Impact of Suramin on Key Pathological Features of Sporadic Alzheimer’s Disease-Derived Forebrain Neurons

Abstract

Background:

Alzheimer’s disease (AD) is characterized by disrupted proteostasis and macroautophagy (hereafter “autophagy”). The pharmacological agent suramin has known autophagy modulation properties with potential efficacy in mitigating AD neuronal pathology.

Objective:

In the present work, we investigate the impact of forebrain neuron exposure to suramin on the Akt/mTOR signaling pathway, a major regulator of autophagy, in comparison with rapamycin and chloroquine. We further investigate the effect of suramin on several AD-related biomarkers in sporadic AD (sAD)-derived forebrain neurons.

Methods:

Neurons differentiated from ReNcell neural progenitors were used to assess the impact of suramin on the Akt/mTOR signaling pathway relative to the autophagy inducer rapamycin and autophagy inhibitor chloroquine. Mature forebrain neurons were differentiated from induced pluripotent stem cells (iPSCs) sourced from a late-onset sAD patient and treated with 100μM suramin for 72 h, followed by assessments for amyloid-β, phosphorylated tau, oxidative/nitrosative stress, and synaptic puncta density.

Results:

Suramin treatment of sAD-derived neurons partially ameliorated the increased p-Tau(S199)/Tau ratio, and fully remediated the increased glutathione to oxidized nitric oxide ratio, observed in untreated sAD-derived neurons relative to healthy controls. These positive results may be due in part to the distinct increases in Akt/mTOR pathway mediator p-p70S6K noted with suramin treatment of both ReNcell-derived and iPSC-derived neurons. Longer term neuronal markers, such as synaptic puncta density, were unaffected by suramin treatment.

Conclusions:

These findings provide initial evidence supporting the potential of suramin to reduce the degree of dysregulation in sAD-derived forebrain neurons in part via the modulation of autophagy.

Keywords

Alzheimer’s disease autophagy induced pluripotent stem cells neurons suramin

INTRODUCTION

Central to current theories regarding the etiology of Alzheimer’s disease (AD) are two major neurophysiological features readily observed in a significant subset of advanced-stage AD patients: neurotoxic amyloid-β (Aβ) plaque accumulations and tau hyperphosphorylation, the latter of which results in neurofibrillary tangle formation [1]. Aggregation of these substrates contributes to chronic neuroinflammation and to the synaptic loss and neuronal degeneration associated with memory loss, cognitive deficits, and behavioral alterations. As such, significant efforts have been applied toward generating and evaluating Aβ- or phosphorylated tau (p-Tau)-targeting therapies. However, aside from recent clinical trials with the Aβ-antibody Lecanemab [2], Aβ- or p-Tau based strategies have predominantly yielded only mild deceleration of disease progress in clinical studies [3]. An increasing body of literature has therefore sought to characterize and address alternative mechanisms and pharmacological targets implicated in AD etiology.

Substantial research has recently been directed toward treating AD through targeting aberrant neuronal macroautophagy. Macroautophagy (hereafter referred to as “autophagy”) is an intricately orchestrated cellular mechanism involving the targeted sequestration of organelles and proteins into vesicles known as autophagosomes. These autophagosomes then fuse with lysosomes to degrade and recycle the sequestered material, thus contributing to proteostasis and healthy cellular function [4]. AD is associated with an accumulation of unresolved autophagosomes in dystrophic neurites, a phenomenon first discovered in post-mortem AD brains in 1967 [5]. This disruption in autophagic homeostasis is believed to underlie, in part, the abnormal Aβ and neurofibrillary tangle formation associated with AD [4, 6]. Disruptions in autophagy are also linked to increased oxidative/nitrosative stress [7, 8] and inflammatory cytokine production [9], both of which are heavily impacted in AD [10 –14]. Indeed, dysfunction of the Akt/mTOR pathway, a major regulatory nexus of autophagy [15 –17], is known to alter reactive oxygen/nitrogen species production through direct modulation of mitochondrial energy production and NADPH oxidase activation [7] and to alter inflammatory cytokine production (e.g., IL-8, IL-1β, TNFα, and IL-6) [9]. Thus, neuronal autophagy, and the Akt/mTOR pathway in particular, have been proposed as targets for AD intervention [18, 19].

