Resveratrol Attenuates the Disruption of Lipid Metabolism Observed in Amyloid Precursor Protein/Presenilin 1 Mouse Brains and Cultured Primary Neurons Exposed to A b

—To examine whether resveratrol (RSV), an activator of silent mating-type information regulation 2 homolog 1 (SIRT1), can reverse the disruption of lipid metabolism caused by b -amyloid peptide (A b ), APP/PS1 mice or cultured primary rat neurons were treated with RSV, suramin (inhibitor of SIRT1), ZLN005, a stimulator of peroxisome proliferator-activated receptor c coactivator-1 a (PGC-1 a ), or PGC-1 a silencing RNA. In the brains of the APP/PS1 mice, expressions of SIRT1, PGC-1 a , low-density lipoprotein receptor (LDLR) and very LDLR (VLDLR) were reduced at the protein and, in some cases, mRNA levels; while the levels of the proprotein convertase subtilisin/kexin type 9 (PCSK9), apolipoprotein E (ApoE), total cholesterol and LDL were all elevated. Interestingly, these changes were reversed by administration of RSV, while being aggravated by suramin. Furthermore, activation of PGC-1 a , but inhibition of SIRT1, decreased the levels of PCSK9 and ApoE, while increased those of LDLR and VLDLR in the neurons exposed to A b , and silencing PGC-1 a , but activation of SIRT1, did not inﬂuence the levels of any of these proteins. These ﬁndings indicate that RSV can attenuate the disruption of lipid metabolism observed in the brains of APP mice and in primary neurons exposed to A b by activating SIRT1, in which the mechanism may involve subsequently aﬀecting PGC-1 a . (cid:1) 2023 The Author(s). Published by Elsevier Ltd on behalf of IBRO. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).


INTRODUCTION
The major changes in the brain associated with insidious and progressive neurodegenerative Alzheimer's disease (AD) include senile plaques formed by deposition of bamyloid peptide (Ab) and neurofibrillary tangles of extensively phosphorylated tau, deposited outside or inside neurons (Chen et al., 2021a). Since the etiology and pathogenesis of AD are not fully understood, preven-tion and treatment of this disease are extremely challenging.
Among the several risk factors linked to lipid homeostasis identified to date (Gro¨sgen et al., 2010), rapid cholesterol biogenesis and abnormal neuronal energy metabolism are strongly implicated as playing critical roles in the pathogenesis of AD (Shah et al., 2017). Moreover, the level of cholesterol in the plasma and brain, as well as expression of Ab in the brain of amyloid precursor protein (APP)/presenilin 1 (PS1) transgenic mice fed a diet rich in cholesterol and fat were elevated (Refolo et al., 2001). In yet another study, reduction in the level of intracellular cholesterol was shown to inhibit the formation of aggregates of Ab (Schneider et al., 2006), perhaps through modulation of lipid rafts in cellular membranes. Indeed, several epidemiological investigations, as well as animal experiments indicate that abnormal lipid metabolism might be involved in the etiology of AD by promoting the generation and aggregation of Ab (Tan et al., 2022).
The level of apolipoprotein E (ApoE), a cholesterol/ lipid transporter in the central nervous system, is one of the most pronounced genetic risk factors for AD (Chen https et al., 2021b), with the presence of the ApoE e4 allele apparently enhancing susceptibility (Mahoney-Sanchez et al., 2016). Furthermore, ApoE accumulates in senile plaques and promotes the deposition of Ab in a dosedependent manner. Thus, ApoE may form a complex with Ab that influences endocytosis mediated by proteins associated with the low-density lipoprotein receptor (LDLR). In connection with AD, Ab may not be degraded completely in lysosomes, thereby accumulating intracellularly and being secreted as microfibrils or a more toxic form which promotes the effects of ApoE (Serrano-Pozo et al., 2021).
In addition, serum levels of triglycerides and cholesterol may play an important role(s) in connection with AD. For instance, the serum level of total cholesterol (TC) and LDL-cholesterol (LDL-C) is elevated in patients with AD, which is proposed to be a potential risk factor for cognitive impairment (Liu et al., 2020a). LDLR and the very LDLR (VLDLR), consisting of several multifunctional cell surface proteins, may mediate signaling events that regulate important aspects of normal brain development, including neuronal positioning and neurotransmission (Beffert et al., 2004), and are suspected of being involved in the pathogenesis of AD (Helbecque and Amouyel, 2000). Disruption of LDLR may impair cellular uptake of both monomers and oligomers of ApoE and Ab42 (Riad et al., 2020). Moreover, the VLDLR gene has been proposed to be linked to the susceptibility for developing AD (Okuizumi et al., 1995).
