Collagen cross-links scale with passive stiffness in dystrophic mouse muscles, but are not altered with administration of a lysyl oxidase inhibitor

Abstract

In Duchenne muscular dystrophy (DMD), a lack of functional dystrophin leads to myofiber instability and progressive muscle damage that results in fibrosis. While fibrosis is primarily characterized by an accumulation of extracellular matrix (ECM) components, there are changes in ECM architecture during fibrosis that relate more closely to functional muscle stiffness. One of these architectural changes in dystrophic muscle is collagen cross-linking, which has been shown to increase the passive muscle stiffness in models of fibrosis including the mdx mouse, a model of DMD. We tested whether the intraperitoneal injections of beta-aminopropionitrile (BAPN), an inhibitor of the cross-linking enzyme lysyl oxidase, would reduce collagen cross-linking and passive stiffness in young and adult mdx mice compared to saline-injected controls. We found no significant differences between BAPN treated and saline treated mice in collagen cross-linking and stiffness parameters. However, we observed that while collagen cross-linking and passive stiffness scaled positively in dystrophic muscles, collagen fiber alignment scaled with passive stiffness distinctly between muscles. We also observed that the dystrophic diaphragm showed the most dramatic fibrosis in terms of collagen content, cross-linking, and stiffness. Overall, we show that while BAPN was not effective at reducing collagen cross-linking, the positive association between collagen cross-linking and stiffness in dystrophic muscles still show cross-linking as a viable target for reducing passive muscle stiffness in DMD or other fibrotic muscle conditions.

Fig 1. BAPN and muscle mechanics in wildtype and mdx.

(A) Representative image of isolated wildtype and mdx diaphragm (DP) muscles for ex vivo mechanical testing. (B) Total body weight was greater for adult mice than young mice in both mdx and wildtype. (C) In soleus, there were significant effects of age and genotype on muscle mass. (D) The specific tension of mdx soleus muscles were significantly lower than wildtype. (E) Elastic stiffness taken as the elastic modulus at 10% strain showed mdx diaphragm was significantly stiffer than the wildtypes, and young soleus was significantly stiffer than adult soleus. (F) Elastic index, related to the amount of force remaining after stress relaxation compared with the peak force, showed significant effect of age and genotype in diaphragm and soleus muscles. * = genotype effect, # = age effects, * = p<0.05 determined by three-way ANOVAs with post-hoc Sidak multiple comparisons tests.

Fig 2. Collagen content and cross-linking.

BAPN did not reduce collagen cross-links and the mdx diaphragm showed the most extensive fibrosis in terms of collagen content and cross-links. (A) The amount of collagen per muscle in wildtype and mdx muscles. The diaphragms and EDL muscle collagen content of mdx mice was significantly increased compared with the wildtype. (B) Insoluble (cross-linked) collagen was higher in mdx and adult muscles compared to wildtype. (C) The percentage of collagen that was cross-linked was higher in the mdx diaphragm and EDL muscles compared to the wildtype. Percent insoluble collagen was higher in the adult diaphragm and lower in the adult EDL compared to the respective young muscles. (D-F) Total collagen, insoluble collagen, and percent insoluble collagen scaled with elastic stiffness across all muscles together and the set of mdx points. * = genotype effect, # = age effects, * = p<0.05 determined by three-way ANOVAs with post-hoc Sidak multiple comparisons tests.

Fig 3. Sirius red collagen architecture.

The mdx diaphragm had a decrease in microECM alignment compared to wildtype, but this was not correlated to a change in elastic stiffness. (A) Representative images of MicroECM alignment (left) and MacroECM deviation (right) in wildtype and mdx diaphragm muscles. (B) The MacroECM deviation showed no significant differences between genotypes or ages in the diaphragm. (C) The quantification of the MicroECM alignment showed significant differences between young and adult in both wildtype and mdx diaphragms. (D) The relationship between MacroECM alignment and elastic stiffness showed no significant correlation overall or in individual muscles. (E) There was no significant relationship between MicroECM deviation and elastic stiffness overall or in any individual muscle. Significance bar represents significant main effect as determined by three-way ANOVA across treatment, genotype, and age. Significance symbol above groups indicates significance from corresponding genotype or age group according to post-hoc Sidak multiple comparisons tests. # = age effects, # = p<0.05.

Fig 4. Second harmonic generation collagen architecture.

Collagen architecture is altered in dystrophic muscle and relates to passive mechanics. (A) Second Harmonic Generation (SHG) representative images of diaphragm and EDL muscle sections from young, adult, wildtype, and mdx mice. Colormap represents angles of each pixel window in the images as analyzed by OrientationJ. Scale bar is 100μm. (B) The mdx EDL had significantly lower collagen deviation (higher alignment) compared to the wildtype. (C) Collagen fiber area was increased in mdx diaphragms compared to wildtype. Adult EDL muscles had higher collagen fiber area than young EDL muscles. (D) All groups combined had a significant positive correlation of collagen deviation and elastic stiffness. The EDL had an independent significant negative correlation. (E) Collagen fiber area correlated positively with elastic stiffness across all muscles. Significance by genotype: *p<0.05and age: #p<0.05.

Fig 5. Multiple linear regressions for elastic stiffness.

Aspects of collagen architecture predict elastic stiffness across muscles and within EDL and soleus. (A) A multiple linear regression model was run in MATLAB to predict elastic stiffness across all muscles and individually within each muscle. The model produced significant predictors collagen deviation, insoluble collagen, and percent insoluble. (B) Collagen deviation and soluble collagen were negative predictors and collagen fiber area was a positive predictor of EDL elastic stiffness. (C) Collagen deviation was a positive predictor of elastic stiffness in the soleus.

Fig 6. Microstructual and mechanical properties of long bones.

(A) Representative images of mid-diaphyseal total cross-sectional area of the femur by μCT. (B) Total cross-sectional area was significantly increased in adult and mdx mice. (C) Cortical bone area was significantly increased in both male and female mice, and female mice also showed decreased area in mdx mice. (D) Representative images of cortical bone area. (E) There was an increase in trabecular thickness in adult mice, and a decrease in mdx mice. BAPN also decreased thickness in adult female wildtype mice. (F) Trabecular bone volume fraction was decreased in mdx mice. (G) Adult mice had significantly increased elastic modulus. (H) Adult mice had significantly increased ultimate stress. Significance by genotype: *p<0.05, age: #p<0.05, and treatment: @p<0.05 determined by three-way ANOVA test with post-hoc Sidak multiple comparisons tests.