Hypoxic Pretreatment Adipose-derived Stem Cell Exosomes Improve Cognition by Delivery Circ-Epc1 and Shifting Microglial M1/M2 Polarization in Alzheimer’s Disease Mice Model and miRNA with real-time Enzyme-linked immunosorbent assay (ELISA) The greater therapeutic effect of hypoxia-pretreated ADSC exosomes at improving cognitive function involves circ-Epc1. Overexpressing miR-770-3p or downregulating TREM2 reversed the effect of circ-Epc1 on shifting M2 microglial under LPS stimulation. Circ-Epc1 containing ADSCs exosomes (circ-Epc1-Exo) increased therapeutic ecacy and improved cognitive function by decreasing neuronal damage and shifting microglia from M1 to M2 in the hippocampus.

Background: Alzheimer’s disease (AD) is the most major dementia in the globe. Increasing evidence informs that exosomes from hypoxic pretreatment adipose-derived stem cells (ADSCs) could therapeutically affect cognitive function in AD-associated pathophysiology. However, their role and regulatory mechanism remain largely unknown. Methods: High-throughput sequencing was used to identify differentially expressed exosomal circRNAs from ADSCs or hypoxia pretreated ADSCs. Luciferase reporter assays and RT-qPCR were used to investigate the relationships between circ-Epc1, miR-770-3p, and TREM2. APP/PS1 double transgenic AD model mice were then utilized to study therapeutic effect regarding circ-Epc1 in ADSCs exosomes. BV2 cells were used to understand the regulatory relationship between circ-Epc1, miR-770-3p, and TREM2 and how these interactions modulated phenotypic transformation and inammatory cytokine expression in microglia. Results: The result show that exosomes from hypoxia pretreatment ADSCs had a greater therapeutic effect at improving cognitive function by decreasing neuronal damage in the hippocampus. High-throughput sequencing found that circ-Epc1 played an important role in hypoxia pretreated ADSC exosomes regarding their ability to improve cognitive function. Luciferase reporter assays showed that TREM2 and miR-770-3p were downstream targets of circ-Epc1. Overexpressing miR-770-3p or downregulating TREM2 reversed the effects of circ-Epc1 on M2 microglia under LPS treatment. In vivo experiments showed that circ-Epc1-containing ADSC exosomes increased the therapeutic effect of exosome at improving cognitive function by decreasing neuronal damage and shifting hippocampal microglia from M1 to M2 polarization. Conclusions: Taken together, the data found that hypoxic pretreatment ADSCs exosomes improve cognition by delivery circ-Epc1 and shifting microglial M1/M2 polarization in alzheimer’s disease mice model. upregulated expression of 10 circRNAs (circ-Rtn4, circ-Hipk2, Circ-Epc1, circ-Asxl1, circ-Atp9b, circ-Xrn2, circ-Sass6, circ-Chd7, circ-Taf4a, and circ-Prdm2) between Exo and HExo. These data demonstrated only circ-Epc1 expression signicantly increased in HExo compared with Exo Bioinformatics analysis advised that circ-Epc1 was derived and cyclized from an exon of Epc1, which is located at Together, these data suggested that circ-Epc1 may play a role in improving cognitive function. Together, these data showed that TREM2 and miR-770-3p were important downstream targets of circ-Epc1 and that circ-Epc1 may regulate TREM2 by sponging miR-770-3p. A previous study found that counteracting immunomodulation of microglia polarization could reduce cognitive aging impairment and AD [17]. In this study, BV2 cells were transfected with circ-Epc1 overexpression vector, miR-770-3p mimic, or si-TREM2 individually or in combination. After exposure to 1 µg/mL LPS for 1 d, the we collected BV2 cells for phenotypic analysis. IF detection of microglial/macrophage polarization was performed using Iba-I + (total microglia), CD11b + (M1 microglia), and CD206 + (M2l microglia). The results showed that upregulating circ-Epc1 decreased M1 microglial differentiation by decreasing CD11b expression. In contrast, overexpressing miR-770-3p or downregulating TREM2 partially reversed M1 microglial differentiation following circ-Epc1 overexpression. We also found that upregulating circ-Epc1 promoted M1 microglial differentiation by increasing CD11b expression. Additionally, overexpressing miR-770-3p or silencing TREM2 partially reversed M2 microglial differentiation in circ-Epc1-overexpressing cells (Fig. 5A-D). ELISA detection showed that miR-770-3p overexpression or silencing TREM2 partially reversed the inhibitory effect of circ-Epc1 to the inammatory factor expressions such like IL-6, TNF-α, and IL-1β (Fig. 5E-G).


