HTS-Compatible Patient-Derived Cell-Based Assay to Identify Small Molecule Modulators of Aberrant Splicing in Myotonic Dystrophy Type 1

Abstract

Myotonic dystrophy type 1 (DM1) is a genetic disorder characterized by muscle wasting, myotonia, cataracts, cardiac arrhythmia, hyperinsulinism and intellectual deficits, and is caused by expansion of a CTG repeat in the 3'UTR of the Dystrophia Myotonica-Protein Kinase (DMPK) gene. The DMPK transcripts containing expanded CUG repeats accumulate in nuclear foci and ultimately cause mis-splicing of secondary genes through the dysregulation of RNA-binding proteins including Muscleblind 1 (MBNL1) and CUG binding protein 1 (CUGBP1). Correction of mis-splicing of genes such as the Skeletal muscle-specific chloride channel 1 (CLCN1), Cardiac troponin T (TNNT2), Insulin receptor (INSR) and Sarcoplasmic/endoplasmic reticulum Ca(2+)ATPase 1 (SERCA1) may alleviate some of the symptoms of DM1; hence identification of small molecule modulators is an important step towards a therapy for DM1 patients. Here we describe the generation of immortalized myoblast cell lines derived from healthy (DMPK CTG(5)) and DM1 patient (DMPK CTG(1000)) fibroblasts by constitutive overexpression of human telomerase reverse transcriptase (hTERT) and inducible overexpression of the Myoblast determination factor (MYOD). MBNL1-containing nuclear foci, mis-splicing events and defective myotube differentiation defects characteristic of DM1 were observed in these cells. A CLCN1 luciferase minigene construct (CLCN1-luc) was stably introduced to monitor intron 2 retention in the DM1 cellular context (a reported splicing defect in DM1). The assay was validated by performing a high-throughput screen (HTS) of ~13,000 low molecular weight compounds against the CLCN1-luc DM1 myoblast cell line, providing an ideal system for conducting HTS to better understand and treat DM1.

Keywords: Alternative splicing; DM1; HTS; cell-based assay.; myotonic dystrophy.

Fig. (1).

Generation of immortalized DM1 fibroblasts, myoblasts and myotubes. (A) RT-PCR confirmed hTERT expression in immortalized WT and DM1 fibroblasts (hTERT FB), fibroblasts containing the MYOD expression construct before (MYOD FB) and after (Myoblast) doxycycline induction, and following serum starvation-induced differentiation (Myotube), but not in primary fibroblasts (Primary FB). Equal levels of cDNA template were confirmed by amplification of the GAPDH housekeeping gene. (B) Western blot analysis confirmed inducible expression of MYOD upon differentiation of immortalized WT and DM1 cell lines into myotubes. β-actin protein expression was used to ensure equivalent amounts of total protein in each sample. (C) Immunostaining of immortalized WT and DM1 myoblasts and myotubes for markers of early (DESMIN) and late stage (MHC) differentiation (green), with malachite green counterstaining (red). Scale bars represent 10µm.

Fig. (2).

Splicing factor expression in immortalized DM1 cells. (A) Western blot analysis of CUGBP1 and MBNL1 splicing factor proteins in immortalized WT and DM1 fibroblasts, myoblasts and myotubes (with Lamin A/C and β-actin loading controls for nuclear and whole cell lysates, respectively). (B) Localization of DMPK mutant RNA to nuclear foci (highlighted by white arrowheads) via in situ hybridization with a Texas Red-conjugated RNA probe containing 7 CUG repeats ((CUG)7-TR), together with MBNL1 immunostaining (green) and Hoechst counterstaining (blue). Scale bars represent 10µm.

Fig. (3).

MBNL1/CUGBP1 target gene expression in immortalized DM1 cell lines. (A) RT-PCR analysis to detect splicing defects in MBNL1/CUGBP1 target genes using primers in flanking exons: TNNT2 exon 5 (E5) inclusion, INSR exon 11 (E11) exclusion and SERCA1 exon 22 (E22) exclusion in DM1 cells (in comparison to WT cells). Equal levels of cDNA template were confirmed by amplification of the GAPDH housekeeping gene. (B) Western blot analysis for CLCN1 skeletal muscle-specific chloride channel expression in immortalized WT and DM1 cell lines. β-actin protein expression was used as a loading control.

Fig. (4).

CLCN1-luc minigene reporter construct splicing and expression in immortalized DM1 cell lines. (A) Schematic of CLCN1 luciferase minigene reporter (CLCN1-luc). A genomic segment of human CLCN1 containing exon 2 to 3 and the intervening intron (solid line) was cloned downstream of the CMV promoter (CMVpr) and in-frame with the luciferase coding sequence in the pLenti6/V5-D-TOPO vector. Dotted lines represent correct splicing (as in WT cells). Intron 2 retention (as in DM1 cells) results in a nonsense mutation, as indicated by TGA. (B) RT-PCR analysis with primers spanning CLCN1 exon 2 to the luciferase coding region, demonstrating complete (+ intron 2) and partial (+ partial intron 2) intron 2 retention of the CLCN1-luc construct in immortalized DM1 cells compared to WT cells (- intron 2), which was partially rescued upon treatment of CLCN1-luc DM1 myoblasts with 10 µM AKT inhibitor Triciribine (C). GAPDH was amplified as a control for equivalent cDNA synthesis. (D) CLCN1-luc activity (RLU = relative luciferase units) 24 h after addition of 10 µM Triciribine to CLCN1-luc WT and DM1 myoblasts (in 384-well format). *** represents a p-value = 1.1 x 10^-5 and * represents a p-value = 2.4 x 10^-3 in two-tailed t-tests. Data points are an average of 3 replicates and error bars represent standard deviation.

Fig. (5).

CLCN1-luc assay miniaturization and HTS validation in immortalized DM1 myoblasts. (A) CLCN1-luc DM1 myoblasts were treated with 10 µM Triciribine for 24 h prior to luciferase activity readout (RLU = relative luciferase units) in 384-well and 1536-well formats. Data points are an average of 3 (384-well) or 6 (1536-well) replicates, and error bars represent standard deviation. (B) Activity trace of HTS of ~13,000 compounds (in triplicate) against CLCN1-luc in DM1 myoblasts in 1536-well format, showing plate median values of DMSO negative control wells (blue), Triciribine positive control wells (red), and test compounds (green), expressed as fold change relative to DMSO. An assay window of 4-fold was maintained throughout the screen (CV = 9% and Z´ factor = 0.5). (C) Histogram of normalized luciferase expression data (green) from HTS of ~13,000 compounds, indicating threshold of >1.3 fold change (relative to DMSO) used to determine a hit rate of 0.26%. Red indicates 10x activity.