Group I and group II introns are two distinct classes of self-splicing RNA molecules that contain their own active site for intron removal and exon ligation (see RNA Splicing). These two classes of ribozymes are distinguished by their mechanisms of splicing and by their unique structures (1).
1. Reaction Mechanisms
Both classes of self-splicing introns perform two consecutive transesterification reactions in the process of exon ligation. The first step of splicing in a group I intron involves nucleophilic attack at the 5′- splice site by the 3′-OH of an exogenous, bound guanosine cofactor (Fig. 1) (reviewed in Ref. 2). This reaction adds the guanosine onto the 5′-end of the intron and releases the 5′-exon. The second step is analogous to the reverse of the first step. Following a conformational rearrangement of the active site in which the exogenous guanosine is replaced by the G nucleotide at the 3′-terminus of the intron, the 5′-exon attacks the 3′-exon boundary. This releases the intron and ligates the 5′ and 3′-exons. Both the first and second steps of the splicing reaction are fully reversible because no net energy is consumed. The result of these reactions is that the flanking exons are ligated and the intron is released as a linear molecule with an uncoded G at the 5′-end.
Figure 1. Two-step reaction mechanisms for the self-splicing group I and group II classes of introns.
Unlike the group I introns, a group II intron utilizes an internal nucleophile for the first step of splicing (reviewed in Ref. 3) (Fig. 1). The 2′-OH of a highly conserved bulged A nucleotide located within domain 6 of the intron attacks the 5′-splice site. This results in release of the 5′-exon and formation of a lariat structure whose 5′-end of the intron is covalently attached to the 2′-OH of the bulged A. Alternatively, the 5′-exon can be released by hydrolysis in which water is the nucleophile. The second step of splicing involves attack by the 5′-exon on the 3′-splice site. This results in ligation of the flanking exons and release of the lariat or linear intron. The reaction catalyzed by the group II intron follows the same mechanism as that employed in the more complex process of messenger RNA precursor splicing catalyzed by the spliceosome (4). For this reason, it has been proposed that group II intron splicing is an evolutionary precursor to pre-mRNA splicing.
2. Structures
Group I and group II introns catalyze their self-splicing reactions by folding into a distinct tertiary structures comprised of many conserved secondary structural elements (5, 6). The hallmarks of a group I intron include a common secondary structure of 10 paired segments (termed P1-P10) (Fig. 2) and several single-stranded joiner (or J) segments between the double helices (5). The "catalytic core" of the intron is made up of about 120 nucleotides that are highly conserved among group I introns isolated from a broad diversity of biological sources. The conserved nucleotides are clustered within paired regions P4, P6, P3, and P7 and the joiner regions J3/4, J4/5, J6/7, and J8/7. There is also a universal requirement for a G • U wobble pair at the 5′-exon boundary. The intron includes a guanosine binding site located in the major groove of the P7 helix, an internal guide RNA sequence that base pairs to the 5′-exon to form the P1 helix, a catalytic cleft within joiner regions J4/5 and J8/7 that orients the P1 helix into the active site for nucleophilic attack, and a terminal G nucleotide that defines the 3′-splice site. Although these elements are quite distant within the intron’s linear sequence, they converge spatially within the tertiary structure to catalyze the two splicing steps.
Figure 2. Secondary structure of the Tetrahymena group I intron showing the conserved helices P1-P10, the 5′ and 3′ splice sites (large arrows), and the guanosine binding site (gray box). The double-headed arrows indicate a few sites of tertiary interaction between different regions of the intron.
An excellent example of this structural convergence was observed within the P4-P6 domain of the Tetrahymena group I intron (7). The crystal structure of this independently folding domain of the intron was recently reported (8). This 160-nucleotide fragment is the largest RNA structure currently known and provides numerous insights into the molecular basis of RNA folding (see RNA Structure). The structure is formed from two long helical segments (helices P6, P4, P5 and helices P5b, P5a) that pack side by side. A bend of about 150° between the helices allows the two helical elements to form two extensive arrays of tertiary interactions. These contacts form between the minor groove of the P4 helix and the A-rich bulge of P5a and between the GAAA tetraloop of P5b and a tetraloop receptor in P6a/6b. Several new motifs of RNA structure and for metal binding were identified within this folding domain, including the "A-platform" and the "ribose zipper" (8, 9). When viewed from the side, the P4-P6 helix is essentially one RNA helix thick. It comprises about half of the intron active site but does not include the P1 helix or the guanosine binding site. Models have been proposed for the complete structures of four different group I introns.
