Alzheimer's Amyloid-β Accelerates Cell Senescence and Suppresses the SIRT1/NRF2 Pathway in Human Microglial Cells

Microglia play important roles in maintenance of brain homeostasis, while due to some pathological stimuli in aging-related neurodegenerative diseases including Alzheimer's disease, they are malfunctioning. Here, we demonstrated that amyloid-β (Aβ) accelerated cell senescence characterized by the upregulation of p21 and PAI-1 as well as senescence-associated beta-galactosidase (SA-β-gal) in human microglial cells. Consistently, Aβ induced the senescence-associated mitochondrial dysfunctions such as repression of ATP production, oxygen consumption rate (OCR), and mitochondrial membrane potential and enhancement of ROS production. Furthermore, Aβ was found to significantly suppress mRNA expression and protein level of Sirtuin-1 (SIRT1), a key regulator of senescence, and inhibit mRNA expression and translocation of NRF2, a critical transcription factor in inflammatory responses, leading to impairment of phagocytosis. Rescue of SIRT1, as expected, could counteract the pathological effects of Aβ. In summary, our findings revealed that Aβ accelerates human microglial senescence mainly through its suppression of the SIRT1/NRF2 pathway and suggested that genetic and pharmaceutical rescue of SIRT1 may provide a potential alternative treatment.


Introduction
Alzheimer's disease (AD) is an age-related neurodegenerative disease and is often characterized by tau aggregation and amyloid-β (Aβ) deposition [1,2]. It is identified that accumulation and aggregation of Aβ drives subsequent pathological events such as neuroinflammation, mitochondrial dysfunction, and cell senescence [3][4][5]. Microglia are the brain's innate immune cells and play important roles in AD [6,7]. Compelling evidence suggests that microglia, in the aged neurodegenerative brain, are activated and recruited to Aβ plaques and secreted proinflammatory cytokines such as IL-1β, IL-6, and TNF-α, which are similar to immunosenescence of macrophages [8]. Cell senescence affects phagocytosis of microglial in AD mice [9]. Cell senescence also leads to microglia dysfunction, resulting in inaccurate response to external stimuli and neurodegeneration worsening [8,10,11]. The prominent feature of microglial senescence includes the morphological alteration described as "dystrophy," [12] telomere shortening [13,14], and functional alterations [8]. During senescence, microglia shift the glycolytic metabolic state featured by the mitochondrial activity [3,15], change their inflammatory profile, increase the immunophenotypic expression, and more importantly, switch from neuroprotective to neurotoxic role when activated [10,[16][17][18]. Aβ, the main contributor of AD, has been suggested to accelerated microglial senescence [13]. However, there is still no direct evidence so far showing the influence of Aβ deposition on human microglial senescence.
Sirtuin-1 (SIRT1) is a NAD-dependent deacetylase that participates in the regulation of cell senescence, metabolism, inflammation, and mitochondrial function [19,20]. Under homeostasis, the expression and activity of SIRT1 is controlled by multiple mechanisms and maintained at normal state [21,22]. However, during aging, metabolic disorder, or neurodegenerative diseases, the expression of SIRT1 is diminished, intensifying oxidative stress potentially [23][24][25]. It is well known that sharp decrease of SIRT1 level is closely related to the accumulation of Aβ and tau proteins in AD patients [26][27][28]. SIRT1 has been suggested to reduce Aβ deposition and toxicity and improved AD pathology based on several earlier in vitro and in vivo studies [29][30][31]. Furthermore, SIRT1 is closely associated with nuclear factor E2-related factor 2 (NRF2), a transcription factor involved in regulating inflammatory responses through activating its downstream genes [32][33][34]. Therefore, regulation of the SIRT1/NRF2 pathway may provide a hopeful way for preventing or treating aging-related neurodegenerative disease.
Here, we used Aβ to induce cellular senescence in human microglial cells. After Aβ stimulation, we found the senescence-related mitochondrial functions were exacerbated significantly. We also detected that Aβ induction affected phagocytosis and ROS production of microglia and downregulated the SIRT1/NRF2 pathway. Overexpression of SIRT1 or using SIRT1 activator such as aspirin can counteract Aβ-induced cellular senescence. In summary, our results suggest that the SIRT1/NRF2 pathway is a therapeutic target for AD-related cellular senescence.

