PSEN1 E280A Cholinergic-like Neurons and Cerebral Spheroids Derived from Mesenchymal Stromal Cells and from Induced Pluripotent Stem Cells Are Neuropathologically Equivalent

Alzheimer’s disease (AD) is a chronic neurological condition characterized by the severe loss of cholinergic neurons. Currently, the incomplete understanding of the loss of neurons has prevented curative treatments for familial AD (FAD). Therefore, modeling FAD in vitro is essential for studying cholinergic vulnerability. Moreover, to expedite the discovery of disease-modifying therapies that delay the onset and slow the progression of AD, we depend on trustworthy disease models. Although highly informative, induced pluripotent stem cell (iPSCs)-derived cholinergic neurons (ChNs) are time-consuming, not cost-effective, and labor-intensive. Other sources for AD modeling are urgently needed. Wild-type and presenilin (PSEN)1 p.E280A fibroblast-derived iPSCs, menstrual blood-derived menstrual stromal cells (MenSCs), and umbilical cord-derived Wharton Jelly’s mesenchymal stromal cells (WJ-MSCs) were cultured in Cholinergic-N-Run and Fast-N-Spheres V2 medium to obtain WT and PSEN 1 E280A cholinergic-like neurons (ChLNs, 2D) and cerebroid spheroids (CSs, 3D), respectively, and to evaluate whether ChLNs/CSs can reproduce FAD pathology. We found that irrespective of tissue source, ChLNs/CSs successfully recapitulated the AD phenotype. PSEN 1 E280A ChLNs/CSs show accumulation of iAPPβ fragments, produce eAβ42, present TAU phosphorylation, display OS markers (e.g., oxDJ-1, p-JUN), show loss of ΔΨm, exhibit cell death markers (e.g., TP53, PUMA, CASP3), and demonstrate dysfunctional Ca2+ influx response to ACh stimuli. However, PSEN 1 E280A 2D and 3D cells derived from MenSCs and WJ-MSCs can reproduce FAD neuropathology more efficiently and faster (11 days) than ChLNs derived from mutant iPSCs (35 days). Mechanistically, MenSCs and WJ-MSCs are equivalent cell types to iPSCs for reproducing FAD in vitro.


Introduction
Alzheimer's disease (AD) is a progressive and chronic neurological condition characterized by loss of memory due to vulnerability and the severe loss of cholinergic neurons from the nucleus basalis magnocellular of Meynert and cholinergic projections to the cortex and hippocampus [1,2]. AD is biologically described not only by the intracellular accumulation of amyloid-beta (iAβ), extracellular amyloid-β (eAβ)-containing plaques, and intracellular hyperphosphorylated tau-containing neurofibrillary tangles [3], but also by dysfunction in the highly interrelated endosomal and lysosomal clearance pathways E280A ChLNs and CSs derived from iPSCs, MenSCs, and WJ-MSCs. We demonstrate for the first time that FAD PSEN1 E280A pathology can be recapitulated in WJ-MSC-and MenSCs-derived ChLNs and CSs in about 11 days, similar to PSEN1 E280A ChLNs derived from iPSCs-NPC in 35 days.

Wild-Type and PSEN 1 E280A MenSCs, and WJ-MSCs Express Comparable Cellular Pluripotential Markers as iPSCs
We wanted to first confirm that wild-type (WT) and mutant iPSCs, MenSCs, and WJ-MSCs displayed typical cellular pluripotential markers. As shown in Figure 1, WT ( Figure  1A) and PSEN 1 E280A iPSCs ( Figure 1B), MenSCs ( Figure 1C,D), and WJ-MSCs ( Figure  1E,F) expressed not only the pluripotent transcription factor octamer binding transcription factor 4 (OCT4, Figure 1A-F), but also the transcription factor Sex determining Region Y-box 2 (SOX 2, Figure 1G-L), Homeobox transcription factor Nanog (NANOG, Figure  1M-R), and Krüppel-like factor 2 (KLF, Figure 1S-X), which are essential for pluripotency and self-renewal of stem cells ( Figure 1Y,Z,AA,1AB). As expected, WT and PSEN 1 E280A iPSCs, MenSCs, and WJ-MSCs differentiated into mesoderm ( Figure 1AC-AH), and ectoderm ( Figure 1AI-AN) germinal layers during embryonic development according to the presence of vimentin and Nestin markers, respectively; however, the endoderm layer was only present in iPSCs ( Figure 1AO-AP) but absent in both WT and mutant MenSCs and WJ-MSCs ( Figure 1AQ-AT) according to the endodermal marker C-X-C chemokine receptor type 4 (CXCR4), which is a specific marker for stromal-derived-factor-1.  The figures represent one out of three independent experiments. The data are expressed as the mean ± SD; significant values were determined by one-way ANOVA with Tukey's post hoc test; ns: not significant. Image magnification, 20×. Representative immunocytochemistry images of mesoderm germ layer stained for Vimentin (AC-AH), ectoderm stained for ectoderm stained for Nestin (AI-AN), and endoderm stained for CXCR4 (AO-AT) in WT and PSEN 1 E280 A mutation from iPSCs, MenScs, and WJ-MSCs. Nuclei are stained with Hoechst (blue). Scale bars 25 µm.