Previous work examining the impact of autophagy modulators in AD has shown that the autophagy inhibitors chloroquine and hydroxychloroquine, which are known to impair lysosome formation, increase Aβ plaque deposition [20 –22] and tau pathology [4]. These results are consistent with an exacerbation of the unresolved autophagosomes associated with AD neurites [5]. In contrast, the autophagy inducer rapamycin, which can increase transport of autophagic cargo to the lysosomal system through inhibition of the Akt/mTOR pathway, appears to be able to prevent onset of AD in 3xTg-AD murine models [23]. Despite these promising results, however, rapamycin administration in this same murine model was unable to reverse existing Aβ plaques, neurofibrillary tangles, and cognitive deficits [23]. Thus, rapamycin may be ineffective in late-stage AD where lysosomal dysfunction is already established [24]. Cumulatively, these studies with chloroquine and rapamycin not only illustrate the potential of autophagy modulation in AD treatment but also highlight the need to investigate alternative autophagy modulation approaches.

In seeking an alternative autophagy modulator, we found the pharmacological agent suramin to be of particular interest. Not only has suramin been shown to influence Akt/mTOR-mediated autophagy [25] and autophagosome/lysosome processing [26, 27], it appears to do so in large part through direct modulation of two core systems consistently disrupted in AD: 1) P2 receptor-mediated purinergic signaling [28] and 2) sirtuin (SIRT) signaling [13, 29]. Specifically, suramin is a broad spectrum P2-receptor antagonist [28], and P2 receptor-based signaling triggered by purines (e.g., ATP, ADP, UTP, UDP) is known to play a profound role in a range of neuroinflammatory disorders [30, 31]. For instance, under conditions of cellular stress or injury, ATP is released into the extracellular milieu, and subsequently interacts with P2 purinergic receptors on a variety of cell types [32]. This interaction, particularly at elevated ATP concentrations, has been shown to elicit various cellular responses, including the activation or suppression of cellular autophagy [33, 34]. Furthermore, AD is associated with a number of changes in neuronal P2 receptor levels [33 –36]. Particularly prominent is an increased ratio of P2X7 to P2Y2 receptors, a change which is characteristic of a range of neuroinflammatory conditions [35 –37] and which has been associated with increased Aβ-mediated neurodegeneration [38, 39].

In addition to serving as a P2 receptor antagonist [28], suramin has recently been found to directly inhibit SIRTs 1,2, and 5 [13, 29], which are histone deacetylases intimately involved in many aging- and neurodegeneration-related disorders, including AD [29]. Indeed, growing evidence supports a central role for SIRTs in modulating neuroinflammation [40]. In particular, SIRT2, an abundant histone deacetylase in the brain [41], has been associated with increased neurodegeneration, and its inhibition has been shown to delay disease progression [42]. Importantly, SIRTs strongly regulate neuronal autophagy and associated mitochondrial function/metabolism and oxidative stress [43 –48], in part though modulating the Akt/mTOR pathway [46].

Despite the known role of purinergic signaling, SIRT activity, and autophagy dysregulation in AD, suramin has seen limited study as a potential AD intervention, and existing AD animal studies indicate a highly context-dependent response of AD pathology to suramin [36, 49]. This contextual dependence appears to be, in part, an issue of dose [36, 49] as well as an issue of variable effects of P2 receptor- and SIRT-inhibition depending on the identity of the P2 receptor or SIRT. For instance, P2X receptor inhibition is generally considered beneficial, and P2Y receptor inhibition, with the exception of P2Y1, non-beneficial, to neuronal health in AD [38, 50]. Similarly, SIRT2 inhibition has been associated with improved neuronal health in AD, whereas SIRT1 inhibition has been associated with increased accumulation of Aβ and tau [11 , 52]. Thus, the fact that suramin antagonizes both P2X and P2Y receptors and inhibits both SIRT1 and SIRT2 may lead to beneficial but also undesired responses, which would require more selective P2 receptor or SIRT antagonism to resolve. That said, a recent study in a Neuro2A AD murine model demonstrated a reduction in BzATP-induced α-secretase activity and subsequent Aβ deposition following low-dose (100μM) suramin treatment [36], indicating that appropriately-dosed suramin may show net benefit even in the complex AD milieu. In addition, suramin has a range of well-studied analogues and derivatives, several of which display increased specificity and therefore could potentially be employed in seeking to improve on observed beneficial effects of suramin [53 –56].

