RNAscope in situ hybridization-based method for detecting DUX4 RNA expression in vitro

Abstract

Facioscapulohumeral muscular dystrophy (FSHD) is among the most common forms of muscular dystrophy. FSHD is caused by aberrant expression of the toxic DUX4 gene in muscle. Detecting endogenous DUX4 in patient tissue using conventional methods can be challenging, due to the low level of DUX4 expression. Therefore, developing simple and trustworthy DUX4 detection methods is an important need in the FSHD field. Here, we describe such a method, which uses the RNAscope assay, an RNA in situ hybridization (ISH) technology. We show that a custom-designed RNAscope assay can detect overexpressed DUX4 mRNA in transfected HEK293 cells and endogenous DUX4 mRNA in FSHD patient-derived myotubes. The RNAscope assay was highly sensitive for tracking reductions in DUX4 mRNA following treatment with our therapeutic mi405 microRNA, suggesting that RNAscope-based DUX4 expression assays could be developed as a prospective outcome measure in therapy trials. This study could set the stage for optimizing and developing a new, rapid RNA ISH-based molecular diagnostic assay for future clinical use in the FSHD field.

Keywords: DUX4; FSHD; RNA; RNA in situ hybridization; RNAscope.

© 2019 Amini Chermahini et al.; Published by Cold Spring Harbor Laboratory Press for the RNA Society.

Figures

FIGURE 1.

RNAscope specifically detected overexpressed DUX4 mRNA in transfected HEK293 cells. (A) DUX4-transfected cells showed punctate brown staining with RNAscope assay, while controls did not. Decreased brown staining with increased doses of DUX4-targeted microRNAs (mi405) demonstrated specificity. (B–D) DUX4 signal after cotransfection of HEK293 cells with CMV.DUX4 and U6.mi405 plasmids at 1:2, 1:4, and 1:8 ratios. (E) Absence of brown DUX4 signal in untransfected HEK293 cell line. (F) RNAscope negative control stain. (G) Housekeeping gene PPIB was detected in all HEK293 cells and served as a positive control for the assay. (H) Cell viability assay demonstrated that mi405 reduced the levels of overexpressed DUX4 mRNA, and significantly protected cells from death at 1:4 and 1:8 weight ratios (N = 3 independent experiments performed in triplicate); P < 0.0001, ANOVA. 40× objective. Scale bar, 50 microns.

FIGURE 2.

RNAscope detects endogenous DUX4 mRNA in 15A FSHD myotubes. FSHD 15A myotubes demonstrated higher amounts of DUX4 mRNA compared to non-FSHD 15V myotubes, as determined by RNAscope staining. Arrows indicate brown punctate signal. (A) Negative control stain. (B) 15V cell line stained with the housekeeping gene PPIB served as a positive control for the assay. (C) DUX4 expression in FSHD 15A myotubes was reduced or absent in (D) 15A cells transfected with U6.mi405 microRNAs. (E) Very weak or absent signal was present in the unaffected 15V cell line alone and in (F) 15V transfected with U6.mi405 plasmid.

FIGURE 3.

Quantification of DUX4 RNAscope signal and DUX4-activated biomarkers. (A) Schematic representation of a cover slip on which FSHD and control myoblasts were grown and differentiated to myotubes. For each sample, a total of 20 image fields were taken from five zones (four quadrants and one central zone). DUX4 RNAscope signal was quantified as demonstrated in the video in Supplemental Figure 2. (B) Percent area of DUX4 signal in one representative experiment (from three independent experiments). Each data point represents total % area per zone. In this experiment, DUX4 signal was significantly elevated in 15A myotubes compared to unaffected 15V controls and to 15A cells transfected with the mi405 therapeutic microRNA. (**) Represents significant differences from untreated 15A FSHD cells (P < 0.01; ANOVA). (C) Data presented here indicate that despite low abundance relative to the entire 15A culture, the ∼1% of cells showing DUX4 signal often expressed high amounts of DUX4 mRNA. Here, each data point represents the number of DUX4 positive foci per field in a representative culture. One DUX4 focus was defined as 16 pixels, which was the minimum visible DUX4 signal. Again, DUX4 signal was absent or very low in unaffected 15V cells, as well as affected 15A cells transfected with mi405 plasmid. (**) Represents significant differences from untreated 15A myotubes (P < 0.01; ANOVA). (D) QPCR assays of DUX4-activated biomarkers, PRAMEF12 and MBD3L2, in 15A and 15V cells. Biomarker expression was used to confirm the specificity of the RNAscope assay and knockdown of endogenous DUX4 by miDUX4.405. Both biomarkers were significantly elevated in affected 15A myotubes, but absent or virtually absent in 15V cells (no signal at cycle 40). PRAMEF12 and MBD3L2 were significantly reduced in 15A cells transfected with miDUX4.405 plasmid. Data were acquired from N = 3 independent experiments, with each QPCR assay performed in triplicate. (**) Represents significant differences from untreated 15A myotubes (P < 0.01; ANOVA).