Wild-Type and PSEN 1 E280A MenSCs and WJ-MSCs Express Neuronal Stem Markers as iPSC-Derived NPC
We initially generated WT and PSEN 1 E280A iPSC-derived neural progenitor cells (NPC) using a well-established stepwise protocol lasting 27 days (Figure 2A,B). The iPSCs ( Figure 2C,D) were successively exposed to different culture formulae (Section 4) to obtain embryonic bodies (EB, Figure 2E,F), pre-NPC ( Figure 2G,H), and NPC ( Figure 2I,J). No evident morphological differences were observed between WT and mutant cells at the different stages of cellular differentiation process (Figure 2C,E,G,I versus Figure 2D,F,H,J).
Then, we evaluated whether WT and PSEN 1 E280A iPSC-derived NPC, MenSCs, and WJ-MSCs expressed neuronal stemness protein markers. Effectively, iPSC-derived NPCs ( Figure 3A  . The figures represent one out of three independent experiments. The data are expressed as the mean ± SD; significant values were determined by one-way ANOVA with Tukey's post hoc test; ns: not significant. Image magnification, 20×. Representative immunocytochemistry images of mesoderm germ layer stained for Vimentin (AC-AH), ectoderm stained for ectoderm stained for Nestin (AI-AN), and endoderm stained for CXCR4 (AO-AT) in WT and PSEN 1 E280 A mutation from iPSCs, MenScs, and WJ-MSCs. Nuclei are stained with Hoechst (blue). Scale bars 25 µm.

Wild-Type and PSEN 1 E280A MenSCs and WJ-MSCs Express Neuronal Stem Markers as iPSC-Derived NPC
We initially generated WT and PSEN 1 E280A iPSC-derived neural progenitor cells (NPC) using a well-established stepwise protocol lasting 27 days (Figure 2A,B). The iPSCs ( Figure 2C,D) were successively exposed to different culture formulae (Section 4) to obtain embryonic bodies (EB, Figure 2E,F), pre-NPC ( Figure 2G,H), and NPC ( Figure 2I,J). No evident morphological differences were observed between WT and mutant cells at the different stages of cellular differentiation process (Figure 2C,E,G,I versus Figure 2D,F,H,J).

