Ctnnd2 KO mice exhibited ASD-like behaviors without defects in physical development, vision, or motor coordination.
As our previously investigated [19], Ctnnd2 KO mice exhibited ASD-like behaviors and intelligence disability. Therefore, we conducted a series of behavioral experiments on KO mice again beginning on PND 51 and further found that KO mice showed autism-like behaviors without defects in physical development, vision, or motor coordination (Fig. 1A). Firstly, a three-chamber sociability assay was performed to assess each animal’s ability to interact with social stimulus. Compared to WT mice, KO mice spent less time interacting with stranger 1(p = 0.0396) and more time in center chamber during the first 10 min (p = 0.00503, Fig. 1B). During the second 10 min, related to WT mice, KO mice spent less time interacting with stranger 2 (p = 0.0315, Fig. 1C). Moreover, by self-comparison, we found that WT mice showed more interests in interacting with Stranger 1 than with object during the first 10 minutes (p = 0.0000254), but KO mice showed less difference (p = 0.011, Fig. 1D: left). WT mice showed more interests in interacting with Stranger 2 than with stranger 1 during the second 10 minutes (p = 0.042), but KO mice not (p = 0.533, Fig. 1D: right). When we compared the amount of time the mice spent interacting with Stranger 1 (p = 0.295) or stranger 2 (p = 0.05) during the first 10 minutes with that during the second 10 minutes, we found a significantly difference again (Fig. 1E). The results of three-chamber test were consistent with those of the juvenile play test: KO mice exhibited reduced interest in socializing with unfamiliar mice (p = 0.0161, Fig. 1F). Each animal was individually placed into a dimly lit, novel, open field in order to assess repetitive and exploratory behaviors. KO mice spent more time engaging in repetitive self-grooming than WT mice did (p༜0.0001, Fig. 1G). KO mice also displayed a significant reduction in numbers of grid crossings (p༜0.0001), center grid crossings (p = 0.003), vertical movements (p = 0.00117), and climbing movements (p = 0.00133, Fig. 1H). These results suggest that Ctnnd2 KO mice displayed behavioral features consistent with the characteristics of ASD.
Because the Ctnnd2 gene is related to general neurodevelopment and myopia, we examined the weight, body length, tail length, foot spread distance, and visual ability of each mouse to exclude the influence of changes in each index on ASD-like behaviors. There was no significant difference in any of these indices in WT and KO mice (Supplementary Fig. 2A-F). Additionally, to investigate the motor ability of KO mice, we counted the total number of entries into the three chambers and conducted a rotarod test. KO mice exhibited the similar number of total entries as WT mice did (Supplementary Fig. 2G, H), and there was no difference in the ability to remain on the rotarod over time between WT and KO mice (Supplementary Fig. 2I). These data suggest that lack of Ctnnd2 did not induce abnormal physical development, changes in visual ability, or motor impairment.
Ctnnd2 KO mice experienced sleep-wake disturbances
Then, we collected in vivo, multichannel EEG-EMG recordings to monitor sleep-wake states in freely behaving mice. We also performed analyses of EEG-EMG recording traces and EEG power spectra during wakefulness, NREM and REM sleep to evaluate sleep-wake states and sleep quality. Based on a 24-hour analysis of the sleep-wake states (Fig. 2A), KO mice exhibited a significant increase in the time of wakefulness (24h: p༜0.0001, light: p = 0.000123, dark: p = 0.0271) and significant reductions in NREM (24h: p = 0.008, light: p = 0.0304, dark: p = 0.0557) and REM (24h: p༜0.0001, light: p = 0.000105, dark: p = 0.0557) sleep, relative to WT mice (Fig. 2B). Next, we analyzed the proportion of time spent in each state during each hour over the 24-h recording period, and significant differences in wakefulness (genotype x time of the day: F(23, 336) = 1.463, p = 0.0905; genotype: F(1, 336) = 21.71, p༜0.0001; time of the day: F(23, 336) = 16.01, p༜0.0001), NREM (genotype x time of the day: F(23, 336) = 1.398, p = 0.1075; genotype: F(1, 336) = 11.15, p = 0.0009; time of the day: F(23, 336) = 15.03, p༜0.0001) and REM (genotype x time of the day: F(23, 336) = 1.805, p = 0.0140; genotype: F(1, 336) = 100.2, p༜0.0001); time of the day: F(23, 336) = 10.25, p༜0.0001) sleep profiles were found between WT and KO mice (Fig. 2C). Starting from ZT 5, KO mice exhibited a significantly abnormal distribution of time spent in REM sleep (Fig. 2C: center); therefore, we chose this time point for the following MT intervention. We surmised that the change in the percentage of each sleep-wake state in mice was caused by the number and mean duration of individual sleep-wake cycles (bouts) [41]. Analyses of the number and mean duration of bouts revealed that, compared with WT mice, KO mice exhibited significantly reduced numbers (24h: p༜0.0001, daytime: p༜0.001, nighttime: p = 0.00402) and mean duration (24h: p = 0.021, daytime: p = 0.018, nighttime: p༜0.05) of bouts of REM sleep (Fig. 2D, E) and an increase in the mean duration of wakefulness (24h: p = 0.0463, daytime: p = 0.007, Fig. 2E). However, there were no significant differences in the number or mean duration of episodes of NREM sleep (Fig. 2D, E). Moreover, in order to evaluate the sleep continuity of KO mice, we also counted the number of microarousals. However, there was no significant difference between WT and KO mice during the 24-hour, light period, or dark period (Supplementary Fig. 3A).
