Evidence that mir-144 is an intrinsically suboptimal Dicer substrate. (A) The structure on the left derives from experimental SHAPE-MaP profiling of pre-mir-144. The stem is short (19bp) and the Dicer cleavage site occurs in a local asymmetric bulge; both features are incompatible with effective processing. The asterisk designates a plausible basepair that was not confirmed in the SHAPE-MaP structure. The blue triangles represent dominant reads from small RNA sequencing; however, the aqua triangle is a well-expressed isomiR of miR-144-5p. The structure on the right is a predicted structure that conforms to known rules for effective Dicer substrates. Its stem length is 21bp, and the Dicer cleavage occurs 2 nts away from the junction of the duplex and the single-stranded loop. The aqua isomiR corresponds well to a 2-nt 3′ overhang from Dicer cleavage of this predicted structure. The free energies (ΔG) of pre-mir-144 isoforms were predicted using RNAstructure. (B) Length variants in which 1–5 bp were inserted into the duplex stem just proximal to the bulge at the Dicer cleavage site. (C) The increasing stem length correlates well with decreasing accumulation of pre-mir-144, indicating compromised biogenesis. Nevertheless, insertion of two bp (144L2bp) resulted in increased mature miR-144-3p from the presumably lower level of hairpin precursor. By contrast, the changes in miR-144 biogenesis were largely segregated from those of miR-451. Effects were only seen with 144L5bp, whose substantially decreased pre-mir-144 was associated with decreased pre-mir-451 and miR-451, as expected from failure of Microprocessor enhancement on suboptimal mir-451. (D) Luciferase sensor assays confirm that 144L2bp maintains high activity, in contrast to all other length mutants that have decreased function. Activity of miR-451 was largely unaffected, except with 144L5bp.
|