WT and PSEN 1 E280A MenSCs and WJ-MSCs Can Transdifferentiate into ChLNs and CSs Similarly to iPSC-Derived NPC
We further cultured WT and mutant iPSC-induced NPC ( Figure 4A (Figures 2A and 4G) to obtain cholinergic-like neurons (ChLNs, Figure 4H-M), and assessed them by flow cytometry analysis ( Figure 4N-P) and immunofluorescence microscopy ( Figure 4Q-Y). The highest percentage of ChLNs was obtained from both WT and mutant WJ-MSCs ( Figure 4P), whereas iPSCs ( Figure 4N) and MenSCs ( Figure 4O) showed no difference. Similar observations were recorded by immunofluorescent microscopy ( Figure 4Q-Y).
When the NPC (obtained at day 27) and MSCs were exposed to Fast-N-Spheres V2 medium ( Figures 2B and 5A), cerebral spheroids (CSs) were clearly identified ( Figure 5B-G) in a process that lasted eight additional days and only 11 days for MenSCs and for WJ-   (Figures 2A and 4G) to obtain cholinergic-like neurons (ChLNs, Figure 4H-M), and assessed them by flow cytometry analysis ( Figure 4N-P) and immunofluorescence microscopy ( Figure 4Q-Y). The highest percentage of ChLNs was obtained from both WT and mutant WJ-MSCs ( Figure 4P), whereas iPSCs ( Figure 4N) and MenSCs ( Figure 4O) showed no difference. Similar observations were recorded by immunofluorescent microscopy ( Figure 4Q-Y).
When the NPC (obtained at day 27) and MSCs were exposed to Fast-N-Spheres V2 medium ( Figures 2B and 5A), cerebral spheroids (CSs) were clearly identified ( Figure 5B-G) in a process that lasted eight additional days and only 11 days for MenSCs and for WJ-MSCs. To further verify the cholinergic phenotype in the CSs, WT and mutant MenSCs-, WJ-MSCs-, and NPC-derived cholinergic cells were assessed for expression of neuronal markers. Figure 5 shows co-expression of the neuronal marker ChAT/VAChT/β III tubulin detected in NPCs-derived CSs ( Figure 5H,I), MenSCs-derived CSs ( Figure 5J,K), and WJ-MSCs-derived CSs ( Figure 5L,M), showing that mutant cells express less β III tubulin ( Figure 5N) and ChAT ( Figure 5P) compared to WT, but no difference was found in the expression of VAChT ( Figure 5O). MSCs. To further verify the cholinergic phenotype in the CSs, WT and mutant MenSCs-, WJ-MSCs-, and NPC-derived cholinergic cells were assessed for expression of neuronal markers. Figure 5 shows co-expression of the neuronal marker ChAT/VAChT/β III tubulin detected in NPCs-derived CSs ( Figure 5H,I), MenSCs-derived CSs ( Figure 5J,K), and WJ-MSCs-derived CSs ( Figure 5L,M), showing that mutant cells express less β III tubulin (Figure 5N) and ChAT ( Figure 5P) compared to WT, but no difference was found in the expression of VAChT ( Figure 5O).
Further analysis shows that the ∆Ψ m in ChLNs derived from WT PSEN 1 NPCs ( Figure 7A

PSEN 1 E280A ChLNs Derived from MenSCs, and WJ-MSCs Show Cell Death Markers of Apoptosis as Mutant NPC-Derived ChLNs
Next, we evaluated whether mutant ChLNs express apoptosis markers. Figure 8 shows that the protein PUMA was constitutively expressed at low levels in WT ChLNs derived from NPC ( Figure 8A  To evaluate whether ChLNs derived from NPCs, MenSCs, and WJ-MSCs were responsive to neurotransmitter stimuli as an assessment of cholinergic neuronal functionality, the WT and mutant ChLNs were exposed to acetylcholine (ACh), and Ca 2+ influx was recorded in fluorescent microscopy. As shown in Figure 9, ACh induced a comparable transient elevation of intracellular Ca 2+ in WT ChLNs derived from NPC ( Figure 9A), Men-SCs ( Figure 9D), and WJ-MSCs ( Figure 9G) with an average fluorescence change (DF/F) of 3.8 ± 0.6-fold, and a mean duration of 40 ± 10 s (n = 20 ChLN cells imaged, N = 3 dishes/each cell source) according to cytoplasmic Ca 2+ response to Fluo-3-mediated imaging ( Figure 9C,F,I). Remarkably, PSEN 1 E280A ChLNs obtained from the three cellular sources almost did not respond to ACh ( Figure 9B   Images were analyzed, and quantitative data were compared (C,F,I). The figures represent one out of three independent experiments. The data are expressed as the mean ± SD; significant values were determined by one-way ANOVA with Tukey's post hoc test; ** p < 0.005; *** p < 0.001. Image magnification 20×. White square area is magnified inset (A',B',D',E',G',H'), magnification 100×. values were determined by one-way ANOVA with Tukey's post hoc test; *** p < 0.001. Image magnification, 20×. Flow cytometry analysis of WT and PSEN 1 E2980A iPSCs::NPC-(Q,T), MenSCs-(R,U), and WJ-MSCs-derived ChLNs (S,V) to identify PUMA (Q-S), p-JUN (Q-S), TP53 (T-V), and CASP3 (T-V). The histograms represent 1 out of 3 independent experiments. The data are expressed as the mean ± SD; significant values were determined by one-way ANOVA with Tukey's post hoc test; *** p < 0.001.

