Exposure of Mesenchymal Stem Cells to an Alzheimer's Disease Environment Enhances Therapeutic Effects

Mesenchymal stem cells (MSCs) have emerged as a promising tool for the treatment of Alzheimer's disease (AD). Previous studies suggested that the coculture of human MSCs with AD in an in vitro model reduced the expression of amyloid-beta 42 (Aβ42) in the medium as well as the overexpression of amyloid-beta- (Aβ-) degrading enzymes such as neprilysin (NEP). We focused on the role of primed MSCs (human Wharton's jelly-derived mesenchymal stem cells (WJ-MSCs) exposed to an AD cell line via a coculture system) in reducing the levels of Aβ and inhibiting cell death. We demonstrated that mouse groups treated with naïve MSCs and primed MSCs showed significant reductions in cell death, ubiquitin conjugate levels, and Aβ levels, but the effects were greater in primed MSCs. Also, mRNA sequencing data analysis indicated that high levels of TGF-β induced primed-MSCs. Furthermore, treatment with TGF-β reduced Aβ expression in an AD transgenic mouse model. These results highlighted AD environmental preconditioning is a promising strategy to reduce cell death and ubiquitin conjugate levels and maintain the stemness of MSCs. Further, these data suggest that human WJ-MSCs exposed to an AD environment may represent a promising and novel therapy for AD.


Introduction
Alzheimer's disease (AD) is a widespread cause of dementia and is an age-related [1,2], progressive, and irreversible neurodegenerative disease [3,4] for which no disease-modifying therapy exists [5,6]. Most of the drugs being developed target Aβ alone [7,8]. The development of a multi-target drug, however, may be more effective given the multiple pathogenic mechanisms involved in AD [9,10].
Prior studies including those reported by our group suggest that mesenchymal stem cells (MSCs) may be a potential treatment for AD [11][12][13][14][15][16]. MSCs secrete proteins that inhibit apoptosis and in ammation, modulate the immune response in damaged tissues, and promote endogenous neurogenesis and neuroprotection. Based on the speci c mechanisms induced and the improved therapeutic outcomes, MSCs show considerable promise [17]. When used to treat AD, MSCs expressed genes related to enhanced extracellular transport and secretion [11-13, 15, 16], which indicates an increase in paracrine activity. These genes are known to exhibit neuroprotective and neurotrophic features such as the inhibition of apoptosis, the regulation of cell proliferation, and the regulation of neurogenesis. Further, our previous study demonstrated that MSCs exposed to cerebrospinal uid (CSF) of AD patients upregulated the genes related to AD treatment while maintaining stemness [18]. Therefore, AD-exposed MSCs enhanced the overall e cacy of MSCs in AD therapy.
In this study, we investigated whether the therapeutic potency of MSCs could be enhanced by exposing them to an AD environment. Therefore, we generated AD-exposed MSCs using a co-culture of MSCs and the APP695-Swedish mutant (K595N/M596L)-expressing H4 cell (H4SW cell) line, which provided an AD environment characterized by high levels of secreted toxic forms of Aβ, such as Aβ1-40 and Aβ1-42 [19,20]. We then analyzed the therapeutic effects of the MSCs following exposure to the AD environment. Furthermore, to identify the genes expressed by conditioned MSCs, which were therapeutically effective in AD, we performed mRNA sequencing analysis of both the naïve and conditioned MSCs.

Wharton's jelly-derived mesenchymal stem cell culture
Wharton's jelly-derived mesenchymal stem cells (WJ-MSCs) were isolated according to the procedure described by Kwon [21]. The WJ-MSCs were cultured according to the standard operating procedures (SOPs) of the Good Manufacturing Practice facility at Samsung Medical Center. Prior to co-culturing with H4SW cells, the WJ-MSCs were detached using a 0.25% trypsin-EDTA solution (Gibco-Invitrogen).

Pre-conditioning MSCs under an AD environment
H4SW cells were cultured and maintained on a 6-well plate compatible with insert wells. Upon reaching 70% con uency, the H4SW cells were co-cultured with 1 × 10 5 WJ-MSCs on 6-well transwell inserts (BD Falcon, USA) for 24 h in serum-free medium at 37 °C with 5% CO 2 .

