Low-dose RT reduced pro-inflammatory cytokine expression and decreased microgliosis in the hippocampus of 5xFAD mice
To determine the effects of low-dose RT on microgliosis the expression of Iba-1, a marker for reactive microglia, was evaluated in the hippocampus from wild type (WT) and 5xFAD mice. The schematic diagram of the study is briefly shown in Fig. 1. As shown in Fig. 2a, 5xFAD-Sham mice showed significantly increased protein levels of Iba-1 compared with that in WT-sham mice (5xFAD-Sham, 2.32 ± 0.16, n = 9; WT-Sham, 1.00 ± 0.15, n = 8). The enhanced Iba-1 expression in the hippocampus was significantly decreased by LMD-CDRT (1.62 ± 0.19, n = 8), and 5xFAD-LD-LDRT mice (2.15 ± 0.18, n = 8) showed only a decreasing tendency compared to 5xFAD-sham mice (Fig. 2a), but statistical significance was not observed.
To determine whether low-dose RT affects the polarization of microglia, the mRNA expression levels of pro-inflammatory and anti-inflammatory cytokines were analyzed. 5xFAD-sham mice showed a significant increase in IL-6 (2.83 ± 0.70), CCL6 (3.95 ± 0.54), and IL-1β (1.49 ± 0.10) mRNA expression compared with that in WT-sham mice (Fig. 2b-e, IL-6, 1.00 ± 0.08; CCL6, 1.00 ± 0.18; C IL-1β, 1.00 ± 0.16). The elevated IL-6 (1.30 ± 0.19) and CCL6 (2.78 ± 0.37) mRNA expression was significantly attenuated in 5xFAD-LD-LDRT (Fig 2b, c). LMD-CDRT also ameliorated the mRNA expression of CCL6 (1.10 ± 0.21) and IL-1β (0.36 ± 0.10) in 5xFAD mice (Fig. 2c, d). However, IL-4 mRNA expression was not significantly different among all mice groups (Fig. 2e,). We also determined the cytokine levels using a proteome profiler array. Among 40 cytokines, we identified 8 cytokines that were significantly increased (CD54, 1.97 ± 0.05; CCL2, 1.63 ± 0.29; CCL4, 1.46 ± 0.21) or showed an increasing tendency (IL-3, 1.24 ± 0.09; IL-16, 1.32 ± 0.18; CXCL9, 1.09 ± 0.13; CXCL10, 1.24 ± 0.17; and CXCL11, 1.25 ± 0.15) in 5xFAD-Sham relative to WT-sham (Fig 3b). After LMD-CDRT, the release of pro-inflammatory cytokines such as IL-3/16 (IL-3, 0.91± 0.09; IL-16, 0.79 ± 0.10), CCL2/4 (CCL2, 1.00 ± 0.09; CCL4, 0.58 ± 0.10), and CXCL9 (1.00 ± 0.09) in 5xFAD mice was significantly reduced (Fig. 3a, b). More interestingly, LD-LDRT diminished the production of inflammatory cytokines such as CD54 (1.35 ± 0.25), IL-3 (1.24 ± 0.09), CXCL9/10 (CXCL9, 0.73 ± 0.14; CXCL10, 0.90± 0.14), CCL2/4 (CCL2, 0.94± 0.09; and CCL4, 0.85 ± 0.13) in 5xFAD mice (Fig. 3a, b). Overall, these results indicate that LD-LDRT and LMD-CDRT attenuated microgliosis and reduced the level of pro-inflammatory cytokines, but not anti-inflammatory cytokines.