PSEN1 E280A ChLNs Derived from WJ-MSCs Secrete Higher Amount of Extracellular Aβ42 Than Mutant ChLNs Derived from NPC and MenSC
Measurement of secreted eAβ42 (expressed as the ratio eAβ42/eAβ40) is critical for early detection and the disease-modifying treatments necessary to combat AD. Therefore, to assess the amounts of eAβ42 and eAβ40 secreted by both WT and mutant ChLNs derived from NPC, MenSCs, and WJ-MSCs, supernatants from culture medium were evaluated by a solid-phase sandwich ELISA according to the standard procedure described in the Section 4. The amount of secreted eAβ40 was almost constant in both WT and mutant MenSCs-and WJ-MSCs-derived ChLNs after 11 days (n = 3 dishes) as a response to ACh treatment (G,H) ACh was puffed into the culture at 0 s (arrow). Then, the Ca 2+ fluorescence of cells was monitored at the indicated times. Color contrast indicates fluorescence intensity: dark blue < light blue < green < yellow < red. (C,F,I) Normalized mean fluorescence signal (∆F/F) over time, indicating temporal cytoplasmic Ca 2+ elevation in response to ACh treatment in PSEN 1 WT and E280A cells. The data are expressed as the mean ± SD; significant values were determined by one-way ANOVA with Tukey's post hoc test; * p < 0.05; ** p < 0.005; *** p < 0.001. Image magnification, 20×.

PSEN1 E280A ChLNs Derived from WJ-MSCs Secrete Higher Amount of Extracellular Aβ42 Than Mutant ChLNs Derived from NPC and MenSC
Measurement of secreted eAβ42 (expressed as the ratio eAβ42/eAβ40) is critical for early detection and the disease-modifying treatments necessary to combat AD. Therefore, to assess the amounts of eAβ42 and eAβ40 secreted by both WT and mutant ChLNs derived from NPC, MenSCs, and WJ-MSCs, supernatants from culture medium were evaluated by a solid-phase sandwich ELISA according to the standard procedure described in the Section 4. The amount of secreted eAβ40 was almost constant in both WT and mutant PSEN 1 E280A ChLNs derived from NPC ( Figure 10A), MenSCs ( Figure 10B), and WJ-MSC ( Figure 10C) ChLNs. However, secreted eAβ42 was significantly higher in the mutant ChLNs when compared to WT ChLNs from all three sources ( Figure 10D-F). Interestingly, PSEN 1 E280A ChLNs derived from WJ-MSCs secreted greater amounts of eAβ42 ( Figure 10F, 133 ± 17 pg/mL) than mutant NPC ( Figure 10D, 53 ± 5 pg/mL), or mutant MenSCs ( Figure 10E, 31 ± 1 pg/mL). As a result, the ratio Aβ42/40 was consistently higher in PSEN 1 E280A ChLNs derived from WJ-MSCs ( Figure 10I, Aβ42/40 = 6) than the other two cellular sources ( Figure 10G, Aβ42/40 = 1.6; and Figure 10H, Aβ42/40 = 1.3). In other words, PSEN 1 E280A WJ-MSCs-derived ChLNs secreted 3.75-and 4.62-folds of Aβ42 compared to mutant ChLNs derived from NPC and MenSCs, respectively. Next, we wondered whether the accumulation of iAPPβ fragments, p-Tau, and oxidation of DJ-1 detected in 2D also occurred in 3D (CSs) structures. Figure 11 shows that no OS, iAPPβf or p-Tau markers were present in WT CSs from NPC ( Figure 11A,A ,A ,B,C,G-I,M-O). In contrast, PSEN 1 E280A NPC-, MenSC-, and WJ-MSCs-derived CSs displayed the typical iAPPβf ( Figure 11D,D ,E,F), OS ( Figure 11D,D ,E,F), and p-Tau ( Figure 11J,J ,K,L) markers, albeit with different levels of intensity ( Figure 11M-O). While mutant WJ-MSCS-derived CSs generated higher levels of iAPPβf i.e., in terms of MFI signal, compared to mutant NPC-derived CSs ( Figure 11M), mutant NPC-derived CSs showed significantly higher levels of oxidized DJ-1 protein ( Figure 11N) compared to either mutant MenSCs-, or mutant WJ-MSCs-derived CSs. Yet, mutant WJ-MSC-derived CSs showed higher levels of p-Tau compared to mutant NPC-derived CSs ( Figure 11O).