Although there are substantially fewer conserved nucleotides, a group II intron also has a distinct structure (6) (Fig. 3). The intron is organized into six double-helical domains (D1-D6) that originate from a central wheel. Each of the domains has a particular function in the activity of the intron, although D1, D5, and D6 form the intron core. Domain 1 is an extended multihelical element that includes two looped regions (termed EBS1 and EBS2) complementary to two sites in the 5′-exon (termed IBS1 and IBS2). Interdomain interactions also allow domain I to serve as the scaffold upon which the intron active site is built. In contrast to domain 1, domain 5 is a relatively small (34-nucleotide) helical element, yet a large percentage of the phylogenetically conserved nucleotides is concentrated in this domain. It constitutes the catalytic center of the group II intron. D5 catalyzes 5′-splice site hydrolysis even when the domain is added trans to the rest of the intron (10). The chemical groups responsible for this catalytic enhancement map into the major groove of the D5 helix in the nucleotides surrounding a conserved G • U wobble pair (11). Domain 6 includes the branch-point adenosine whose 2′-OH nucleophilically attacks the 5′-splice site.
Figure 3. Schematic secondary structure of the group II intron. The bulged A, which acts as the nucleophile in the first step of splicing, is circled in Domain 6. The complementary sequences (EBS and IBS) between Domain I of the intron and the 5′-exon are shown as shaded lines. Arrows represent the splice sites.
3. Biological Occurrence
Group I introns are found within genes for mRNA, ribosomal RNA, and transfer RNA (1). They are extremely widespread across phylogeny and have been found in mitochondrial, chloroplast, and nuclear genomes of diverse eukaryotes, although they have not yet been observed in vertebrates. The discovery of a group I intron in T4 bacteriophage was the first example of RNA splicing observed in a prokaryote. Group I introns have also been found in eubacterial genomes. There are currently more than 450 examples of group I introns in the genomic databases. All known group II introns are located within eukaryotic organelles, including plant and fungal mitochondrial DNA and the majority of introns in plant chloroplasts (1).
4. Metals in Folding and Catalysis
Both group I and group II introns are metalloenzymes, which require divalent metal cations for activity (12). Although the metal specificity varies substantially among group I introns, it is usually satisfied by Mg or Mn . Group II splicing is substantially less efficient and requires nonphysiological concentrations of monovalent and divalent cations (as high as 2.0 M KCl and 100 mM MgCy, which play both structural and catalytic roles. A dramatic example of divalent metals in RNA folding is in the P5abc subdomain of the Tetrahymena group I intron. Three Mg ions coordinate to separate phosphate groups within the subdomain (13). This allows the RNA to fold inside-out, that is, the phosphates point into the structure and the nucleotide bases point out to the solvent. Some structural metals in RNA can often be substituted with polycations, such as spermidine or cobalt hexamine, which emphasizes the importance of charge neutralization in RNA folding.
Biochemical evidence has implicated two metal ions in the chemical transition state of group I intron splicing (14, 15). One of these metals activates the nucleophile, and the second stabilizes the leaving group during the transesterification reaction. These metal-binding sites are highly selective and cannot be substituted with a generic polycation. A two-metal active site has also been proposed for the group II intron reaction mechanism, although the evidence for this mechanism is not as complete (11). A ribozyme active site that has two metal ions is analogous to those seen in protein polymerases that catalyze the transesterification reactions of replication and transcription (16).
5. Accessory Protein Factors
Although some group I and group II introns undergo efficient self-splicing in vitro, several have no in vitro splicing activity. This is true of the majority of the group II introns that have been isolated. Splicing of these introns is likely to be promoted by accessory protein factors that assist the RNA in forming the appropriate active structure. For example, the group I intron in Neurospora pre-rRNA has no self-splicing activity unless complexed with the CYT-18 accessory protein (17). Even the yeast mitochondrial group II introns that have in vitro splicing activity require nuclear genes to splice efficiently in vivo (18).
6. Multiple-Turnover Catalysis
Both the group I and group II introns can be converted into ribozymes capable of multiple-turnover catalysis. The group I intron was converted to a true enzyme by eliminating the 3′-splice site and breaking the connection of the 5′-exon to the rest of the intron. An oligonucleotide analog of the 5′-exon was added to the ribozyme as a substrate (19) where it bound the internal guide sequence within the intron and was cleaved in a reaction equivalent to the first step of splicing. In this form, the ribozyme is not covalently altered during the reaction, and the RNA can cleave multiple substrates. Under multiple-turnover conditions where the substrate is in excess of the ribozyme concentration, the rate-limiting step is not the chemical reaction but release of the 5′-exon (20). A multiple-turnover form of the group II intron was created by deleting the 3′-exon and domains 4 and 5 (10). Upon adding independent domain 5, 5′-splice site hydrolysis of the truncated RNA was catalyzed. Analysis of this construct demonstrated that the chemical step is slower than the association or dissociation rates for substrate binding (21).