In the present work, we therefore propose to further characterize the effects of low-dose 100μM suramin on neuronal health with a focus on its effect on Akt/mTOR-mediated autophagy, oxidative/nitrosative stress, and Aβ_42/40 and p-Tau/Tau levels. This relatively low-dose of suramin avoids the cell death often associated with suramin levels greater than 200μM [57] while fully inhibiting P2X1, P2X2, P2X3, P2X5 and partially inhibiting P2X4 and P2X7 [55]; fully inhibiting P2Y1, P2Y2, P2Y11, P2Y12, P2Y14 and partially inhibiting P2Y6 [56, 58]; and fully inhibiting SIRTs 1, 2, and 5 [45, 59], thus serving as a starting point for dose refinement or analogue selection. Herein, the effects of suramin on the Akt/mTOR pathway will be placed within the context of standard autophagy modulators chloroquine and rapamycin using forebrain neurons derived from the well-characterized ReNcell human neural progenitor cell (NPC) line. We will also evaluate the capacity of 100μM suramin to attenuate key AD pathological features in forebrain neurons derived from induced pluripotent stem cells (iPSCs) sourced from an AD patient relative to a matched healthy control. In our iPSC studies, we have chosen to focus on late-onset sporadic AD (sAD) rather than familial AD given 1) that familial AD, by most estimates, represents less than 1% of all AD cases [60] and 2) that recent literature suggests that therapeutic approaches optimized for familial AD might not be efficacious for sAD [61].

METHODS

Comparative analysis of Akt/mTOR modulation by suramin, chloroquine, and rapamycin in ReNcell-derived forebrain neurons

ReNcell culture, differentiation, and treatment

To evaluate the effect of suramin on neuronal autophagic processes, ReNcells (Millipore Sigma), a well-characterized and widely-published NPC line, was subjected to neuronal differentiation according to the manufacturer’s protocol. In brief, ReNcells were plated at 2×10⁴ cells/cm² on growth factor reduced (GFR) Matrigel-coated cultureware in expansion medium containing Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F-12; Gibco), 1X B-27 supplement (Gibco), 2μg/mL heparin (Alfa Aesar), 20 ng/mL fibroblast growth factor-2 (FGF-2; PeproTech), and 20 ng/mL epidermal growth factor (EGF; PeproTech). On Day 0, differentiation was initiated by passaging with TrypLE Express and replating at 5×10⁴ cells/cm² on GFR Matrigel-coated cultureware in expansion medium without growth factors. Half-medium changes were performed every other day. On Day 21, a subset of neuron cultures were set aside for immunocytochemistry (ICC) validation. To characterize the effects of suramin treatment in the context of known autophagy modulators, Day 23 neurons were treated in parallel with either 100μM suramin (Tocris), 40μM chloroquine (New England Biolabs), or 500 nM rapamycin (New England Biolabs) for 24 h. The 24 h time point and the concentrations of suramin, rapamycin, and chloroquine were each selected to be relatively low-dose and to allow maximal contextualization within existing literature [62 –64]. In a subset of samples, suramin application was extended to 72 h to mimic intended treatment conditions for the sAD iPSC-derived neurons.

