Regulation of DMD pathology by an ankyrin-encoded miRNA

Abstract

Background: Duchenne muscular dystrophy (DMD) is an X-linked myopathy resulting from the production of a nonfunctional dystrophin protein. MicroRNA (miRNA) are small 21- to 24-nucleotide RNA that can regulate both individual genes and entire cell signaling pathways. Previously, we identified several mRNA, both muscle-enriched and inflammation-induced, that are dysregulated in the skeletal muscles of DMD patients. One particularly muscle-enriched miRNA, miR-486, is significantly downregulated in dystrophin-deficient mouse and human skeletal muscles. miR-486 is embedded within the ANKYRIN1(ANK1) gene locus, which is transcribed as either a long (erythroid-enriched) or a short (heart muscle- and skeletal muscle-enriched) isoform, depending on the cell and tissue types.

Results: Inhibition of miR-486 in normal muscle myoblasts results in inhibited migration and failure to repair a wound in primary myoblast cell cultures. Conversely, overexpression of miR-486 in primary myoblast cell cultures results in increased proliferation with no changes in cellular apoptosis. Using bioinformatics and miRNA reporter assays, we have identified platelet-derived growth factor receptor β, along with several other downstream targets of the phosphatase and tensin homolog deleted on chromosome 10/AKT (PTEN/AKT) pathway, as being modulated by miR-486. The generation of muscle-specific transgenic mice that overexpress miR-486 revealed that miR-486 alters the cell cycle kinetics of regenerated myofibers in vivo, as these mice had impaired muscle regeneration.

Conclusions: These studies demonstrate a link for miR-486 as a regulator of the PTEN/AKT pathway in dystrophin-deficient muscle and an important factor in the regulation of DMD muscle pathology.

Figure 1

miR-486 is a highly conserved mammalian miRNA that is differentially expressed in DMD skeletal muscle. (A) miR-486 is conserved across mammals and is under the transcriptional regulation of the ankyrin 1 locus. The ANK1.5 isoform has a unique start exon (exon 39A; blue) that is transcribed independently of the ANK1.1 isoform and lacks the ankyrin repeat domain. (B) miR-486 is significantly reduced in DMD human skeletal muscle biopsies, but not in the milder BMD muscle biopsies, as compared to normal muscle controls. P = 0.03. The y-axis represents mature miR-486 fold levels normalized to U6 snRNA housekeeping controls. Error bars indicate SEM. Normal biopsies: n = 5, DMD biopsies: n = 5 and BMD biopsies: n = 3. (C) miR-486 is significantly upregulated during normal human myogenic differentiation, while the DMD myoblasts express miR-486 at significantly lower levels. The y-axis represents mature miR-486 fold levels normalized to U6 snRNA loading controls, and the x-axis represents the time of myogenic differentiation from 50% confluency (proliferating myoblasts) until day 4 differentiation (multinucleated myotube formation). At 90% confluency (day 0 differentiation), myoblasts were exposed to differentiation medium (2% horse serum). Blue lines represent normal human myoblasts, and red lines represent DMD myoblasts. (D) miR-486 is upregulated during skeletal muscle regeneration following CTX-induced TA muscle injury. Blue lines indicate wild-type C57B6/J mice, and red bars indicate mdx^5cv mice. The y-axis represents mature miR-486 fold levels normalized to U6 snRNA housekeeping controls. Error bars indicate SEM. *P < 0.005 and **P < 0.05.

Figure 2

Overexpression and inhibition of miR-486 causes profound cellular changes in myoblasts and myotubes. (A) Phase contrast and GFP images of normal human myoblasts overexpressing miR-486, anti-miR-486 and a scrambled miRNA negative control virus are shown. Note that the myoblasts overexpressing the miR-486 inhibitor (anti-miR-486) appear rounded and have not flattened out (inset). (B) Day 4 differentiated normal human myotubes infected with the same lentiviral vectors are shown. Note the aggregate of clumped myotubes (inset) in cells infected with the anti-miR-486 lentivirus. (C) Higher magnification (×40) of normal human myoblasts infected with lentivirus expressing scrambled miRNA, miR-486 and anti-miR-486. The cells show detection of the GFP reporter and 4',6-diamidino-2-phenylindole (DAPI) to detect the DNA. Note that the anti-miR-486-infected myoblasts are rounded (arrowheads) and increased in overall size compared to the scrambled miRNA and miR-486-overexpressing myoblasts. (D) MF20 (red) staining of normal human myoblasts overexpressing scrambled miRNA, miR-486 or anti-miR-486 lentivirus. Chart indicates reduced levels of MF20⁺ myoblasts that express anti-miR-486 lentivirus; conversely, myoblasts with high levels of miR-486 have increased amounts of MF20⁺ myoblasts. Ten fields of GFP-positive (not shown) and DAPI-positive cells were counted for the presence of MF20⁺ staining. Values were normalized to the scrambled miRNA. *P < 0.005 and **P < 0.05. Scale bar indicates 35 μm. (E) FACS histogram reveals an increase in cell size in myoblasts infected with anti-miR-486 lentivirus. The x-axis represents forward scatter height (FSC-H), and the y-axis represents the percentage cell size relative to the control scrambled miRNA myoblasts. FSC-H is quantified in the table that shows that anti-miR-486 (blue line) have increased in cell size compared to myoblasts overexpressing miR-486 (green line) and scrambled miRNA controls (red line). *P = 0.028 ± SEM.

