Effects on Autophagy of Moxibustion at Governor Vessel Acupoints in APP/PS1double-Transgenic Alzheimer's Disease Mice through the lncRNA Six3os1/miR-511-3p/AKT3 Molecular Axis

Objective To explore the effect and mechanism of moxibustion at acupoints of the governor vessel on lncRNA Six3os1 in amyloid precursor protein/presenilin1 (APP/PS1) double-transgenic Alzheimer's disease (AD) mice. Methods Twenty-four specific pathogen-free and APP/PS1 double-transgenic male mice were randomly allocated into the AD model and moxibustion groups, with 12 cases in each group. Twelve syngeneic C57BL/6J mice were selected as the control group. Mice in the moxibustion group received aconite cake-separated moxibustion at the Baihui acupoint. Suspension moxibustion was applied at Fengfu and Dazhui for 15 minutes each day. All treatments were conducted over two weeks. Control and AD model mice were routinely fed without any intervention. Behavioral observation tests were conducted before and after the intervention. The autophagosome in the hippocampus was observed using transmission electron microscopy. Immunohistochemistry was performed to detect Aβ1-42 expression. LC3B and P62 expressions were evaluated by immunofluorescence. The expression levels of the lncRNAs Six3os1, miR-511-3p, and AKT3 were detected by qRT-PCR. The differential expression of PI-3K, AKT3, mTOR, LC3B-II/I, and P62 proteins in the hippocampus was detected by western blot. The dual-luciferase assay was undertaken to examine the targeting relationships of the lncRNAs Six3os1, miR-511-3p, and AKT3. Results Compared with the control group, the AD model showed higher escape latency in the Morris Water Maze and reduced autophagic vacuoles in the cytoplasm of hippocampal neurons (both p < 0.01). Compared with the control group, the AD model showed higher expression of Aβ1-42, the lncRNAs Six3os1, PI-3K, mTOR, P62, and AKT3 protein (all p < 0.01); but lower mir-511-3p and LC3B (both p < 0.01). Compared with the AD model group, the moxibustion group had a shorter escape latency, more autophagic bubbles in the hippocampus, and lower expression of positive Aβ1-42, the lncRNAs Six3os1, PI-3K, mTOR, P62, and AKT3 protein (all p < 0.01). In contrast, the levels of miR-511-3p and LC3B proteins were considerably increased in the moxibustion group compared to the AD model group (both p < 0.01). Based on the dual-luciferase assay, there was a targeting link among the lncRNAs Six3os1, miR-511-3p, and AKT3. Conclusion Moxibustion at acupoints of the governor vessel can suppress the lncRNA Six3os1 expression, promote cell autophagy, accelerate Aβ1-42 clearance and alleviate cognitive dysfunction of AD mediated by the PI3K/AKT/mTOR signaling pathway through the lncRNA Six3os1/miR-511-3p/AKT3 axis.


Introduction
Alzheimer's disease (AD) is a multifactorial and irreversible neurodegenerative disease that accounts for 50-70% of dementia cases [1]. Pharmacological therapies used to treat AD can relieve symptoms but do not reverse disease progression [2]. e failure of a range of clinical agents has led some to question the amyloid-β (Aβ) pathophysiological hypothesis of AD [3]. However, anti-Aβ drugs such as aducanumab and ALZ-801 have shown encouraging outcomes in phase 3 trials and have confirmed amyloid as a viable therapeutic target [4,5]. Studies have established that dysfunctional autophagy is involved in neurodegenerative disease and that its induction can accelerate the clearance of abnormally accumulated Aβ, thereby improving cognitive function in AD [6]. Furthermore, an increasing number of studies have revealed that autophagy is implicated in the etiology and progression of AD [7,8]. erefore, the identification of key regulators of autophagy is essential for AD treatment.
LncRNAs are noncoding RNAs > 200 nucleotides in length that can be widely distributed in the nucleus; they have emerged as critical regulators of numerous basic biological activities [9]. Studies have shown that lncRNAs can be represented as molecular sponges to target miRNAs and influence cell autophagy directly [10]. Xu et al. found that overexpression of the lncRNA H19 affects the normal activity of the PI3K/Akt/mTOR signaling pathway, impairing impair cell survival and increasing cell autophagy.
Our earlier study found that moxibustion effectively treated AD by improving patient cognitive function and daily tasks. Underlying these clinical effects, moxibustion may act by inhibiting PI3K/AKT/mTOR and P38 MAPK signaling pathways to enhance cell autophagy and accelerate Aβ clearance [11][12][13]. We examined gene expression in the hippocampal tissue of APP/PS1 double-transgenic mice by utilizing high-throughput sequencing technology. We aimed to determine the differential expression of lncRNAs and mRNAs before and after the moxibustion treatment. Functional enrichment analyses were also performed to enrich the PI3K/AKT/mTOR signaling pathway to construct a ceRNA regulation network and eventually screen out the crucial lncRNA Six3os1. Next, we evaluated the specific effects and mechanisms of moxibustion in modulating the PI3K/AKT/mTOR signaling pathway through lncRNA Six3os1 to provide reliable molecular markers and targets for the clinical diagnosis and treatment of AD.