After the selected treatment time, ReNcell neurons were lysed via addition of ice-cold 0.5% Triton X-100 in DPBS + Halt^® Protease and Phosphatase Inhibitor (Thermo Scientific 78440), followed by gentle scraping. Lysates were then transferred to 1.7 mL microfuge tubes (VWR) and vortexed for 10 min. Finally, lysates were centrifuged (16,000×g, 20 min) at 4°C and the supernatants were extracted and stored at –80°C for subsequent proteomic analyses.

Immunocytochemistry analysis of ReNcell differentiation

ReNcell forebrain neuronal differentiation was assessed via ICC for forebrain neuron-specific transcription factor FOXG1, neuron-specific transcription factor NeuN, mature neuronal structural protein MAP-2, and astrocyte-specific marker S100β (Supplementary Figure 1). Following 21 days of neuronal maturation, cells were washed twice with DPBS (Corning 21-031-CV), then fixed in neutral-buffered formalin solution at room temperature (RT) overnight. The next day, cells were permeabilized using 0.1% Triton X-100 (Fisher Scientific BP151-100) in DPBS, then blocked with Background Terminator solution (BioCare Medical BT967L) (1 h, RT). Cells were then incubated in primary antibody solutions comprised of 5% bovine serum albumin (BSA; Fisher Scientific BP1600-100) in DPBS (overnight, 4°C). The following day, cells were then incubated in secondary antibodies (5% BSA in DPBS; 1 h, RT) and counterstained with DAPI (Invitrogen D3571). All incubation steps were performed using an orbital shaker (VWR) set to 80 RPM, with intervening wash steps, as appropriate. Fluorescence images were captured relative to negative controls using a Leica TCS SP8 STED confocal microscope using a 10X objective, then processed using ImageJ. Primary and secondary antibody information can be found in Supplementary Table 1.

Assessment of cell density and autophagy modulation via Akt/mTOR signaling pathway

The DNA content of the sample lysates was measured using the PicoGreen Assay (Invitrogen) as a standard indicator of net cell proliferation and loss across treatment groups. As an indicator of Akt/mTOR pathway-mediated autophagy (Fig. 1A), the levels of p-PTEN(Ser380), p-Akt(Ser473), p-mTOR(Ser2448), p-GSK3α/β [p-GSK3α(Ser21)+ p-GSK3β(Ser9)], and p-p70S6K(Thr412) were measured using a Millipore multiplex bead-based immunoassay. For each sample, net median fluorescence intensities corresponding to analyte concentrations were normalized to scaled DNA content.

Fig. 1

Suramin impacts the phosphorylation state of the Akt/mTOR pathway differentially from chloroquine and rapamycin. A) Schematic of the Akt/mTOR pathway adapted from Millipore. Evaluated pathway markers are shown in blue. B) PLS-DA of Akt/mTOR multiplex bead-based immunoassay results from ReNcell-derived neurons, n = 6–10 for each group. Untreated (CTL), suramin-treated (SURA), chloroquine-treated (CHL), and rapamycin (RAP)-treated ReNcell-derived neurons are shown in maroon, red, blue, and gold, respectively. Cluster centroids (means) are shown as squares. The elliptical space around each centroid represents the space correlating to 2 standard deviations from the centroid mean. Calculated Mahalanobis distances, Hotelling’s t-squared statistics, and factor loadings corresponding to this PLS-DA are presented in Supplementary Table 3. C) Relative expression of phosphorylated Akt/mTOR pathway markers identified as significantly contributing (>20%) to differences among groups along the Latent X1 axis in (B). D) Relative expression of phosphorylated Akt/mTOR pathway markers identified as significantly contributing (>20%) to differences among groups along the Latent X2 axis in (B). ^*, #, $ denote a significant difference from the CTL, SURA, and CHL groups, respectively (p < 0.05). Error bars represent SEM. White dots represent individual sample values.

