Immortalized human myotonic dystrophy muscle cell lines to assess therapeutic compounds

Abstract

Myotonic dystrophy type 1 (DM1) and type 2 (DM2) are autosomal dominant neuromuscular diseases caused by microsatellite expansions and belong to the family of RNA-dominant disorders. Availability of cellular models in which the DM mutation is expressed within its natural context is essential to facilitate efforts to identify new therapeutic compounds. Here, we generated immortalized DM1 and DM2 human muscle cell lines that display nuclear RNA aggregates of expanded repeats, a hallmark of myotonic dystrophy. Selected clones of DM1 and DM2 immortalized myoblasts behave as parental primary myoblasts with a reduced fusion capacity of immortalized DM1 myoblasts when compared with control and DM2 cells. Alternative splicing defects were observed in differentiated DM1 muscle cell lines, but not in DM2 lines. Splicing alterations did not result from differentiation delay because similar changes were found in immortalized DM1 transdifferentiated fibroblasts in which myogenic differentiation has been forced by overexpression of MYOD1. As a proof-of-concept, we show that antisense approaches alleviate disease-associated defects, and an RNA-seq analysis confirmed that the vast majority of mis-spliced events in immortalized DM1 muscle cells were affected by antisense treatment, with half of them significantly rescued in treated DM1 cells. Immortalized DM1 muscle cell lines displaying characteristic disease-associated molecular features such as nuclear RNA aggregates and splicing defects can be used as robust readouts for the screening of therapeutic compounds. Therefore, immortalized DM1 and DM2 muscle cell lines represent new models and tools to investigate molecular pathophysiological mechanisms and evaluate the in vitro effects of compounds on RNA toxicity associated with myotonic dystrophy mutations.

Keywords: Alternative splicing; Expanded repeats; Muscle cell line; Myotonic dystrophy; Nuclear aggregates; Therapeutic compounds.

Fig. 1.

Features of primary DM muscle cells. (A) Control (Ctrl), DM1 and DM2 primary myoblasts were differentiated for 4 days and then fixed for immunofluorescence analysis using desmin antibody (green) and nuclei staining (blue) with Hoechst. (B) The fusion index of differentiated primary muscle cell cultures determined by the percentage of the number of nuclei in myotubes (>2 myonuclei) to the total number of nuclei in desmin-positive cells (n=4 for Ctrl, n=3 for DM1 and DM2, >1000 nuclei counted per experiment). (C) RT-PCR analysis of alternative splicing profile of BIN1 exon 11, DMD exon 78 and LDB3 exon 11 for differentiated cultures of Ctrl, DM1 and DM2 primary muscle cells (n=3 in duplicate for each condition). Data are expressed as mean±s.e.m. Comparison with one-way ANOVA with Newman-Keuls post-test; **P<0.01.

Fig. 2.

Characterization of DM myoblast cell lines. (A) Representative cultures of Ctrl (cl4), DM1 (cl5) and DM2 (cl26) immortalized myoblasts differentiated for 4 days and then fixed for immunofluorescence analysis using desmin antibody (green) and Hoechst nuclear staining (blue). (B) The fusion index of differentiated immortal myoblast cultures was determined by the percentage of the number of nuclei in myotubes (>2 myonuclei) to the total number of nuclei in desmin-positive cells (n=3 for each condition, >1000 nuclei counted per experiment). (C) Nuclear RNA aggregates of expanded-CUG or -CCUG repeats in immortalized DM1 and DM2 myoblasts were detected by FISH using a Cy3-(CAG)7 or Cy3-(CCAG)5 probe (red), respectively. (D) Quantification of alternative splicing profile of BIN1 exon 11, DMD exon 78 and LDB3 exon 11 by RT-PCR analysis in differentiated cultures of Ctrl, DM1 and DM2 immortal myoblast cell lines (n=3 for each condition). (E) Southern blot analysis using a DMPK probe to determine the size of the CTG expansion in primary DM1 muscle cells (Pr) and two clones of immortalized DM1 myoblasts (Cl5 and Cl6) derived from the same primary muscle cell culture. Data are expressed as mean±s.e.m. Comparison with one-way ANOVA with Newman-Keuls post-test; **P<0.01, ***P<0.001.

Fig. 3.