Aβ Peptide Preparation.
The Aβ peptides (Chinese peptide) were prepared according to the protocols described previously [34,35]. In brief, Aβ peptides were dissolved in HFIP (Sigma, #105228) to a final concentration of 1 mM, the HFIPtreated Aβ peptides were resolved in DMSO and then diluted to a concentration of 100 μM with DMEM/F12 phenol-red free medium and incubated at 4°C for 24 h. After centrifugation at 12,000 g for 10 min, the supernatant with soluble Aβ was added to cultures. Aβ 42-1 peptides (Beyotime, #P9005) were used as a negative control. Aβ 42-1 peptides were prepared using the same protocol. The concentration of Aβ 42-1 was 10 μM. In this paper, Aβ presented Aβ 1-42 .
2.3. SA-β-Gal Staining. Senescence-associated β-galactosidase (SA-β-gal) activity was performed using the SA-β-gal staining kit (Beyotime, #C0602), according to the manufacturer's instructions. In brief, cells were plated in the density of 40,000 cells per well into a 24-well plate. After 24 h seed-ing, cells were treated with Aβ for 72 h, then the cells were fixed with 4% formaldehyde in PBS for 15 min, and the fixed cells were stained with SA-β-Gal staining solution at 37°C for 15 h. The percentage of positively stained cells were calculated based on three replicates.

Western
Blotting. Western blotting was performed as described previously [34]. Briefly, 20 μg samples were loaded and separated on 10% or 12% SDS-PAGE and transferred to PVDF membranes. Membranes were blocked with 5% milk in TBST for 1 h, we used TBST to dilute primary and secondary antibodies. Membranes were incubated with primary antibody overnight at 4°C, washed in TBST, and incubated with HRP-conjugated secondary antibody for 60 min. The proteins of interest were performed using an ECL western blot detection kit (Bio-Rad). ImageJ software was used to evaluate the densitometry. Actin or proliferating cell nuclear antigen (PCNA) was used as loading control. Antibodies used were as follows: p53 (Beyotime, #AF7671), PAI-1 (Cell 2.9. Phagocytosis Assay. HMC3 cells were plated into 96-well plates at 9,000 cells per well and cultured overnight and then treated with Aβ for 72 h. After that, the cells were washed with PBS and the medium were changed to FBS-free DMEM alone at 37°C for 6 h. The fluorescent latex beads (Sigma, #L1030) were preopsonised in 50% FBS and PBS, and the beads were loaded to the cells at concentrations of 20 beads per cell and incubated at 37°C for 3 h. After Aβ uptake, the cells were processed for immunofluorescence using Zeiss confocal laser scanning microscope (Zeiss 880 Airyscan).
Hoechst was used to stain the nuclei. Lastly, the fluorescence intensity was also detected using the BioTek SynergyNEO (BioTek, USA) at 485 nm excitation/538 nm emission.

Nuclear and Cytoplasmic
Extraction. HMC3 cells were cultured in 6 cm plates, grew for 24 h, and then were treated with Aβ for indicated time. Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime, #P0028) was used in this experiment according to the manufacturer's instructions. In brief, 80 μl Buffer A was added to cells for 10 min, next 5 μl Buffer B was added, then the cells were centrifugated at 12,000 g for 5 min, and the supernatant was the cytoplasmic protein. The precipitation was resolved with 25 μl Buffer C for 30 min and centrifuged at 16,000 g for 10 min at 4°C to obtain the nuclear fraction.

Aβ Induced Senescence Gene Activation in Human
Microglial Cells. Several studies have shown strong evidences that cellular senescence increased significantly in AD mice [33,35,36]. Here, we treated human microglial cells HMC3 with 10 μM Aβ for different times and evaluated the gene expression of senescence. Compared with the Aβ 42-1 treatment control group, cells treated with Aβ displayed significantly higher expression of senescenceassociated genes (Figures 1(a)-1(d)). Similarly, the protein level of p53, PAI-1, and p21 was detected, showing that PAI-1 and p21 was markedly increased at 72 h with 10 μM Aβ stimulation, but there was no obvious change about the level of p53 (Figures 1(e)-1(h)). Thus, treatment of Aβ for 72 h was applied for the subsequent experiments. We also 3 Oxidative Medicine and Cellular Longevity  ). Furthermore, we performed a senescence-associated beta-galactosidase (SA-β-gal) assay to confirm the senescence phenotype of HMC3 cells. As shown in Figures 1(m) and 1(n), an increased percentage of SA-β-gal-positive cells is observed in cell culture treated with 10 μM Aβ. We also evaluated SASP markers such as TNF-α, IL-1β, and IL-6 by qPCR (Figures 1(o)-1(q)).These results indicated that Aβ induced senescence in human microglial cells.

