Topological crossing in the misfolded Tetrahymena ribozyme resolved by cryo-EM

The Tetrahymena group I intron has been a key system in the understanding of RNA folding and misfolding. The molecule folds into a long-lived misfolded intermediate (M) in vitro, which has been known to form extensive native-like secondary and tertiary structures but is separated by an unknown kinetic barrier from the native state (N). Here, we used cryogenic electron microscopy (cryo-EM) to resolve misfolded structures of the Tetrahymena L-21 ScaI ribozyme. Maps of three M substates (M1, M2, M3) and one N state were achieved from a single specimen with overall resolutions of 3.5 Å, 3.8 Å, 4.0 Å, and 3.0 Å, respectively. Comparisons of the structures reveal that all the M substates are highly similar to N, except for rotation of a core helix P7 that harbors the ribozyme’s guanosine binding site and the crossing of the strands J7/3 and J8/7 that connect P7 to the other elements in the ribozyme core. This topological difference between the M substates and N state explains the failure of 5′-splice site substrate docking in M, supports a topological isomer model for the slow refolding of M to N due to a trapped strand crossing, and suggests pathways for M-to-N refolding.

The Tetrahymena thermophila group I self-splicing intron was the first RNA-only system to be identified as a catalytic RNA (1). RNAs need to be precisely folded into specific three-dimensional (3D) structures to function in biological processes such as protein translation and splicing but appear highly susceptible to misfolding (2). Like many structured RNAs, the Tetrahymena ribozyme assembles via a variety of folding pathways, many of which end up in nonnative kinetic traps (3). Under standard conditions in vitro, a small fraction folds to the native (N) state within 1 min, while the remaining fraction forms a long-lived misfolded intermediate (M) that refolds to the N state on a time scale of hours (3–5). The structural mechanism of this conformational switch remains unknown.The N state comprises an intricate catalytic core involving numerous strands and helices that bind the substrates of the splicing reaction along with an extensive ring of peripheral elements that “locks in” this core (6). Biochemical data indicate that the secondary and tertiary structures of M bear a close resemblance to N. Nevertheless, it has remained paradoxical that the transition from M to N is so slow when the states are so structurally similar to each other. A potential resolution to the paradox comes from a proposal that M and N are topologically distinct, with at least one pair of strands crossed in the core (7). The transition from the M state to the N state then cannot occur through a fast local rearrangement because of stereochemical constraints but instead requires unfolding of the ribozyme’s peripheral regions to release constraints on movement of core elements. Rearrangements of relative positions of core strands and helices in this unlocked state lead to the native-like topology, after which refolding of the peripheral tertiary contacts locks in the N state (3, 7–10). While this topological model is compelling and consistent with a wealth of biochemical data, the 3D structure of M has not been resolved, and it remains unknown which, if any, of the possible strand topological crossings are present in M.Here, we performed single-particle cryogenic electron microscopy (cryo-EM) to obtain structures of the M and N states. The structures provide unequivocal evidence for the topological isomer model, explain a large body of prior biochemical data and, by pinpointing a specific topological crossing, suggest possible pathways for transition from M to N. Our results illustrate the growing ability of cryo-EM to reveal complex structural changes in RNA folding pathways.