Fig. 3: Discovery of a conserved 5′-G-favoured binding pocket for precise RNA cleavage.
From: DICER cleavage fidelity is governed by 5′-end binding pockets
a, Binding pockets in the DICER–26S-UG complex. Left, cryo-EM model of DICER bound to 26S-UG with colour-coded domains. Left middle, electrostatic surface view of the PAZ–platform region, showing the paths of the 5′- and 3′-ends. Right middle, close-up view of the 3′-end pocket, with A64 and labelled residues. Right, 5′-end pocket, showing U1 contacts with R821 and R1003. b, Binding pockets in the DICER–26S-GU complex. Panels as in a. The 3′-end pocket accommodates A64, whereas the 5′-end pocket binds to G1, with residues D991 and H992 indicated. c, The alignment of the RNA backbone is influenced by 5′-end binding. In DICER–26S-GU, the RNA backbone shifts one base pair upwards towards the catalytic centre, compared with DICER–26S-UG. d,e, In vitro dicing of pre-mir-517a_GU (d) and pre-mir-517a_CU (e) by wild-type (WT) and mutant DICER. Left, denaturing gels showing DC21 and DC22 products. Middle, quantification of cleavage accuracy; significance by two-tailed, two-sample t-test (***P < 0.001, ****P < 0.0001; NS, not significant). Right, pre-miRNA schematics with cleavage sites. f, Cryo-EM map of DICER(D991G/H992G)–26S-GU, with colour-coded domains and RNA duplex (26S-GU) indicated. g, Electrostatic surface view of the active-site region of the mutant DICER, highlighting the bound duplex and the 5′- and 3′-strand positions. h, Close-up view of the PAZ domain and the terminal pair G1–U62. i, Proposed mechanism: distinct 5′ binding pockets (5′-G-favoured for DC21 and 5′-U-favoured for DC22) determine cleavage-site selection.