Flow-cytometry analysis for validating re-conditioned WJ-MSCs
After co-culture, the pre-conditioned WJ-MSCs at passage ve were detached using a 0.25% trypsin-EDTA solution and harvested in a 15 mL conical tube. After centrifugation, the WJ-MSCs were washed and resuspended in phosphate-buffered saline (PBS) with 2% FBS to block nonspeci c binding sites.
Immunophenotypic analysis of the pre-conditioned WJ-MSCs was performed according to the MSC criteria of the International Society for Cell Therapy (ISCT) [22] via ow cytometry to determine the expression of the following markers: CD44, CD73, CD90, CD105, CD14, CD11b, HLA-DR (MHC-II), CD34, CD45, and CD19 (BD Biosciences, USA). At least 10,000 events were acquired on a BD FACSVerse (BD Biosciences, NJ, USA), and the results were analyzed with BD FACSuite software version 10 (BD Biosciences, USA). The differentiation of pre-conditioned WJ-MSCs was analyzed according to the procedure outlined in a previous report [21].
2.5 H4SW cell co-culture with pre-conditioned WJ-MSCs At 70% con uency, H4SW cells (in the lower chamber of the Transwell unit) were co-cultured with 1 × 10 5 pre-conditioned WJ-MSCs seeded on 6-well transwell inserts (BD Falcon) for 24 h under serum-free conditions at 37 °C with 5% CO 2 . Naive WJ-MSCs were co-cultured with H4SW cells as a control group.
After co-culture for 24 h, the H4SW cells were harvested and rapidly frozen for further analysis.

Intraventricular injection of WJ-MSCs and TGF β into 5XFAD mice
A 12-month-old transgenic mouse model of AD, 5xFAD (MMRC #04848) was used in this study. The mice were purchased from the Jackson Laboratory (Bar Harbor, ME, USA). Experimental animals were divided into ve groups: wild-type (WT), 5xFAD (sham), +naïve MSC (naïve MSCs were injected into 5xFAD mice), +primed MSCs (primed MSCs were injected into 5xFAD mice), and + TGF β (recombinant TGF-β proteins were injected into 5xFAD mice). Before injecting WJ-MSCs, all the mice were anesthetized and maintained on 5% iso urane with 2% iso urane inhalation during the surgical procedure. After shaving and sterilizing the surgical site with povidone-iodine, a skin incision approximately 1 cm in length was made. Using a microdrill, a small burr hole was made at the following coordinates (right lateral ventricle): A/P-0.4 mm, M/L + 1.0 mm, and D/V-2.3 mm from the bregma. WJ-MSCs (1 × 10 5 cells) suspended in 3 µL of phenolred MEM-alpha or 3 µL of TGF-β (10 ng/mL) were injected into the right lateral ventricle at a rate of 1 µL per min with a 15 min delay using a Hamilton syringe (Hamilton Company, NV, USA). The needle was carefully removed after the injection was complete, and the skin was sutured, followed by sterilization of the area. All mice were euthanized one week after administration.

Brain tissue preparation
One week after the injection of WJ-MSCs, all mice were anesthetized with iso urane, followed by cardiac perfusion. The brain tissue from the mice was harvested and divided in half along the longitudinal ssure. The harvested brain tissues were frozen in liquid nitrogen and stored at -80 °C for Western blots and enzyme-linked immunosorbent assay (ELISA) analysis, or xed in 4% paraformaldehyde for histological analysis.
The ELISA tests were performed with an Aβ42 ELISA kit (Wako, Cambridge, UK) and an SRGN ELISA kit (LifeSpan BioSciences, Washington, USA) according to the manufacturers' instructions.

Thio avin-S staining
Fixed brain tissues were embedded in para n and 4-µm-thick coronal sections were prepared. To detect Aβ, thio avin-S staining was performed according to the manufacturer's instructions. All slides were depara nized by serial hydration using an graded ethanol series, followed by treatment of the slides with 1% ltered thio avin-S (Sigma-Aldrich) and washing. The mounted slides were stored at 4 °C before uorescence microscopy imaging (Nikon, Shinagawa, Tokyo, Japan).

RNA isolation
Total RNA was isolated using TRIzol reagent (Invitrogen). The RNA quality was assessed by an Agilent 2100 bioanalyzer using the RNA 6000 Nano Chip (Agilent Technologies, Amstelveen, The Netherlands), and RNA quanti cation was performed using an ND-2000 Spectrophotometer (Thermo Inc., DE, USA).