Discussion
Given their resemblance to embryonic stem cells (ESCs), human iPSCs [31] have been instrumental for AD modeling in vitro [32] to understand the underlying mechanisms of the disease (e.g., [33]), test potential drugs [34], and develop personalized therapies [35]. Here, we obtained iPSCs from a WT PSEN 1 subject and from a patient with FAD bearing the mutation PSEN 1 E280A and used both as reference 2D and 3D tissue cultures for revealing the natural neuropathology of FAD and for comparative purposes. Like iPSCs, we established that multipotent MenSCs and WJ-MSCs: (i) expressed the pluripotent-associated markers OCT4, SOX2, NANOG, and KLF4; (ii) differentiated into cells of ectoderm and mesoderm germ layers; (iii) highly expressed neuronal stem marker nestin, and neural stem and progenitor cell marker SOX2; (iv) transdifferentiated into ChLNs, which expressed the cholinergic marker ChAT/VAChT (56 ± 6 & 79 ± 4%); and (vi) transdifferentiated into CSs. Taken together, these observations suggest that MenSCs and WJ-MSCs might be developmentally equivalent to iPSCs. Interestingly, the protocol to obtain ChLNs and CSs from MenSCs and WJ-MSCs lasts 11 days, whereas the protocol for obtaining neurons from iPSCs takes no less than 35 days. Therefore, using both Cholinergic-N-Run [29] and Fast-N-Spheres V2 [30] medium protocols, at least 24 days are free from laboratory work (i.e., a 68% reduction in labor time). It was concluded that under the present conditions, both culture media are highly efficient as inducers of ChLNs. Since cholinergic neurons that form the nucleus basalis of Meynert are the most vulnerable to AD pathology [36], and therefore more severely lost, we considered that most of the 2D and 3D ChLNs in culture are reasonably homologous to ChNs in vivo.
In the present work, we report for the first time that iPSC::NPC-derived (planar or 2D culture) PSEN 1 E280A ChLNs and (3D culture) CSs displayed the neuropathological markers of AD, such as iAPPβf, and the secretion of high amounts of eAβ and p-TAU. Similar observations were recorded in 2D and 3D PSEN 1 E280A ChLNs derived from MenSCs (this work) and WJ-MSCs [29,30]. These observations suggest that the abnormal intracellular accumulation of Aβ seen in PSEN 1 E280A ChLNs and CSs is the earliest pathological event of a continuous process from an initial accumulation of iAPPβf to TAU phosphorylation [37] to the well-established extracellular Aβ aggregation, culminating in the formation of amyloid plaques [38] and cholinergic cell death. It was concluded that iAPPβf plays an important role in the genesis of PSEN 1 FAD [5]. Although several drugs that remove eAβ have failed to demonstrate clinical efficacy, including PSEN 1 E280A-derived Aβ [39], our in vitro data suggest that other alternative therapies against iAβ [40] or Tau protein [41] might be tested. Interestingly, we found oxidation of sensor protein DJ-1 (Cys 106 -SO 3 ) concomitantly with iAPPβf accumulation in PSEN1 E280A iPSCs::NPC-derived ChLNs and CSs. Like mutant iPSCs::NPC-derived neurons, mutant ChLNs derived from MenSCs and WJ-MSCs also consistently showed DJ-1 (Cys 106 -SO 3 ) together with iAPPβf. However, how exactly these two phenomena relate to each other is not yet established. One possible explanation is that iAPPβf directly or indirectly causes mitochondrial electron transport disruption [42,43], thereby bursting into ROS production (e.g., H 2 O 2 ), which in turn might specifically oxidize DJ-1 [44]. In accordance with this assumption, we found that mutant ChLNs and CSs generated ROS (H 2 O 2 ), and induced loss of ∆Ψ m concomitantly with Cys 106 -SO 3 . Whatever the connection might be, we consistently found that neurons expressed amyloidogenic Aβ, mitochondrial damage, ROS generation, and OS in planar ChLNs and (3D) CSs derived from the three biological sources (iPSCs::NPC, MenSCs, WJ-MSCs). Given that DJ-1 is not only an important stress sensor protein, but also modulates several signaling cellular pathways [45], it becomes a potential biomarker and therapeutic target in AD [46,47].
