Magnetic resonance elastography reveals altered brain viscoelasticity in experimental autoimmune encephalomyelitis

Abstract

Cerebral magnetic resonance elastography (MRE) measures the viscoelastic properties of brain tissues in vivo. It was recently shown that brain viscoelasticity is reduced in patients with multiple sclerosis (MS), highlighting the potential of cerebral MRE to detect tissue pathology during neuroinflammation. To further investigate the relationship between inflammation and brain viscoelasticity, we applied MRE to a mouse model of MS, experimental autoimmune encephalomyelitis (EAE). EAE was induced and monitored by MRE in a 7-tesla animal MRI scanner over 4 weeks. At the peak of the disease (day 14 after immunization), we detected a significant decrease in both the storage modulus (G') and the loss modulus (G″), indicating that both the elasticity and the viscosity of the brain are reduced during acute inflammation. Interestingly, these parameters normalized at a later time point (day 28) corresponding to the clinical recovery phase. Consistent with this, we observed a clear correlation between viscoelastic tissue alteration and the magnitude of perivascular T cell infiltration at both day 14 and day 28. Hence, acute neuroinflammation is associated with reduced mechanical cohesion of brain tissues. Moreover, the reduction of brain viscoelasticity appears to be a reversible process, which is restored when inflammation resolves. For the first time, our study has demonstrated the applicability of cerebral MRE in EAE, and showed that this novel imaging technology is highly sensitive to early tissue alterations resulting from the inflammatory processes. Thus, MRE may serve to monitor early stages of perivascular immune infiltration during neuroinflammation.

Keywords: Experimental autoimmune encephalomyelitis; Magnetic resonance elastography; Neuroinflammation.

Fig. S1

Schematic of MRE experimental design.

Fig. S2

Elastography parameters are not correlated with EAE severity. The cumulative disease activity did not show any statistically significant correlation with any of the elastography parameters (Pearson correlation). Depicted here are representative data from the EAE recovery phase experiment (shown in Fig. 3).

Fig. S3

The magnitude of CD3 epsilon expression in cerebellum is significantly correlated with EAE disease severity (Spearman r = 0.6397, p = 0.0186).

Fig. 1

Hypothesis, experimental setup and baseline measurements for studying EAE with MRE in the mouse. A. Conceptual diagram illustrating the interpretation of the elastography parameters. A reduction in the storage (G′), and loss (G″) moduli, while maintaining a constant ratio of these terms, is indicative of reduced mechanical rigidity of the tissue, which can be associated with acute pathological processes leading to tissue alteration. A change in the ratio G″/G′ reflects alteration of the structural complexity of the tissue architecture, which can be associated with severe tissue degradation in pathological states. B. A schematic of the mouse MRE apparatus: (A) driving coil; (B) carbon fiber piston; (C) respiratory mask; (D) rubber bearing; (E) retaining bracket; (F) mouse bed; (G) magnet bore; (H) plastic disk; (I) Lorentz coil. C. Representative wave (deflection) images and complex modulus maps, superimposed on T2-weighted anatomical scans of the mouse head. D. Plots of the storage and loss moduli of healthy mice show no significant changes over 28 days.

Fig. 2

Brain viscoelasticity is reduced during acute inflammation. The EAE clinical course and plot of cumulative disease activity are shown in A and B. Significant reductions in the storage modulus (elasticity) C, loss modulus (viscosity) D, and the magnitude modulus E at day 14 are seen in EAE mice, but not in controls. No change is seen in the ratio of G″/G′, F. n = 10 and n = 14 for controls and EAE group, respectively.

Fig. 3

Brain viscoelasticity normalizes during the recovery phase of EAE. The EAE clinical course and plot of cumulative disease activity are shown in A and B. In EAE mice, the significant reductions in the storage modulus (elasticity) C, loss modulus (viscosity) D, and the magnitude modulus E seen at day 14 are normalized at day 28, and no longer significantly different from the baseline values. No change is seen in the ratio of G″/G′, F. n = 10 and n = 10 for controls and EAE group.

Fig. 4

Histological features in EAE. Representative images of hematoxylin and eosin (H&E) staining show absence of pathology in healthy controls, in contrast to robust inflammatory lesions (arrows) in the cerebellum white matter at the peak of the disease (d14), and relatively milder lesions from the recovery phase (day 28). Staining with luxol fast blue (LFB) indicates that the cerebellar lesions (arrows) are not associated with demyelination.

Fig. 5

Elastographic alterations correlate with CD3 epsilon expression. Immunohistochemical staining for CD3 reveals the presence of T lymphocytes in mild, A, and severe, B cerebellar white matter lesions (arrows). The level of CD3 epsilon mRNA in the cerebellum (from both the d14 and d28 time points) detected by quantitative real-time PCR is significantly correlated with the magnitude of the storage modulus (elasticity) C, loss modulus (viscosity) D, and the magnitude modulus E values, but not with the ratio of G″/G′, F.