H2). Cleavage of the target RNA occurs between the N and G nucleotides on the substrate as indicated. (E) The HDV ribozyme. The HDV ribozyme forms 7 or 8 base pairs with its target RNA and cleavage occurs just 5' of this base pairing interaction.

that each employs to perform catalysis varies. Hammerhead, hairpin, and the HDV ribozymes are only 30 to 80 nucleotides in length and form cleavage products with 2',3' cyclic phosphate and 5'-hydroxyl termini. By contrast, catalytic RNAs derived from group I introns and RNase P are typically greater than 200 nucleotides in length and both cleave target RNAs to generate products with 3'-hydroxyl and 5'-phosphate termini. Each class of ribozyme is discussed in more detail below.

A. The Group I Intron from T. thermophila

The intervening sequence (IVS) found in nuclear precursor rRNA transcripts from T. thermophila is one of the most well-characterized catalytic RNAs. This IVS is a member of a growing family of group I introns that have common structural and functional features. The Tetrahymena intron is naturally 413-nucleotides long and is found in the middle of the 26S rRNA gene. The IVS is transcribed as part of the rRNA precursor and excises itself by performing a cleavage and a ligation reaction to form a functional rRNA without the aid of proteins (1,15). Prior to the discovery of this self-splicing reaction, RNA was believed to be only a carrier of genetic information or a scaffold for protein binding and not able to perform catalysis on it own.

B. The Self-splicing Reaction of the Tetrahymena Group I Intron

Comparative sequence analysis of several group I introns (16,17) as well as mutational analyses (18-26) have been employed to develop a phylogenetically conserved prediction of the secondary structure of group I introns consisting of a set of paired regions, P1-P9. More recently, X-ray crystallography and chemical probing studies have revealed that the Tetrahymena intron adopts a particular 3- dimensional structure using several tertiary interactions (27-34) and contains a catalytic core surrounded by a close-packed layer of RNA helices (32-34). Several studies have shown that this folded RNA structure participates directly in self-splicing (35,36).

The Tetrahymena intron excises itself and ligates together its flanking exons by performing 2 consecutive transesterifi-cation reactions (2,15). The first step of splicing is a cleavage reaction that is initiated by a free guanosine that is bound by the intron. This guanosine serves as a nucleophile attacking the 5' splice site and is covalently attached to the 5' end of the intron (Fig. 2A). The recognition element that defines the exact site of guanosine attack is a G-U wobble base pair that is highly conserved among group I introns. The G-U base pair is part of a short duplex called P1. The P1 duplex includes base pairing between the last 6 nucleo-tides of the 5' exon and sequences within the intron called the internal guide sequence (IGS) or the 5' exon-binding site. In the second step of splicing, the newly generated 3'-hydroxyl group, at the 3' end of the cleaved 5' exon, attacks the phosphorous atom at the 3' splice site, resulting in the ligation of the 5' and 3' exons and the excision of

Figure 2 Group I ribozyme-mediated self- and trans-splicing. (A) Self-splicing is initiated by the attack of the 5' splice site by an intronbound guanosine. This cleavage occurs just 3' of the uridine shown in red that is involved in a G-U wobble base pair. The group I intron (green) holds onto the 5' cleavage product via base pairs formed between the internal guide sequence of the ribozyme and the 3' end of the 5' exon. In the second step of splicing, the intron attaches the cleaved 5' exon (blue) onto the 3' exon (yellow) and liberates itself. (B) During trans-splicing, the ribozyme binds to a sequence in a target RNA (5'-NNNNNU-3') via base pairing through its internal guide sequence (5'-gn'n'n'n'n'n'-3'). The ribozyme cleaves the target RNA at the reactive uridine (red), releases the downstream cleavage product (light blue) and ligates a 3' exon (yellow) onto the upstream cleavage product (dark blue). See the color insert for a color version of this figure.

Figure 2 Group I ribozyme-mediated self- and trans-splicing. (A) Self-splicing is initiated by the attack of the 5' splice site by an intronbound guanosine. This cleavage occurs just 3' of the uridine shown in red that is involved in a G-U wobble base pair. The group I intron (green) holds onto the 5' cleavage product via base pairs formed between the internal guide sequence of the ribozyme and the 3' end of the 5' exon. In the second step of splicing, the intron attaches the cleaved 5' exon (blue) onto the 3' exon (yellow) and liberates itself. (B) During trans-splicing, the ribozyme binds to a sequence in a target RNA (5'-NNNNNU-3') via base pairing through its internal guide sequence (5'-gn'n'n'n'n'n'-3'). The ribozyme cleaves the target RNA at the reactive uridine (red), releases the downstream cleavage product (light blue) and ligates a 3' exon (yellow) onto the upstream cleavage product (dark blue). See the color insert for a color version of this figure.

the intron. The excised group I intron maintains its ability to make and break phosphodiester bonds. However, because the group I intron is not regenerated in its original form following self-splicing, this catalytic RNA is not a true enzyme in the strictest sense (37). Subsequently, shortened forms of the Tetrahymena group I intron that lack exon sequences were generated that fulfill the definition of a true enzyme in that they can perform multiple turnover reactions without being modified in the process (38).