Library preparation and QuantSeq 3' mRNA sequencing
Libraries were constructed from the control and test RNAs using a QuantSeq 3' mRNA-Seq Library Prep Kit (Lexogen, Inc., Austria) according to the manufacturer's instructions. In brief, 500 ng of each total RNA sample was prepared and an oligo-dT primer containing an Illumina-compatible sequence at its 5' end was hybridized to the RNA, and reverse transcription was performed. After degradation of the RNA template, the second-strand synthesis was initiated by a random primer containing an Illuminacompatible linker sequence at its 5' end. The double-stranded library was puri ed by magnetic beads to remove all reaction components. The library was ampli ed to add the complete adapter sequences required for cluster generation. The nished library was puri ed from the PCR components. Highthroughput sequencing was performed via single-end 75 sequencing using NextSeq 500 (Illumina, Inc., USA).

QuantSeq 3' mRNA sequencing data analysis
The QuantSeq 3' mRNA-Seq reads were aligned using Bowtie2 [24]. Bowtie2 indices were either generated from the genome assembly sequence or the representative transcript sequences for aligning with the genome and transcriptome. The aligned le was used to assemble the transcripts, estimate their abundance, and detect the differential expression of genes. Differentially expressed genes were determined based on unique counts and multiple alignments using Bedtools [25]. The RT (read count) data were processed based on the quantile-quantile normalization method using EdgeR within R software [26] using Bioconductor [27]. Gene classi cation was based on searches conducted in the DAVID

Statistical analyses
All values are presented as the mean ± standard error of the mean (S.E.M). One-way ANOVA was used to assess signi cance and a p-value of ≤ 0.05 was considered statistically signi cant. IBM SPSS software version 21.0 was used for all analyses.

Primed MSCs show anti-apoptotic effects in the H4 Swedish cell line under serum starvation
To evaluate the therapeutic e cacy of primed-MSCs, H4 Swedish cells (H4SWs) were co-cultured with primed MSCs for 24 h (Fig. 1A). Apoptosis was observed when the H4SW cells were in the serumstarvation state for 24 h (H4SW only). However, when naïve MSCs or primed MSCs were co-cultured with H4SW cells, cell death was inhibited ( Fig. 2A). Following co-culture, the number of viable cells was counted. Compared to the H4SW cells-only group, more viable cells were observed in the + naïve MSCs and + primed MSCs groups (Fig. 2B). The anti-apoptotic effects in the H4SW cell model were the highest in the + primed MSC group. Next, Western blot analysis was performed to con rm the anti-apoptotic effect of naïve and primed MSCs. The expression of cell death markers, cleaved PARP, and cleaved caspase-3 was decreased when H4SW cells were co-cultured with naïve MSCs and primed MSCs (Fig. 2C, D). Based on the densitometric analysis, the levels of cleaved PARP and caspase-3 were signi cantly decreased in the + naïve MSC (2.2-and 1.4-fold changes, respectively) and + primed MSC groups (2.7and 1.8-fold changes, respectively). From these results, we con rmed that primed MSCs exhibited stronger anti-apoptotic effects on H4SW AD cells than naïve MSCs in the in vitro model.

Primed MSCs show in vitro therapeutic effects against
Alzheimer's disease Next, we performed Western blot analysis to con rm the therapeutic e cacy of primed MSCs on AD pathology, especially Aβ and ubiquitin conjugates. Aβ is the most well-known pathological hallmark of AD. Ubiquitin conjugates are negatively correlated with 26S proteasome activity, which means that impaired 26S proteasome activity results in the accumulation of Aβ, hyper-phosphorylated tau, and ubiquitin conjugates in the AD brain. Therefore, along with Aβ, the level of ubiquitin conjugates was measured in this study as another hallmark of AD.
Primed MSCs were co-cultured with H4SWs in vitro for 24 h to evaluate the therapeutic e cacy against AD symptoms (Fig. 3). Following co-culture, the level of Aβ in the conditioned media was measured by ELISA (Fig. 3A). Secreted Aβ was signi cantly reduced under the + primed MSC condition compared to the control H4SW cells (1.6-fold change). However, naïve MSCs did not show a statistically signi cant anti-Aβ effect. Next, the cumulative changes in the levels of ubiquitin conjugates were analyzed by Western blots (Fig. 3B) and the intensity of the bands was quanti ed (Fig. 3C). In the AD in vitro model (H4SW cells), more ubiquitin conjugates accumulated in the cytosol than in the normal cell line (H4). However, the level of ubiquitin conjugates was signi cantly attenuated in both + naïve MSCs (1.2-fold change) and + primed MSCs (1.4-fold change). In particular, primed MSCs showed enhanced therapeutic effects by attenuating ubiquitin conjugate accumulation. This demonstrated that primed MSCs successfully reduced the level of Aβ and ubiquitin conjugates in the AD in vitro model and that this effect was better than that of naïve MSCs.
In addition, the differences in gene expression (APPSW, BACE1, and IGFBP3) in H4 and H4SW cells cocultured with naïve MSCs or primed MSCs were analyzed. The analysis revealed that the dysregulated genes in the H4SW AD in vitro model were altered toward normal conditions (H4 cells) after co-culture with naïve MSCs and primed MSCs. Between the two MSCs, primed MSCs showed better alteration ( Supplementary Fig. 1).