Proprotein convertase subtilisin/kexin type 9 (PCSK9), a serine protease, regulates the expression of LDLR by binding to both this receptor and its ligand LDL-C, thereby targeting them for lysosomal destruction. Since this enzyme also promotes the degradation of VLDLR and lipoprotein receptor-related protein 1 (Banerjee et al., 2016), PCSK9 would appear to be an attractive target for the treatment of familial hypercholesterolemia and other refractory dyslipidaemias. Indeed, PCSK9 reduces the expression of LDLR during brain development, which provides support for the hypothesis that the degradation of lipoprotein receptors by PCSK9 attenuates cholesterol uptake (Rousselet et al., 2011). Importantly, regulators of lipid metabolism are already highly effective, commercially available drugs and may offer new and effective options for treatment of the devastating disorder of AD (Di Paolo and Kim, 2011).
In recent years, common risk loci for AD have been identified, one of which is the silent mating-type information regulation 2 homolog 1 (SIRT1) (Rana et al., 2019). Among the members of the sirtuin family examined most extensively to date (SIRT1-7), SIRT1 appears to be the major regulator of cellular stress responses (Simmons et al., 2015). It has been proposed that the pathway involving SIRT and the peroxisome proliferator-activated receptor c coactivator-1a (PGC-1a) are closely involved in APP metabolism and Ab accumulation in the brains of patients with AD (Yin et al., 2021). In this context, we observed previously that activation of SIRT1 upregulates the expression of its downstream gene PGC-1a, inhibits the production of Ab, stimulates non-amyloidogenic path-ways and reduces accumulation of Ab (Dong et al., 2020). Moreover, SIRT1 is also involved in regulating lipid metabolism (Simmons et al., 2015). When cells are under stress, SIRT1 promotes oxidation of fatty acid by regulating the expression of enzymes for catalyzing gluconeogenesis and glycolysis, a process that involves the PGC-1a transcription factor in the case of chronic starvation.
In animal models of AD, SIRT1 can diminish the memory impairment, synaptic dysfunctional and neurodegeneration induced by Ab though altering cholesterol biosynthesis. Accordingly, these pathological changes are enhanced upon silencing SIRT1, which elevates the expression of sterol-regulatory elementbinding proteins and disrupts cholesterol metabolism (Shah et al., 2017). Moreover, cellular resistance to the toxic effects of Ab in the brain of a murine model of AD maintained on a high-cholesterol diet can be augmented by increasing the expression of SIRT1 through activation of PGC-1a, which significantly lowers the levels of cholesterol and free fatty acids in the systemic circulation (Hsieh et al., 2018).
Resveratrol (RSV), a small polyphenol that upregulates the enzymatic activity of SIRT1, improves cholesterol metabolism and resistance to the obesity and metabolic syndrome induced by a high-fat diet (Baur et al., 2006). In addition, RSV could indirectly activate SIRT1 expression (Beher et al., 2009), that its role may be related to the activation of substrates of SIRT1 (Kaeberlein et al., 2005), including nicotinamide adenine dinucleotide (NAD+) and PGC-1a (Higashida et al., 2013), data from these in vitro and vivo studies indicated that RSV could increase the level of NAD+ (the substrates of SIRT1) (Dong et al., 2020). Furthermore, this compound exerts neuroprotection by maintaining cholesterol homeostasis and reducing Ab deposition through enhanced expression SIRT1 (Sathya et al., 2017). The decreases in the total plasma levels of TC and LDL following activation of SIRT1 by PSV involve Pcsk9 (Miranda et al., 2015). The activated SIRT1 appears to increase hepatic expression of LDLR by inhibiting its PCSK9mediated degradation (Miranda et al., 2015), although not all findings are consistent with this hypothesis (Reynolds et al., 2010).
Clearly, a more detailed characterization of the correlation between SIRT1 and lipid homeostasis in the context of AD is required. Here, we examined the influence of RSV on lipid metabolism in brains of APP/ PS1 double-transgenic mice and primary cultured neurons exposed to Ab. Our findings indicate that activation of SIRT1 by RSV, through its downstream target gene PGC-1a, attenuates the disruption in lipid metabolism caused by Ab.