Animal Models.
Six-month-old APP/PS1 double transgenic and specific pathogen-free male AD mice were provided by Nanjing Junke Bioengineering Co., Ltd.
[License: SCXK (Su) 2020-0009]. e average body weight was 28 ± 2 g. e animals were raised in a clean animal room at the Science Experimental Center of the Anhui University of Chinese Medicine. Each animal was kept in a separate cage clarified at 23 ± 2°C, 50 ± 5% humidity, and in a 12-hour light-dark cycle. Approval for the study was provided by the Anhui University of Chinese Medicine (NO: AHUCMmouse-2021042).
e Morris water maze spatial learning test was performed one week following adaptive feeding; animals who did not swim or exhibited significant differences in test scores from other mice were excluded from the experiment. e remaining 24 animals were randomly but evenly divided into the AD model and moxibustion groups.

Experimental Intervention.
e acupoints Baihui, Dazhui, and Fengfu were selected according to the Nomenclature and Location of Acupuncture Points for Laboratory Animals [14]. Animals were shaved at the intervention site and acupoints were marked. Several aconite cakes with a diameter of ∼1 cm and a thickness of 4-6 mm were prepared in advance and multiple small holes were made on their surface with a toothpick. Mice were secured in a restraint with the head and neck exposed. In the moxibustion-treated group, the moxibustion bar was lit and the lit end was placed on the aconite cake, which was then positioned at Baihui for 15 minutes every day. Dazhui and Fengfu were treated with suspension moxibustion for 15 minutes per day at a distance of 2-3 cm from the skin. To control for radiant heat, one end of a temperature probe was affixed next to the Fengfu acupoint and the moxibustion temperature was maintained at 44-46°C (the same protocol was followed at Dazhui acupoints).
e control and AD model groups were routinely housed under comparable conditions but received no intervention. All mouse groups were treated once daily for two weeks. Trained professionals from the Anhui Acupuncture Hospital performed the procedures.

Morris Water Maze Test of Spatial Localization.
e water maze test was performed before and after the intervention. e laboratory was sheltered from light, and the room temperature was maintained at 24-25°C with the water temperature at 24 ± 2°C. Mice were continuously trained for four days before data collection. On the first day, the mice were placed in the water for two minutes to adapt to the surroundings. Each mouse received two training sessions, separated by 4 hours, per day from the second day. Each time, they were introduced to the water from a different point. e latency time (from placement until arrival at the platform) was recorded. If the mouse did not find the platform within 2 minutes, the latency was recorded as 2 minutes. e mice were permitted to stay on the platform for 30 seconds whether or not they found the platform within 2 min. Following the training, the mice were tested to establish a baseline and again after the treatment.

Hippocampal Autophagy.
Brains were harvested from six mice and the hippocampi were isolated; harvesting was performed on an ice plate. e hippocampal CA1 region was excised and stored in an electron microscopy solution for subsequent electron microscopy analysis; the remaining hippocampal tissues were preserved at −80°C for gene and protein quantification. Several isolated fresh sections (∼1 mm 3 ) of the hippocampal CA1 region were rinsed, fixed, dehydrated, and embedded for ultrathin sectioning (60 nm thickness). ree sections were taken from the mice of each group. e morphology of neuronal cell structures, autophagic vacuoles, autolysosomes, and lysosomes was examined and photographed using the JEM1400 transmission electron microscope after double staining (lead and uranium staining). ree to five fields were randomly selected from each section for analysis.

