Use of HSALR female mice as a model for the study of myotonic dystrophy type I

Abstract

HSA^LR mice are the most broadly used animal model for studying myotonic dystrophy type I (DM1). However, so far, HSA^LR preclinical studies have often excluded female mice or failed to document the biological sex of the animals. This leaves an unwanted knowledge gap concerning the differential development of DM1 in males and females, particularly considering that the disease has a different clinical presentation in men and women. Here we compared typical functional measurements, histological features, molecular phenotypes and biochemical plasma profiles in the muscles of male and female HSA^LR mice in search of any significant between-sex differences that could justify this exclusion of female mice in HSA^LR studies and, critically, in candidate therapy assays performed with this model. We found no fundamental differences between HSA^LR males and females during disease development. Both sexes presented comparable functional and tissue phenotypes, with similar molecular muscle profiles. The only sex differences and significant interactions observed were in plasma biochemical parameters, which are also intrinsically variable in patients with DM1. In addition, we tested the influence of age on these measurements. We therefore suggest including female HSA^LR mice in regular DM1 studies, and recommend documenting the sex of animals, especially in studies focusing on metabolic alterations. This will allow researchers to detect and report any potential differences between male and female HSA^LR mice, especially regarding the efficacy of experimental treatments that could be relevant to patients with DM1.

Fig. 1. Organismal-level phenotypes.

a,b, The size of male (a) and female (b) HSA^LR animals is increased compared with control WT. c, The human HSA transgene is expressed in the quadriceps and gastrocnemius of HSA^LR mice, but not in the muscles of WT mice. Slight but significant differences are detected in HSA transgene expression between female and male HSA^LR muscles (n = 10, n = 15, n = 10, n = 23, for WT males and females and HSA^LR males and females respectively). This observation explains the effect of sex and interaction genotype × sex observed with ANOVA (Supplementary Fig. 2). d, Strong myotonia is observed in HSA^LR animals (n = 27, n = 3, n = 42, n = 11). e,f, Weight measurements show that females are smaller than males (n = 37, n = 15, n = 42, n = 19) (e), and normalization to WT weight shows that weight increases consistently in both male and female HSA^LR animals (f). g, Forelimb strength, measured in seconds, is comparable between males and females and significantly reduced in HSA^LR animals (n = 14, n = 15, n = 17, n = 19). h, Forelimb strength normalized to weight shows significant differences between sexes and genotypes. The boxes indicate the two central quartiles, the midline represents median values and the whiskers indicate the minimum and maximum data value, excluding outliers. Statistical analysis was performed using two-way ANOVA (WT versus HSA^LR, males versus females). Differences between WT and HSA^LR are represented as black asterisks. Differences between sexes were confirmed using the Wilcoxon test and are represented as blue asterisks. ***P < 0.001; **P < 0.01; *P < 0.05.

Fig. 2. Muscle histology in male and female WT and HSA^LR mice.

a, Representative bright-field microscopy pictures (200× magnification; scale bar, 50 µm) of hematoxylin–eosin staining of quadriceps sections showing an increase in the amount of central nuclei (black arrows) in HSA^LR mice. b, The percentage of fibers with central nuclei is dramatically increased in HSA^LR muscles (quadriceps and gastrocnemius combined; WT males n = 21, WT females n = 19, HSA^LR males n = 37, HSA^LR females n = 24). c, Representative confocal images (gastrocnemius; 400× magnification) showing the detection of CUG repeats (green) and agglutinin (red) as a marker of cell membrane. d, The percentage of nuclei with foci in the muscles studied (quadriceps and gastrocnemius combined). The individual quantification of each slide is shown (seven slides per muscle per mouse; WT males n = 5, WT females n = 5, HSA^LR males n = 5, HSA^LR females n = 5). e, Representative confocal images (gastrocnemius; 400× magnification) where MBNL1 was stained (green). f, MBNL1 fluorescence was quantified (quadriceps and gastrocnemius combined). The mean quantification of the slides is shown (seven slides per muscle per mouse; WT males n = 5, WT females n = 5, HSA^LR males n = 5, HSA^LR females n = 5). The boxes indicate the two central quartiles, the midline represents median values and the whiskers indicate the minimum and maximum data value, excluding outliers. Statistical analysis was performed using two-way ANOVA (WT versus HSA^LR, males versus females). Differences between WT and HSA^LR are represented as black asterisks. Differences between sexes were confirmed using the Wilcoxon test and are represented as blue asterisks. ***P < 0.001; **P < 0.01; *P < 0.05.

Fig. 3. Muscleblind family gene expression analysis.

a, Mbnl1 gene expression, measured by qRT–PCR, is significantly increased in HSA^LR muscles compared with WT. b, By contrast, Mbnl2 expression is reduced in HSA^LR muscles. c, MBNL1 protein levels, as measured by quantitative dot blot, show a significant decrease in HSA^LR muscles compared with WT. d, miR-218 levels are increased in the muscles of HSA^LR animals compared with WT. The boxes indicate the two central quartiles, the midline represents median values and the whiskers indicate the minimum and maximum data value, excluding outliers. e,f, Representative western blot membranes depicting MBNL1 protein levels in quadriceps (e) and gastrocnemius (f). Statistical analysis was performed using two-way ANOVA (black asterisks) ***P < 0.001; *P < 0.05. Sample sizes for the whole figure: n = 10, n = 15, n = 17 and n = 24 for WT males and females and HSA^LR males and females, respectively. Quadriceps and gastrocnemius were analyzed for each individual.

Fig. 4. Alternative splicing defects.

Comparison of typical missplicing events in HSA^LR muscles in males (green) and females (orange). a–f, Percentages of abnormal inclusion of exon 7 in Nfix (a), exon 5 in Mbnl1 (b) and exon 7a in Clcn1 (c) transcripts and abnormal exclusion of exon 22 in Atp2A1 (d), exon 11 in Bin1 (e) and exon 29 in Cacna1s (f). The boxes indicate the two central quartiles, the midline represents median values and the whiskers indicate the minimum and maximum data value, excluding outliers. g, Representative images of agarose gel electrophoresis from semiquantitative PCR determinations of the indicated alternative exons (Ex). Statistical analysis was performed using two-way ANOVA (black asterisks). Minor effects are observed due to the sex of the animals, which are confirmed only in Atp2a1 using the Wilcoxon test (blue asterisks) ***P < 0.001; **P < 0.01; *P < 0.05. Sample sizes for the whole figure: n = 10, n = 15, n = 17 and n = 24 for WT males and females and HSA^LR males and females, respectively, except Clcn1 (n = 7, n = 14, n = 17 and n = 24 for WT males and females and HSA^LR males and females, respectively). Quadriceps and gastrocnemius were both analyzed for each individual.

Fig. 5. Plasma biochemical parameters.

a–f, Plasma levels of cholesterol (a), triglycerides (b), CPK (c), LDH (d), AST (e) and glucose (f) were analyzed in WT and HSA^LR mice. The boxes indicate the two central quartiles, the midline represents median values and the whiskers indicate the minimum and maximum data value, excluding outliers. Statistical analysis was performed using two-way ANOVA (black asterisks). Detected differences between sexes were confirmed using the Wilcoxon test (blue asterisks) ***P < 0.001; **P < 0.01; *P < 0.05. Sample sizes for the whole figure: n = 12, n = 13, n = 19 and n = 13 for WT males and females and HSA^LR males and females, respectively.