Analysis of suramin treatment on forebrain neurons derived from iPSCs from sporadic AD relative to healthy controls

iPSC culture, differentiation, and treatment

Two iPSC lines (one clone derived from a patient with clinically diagnosed late-onset sAD (GM24666 [65]; 83-year-old Caucasian male; Tufts University) and the other clone derived from a healthy control (ND41866*C [66]; 64-year-old Caucasian male; Tufts University) were separately maintained on cultureware coated with human embryonic stem cell (hESC)-qualified Matrigel (Corning) in mTeSR Plus medium (STEMCELL Technologies), according to the manufacturer’s protocols. Undifferentiated iPSC colonies were selectively passaged using ReLeSR (STEMCELL Technologies) before replating onto freshly coated cultureware. Full medium changes were performed every other day. Cell culture supernatants were routinely tested for the presence of mycoplasma using the MycoProbe Mycoplasma Detection Kit (R&D Systems CUL001B). At P29 and P45 for the GM24666 and ND41866*C lines, respectively, cells were passaged and either plated in parallel for ICC to confirm pluripotency or applied in parallel to terminal forebrain neuron differentiation. Both iPSC lines implemented herein were found to be >95% Nanog⁺ and SOX-2/OCT-4^+/+ via ICC prior to the onset of differentiation (Supplementary Figure 2).

Forebrain neuron differentiation was performed for the iPSC lines in parallel using commercially available kits in accordance with the manufacturer’s protocols. Briefly, NPC differentiation was achieved via the STEMdiff™ SMADi Neural Induction Kit (STEMCELL Technologies 08581): iPSCs were single-cell passaged using TrypLE Express (Gibco) and plated in GFR Matrigel-coated cultureware at a density of 2×10⁵ cells/cm² in STEMdiff™ Neural Induction Medium containing the kit-provided SMADi Supplement and 2μM thiazovivin (STEMCELL Technologies). The following day, spent medium was replaced with fresh Neural Induction Medium without thiazovivin. Full medium changes were performed daily thereafter. Cells were passaged thrice, under these conditions, every 6 days, with cells reaching monolayer confluence with each passage. Thiazovivin (2μM) was applied following each passage to promote cell viability and, each time, was removed the following day.

On Day 18 (Passage 3) of neural induction, NPCs were applied toward terminal neuronal differentiation using the STEMdiff™ Forebrain Neuron Differentiation Kit (STEMCELL Technologies 08600) and the STEMdiff™ Forebrain Neuron Maturation Kit (STEMCELL Technologies 08605). Briefly, Day 18 NPCs were single-cell passaged using TrypLE Express and plated at a density of 1×10⁵ cells/cm² on GFR Matrigel-coated cultureware in STEMdiff™ Neural Induction Medium + SMADi + 2μM thiazovivin. The next day, spent medium was aspirated and replaced with STEMdiff™ Forebrain Neuron Differentiation Medium without thiazovivin, with daily medium changes thereafter. On Day 25 of neural induction, cells were single-cell passaged using TrypLE Express and plated at either 5×10⁴ cells/cm² (proteomic assessments) or 2×10⁴ cells/cm² (ICC) on GFR Matrigel-coated cultureware in STEMdiff™ Forebrain Neuron Maturation Medium + 2μM thiazovivin. The next day, spent medium was aspirated and replaced with Maturation Medium without thiazovivin. Half medium changes were applied every other day thereafter for 21 days.

Although STEMCELL Technologies protocols identify 8 days as a minimum necessary maturation period for forebrain neurons (β3-tubulin⁺ and FOXG1⁺), we chose to implement 21 days of maturation based on manuscripts following a similar SMAD-based inhibition strategy [67]. These studies have reported 21 days of maturation following from the NPC stage as sufficient to produce a mixed population of excitatory and inhibitory cortical neurons, a significant portion of which were capable of firing spontaneous and repetitive action potentials [67]. Thus, after 21 days of maturation (Day 46), a subset of sAD neurons was treated with 100μM suramin (Tocris) for 72 h via full medium change—all other groups were simultaneously subjected to a full medium change with fresh Maturation Medium. The suramin treatment conditions employed herein were selected for consistency with recent literature separately evaluating the impacts of broad P2 purinergic receptor antagonism via suramin on a variety of neural cell types [68 –70].

