miRNP proteins and miRNA cosediment with polyribosomes. Lysates from human Weri cells prepared in the presence of cycloheximide (to preserve polyribosomes; A) or EDTA (to disrupt polyribosomes; B), were sedimented on 15%–40% sucrose gradients. Fractions were collected and aliquots of equal volume were analyzed as shown. Top and bottom of gradients are shown. RNA from each fraction was resolved on agarose gels and stained with ethidium bromide to monitor the integrity of the RNA and the presence of polyribosomes; ribosomal RNAs (28S and 18S) are indicated. RNA from indicated fractions was analyzed by Northern blots (NB) using a miR-124a specific probe. Aliquots (equal volume) from indicated fractions were analyzed by Western blots (WB) with antibodies against the Argonaute protein eIF2C2 (8C7), Gemin3 (12H12), Gemin4 (17D10), poly(A) binding protein (PABP; 10E10), Dicer (α-DICER), SMN (2B1), and Gemin2 (2E17). PABP was used as a positive control to monitor the presence of mRNAs in fast-sedimenting (polyribosome-containing) fractions. SMN and Gemin2 proteins do not associate with miRNPs and are not present in polyribosome-containing fractions.

A miRNA recognition element (MRE) for let-7b, naturally found in the 3′ UTR of human and mouse lin-28 mRNAs, is sufficient to confer let-7b-mediated translational repression to a luciferase reporter construct. (A) Potential base-pairing between LIN-28 wild-type or mutant (LIN-28, M1) MREs with let-7b is shown. Weri cells were cotransfected with indicated renilla luciferase (RL) constructs along with firefly luciferase (FL). Results shown are average values (with standard deviations) of normalized RL/FL activities obtained from three experiments. (B) Weri cells were transfected with indicated constructs, total RNA was isolated, and RL mRNA was visualized with Northern blotting; the ethidium bromide-stained gel is shown underneath. (C) Endogenous, human lin-28 mRNA associates with miRNPs in polyribosome-containing fractions only. Immunoprecipitations were performed either from polyribosomal or light fractions with anti-eIF2C2 (8C7) or nonimmune mouse IgG. RNA was isolated and amplified with RT-PCR using primers specific for lin-28 or β-actin mRNAs. (T) ~10% of the total lysate from polyribosomal or light fractions.

Demonstration of a physical association between miRNP proteins and a miRNA (let-7b) in polyribosome-containing fractions. Weri cell lysates were separated in heavy, polyribosome-containing (P) and non-polyribosome-containing, light fractions (L). (A) Western blot (WB) against PABP. (B) Immunoprecipitations were performed from P and L fractions with an antibody against eIF2C2 (8C7) or nonimmune mouse IgG as a negative control. Approximately 10% of the 8C7 and IgG immunoprecipitates from the polyribosomal pellet and ~50% of the 8C7 immunoprecipitate from the light fraction were loaded on the gel and analyzed by Western blots with anti-Gemin3 (12H12) and anti-Gemin4 (17D10) antibodies. Asterisk shows nondissociated IgG molecules. (C) Immunoprecipitations were performed from the polyribosomal fraction with 8C7, or nonimmune mouse IgG; RNA was isolated and analyzed by Northern blots using oligonucleotide probes against let-7b or 28S rRNA. (T) ~10% of the total lysate from polyribosomal pellet.

Proposed model for miRNA:target mRNA association. We hypothesize that miRNAs and the miRNP proteins associate with their mRNA targets in polyribosomes. Additional, as yet unidentified factors (shown in green), may also associate with the polyribosomal miRNP. We favor a direct miRNA:mRNA association guided by miRNP proteins; however, direct interactions between miRNP proteins and the translational machinery (depicted with an arched line) are also likely to occur.