Adiponectin improves amyloid‐β 31‐35‐induced circadian rhythm disorder in mice

Abstract Adiponectin is an adipocyte‐derived hormone, which is closely associated with the development of Alzheimer's disease (AD) and has potential preventive and therapeutic significance. In the present study, we explored the relationship between adiponectin and circadian rhythm disorder in AD, the effect of adiponectin on the abnormal expression of Bmal1 mRNA/protein induced by amyloid‐β protein 31‐35 (Aβ31‐35), and the underlying mechanism of action. We found that adiponectin‐knockout mice exhibited amyloid‐β deposition, circadian rhythm disorders and abnormal expression of Bmal1. Adiponectin ameliorated the abnormal expression of the Bmal1 mRNA/protein caused by Aβ31‐35 by inhibiting the activity of glycogen synthase kinase 3β (GSK3β). These results suggest that adiponectin deficiency could induce circadian rhythm disorders and abnormal expression of the Bmal1 mRNA/protein, whilst exogenous administration of adiponectin may improve Aβ31‐35‐induced abnormal expression of Bmal1 by inhibiting the activity of GSK3β, thus providing a novel idea for the treatment of AD.


| INTRODUC TI ON
Since the 21st century, Alzheimer's disease (AD) has become a serious health problem that affects ageing populations worldwide.
Circadian rhythm disorder occurs in the early stage of AD, and this can induce impairment of learning and memory in AD. 1,2 Therefore, circadian rhythm disorders play a vital role in the development of AD. Further studies have found that circadian rhythm disorders are closely related to the extracellular aggregates of amyloidβ protein (Aβ) in the brain, which plays a causative role in AD pathogenesis. 3 Our previous study found that the intrahippocampal injection of amyloidβ protein 31-35 (Aβ31-35) resulted in circadian rhythm disorder in C57BL/6 mice. 4 Circadian rhythms are rhythmic oscillations that spontaneously form in organisms. Their maintenance depends on the transcriptional-translational feedback loop composed of a series of clock genes and proteins, amongst that Bmal1 is an important positive regulator. 5 Studies have found that Bmal1 −/− mice lose circadian rhythmicity at the behavioural and molecular levels, 6 and the tripletransgenic AD mouse model exhibits Aβ deposition in the brain and abnormal expression of Bmal1. 7 Our previous study found that Aβ31-35 induces abnormal expression of Bmal1 mRNA/protein in HT22 cells. 8 However, there is still no effective measure to reverse the Aβ31-35-induced abnormal expression of Bmal1, and the underlying mechanism is not yet clear.
Studies have shown that AD is closely related to type 2 diabetes mellitus (T2DM) in pathogenesis. Insulin resistance is a common pathophysiological characteristic of these two diseases and plays a significant role in the development of AD. Adiponectin (APN) is an adipocytokine secreted mainly by adipocytes, with insulinsensitizing effects via the activation of insulin signalling pathways. 9 Clinical studies have shown that circulating adiponectin levels are decreased in patients with mild cognitive impairment and AD. 10 Recent studies have also found that APN signal transduction defects are sufficient to induce AD-like phenotypes in mice, including Aβ deposition, tau protein hyperphosphorylation, synaptic loss and neuronal apoptosis. 11,12 APN can enhance insulin sensitivity in SH-SY5Y cells by activating AdipoR1 and APN signalling to alleviate neuropathological deficits and clinical manifestations in APP/PS1 mice, such as Aβ aggregation, synapse dysfunction, memory and cognitive deficits. 11,13 These results reveal that APN is closely associated with the development of AD and has potential preventive and therapeutic significance for AD. In contrast, the role of APN in AD circadian rhythm disorder remains uncertain, and whether APN can improve the abnormal expression of Bmal1 mRNA/protein caused by Aβ31- 35 has not been documented.
Adiponectin has also been reported to regulate insulin sensitivity to activate insulin signalling and inhibit glycogen synthase kinase 3β (GSK3β) activity. 11 GSK3β is a serine-threonine kinase involved in the regulation of circadian rhythm, 14 and it is closely related to the regulation of Bmal1. Studies have shown that GSK3β can directly phosphorylate and degrade BMAL1. 15 Studies have revealed that abnormal deposition of Aβ can increase the activity of GSK3β. 16 However, whether GSK3β activation induced by Aβ affects the expression of Bmal1 mRNA/protein and whether APN can improve the abnormal expression of Bmal1 induced by Aβ31-35 by inhibiting the activity of GSK3β are still unclear. This study explored the relationship between APN and circadian rhythm disorder in AD and the effect of APN on Aβ31-35-induced abnormal expression of Bmal1 mRNA/protein and its possible mechanism.
2 | MATERIAL S AND ME THODS

| Experimental animals
All experimental procedures were approved by the Ethics Committee

| Polymerase chain reaction
Polymerase chain reaction (PCR) was used to confirm the genotype of the experimental mice. Tail tissues (approximately 0.5 cm) were digested overnight in 500 μl lysis buffer containing 5 μl proteinase K. The digested tissue was added to 500 μl of phenol/chloroform mixed solution (equal volume) and centrifuged at 4°C and 13523g  bands were 326 bp for wild-type mice and 531 bp for homozygous APN-KO, and PCR products of heterozygous mice had two DNA bands located at 531 bp and 326 bp ( Figure S1).