Introduction
Alzheimer's disease (AD) is the most common contribution regarding dementia in the globe and is being more prevalent due to human aging population; it thus constitutes a big challenge to health care systems [1,2]. The hallmark pathological feature of AD is the deposition of β-amyloid (Aβ), which has strong neurotoxicity in brain tissue and leads to cognitive impairment [3,4]. AD is a microglia-mediated neuroin ammatory disease. More evidence inferrs that microglial activation in central nervous system is heterogeneous, which could be categorized to disparate classes: the M1 and M2 phenotypes. Due to the phenotype activated, microglia could generate either neuroprotective or cytotoxic effects [5]. The different microglial phenotypes vary according to disease stage and severity. The ability to control stage-speci c switching of M1/M2 phenotypes in appropriate time windows might supply therapeutic bene ts to AD patients [6,7].
Mesenchymal stem cells (MSCs) are a family of adult stem cell that could produce a large number of multivesicular bodies that are secreted in the form of exosomes. Exosomes vary in size, with diameters of approximately 30-150 nm, which can cross blood-brain barrier. Exosomes can carry a large number of non-coding (nc) RNAs including circRNAs to the brain to alter physiology. Exosomes derived from MSCs play regulatory roles on AD [8][9][10]. A former investigation illustrated that when adipose-derived stem cell (ADSC) exosomes were injected intravenously, they improved learning and memory capabilities signi cantly, which decremented plaque deposition and Aβ levels to normalize in ammatory cytokine levels [11]. Present study discovered that hypoxia pretreated ADSC exosomes had increased treatment e cacy for AD-induced nerve damage and cognitive impairment in the hippocampus. The subsequent aim of current study was to illustrate the role and regulatory mechanism regarding ADSC exosomes on AD.

Animals and ethics statement
We obtained APP/PS1 double transgenic mice and B6C3-Tg (APPswe, PSEN1dE9) 85Dbo/J (original species No. 004462) from Jax Laboratory (Bar Harbor, ME, USA). Our lab maintained transgenic mice on a standard 12-h light/dark cycle under constant temperature with free access to water and food. Animal ethics committee in Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine approved this study. All protocols followed "Guide for the Care and Use of Laboratory Animals" from National Institutes of Health (Bethesda, MD, USA), and we made all efforts to minimize the number of animals utilized and any discomfort experiments.

ADSCs multilineage differentiation
To validate ADSCs multilineage differentiation, third-passage mouse ADSCs were cultured in adipogenic differentiation medium (Sigma-Aldrich), which were stained with oil red O after 2 w, or cultured in osteogenic differentiation medium (Sigma-Aldrich) and stained with alizarin red after 3 w.

ADSC-derived exosome isolation and identi cation
After reaching 80%~90% con uency, our lab rinsed ADSCs with PBS and cultured them in FBS-free endothelial cell growth medium (EGM)-2MV supplemented with 1× serum replacement solution (PeproTech, Cranbury, NJ, USA) for another 2 d. We erased conditioned culture medium and centrifuged it at 300 × g for 10 min and then at 2000 × g for 10 min to erase apoptotic cells and cellular debris. Brie y, we removed cell debris and large membrane vesicles by sequential centrifugation at 300 × g for 10 min, 2,000 × g for 10 min, and 10,000 × g for 0.5 hr, followed by ltration through 0.22-μm syringe lters. Afterwards, our lab transferred supernatant that cleared to fresh tube and spun it at 100,000 × g for 70 min. Then, we completely removed supernatant, and washed the pellet with PBS to collect exosomes. We characterized exosomes via transmission electron microscopy and western blotting, and the size was determined by dynamic light scattering using nanoparticle tracking analysis (NTA; NanoSight, Malvern, Worcestershire, UK).
Strand-speci c high-throughput RNA-Seq library construction Our team extracted total RNA from ADSCs exosomes (Exo) and hypoxia-pretreated ADSC exosomes (HExo) with or without hyperglycemia pretreatment through TRIzol Reagent (Invitrogen, Carlsbad, CA, USA). About 3 μg total RNA from every sample was subjected to VAHTS Total RNA-seq (H/M/R) Library Prep Kit from Illumina (Vazyme Biotech Co., Ltd, Nanjing, China). In this way, we removed ribosomal RNA and retained other types of RNA including ncRNA and mRNA. Our lab treated puri ed RNA with 40 U RNase R (Epicenter, Madison, WI, USA) at 37°C for 3 h, followed by TRIzol puri cation. Our lab made RNAseq libraries via KAPA Stranded RNA-Seq Library Prep Kit (Roche, Basel, Switzerland), which were subjected to deep sequencing with an Illumina HiSeq 4000 at Aksomics, Inc. (Shanghai, China).