Evaluation of therapeutic e cacy of primed MSCs in 5xFAD mice
To evaluate the e cacy of primed MSCs in AD, we performed an in vivo experiment using 5xFAD AD transgenic mouse. The experimental animals (12 months old) were divided into four groups: wild-type control (WT), transgenic control (sham), naïve-MSC, and primed MSC. We injected 1 × 10 5 WJ-MSCs into the right lateral ventricle. One week after injection, the mice were euthanized and brain tissues were harvested. First, the anti-apoptotic effect of primed MSCs was assessed by cleaved caspase-3 Western blot analysis (Fig. 4A). When compared to the WT mice, the 5xFAD mice showed increases in cleaved caspase-3, indicating neuronal death in the brain, whereas both the naïve MSCs and primed MSCs signi cantly reduced cleaved caspase-3 levels in the brain. Next, Aβ accumulation in the brain was measured by Western blots and thio avin-S staining (Fig. 4B). Compared to the WT control, the deposition of Aβ in the brain was observed in 5xFAD mice. The groups injected with naïve MSCs and primed MSCs showed decreases in Aβ accumulation. Primed MSCs, in particular, attenuated Aβ accumulation more effectively than naïve MSCs, which was con rmed by thio avin-S staining. Thio avin-S staining (Fig. 4C) revealed extensive Aβ (green) deposits in the cortex and hippocampal regions of the 5xFAD transgenic mouse control group (sham). Strikingly, the amount of Aβ in the cortex and hippocampus was reduced in the groups injected with naïve and primed MSCs, and the primed-MSC group showed better therapeutic e cacy.
Then, we quanti ed the number of naïve MSCs or primed MSCs in the 5xFAD brains via real-time quantitative PCR analysis using a human-speci c ALU primer (Fig. 4E). The absolute number of MSCs was determined based on the standard curve (linear regression R 2 = 0.992, Fig. 4D). Approximately 2000 remaining cells were found in the mice injected with naïve MSCs, whereas increased numbers of primed MSCs were detected (2.4-fold change). Based on these results, the primed MSCs showed both enhanced therapeutic effects and increased cell survival in vivo.

Primed MSCs differ from naïve MSCs in mRNA expression
An RNA microarray was performed to identify the changes in mRNA expression in the primed MSCs (Fig. 5) and a scatterplot was derived from the raw data (Fig. 5A). In Fig. 5A, the upregulated genes in the primed MSCs were compared to the naïve MSCs and are shown in red and the downregulated genes are shown in blue. The Euclidean distance clustering of the signi cant genes analyzed by MeV software is presented as log-transformed data in Fig. 5B. The 38 upregulated genes were clustered as upregulated.
Among these genes, we screened TGF-β, whose expression was increased over 3.0-fold in primed MSCs compared to the levels in naïve MSCs. Furthermore, the upregulation of TGF-β expression in primed MSC was con rmed via quantitative real-time PCR. The results showed that the primed MSCs expressed TGF-β at levels 3.2-fold higher than those in naïve MSCs (Fig. 5C).
These results demonstrated that TGF-β, which is highly secreted by primed MSC, could be a key molecule for therapeutic e cacy on AD.