On Day 49, forebrain neurons were either applied toward ICC characterization or lysed via addition of ice-cold 0.5% Triton X-100 in DPBS + Halt^® Protease and Phosphatase Inhibitor (Thermo Scientific 78440) as described for ReNcell cultures. Lysates were stored at –80°C for subsequent biochemical analyses.

Immunostaining and semi-quantitative image analysis

Following 21 days of neuronal maturation, iPSC-derived cultures were fixed in neutral-buffered formalin solution at RT overnight and then subjected to ICC for forebrain neuron markers FOXG1, NeuN, and MAP-2, as well as for off-target astrocyte marker S100β. In addition, levels of the excitatory synaptic marker GluA2/3/4, the inhibitory synaptic marker VGAT, and the pre-synaptic marker synaptophysin (Syp) were examined by ICC, as well as levels of P2X7 and P2Y2 receptors. Primary and secondary antibody information can be found in Supplementary Table 1. Fluorescence images (1024 X 1024 px.) were captured using a Leica TCS SP8 STED confocal microscope using a 10X objective relative to negative controls, then processed using ImageJ.

Forebrain neuronal images were converted to 8-bit grayscale format prior to analysis using ImageJ. To account for potential variations in cell density across different regions of interest (ROIs) and experimental groups, relative Syp, P2X7, and P2Y2 expression levels were assessed by dividing the integrated density of the unprocessed Syp images by that of the associated binarized DAPI counterstain images on a per-ROI basis, for a total of 5 ROIs per sample. Binarization was performed on the 8-bit grayscale DAPI counterstain images by first denoising each image through the sequential application of median and low-pass filters (r = 1 pixel), then applying a Phansalkar threshold (r = 250 pixels) to each image. The resulting binarized DAPI images were further denoised using a binary opening operation before calculation of integrated density of the resulting images.

Syp staining was also used to semi-quantitatively evaluate synaptic puncta density using the SynQuant ImageJ plugin, a tool designed for the quantitative analysis of synaptic puncta [71]. This plugin employs an unsupervised, probability principled machine-learning approach to classify synapse candidates based on parameters including size, local contrast, and noise level. Each candidate is consequently assigned a value that represents the likelihood of accurate synapse detection. This value is then compared to a user-defined z-score for thresholding. The following parameters were applied in identifying puncta from 1024×1024 pixel images: Minimum Particle Size = 10, Maximum Particle Size = 200, Minimum Fill = 0.5, Maximum WH Ratio = 4, Extended Distance = 0, and Z-Distance Multiplier = 1. To ensure the robustness and reproducibility of our findings, we incorporated a false discovery rate (FDR) control with a q-value of 0.05, corresponding to an FDR of 5%. Following synapse identification, we quantified the relative number of Syp⁺ puncta within each ROI, for a total of 5 ROIs per sample. As previously described, we likewise normalized the raw count of Syp⁺ puncta to the integrated densities of the corresponding, binarized DAPI counterstain images on a per-ROI basis.

Aβ_42/40 and p-Tau(S199)/total tau enzyme-linked immunosorbent assays (ELISAs)

ELISA kits were implemented on neuronal lysates to determine the concentrations of Aβ₄₀ (Invitrogen KHB3481) and Aβ₄₂ (Invitrogen KHB3441). Cell culture supernatants were analyzed by ELISA for p-Tau(S199) (Invitrogen KHB7041) and total tau (Invitrogen KHB0041), the ratio of which is hereafter referred to as “p-Tau/Tau”. All assays were performed in accordance with the manufacturer’s protocols. Blank lysis buffer or medium, in addition to the kit-provided standard dilution buffers, were utilized to dilute the kit-provided standards. Standard development was monitored during chromogen incubation via a Synergy HTX Multi-Mode Reader (BioTek) at 620 nm, followed by addition of kit stop solution and acquisition of sample absorbance at 450 nm.