Figure 3

miR-486 is essential for normal myoblast fusion, cell cycle kinetics and viability in human skeletal myoblasts. (A) and (B) Myoblast fusion assay reveals significantly lower fusion (shown as Fusion Index % of cells with 2 or more nuclei) in miR-486-inhibited (anti-miR-486) cells. Normal and DMD myoblasts were infected at 50% confluency with lentivirus expressing miR-486, miR-486 inhibitor (anti-miR) or a scrambled miRNA control virus and allowed to differentiate to 90% confluency. Note the decreased levels of fusion in both the normal and DMD myoblasts at two time points (day 2 (Figure 2A) and day 4 (Figure 2B) of differentiation) when infected with anti-miR-486. (C) Increased cellular apoptosis measured using a caspase 3/7 enzymatic assay in normal and DMD myoblasts overexpressing (miR-486) or inhibiting (anti-miR-486) miR-486. DMD myoblasts overall had significantly higher caspase 3/7 activity compared with normal human myoblasts across all samples. The y-axis indicates average RLUs (that is, caspase 3/7 activity), and the x-axis indicates the condition of the myoblasts. Mock and scrambled miRNA-GFP-infected myoblasts served as negative controls (n = 3 replicates/cohort). Cells treated with hydrogen peroxide (H₂O₂) for 12 hours served as positive controls. (D) Ki-67 levels measured by immunofluorescence in normal and DMD myoblasts which have increased miR-486 expression, reduced miR-486 expression or expression of scrambled miRNA (negative control). Note the decreased levels of Ki-67 in myoblasts in which miR-486 expression is knocked down. *P < 0.005 and **P < 0.05 for comparisons between scrambled miRNA (control) versus miR-486- and/or anti-miR-486-infected myoblasts.

Figure 4

Members of the PTEN/AKT signaling pathway and splicing factors are direct downstream targets of miR-486 in skeletal muscle. (A) Schematic shows the location of the 3'UTR miR-486 target cloned downstream of the luciferase gene reporter with a poly(A) tail. (B) Luciferase miRNA reporter assay of several of the 3'UTR of the predicted miR-486 direct downstream targets. The lower graph shows results from the mutation of the miR-486 seed site, which ablates miR-486 binding and functions to derepress luciferase expression. The y-axis displays relative RLU activity normalized to 100% activation of the reporter transfected alone (vector alone). Error bars indicate SEM (n = 3 replicates, with the experiment performed on three separate occasions using HEK293T cells). *P < 0.005 and **P < 0.05. (C) Western blots of myoblasts overexpressing miR-486 or scrambled miRNA controls. Anti-β-tubulin served as a loading control. Western blot analysis of alternately splicing factors SFRS1 and SFRS3 was performed from myotube lysates from which both proteins are most abundantly expressed.

Figure 5

Transgenic miR-486 overexpression in mice results in abnormal skeletal muscle regeneration following CTX-induced TA injury. (A) Diagram indicating the construct used to generate MCK-miR-486-transgenic mice. Real-time qPCR of miR-486 expression levels in the TA muscle was analyzed from three separate wild-type (black bars) and miR-486 Tg (red bars) adult mice. Expression levels were normalized to U6 snRNA levels and compared to wild-type (control) levels. (B) Mouse weights of wild-type and miR-486 Tg mice versus age (in months) are shown. Three mouse cohorts of male and female littermates were weighed from 1, 3, 6, 9, 12 and 15 months of age (n = 5 male or female mice per cohort). *P < 0.005 and **P < 0.05. (C) H & E-stained histological sections of wild-type and miR-486 Tg mice on day 0 (uninjured), 7 and 14 post-CTX-induced TA injury. Arrowheads indicate centralized myonuclei. Scale bars = 20 μm. (D) Percentages of total centralized myonuclei during CTX-induced TA injury in wild-type and miR-486 Tg mice. Percentages are based on counts of 200 myofibers in three separate mice per cohort at three time points (0, 7 and 14 days post-CTX injury). *P < 0.005 and **P < 0.05.

Figure 6

miR-486 overexpression regulates skeletal muscle PTEN/AKT signaling in vivo. Real-time qPCR measuring levels of Pdgfrβ, Pten, Foxo1, Pik3r1 (p85α), p21 (Cdkn1a) and p27 (Cdkn1b) in mouse skeletal muscle during the time course of skeletal muscle regeneration (day 0 (uninjured) to day 14 post-CTX TA injection) in wild-type (black bars) and miR-486 Tg (green bars) mice. Three muscles per time point per day of injury were analyzed in triplicate. Error bars indicate SEM. Relative mRNA expression levels were normalized to 18 s ribosome (18sRib) loading control and day 0 (uninjured) wild-type muscle. *P < 0.05 wild-type vs. miR-486 Tg values.

Figure 7

miR-486 is an important regulator of PTEN/AKT signaling in normal and dystrophic muscle. Schematic of miR-486 regulation of PTEN/AKT pathway upstream and downstream signaling components in skeletal muscle is shown. Regulation of the PTEN/AKT pathway in turn regulates FOXO1, which is a well-characterized regulator of p21 and p27 in skeletal muscle.