Experimental animals
Five 4-month-old B6. Cg-Tg (APPswe, PSEN1dE9) mice with an 85Dbo/Mmjax background and 15 wild-type (WT) mice of the same strain (all with a body weight of 20-30 g) were purchased from the Shanghai Nanfang Biological Technology Development Co. Ltd., China. All animals were first acclimatized at a humidity of 30-55% and temperature of 22-25°C for one week, following which each female APP/PS1 mouse underwent natural mating with three WT males. At a postnatal age of 12-20 days, the tip of the tail of each pup was cut off and DNA extracted for genotyping by the polymerase chain reaction (PCR), with detection of the products by 1.5% agarose gel electrophoresis. The PCR primers for target transcripts were designed on the basis of the complete cDNA sequences in GenBank. Using genomic DNA as a template for amplification, bands approximately 400 bp and 600 bp in size, consistent with the size of the APP and PS1 genes, respectively, were seen in the transgenic, but not WT mice (Dong et al., 2018). Starting at the age of 8 months, the APP/PS1 and WT mice received RSV (20 mg/kg), suramin (20 mg/kg) or physiological saline by gavage once daily for two months (Brandely et al., 1986;Xu et al., 2015). All of these animal experiments were pre-approved by the Ethical Committee of Guizhou Medical University, China (No. 1702110).

Cell cultures
Primary hippocampal neurons were prepared from the brains of neonatal Sprague-Dawley (SD) rats by the procedure described in detail previously (Nunez, 2008;Pacifici and Peruzzi, 2012). In brief, the hippocampus was digested with 0.25% trypsin for 10 min at 37°C and the digested tissue disrupted by repeated suction through fire-polished glass pipettes in 2 ml Neurobasal/B27 complete medium. The upper part of the resulting suspension containing single cells was transferred into new tube and the cells counted with trypan blue exclusion and thereafter placed onto 96-or 6-well poly-L-lysine-coated plates at a density of approximately 0.1-1.0 Â 10 4 /cm 2 . The neurons were maintained under a humidified atmosphere containing 5% CO 2 at 37°C with replacement of half of the medium once every 3 days. The purity of these primary neurons was evaluated by immunofluorescent doublestaining with the appropriate antibodies and revealed that approximately 90% of these cells were neurons (Dong et al., 2018). After 10 days of incubation, the medium was replaced with neurobasal medium lacking B27 and the neurons then prepared for various treatments.
Treatment of primary cultures of neurons with AbO, RSV, suramin, ZLN-005 or PGC-1a siRNA transfection AbO was prepared utilizing a published procedure (Klein, 2002). The primary neurons were seeded onto 96-, 12-or 6-well PLL-coated plates and B27 withdrawn from the culture medium 2 hr prior to treatment. In order to determine the optimal conditions for survival in starvation media, the cells were first exposed to different concentrations of AbO (0-2 mM), RSV (0-150 mM) or suramin (0-500 mg/ml), after treatment for 24, 10 ll of CCK-8 reagent was added to each well and then cultured for 2 hours. All experiments were performed in triplicate. The absorbance was analyzed at 450 nm using the Spectophotometer Varioskan LUX type:3020 (ThermoFisher) using wells without cells as blanks. The proliferation of cells was expressed by the absorbance (Dong et al., 2020). The viabilities observed were used to choose suitable concentrations of and incubation times with AbO, RSV, or suramin.
For transfection with PGC-1a siRNA, cultured neurons were first washed twice with 2 ml transfection medium, following which 0.8 ml of this same medium containing the siRNA transfection reagent was added to each tube for 5, 6, 7, 12 or 24 hr at 37°C in a CO 2 incubator (Dong et al., 2020). Thereafter, 2 ml Neurobasal/B27 complete medium was added and the cells incubated for an additional 24 hr. In addition, cultured neurons were treated with ZLN-005 (0-100 mM), an activator of PGC-1a activator, for 24 hr (Dong et al., 2020).
Assessing spatial learning and memory in mice Spatial learning and memory in mice of 10 months of age were evaluated using the Morris Water Maze test (Morris, 1984). This maze consists of a circular pool (130 cm in diameter) with dark walls filled with tap-water colored white by milk. An escape platform (9 cm in diameter) made of stainless steel with white walls is submerged 0.5 cm below the surface of the water. During the navigation test, the time required for the mouse to locate the escape platform (escape latency) was determined with Videotrack Software (View Point) and after locating this platform the animal was allowed to remain seated on it for 2 sec. Mice who failed to find the platform within 60 s were guided there and then allowed to remain for 2 sec as well, with their escape latency being recorded as 60 sec. Each mouse was tested 4 times a day, with 5-7 min intervals of rest, on each of 4 consecutive days and the results of the 4 trials on each individual day averaged for statistical analysis. On day 5, the platform was removed and the time spent swimming to and staying at the original location of the platform, as well as the number of times the quadrant of the tank in which the platform had been located was crossed were recorded. All of these behavioral tests were conducted in a quiet environment with subdued lighting.