Morris water maze (MWM) test
Our lab tested memory and learning function with MWM [12]. Operator blinded to treatment groups performed all examinations. The apparatus consisted of a round steel pool (diameter: 122 cm; height: 60 cm) that was lled with water to 1 cm higher than the platform (diameter: 10 cm; depth: 30 cm) top. Blue curtain with cues surrounded the pool, which was put in an isolated room (20°C, 60% humidity). We maintained water at 21°C and opaci ed it via inserting titanium dioxide.
Our team performed testing for 5 d. The rst 4 d (P40-P43) comprised a place navigation (reference memory) test including 16 training trials (4 trials per day for 4 d, with 30-40 min inter-trial interval). At the start of every trial, we put mice in the water facing the wall in various starting locations (south, north, west or east), which were allowed 1 min to discover and 15 s to stay on top of the hidden platform. If the mouse could not locate the platform within 1 min, it was guided to and allowed to stay on the platform for 15 s. We employed video tracking system to track swimming activity regarding every mouse. Escape latency, say, timing from placement into the water to staying on the platform, was tracked. We performed spatial probe test where the platform was moved out of the pool. We placed the animal in opposite quadrant and allowed it to swim freely for 2 min. We tracked platform crossing numbers. Our team analyzed data via motion-detection software for MWM test (Shanghai Mobile Datum Information Technology Co., Shanghai, China).
For phenotypic analyses, we transfected BV2 cells with a small interfering (si)RNA against TREM2, a circ-Epc1 overexpression plasmid, or a miR-770-3p overexpression plasmid (mimic) (GeneCopoeia, Shanghai, China) through Lipofectamine 2000 (Invitrogen) following standard procedures. We used cells for further experimentation after 2 d and exposed them to LPS (1 μg/mL) for 1 d before phenotypic analysis.
RNA and miRNA extraction with real-time (RT)-PCR Our lab isolated total RNA from serum, cells, or brain tissues via TRIzol reagent. We synthesized rst strand cDNA through PrimeScript RT Master Mix (Perfect Real Time) Kit (RR036A, Shiga, Japan), which was leveraged for RT-PCR, along with reverse and forward primers and Power SYBR Green PCR Master Mix (Life Technologies, Carlsbad, CA, USA). U6 and GAPDH were utilized as internal controls. We analyzed data via 2 −ΔΔCt method.

Luciferase reporter assays
We cloned putative miR-770-3p binding site in 3′-UTR of target gene TREM2 and circ-Epc1 (Wt or Mut) into psi-CHECK vector (Promega, Madison, WI, USA) downstream of re y luciferase 3'-UTR or circ-Epc1 as primary luciferase signal with Renilla luciferase as normalization signal, these vectors were termed TREM2-Wt/circ-Epc1-Wt and TREM2-Mut/circ-Epc1-Mut, respectively. psi-CHECK vector provided Renilla luciferase signal as normalization to compensate the differences between transfection and harvesting e ciencies. We performed transfection into HEK293 cells through Lipofectamine 2000 (Invitrogen). Our lab detected Renilla and re y luciferase activities 1 d after transfection with Dual-Luciferase Reporter Assay System (Promega) through luminometer (Molecular Devices, San Jose, CA, USA). Our lab detected relative Renilla luciferase activities following the manufacturer (Promega) instructions.

Immunohistochemistry (IHC) and immuno uorescence (IF) analyses
Our team xed brain tissue samples in 10% formalin solution, embedded them in para n, which were sectioned at 5 μm. Our lab stained tissue sections with TUNEL detection kit (Zeiss, Oberkochen, Germany) for apoptosis evaluation. Our lab performed IF staining for Iba-I, CD11b, and CD206 to validate microglial polarization. We analyzed results through Axiophot light microscope (Zeiss) or a uorescence microscope (Nikon, Tokyo, Japan), which were photographed with digital camera.

