3.1. Edaravone improved spatial learning memory of P301L mice in MWM test
To assess the effect of EDR on spatial reference learning, vehicle and EDR treated mice with their WT littermates were tested in MWM. Results showed that the average escape latency during all hidden platform trials did not differ between groups. All groups learned to locate the hidden platform as evident by decrease in latency time during the acquisition phase {14 months (n = 9–14/group): time effect: F (1.757, 15.81) = 0.9567; p = 0.3944, treatment effect: F (2, 9) = 0.2654; p = 0.7727, treatment/time effect: F (6, 27) = 0.3892; p = 0.8794, {25 months [n = 4/group]: time effect: F (2.748, 24.74) = 1.848; p = 0.1681, treatment effect: F (2, 9) = 3.746; p = 0.0655, treatment/time effect: F (6, 27) = 1.656; p = 0.1705} (Fig. 1a and 1b). However, EDR treated mice and WT littermates showed reduced average escape latency compared to vehicle controls.
In the probe trial, 14 months old control P301L mice spent less time in the target quadrant zone compared to WT (F(2,29) = 3.821; p = 0.033) and showed the least number of platform site crossing (F(2,29) = 0.8158; p = 0.452). EDR treated mice performed better than vehicle controls as there was no difference in the time spent in the target zone between WT and EDR mice (p = 0.6891) (Fig. 1c, 1e). No difference in the time spent in the target zone was found in 25-month-old mice (F (2,9) = 0.1265, p = 0.8827). However, EDR treated mice showed significantly higher number of platform site crossing compared to vehicle control group (F (2, 9) = 7.929; p = 0.0236), and nearly equal to WT group (p = 0.943). These results suggest that EDR improved acquisition and search accuracy of P301L mice (Fig. 1d, 1f).
3.2 Edaravone altered the exploratory behavior of P301L mice in open field and elevated plus maze tests
Mice were tested in open field and elevated plus maze to evaluate the effect of Edaravone treatment on anxiety, exploratory and locomotor behavior. In open field test, control P301L mice showed significantly longer immobility time % (F (2,9) = 5.336, p = 0.0256) compared to WT littermates. They also displayed less tendency to explore the central (anxious) zone (less time in central zone) (F(2,9) = 7.576, p = 0.0139), and travelled less distance (F(2,9) = 7.307, p = 0.005), with less average speed (F(2,9) = 7.193, p = 0.0046) compared to WT mice. They also showed less tendency of exploration as evident by reduced rearing behavior (F(2,9) = 3.947, p = 0.038). EDR treatment did not improve or reverse such behavior in the 14 months old group. The results of immobility time and the time spent in the central zone were similar among younger mice and no difference was found between vehicle and EDR groups. At the old age group, both control and EDR treated P301L mice spent significantly less time in central zone compared to WT mice (F (2, 9) = 9.201; p = 0.0089 for control P301L vs WT, p = 0.0175 for EDR treated P301L vs WT). But there was no difference between vehicle and EDR treated mice (p = 0.89). Also, no significant difference was observed between P301L and WT mice for total distance, average speed, rearing episodes, number of zone transitions (line crossing) and immobility time %, possibly due to small sample size. (Fig. 2a-f).
In the elevated plus maze task, no significant difference was found between WT and P301L mice at both age groups regarding time (14 months old group: F (2, 29) = 1.742; p = 0.193, 25 months old group: F (2, 9) = 2.244; p = 0.1619) and entries to open arms (14 months old group: F(2, 29) = 2.162; p = 0.133, 25 months old group: F (2, 9) = 0.855; p = 0.457), distance (14 months old group: F(2, 29) = 0.5682; p = 0.5727, 25 months old group: F (2, 9) = 0.965; p = 0.417), and total number of head dips (14 months old group: F(2, 29) = 2.923; p = 0.0697, 25 months old group: F (2, 9) = 0.309; p = 0.742). However, we observed that control P301L mice spent more time in the open arms and showed overall higher exportation and spontaneous locomotion. In contrast, EDR treated littermates tended to spend more time in the closed arms similar to WT littermates which may indicate that they can recognize their starting location with more protected head dips. We also found that EDR significantly increased the percentage of protected head dips of P301L mice at 25 months old group compared to vehicle controls (F (2, 9) = 9.575; p = 0.011 vs controls) (Fig. 2g-k).