Quantification of SIRT1 and PCSK9 mRNAs by quantitative real-time PCR
Total RNA was isolated from hippocampus tissues of mouse brains of mice or from primary cultured neurons with the Trizol reagents (Invitrogen, USA), using DNase I to remove residual genomic DNA. Thereafter, 3 mg total RNA was converted with the first-strand cDNA synthesis kit (Promega, USA) and oligo-d (T)18 primers following the protocol recommended by the manufacturer (Dong et al., 2018(Dong et al., , 2020. The real-time PCR primers for target transcripts were designed on the basis of the complete cDNA sequences in GenBank (accession numbers: NM_001372090.  Table 2).
Quantitative real-time PCR was carried out utilizing the ABI PRISM 7300 Sequence Detection System (Applied Biosystems, USA) and the results analyzed with the GeneAmp7300 SDS software. In brief, the 10-ml reaction mixture contained 1 ml first-strand cDNA, 5 ml 2 Â SYBR Green Master (Rox) Mix, 0.5 ml each of the forward and reverse primers (10 M), 3 ml DNase and RNase-free H 2 O (Dong et al., 2018(Dong et al., , 2020. The thermal cycling conditions were 2 min at 50°C and 10 min at 95°C, followed by 40 cycles at 95°C for 15 sec and then 1 min at 60°C (Dong et al., 2018(Dong et al., , 2020. The levels of the SIRT1 and b-actin transcripts were calculated as 2 ÀDDCT , where DCT represents the difference between the cycle threshold (CT) values for the target gene and actin beta (Dong et al., 2018(Dong et al., , 2020. Determination of the levels of SIRT1 and PCSK9 in primary neurons by immunofluorescence Double immunofluorescent labeling for SIRT1 and PCSK9 involved pretreatment of the sections with 0.3% Triton-X100 for 10 min, followed by incubation in 2% fetal bovine serum in 1 Â PBST. Subsequently, the sections were incubated with primary antibodies (anti-SIRT1 or -PCSK9) overnight at 4°C, followed by, as the secondary antibody, 488-conjugated goat anti-rabbit antibodies (diluted 1:300) (Dong et al., 2018(Dong et al., , 2020. As a control, the primary serum was omitted, resulting in no detectable staining. Staining for SIRT1 and PCSK9 was monitored with an OLYMPUS laser scanning confocal microscope (Japan). The images were digitized and quantified using Image J program (Park et al., 2017).

Contents of TC and LDL in brain tissue and serum
Brain tissue (0.05 or 0.1 g) was homogenized in 1 Â PBS containing complete protease inhibitors in Eppendorf tubes. Serum was collected by centrifugation of whole blood and then placed into Eppendorf tubes. Circulating TC and LDL levels were determined in brains and serum using the Spectophotometer Varioskan LUX type:3020 (ThermoFisher). TC and LDL levels in brains and serum were measured using the appropriate kits (KeyGEN BioTECH, China) according to manufacturer's instructions.

Statistical analysis
The values for the different groups of mice and primary neurons are presented as means ± SD. Statistical analysis was performed employing analysis of variance (ANOVA), followed by Student-Newman-Kenl's test or the two-paired Student's t test, utilizing the SPSS22.0 software (SPSS Inc., USA). P values of <0.05 were regarded statistically significant.

RESULTS
Viability of primary cultured neurons exposed to AbO, RSV and suramin As evaluated utilizing the CCK-8 test, the viability of cultured primary neurons was significantly reduced upon treatment with !0.5 mM AbO (Fig. 1A) for 48 hr or with !20 mM RSV (Fig. 1B) or !300 mg/ml suramin (Fig. 1C) for 24 hr. On the basis of these findings, the concentration of and time of exposure to each factor was chosen (see further below). Here, suramin, which binds to the nicotinamide pocket of sirtuins, was used to inhibit these proteins, in contrast to their activation by RSV.
Levels of the PGC-1a protein in cultured primary neurons treated with ZLN-005 or subjected to siRNA transfection As determined by Western blotting, the level of the PGC-1a protein in cultured primary neurons was increased by treatment with 20 mM ZLN-005 ( Fig. 2A) and reduced following transfection with siRNA ( Fig. 2B) for 24 hr.