Characterization of DM1 and Ctrl myotube cultures derived from conditional MyoD-converted fibroblast cell lines. (A) Desmin (Desm) immunofluorescence staining (green) of myotube cultures derived from conditional MYOD1 converted Ctrl (left) and DM1 (right) fibroblast cell lines under permissive conditions to express MYOD1. (B) FISH using a Cy3-CAG7 probe (red) and MBNL1 immunostaining (green) showing the colocalization of MBNL1 with nuclear CUGexp-RNA aggregates in muscle converted DM1 immortalized fibroblasts. Hoe, Hoechst staining. (C,D) RT-PCR analysis and quantification of alternative splicing changes in ATP2A1, BIN1, IR, LDB3, MBNL1 and TNNT2 transcripts extracted from myotube cultures derived from converted DM1 and Ctrl fibroblast cell lines (n=4 for each condition). Data are expressed as mean±s.e.m. Comparison with Mann-Whitney test; **P<0.01.

Fig. 4.

(CAG)7-ASO reverses CUGexp-RNA toxicity in immortalized muscle DM1 cells. (A) Percentage of immortalized DM1 myoblasts containing nuclear CUGexp-RNA aggregates (foci) following 2′OMe-(CAG)7 treatment (n=4). Data are expressed as mean±s.e.m. Comparison with Mann-Whitney test; ***P<0.001. (B) FISH immunofluorescence staining using a Cy3-(CAG)7 probe and an anti-MBNL1 antibody to detect nuclear foci (red) and MBNL1 (green), respectively, in immortalized DM1-cl5 myoblasts treated with 2′OMe-(CAG)7 ASOs compared with non-treated (NT) cells. (C) FISH experiment to detect CUGexp-RNA foci (red) in the nucleus (blue) of myotubes derived from converted DM1 fibroblast cell lines treated with 2′OMe-(CAG)7 ASOs. (D) Quantification of MBNL1 exon 7, LDB3 exon 11 and BIN1 exon 11 splicing profiles determined by RT-PCR in myotubes cultures of converted Ctrl and DM1 fibroblast cell lines treated or not with 2′OMe-(CAG)7 ASOs (n=8, except for BIN1 where n=6). Data are expressed as mean±s.e.m. Comparison with one-way ANOVA with Newman-Keuls post-test; **P<0.01, ***P<0.001; ns, not significant.

Fig. 5.

hU7-(CAG)15 reverses CUGexp-RNA toxicity in DM1 myoblast cell lines. (A) Differentiation of DM1-cl5 myoblast cell line expressing hU7-(CAG)15 antisense transcripts was assessed by desmin (green) and Hoechst (blue) immunostaining and compared with differentiated DM1-cl5 and Ctrl cells. (B) Fusion index of differentiated cultures of immortalized DM1 myoblasts transduced or not with hU7-(CAG)15 lentiviral vectors compared with Ctrl-cl4 (n=4 for each condition, >1000 nuclei counted per experiment). Data are expressed as mean±s.e.m. Comparison with Mann-Whitney test; *P<0.05. (C) Nuclear CUGexp-RNA aggregates in differentiated DM1-cl5 myoblast cell line expressing hU7-(CAG)15. (D) RT-PCR analysis of BIN1 exon 11, DMD exon 78 and LDB3 exon 11 splicing profiles in differentiated cultures of immortalized Ctrl-Cl4 and DM1-cl5 myoblasts transduced or not with hU7-(CAG)15 lentiviral vectors (n=6 for each condition). Data are expressed as mean±s.e.m. Comparison with one-way ANOVA with Newman-Keuls post-test; ***P<0.001.

Fig. 6.

Effect of hU7-(CAG)15 antisense oligonucleotide on global DM1-splicing changes. Total RNA extracted from Ctrl-cl4, DM1-cl5 and hU7-(CAG)15-treated DM1-cl5 differentiated myoblasts analyzed with a paired-end RNA sequencing (RNA-seq). (A) Repartition of 349 mis-splicing events in DM1-cl5 compared with Ctrl-cl4 cells (number of events; percentage compared with total number of events). (B) Box plot representing splicing inclusion or EUC (exon usage coefficient, determined with DEX-seq software) of the 349 splicing events (exon_bin) regrouped in excluded exons (n=245, left panel) and included exons (n=104, right panel) in Ctrl-cl4, DM1-cl5 and DM1-cl5 treated with hU7-(CAG)15. In box plot, lower whisker represents smallest observation greater than or equal to lower hinge-1.5x interquartile range (IQR); lower hinge is 25% quantile, middle line is median; upper hinge is 75% quantile; upper whisker is largest observation less than or equal to upper hinge+1.5xIQR. (C) Analysis of individual exon_bin exclusion (left panel) or inclusion (right panel) event in DM1-cl5 cells compared with Ctrl-cl4 cells. More than 50% of events are fully normalized (red), 28% are partially restored (blue) and around 20% are not significantly modulated (green) by treatment with hU7-(CAG)15.