3.3 Edaravone did not affect the working memory in Y maze novel arm test and spontaneous alternation test
In order to investigate the effect of EDR on spatial and short-term working memory, mice were tested in the Y maze test. In the spontaneous alternation task, all mice in the 14 months old group showed the same percentage of spontaneous alternation (F(2, 29) = 0.7303, p = 0.4904). However, in the 14 months old group, WT mice showed more total arm entries compared to control (F(2, 29) = 4.714, p = 0.08) and significantly more than EDR treated mice (p = 0.02). Similarly, in the 25 months old age group, there was no significant difference in spontaneous alternation % among groups (F(2, 9) = 2.928, p = 0.1048), but control P301L mice showed the highest alternation (70%) while WT and EDR treated mice alternated between 50 and 55%. On the other hand, the total number of arm entries was significantly higher in WT mice compared to control P301L mice (F(2, 9) = 4.344, p = 0.042), while EDR treatment slightly increased it but the difference was not statistically significant compared to both WT (p = 0.580) and vehicle treated mice (p = 0.199) (Fig. 3a, b). In the Y maze novel arm task, there was no significant difference in novel arm time (14 months old group: F(2, 29) = 0.3917; p = 0.679, 25 months old group: (F(2, 9) = 1.145, p = 0.361) or entries (14 months old group: F(2, 29) = 0.08491; p = 0.919, 25 months old group: F(2, 9) = 0.3302, p = 0.727) between WT and P301L mice in both age groups (Fig. 3c, d).
3.4. EDR improved the recognition memory in P301L mice
Novel object recognition test was used to evaluate the effect of EDR on the animal tendency to explore novel objects over familiar ones. Results showed that there was no significant difference in the recognition index between groups in the 14 months old group (F(2, 29) = 0.7909, p = 0.4630). However, in the 25 months old group, control P301L mice showed less interaction with the novel object (RI = 35%) compared to WT mice (RI = 51%), while EDR significantly improved the recognition memory of P301L treated mice (F(2, 9) = 5.099, p = 0.0269, RI = 70%) compared to their control littermates and performed even better than WT mice (p = 0.2429) (Fig. 4a).
3.5. EDR attenuated motor deficits in P301L mice
In order to test the effect of EDR on the motor deficits observed in P301L mice, limb clasping test was carried out. In both age groups, control P301L mice showed significantly higher limb clasping score compared to WT (14 months old: F (2, 29) = 15.20, p < 0.0001, 25 months old: F (2, 9) = 8.471, p = 0.007) that agrees with previous findings observed in that model [49]. EDR reduced the limb clasping score compared to vehicle controls (14 months old: p = 0.0217, 25 months old: p = 0.1541) (Fig. 4b). These results suggest that EDR rescued motor function in P301L mice.
3.6. Effect of EDR on phosphorylated Tau expression in P301L mice
To study the effect of EDR treatment on Tau pathology, we did immunostaining and western blotting of whole brain homogenate for epitope pS396 Tau and AT8 (an antibody for phosphorylated Tau at Ser202/Thr205). Given the initial characterization of P301L mice, Tau aggregates can be detected in several brain areas; early in the basolateral nucleus (BLA) of amygdala and later in the hippocampus (< 18 months old) [37, 35]. The expression of endogenous mouse tau and mutant human tau was detected by mouse tau 5 and HT7 antibodies, respectively. While neurons of WT littermates did not show any immunoreactivity towards HT7 antibody (data not shown). IHC revealed that EDR treatment significantly reduced the immunoreactivity of pS396 in the hippocampus (14 months old: t(34) = 2.309, p = 0.027; 25 months old: t(14) = 3.681, p = 0.0025) and amygdala (age = 14 months old: t(35) = 1.692, p = 0.099, age = 25 months old: t(14) = 3.681, p = 0.0293) compared to controls (Fig. 5A, 7A). In the same context, western blotting also showed modest reduction in the expression levels of pS396 and AT8 compared to total HT7, though, seemed to be less sensitive than IHC to detect any change in the total brain homogenate. The expression levels of total mouse tau and mutated human tau were not altered in control and treated mice. But, interestingly, HT7 expression level was significantly reduced in 25 months old EDR treated mice compared to their control littermates (t(14) = 3.681, p < 0.0407, Fig. 6a and 8a). We also performed two-way ANOVA to compare the hyperphosphorylated Tau expression levels between the two age groups. We found significant increase in levels of pS396 Tau and AT8 in the older age group compared to both vehicle and EDR treated 14 months old P301L mice, that indicated the buildup of the pathology with aging. Interestingly, EDR significantly reduced the expressions levels of pS396 phosphorylated Tau F (3, 14) = 17.48, p < 0.0001) and AT8 (F (3, 14) = 17.12, p < 0.0001) in the 25 months old P301L mice compared to the age matched vehicle controls (please refer to supplement Figs. 14,15 and 16 in supporting information).
