Sequence requirements for micro RNA processing and function in human cells

Abstract

Most eukaryotes encode a substantial number of small noncoding RNAs termed micro RNAs (miRNAs). Previously, we have demonstrated that miR-30, a 22-nucleotide human miRNA, can be processed from a longer transcript bearing the proposed miR-30 stem-loop precursor and can translationally inhibit an mRNA-bearing artificial target sites. We also demonstrated that the miR-30 precursor stem can be substituted with a heterologous stem, which can be processed to yield novel miRNAs and can block the expression of endogenous mRNAs. Here, we show that a second human miRNA, termed miR-21, can also be effectively expressed when its precursor forms part of a longer mRNA. For both miR-30 and miR-21, mature miRNA production was highly dependent on the integrity of the precursor RNA stem, although the underlying sequence had little effect. In contrast, the sequence of the terminal loop affected miRNA production only moderately. Processing of the initial, miR-30-containing transcript led to the production of not only mature miR-30 but also to the largely nuclear excision of an approximately 65-nucleotide RNA that is likely to represent an important intermediate in miR-30 processing. Consistent with this hypothesis, mutations that affected mature miR-30 production inhibited expression of this miR-30 pre-miRNA to an equivalent degree. Although point mutations could block the ability of both miR-30 and miR-21 to inhibit the translation of mRNAs bearing multiple artificial miRNA target sites, single point mutations only attenuated the miRNA-mediated inhibition of genes bearing single, fully complementary targets. These results suggest that miRNAs, and the closely similar small interfering RNAs, cannot totally discriminate between RNA targets differing by a single nucleotide.

FIGURE 1.

Analysis of miR-30 expression level. (A) A schematic of pCMV-miR-30. The predicted miR-30 precursor (Table 1 ▶) is flanked by XhoI sites, and the basal stem is, therefore, predicted to be extended by 6 bp. PA site, polyadenylation site. (B) Primer extension assay to detect miR-30 in transfected 293T cells. Lane 1, RNA from cells transfected with pBC12/CMV; lane 2, pCMV-miR-30; lanes 3–7, various miR-30 mutants. (C) Northern analysis of miR-30 expression. Lane 1, pBC12/CMV; lane 2, pCMV-miR-30; lanes 3–5, various miR-30 mutants. The arrow indicates the predicted position of the mature miR-30 and the arrowhead indicates the ∼65-nt mir-30 pre-miRNA on the Northern blot. Positions of known DNA markers are shown at left.

FIGURE 2.

Biological activity of miR-30 mutants. (A) Schematic of the indicator construct with four copies of a miR-30 target site (black boxes). The predicted base-pairing between miR-30 and the mRNA target, and the position of three miR-30 mutations are shown. (B) CAT expression in 293T cells cotransfected with the indicator plasmid pcRev and various miR-30 expression plasmids, as well as with the pBC12/CMV/β-gal internal control plasmid. CAT activities were normalized using the β-gal activities, and set at 1 for cells transfected with the pBC12/CMV negative control plasmid (marked as −). Error bars represent standard deviations.

FIGURE 3.

Analysis of miR-21 mutants. (A) Primer extension to detect mature miR-21 (lanes 1–9) or anti-miR-21 (lanes 10–12) in transfected cells. −, Cells transfected with pBC12/CMV (lane 1); +, 20 pg of a DNA oligonucleotide encoding mature miR-21 (lane 9) or the predicted anti-miR-21 (lane 12). Arrows indicate the expected position of the mature miRNA. (B) Schematic of the CAT indicator plasmid containing four miR-21 target sites (black boxes). The predicted partial base-pairing between miR-21 and its artificially designed target site, and the positions of two miR-21 point mutations, miR-21(C3) and miR-21(U), are shown. (C) CAT assay. See Figure 2B ▶ legend for details.

FIGURE 4.

Inhibition of luciferase expression by miR-30 and miR-21. (A) Schematic of the luciferase indicator plasmids bearing four miR-30 or miR-21 target sequences (black boxes). (B) miR-30 inhibits the expression of the luciferase indicator bearing 4xmiR-30 target sites. A dual luciferase assay was performed, and the ratio of firefly luciferase to the internal control Renilla luciferase, from cells cotransfected with the parental pBC12/CMV vector, was set at 1 (shown by −). (C) miR-21 inhibits the expression of the luciferase indicator encoding 4xmiR-21 target sites. Descriptions are the same as in B. (D,E) Northern analyses detecting the luc indicator mRNA, or the β-gal internal control mRNA, in 293T cells cotransfected with the pBC12/CMV control (−) or the indicated miRNA expression plasmid.

FIGURE 5.

Mutational analysis of the miR-30-luc precursor. (A) Design of miR-30-luc. Sequences underlined are derived from the firefly luciferase mRNA, with the bottom strand being antisense to the mRNA. Single point mutations are indicated. (B) Expression of mature miR-30-luc variants as determined by primer extension. −, Cells transfected with pCMV/miR-21-luc (lane 1); +, 20 pg of an oligonucleotide encoding the intended miR-30-luc, the underlined bottom strand in A (lane 3). Arrow indicates the expected position of miR-30-luc. (C) Effect of miR-30-luc on luciferase expression. 293T cells were transfected with pCMV/luc, pRL-CMV, and plasmids encoding a wild-type or mutant miR-30-luc precursor. A dual luciferase assay was performed 2 d later. The ratio of firefly luciferase to Renilla luciferase observed in cells cotransfected with the parental pBC12/CMV vector was set at 1 (shown by −).

FIGURE 6.

Subcellular localization of miRNAs. (A) Schematic of pgTat-miR-30. (B) Northern blotting for miR-30. Positions of DNA markers are shown in the left. The arrow marks the position of the mature miRNA and the arrowhead indicates the ∼65-nt pre-miRNA. Lane 1, RNA from mock transfected cells; lane 2, nuclear RNA from cells transfected with pgTat-miR-30; lane 3, cytoplasmic RNA from cells transfected with pgTat-miR-30; lane 4, nuclear RNA from cells transfected with pgTat-miR-30 and pcRev; lane 5, cytoplasmic RNA from cells transfected with pgTat-miR-30 and pcRev; and lane 6, total RNA from cells transfected with pCMV-miR-30. (C) Northern blot containing nuclear RNA (N) and cytoplasmic RNA (C) from nontransfected 293T cells was probed with an anti-miR-30-specific oligonucleotide. Labeling is the same as in B. Although detection of the endogenous anti-miR-30 required a much longer exposure than did detection of the exogenously expressed miR-30, it remains unclear why we were consistently unable to detect an endogenous miR-30 precursor signal in lane 1 of B.

FIGURE 7.

Two miRNAs can be expressed from a single transcript. Schematic of the CMV-IE-based pCMV/miR-30-miR-30-E7 and H1 promoter-based pH1/miR-30-miR-30-E7 miRNA expression plasmids. Transfected 293T cells produced both mature miR-30 and miR-30-E7, as detected by primer extension (shown by arrows). −, RNA from mock transfected cells.