Aβ Accelerated Mitochondrial Dysfunction in Microglia.
Recent studies revealed that cellular senescence is associated with mitochondrial defects [37][38][39]. We therefore assessed the effects of Aβ on mitochondrial functions in HMC3 cells. To evaluate mitochondrial functions, we tested the oxygen consumption rate (OCR) in HMC3 after treatment with oligomycin (ATP synthase inhibitor), FCCP (H+ ionophore), or rotenone and antimycin A (electron-transport chain inhibitor). These results revealed that Aβ treatment significantly diminished the maximal respiratory capacity of mitochondria and ATP production in microglia compared with vehicle. Aβ treatment also reduced basal respiration and spare capacity OCR, but no statistically significant change was observed (Figures 2(a)-2(e)), Aβ 42-1 , as the negative control, had no obvious effect. Furthermore, we tested whether Aβ could induce the loss of mitochondrial membrane potential (MMP). In this study, we used JC-1 probe to evaluate MMP in HMC3 cells. Red fluorescence and green fluorescence characterized high and low mitochondrial membrane permeability, respectively, and the ratio could signify the change of MMP. Cells treated with Aβ increased green fluorescence intensity (Figure 2(f)) and reduced the red/green fluorescence (Figure 2(g)), indicating Aβ induced depolarization. Taken together, our data indicated that Aβ induced mitochondrial dysfunctions through reduction of OCR and MMP.

Aβ Decreased Microglial Phagocytosis and Increased ROS
Production. Phagocytosis, one of the most important feature of microglia, has been reported to be decreased significantly in AD mice [3,[40][41][42]. Moreover, recent studies showed that the phagocytic activity of mouse primary microglial cells was markedly decreased with Aβ stimulation, which is associated with mitochondrial dysfunction concluding reduction of OCR [3,42]. However, whether Aβ could affect phagocytosis in human microglial cells is unknown. Here, we treated HMC3 cells with Aβ for 72 h and mixed with fluorescent latex beads for 3 h. The phagocytic capacity was assessed by confocal microscope (Figures 3(a)-3(c)). And also, the average cell fluorescence intensity was detected at 485 nm excitation/538 nm emission using the a BioTek fluorescence reader (Figure 3(d)). Aβ also decreased human microglial paghocytosis by flow cytometry (Supplementary Figure 1A). The results revealed Aβ significantly reduced phagocytic capacity (Figures 3(a)-3(d)). Inflammatory responses, another vital feature of microglia, increase dramatically in AD mice and AD patients. Aβ could (o-q) Senescence-associated proinflammatory cytokines such as TNF-α, IL-1β, and IL-6 were detected by qPCR. The data are presented as mean ± SEM, n ≥ 3 independent experiments; * p < 0:05, * * p < 0:01, * * * p < 0:001, and * * * * p < 0:0001, analyzed by one-way ANOVA followed by Bonferroni's test. 6 Oxidative Medicine and Cellular Longevity  The cells were stained with JC-1 dye and then captured by Zeiss 880 microscope (f), and the fluorescence intensity in (f) was quantified using BioTek reader (g). Scale bars, 100 μm. The data are presented as mean ± SEM, n ≥ 3 independent experiments, * * p < 0:01 and * * * * p < 0:0001, analyzed by one-way ANOVA followed by Bonferroni's test. (e, f) ROS production in HMC3 cells was investigated with the DCFH-DA probe. The pictures were obtained by Zeiss microscope. Scale bars, 100 μm. The fluorescence intensity in (e) was quantified using BioTek reader (f). The data are presented as mean ± SEM, n ≥ 3 independent experiments; * * p < 0:01, * * * p < 0:001, and * * * * p < 0:0001, analyzed by one-way ANOVA followed by Bonferroni's test. 8 Oxidative Medicine and Cellular Longevity   . (h, i) The quantification of relative protein levels in (g). The data are presented as mean ± SEM, n ≥ 3 independent experiments; * p < 0:05 and * * p < 0:01, analyzed by one-way ANOVA followed by Bonferroni's test. 9 Oxidative Medicine and Cellular Longevity  induce ROS generation, thus causing oxidative stress of microglia. Here, we stimulated human microglial cells with Aβ for 72 h and assessed the intracellular ROS level by staining with the DCFH-DA probe. The probe has no fluorescence and can pass through plasma membrane freely and produce fluorescent DCF when oxidized by ROS. Results showed that treatment with Aβ markedly increased DCF fluorescence (Figures 3(e) and 3(f)). Taken together, these results indicated Aβ significantly impaired phagocytic capacity and increased ROS production in human microglia cells.

