Visual acuity (VA) in Fingolimod-treated mice was better than that in saline-treated mice
There was no VA difference between fingolimod-treated (0.34 ± 0.03 c/d, n = 16 eyes, from 8 EAE mice) and saline-treated groups (0.36 ± 0.04 c/d, n = 18 eyes, from 9 EAE mice) at baseline before treatment. Over the treatment period, the VA of fingolimod-treated mice was significantly better than saline-treated mice from 3 days to 10 weeks after treatment onset (p < 0.05, Fig. 1B).
DBSI detected the reduced inflammation in fingolimod-treated optic nerves
Representative DBSI non-restricted isotropic fraction (putatively reflecting vasogenic edema) and restricted isotropic fraction (putatively reflecting cellularity) maps of optic nerves were presented to demonstrate the evolution of these markers at baseline (before immunization), 2, 6, and 10 weeks after treatment (Fig. 2A).
In the group of saline-treated mice with ON, non-restricted isotropic fraction of optic nerves significantly increased from baseline by 133% (p = 0.020), 198% (p = 0.001), and 229% (p < 0.001) at 2, 6, and 10 weeks after treatment, respectively (Fig. 2B). Restricted isotropic fraction of optic nerves significantly increased from the baseline by 201% (p < 0.001), 98% (p = 0.070), and 164% (p = 0.002) at 2, 6, and 10 weeks after placebo treatment, respectively (Fig. 2C). In the fingolimod treatment group, non-restricted isotropic fraction increased from baseline by 21% (p = 0.740), 48% (p = 0.440), and 63% (p = 0.320) at 2, 6, and 10 weeks after treatment (Fig. 2B), while restricted isotropic fraction increased from baseline by 72% (p = 0.180), 28% (p = 0.590), and 51% (p = 0.340) at 2, 6, and 10 weeks (Fig. 2C).
In comparison to the saline-treated group, the fingolimod-treated mice exhibited significantly lower non-restricted isotropic fraction at 2 (60% lower, p = 0.070), 6 (62% lower, p = 0.020) and 10 weeks (63% lower, p = 0.010) after treatment (Fig. 2B). Although the increase in restricted isotropic fraction was numerically greater for the placebo group, differences between the two treatment groups were not statistically significant.
DBSI detected reduced axon loss in Fingolimod-treated optic nerves
Representative fiber fraction (the total diffusion signal from anisotropic axonal fiber bundles) maps of optic nerves at baseline, 2, 6, and 10 weeks after treatment were presented to demonstrate the evolution of fiber fraction (Fig. 3A).
In the saline-treated group, DBSI fiber fraction (putative marker of apparent axonal density) of the optic nerves decreased by 10% (p = 0.140), 24% (p = 0.002) and 25% (p = 0.003) at 2, 6, and 10 weeks, respectively, when compared to baseline before immunization (Fig. 3B). DWI-derived volumes of optic nerves decreased by 23% (p = 0.004), 37% (p < 0.001) and 25% (p = 0.002) at 2, 6, and 10 weeks, respectively, from baseline before immunization (Fig. 3C). DBSI-derived axonal volume (putative marker of axonal content) decreased by 28% (p = 0.006), 45% (p < 0.001), and 36% (p < 0.001) at 2, 6, and 10 weeks, respectively, when compared to baseline before immunization (Fig. 3D). In the fingolimod-treated group, no significant difference from baseline was observed for DBSI fiber fraction, DWI-derived optic nerve volume, or DBSI axonal volume at any time point after treatment.
Compared to saline-treated mice, the fingolimod-treated group exhibited significantly higher DBSI fiber fraction at 6 weeks (38% higher, p = 0.020) and 10 weeks (Fig. 3B, 33% higher, p = 0.030), with significantly lower DWI-derived optic nerve volumes at 6 weeks (35% higher, p = 0.02) and at 10 weeks (39% higher, p = 0.005, Fig. 3C). DBSI-derived axonal volumes in the fingolimod-treated optic nerves were higher than for saline-treated optic nerves at 2 weeks (25% higher, p = 0.400), 6 weeks (62% higher, p = 0.020) and at 10 weeks (64% higher, p = 0.007) after treatment commencement (Fig. 3D).
DBSI detection of diminished injury to residual axons in Fingolimod-treated optic nerves
Figure 4A shows representative DBSI axial diffusivity (λ∥) and radial diffusivity (λ⊥) maps at baseline, 2, 6, and 10 weeks after treatment for each group. Compared to baseline, for saline-treated mice DBSI λ∥ significantly decreased by 17% (p < 0.001), 14% (p < 0.001), and 12% (p = 0.001) at 2, 6, and 10 weeks (Fig. 4B), while DBSI λ⊥ significantly increased by 45% (p = 0.015), 62% (p = 0.001) and 75% (p < 0.001) at 2, 6, and 10 weeks (Fig. 4C), respectively. In the fingolimod treatment group, DBSI λ∥ decreased by 7% (p = 0.040), 7% (p = 0.033), 3% (p = 0.330) at 2, 6, 10 weeks (Fig. 4B), while DBSI λ⊥ increased from baseline by 14% (p = 0.400), 15% (p = 0.400) and 9% (p = 0.600) at 2, 6 and 10 weeks after treatment (Fig. 4C).
Compared the difference between groups (saline vs. fingolimod), fingolimod-treated group showed no significant differences on DBSI λ∥ for all the time points while DBSI λ⊥ decreased by 38% at 10 weeks (p = 0.010, Fig. 4C).
Immunohistochemistry (IHC) staining of optic nerves
Mice were euthanized at the end of the treatment period for histologic assessment of the optic nerves. IHC staining of total neurofilaments (SMI-312, stains both injured and intact axons), phosphorylated neurofilaments (SMI-31, stains intact axons), and myelin basic protein (MBP, stains myelin) revealed that the saline-treated group exhibited less staining for SMI-312 (Fig. 5A), MBP (Fig. 5D), and SMI-31 (Fig. 5G), when compared to Fingolimod-treated group (Fig. 5B, E, H). The apparent optic nerve sizes of saline-treated group were smaller than those of Fingolimod-treated group (Figs. 5A - I). However, compared to one representative wildtype (WT) optic nerves (10-week-old C57BL/6 female mouse optic nerve), fingolimod-treated group showed enlarged staining spots (Fig. 5b, e, h; blue arrows), suggesting swollen axons or myelin debris aggregation. Moreover, increased positive DAPI staining areas (cell nuclei) were also seen in both saline and Fingolimod treatment groups compared to WT optic nerve, suggesting increased cellularity (Figs. 5J, K).
DBSI λ∥, DBSI λ⊥, DBSI-derived axon volume and DBSI restricted fraction correlated well with SMI-31 area fraction (y = 1.7475x + 1.4709, slope of standard error = 0.4246, Fig. 6A), MBP area fraction (y = -0.7347x + 0.3794, slope of standard error = 0.1556, Fig. 6B), SMI-312 area fraction (y = 5.0633x + 0.0303, slope of standard error = 0.8941, Fig. 6C) and DAPI density (y = 9.585 × 10− 6 x + 0.03, slope of standard error = 1.506 × 10− 6, Fig; 6D), respectively in the ON-affected optic nerves.