Assessment of oxidative/nitrosative stress via measurement of glutathione and oxidized nitric oxide

Quantification of reduced glutathione (GSH), considered to be the primary antioxidant of the body, in neuronal lysates was carried out using a GSH fluorometric detection assay (Abcam ab205811), wherein lysate GSH was reacted with a fluorescent thiol green indicator. Sample fluorescence was recorded at 490/520 nm excitation/emission wavelength via a Synergy HTX Multi-Mode Reader. Standard curves generated using the kit-provided GSH standards were used to determine GSH levels. Levels of oxidized nitric oxide (NO), a major regulator of cellular stress, were measured in cell culture supernatants by assessing total nitrate and nitrite [72]. Total nitrate/nitrite levels (hereafter referred to as oxidized NO or “oxNO”) were measured using a Nitrate/Nitrite colorimetric assay kit (Abnova KA1342) according to the manufacturer’s instructions.

Assessment of cell density and autophagy modulation via Akt/mTOR signaling pathway

The DNA content of the sample lysates was measured using the PicoGreen Assay (Invitrogen) as a standard indicator of net cell proliferation and loss across treatment groups. As an indicator of Akt/mTOR pathway-mediated autophagy (Fig. 1A), the levels of p-PTEN(Ser380), p-Akt(Ser473), p-mTOR(Ser2448), p-GSK3α/β [p-GSK3α(Ser21)+ p-GSK3β(Ser9)], and p-p70S6K(Thr412) were measured using a Millipore multiplex immunoassay. For each sample, net median fluorescence intensities corresponding to analyte concentrations were normalized to scaled DNA content.

Data analysis

Univariate data analysis

Univariate data are reported as mean ± standard error of the mean (SEM). For biochemical results, statistical comparisons between groups were performed using GraphPad Prism with a standard α level of 0.05. Standard one-way ANOVAs were implemented to compare sample means, followed by post hoc Tukey’s multiple comparison testing for pairwise sample means. For semi-quantitative ICC statistical comparisons, a linear mixed-effects model (SPSS) was applied to the resulting data sets to minimize the potential for within-subject correlation to bias between-group assessments. All comparisons with p < 0.05 were considered statistically significant.

Partial least squares discriminant analysis (PLS-DA)

To more holistically evaluate group differences, we performed data dimensionality reduction using PLS-DA on the Akt/mTOR pathway and AD marker results. This multivariate statistical method combines the benefits of Partial Least Squares (PLS) regression and Linear Discriminant Analysis (LDA), providing an effective framework for analyzing high-dimensional, colinear datasets [73, 74]. Unlike Principal Component Analysis, which is designed to find the optimal low-dimensional representation of the variance in the observed data, PLS-DA operates under the assumption of certain underlying variables that are systematically correlated with observed measurements. PLS-DA thereby uncovers the underlying latent structure (the latent factors) guiding the observed data. This is done by forming linear combinations of observed variables with the aim of maximizing the covariance between these input variables and the reduced set of latent factors. These computed latent factors then serve as a robust basis for gauging the concordance between different experimental groups.

Briefly, PLS-DA was performed using values obtained from various data sets. All included measurements were subjected to standard score normalization to ensure homogeneous scaling between observed variables. In the current work, it was determined via Leave-One-Out Cross-Validation (LOOCV) that two latent variables were optimal for all PLS-DA plots presented, corresponding to the number of latent variables where the Root Mean Square Error of Cross-Validation (RMSECV) was minimized. After calculating the optimal observed factor loadings (λ) corresponding to each latent variable [75], each sample could be represented in a two-dimensional space, with the X-axis being represented by Latent X1 and the Y-axis by Latent X2, with the magnitude of squared factor loadings (λ²) being representative of the strength of the relationship between a factor and the latent variable.