3.7. EDR treatment altered GSK-mediated Tau phosphorylation
To assess the effect of EDR on the activity of Glycogen synthase kinase 3β (GSK-3β), we performed western blotting for the total protein level and its phosphorylated form at Ser9 (pS9-GSK3β). GSK-3β is a serine/threonine kinase and involved in aberrant Tau phosphorylation in neurodegenerative diseases [50, 51]. Western blot revealed that P301L mice showed significantly reduced levels of the inactive pS9-GSK3β (inhibited) form to the total protein compared to age matched WT mice in the 14 months old group, but no difference among all groups in the older age (14 months old: F (2, 23) = 14.38, p < 0.0001; 25 months old: F (2, 8) = 0.1093, p = 0.8978). However, total GSK-3β protein levels did not change. On the hand, EDR did not increase the levels of the phosphorylated form of GSK-3β in treated P301L mice compared to vehicle controls. This might indicate that the effect of EDR treatment on Tau hyperphosphorylation is mediated through different mechanisms but not mainly through altering the activity of GSK-3β (Fig. 6b and 8b). Performing Two-way ANOVA for samples run on same gel to compare the expression levels of pS9-GSK3β among groups of different ages, we found significant increase in the levels of the phosphorylated form to the total protein levels in the EDR treated older P301L mice compared to age matched vehicle controls and to both vehicle and EDR treated younger P301L mice (F (3, 14) = 14.74, p = 0.0001 please refer to supplement Fig. 19 in supporting information).
3.8. EDR treatment reduced neuroinflammation in the brains of P301L mice
In order to study effect of EDR on neuroinflammation, we examined the expression levels of glial fibrillary acidic protein (GFAP); a marker of astrocytosis and CD45; a leukocyte common antigen and Iba-1; both detect activated microglia. IHC showed that the density of GFAP-immunoreactive astrocytes was reduced in EDR treated mice from 9.4–4.7% (t(6) = 3.760, p = 0.0094) in the hippocampus and from 7.1–4.4% in the amygdala (t(6) = 3.569, p = 0.0118, Fig. 7C) at 25 months old, and showed a trend towards reduction in the hippocampus of 14 months old group compared to controls (t(16) = 0.8769, p = 0.393, Fig. 5C). In addition, EDR also significantly reduced CD45 positive microglia in the brains of old P301L mice (t(6) = 2.576, p = 0.042, Fig. 7E). Western blotting of the whole brain homogenate showed subtle changes in GFAP expression levels of in both age groups (Fig. 6c and 8c), and significant reduction in Iba-1 expression in 25 months age P301L group compared to age matched controls (F(2,9) = 6.412, p = 0.0186, Fig. 8c). These results suggest that EDR oral formulation displayed some anti-inflammatory effects, which is more obvious in the older age group as Tau pathology progresses with aging.
3.9 EDR treatment attenuates loss of synaptic plasticity and protected neurons in the brain of P301L mice
Then, we studied the effect of EDR on the expression levels of some synaptic proteins including post synaptic density-95 (PSD-95), vesicle-associated membrane protein-2 (VAMP-2) and synaptophysin. Immunoblotting results showed that EDR oral formulation significantly increased the expression of synaptophysin in the 14 months old group (F (3, 22) = 4.947, p = 0.0089 vs age matched controls, two-way ANOVA, supplement Fig. 18 ), while slightly improved the expression levels of other synaptic proteins (Fig. 6d and 8d). To study the effect of EDR on neuron cell viability, we performed NeuN western blotting and IHC to detect nuclei of mature neurons in aged mice when massive neuron loss is evident. No significant difference was found in NeuN immunoreactivity (calculated as % of area fraction) in all regions of the hippocampus, dentate gyrus (DG) and neocortex of EDR treated P301L mice compared to controls (Suppl. Figure 2), although data showed a trend towards improvement in EDR treated mice. Western blotting also showed no differences in NeuN protein expression between WT, controls and EDR treated in both age groups (Figs. 6c and 8c). These results might suggest that EDR oral formulation was able to improve synaptic function and, to some extent, inhibit neurotoxicity in aged P301L mice.
3.10. EDR alleviates oxidative stress in P301L mice
4-hydroxynonenal (4-HNE) and 3-nitrotyrosine (3-NT) protein adducts are considered a biomarker of oxidative and nitrosative damage [52, 53]. Results showed that EDR produced small but significant reduction in the levels of 4-HNE (Fig. 6t, t(16) = 6.728, p < 0.0001, 14 months old) and 3-nitrotyrosin (3-NT) (Fig. 8t, t(6) = 5.979, p = 0.001, 25 months old) in the brains P301L mice as evident by western blotting.