Enzyme-linked immunosorbent assay (ELISA)
We collected cell culture medium after the treatments that mentioned above. We utilized ELISA kits (R&D Systems, Minneapolis, MN) to obtain the interleukin (IL)-1β, IL-6, and tumor necrosis factor (TNF)-α levels following the instructions.

Statistical analysis
Continuous variations are denoted as mean ± standard deviation (SD). We used one-way analysis of variance for multiple comparisons via GraphPad Prism (GraphPad Software, Inc., La Jolla, CA, USA). A Pvalue ≤0.05 inferred statistically signi cance.

ADSC-exosomes characterization
Former investigations reported that exosomes from ADSCs reduce Aβ pathology and neuronal cell apoptosis in a transgenic mouse model of AD [13]. Nevertheless, the underlying mechanism is unclear. In current study, our lab isolated ADSCs to con rm the typical cobblestone-like morphology (Fig. 1A). IF staining was positive for mesenchymal cell markers CD90, CD29, CD44, and CD105, and negative for endothelial marker vWF (Fig. 1B-G). Oil-red-O (Fig. 1H) and alizarin red (Fig. 1I) staining data validated that ADSCs could differentiate into various lineages such as osteoblasts and adipocytes.
Western blot analysis of exosome lysates demonstrated positive expression regarding exosomal proteins CD81 and CD63 in both normal and hypoxia-pretreated ADSCs (Fig. 1J). Transmission electron microscopy illustrated that ADSC-exosomes exhibited characteristic cup-shaped morphology (Fig. 1K). We quanti ed size of ADSC-exosomes by Zetasizer Nano. The mean vesicle diameter was 80-130 nm (Fig. 1L), which was consistent with exosomes that formerly described [14]. The data inferred that the nanoparticles were exosomes.
Exosomes from hypoxia-pretreated ADSCs improved cognitive function by decreasing neuronal damage in the hippocampus.
To calculate exosome presences derived from ADSCs in cortex and hippocampus, we observed brain slices under a uorescence microscope 5 h after injection. We discovered DiI-labeled exosomes in hippocampus and cortex ( Fig. 2A). ELISA detection show in ammatory factor IL-6, TNF-α, and IL-1β expressions in brain tissue. These results showed that exosome treatment inhibited in ammatory factor expression. Exosomes from hypoxia-pretreated ADSCs (HExo) had more therapeutic effect and decreased IL-6, TNF-α, and IL-1β expression to a greater extent (Fig. 2B-D). IHC show that HExo suppressed nerve apoptosis in the hippocampus (Fig. 2E-F) more than other treatments. To explore behavioral consequences of exosomes in the mouse model of AD, we assessed spatial learning and memory with MWM (Fig. 2G). Exo treatment decreased escape latency more than the untreated AD mice. HExo had a greater therapeutic effect in decreasing escape latency than normal exosomes. Furthermore, during spatial probe test, the platform crossing number decremented in Exo treatment group, particularly in the HExo treatment group (Fig. 2H).
The greater therapeutic effect of hypoxia-pretreated ADSC exosomes at improving cognitive function involves circ-Epc1.
Circ-Epc1 containing ADSCs exosomes (circ-Epc1-Exo) increased therapeutic e cacy and improved cognitive function by decreasing neuronal damage and shifting microglia from M1 to M2 in the hippocampus.
To illustrate the protective effect of circ-Epc1 on cognitive function, APP/PS1 double transgenic mice received PBS, HExo, or circ-Epc1-Exo treatment. IHC showed that circ-Epc1-Exo treatment had a greater therapeutic effect than HExo at decreasing hippocampal apoptosis (Fig. 6A-B). To elucidate the behavioral consequences of exosome treatment in the mouse model of AD, our lab assessed spatial learning and memory via MWM (Fig. 6C). Circ-Epc1-Exo treatment had a greater therapeutic effect than HExo at decreasing escape latency compared with untreated AD mice. Furthermore, during spatial probe test, the platform crossing number decremented in HExo treatment group, particularly in the circ-Epc1-Exo treatment group (Fig. 6D).
ELISA detection showed that circ-Epc1-Exo treatment had considerable therapeutic effect than HExo on the in ammatory factor expressions such like TNF-α, IL-1β and IL-6 ( Fig. 6I-K). RT-qPCR detection showed that circ-Epc1-Exo treatment also had a greater therapeutic effect than HExo at increasing miR-770-3p and decreasing TREM2 expression in hippocampal tissues (Fig. 6L-M).

