Comparison of miR-122 inhibitor strategies in cultured cells. (a) miRNA inhibitor constructs. (b) Pairing of inhibitors (black) to miR-122 (red). (c) Plasmid harboring nLacZ reporter gene with one or three sites complementary to miR-122 was co-transfected into HuH-7 cells with pTBG-driven firefly luciferase (FLuc) and either control, anti-miR-122 sponge or U6-driven anti-miR-122 TuD plasmid. The cells were stained for nLacZ expression 48 h after transfection, and the proportion of blue cells were counted and reported relative to a control reporter lacking miR-122-binding sites. n = 3 independent experiments. (d) Reporter plasmid expressing nLacZ mRNA containing 3 miR-122-binding sites was cotransfected into HuH-7 cells with a U6-driven sponge-, miRZip-, TuD-expressing, or empty (control) plasmid. n = 4 . (e) HEK 293 cells were transfected with an nLacZ reporter plasmid containing three fully complementary miR-122-binding sites together with constructs expressing anti-let-7 or anti-miR-122 TuD transcribed from a U6 promoter or anti-miR-122 sponge or anti-let-7 sponge transcribed from an SV40 promoter, as well as different amounts of a plasmid producing pri-miR-122 RNA. Percentages of nLacZ positive cells relative to the control (nLacZ without miR-122-binding sites), were determined after 48 h (c, d, and e). n = 3. (f) HuH-7 cells were transfected with plasmid expressing control Renilla reniformis luciferase (RLuc) and FLuc bearing seven miR-122-binding sites or seven mutant sites; these were either alone or in the presence of plasmid expressing anti-miR-122, anti-let-7 or a scrambled TuD RNA control. Luciferase activity was assayed after 24 h and is presented as the mean ratio of RLuc to Fluc ± s.d. (n = 3). (g and h) Evaluation of let-7 inhibitor constructs in HeLa cells. Total RNA and protein were prepared from HeLa cells transfected with plasmids expressing either anti-miR-122 or anti-let-7 TuD or anti-let-7 sponge or control plasmid. Relative Dicer mRNA abundance was measured by qRT-PCR (g) and Dicer protein abundance by western blotting (h). Values are mean ± s.d. (n = 3).

Real-time monitoring of endogenous miRNA activity using miRNA sensor system. (a) Schematic presentation of GLuc-expressing vectors. CB, chicken β actin promoter with CMV enhancer. AAV vector plasmids were transfected into HuH-7 (n = 6 independent trials; b) or HeLa cells (n ≥ 17; c). GLuc activity was measured after 48 h. (d, e) C57BL/6J mice were administered 1 × 10¹² genome copies of scAAV9 per animal by tail vein injection. Blood was collected at the indicated times and assayed for GLuc activity, reported as mean ± s.d., relative to injected GLuc vector lacking both the TuD expression cassette and 3′ UTR miRNA-binding sites. n = 4 for each group.

Analysis of miRNA expression in liver from mice administered scAAV9 expressing anti-miRNA TuD. C57BL/6J mice were injected via tail vein with 1 × 10¹² genome copies of control, anti-miR-122 or anti-let-7 TuD expressing vectors. Animals were sacrificed four weeks later, and the abundance of let-7, miR-122, miR-26a, miR-22 and U6 in total liver RNA was measured by qRT-PCR (a; n = 3 per group) and northern hybridization (b; n = 4 per group). Values are mean ± s.d. U6 RNA provided a loading control. (c, d) High-throughput sequencing of total liver small RNA was used to determine the length and abundance of genome-matching (c) or prefix-matching (d) miR-122 isoforms four weeks after injection. The most abundant non-genome matching nucleotides added to the 3′ end of miR-122 fragments are indicated in the grey boxes.2(e) Nucleotide differences among the nine let-7 paralogs expressed in liver are indicated in black and their pairing to the TuD RNA is shown. The seed sequence, which determines miRNA target specificity, is underlined. The TuD targeting let-7 decreased the abundance of full-length let-7 and increased the abundance of prefix-matching let-7 sequence reads relative to the control. Isoforms are shown in black whose genome-matching reads decreased and whose prefix-matching reads increased. MM: total number of mismatches between each let-7 paralog and the TuD.

Analysis of TuD-directed inhibition of miR-122 in HuH-7 cells. (a) Anti-miR-122 TuDs used. (b) Dose-response curves for inhibition of miR-122 function in HuH-7 cells. Cells were transfected with constructs expressing the TuDs, together with a psiCHECK-2 reporter plasmid bearing three sites complementary to miR-122. Values are mean ± s.d. (n = 3). (c) Northern hybridization analysis of the experiment in (b).

Expression in TuD-treated mice of previously validated regulatory targets of miR-122 and let-7. scAAV9 vectors (control, TuD targeting miR-122 or let-7) were injected by tail vein into C57BL/6J mice (1 × 10¹² genome copies each). The mice were sacrificed four weeks later and the expression of representative endogenous targets of miR-122 (Aldolase A, Cyclin G1, Tmed3, Hfe2, and Cat-1 mRNA) and let-7 (Kras, Hras, Nras, and Dicer mRNA) analyzed in liver (left) or heart (right) by qRT-PCR (a) or in liver by western blotting (b). Values are mean ± s.d. comparing TuD-treated to control scAAV-treated mice; each group used four (a) or three (b) mice.

Change in cholesterol profiles of wild-type C57BL/6J mice after administration of scAAV expressing both the TuD targeting miR-122 and the GLuc reporter bearing miR-122-binding sites, relative to control GLuc reporter lacking the TuD and the miR-122 binding sites. (a) Four-to-six week–old male wild-type C57BL/6J mice were intravenously injected with 1 × 10¹² genome copies of scAAV9 per mouse. Serum levels of total cholesterol, high-density lipoprotein (HDL) and low-density lipoprotein (LDL) were measured at different times after injection. (b) The serum transaminases aspartate, aminotransferase (ASL) and alanine aminotransferase (ALT) were assayed to assess liver toxicity. Values are mean ± s.d.; n = 4, control and anti-miR-122 TuD groups and n = 3, anti-let-7 TuD group.