C. The Trans-cleaving Reaction of the Tetrahymena Ribozyme

Shortened versions of the intervening sequence from Tetrahymena that lack the first 19 or 21 nucleotides (called L-19 or L-21) can catalyze the cleavage of oligonucleotide substrates with multiple turnover (38). Moreover, the rate enhancement achieved by this shortened form of the intron is within the range of values achieved by protein enzymes, such as EcoRI, that catalyze sequence-specific cleavage of nucleic acids. The catalytic mechanism employed by the shortened form of the Tetrahymena ribozyme in this trans-cleavage reaction is quite similar to that used by the full-length intron in the cleavage step of self-splicing with a few exceptions (39). First, the 5' exon sequences preceding the IGS have been removed in the shortened form of the intron. Therefore, the ribozyme must bind to a target RNA that is present in trans (Fig. 2B). As in the case of the self-splicing reaction, a wobble G-U base pair is required at the cleavage site and the IGS of the ribozyme must be complementary to the sequence found on the substrate RNA just 5' of the reactive uridine residue (40). For the wildtype IGS (5'-GGAGGG-3'), binding would occur at nucleo-tides 5'-CCCUCU-3' within the substrate RNA. Cleavage occurs just 3' of the uridine residue on the substrate RNA at the reconstructed G-U base pair (Fig. 2B).

As noted, substrate recognition and trans-cleavage require base pairing between the IGS on the ribozyme and the RNA substrate. Substrate specificity can be manipulated by altering the sequence of the ribozyme's IGS to make it complementary to any target RNA molecule (40-42). Moreover, no specific sequence requirements exist for the IGS, except that it must contain a guanosine residue at the reaction site. Thus, by altering the guide sequence of the L-21 version of the Tetrahymena catalytic RNA, a ribozyme can be created that can be employed to recognize and cleave a target RNA following any uridine residue (Fig. 2B).

The L-21 form of the Tetrahymena ribozyme binds to a 6-nucleotide long substrate RNA 103- to 104-fold tighter than would be predicted by base pairing binding energy alone (43-45). Studies on the L-21 ribozyme, as well as the self-splicing form of the intron, suggest that tertiary interactions contribute to the ribozyme substrate-binding energy. Specific tertiary interactions have been identified that involve 2'-hy-droxyl groups on the ribose backbone of the substrate and the intron (46-49). This tight binding between the ribozyme and RNA substrate limits the substrate specificity of the Tetrahy-mena ribozyme because both matched substrates, which are only 6 nucleotides long, and substrates that form single base pair mismatches with the IGS serve as excellent substrates for the ribozyme (44,45). Under conditions of saturating gua-nosine and 10 mM MgCl2, such RNA substrates are bound so tightly that the ribozyme goes on to cleave essentially every RNA that it binds. Therefore, the substrate specificity of this ribozyme will probably have to be improved if it is to become useful for gene therapy because any 6 nucleotide sequence would be expected to be present in many cellular RNAs. Fortunately, several logical approaches exist that can potentially be employed to enhance the substrate specificity of the Tetrahymena ribozyme (50,51).

If the specificity of the group I ribozyme proves very difficult to improve, then trans-cleaving ribozymes derived from group II introns may represent an alternative catalytic RNA motif that may be able to achieve the high levels of substrate specificity that might be required for gene therapy applications. At least some of these group II autocatalytic RNAs, which are present in organelles of plants, lower eukaryotes, and prokaryotes, do not appear to form additional tertiary interactions with their substrates (52). Thus, group II introns may be particularly adept at recognizing only the intended RNA sequence in the pool of cellular transcripts.

D. The Trans-splicing Reaction of the Tetrahymena Ribozyme

In addition to performing a trans-cleavage reaction, the Tetrahymena ribozyme can also mediate targeted trans-splicing by employing intermolecular cleavage and ligation reactions (41,53). During trans-splicing, a ribozyme with a 3' exon attached to its 3' end recognizes a target RNA (5' exon) by complementary base pairing as in trans-cleavage (Fig. 2B). The ribozyme then cleaves its target RNA as usual at a site immediately 3' of a conserved G-U base pair formed between a guanosine nucleotide at the 5' end of the IGS and a uridine nucleotide within the substrate RNA. The sequences downstream of the cleavage site (3' cleavage product) are then released by the ribozyme. The 3' end of the 5' cleavage product is then attached to the 3' exon, which is originally appended to the ribozyme, to generate the ligated product (Fig. 2B).