Low-dose RT reduced mature amyloid plaques, but did not affect the number of amyloid plaques in the hippocampi of 5xFAD mice
Because low-dose RT reduced the release of pro-inflammatory cytokines in 5xFAD mice, we speculated that low-dose RT may attenuate accumulation of amyloid plaques by inhibiting microglial activation. The number and size of amyloid plaques was measured after 6E10 staining to identify the effect of low-dose RT on amyloid plaque burden (Fig. 4a). The number of amyloid plaques was not significantly different among all mice groups (Fig. 4b). However, the area of the amyloid plaques in the hippocampus of 5xFAD mice was significantly decreased by LD-LDRT and LMD-CDRT (Fig. 4c; Sham, 0.0234 ± 1.20e-3; LD-LDRT, 0.0199 ± 9.21e-4; LMD-CDRT, 0.0196 ± 1.0425e-3). Amyloid plaques have been previously classified into different types by area (21). Therefore, we classified the amyloid plaques in the hippocampus based on their areas size and found that the number of amyloid plaques larger than 100 nm2 were significantly reduced by approximately 12% with LD-LDRT and LMD-CDRT (Fig. 4d; Sham, 30.85 ± 1.82; LD-LDRT, 27.22 ± 0.60; LMD-CDRT 25.64 ± 0.39). These data demonstrate that LD-LDRT and LMD-CDRT diminished the mature amyloid plaques in the hippocampus of 5xFAD mice.
Low-dose RT rescued the cognitive impairments in 5xFAD mice
Accumulation of amyloid plaques, a major neuropathological characteristic, together with neurofibrillary tangles, contributes to cognitive impairment, which is one of the major symptoms in AD (22, 23). To confirm the impact of RT on cognitive impairment, a spontaneous Y-maze test was performed before and after low-dose RT, to assess cognitive function and to determine a spontaneous ratio indicating spatial memory (Fig. 5a). Before low-dose RT, the spontaneous alternation ratio in 5xFAD-Sham (0.60 ± 0.02), LD-LDRT (0.54 ± 0.03), and LMD-CDRT (0.51 ± 0.18), mice were compared to that in WT-Sham (Fig. 5b, 0.70 ± 0.03). Surprisingly, cognitive impairment was recovered by LD-LDRT (1st week, 0.62 ± 0.05; 2nd week, 0.60 ± 0.02, 3rd week, 0.64 ± 0.06; 4th week, 0.60 ± 0.04) and LMD-CDRT (1st week, 0.62 ± 0.05; 2nd week, 0.60 ± 0.02, 3rd week, 0.58 ± 0.03; 4th week, 0.55 ± 0.03) and its effect lasted for 4 weeks (Fig. 5c). Both body and brain weight showed no significant differences between Sham and RT group mice (Fig. 5d and 5e). These results suggest that LD-LDRT and LMD-CDRT attenuates cognitive impairment in AD without significant adverse effects.
Low-dose RT did not affect neuronal death or synaptic plasticity-related proteins, but reduced reactive astrocytes in 5xFAD mice
Neuronal cell death and impaired synaptic plasticity have been associated with the accumulation of amyloid plaques or cognitive impairment in AD (24, 25). To examine whether low-dose RT could protect neuronal cell death or change synaptic plasticity by decreasing amyloid plaque burden, we analyzed neuronal loss and synaptic density in the hippocampus by examining the protein level of NeuN and PSD-95, which are markers for neurons and postsynapses, respectively. The protein level of NeuN (1.02 ± 0.08) and PSD-95 (0.43 ± 0.08) was significantly reduced in 5xFAD-Sham mice. The expression of both NeuN (LD-LDRT, 0.93 ± 0.04; LMD-CDRT 0.85 ± 0.09) and PSD-95 (LD-LDRT, 0.57 ± 0.10; LMD-CDRT 0.38 ± 0.08) was not affected by low-dose RT (Fig. 6a). Taken together, low-dose RT did not affect neuronal death or impaired synaptic plasticity even though it reduced the amyloid plaque burden. Next, we examined the astrocyte reactivity after low-dose RT in 5xFAD mice. Reactive astrocytes that are close to amyloid plaques are associated with the development and progression of AD (26). The increase in GFAP expression observed in 5xFAD-sham mice (1.92 ± 0.21) was also attenuated in 5xFAD-LD-LDRT (0.95± 0.17), but not in 5xFAD-LMD-CDRT (Fig. 6b).