Therapeutic e cacy of TGF-β in 5xFAD mice
To determine the role of TGF-β, especially in anti-apoptosis and anti-Aβ, recombinant protein was injected into the lateral ventricle of 5xFAD mice followed by euthanasia one-week later (Fig. 6). Western blot analysis revealed that cleaved caspase-3 was increased in the 5xFAD mice compared to the WT controls but was signi cantly decreased in the TGF-β group (Fig. 6A). Next, the anti-Aβ effect of TGF-β was measured by Aβ Western blot analysis and thio avin-S staining (Figs. 6B and C). The deposition of Aβ in 5xFAD mice was reduced following treatment with TGF-β (Fig. 6B). However, this observation was not replicated and no statistical signi cance was observed in the histological analysis of thio avin-S staining (Fig. 6C).

SRGN secretion by H4SW cells: A potential preconditioning factor inducing AD
Next, which molecule caused primed MSCs to secrete TGF-β was investigated. To identify the potential candidates responsible for priming the MSCs, the gene expression pro les of H4 and H4SW cells were analyzed (Fig. 7). The red dots in the gure denote increased mRNA expression of the H4SW cells compared to the H4 cells, and the green dots indicate decreased expression (Fig. 7A). A total of six genes highly upregulated in the H4SW cells were selected and clustering using the Euclidean distance measurements of signi cant genes was conducted (Fig. 7B). Next, the amount of secreted serglycin (SGRN) proteins was measured in the conditioned media. In the H4SW cells, the level of SGRN protein was signi cantly elevated compared to the H4 cells, suggesting that the SGRN protein may represent the AD microenvironment and potentially act as the main inducer of primed MSCs.

SGRN is the main effector of primed MSCs
To con rm whether SGRN was the main inducer of primed MSCs, various concentrations of SGRN protein were used to treat naïve MSCs (Fig. 8). After the treatment of the naïve MSCs with SGRN for 24 h, the TGF-β mRNA expression in naïve MSCs was measured via quantitative real-time PCR analysis. Signi cant increases in TGF-β mRNA expression were observed, except at 10 ng/mL (Fig. 8A). The peak was observed at 2 mg/mL SGRN treatment. Next, the therapeutic potential of SGRN-treated MSCs (SGRN MSCs) was brie y assessed (Fig. 8B). H4SW cells were co-cultured with naïve MSCs or SRGN MSCs for 24 h and then the CCK assay was conducted to con rm the anti-apoptotic effect of naïve and SRGN MSCs on H4SW cells. Cell death was signi cantly inhibited when H4SW cells were co-cultured with naïve MSCs and SGRN MSCs. This suggests that SGRN secreted by H4SW cells or the AD microenvironment is an inducer of primed MSCs.

