Synopsis

We used time-dependent diffusion, with diffusion times up to 500ms to assess the muscle tissue microstructure in two different mouse models of Duchenne muscular dystrophy (DMD). We found significant differences in diffusivities between DMD models and the genetic background matched wild types in diffusion times up to 100ms. We modeled diffusion time-dependence with the random permeable barrier model (RPBM), which yielded surface-to-volume ratios indicative of fiber size and compared these values to gold-standard histology of the same muscles. The RPBM predicted relative differences in fiber dimensions between mdx and wild type within each genetic background.

Introduction

Method

Sixteen-week-old male DMD mice with two different genetic backgrounds (C57BL/10ScSnJ and DBA2/J) and corresponding wild-types (WT), consisting of n=18 mdx-BL10, n=22 WT-BL10, n=20 mdx-DBA and n=18 WT-DBA mice, were studied using a Bruker 7T PharmaScan and a 23mm resonator. The DBA2/J background resulted in smaller muscle fibers and increased muscle fibrosis in mdx-DBA mice compared to both the genetically-matched wild-type (WT-DBA) and mdx-BL10 mice, due to Anxa6 and Ltbp4 gene variants involved in regeneration and fibrosis.

A stimulated echo diffusion sequence with EPI readout was used. For all mdx-BL10 and WT-BL10 mice and two mdx-DBA mice, diffusion times were Δ=[20 100 200 300 400 500]ms and for all other mdx-DBA and WT-DBA mice, diffusion times were Δ=[20 100 200 500]ms. FOV was 20x10x1mm; NA=8; matrix=128x64; 2-segmented EPI; Partial Fourier=1.5. Diffusion tensors were calculated from two unweighted and b=300s/mm2 weighted images in 8 directions (effective b-values ranged from 321 to 573 s/mm²). The axial and radial (λ_⊥=(λ₂+λ₃)/2) diffusivities were averaged across the gastrocnemius muscle (figure 2) and modeled with the RPBM.

Mice were sacrificed the day after MRI and the gastrocnemius muscle was snap-frozen using cooled 2-methylbutane. Sections (8 µm) were stained for laminin and myofibers were segmented into individual objects with the BZ-X Analyzer (Keyence) and their perimeter-to-area ratio was calculated and averaged per mouse. 17 mdx-BL10, 21 WT-BL10, 10 mdx-DBA and 1 WT-DBA mice have been histologically processed so far.

Results

Diffusivity [MD,λ_1-3] decreased, while fractional anisotropy (FA) increased, with increasing Δ for all mouse strains. At Δ=20ms FA was different between mdx mice and their genetic matched wild type controls, at Δ=20 λ₃ was different in mdx-DBA mice and at Δ=100 FA was different in mdx-BL10 mice compared to their controls (Figure 3).

The histologically assessed perimeter-to-area ratio was significantly increased in mdx-BL10 compared to WT-BL10 mice (19.49+1.21% vs 13.51+1.60%; P<0.001), as was the corresponding RPBM surface-to-volume ratio (23.61+6.75% vs 19.27+2.71%; P=0.02), see figure 4. Also in mdx-DBA mice, the surface-to-volume ratio was significantly increased compared to WT-DBA mice (20.50+6.64% vs 16.13+2.60%, P=0.03). Correlation plots (figure 5) show a high correspondence between histology and the RPBM model for the BL10 background (r=0.44, P<0.01). For DBA mice, the correlation is less, due to the fact that wild-type histology still needs to be added; therefore the dynamic range of fiber sizes in the DBA data is at the moment too small.

Discussion

The most different DTI measure was FA, able at short diffusion times to differentiate between mdx and wild type mice. By increasing diffusion time, the ability to measure restriction increases; it is likely that in mdx mice the restricted effects of smaller muscle fibers on diffusivities is counteracted by the increased permeability of the muscle cells.

The RPBM surface-to-volume ratio differentiated between mdx and wild type mice in each genetic background. From the data, it is clear that strain-dependent effects are present beyond differences in fiber size. The mdx-DBA model is characterized by increased muscle fibrosis(8) and the RPBM reflects both the intracellular and extracellular space as single unit and this oversimplification might have biased the surface-to-volume ratio in this mouse model. Secondly, the membrane permeability may be different for each genetic background. We are currently performing histology for membrane leakage in these models (Evans Blue infiltration), after which we can also evaluate the RPBM model for permeability.

Conclusion

Acknowledgements

References

1. Briguet A, Courdier-Fruh I, Foster M, Meier T, Magyar JP. Histological parameters for the quantitative assessment of muscular dystrophy in the mdx-mouse. Neuromuscul Disord. 2004;14:675–82. 2. Vohra R, Batra A, Forbes SC, Vandenborne K, Walter GA. Magnetic Resonance Monitoring of Disease Progression in mdx Mice on Different Genetic Backgrounds. Am J Pathol. Elsevier; 2017 Sep;187(9):2060–70.

3. Wu JS, Li J, Greenman RL, Bennett D, Geisbush T, Rutkove SB. Assessment OF aged mdx mice by electrical impedance myography and magnetic resonance imaging. Muscle and Nerve. 2015;52(4):598–604. 4. Park J, Wicki J, Knoblaugh SE, Chamberlain JS, Lee D. Multi-Parametric MRI at 14T for Muscular Dystrophy Mice Treated with AAV Vector-Mediated Gene Therapy. PLoS One. 2015;10(4):e0124914.

5. McMillan AB, Shi D, Pratt SJP, Lovering RM. Diffusion tensor MRI to assess damage in healthy and dystrophic skeletal muscle after lengthening contractions. J Biomed Biotechnol. 2011 Jan;2011:970726.

6. Fieremans E, Lemberskiy G, Veraart J, Sigmund EE, Gyftopoulos S, Novikov DS. In vivo measurement of membrane permeability and myofiber size in human muscle using time-dependent diffusion tensor imaging and the random permeable barrier model. NMR Biomed. 2017 Mar;30(3).

7. Novikov DS, Fieremans E, Jensen JH, Helpern JA. Random walks with barriers. Nat Phys. 2011 Jun;7(6):508–14.

8. Fukada S, Morikawa D, Yamamoto Y, Yoshida T, Sumie N, Yamaguchi M, et al. Genetic background affects properties of satellite cells and mdx phenotypes. Am J Pathol. 2010;176(5):2414–24.