| HT22 cell culture
The mouse hippocampal nerve cell line (HT22) were purchased from Guangzhou Jennio Biotechnology Co., Ltd. The HT22 cells were cultured in Dulbecco's Modified Eagle Medium complete medium (HyClone) supplemented with 10% foetal bovine serum (FBS; Sciencell) and were kept in a constant-temperature incubator at 37°C and 5% CO 2 . After adhesion, cells with 80% density and adequate growth conditions were selected for the synchronization treatment in each group. The complete medium for culturing cells was replaced with a starvation medium supplemented with 1% FBS.

| Immunohistochemical staining
The full-length Aβ1-42 is more neurotoxic and immunohistochemical staining was used to detect Aβ1-42 deposition in 12-month-old APN-KO mice and C57BL/6 mice. The brain tissue located 4 mm be-

| Wheel-running behavioural test
The circadian rhythm of each group of male mice (n = 6) was evaluated using a wheel-running behavioural test. Power analysis was performed to evaluate the sample size for behavioural animal experiments. The 4-month-old APN-KO mice and C57BL/6 mice were placed in a wellventilated running wheel device at a temperature of 22 ± 2°C and humidity of 35%-55%. The lighting environment was set to 12 h of light and 12 h of darkness (Light-Dark, LD) for 1 week; that is, the lights were turned on at 6:00 and turned off at 18:00. Then, the environment was changed to constant darkness for 2 weeks. Due to the lack of light, the endogenous biological rhythms of the animals were represented by circadian time (CT). 19 The length of each circadian cycle was divided into 24 equal parts, and each part was 1 CT. The time at which the mice started their daily activities was defined as CT12. 4 The running wheel activity was recorded using the VitalView system at a frequency of every 5 min. The running wheel data were analysed using ActiView software, accompanied by the acquisition of the original map of the wheel-running activity, free-running cycle and day and night activities. Upon the termination of wheel-running, the mice were decapitated at CT4, CT8, CT12, CT16, CT20 and CT24, and the hippocampal tissue was peeled off on ice to further detect the expression of Bmal1 mRNA/protein ( Figure 1A).

| Real-time PCR
Bmal1 mRNA expression levels were detected using realtime PCR at different CT points. Total RNA from mouse hippocampus and HT22 cells was extracted by the Trizol method and reverse transcribed to cDNA and then specifically amplified using the SYBR Green kit. The corresponding primer design was as follows: Bmal1 (Gen-Bank ID NM_001243048.1), forward: 5′-ACGACATAGGACACCTCGCAGA-3′, reverse: 5′-CAACAATCTCCACTTTGCCACTG-3′. All data were standardized with the expression of GAPDH at CT4 in the control group, and the target gene mRNA was quantified using the 2 −ΔΔCt method.

| Western blotting
The expression of BMAL1 protein was detected by western blotting. The mouse hippocampal and HT22 cells were lysed on ice for 1.5 h with RIPA lysis buffer, and the supernatant was extracted after centrifugation at 13523g for 15 min. The protein concentration was quantified using the bicinchoninic acid (BCA) method, and the protein content was quantified as 40 μg. The protein was completely denatured after adding 5 × loading buffer and heating at 100°C for 10 min. The protein samples were subjected to SDS-PAGE and then transferred to a polyvinylidene fluoride membrane. The membrane was blocked with 5% skimmed milk at room temperature for 2 h and then incubated with primary antibodies against BMAL1 (Abcam), pGSK3β S9 (BBI), GSK3β (BBI), β-actin and GAPDH overnight at 4°C.
After washing with Tris Buffered Saline with Tween (TBST), the membrane was incubated with the corresponding secondary antibody for 2 h at room temperature, followed by washing with TBST again. Images were exposed and captured using a gel imaging system. Western blot data were quantified using ImageJ software.

| Statistical analysis
Statistical analysis was performed using SPSS software (version 16.0). The normal distribution of measured data was presented as group mean ± standard error of mean. Statistical analyses were performed using one-way analysis of variance for multiple group comparisons and a least significant difference t test for comparison between groups. The results were presented as α = 0.05, and statistical significance was set at p < 0.05.