Discussion And Conclusions
The AD prevalence has incremented in recent years, leading to heavy burden regarding economies, families, and the whole society. Currently, there are no e cient therapeutic drugs for AD. Stem cell exosome therapy is a preferrable non-pharmacotherapy treatment because of its wide positive effect range, fewer side effects, and low economic burden [18][19][20]. Previous studies have found that ADSC exosome treatment can reduce Aβ40 and Aβ42 levels together with the Aβ42/40 ratio of AD cells [13]. Current study suggested that hypoxia-pretreated ADSC exosome treatment had a greater therapeutic effect than ADSC exosomes at decreasing apoptosis and in ammatory cytokines in the hippocampus. Cognitive function tests found that hypoxia-pretreated ADSC exosome treatment had greater effects than treatment with ADSC exosomes at recovering AD-induced cognitive impairment.
To verify the regulatory mechanism that underlined how ADSC exosomes alter AD pathophysiology, we produced high-throughput sequencing of exosomes from different treatment groups. The results showed that circRNAs were abnormally expressed in exosomes from hypoxia-pretreated ADSCs when compared with untreated ADSC exosomes. RT-qPCR detection showed that circ-Epc1 levels were signi cantly increased in hypoxia-pretreated ADSC exosomes. Circ-Epc1 is derived and cyclized from part of the Epc1 gene, which is located at chr18:6448902-6450637. Overall, our data suggested that circ-Epc1 may play a role in improving cognitive function.
CircRNAs are a particular family of noncoding single-stranded highly stable ribonucleic acid molecules that are abundant in eukaryotic transcriptome. CircRNAs are enriched signi cantly in human retinal and brain tissues [15]. Previous studies have found that circRNAs contribute to AD pathogenesis [21,22].
Another study found that circRNAs regulate gene expression by sponging miRNAs [23]. In this study, we found that circ-Epc1 interacted with miR-770-3p using a luciferase reporter system. The literature reports that miR-770-3p expression signi cantly increases with aging [24]. MiR-770-3p expression can facilitate apoptosis by inducing in ammation [25]. Further studies found that miR-770-3p interacts with triggering receptor that expressed in myeloid cells 2 (TREM2), which was also veri ed by our luciferase reporter assays. TREM2 is a receptor only expressed by microglia in brain, where it changes microglial immune homeostasis. Human genetic investigations demonstrated that loss-of-function mutations in TREM2 signaling are highly associated with elevated risk of age-related neurodegenerative traits such like AD. TREM2 loss confers resilience to synaptic and cognitive impairment in aged mice [26]. TREM2 also ameliorates neuroin ammatory responses and cognitive impairment via the PI3K/AKT/FoxO3a pathway in AD mice [27].
In this study, the data found that overexpressing circ-Epc1 promoted TREM2 and decreased miR-770-3p expression. Overexpressing miR-770-3p also decreased TREM2 expression. TREM2 overexpression did not in uence circ-Epc1 or miR-770-3p expression, suggesting that TREM2 and miR-770-3p were downstream targets of circ-Epc1. Circ-Epc1 regulated TREM2 by sponging miR-770-3p. Increasingly, studies are nding that AD induces microglial activation and transforms their polarization towards the M1 phenotype, which is associated with the production of proin ammatory cytokines, eventually leading to nerve cell damage in the hippocampus and cognitive impairment [28][29][30]. In our study we found that circ-Epc1 expression promoted M2 microglial phenotypes under LPS stimulation. However, overexpressing miR-770-3p or downregulating TREM2 reversed the promotion effects of circ-Epc1 on shifting the M2 microglial phenotype under LPS. Our in vivo experiments also found that circ-Epc1containing ADSC exosomes partially rescued AD-induced cognition impairment by shifting the microglial phenotype from M1 to M2, which resulted in decreased expression of in ammatory cytokines and apoptosis of hippocampal neurons.

Conclusions
Taken together, we found that hypoxic pretreatment adipose-derived stem cell exosomes improve cognition by delivery circ-Epc1 and shifting microglial M1/M2 polarization in alzheimer's disease mice model. Further work is necessary to verify and expand the ndings, which might ultimately enable us to completely elucidate the AD mechanisms.

Declarations
Ethical Approval and Consent to participate Not applicable.

Consent for publication
All authors agree to publish the article.