Muscle-specific increased expression of JAG1 improves skeletal muscle phenotype in dystrophin-deficient mice

Abstract

Therapeutic strategies for Duchenne Muscular Dystrophy (DMD) will likely require complementary approaches. One possibility is to explore genetic modifiers that improve muscle regeneration and function. The beneficial effects of the overexpression of Jagged-1 were described in escaper golden retriever muscular dystrophy (GRMD) dogs that had a near-normal life and validated in dystrophin-deficient zebrafish (1). To clarify the underlying biology of JAG1 overexpression in dystrophic muscles, we generated a transgenic mouse (mdx^5cv-JAG1) model that lacks dystrophin and overexpresses human JAG1 in striated muscles. Skeletal muscles from mdx^5cv-JAG1 and mdx^5cv mice were studied at one, four, and twelve-month time points. JAG1 expression in mdx^5cv-JAG1 increased by three to five times compared to mdx^5cv. Consequently, mdx^5cv-JAG1 muscles were significantly bigger and stronger than dystrophic controls, along with an increased number of myofibers. Proteomics data show increased dysferlin in mdx^5cv-JAG1 muscles and an association of Nsd1 with the phenotype. Our data supports the positive effect of JAG1 overexpression in dystrophic muscles.

Keywords: Biological Sciences; Duchenne Muscular Dystrophy; Jagged-1; genetic modifier; mouse model; muscles; physiology.

Figure 1.. mdx^5cv-JAG1 mice overexpress Jagged-1 across life.

(A) The top portion shows the construct used for our mdx^5cv controls, which had human JAG1 transcription blocked by a loxP-flanked STOP cassette. The bottom portion shows the final construct in which muscle creatine kinase drives the expression of Cre recombinase and allows overexpression of JAG1 and luciferase. (B) Schematic diagram of the breeding strategy to generate mice with striated muscle-specific Jagged-1 overexpression. Mdx^5cv;R26-LSL-JAG1 mice crossed with MCK-Cre mice will produce Mdx^5cv;R26-LSL-JAG1^Tg/WT,MCK-Cre offspring, which overexpress Jagged-1 in skeletal and cardiac muscle specific. (C) Western blot images showed increased expression of Jagged-1 at the three time points tested. Intensity changes were calculated using the Ponceau as loading control for each timepoint. All original membranes are in Supplemental Figure 1. (D) Immunofluorescence images of TA cross-sections at four months stained for DAPI (Cyan), jagged- 1 (Yellow), and laminin (Magenta). (E) Hematoxylin and Eosin (H&E) staining at the three timepoints tested (Columns). The top and bottom rows show images from mdx^5cv and mdx^5cv-JAG1, respectively. Scale bars are 250 μm.

Figure 2.. Changes in body and muscle mass over time.

(A - F) Plots for whole body mass, dissected quadriceps mass, quadriceps mass normalized per body mass, dissected tibialis anterior (TA) mass, TA mass normalized per body mass, and TA cross-section area for one-, four, and twelve months. The mdx^5cv and mdx^5cv-JAG1 groups are in red and blue, respectively. Plots show mean ±SEM and p-value. Each gray dot represents data collected from one animal.

Figure 3.. mdx^5cv-JAG1’s sections were bigger and showed differences in fiber type and size.