| Adiponectin-knockout mice exhibited Aβ deposition and circadian rhythm disorders
To explore the correlation between APN and AD, we first used immunohistochemical staining to detect Aβ1-42 deposition in the hippocampus of 12-month-old C57BL/6 mice and APN-KO mice.
An aggregation of brownish-yellow granules was observed in the hippocampus of APN-KO mice when compared to C57BL/6 mice ( Figure 2), suggesting that APN deficiency could induce abnormal Aβ deposition in the hippocampus.
Subsequently, we selected 4-month-old C57BL/6 mice and APN-KO mice to conduct wheel-running experiments to clarify the effect of APN deficiency on circadian rhythm. The results showed that the mice in the control group displayed rhythmic wheel-running activity with clearly demarcated movement and rest phases. The activities mainly occurred during subjective nights, and the starting time was relatively fixed ( Figure 3A). The ratio of subjective daytime activity to total activity was 27.34 ± 9.36%, and the free-running period was 23.04 ± 0.39 h ( Figure 3B,C). Conversely, APN-KO mice displayed circadian rhythm disorder that was manifested by changes in the starting time of daily activities, increased subjective daytime activities and reduced subjective night activities ( Figure 3A). The ratio of subjective daytime activity to total activity increased significantly ( Figure 3B), and the free-running period was prolonged (p < 0.05) ( Figure 3C). Collectively, these results showed that APN deficiency could induce circadian rhythm disorders in C57BL/6 mice.

| Abnormal expression of Bmal1 mRNA/ protein in the hippocampus of APN-KO mice
To explore the effect of APN deficiency on the expression of the showed that the expression of Bmal1 mRNA in the control group was relatively high at CT4, CT12, CT20 and CT24, with a peak at CT20, whilst the expression was relatively low at CT8 and CT16, with a trough at CT8. The rhythmic expression of Bmal1 mRNA in APN-KO mice was abnormal, showing relatively high expression at CT4, CT8, CT20 and CT24, with a peak at CT24, and relatively low expression at CT12 and CT16, with a trough at CT12. The Bmal1 mRNA expression level at CT12 was significantly lower than that at CT12 in the control group (p < 0.05) ( Figure 4A,B).
We then examined the expression of the BMAL1 protein. The data showed that the expression level of BMAL1 protein was the highest at CT24 in the control group, whilst it was the highest at CT4 in the APN-KO group. Compared with that in the control group, the expression of BMAL1 protein in APN-KO mice was significantly increased, which was statistically significant at CT4, CT12, CT16 and CT20 (p < 0.05) ( Figure 4C-E). These results suggest that APN deficiency could induce abnormal expression of the Bmal1 mRNA/protein in the hippocampus.

Aβ31-35 in HT22 hippocampal neurons cells
To explore whether APN could improve the abnormal expression of

| DISCUSS ION
In the present study, APN-KO mice showed Aβ deposition, circadian rhythm disturbance, and abnormal expression of the Bmal1 mRNA/ Several studies have found a potential relationship between AD and T2DM. In T2DM patients, the grey matter content of the frontotemporal area and the volume of the hippocampus decreased.
The risk of cognitive impairment and development of AD in T2DM patients is 1.5 to 2 times higher than that in patients without T2DM. 30,31 Meanwhile, approximately 80% of the AD patients have T2DM or impaired glucose tolerance. 32 Insulin resistance is a common pathophysiological feature of AD and T2DM. Insulin resistance is a reduced sensitivity of body tissues to insulin, which is one of the earliest and most significant metabolic defects in T2DM. 33 Similar insulin response defects have also been observed in AD patients and animal models. 31,34 In addition, neuronal insulin signalling is closely associated with Aβ deposition, tau protein phosphorylation, synaptic plasticity and memory function. 35 When the circadian rhythm is disturbed, the rhythm indicators such as periodicity, phase and amplitude will alter, leading to significant changes in Bmal1 mRNA at some CTs, but no significant difference at some CTs. We hypothesized that the Bmal1 mRNA expression abnormalities at different CTs appear to be selective after Aβ31-35 treatment. The expression of the BMAL1 protein at CT20 was consistent with that at the gene level. We found that APN can reverse the abnormal expression of the Bmal1 mRNA/protein induced by  Increasing evidence has suggested that GSK3β plays an important role in the maintenance of circadian rhythm, and the change in its activity is closely related to the regulation of Bmal1. A key feature of GSK3β is that it is active in its default state and that it is inactivated by phosphorylation Ser-9 for GSK3β (pGSK3β S9 ). The ratio of pGSK3β S9 to GSK3β is an important indicator of GSK3β activity.
When the level of pGSK3β S9 decreased and the ratio of pGSK3β S9 to GSK3β decreased, the activity of GSK3β increased; otherwise, the activity was inhibited. 51,52 The increase in GSK3β activity can destroy the circadian rhythm of BMAL1 protein expression, 51 whilst the inhibition of GSK3β activity can enhance the stability of BMAL1 protein and increase its expression level. 15 In addition, GSK3β phosphorylates and stabilizes the orphan nuclear receptor REV-ERBα, a negative component of the circadian clock. Inhibition of GSK3β activity leads to the degradation of REV-ERBα and activates Bmal1 transcription. 53 Therefore, GSK3β is critical for rhythmic Bmal1 expression. Another study has shown that Aβ can induce an increase in GSK3β activity. 16 In this study, we also found that GSK3β activity increased significantly after treatment with 5 μmol/L Aβ31-35 in HT22 cells. Next, we investigated whether GSK3β activated by which provides a novel approach for the treatment of AD.

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

AUTH O R CO NTR I B UTI O N S
Yuan Yuan: Investigation (equal); Writing-original draft (equal).

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.