Spatial learning and memory
The escape latency time of APP/PS1 mice was clearly longer than that of the WT animals (Fig. 3A), indicating the expected impairment of spatial learning in the former. When the platform had been removed, the APP/ PS1 animals crossed its original position a larger number of times (Fig. 3B), but spent less time near this position (Fig. 3C), demonstrating their poorer spatial memory. In addition, the total distance of APP/PS1 mice was also longer and the average speed was slower than that of the WT mice (Fig. 3D, E). Interestingly, treatment with RSV attenuated the decreased spatial learning and memory exhibited by the transgenic mice, while suramin augmented this impairment (Fig. 3).
The levels of SIRT1 and PCSK9 mRNA and the corresponding proteins in the hippocampus of mice and primary neurons In comparison to WT mice and untreated neurons, as evaluated by Western blotting, the levels of the SIRT1 protein (Fig. 4A, B) and corresponding mRNA (Fig. 4C, D) in the hippocampus of APP/ SP1 mice and primary neurons exposed to AbO were significantly reduced, whereas the opposite was observed for PCSK9 ( Fig. 5A-D (Figs. 4  and 5). In addition, RSV increased the level of SIRT1, (Fig. 4A and C) and reduced PCSK9 in brains of the control mice, while suramin had opposite effects, (Fig. 5A and C).
The levels of the PGC-1a, ApoE, LDLR and VLDLR proteins in the hippocampus of mice and cultured primary rat neurons As determined by Western blotting, the levels of PGC-1a (Fig. 6A, E), LDLR (Fig. 6C, G) and VLDLR (Fig. 6D, H) in the hippocampus of the APP/PS1 mice and in cultured  The total distance; (E) The average speed. The values presented are means ± SD for 8 mice. *P < 0.05 and ** P < 0.01 in comparison to WT; # P < 0.05 and ## P < 0.01 in comparison to APP/PS1, as determined by analysis of variance (ANOVA), followed by Student-Newman-Keul's test. primary rat neurons exposed to AbO were significantly lower, while the level of ApoE (Fig. 6B, F) was higher than in WT mice. Furthermore, RSV attenuated, but suramin enhanced these differences as well (Fig. 6). RSV alone increased levels of PGC-1a, LDLR and VLDLR, but decreased ApoE in brains of the control mice, while suramin had opposite effects (Fig. 6A-D).

The levels of TC and LDL in the serum and brains of APP/PS1 mice
The levels of TC and LDL in the serum (Fig. 7A, B) and brain tissue (Fig. 7C, D) isolated from APP/PS1 mice were significantly lower than those in WT animals. Once again, RSV attenuated, whereas suramin augmented these differences. RSV alone decreased TC content, but increased LDL in serum of brains of the control mice, while suramin had opposite effects (Fig. 7).
The levels of the PCSK9, ApoE, LDLR and VLDLR proteins in cultured primary rat neurons transfected with PGC-1a siRNA or exposed to RSV, suramin or ZLN-005 Transfection of the neurons with PGC-1a siRNA enhanced their levels of PCSK9 (Fig. 8A) and ApoE (Fig. 8B) in primary cultured neurons, while reducing LDLR (Fig. 8C) and VLDLR (Fig. 8D). Although RSV alone decrease the levels of PCSK9 and ApoE, and increased LDLR and VLDLR in control neurons, it did not change the effects of PGC-1a siRNA transfection on these lipids (Fig. 8A-D).
Interestingly, ZLN005, conferring a protective effect by activating PGC-1a Liu et al., 2020b;Zhu et al., 2022), reduced the levels of PCSK9 (Fig. 8E) and ApoE (Fig. 8F) and increased LDLR (Fig. 8G) and VLDLR (Fig. 8H) in primary neurons. Whereas, suramin had the Fig. 4. Effects of RSV (20 mM) or suramin (300 mg/ml) on the levels of the SIRT1 protein and corresponding mRNA in the hippocampus of the brain of 10-month-old mice and in cultured primary rat neurons treated with AbO (0.5 mM). (A, C) The results of Western blotting and real-time PCR, respectively, on mouse brain tissue. (B, D) The results of Western blotting and real-time PCR, respectively, on cultured primary neurons. Below graphs A and B, representative Western blots are shown. (E-I) Immunostaining of untreated neurons, and neurons treated with AbO, AbO + RSV, AbO + suramin, respectively, for SIRT1 (green, indicated by the arrows). Scale bar = 50 mm. (I) quantification of the immunostaining shown in E-H. The values presented are the means ± SD for 8 mice or independent experiments (I). *P < 0.05 and ** P < 0.01 in comparison to the corresponding WT animals or untreated cells (controls); # P < 0.05 and ## P < 0.01 in comparison to APP/PS1 or AbO, as determined by analysis of variance (ANOVA), followed by Student-Newman-Keul's test.
opposite results on these lipids and did not influence the effects of ZLN005 (Fig. 8E-H).

