Heart-specific overexpression of CUGBP1 reproduces functional and molecular abnormalities of myotonic dystrophy type 1

Abstract

Myotonic dystrophy type 1 (DM1) is caused by a CTG expansion within the 3'-untranslated region of the DMPK gene. The predominant mechanism of pathogenesis is a toxic gain of function of CUG repeat containing RNA transcribed from the expanded allele. The molecular mechanisms by which the RNA containing expanded repeats produce pathogenic effects include: sequestration of muscleblind-like 1 (MBNL1) protein and up-regulation of CUG binding protein 1 (CUGBP1). MBNL1 and CUGBP1 are RNA binding proteins that regulate alternative splicing transitions during development. Altered functions of these proteins in DM1 lead to misregulated splicing of their target genes, resulting in several features of the disease. The role of MBNL1 depletion in DM1 is well established through a mouse knock-out model that reproduces many disease features. Here we directly test the hypothesis that CUGBP1 up-regulation also contributes to manifestations of DM1. Using tetracycline-inducible CUGBP1 and heart-specific reverse tetracycline trans-activator transgenes, we expressed human CUGBP1 in adult mouse heart. Our results demonstrate that up-regulation of CUGBP1 is sufficient to reproduce molecular, histopathological and functional changes observed in a previously described DM1 mouse model that expresses expanded CUG RNA repeats as well as in individuals with DM1. These results strongly support a role for CUGBP1 up-regulation in DM1 pathogenesis.

Figure 1.

Generation of MHC-rtTA/TRECUGBP1 bitransgenic mice that inducibly overexpress human CUGBP1 in heart. (A) The TRECUGBP1 construct, shown to scale, expresses the human LYLQ isoform of CUGBP1 containing an N-terminal Flag tag driven by a TRE and a CMV minimal promoter. (B) Anti-Flag western blot using heart protein extracts showing a time course following transgene induction. MHC-rtTA/TRECUGBP1 bitransgenic animals from two TRECUGBP1 founder lines (3413 and 3481) sacrificed at 2, 4 and 7 days after 2 g doxycycline/kg food was initiated. (C) Quantification of CUGBP1 expression in MHC-rtTA/TRECUGBP1 animals from the 3413 line. Western blot analysis using anti-CUGBP1 antibodies of 12.5, 25 and 50 µg of protein extracts from hearts of induced (8 days) and uninduced MHC-rtTA/TRECUGBP1 bitransgenic animals, and MHC-rtTA animals were given doxycycline. Bitransgenic animals receiving doxycycline express total CUGBP1 levels four fold or higher compared with control mice.

Figure 2.

Induced bitransgenic MHC-rtTA/TRECUGBP1 animals exhibit enlarged hearts and histopathological features of DM1. (A) Hearts from MHC-rtTA/TRECUGBP1 and MHC-rtTA mice that did or did not receive 2 g/kg doxycycline food as indicated. (B) Percent heart weight to body weight was significantly increased in bitransgenic mice given doxycycline (+) compared with controls (−) (**P < 0.001, also see Table 1). Animals were induced 11 days ± 1.0. (C) Histological analysis revealed widespread degeneration, necrosis and loss of myocardial fibers in animals induced to overexpress CUGBP1 (MHC/TRECUGBP + dox). A moderate mixed inflammatory infiltrate was associated with these areas. No significant lesions were observed in hearts of bitransgenic mice not receiving doxycycline (MHC/TRECUGBP − dox). Minor degenerative changes were seen in MHC-rtTA mice receiving doxycycline; however, cardiac muscle fiber striations were still visible in these degenerative areas. Hematoxylin and eosin as well as Trichrome staining (MCC-rtTA + dox H&E 40× and tri 40×) revealed degenerating muscle fibers in a zone near the middle of the ventricular wall (note scattered individual more darkly stained fibers). The scale bars indicate indicate 1 mm, 500 µm and 50 µm for 1.25×, 4× and 40× images, respectively.

Figure 3.

Alternative splicing events responsive to CUGBP1 but not MBNL1 are altered in DM1 heart tissue. (A) Diagram of RT–PCR analysis of alternative splicing. Primers located in exons that flank the alternative exon detect mRNAs that include or skip the variable region. (B) RT–PCR results from heart tissues from three mice in which heart-specific expression of DMPK-CUG960 RNA is induced by Cre-mediated recombination (EpA960/MCM + tam), compared with uninduced bitransgenic animals (EpA960/MCM − tam) and MCM animals expressing only Cre given tamoxifen (MCM + tam). Splicing events responsive to CUGBP1 but not MBNL1 are c10orf97, Mfn2, Capzb and Ppfibp1. cTNT and Ablim1 are responsive to MBNL1. Percent of mRNAs including the variable region(s) are calculated. RT–PCR gels are shown in Supplementary Material, Figure S1. (C) Splicing events responsive to CUGBP1 but not MBNL1 are altered in DM1 heart tissue but not non-DM1-dilated cardiomyopathies. For (B) and (C), each histogram represents the percent of mRNA containing the variable region from individual heart samples and the lane order is the same for all six splicing events tested. Analysis of splicing in fetal and adult heart demonstrates transitions that are conserved between mouse and human heart development (16). RT–PCR gels are shown in Supplementary Material, Figure S2.

Figure 4.

M-mode echocardiography analysis indicates that mice overexpressing CUGBP1 exhibit severe functional abnormalities. (A) Representative M-mode echocardiograms before doxycycline administration and at day 8 after doxycycline initiation for bitransgenic MHC-rtTA/TRECUGBP1 and control MHC-rtTA mice. Tracings are from the same two animals before and after dox administration. (B–D) Bitransgenic mice fed doxycycline-containing food showed significantly decreased ejection fractions and significantly increased LVIDs and LVIDd, respectively, when compared with control mice (**P < 0.001 for all results shown). Uninduced bitransgenic animals and MHC-rtTA mice fed 2 g/kg doxycycline in food were grouped for statistical analyses.

Figure 5.

CUGBP1 overexpression in heart produces electrocardiography abnormalities. (A) Representative electrocardiograms are from the same animals before and after doxycycline administration for bitransgenic MHC-rtTA/TRECUGBP1 (+dox) and control MHC-rtTA mice (+dox). (B) Maximal change in PR interval and QRS complex duration from doxycycline day zero is significantly increased in induced bitransgenic mice compared with controls (*P < 0.05).