Exportin-5 mediates the nuclear export of pre-microRNAs and short hairpin RNAs

Abstract

MicroRNAs (miRNAs) are initially expressed as long transcripts that are processed in the nucleus to yield approximately 65-nucleotide (nt) RNA hairpin intermediates, termed pre-miRNAs, that are exported to the cytoplasm for additional processing to yield mature, approximately 22-nt miRNAs. Here, we demonstrate that human pre-miRNA nuclear export, and miRNA function, are dependent on Exportin-5. Exportin-5 can bind pre-miRNAs specifically in vitro, but only in the presence of the Ran-GTP cofactor. Short hairpin RNAs, artificial pre-miRNA analogs used to express small interfering RNAs, also depend on Exportin-5 for nuclear export. Together, these findings define an additional cellular cofactor required for miRNA biogenesis and function.

Figure 1.

Knockdown of endogenous human Exp5 mRNA and protein expression by RNA interference. (A) 293T cells were transfected at 0, 36, and 60 h with the siExp5 RNA duplex, with the siTat duplex as a negative control, or mock-transfected. At 96 h, Exp5 mRNA expression levels were determined by Northern analysis. Ribosomal RNA served as a loading control. (B) Similar to panel A, except that 293T cells were cotransfected with pBC12/MS-HA-Ran, which expresses an HA-tagged Ran protein, at 60 h. Western analysis was performed at 96 h using a rabbit polyclonal anti-Exp5 antiserum or an HA-specific mouse monoclonal antibody.

Figure 2.

Loss of Exp5 expression specifically relieves inhibition of gene expression caused by pre-miRNAs or shRNAs. (A) Schematic representation of the pCMV-luc-miR-30(P) indicator plasmid, and of the different miR-30 RNA variants used. (B) At 60 h, coincident with the third siExp5 or mock transfection, 293T cells were cotransfected with the pCMV-luc-miR-30(P) indicator plasmid, a Renilla luciferase internal control plasmid, the indicated pBC12/CMV- or pSuper-derived control or miRNA expression plasmids, or the synthetic miR-30(siRNA) RNA duplex. Firefly and Renilla luciferase expression levels were determined at 96 h and adjusted for minor variations in the Renilla internal control. Pos (Positive control) refers to mock- or siExp5-transfected cultures that had been cotransfected with pCMV-luc-miR-30(P) and the Renilla and pBC12/CMV control plasmids. Data are presented relative to the firefly luciferase activity detected in the mock-transfected positive control culture, which was set at 100. Average of three experiments with standard deviation indicated. (C) These data were generated as described in panel B except that the indicator plasmids pCMV-luc-miR-21(P) and pCMV-luc-random were used. In addition, this panel used cells transfected with the siTat duplex as the control for siExp5. (D) These data were obtained as described in B, using the pCMV-luc-miR-30(P) indicator plasmid, and are presented as shown in B.

Figure 3.

RNAi of Exp5 exerts a specific phenotype. (A) Sequence of the Exp5 siRNA (lower strand) and of the mRNA target (upper strand) in wild-type Exp5 and in the Exp5M mutant. (B) Coincident with the third siRNA transfection, here using either siExp5 or siLuc as a control, cells were cotransfected with plasmids expressing HA-tagged versions of wild-type or mutant Exp5 or Ran. A further 36 h later, HA-tagged protein levels were determined by Western analysis. (C) This transfection experiment was performed as described in Fig. 2B, except that at 60 h the 293T cells were also cotransfected with plasmids expressing wild-type Exp5 or the Exp5M mutant. All cultures were treated with the siExp5 RNA duplex. The Pos (Positive control) culture here refers to cells transfected with the pCMV-luc-miR-30(P) indicator, the Renilla and pBC12/CMV control plasmids, and the siExp5 RNA duplex. Data are given relative to the Pos culture cotransfected with the wild-type Exp5 expression plasmid, which was set at 100.

Figure 4.

Analysis of miR-30 RNA expression. 293T cells were transfected with siExp5 or mock-transfected. At 60 h, coincident with the third siExp5 or mock transfection, cells were also cotransfected with pCMV-miR-30; 36 h later, total and cytoplasmic RNA fractions were isolated and subjected to Northern analysis. The location of the pri-miR-30, pre-miR-30, and mature miR-30 RNAs is indicated. 5S rRNA served as a loading control.

Figure 5.

Specific binding of pre-miR-30 by Exp5 in the presence of Ran-GTP. A ³²P-labeled pre-miR-30 RNA probe was incubated with recombinant Exp5-His and/or recombinant Ran-GTP in the presence or absence of the indicated fold excess of various unlabeled competitor RNAs. Protein:RNA complexes were detected by nondenaturing gel electrophoresis and autoradiography.