As with the trans-cleavage reaction, targeted trans-splicing is very malleable. In principle, any uridine residue in an RNA molecule can be targeted for trans-splicing by simply making the nucleotides in the ribozyme's IGS complementary to the nucleotides that precede an available uridine residue in a targeted RNA. No specific sequence requirements for the 3' exon are know to exist for this splicing reaction. Thus, essentially any RNA sequence can be employed as a 3' exon in this reaction and spliced onto a targeted 5' exon as long as the 3' exon sequences do not inhibit the ribozyme from folding into a catalytically competent conformation.

E. The RNase P Ribozyme

RNase P, unlike the other 4 ribozymes discussed in this chapter, is the only catalytic RNA that is naturally a true enzyme. RNase P is found in both prokaryotic and eukaryotic cells, where it catalyzes the removal of the 5' leader sequences from the variety of precursor tRNAs (3,4). This catalytic RNA is naturally part of a ribonucleoprotein (RNP) and in the case of RNase P isolated from Escherichia coli the RNP consist of a 377-nucleotide RNA subunit (M1) and a 119-amino acid protein (Fig. 1B). Although it was initially believed that the holoenzyme was required to perform catalysis in vitro, M1 RNA preparations from E. coli (54) as well as in vitro-tran-scribed versions of M1 RNA (55) are able to cleave tRNA precursors with multiple turnover in the presence of high concentrations of magnesium in the test tube. RNase P cleaves substrate RNA by hydrolysis to generate 5'-phosphate and 3'-hydroxyl termini. Although the RNA subunit of RNase P alone is catalytic, the protein cofactor facilitates the RNA processing reaction and allows it to proceed efficiently under physiologically relevant conditions (55).

When processing its natural pre-tRNA substrate, RNase P removes the 5' leader sequences from the end of the precursor transcript. Cleavage occurs specifically and accurately just 5' of the first nucleotide in the mature tRNA, even though only a small degree of sequence conservation exists between different pre-tRNA species. This observation suggested that some facet of the 3-dimensional structure of the precursor tRNAs is the feature of the transcript that is recognized by RNase P. Mutagenesis studies supported this hypothesis because disruption of tRNA folding was shown to decrease the rate of RNase P-mediated cleavage of substrate RNAs. However, the full tertiary structure of tRNA is not required for RNase P recognition and RNA processing (56). Rather the ribozyme appears to recognize a short RNA duplex similar to the acceptor stem of a tRNA just upstream of an unpaired CCA sequence found on the 3' end of partially processed tRNA transcripts (56). The structure recognized by RNase P can be approximated by a short RNA fragment, termed the external guide sequence or EGS, that is complementary to a substrate (Fig. 1B). RNase P will cleave single-stranded 5' leader sequences adjacent to any double-stranded RNA duplex as long as the unpaired CCA nucleotides are present at the 3'end of the EGS (Fig. 1B) (56). Thus, through the use of EGS oligonucleotides, RNase P can in principle be targeted to cleave any target RNA.

F. The Hammerhead Ribozyme

The hammerhead ribozyme is a catalytic RNA motif that was originally derived by comparing the self-cleavage domains from a number of naturally occurring viroid and satellite RNAs that replicate in plants (5). The self-cleaving consensus domain consists of a highly conserved catalytic region and 3 helices, and has been shown to have sequence-specific ribo-nuclease activity (Fig. 1C). Subsequently, a hammerhead domain of less than 60 nucleotides was shown to be sufficient for cleavage (57,58), and two separate oligonucleotides that assembled into a hammerhead structure were shown to mediate a trans-cleavage reaction (59,60).

Crystallographic studies of the hammerhead ribozyme have demonstrated that the tertiary structure of the hammerhead appears to be ''Y shaped'' or like that of a ''wishbone'' (61,62). Helices I and II are in close juxtaposition, whereas helix III is at the bottom of the molecule. However, all 3 are A-form helices. As shown in Fig. 1C, the minimal structural requirements for hammerhead-catalyzed cleavage include two single-stranded regions that contain 9 highly conserved nu-cleotide sequences, 3 helices, and the nucleotides GUN immediately 5' of the cleavage site in the substrate RNA. Results of mutagenesis and kinetic studies suggest that the conserved residues comprise the catalytic core of the ribozyme and are required for cleavage (63,64). Helices I and III, which flank the cleavage site, are formed by base pairing with the substrate. This base pairing interaction is extremely important not only because it holds the ribozyme and substrate together, but also because it precisely positions the ribozyme relative to the cleavage site. The most efficient cleavage has been observed with GUC, GUA, or GUU at the cleavage site although some cleavage also occurs after CUC, UUC, and AUC. The hammerhead catalyzes transesterification of the 3', 5'phosphodies-

ter bond at the cleavage site, which results in the production of RNA with 2', 3' cyclic phosphate and 5'-hydroxyl termini.