Aβ1-42 Protein Expression.
ree mice from each group were randomly selected and anesthetized by intraperitoneal injection with 0.3% pentobarbital sodium (30 mg/kg). e brains were fixed after perfusion with 4% paraformaldehyde. Sections of 4 μm thickness were made from prepared paraffin blocks, then deparaffinized and hydrated for antigen retrieval. e sections were then rinsed ×3 in phosphate-buffered saline ×5 min, blocked, and incubated in Aβ (1 : 1000) primary antibody overnight at 4°C. e secondary antibody (1 : 5000) was added dropwise to the sections for incubation at 37°C for 40 minutes before rinsing. DAB was added dropwise to the sections and the chromogenic time was adjusted under the microscope. e sections were washed following appropriate color development. e sections were rewashed after hematoxylin counterstaining for two to five minutes, after which the sections were washed after blueing with lithium carbonate solution for 30 seconds. After dehydration and xylene-induced tissue transparency, neutral gum was then added and the sections were coverslipped for imaging. e images were observed and captured under a high-power (×400) microscope to determine the mean absorbance values of the positive staining.

Hippocampal Expression of lncRNAs Six3os1
, miR-511-3p, and AKT3 in APP/PS1 Mice. -80°C frozen hippocampal tissues weighing 50-60 mg were weighed, chopped, and total RNA was extracted using TRIzol-chloroform-isopropanolethanol. According to the instructions, RNA was reverse transcribed into cDNA using a PrimeScript ™ RT reagent Kit with gDNA Eraser (TaKaRa). e PCR reaction system consisted of 5 μL of 2 × SYBR Green mixture, 1 μL of each upstream and downstream primers, 1 μL cDNA, and 2 μL of nuclease (10 μL of the final mixture). e PCR parameters for lncRNAs Six3os1, miR-511-3p, and AKT3 were as follows: 95°C for 1 minute, 95°C for 20 seconds, and 60°C for 1 minute. e fluorescence signals were acquired over 40 cycles with three technical replicates per sample. β-actin was used as the internal reference gene and the results were analyzed by using the 2 −ΔΔCt method after appropriate quality checking (e.g. melt curves). Primer sequences are shown in Table 1.

LC3B and P62 Hippocampal Expression.
ree mice in each group were randomly selected and their brains were harvested and cut into two along the sagittal axis. e fresh brain tissues were placed on a frozen plate and embedded in OCT until the tissues were covered entirely, following which the tissues were stored at −80°C refrigerators until frozen. e brains were sectioned and rinsed. Antigen retrieval was done in a pressure-cooker then goat serum blocking solution was added dropwise and incubated at 37°C. e primary antibody (LC3B/P62) was added dropwise and incubated at 37°C for 60 minutes. e secondary antibody (goat antirabbit; 1 : 400) was added dropwise, capped, and incubated at 37°C in the dark for 30 minutes. en, an anti-fluorescent quench blocking agent was added (containing DAPI) and the fluorescent sections were scanned by using Pannoramic MIDI.

Hippocampal
Expression of PI-3K, AKT3, mTOR, LC3B-II/I, and P62 Protein. Approximately 100 mg of hippocampal tissue, frozen at −80°C, was lysed in 1 ml of RIPA cell lysis buffer + 1 mM PMSF protease inhibitor. e samples were centrifuged at 1,200 RCF for 10 min at 4°C, and the supernatants were collected for protein extraction. e proteins were tested in seven consecutive steps including denaturation, loading electrophoresis, transmembrane, blocking with 5 percent fat-free milk powder for 2 hours at room temperature, and incubation with primary antibodies (PI-3K, mTOR, and LC3B-II/I at 1 : 1000; AKT3 at 1 : 500) overnight at 4°C. e next day the membranes were washed and incubated with secondary antibodies (goat anti-mouse -IgG and goat anti-rabbit IgG at 1 : 1000) at room temperature for 1.2 hours. e ECL luminescence kit was used for detection on film with appropriate exposure and film development. Image J software was used to analyze film strips. GAPDH was used as the reference protein.