It is well known that sleep quality is related to the normalized EEG power spectrum of each state [42]; therefore, we compared the normalized EEG power spectrum (Fig. 2A, D, G) of Wake (genotype x frequency: F(60, 854) = 3.362, p༜0.0001; genotype: F(1, 854) = 3.104e-008, p = 0.9999; frequency: F(60, 854) = 71.79, p༜0.0001, Fig. 3B), NREM (genotype x frequency: F(60, 854) = 1.724, p = 0.0007; genotype: F(1, 854) = 1.879e-009, p > 0.9999; frequency: F(60, 854) = 96.8, p༜0.0001, Fig. 3E) and REM (genotype x frequency: F(60, 854) = 2.839, p༜0.0001; genotype: F(1, 854) = 7.261e-011", p > 0.9999; frequency: F(60, 854) = 90.02, p༜0.0001, Fig. 3H) sleep during the daytime period because KO mice spent more time sleeping during the daytime. KO mice exhibited an increase in the 0.5 to 4 Hz frequency band (delta power) (p = 0.0308) and a decrease in the 4 to 10 Hz frequency band (p = 0.00371) of wakefulness (Fig. 3C). KO mice also exhibited a higher frequency (0.5–4 Hz) (p = 0.0291) and a lower frequency (4–10 Hz) of activity (pronounced theta rhythm) (p = 0.0257) during REM sleep (Fig. 3I). However, there was no significant difference in NREM sleep (Fig. 3F). Collectively, these results suggest that relative to WT mice, KO mice exhibited sleep-wake disturbances characterized by an increase in the time of wakefulness and a decrease in the time of NREM and REM sleep. KO mice also exhibited abnormal sleep quality.
MT ameliorated sleep-wake dysfunction to some extent in Ctnnd2 KO mice
We assessed whether oral administration of MT could correct sleep-wake dysfunction in KO mice. According to the EEG-EMG recordings (Supplementary Fig. 4A), MT administration significantly increased the proportion of time spent in REM sleep for KO mice at all concentrations (24 h: F = 18.17, p༜0.0001; 5 mg: p = 0.0024; 10 mg: p༜0.0001; 50 mg: p = 0.0002 (left); daytime: F = 8.032, p = 0.0005; 5 mg: p = 0.0204; 10 mg: p = 0.0002; 50 mg: p = 0.0023) and nighttime: F = 11.76, p༜0.0001; 10 mg: p༜0.0001; 50 mg: p = 0.0413(right), Fig. 4A). By analyzing the proportion of time spent in each state separately during each hour over the 24-hour recording period, we found that MT administration at the concentration of 10 mg/kg significantly improved sleep-wake disturbances experienced by KO mice at ZT5, 6, 12, 18 and 20. (5mg: genotype x time of the day: F(23, 336) = 1.286, p = 0.1728; genotype: F(1, 336) = 22.24, p༜0.0001; time of the day: F(23, 336) = 10.94, p༜0.0001. 10mg: genotype x time of the day: F(23, 336) = 1.6, p = 0.0413; genotype: F(1, 336) = 56.77, p༜0.0001; time of the day: F(23, 336) = 8.513, p༜0.0001. 50mg: genotype x time of the day: F(23, 336) = 2.764, p༜0.0001; genotype: F(1, 336) = 30.8, p༜0.0001; time of the day: F(23, 336) = 10.26, p༜0.0001, Fig. 4B). However, MT did not affect the time spent in wakefulness or NREM sleep (Supplementary Fig. 5A-D). Additionally, three dosages of MT increased the number of bouts of REM sleep in the 24-h period (24 h: F = 10.07, p༜0.0001; 5 mg: p = 0.0395; 10 mg: p༜0.001; 50 mg: p = 0.0081 (left); daytime: F = 3.664, p = 0.0241; 10 mg: p = 0.0131; 50 mg: p = 0.0407) and nighttime: F = 8.079, p = 0.0005; 10 mg: p = 0.0002 (right), Fig. 4C), whereas 5 mg/kg MT decreased the mean duration of wakefulness bouts during light period (F = 3.572, p = 0.0264; 5 mg: p = 0.0133, Fig. 4D). We also observed an increase in the mean duration of REM sleep bouts after MT administration (24 h: F = 7.626, p = 0.0007; 5 mg: p = 0.0089; 10 mg: p = 0.0005; 50 mg: p = 0.0018 (left), daytime: F = 4.279, p = 0.0132; 10 mg: p = 0.0064; 50 mg: p = 0.0289 and nighttime: F = 4.043, p = 0.0166; 5 mg: p = 0.0089; 50 mg: p = 0.034 (right), Fig. 4E). There were no changes in the number of wakefulness bouts (Supplementary Fig. 6A) and in the number or mean duration of NREM sleep bouts (Supplementary Fig. 6B, C). At the same time, the number of microarousals was also no significantly difference between KO and KO treated with three doses of MT groups (Supplementary Fig. 6D).