Interestingly, PSEN 1 E280A ChLNs derived from iPSCs::NPC, MenSCs, and WJ-MSCs reliably showed up-regulation of JUN, TP53, PUMA, and activation of CASP3, all involved in intrinsic apoptosis [48]. Accordingly, the iAPPβf generated H 2 O 2 either oxidizes DJ-1 or can function as a second messenger [49], thereby activating other redox signaling proteins (e.g., apoptosis signal-regulating kinase 1 [50,51]), in turn activating the c-Jun N-terminal Kinase (JNK) pathway [52]. Remarkably, JNK can activate the transcription factor JUN [52], the transcription factor TP53 [53], and phosphorylate the protein Tau [54]. Therefore, JNK might also be a potential therapeutic target for AD [55]. Remarkably, both JUN [56] and TP53 [57,58] transactivate BH3-only protein PUMA, which in turn directly or indirectly induces loss of ∆Ψ m through the Bcl-2 pro-apoptotic protein Bax [59]. This last event on mitochondria leads to the release of apoptogenic proteins [60] and the activation of the cellular end executer protein CASP3, which is responsible for neuronal dismantling. Taken together, these results comply with the notion that iAPPβf induces a cascade of events involving OS-signaling, mitochondrial depolarization, p-Tau, and apoptosis [29,61,62] in ChLNs derived from the three biological sources examined. Interestingly, it has been shown that PSEN1 E280A ChLNs derived from WJ-MSCs did not respond to ACh-induced Ca 2+ influx, most probably due to eAβ42 interaction with nicotinic acetylcholine receptors [63]. We report for the first time that iPSCs::NPC-and MenSCs-derived PSEN 1 E280A ChLNs were resilient to ACh-induced Ca 2+ influx. Together, these results suggest that eAβ42 affects neuronal Ca 2+ flux in PSEN 1 E280A independently of the cellular source tested.
Despite the use of iPSCs, these cells present several disadvantages. For instance, reprogramming highly depends on the efficient delivery and the suitable expression of Yamanaka factors, i.e., OSKM, into specific cell types (e.g., fibroblasts, cord blood CD133+ cells, peripheral blood mononuclear cells). Furthermore, the protocol(s) for obtaining reprogrammed cells is rather a slow and vulnerable process that may be affected by several factors (e.g., biopsy, cell type, particular culture conditions, long culture period) that hinder the efficiency, reproducibility, and quality of the resulting iPSCs (e.g., [64][65][66]). Therefore, replacement of iPSCs with a more manageable, cost-effective, time-saving biologic source that offers similarities to iPSCs is highly desirable.
In conclusion, we have demonstrated that PSEN 1 E280A ChLNs and cholinergic CSs derived from MenSCs, WJ-MSCs, and iPSCs can reliably reproduce the neuropathology of FAD in vitro, and therefore, they are cellularly and biochemically equivalent. Accordingly, PSEN 1 E280A iPSCs can be interchangeable with PSEN 1 E280A MenSCs and PSEN 1 E280A WJ-MSCs. Furthermore, they expressed the typical cellular hallmarks of AD, i.e., eAβ42 and p-TAU. Additionally, the presence of iAβ, oxidative markers DJ-1 Cys 106 -SO 3 and p-JUN, loss of ∆Ψ m , apoptosis markers TP53, PUMA, and CASP3, and dysfunctional ACh-induced Ca 2+ influx were observed in mutant ChLNs and CSs in 11 days, whereas at least 35 days of culture are necessary to reproduce AD markers from mutant iPSCs::NPCs. This labor time-gap in favor of MenSCs and WJ-MSCs makes these biological sources much more attractive not only for modeling FAD but also for speeding up drug discoveries (e.g., antioxidant compounds [67]). However, to fully validate our findings, dissecting transcriptomic signatures of cholinergic neuronal differentiation using PSEN 1 E280A iPSCs (as reference tissue [68]), MenSCs and WJ-MSCs is warranted. Whatever the cause, cholinergic degeneration remains one of the earliest, most severe, and most consistent cellular changes in AD. Therefore, studying the cellular and molecular changes in cholinergic neurons may provide clues to the pathogenesis and treatment of this disorder [69]. Moreover, MSC-based therapy might be a promising alternative for the treatment of AD [70,71].

