‘Hidden’ states are pervasive in RNA folding: detection and dissection through mutate‐and‐map experiments

Experiments that probe RNA biophysical behavior are typically interpreted in terms of the smallest number of states that account for observed data. In contrast, extensive computational work has suggested that most RNAs sample many dramatically different secondary structures at equilibrium frequencies of at least a few percent. Data bearing on this prediction are scarce. We will present detailed evidence from a novel experimental approach indicating that such alternative states are pervasive but have been missed in prior experiments. We have applied the mutate‐and‐map method, which couples comprehensive single‐nucleotide mutagenesis and single‐nucleotide‐resolution chemical mapping, to a dozen systems including riboswitches, the putative noncoding RNA HAR1, and RNAs of random sequence. The data are strikingly information rich, in some cases yielding not just models of alternative structures but also their energy differences: a view of the low‐lying energy landscape. Understanding whether these newly revealed states are tolerated, destabilized, or leveraged by functional RNAs in the cellular milieu is a critical next challenge.