DISCUSSION
The lack of effective approaches to the treatment of AD has motivated considerable research in this area and, interestingly, sirtuin may be a potential target in this context (Martino Adami et al., 2017). SIRT1 has recently been shown to attenuate the amyloid transition of APP, both in cell cultures and in transgenic mouse models of AD, by increasing production of a-secretase. This enzyme catalyzes the non-amyloidogenic processing of APP, thereby reducing the accumulation of pathological Ab (Bonda et al., 2011).
In our present experiments, we observed the impaired spatial learning and memory in the mice carrying the APP/ PS1 double mutation. Spatial learning and memory deficits, as assessed by Morris Water Maze test, are the most commonly described cognitive alterations in any transgenic AD animal model and the APP/PS1 mouse exhibits alterations of the cognitive deficit as early as 5-7 months of age (Lewis et al., 2010;Brose et al., 2018). These findings indicate that high levels of Ab can influence cognitive functions and alter the morphology of neurons, results similar to those we have reported previously (Dong et al., 2020). The underlying mechanism(s) may involve the lowered levels of SIRT1 in the brains of the transgenic mice. In AD animal studies, some authors also investigated the changes of motor behavior and found that ischemic injury resulted in reduced motor coordination and impaired spatial learning and memory by inducing transient global ischemia to young adult APP/PS1 mice (Kemppainen et al., 2014), which may bring more attention for the change.
The ability of SIRT1 to modulate synaptic plasticity and memory formation (Gao et al., 2010) makes it a promising target for the treatment of cognitive impairment (Cao et al., 2017). We demonstrated earlier that a reduction in SIRT1 expression leads to excessive deposition of Ab and elevates the number of senile plaques in the The results of Western blotting and real-time PCR, respectively, on cultured primary neurons. Below graphs A and B, representative Western blots are shown. (E-I) immunostaining of untreated neurons, and neurons treated with AbO, AbO + RSV AbO + suramin, respectively, for PCSK9 (green, indicated by the arrows). Scale bar = 50 mm. I: quantification of the immunostaining shown in E-H. Scale bar = 50 mm. The values presented are means ± SD of 8 mice or independent experiments (I). *P < 0.05 and ** P < 0.01 in comparison to the corresponding WT animals or untreated cells (controls), # P < 0.05 and ## P < 0.01 in comparison to APP/PS1 or AbO, as determined by analysis of variance (ANOVA), followed by Student-Newman-Keul's test.
brains of APP/PS1 mice (Dong et al., 2020). In contrast, when SIRT1 was activated by RSV, the number of senile plaques was significantly reduced and the spatial learning and memory of the animals improved significantly. In addition, RSV inhibited the formation of Ab42 fibrils and cytotoxicity in a dose-dependent manner; attenuated Ab-induced cell death; promoted the clearance of Ab; and improved memory by lessening cognitive impairment (Feng et al., 2009;Porquet et al., 2014). However, it's more interesting to understand what is the possible mechanism underlying that RVS not only stimulates the activity of SIRT1, but also influences the expression of SIRT1 (Kim et al., 2018;Wang et al., 2018;Yang et al., 2019), which is worthy of further study.