Evaluation of Intracellular Hydrogen Peroxide (H 2 O 2 ) by Fluorescence Microscopy
The levels of intracellular H 2 O 2 , were determined according to ref. [29]. Briefly, cells were left in neural medium (NM) for 0 and 4 days or in minimal culture media (mCM) for 4 days. Then, the cells (5 × 10 3 ) were incubated with 2 ,7 -dichlorofluorescein diacetate reagent (5 µM, DCFH 2 -DA; Cat# D399, Invitrogen, Waltham, MA, USA) for 30 min at 37 • C in the dark. Cells were then washed, and DCF fluorescence intensity was determined by analysis of fluorescence microscopy images. The assessment was repeated three times in independent experiments, blind to the experimenter.

Analysis of Mitochondrial Membrane Potential (∆Ψ m ) by Fluorescence Microscopy
The ChLNs were left in NM for 0-4 days or in mCM for 4 days. Then, the cells (5 × 10 3 ) were incubated with the passively diffusing and active mitochondria-accumulating dye deep red MitoTracker compound (20 nM, final concentration) for 20 min at RT in the dark (cat# M22426, Invitrogen, Waltham, MA, USA) according to ref. [29]. The assessment was repeated three times in independent experiments.

Intracellular Calcium Imaging
Intracellular calcium (Ca 2+ ) concentration changes evoked by cholinergic stimulation were assessed according to refs. [72,73] with minor modifications. Briefly, for the measurement, the fluorescent dye Fluo-3 (Fluo-3 AM; cat: F1242 Thermo Fisher Scientific, Santa Fe, NM, USA) was employed. Intracellular Ca 2+ transients were evoked by acetylcholine (Ach; Cat# A2661, Sigma-Aldrich, St. Louis, MO, USA); 1 mM final concentration) at 4 days post differentiation. The amplitudes of the Ca 2+ -related fluorescence transients were expressed relative to the resting fluorescence (∆F/F) and were calculated by the following formula: ∆F/F = (F maximum − F resting )/(F resting − F background ). For the calculation of the fluorescence intensities, ImageJ was used. The assessment was repeated three times in independent experiments, blind to the experimenter.

Measurement of Aβ 1-40 and Aβ 1-42 Peptides in Culture Medium
The levels of Aβ 1-40 and Aβ 1-42 peptides were measured according to a previous report with minor modifications. Briefly, WT and PSEN1 E280A cells were left in NM or mCM for 4 days. Then, 100 µL of supernatants were collected, and the levels of secreted Aβ1-40 and Aβ1-42 peptides were determined by a solid-phase sandwich ELISA (Cat# 150496 and Cat# 574166, respectively, Abbexa, Cambridge, UK) following the manufacturer's instructions. The assessment was repeated three times in independent experiments, blind to the experimenter.

Data Analysis
In this experimental design, two vials of each specimen/origin were thawed (PSEN1-WT and -E280A), cultured, and the cell suspension was pipetted at a standardized cellular density of 2.6 × 10 4 cells/cm 2 into different wells of a 24-well plate. Experiments were conducted in triplicate wells, according to ref. [29]. The data from individual replicate wells were averaged to yield a value of n = 1 for that experiment, and this was repeated on three occasions, blind to the experimenter and/or flow cytometer analyst, for a final value of n = 3 [74]. The data were analyzed according to ref. [29]. The statistical significance was determined by one-way analysis of variance (ANOVA) followed by Tukey's post hoc comparison calculated with GraphPad Prism 5.0 software (https://www.graphpad.com/; accessed on 5 February 2023). Differences between groups were only deemed significant when a p-value of <0.05 (*), <0.001 (**) and <0.001 (***). All data are illustrated as the mean ± S.D.