A passive protein environment offsets RNA folding energetics in cells

RNA structure can be predicted from sequence in dilute solution, but these predictions often fail in cells. Many cellular RNAs keep their in vitro folds, whereas others are substantially less structured. We combined single-molecule FRET with Xenopus oocyte extract to measure RNA folding and duplex formation under cell-like conditions. Extract proteins passively suppressed base pairing: they slowed apparent strand association while leaving duplex dissociation largely unchanged, without sequence specificity or ATP consumption. A competitive binding model captured this activity as an effective folding penalty of ~0.45 kcal/mol per nucleotide. Similar activity occurred in HeLa lysate. Applied to in-cell chemical probing data, this correction stratified reactivity across ~42,000 stem-loops, whereas predictive models established in dilute solution classified nearly all as stably folded. Our results identify a passive protein-mediated mechanism as a key contributor to cellular RNA folding, shifting the balance between paired and unpaired states without changing the underlying base-pairing rules.