Interestingly, we found here that stimulation of the activity of SIRT1 by RSV also improved the spatial learning and memory of mice carrying the APP mutation; and made cultured primary rat neurons exposed to Ab more viable. In contrast, suramin, an inhibitor of SIRT1 inhibitor, worsened the symptoms of neurodegeneration in these same mice and cells, providing further support for the neuroprotective action of SIRT1. In our previous research (Dong et al., 2018), we have reported that examination of the brains of 6-or 10-month-old APP/PS1 mice revealed Ab-  immunoreactive senile plaques distributed throughout the cortex and hippocampus of the mice. Furthermore, the numbers and size of these plaques were markedly reduced in both the cortex and hippocampus by exposure to RSV and increased by suramin. In addition, up-regulation of the expression of SIRT1 and PGC-1a improves learning and memory in a rat model of AD . Screening of a pharmacophore library of SIRT1 activators revealed that RSV appears to be the most potent (Sawda et al., 2017). Preclinical findings also support the proposal that RSV may play a valuable role in connection with the treatment and prevention of neurodegenerative diseases, including AD (Sun et al., 2010). RSV can protect neurons from reactive oxygen species (ROS), hydrogen peroxide free radicals, nitric oxide, Ab, and other intra-and extracellular toxins associated with neurodegenerative disorders (Gra¨ff et al., 2013). In addition, this compound may reduce production of Ab by inhibiting enzymes that generate ROS (Anekonda, 2006). All of these observations are consistent with our own experimental data (Dong et al., 2018(Dong et al., , 2020. Ab multimers, which play a key role in the pathogenesis of AD, are closely related to cholesterol homeostasis in the brain and, in particular, LDL-C metabolism (Allinquant et al., 2014). The level of cholesterol can influence the neurotoxicity of Ab at various stages, including the hydrolytic processing of APP and the activities of a, b, and c-secretases involved in this pro-cess. Furthermore, cholesterol mediates the transport, degradation and clearance of Ab metabolism .
The mechanism by which cholesterol exerts an impact on Ab production is not fully understood. It has been proposed that cholesterol-mediated alterations in membrane properties, including stiffness and fluidity, change the activities of membrane-bound proteins and enzymes, including secretases (Shobab et al., 2005). Intracellular low-density cholesterol inhibits Ab production, while higher levels of Ab may further enhance extracellular levels of cholesterol (Abad-Rodriguez et al., 2004).
Moreover, APP and Ab are thought to regulate cholesterol transport. TG and LDL-C can bind to membrane cholesterol and transport Ab into the cell, whereas the interaction of ApoE and high-density lipoprotein with membrane cholesterol result in outflux of Ab (Michikawa et al., 2000). In another study, high intracellular levels of cholesterol were shown to affect the functioning of ApoE, inhibiting intracellular degradation of Ab and its efflux through capillaries in the blood-brain barrier (Lee et al., 2012).
In other words, the balance between extra-and intracellular levels of cholesterol levels influences Ab production and disruption of cholesterol homeostasis by Ab is one important mechanism underlying neuronal degeneration and the development of AD. In this current investigation, we found that the levels of TC in the brain Fig. 8. Levels of the PCSK9, ApoE, LDLR and VLDLR proteins in cultured primary rat neurons exposed to RSV (20 mM), suramin (300 mg/ml), or ZLN-005 (20 mM) or subjected to transfection with PGC-1asiRNA, as determined by Western blotting. (A, E) PCSK9; (B, F) ApoE; (C, G) LDLR; and (D, H) VLDLR. Below each graph, a representative Western blot is shown. The values presented are the means ± SD of three independent experiments. *P < 0.05 and ** P < 0.01 compared to untreated cells (control); ## P < 0.01 compared to cells treated with ZLN-005 or transfected with siRNA, as determined employing analysis of variance (ANOVA), followed by Student Newman-Keul's test. and serum and expression of ApoE in the hippocampus of APP/PS1 transgenic mice were all significantly increased, whereas the levels of LDL, LDLR and VLDLR were lowered, indicating that AD is associated with disruption of cholesterol metabolism.
PCSK9 has been reported to be involved in neuronal differentiation, metabolism of members of the family of LDL receptors, inflammation, and apoptosis in the brain, although several in vitro and in vivo studies have arrived at opposite conclusions about this involvement (O'Connell and Lohoff, 2020). Regulation of plasma levels of PCSK9, LDLR, and LDL is closely linked, because PCSK9 is cleared from the plasma primarily by binding to LDLR, while simultaneously promoting LDLR degradation (Tavori et al., 2013). PCSK9 interacts with several receptors in the brains that transport cholesterol into neurons, including the ApoE receptor 2, LDLR and VLDLR (Adorni et al., 2019).
Even though PCSK9 is expressed at low levels in the adult brain, this expression is strongly upregulated in disease states. Thus, the levels of both PCSK9 mRNA and the corresponding protein in the frontal cortex of patients with late-onset AD are elevated (Picard et al., 2019). Furthermore, the enhanced content of PCSK9 in the cerebrospinal fluid (CSF) of AD patients is positively correlated with the levels of ApoE, as well as of the specific biomarker for AD, i.e., Ab in this same fluid (Picard et al., 2019). Consistent with these findings, we observed here that the levels of both PCSK9 mRNA and protein in the brains of APP/PS1 mice and in cultured primary rat neurons exposed to AbO were significantly elevated.