e Targeting Relationship of lncRNAs Six3os1
, miR-511-3p, and AKT3. We used the starbase database (https:// starbase.sysu.edu.cn/) to predict binding sites and construct dual-luciferase reporter vectors, including Six3os1-wt, Six3os1-mut, AKT3-wt, and AKT3-mut. 293T cells and Evidence-Based Complementary and Alternative Medicine target plasmids previously prepared for transfection were dispensed into 96-well plates at 50-70% confluence. e target plasmid was fully mixed with 5 pmol of miR-511-3p (Negative Control, NC) at room temperature (solution A), following which 10 μL DMEM was mixed thoroughly with 0.3 μL of transfection reagent (HANBIO product with a concentration of 0.8 mg/mL) at room temperature for 5 min (solution B). Solution A was thoroughly mixed with solution B at room temperature for 20 min. e cell media were refreshed before transfection, following which the transfection mixture was added to the mix and incubated at 37°C with 5% CO 2 . After six hours, the media were exchanged and the cells were incubated for 48 hours. Luciferase activity was assayed by following the instructions of the Promega Dual-Luciferase assay kit.

Statistical
Analysis. SPSS version 23.0 was used for all statistical analyses. Data were expressed as mean ± standard deviation (mean ± SD). One-way analysis of variance was used to test differences between multiple groups and the least significant difference was used to examine group differences. Data that were non-normally distributed or showed signs of heteroscedasticity were analyzed by using the Kruskal-Wallis H test. Statistical significance was determined by p < 0.05. Figure 1, before treatment, the escape latency was markedly longer in the AD model group of the double transgenic mice compared with the control mice (p < 0.01). ere was no significant difference between the AD model and moxibustion groups (p > 0.05). After the intervention, the escape latency of AD mice in the model group was longer compared with the control group (p < 0.01). e escape latency was substantially shorter in the moxibustion group than in the AD model group (p < 0.01). Figure 2, organelles were well arranged with clear and complete structures. More autophagic vacuoles and autophagosomes were observed in the hippocampal cytoplasm of the control mice. Deformed and atrophied organelles were detected in the AD model group and the autophagic vacuoles were significantly reduced. Organelles were abundant and autophagic vacuoles and autophagosomes were increased in the moxibustion group compared with the AD model group. Figure 3, Aβ 1-42 protein content was markedly increased in the AD model group compared with the control group, as determined by immunohistochemistry (p < 0.01). Aβ 1-42 content was significantly decreased in the moxibustion group compared to the AD model group (p < 0.01).

Hippocampal Expression of lncRNAs Six3os1
, miR-511-3p, and AKT3. As shown in Figure 4, compared with the control group, in the AD model the expression of the lncRNAs Six3os1 and AKT3 was increased while miR-511-3p expression was lower, as determined by qRT-PCR (both p < 0.01). In contrast, compared to the AD model group, in the moxibustion group, the expression of the lncRNAs Six3os1 and AKT3 was decreased and miR-511-3p expression was significantly increased (both p < 0.01).    Figure 4: Expression of the lncRNAs Six3os1, miR-511-3p, and AKT3 in the hippocampus (mean ± SD, n � 6). Note: * * indicates significant difference versus the control group, at p < 0.01; ## indicates significant difference versus the AD model group, at p < 0.01. Figure 5, compared with the control group, in the AD model the LC3B-II/I ratio was markedly decreased, while P62 was significantly increased, as determined via WB (both p < 0.05). Compared with the model group, in the moxibustion group, the LC3B-II/I ratio was greatly increased, while P62 expression was increased (both p < 0.05).

Hippocampal Expression of LC3B and P62. As shown in
As shown in Figure 6, the fluorescent LC3B signal was attenuated in the AD model and moxibustion groups compared with the control group. At the same time, P62 was significantly enhanced in the AD model group (both p < 0.01). Compared with the model group, in the moxibustion group, LC3B fluorescence was significantly enhanced and P62 was greatly attenuated (both p < 0.01).

