Diffusion tensor imaging detects treatment effects of FTY720 in experimental autoimmune encephalomyelitis mice

Abstract

Fingolimod (FTY720) is an orally available sphingosine-1-phosphate (S1P) receptor modulator reducing relapse frequency in patients with relapsing-remitting multiple sclerosis (RRMS). In addition to immunosuppression, neuronal protection by FTY720 has also been suggested, but remains controversial. Axial and radial diffusivities derived from in vivo diffusion tensor imaging (DTI) were employed as noninvasive biomarkers of axonal injury and demyelination to assess axonal protection by FTY720 in experimental autoimmune encephalomyelitis (EAE) mice. EAE was induced through active immunization of C57BL/6 mice using myelin oligodendrocyte glycoprotein peptide 35-55 (MOG(35-55)). We evaluated both the prophylactic and therapeutic treatment effect of FTY720 at doses of 3 and 10 mg/kg on EAE mice by daily clinical scoring and end-point in vivo DTI. Prophylactic administration of FTY720 suppressed the disease onset and prevented axon and myelin damage when compared with EAE mice without treatment. Therapeutic treatment by FTY720 did not prevent EAE onset, but reduced disease severity, improving axial and radial diffusivity towards the control values without statistical significance. Consistent with previous findings, in vivo DTI-derived axial and radial diffusivity correlated with clinical scores in EAE mice. The results support the use of in vivo DTI as an effective outcome measure for preclinical drug development.

Keywords: FTY720; axial diffusivity; axonal injury; demyelination; diffusion tensor imaging; experimental autoimmune encephalomyelitis; multiple sclerosis; radial diffusivity.

Figure 1. Sagittal and coronal scout images of mouse vertebral column

Mouse vertebral discs can be seen on the sagittal image while ribs can be seen on the coronal images. Discs and ribs were used as references to determine target axial image plan.

Figure 2. Region of interest (ROI) of spinal cord ventrolateral white matter (VLWM).

ROIs of spinal cord L4 level were shown on B0 image, diffusion-weighted (DW) image, λ_┴ map, λ_║ map and RA map. The boundary between white matter and cerebrospinal fluid was identified on the B0 image (no diffusion weighting) and that of white and grey matter was identified on one of the raw diffusion weighted image (-1,1,0). DTI-derived λ_║, λ_┴ and RA maps were used as references while defining VLWM ROIs.

Figure 3. EAE disease course (a, prophylactic and c, therapeutic) and group-averaged clinical score at end point (b, prophylactic and d, therapeutic)

In prophylactic study, FTY720 in both doses prevented the EAE onset and suppressed the disease throughout the course of study (a). Mean clinical scores of both treated groups decreased to values less than 1 without statistical significance (b). In therapeutic study, 3 mg/kg treatment after the disease onset delayed the disease peak while 10 mg/kg treatment ameliorated the disease severity (c). Lower mean CS was also seen in both treatment groups (without reaching statistical significance). Open circle: non-treated EAE group; open square: FTY720 3 mg/kg group; open triangle: FTY720 10 mg/kg group. Error bar represents standard error in Figure 3a and 3c, standard deviation in Figure 3b and 3d.

Figure 4. DTI parameter maps and values of each group in prophylactic study end point

Color-coded DTI maps of spinal cord L4 level are shown for a representative mouse from each of the four groups (a). Improvement in VLWM integrity was seen in treated mice cords marked by normalized λ_║, λ_┴ and RA. Group-averaged λ_║ (b), λ_┴ (c) and RA (d) show the same trend. Mean RA and λ_║ of both FTY720 treated groups are significantly higher than that of vehicle treated group (P<0.05). Mean λ_┴ of 3 mg/kg treatment group is significantly lower than that of vehicle treated group (P<0.05). The 10 mg/kg treatment group mean does not show statistical difference. No statistical difference exists between control group and either of the FTY720 treated groups in RA, λ_║, or λ_┴. Error bar represents standard deviation. *P< 0.05, **P<0.01, ***P<0.001 and ****P<0.0001.

Figure 5. DTI parameter maps and values of each group in therapeutic study end point

Color-coded DTI maps of spinal cord L4 level are shown for a representative mouse from each of the four groups (a). Treated groups still shown increased λ_║, increased RA, and decreased λ_┴ in the VLWM region compared to those of vehicle treated EAE group (a). However, no statistical improvement was detected in either FTY720 treated group in λ_║ (b), λ_┴ (c), or RA (d), (P>0.05). Error bar represents standard deviation. *P< 0.05, **P<0.01, ***P<0.001 and ****P<0.0001.

Figure 6. SMI-31 (a) and MBP (b) staining of L4 spinal cord sections from representative mice in prophylactic study

Substantially decreased SMI-31-positive axons (a) and MBP-positive myelin rings (b) can be seen in the vehicle treated mouse cord. The density and morphology of SMI-31-positive axons in both representative FTY720 treated cords resemble those in shame cord (a). Comparable myelin density between sham mouse cord and FTY720 3mg/kg treated mouse cord can also be seen. On the other hand, the 10 mg/kg treated cord show mild decrease in myelin density (b). Scale bar: 50µm.

Figure 7. SMI31 (a) and MBP (b) staining of L4 spinal cord sections from representative mice in therapeutic study

Drastic loss of SMI31-positive axons (a) and MBP-positive myelin rings (b) can be observed in the vehicle treated mouse cords. Therapeutic treatment at both doses helped preserving axon density in the VLWM (a). However, axon loss proximity to the meninges can still be seen in 3 and 10 mg/kg treated cords (a). Similarly, FTY720 treated cords show partially improved myelin density comparing to vehicle treated cord (b). Scale bar: 50µm.