(A) Northern blots documenting miRNA expression levels in normal liver and tumors from tet-o-MYC; LAP-tTA mice and in uninfected or retrovirally-infected HepG2 cells. tRNA^Lys served as a loading control.
(B) Cell-cycle profiles of retrovirally-infected HepG2 cells as determined by propidium-iodide (PI) staining and flow cytometry. Numbers over each histogram indicate the percentage of cells in G₁, S, and G₂ cell-cycle phases.
(C) Cell-cycle profiles of retrovirally-infected HepG2 cells following treatment with nocodazole (Noc). Numbers over each histogram indicate the percentage of cells remaining in G₁.

(A) Schematic representation of scAAV vectors used in this study depicting locations of inverted terminal repeats (ITRs), elongation factor 1 α promoter (EF1α), miRNA (shown in hairpin form), and enhanced green fluorescent protein (eGFP) open reading frame.
(B) Northern blot of transiently-transfected HeLa cells demonstrating equivalent levels of miR-26a when expressed from an intronic (scAAV.miR26a.eGFP) or exonic (scAAV.miR26a) context. Co-transfection with a miR-122a expression plasmid (pcDNA-miR-122a) provided a control for transfection efficiency while tRNA^Lys levels documented equal loading.
(C) Fluorescent microscopy showing eGFP expression in HeLa cells transiently-transfected with the indicated AAV vectors.
(D) Northern blots showing expression of miRNAs in livers 21 days following administration of the indicated AAV vectors.
(E) Fluorescent microscopy showing efficient transduction of hepatocytes, as indicated by eGFP expression, 21 days following AAV administration.

(A) Expression of miR-26a in tumors in AAV-transduced animals. qPCR was used to measure miRNA abundance in tumors 5 or 10 days following administration of scAAV8.eGFP (eGFP) or scAAV8.miR26a.eGFP (miR26a.eGFP). All values were normalized to 18S rRNA expression. Normal liver expression and tumor expression lines were derived from qPCR analysis of miR-26a levels in samples shown in Figure 1A. Each box represents the range of expression observed. The ends of the boxes represent the 25th and 75th percentiles, the bars indicate the 10th and 90th percentiles, and the median is depicted by a horizontal line within the boxes.
(B) Fluorescence microscopy of tumor sections demonstrating GFP expression in tumor cells in AAV-transduced animals.

(A) Northern blot analysis of miRNA expression in normal liver (N) or tumor tissue (T) from mice of the indicated genotypes. Graphs depict relative quantification of miRNA levels normalized to tRNA^Lys abundance.
(B) qPCR analysis of miR-26a expression in human HCC and normal liver biopsies. miRNA abundance was normalized to 18S rRNA expression. p value calculated by two-tailed t test.
(C) miR-26a expression in individual HCC tumors relative to expression in paired normal liver samples.

(A–B) Western blots documenting abundance of cyclins D2 and E2 in retrovirally-infected HepG2 cells. Relative quantification of band intensities, normalized to tubulin levels, are shown below blots.
(C–D) Sequence and evolutionary conservation of the miR-26a binding sites in the 3' UTRs of transcripts encoding cyclin D2 (CCND2) and cyclin E2 (CCNE2). Mutations introduced into luciferase reporter constructs are shown in red.
(E–F) Relative firefly luciferase activity derived from CCND2 (E) and CCNE2 (F) 3' UTR reporter constructs following transfection into HepG2 cells alone or in combination with miR-18a or miR-26a synthetic miRNA mimics. All values were normalized to renilla luciferase activity produced from a co-transfected control plasmid. Error bars represent standard deviations from 3 independent transfections. *, p<0.05; **, p<0.01 (two-tailed t test).

(A) Time-line of miR-26a therapeutic delivery experiment.
(B) Gross tumor burden of livers from miR-26a-treated and control animals, as determined by quantification of tumor area using the ImageJ software package. The mean tumor burden in each treatment group is indicated by horizontal lines. Data points highlighted by asterisks represent animals that exhibited low AAV transduction efficiency (see Fig. S5). p value calculated by two-tailed t-test.
(C) Representative images of livers from miR-26a-treated and control animals.
(D) Liver:body weight ratios of miR-26a-treated and control animals. A chi-square statistic was used to compare the fraction of animals in each treatment group with a liver:body weight ratio above 0.1 (indicated by horizontal line). Data points highlighted by asterisks represent animals that exhibited low AAV transduction efficiency (see Fig. S5).

(A) Representative DAPI and Ki67-stained sections from miR-26a-treated and control tet-o-MYC; LAP-tTA animals 5 days after AAV administration showing tumors (outlined with dotted line) and adjacent normal-appearing liver.
(B) Quantification of Ki67 staining in tumors from miR-26a-treated and control animals. Olympus Slidebook 4.2 was used to quantify the Ki67 fluorescence intensity in tumors in 3–5 randomly chosen fields per animal (n=2–4 animals per treatment per timepoint). The mean Ki67 fluorescence intensity per condition is plotted with error bars representing standard deviations.
(C) Representative DAPI and TUNEL-stained sections from miR-26a-treated and control animals 5 days after AAV administration. Tumors are outlined with dotted lines.
(D) Quantification of TUNEL staining in tumors from miR-26a-treated and control animals. ImageJ was used to quantify the TUNEL positive area in tumors in 6 randomly chosen fields per animal (n=2–4 animals per treatment per time-point). The mean TUNEL positive area per condition is plotted with error bars representing standard deviations.