Hippocampal Protein Expression of PI3K/AKT/mTOR
Pathway-Related PI3K, AKT3, and mTOR. As shown in Figure 7, the expressions of PI3K, AKT3, and mTOR proteins were significantly increased in the model group compared to the control group, as demonstrated by WB (all p < 0.01). Conversely, compared with the model group, the moxibustion group showed lower expression of PI3K, AKT3, and mTOR proteins (all p < 0.01).

Discussion
AD is a type of dementia. In Chinese Medicine, several factors are recognized in older individuals as being pathogenic for AD including dysfunction of zang-fu organs, marrow sea deficiency, promotion of kidney Yang, and poor function of warmth. Professor CAI Shengchao, a renowned traditional Chinese physician posited the hypothesis underlying our study; that is, that the governor vessel connects to brain function, inline with the heart and kidney, and passes through the Ren Channel, where all are organically connected with the function of the house of the soul. erefore, it was proposed that dementia treatment should start from the shen-braingovernor vessel-kidney-Ren Channel axis [15]. In our experiment, we selected the three acupoints Baihui, Fengfu, and Dahui from the governor vessel, with the intent of thriving the governor vessel, filling the marrow sea, resolving the phlegm, and opening the brain orifices. e pathogenesis of AD is complicated and includes the abnormal accumulation of Aβ, Tau protein hyperphosphorylation, and neuroinflammation [16,17]. Furthermore, the deposition of amyloid-β produces further pathogenic cascades that underlie the primary mechanisms for AD etiology and progress [18]. Gene mutations in amyloid precursor protein (APP), presenilin-1 (PS-1), and presenilin-2 (PS-2) can mediate the Aβ hypersecretion and eventually deposit to form senile plaques, which trigger neuronal damage and death [19]. In our study, we observed that in the AD model, the escape latency in the Morris Water Maze was significantly prolonged and Aβ 1-42 expression was significantly higher than in the control group. e  Autophagy is a primary metabolic process in eukaryotic cells that utilize lysosomes to degrade damaged organelles and aberrant proteins to maintain cell homeostasis [20]. As a landmark protein of autophagy, LC3B is involved in forming early autophagic vacuoles and can reflect the extent of autophagy [21]. e autophagic substrate P62 is continuously consumed during autophagy formation and reflects autophagic activity [22]. It has been demonstrated that activation of autophagy can reduce Aβ levels [23]. For example, Chen et al. reported several biochemical alterations that can enhance autophagic activity, including downregulation of miR-331-3p and miR-9-5p which markedly attenuated the accumulation of Aβ in mice with early AD [24]. Wu et al. reported that SIRT5 overexpression can ameliorate AD progression in vitro and in vivo by activating autophagic mechanisms to clear Aβ protein [25].
AKT, as a serine/threonine protein kinase, is a central effector molecule of the phosphoinositide 3-kinases/protein kinase B (PI3K/AKT) signaling pathway with its location at its hub [26]. AKT can influence downstream mTOR signaling, thereby contributing to the PI3K/AKT/mTOR pathway. e PI3K/AKT/mTOR signaling pathway is one of the critical pathways that regulate autophagy; inhibition of this pathway can activate autophagy in the AD mouse model [27]. Here we found that AD mice exhibited reduced autophagic vesicles and autophagosomes, indicating defective autophagy and promotion of Aβ deposition. e expression of PI3K, AKT3, mTOR, and P62 protein was decreased after moxibustion compared with the AD model group. In contrast, the autophagy marker protein LC3B was increased, indicating that moxibustion might inhibit the PI3K/KT/mTOR signaling pathway to promote autophagy and reduce Aβ deposition; this, in turn, improved the cognitive function of the treated mice. is outcome was consistent with prior research [11].

Conclusions
In summary, we speculated that moxibustion can regulate the lncRNA Six3os1 and inhibit the PI3K/Akt/mTOR signaling pathway via miR-511-3, thus modulating downstream target proteins to promote cell autophagy.
ose effects contributed to the acceleration of Aβ 1-42 clearance in the hippocampus, reduced neuronal damage, and alleviated cognitive dysfunction. ese data provide support for the lncRNA Six3os1 as a molecular marker and target for the diagnosis and treatment of AD [31].
Data Availability e data that support the findings of this study are available from the corresponding author upon reasonable request.
Evidence-Based Complementary and Alternative Medicine 9