Finally, MT administration at all concentrations had no significant effect on the normalized EEG power spectrum in sleep-wake states (Supplementary Fig. 7A-F). These results suggest that MT attenuated sleep dysfunction in KO mice by promoting sleep, especially REM sleep, rather than by influencing sleep quality.
MT improved core symptoms of ASD in Ctnnd2 KO mice
We assessed changes in ASD-like behaviors after the oral administration of MT (Fig. 5A). During the three-chamber sociability assay, the oral administration of three dosages of MT led to an increase in the time spent interacting with Stranger 1 during the first 10 minutes (F = 3.543, p = 0.0272; 5 mg: p = 0.0279; 10 mg: p = 0.0446; 50 mg: p = 0.0315 (left), Fig. 5B), and the 10 mg/kg MT group exhibited a similar result in the time spent interacting with Stranger 2 during the second 10 minutes (F = 3.746, p = 0.0222; 10 mg: p = 0.0084 (right), Fig. 5B). Moreover, by self-comparison, 5, 10 and 50 mg/kg groups showed more interests in interacting with Stranger 1 than with object during the first 10 minutes (5mg: p༜0.001, 10mg: p = 0.0000154, 50mg: p = 0.00349 (left), Fig. 5B), and only 10 mg/kg group showed more interests in interacting with Stranger 2 than with stranger 1 during the second 10 minutes (10mg: p = 0.000633 (right), Fig. 5B). When we compared the time spent by mice interacting with Stranger 1 or stranger 2 during the first 10 minutes with that during the second 10 minutes, we found no significantly difference between KO mice and mice treated with MT (Supplementary Fig. 8A). During the direct social interaction test, the mice treated with 5 and 10 mg/kg MT displayed more interest in interacting with the stranger (F = 3.711, p = 0.023; 5 mg: p = 0.0357; 10 mg: p = 0.0151, Fig. 5C). Additionally, MT treatment reduced the time spent engaging in self-grooming at all concentrations (F = 77.8, p༜0.0001; 5, 10 and 50 mg: p༜0.0001, Fig. 5D), although only treatment at 10 mg/kg increased both the numbers of grid (F = 3.781, p = 0.0214; 10 mg: p = 0.0148) and center grid crossings (F = 3.124, p = 0.0416; 10 mg: p = 0.0173) during the open-field test (Fig. 5E). These findings indicate that oral administration of MT improved ASD-like behaviors, including social interaction ability, restrictive behavior and exploration capability.
MT improved core symptoms of ASD in Ctnnd2 KO mice by ameliorating sleep-wake dysfunction
Some researchers have reported that poor sleep exacerbates undesired daytime behaviors experienced with autism [9]. On the other hand, in addition to improving sleep, MT may exert neuroprotective effects via various mechanisms. Thus, we aimed to verify the correlation between sleep and core symptoms of ASD in KO mice. We treated KO mice with 10 mg/kg MT, an effective dose according to our above results, and simultaneously induced SR for 28 consecutive days to examine whether SR could reverse the effects of MT (Fig. 6A). Interestingly, during the three-chamber test, compared with the 10 mg/kg MT group, the 10 mg/kg MT with SR group exhibited reduced interaction with Stranger 1 during the first 10 minutes (F = 10.77, p = 0.0004; 10 mg + SR: p = 0.0002; self-comparison: 10mg: p༜0.0001(left), Fig. 6B) and reduced interaction with Stranger 2 during the second 10 minutes (F = 19.66, p༜0.0001; 10 mg + SR: p༜0.0001; self-comparison: 10mg: p༜0.001 (right), Fig. 6B). During the direct social interaction test, the 10 mg/kg MT with SR group exhibited reduced interaction with the stranger mouse compared to the 10 mg/kg MT group (F = 6.918, p = 0.0041; 10 mg + SR: p = 0.0031, Fig. 6C). Compared with KO mice treated with 10 mg/kg MT, KO mice treated with 10 mg/kg MT and subjected to SR spent more time engaging in self-grooming activities (F = 72.55, p༜0.0001; 10 mg + SR: p༜0.0001, Fig. 6D). During the open-field test, mice in the 10 mg/kg group subjected to SD exhibited decreased numbers of grid crossings (F = 6.108, p = 0.0069; 10 mg + SR: p = 0.0044) and center grid crossings (F = 6.055, p = 0.0072; 10 mg + SR: p = 0.0049) when compared to those received 10 mg/kg MT without SR (Fig. 6E). Collectively, these results indicate that MT improved ASD-like behaviors by attenuating sleep-wake dysfunction of KO mice.