(A) Overlay of stained sections obtained from the thickest portion of the TA muscle. The left and right rows show sections for mdx^5cv and mdx^5cv-JAG1 at one-, four-, and twelve-month timepoints. Sections were stained for laminin (Magenta), myosin heavy chain type I, type IIA (yellow), and IIB (Cyan). Scale bars are 1000 μm. (B) Histograms comparing the normalized density of fibers from 10 μm to 100 μm between mdx^5cv (red) and mdx^5cv-JAG1 (blue). (C) Same as the figure 3B but organized by genotype to show the temporal progression of fiber density per minimum Feret. (D) Comparison of the total number of fibers per section at the three timepoints, and the number of fibers of each fiber type for all muscle sections from mdx^5cv and mdx^5cv-JAG1 mice. Type I fibers were excluded since they are less than 1% of the fiber type composition. Not stained fibers were considered type IIX. (E) Histograms comparing the normalized density of type IIB fibers from 10 μm to 100 μm. (F) mdx^5cv-JAG1 mice have an increased number of newly formed fibers at one month. From left to right, the figure shows laminin staining in magenta, nuclear DNA staining with DAPI in cyan, embryonic myosin heavy chain (eMyHC) in yellow, and the overlay of all channels. From top to bottom, the figure intercalates two different mice from each group, i.e. mdx^5cv-JAG1 and mdx^5cv. The scale bar is 500 μm. Additional eMyHC from the other four- and twelve-month-old time points are shown in Supplemental Figure 3. Error bars are SEM and level of significance between genotypes in the same age is indicated as * = p < 0.05 and ** = p < 0.01. Level of significance between time points of the same genotype is indicated as ^§ = p < 0.05, ^§§ = p < 0.01, and ^§§§ = p < 0.001 – the blue color indicates the mdx^5cv-JAG1 genotype, no difference was observed in the mdx^5cv genotype.

Figure 4.. Overexpression of Jagged-1 increases whole-muscle force over time.

(A) Peak force of TA muscles from each animal per time point. (B) Specific force (Peak force normalized by cross-section area (CSA)). (C) Representative force records of both genotypes tested. (D) Absolute force of TA muscles vs. frequency of stimulation. Error bars represent SEM and the shaded area indicates the estimated 95th confidence interval. For each curve, an F-test was used to compare a genotype-specific fitting model against a shared model across genotypes. A genotype-specific model provided the best fit for data at the four- and twelve-month time points (F(4,289) = 12.47, p < 0.001 and F(4,253) = 7.71, p < 0.0001, respectively), but not at the one-month time point (F(4,124) = 1.66, p = 0.163). (E) Similar to D, but for specific force. A genotype-specific model provided the best fit for data at the four- and twelve-month time points (F(4,289) = 7.00, p < 0.0001 and F(4,253) = 3.43, p = 0.0095, respectively), but not at the one-month time point (F(4,124) = 0.16, p = 0.959). (F) Change in isometric force of TA muscles after the eccentric contraction protocol. Values are the post-protocol isometric force expressed relative to the pre-protocol isometric force. Supplemental Figure 4 shows the progressive decline in force obtained during these experiments. (G) Jagged-1 expression in quadriceps muscles assessed by Western blot. Intensity of chemiluminescence was normalized to whole protein lysate content (i.e. Ponceau staining intensity). Each dot represents Western blot analysis performed in an individual mouse. Large discrepancies were observed within the mdx^5cv-Jag1 group at 12 months of age in which most mice display increased Jagged-1 expression as similarly seen at 1 and 4 months of age (3-5-fold increase), while a small cohort of mice displays 10-fold increased expression (referred to as supra group). (H-K) Peak force, specific force, force-frequency, and specific force-frequency relationships when animals with supra-abundance of Jagged-1 are considered as separate group. A two-way ANOVA showed a significant effect for level of jagged-1 abundance in Force and Force/CSA (p = 0.000958 and p = 0.00435 respectively). Bonferroni-adjusted pairwise comparing Force (Fig. H) showed significant differences between Baseline vs. Increased (p.adj= 0.0284) and Increased vs. Supra (p.adj= 0.00288), whereas Baseline vs. Supra was not significant (p.adj= 0.213). The same statistical test for Force/CSA (Fig. I) showed differences between Baseline vs. Supra (p.adj= 0.00142) and Increased vs. Supra (p.adj= 0.0131), whereas Baseline vs. Increased was not significant (p.adj= 1.000). Error bars are SEM and shaded area is the estimated 95^th confidence interval, and level of significance is indicated as * = p < 0.05, ** = p < 0.01, and *** = p < 0.001.

Figure 5.. Proteomics using quadriceps muscles dissected from all the time points studied.

(A-C) Each point in the volcano plot show the fold change of a protein comparing mdx^5cv-JAG1 over mdx^5cv. Orange and blue dots show all proteins with statistically significant changes in expression and with fold change > 1.5. Magenta points show specific cases described in the main text. (D) Heat map showing the Z-score normalized per line of selected important in the context of this study.