At present, the relationship between SIRT1 and PCSK9 in connection with the development of AD remains unclear. It is known that activated SIRT1 reduces the formation of senile plaques in the brain by reducing PCSK9-mediated degradation of LDLR degradation, while at the same time stimulating expression of this receptor (Miranda et al., 2015;Zimetti et al., 2017). Reduction of PCSK9 secretion, a novel effect of SIRT1 activity, is currently considered to be a very promising approach to attenuating plasma levels of TG, increasing LDL levels and treating atherosclerosis (Miranda et al., 2015).
In the pathogenesis of AD, lipid metabolism disorder is closely related to abnormal amyloid protein metabolism. Pathogenically, ApoE seeds Ab plaques in the brain with ApoE4 driving earlier and more abundant amyloids (Raulin et al., 2022). Strong evidence from clinical and basic research suggests that a major pathway by which ApoE4 increases the risk of AD is by driving earlier and more abundant amyloid pathology (Yamazaki et al., 2019). PCSK9 may degrade b-site APP-cleaving enzyme 1 (BACE1), the key enzyme cleaving APP to generate Ab (Wu et al., 2014). By using an established blood-brain barrier model, the reduced LRP1-mediated brain-toblood Ab clearance was due to PCSK9 across different endothelial monolayer in vitro (Mazura et al., 2022). Interestingly, lack of LDLR increased the susceptibility to Abinduced neurotoxicity in mice (de Oliveira et al., 2014). In addition, among the LDLR family members, LRPRrelated protein-1 (LRP1) is considered as an important therapeutic target in AD since it regulates brain and systemic clearance of Ab (Zlokovic et al., 2010).
In our study, administration of RSV to APP transgenic mice lowered the levels of PCSK9 and ApoE in the brains of these animals, while elevating LDL levels and expression of LDLR and VLDLR. These results indicate that pharmacological activation of SIRT1 by RSV might attenuate the abnormally rapid secretion of PCSK9 from and production of TC in mouse brains that express high levels of APP. In contrast, when the activity of SIRT1 was inhibited with suramin, expression of both PCSK9 and ApoE was clearly increased, while expression of LDLR and VLDLR was lowered even further, providing additional evidence that SIRT1 plays an essential role in regulating cholesterol homeostasis.
The transcription factor PGC-1a is known to act downstream of SIRT1 (Park et al., 2014). We have previously shown that SIRT1 can regulate the activity of PGC-1a, reduce the accumulation of Ab, help preserve mitochondrial functions and inhibit cellular apoptosis (Dong et al., 2020). ZLN005, the PGC-1a activator, exhibits neuroprotective effects under ischemic conditions and molecular mechanisms possibly involving activation of PGC-1a signaling pathway and cellular antioxidant systems . ZLN005 efficiently protects retinal pigment epithelium cells from oxidative damage through selective induction of PGC-1a and its target antioxidant enzymes (Satish et al., 2018;Wang et al., 2021). Here, we observed that in cultured primary rat neurons ZLN005 significantly decreased the expression of PCSK9 and ApoE, while increasing the levels of LDLR and VLDLR. Interestingly, when these same cells were transfected with siPGC-1a, these changes were not reversed by reactivation of SIRT1. It seems likely that following activation of PGC-1a after RSV, the relationship between lipid metabolism and the amyloidogenic pathway of APP metabolism designed to maintain lipid homeostasis involves primarily stimulation of PGC-1a.
In conclusion, the mice that express APP at high levels exhibited impaired spatial learning and memory. In addition, their brain protein and/or mRNA levels of SIRT1, PGC-1a, LDLR and VLDLR were lowered, and the the contents of LDL in the serum or brains were also decreased; levels of PCSK9 and ApoE protein elevated; and contents of TC enhanced. These changes were reversed by administration of RSV, while being aggravated by suramin. Analogous results were obtained with cultured primary rat neurons exposed to AbO. Altogether, our present findings, in combination with previous reports, indicate that RSV attenuates the disruption of lipid metabolism in the brains of APP/PS1 mice and in primary neurons exposed to AbO by activating SIRT1, a process in which PGC-1a may play a key role. As an activator of SIRT1, RSV might be able to slow down the pathological development of AD.

DECLARATION OF COMPETING INTEREST
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.