Oxidative Medicine and Cellular Longevity
3.4. Aβ Downregulated the SIRT1/NRF2 Pathway in the Cells. SIRT1, which is a NAD+-dependent deacetylase, has been reported to play an important role in age-related neu-rodegenerative diseases [19,43,44]. Recent studies have shown that the expression of SIRT1 was decreased markedly in AD patients [26][27][28]. Here, we wanted to understand whether SIRT1 took part in Aβ-induced microglial senescence. We treated cells with Aβ for different times and found that at 72 h, Aβ reduced the mRNA expression of SIRT1 and downregulated the protein level of SIRT1 (Figures 4(a)-4(e)). SIRT5, another SIRT family protein, was not affected after Aβ stimulation. Furthermore, accumulating evidence showed that SIRT1 is involved in the activation of nuclear factor E2-related factor 2 (NRF2) [32,33]. NRF2 can be served as a sensor of oxidative stress. Next, we investigated whether Aβ treatment affected nuclear translocation of NRF2. The results indicated that Aβ reduced the mRNA expression of NRF2 (Figure 4(f)).   Figure 2C and D). These results detected that SIRT1 may be involved in human microglial senescence.

Overexpression of SIRT1 Rescues Aβ-Induced Senescence and Mitochondrial Dysfuntions.
Since the protein level of SIRT1 has been reduced with Aβ treatment, we next tested whether overexpression of SIRT1 could rescue Aβ defects including senescence, mitochondrial disability, and microglial dysfunctions. Here, we transfected HMC3 cells with SIRT1 plasmid or FUGW plasmid. As shown in  Figure 3A). Similarly, Aβ-induced ROS production was significantly rescued by SIRT1 overexpression (Figure 5(h) and Supplementary Figure 3B). Moreover, SIRT1 treatment involved in the enhancement of Aβ phagocytosis ( Figure 5(i) and and Supplementary Figure 3C). Together, these data suggested that SIRT1 protein was dispensable for the Aβmediated cell senescence, mitochondrial dysfunctions, and microglial state.

Aspirin Alleviates Aβ-Induced Senescence and Mitochondrial Dysfunctions via Upregulation of SIRT1
Pathway. In this paper, we wanted to find some drugs which could relieve microglial cellular senescence. Surprisingly, treatment of aspirin in HMC3 was found to rescue cellular senescence after Aβ stimulation (Figures 6(a) and 6(b)). Aspirin is a common drug, which was widely used for treating pain, fever, inflammation, and cardiovascular diseases [45][46][47]. Previous studies showed that aspirin could activate SIRT1 in liver cells and in endothelial cells [48,49]. Here, pretreatment with aspirin for 4 h did lower senescenceassociated protein levels obviously (Figures 6(a) and 6(b)). Meanwhile, 100 μM aspirin markedly increased SIRT1 level.
To assess the effect of aspirin on senescent microglia cells, SA-β-gal activity was detected and there was a significant reduction in the number of SA-β-gal-positive cells in aspirin-treated cells (Figures 6(c) and 6(d)). We also investigated whether aspirin could affect mitochondrial functions and found that aspirin increased mitochondrial membrane potential using JC-1 probe (Figure 6(e) and Supplementary Figure 4A). Lastly, 100 μM aspirin significantly inhibited ROS production (Figure 6(f) and Supplementary Figure 4B) and increased phagocytic capacity in HMC3 cells (Figure 6(g) and Supplementary Figure 4C). In summary, aspirin may be a potential drug in aging-related neurodegenerative diseases through the SIRT1 pathway.