Discussion
Recent advances have demonstrated the promising therapeutic role of MSCs in AD [12,17]. Because AD remains a major cause of morbidity and mortality, signi cant effort has been directed toward Aβ removal via stem cell transplantation [13,28]. The therapeutic properties of MSCs are largely related to their antiapoptotic and anti-in ammatory abilities, which have been con rmed both in vivo and in vitro [13,29,30]. However, the low survival rates of MSCs in vivo are a challenge and the bene ts of MSCs are mediated by unde ned mechanisms [31][32][33].
Various modi cations of MSCs have been attempted to improve their survival rates and therapeutic e cacy [31,32,34,35]. Attempts to improve stem cell survival, metabolism, or migration ability have focused on genetic modi cations to knock-out or knock-in speci c genes [36][37][38]. However, the clinical application of genetically-modi ed MSCs is associated with the risk of unexpected genetic mutations resulting in tumor formation [39]. In another approach, biocompatible scaffolds as an alternative to encapsulated MSCs have been developed to improve the survival and engraftment rates [40]. This method facilitated clinical application but did not improve the e cacy of MSCs.
In recent years, pre-conditioning methods that attempted to improve the e cacy of MSCs have also been in the spotlight [41][42][43]. Pre-conditioning aims to promote cell proliferation [43], improve migratory ability [43], and enhance protein secretion [44]. Unlike genetic modi cations, pre-conditioning can be achieved by exposing MSCs to speci c microenvironments. Compared with genetic modi cations, pre-conditioning enhanced therapeutic e cacy while maintaining the genotype of the cells [45]. A number of approaches have been proposed to make pre-conditioned MSCs. Pre-conditioning by hypoxia [46], in ammatory stimuli [42,45], or other factors [42] are strategies designed to enhance the survival and effectiveness of MSCs post-transplantation. In this study, we pre-conditioned MSCs using Aβ, the most important hallmark of Alzheimer's disease and used H4SW cells for pre-conditioning through endogenous Aβ.
H4SW cells are a stable cell line whereby the amyloid precursor protein (APP) Swedish mutation was introduced into a human glioblastoma cell line. APP is an integral membrane protein of neuronal cells involved in synaptic formation, synpatic plasticity, and ion export. APP, expressed in cell membranes, is usually cleavaged byα-secretase. However, mutations in APP protein or PSEN1/PSEN2 increases the change for APP to be cleavaged by β-, and γ-secretase, resulting in high levels of Aβ production in the brain. The Aβ produced is considered the causative substance of Alzheimer's disease, as it forms oligomer aggregates and Aβ plaques, resulting in neuronal toxicity and ultimately, the death of neuronal cells. APP Swedish, which is adjacent to the β-secretase site in APP, is one of the well-known genetic mutations in familial Alzheimer's disease, resulting in increased total Aβ production [47][48][49]. Therefore, the research model for AD with an APP Swedish mutation is now widely used [50][51][52][53], and the H4SW cell line is called the AD in vitro model [20]. Moreover, Aβ accumulated in the brain of AD patients activates glia cells, which are known to eliminate Aβ and have neuroprotective effects [54][55][56][57]. MSCs do not remove Aβ itself when exposed to AD but secrete proteins that can stimulate neurons or glial cells through paracrine action [58]. Therefore, we propose that the H4SW cell line was suitable for this study because the therapeutic e cacy of MSCs can be evaluated by measuring the reduction in Aβ deposits by stimulated H4SW cells.
When H4SW cells were co-cultured with primed MSCs, decreases in the level of Aβ and ubiquitin conjugates were observed in the H4SW cells (Figs. 2 and 3). In addition, when primed MSCs were administered directly into the brain of 5xFAD mice, an AD in vivo model, primed MSCs showed the therapeutic effects of suppressing neuronal death and promoting Aβ clearance (Fig. 4). Messenger RNA sequencing con rmed that SGRN secreted by H4SW cells promoted TGF-β secretion by MSCs, TGF-β protein had the same anti-cell death and anti-Aβ effects as primed MSCs, and SRGN-treated MSCs showed anti-cell death effects (Figs. 5,6,7,8). It is known that the secretion of SRGN is increased when an in ammatory reaction occurs [59]. Heparin sulfate proteoglycan, which contains SRGN, was responsible for promoting the brillization of Aβ and tau proteins [60]. Interestingly, it was also reported that SRGN gene expression and protein expression were signi cantly increased in AD patients compared to normal controls [61]. Thus, SRGN may be thought of as a possible biomarker for AD, suggesting that the pre-conditioning of SRGN in MSCs may be a possible to generate enhance MSC for AD treatment. Additionally, TGF-β is highly expressed in primed MSCs or SRGN-treated MSCs, and the signaling pathway associated with TGF-β is impaired in AD [62] and TGF-β itself showed neuroprotective effects [63]. Therefore, the results of this and previous studies suggest that TGF-β, highly secreted by primed MSCs, can have therapeutic e cacy in AD.
A particularly noteworthy nding is that when MSCs were exposed to an AD microenvironment, SRGN secreted locally in the Alzheimer's brain was recognized by the MSCs, which were induced to increase the expression of TGF-β, promoting therapeutic e cacy. As far as we know, this is the rst study to generate pre-conditioned MSC using a possible biomarker for the target disease. Like the concept of vaccination, we can make MSCs in a ready-to-ght state, promoting the secretion of effective proteins by exposing them to the target disease microenvironment in advance.
Our study had several limitations. First of all, the exact mechanism of action of SGRN, MSC, and TGF-β was not elucidated. Second, the recovery of cognition in the AD in vivo model was not studied. After the injection of primed MSCs, TGF-β, or SRGN MSCs, a long-term follow-up must be observed. Finally, the optimization of signaling factors and their combinations used in MSC preconditioning requires further investigation. Studies based on preconditioned MSCs should be conducted to enhance the therapeutic capacity of MSCs and expand the platform developed in this study.

Conclusions
In summary, we report that AD environmental preconditioning is a promising strategy to reduce cell death and ubiquitin levels while maintaining the stemness and characteristics of MSCs. Further, these data suggest that human WJ-MSCs exposed to an AD cell model in vitro may represent a promising and novel therapy for AD.