The 413 nucleotide self-splicing group I intron from Tetrahymena thermophila pre-rRNA contains a 160 nucleotide independently folding domain of RNA tertiary structure, the P4-P6 domain. This domain consists of sequence elements highly conserved among group I introns (P4 and P6) and peripheral extensions conserved in certain subgroups of these introns (P5abc and P6ab). The effect of mutation of selected bases on the formation of domain structure was analyzed using two probes: solvent-based Fe(II)-EDTA, which monitors backbone accessibility and dimethyl sulfate, which monitors availability of N(1) of adenine and N(3) of cytosine. A GAAA tetraloop and an adenosine-rich bulge were found to stabilize domain tertiary structure in a sequence-specific manner. A single base change in the GAAA tetraloop disrupted Fe(II)-EDTA protection both locally and in P6a, and a specific base-pair substitution in P6a similarly disrupted protection locally and in the tetraloop; thus remote elements of the secondary structure are linked in tertiary structure. Our model of the domain's tertiary structure is refined to include this long-range tertiary interaction. The interaction requires severe bending of the domain RNA such that sequences separated by approximately 50 bases of largely double-stranded RNA are in proximity in the tertiary structure. The bending causes or allows for contact between sequences of the conserved core and sequences of the P5 extension. Thus the P5 extension may serve to stabilize the structure of the intron core in vivo.

The P4-P6 domain RNA of the group I intron of Tetrahymena thermophila has previously been shown by chemical probing to be an independently folding domain of the intron's tertiary structure. To directly visualize this tertiary structure, the P4-P6 domain and two folding defective mutants were prepared for high-resolution electron microscopy using tungsten shadowcasting. In the presence of Mg2+, the P4-P6 domain predominantly consists of compact molecules, while the two mutant RNAs are nearly all rod-like molecules. The measured length of the rod-like molecules is 64 (+/- 6) bp, which agrees closely with the length expected for molecules containing secondary structure only. In the absence of Mg2+, the P4-P6 domain contains threefold or tenfold fewer compact structures (depending on the mounting procedures) than in the presence of Mg2+. These results provide direct evidence for the overall shape of the tertiary structure proposed on the basis of biochemical experiment, and they confirm the Mg2+ dependence of tertiary folding. An equilibrium between the extended (rod-like) and the compact structures is suggested, with the concentration of bound Mg2+ and different mounting methods influencing the direction of the equilibrium. The entire group I ribozyme (L-21 Sca I RNA) was also examined by electron microscopy in the presence of Mg2+, and was revealed to have a compact shape. These studies present a direct demonstration of long-range interactions in a catalytic RNA molecule.

The L‐21 Tetrahymena ribozyme, an RNA molecule with sequence‐specific endoribonuclease activity derived from a self‐splicing group I intron, provides a model system for studying the RNA folding problem. A 160 nucleotide, independently folding domain of tertiary structure (the P4‐P6 domain) comprises about half of the ribozyme. We now apply Fe(II)‐EDTA cleavage to mutants of the ribozyme to explore the role of individual structural elements in tertiary folding of the RNA at equilibrium. Deletion of peripheral elements near the 3′ end of the ribozyme destabilizes a region of the catalytic core (P3‐P7) without altering the folding of the P4‐P6 domain. Three different mutations within the P4‐P6 domain that destabilize its folding also shift the folding of the P3‐P7 region of the catalytic core to higher MgCl2 concentrations. We conclude that the role of the extended P4‐P6 domain and of the 3′‐terminal peripheral elements is at least in part to stabilize the catalytic core. The organization of RNA into independently folding domains of tertiary structure may be common in large RNAs, including ribosomal RNAs. Furthermore, the observation of domain‐domain interactions in a catalytic RNA supports the feasibility of a primitive spliceosome without any proteins.

Coaxial stacking of helical elements is a determinant of three-dimensional structure in RNA. In the catalytic center of the Tetrahymena group I intron, helices P4 and P6 are part of a tertiary structural domain that folds independently of the remainder of the intron. When P4 and P6 were fused with a phosphodiester linkage, the resulting RNA retained the detailed tertiary interactions characteristic of the native P4-P6 domain and even required lower magnesium ion concentrations for folding. These results indicate that P4 and P6 are coaxial in the P4-P6 domain and, therefore, in the native ribozyme. Helix fusion could provide a general method for identifying pairs of coaxially stacked helices in biological RNA molecules.

In catalysis by group I introns, the helix (P1) containing the RNA cleavage site must be positioned next to the guanosine binding site. We have identified a conserved adenine in the catalytic core that contributes to the stability of this arrangement and propose that it accepts a hydrogen bond from a specific 2'-OH in P1. Such base-backbone tertiary interactions may be generally important to the organization of RNA tertiary structure.

A shortened form of the intervening sequence of the self-splicing RNA from Tetrahymena thermophila catalyzes sequence-specific cleavage of RNA. Cleavage site selection involves a base-pairing interaction between the substrate RNA and a binding site within the intervening sequence. Single-base changes in this binding site were previously shown to alter substrate specificity in a predictable manner. To examine the generality with which substrate specificity can be altered, six variant catalytic RNAs (ribozymes) have been produced with two- or three-base changes in the active site. Each ribozyme cleaves its predicted substrate. The conditions required for good reactivity and for discrimination against cleavage at mismatched sites vary and were independently determined for each ribozyme.