Discussion
AD is a neurodegenerative disease mainly characterized by the progressive aggregation of Aβ [50,51]. The microglia play an important role in the maintenance of brain homeostasis [52,53]. Recent studies indicated that microglia can be The fluorescent latex beads were added to the medium and incubated at 37°C for 3 h. The fluorescence intensity was quantified using BioTek reader. The data are presented as mean ± SEM, n ≥ 3 independent experiments; * p < 0:05, * * p < 0:01, * * * p < 0:001, and * * * * p < 0:0001, analyzed by one-way ANOVA followed by Bonferroni's test.
13 Oxidative Medicine and Cellular Longevity categorized into two opposite types: toxic phenotype and protective phenotype [54,55]. Toxic microglia produce chemokines and cytokines such as CCL2, IL-1β, IL-6, IL-12, and TNF-α and generate nitrogen species and reactive oxygen. However, protective microglia produce antiinflammatory cytokines such as IL-10 and TGF-β and growth factors. The dynamic changes of toxic/protective phenotypes are critically associated with AD. Endogenous stimuli including Aβ and tau may persistently activate proinflammatory responses and finally aggravate progression of neurodegenerative disease [35]. The expression of proinflammatory cytokines is one of the hallmarks of cellular senescence [37]. In the present study, we demonstrate that Aβ could induce microglial senescence. Srinivasan et al. revealed Alzheimer's patient microglia exhibited enhanced aging [56]. We also find that Aβ aggravate senescenceassociated mitochondrial dysfunctions and impair microglial functions. Interestingly, we revealed that the SIRT1/NRF2 pathway is partly reduced by Aβ stimulation. Notably, overexpression of SIRT1 or use SIRT1 activator such as aspirin may rescue Aβ defects.
Increasing evidences point out that mitochondrial dysfunction is one of the hallmarks of aging [57,58], attributing to the accumulation of mtDNA mutations, damaged fission and fusion behavior, weakened membrane potential, abnormal metabolism, and defective electron transport chain (ETC) function in mitochondrial. Mitochondrial dysfunctions in microglia has been linked to the development of agingrelated neurodegenerative diseases such as AD [59][60][61]. The high level of reactive oxygen species (ROS) and loss of mitochondrial membrane potential have been observed in human microglial cells with Aβ stimulation in our work. Previous studies reported that a metabolic switch from mitochondrial OXPHOS to anaerobic glycolysis in Aβ-treated primary mouse microglia is associated with microglia phagocytosis [3,42]. Here, we detected Aβ treatment significantly reduced OCR levels in human microglia cells and impaired the capacity of microglia phagocytosis. Thus, therapies targeting basic mitochondrial processes hold great promise.
Sirtuins are class III histone deacylases possessing outstanding properties in preventing diseases and reversing some aspects of aging [62][63][64]. SIRT1 has been shown to regulate cellular metabolism by acting as a cellular sensor [65]. Increasing studies indicate that the expression of SIRT1 is significantly diminished in aging, metabolic, and neurodegenerative diseases, leading to oxidative stress [66]. Importantly, the protein level of SIRT1 is also decreased dramatically in AD mice brains and AD patients, which is closely related to the accumulation of Aβ and tau proteins than in normal aging individuals [26,67]. Here, we found SIRT1 was decreased after Aβ stimulation in human microglia cells. At the same time, SIRT1-associated NRF2 nuclear translocation was also reduced after Aβ treatment. Aspirin is one of the most widely used treatments for cardiovascular disease. Aspirin has obviously antiinflammatory function. We found that aspirin could increase SIRT1 production and alleviate human microglial senescence. Aspirin has been reported to reduce amyloid plaque in a mouse model of AD [68]. However, aspirin does not reduce the risk of Alzheimer's disease in clinical trial [69]. There are many differences between cell lines and human. Therefore, activating the SIRT1/NRF2 pathway may provide a promising way for prevention and treatment of aging-related neurodegenerative diseases.

Data Availability
The data, methods, and study materials used to conduct the research will be available from the corresponding authors on reasonable request. fluorescent latex beads to detect phagocytic capacity. The cells were captured by Zeiss 880 microscope. Scale bars, 100 μm. Figure 4. Aspirin alleviates Aβ-induced mitochondrial dysfunction, ROS production and phagocytic capacity. (A) HMC3 cells were stained with JC-1 dye and then captured by Zeiss 880 microscope. Scale bars, 100 μm. (B) HMC3 cells were stained with DCFH-DA probe to detect ROS production by Zeiss microscope. Scale bars, 100 μm. (C) HMC3 cells were mixed with the fluorescent latex beads. The cells were captured by Zeiss 880 microscope. Scale bars, 100 μm. (Supplementary Materials)