Much effort has been focused on developing hammerhead ribozymes into useful therapeutic agents. The hammerhead's small size and simple secondary structure, containing helices

1 and III, which can be made to base pair with virtually any substrate RNA, has allowed a great number of investigators to design hammerhead ribozymes to target any RNA molecule for cleavage and destruction (discussed in detail below). In particular, efforts have been directed at optimizing the interaction between the ribozyme and its substrate because both the length and base composition of complementary helices I and III can affect substrate specificity, ribozyme substrate affinity, and rate of reaction turnover.

G. The Hairpin Ribozyme

As is the case for the hammerhead ribozyme, the hairpin ribo-zyme represents a catalytic RNA motif that is derived from RNA associated with a plant pathogen (5). This small catalytic RNA was discovered in the 359-nucleotide long negative strand satellite RNA of Tobacco Ringspot virus [(- )sTRSV], which was shown to mediate a self-catalyzed cleavage reaction as part of its replication pathway (57,65).

A minimal catalytic domain for this RNA molecule has been identified (66,67), which consists of a 50-base RNA catalyst that efficiently cleaves an RNA substrate containing 14 bases of satellite RNA sequence (68). Features of secondary structure within this domain, defined from minimum energy RNA folding calculations and supported by mutagenesis studies, include 4 helices, 2 of which are formed by base pairing between the RNA catalyst and the substrate (Fig. 1D) (69). These 2 helices form part of the substrate recognition site and flank a 4-base loop within the substrate (5'-AGUC-3') containing the cleavage site. For the wild-type (- )sTRSV, cleavage occurs between the nucleotides, A and G, by transes-terification generating a 5' fragment with 2', 3' cyclic phosphate termini and a 3 fragment with a 5 -hydroxyl termini. Trans-cleavage has also been observed in vitro with multiple turnover when substrate RNAs are added to the hairpin ribo-zyme as separate transcripts (68,69).

The hairpin catalytic RNA motif can be designed to target a great variety of RNA molecules for cleavage because only

2 sequence requirements exist for this reaction. First, to maintain catalytic activity complementary base pairing between the ribozyme and substrate must occur. Single base pair mismatches at the 10 positions included in the 2 flanking helices can result in the loss of catalytic activity (69), although single base pair mismatches distal to the cleavage site appear to be tolerated (70). In addition, it has been noted that the composition of base paired substitutions in these helices can have a wide range of affects on the kinetic properties of the ribozyme (69), suggesting that base pair substitutions should be optimized for each application. The second sequence requirement for the hairpin cleavage reaction involves the nucleotides that compose the target site. Optimal substrate cleavage occurs with the nucleotides GUC immediately 3 of the cleavage site.

The guanosine residue appears to be essential and is believed to be directly involved in catalysis (71). Moreover, catalytic activity has been shown to vary widely when nucleotide substitutions are made at the other 3 positions (71).

H. The HDV Ribozyme

HDV is a 1700-nucleotide, covalently closed circular RNA that is associated with hepatitis B virus infection in certain patients. This animal RNA virus undergoes autocatalytic self-cleavage as part of its replication cycle (5). A minimal self-cleaving RNA motif have been determined for the HDV catalytic RNA (72,73), which includes approximately 85 nucleotides from both the genomic (74) and antigenomic RNAs (75). Features of secondary structure were proposed based on nuclease probing and/or site-directed mutagenesis, and were found to be similar for both the genomic and antigenomic self-cleaving sequence elements (75). More recently, X-ray crystal structure studies have been employed to solve the 3-dimensional structure of the genomic HDV ribozyme (76). Such analysis demonstrates that the ribozyme forms 4 stems, 2 of which (stems I and II) generate a tertiary interaction called a pseudoknot (Fig. 1E) (75,76). Stems II, III, and IV appear to be important for stabilizing the catalytic form of the ribozyme, whereas formation of stem I is required for efficient cleavage (75).

The self-cleavage reaction mediated by the HDV ribozyme occurs by transesterification and is dependent on divalent cations. Cleavage products have 5'-hydroxyl and 2',3' cyclic phosphate termini. The rates of HDV self-cleavage in vitro using either in vitro-transcribed HDV RNA or HDV RNA isolated from infected tissue appears to be very slow. However, addition of urea or formamide has been shown to increase the rate of cleavage as much as 50-fold, suggesting that these denaturants may be mimicking a viral or cellular RNA-binding or unwinding factor that facilitates cleavage in vivo (77). A derivative of the catalytic RNA motif from HDV has been engineered to catalyze cleavage reactions in trans with multiple turn over (78,79). Target site recognition is dependent on the formation of 7 to 8 base pairs formed between the target RNA and the HDV ribozyme 3' of the cleavage site (79).

Was this article helpful?

0 0

Post a comment