Distinct treatment groups clustered separately in this 2D sample space, with each cluster being represented graphically by a centroid and distribution space (i.e., an elliptical region surrounding the group centroid encompassing 2 group standard deviations from the centroid mean). To evaluate if the separation among various group centroids was statistically significant, the Mahalanobis distance, a metric used to quantify the proximity between a point in n-dimensional space and the mean of a distribution, was calculated. The Mahalanobis distance is distinct from the Euclidean distance as it takes into consideration the covariance matrix between variables and can account for different variables following different scales. Hotteling’s t² statistics were then computed for the Mahalanobis distances between cluster centroids (p < 0.05). All dimensionality reduction and factor analyses were performed using MATLAB (2022b) and RStudio. The resulting PLS-DA cluster plots were then visualized using PyCharm (2022.3.2).

RESULTS

Our goals in the present work are two-fold: 1) to assess the effect of low-dose suramin on autophagy using ReNcell differentiated neurons in the context of known autophagy modulators rapamycin and chloroquine, and 2) to evaluate the capacity of low-dose suramin to attenuate pathological AD features in forebrain neurons differentiated from iPSCs sourced from a late-onset sporadic AD (sAD) patient relative to a matched healthy control.

Suramin effects relative to well-characterized modulators of autophagy

Although suramin is known to impact autophagic processes [26, 27], its specific effects on neuronal autophagy-associated signaling pathways have seen limited investigation. In these studies, the Akt/mTOR signaling pathway was chosen for investigation given its fundamental role in autophagy regulation as well as in purinergic and SIRT signaling [37 , 76]. Forebrain neurons derived from the ReNcell line were utilized for these studies, as ReNcells are highly characterized in literature and serve as a robust and reproducible entry model of forebrain neuron behavior. To characterize the effects of suramin treatment on neuronal autophagy in the context of the known autophagy modulators, mature ReNcell-derived neurons were treated with 100μM suramin, 40μM chloroquine, or 500 nM rapamycin for 24 h. The duration and concentrations of the treatment conditions were selected to be consistent with recent literature probing the impact of chloroquine and rapamycin on cellular autophagy [62 , 77–81]. In addition to their prior use in AD research [4 , 20–23], chloroquine and rapamycin are integral pharmacological tools in autophagy research due to their contrasting regulatory effects on this cellular process. Specifically, chloroquine inhibits autophagy by disrupting lysosomal acidification and autophagosome/lysosome fusion [20 –22]. Conversely, rapamycin induces autophagy by inhibiting the mTOR complex –a serine/threonine kinase that normally suppresses autophagy –and promoting autophagosome formation [82]. These opposing effects on autophagy will allow us to place the impact of suramin in a known context.

Following confirmation of retained cell viability following 24 h of treatment (Supplementary Figure 3), alterations in the phosphorylation state of Akt/mTOR signaling intermediates were analyzed for the suramin, chloroquine, and rapamycin groups relative to untreated controls. Full data from the Akt/mTOR immunoassay can be found in Supplementary Table 2. Toward achieving a more comprehensive analysis of this multifactorial dataset, we employed PLS-DA to collapse the 5-dimensional data [p-PTEN, p-Akt, p-mTOR, p-GSK3α/β, and p-p70S6K] onto a two-dimensional space (Fig. 1B). PLS-DA allows patterns among variables in complex datasets to be uncovered in a statistically robust framework and for identified patterns to be readily visualized. PLS-DA of the Akt/mTOR results revealed that the impact of suramin (red cluster) on neuronal autophagy was distinct from either the 40μM chloroquine (blue cluster; classical autophagic inhibition) or 500 nM rapamycin (gold cluster; classical autophagy induction) treatment. The contribution, or loading (λ), for each factor for each axis of the PLS-DA plot is provided in Supplementary Table 3.

In terms of distinguishing suramin from chloroquine or rapamycin treatment, Latent X1 was the primary separating axis (Fig. 1B), with the two major factors (each contributing >20% to differences among groups along this axis) being p-p70S6K and p-mTOR (Supplementary Table 3). The primary distinguishing axis for suramin relative to the untreated control (maroon cluster) was Latent X2, which was dominated by p-PTEN (Supplementary Table 3). To examine the specific effects of various treatments on these factors more closely, conventional univariate comparisons for p-p70S6K, p-mTOR, and p-PTEN are shown in Fig. 1C and 1D. Notably, p-p70S6K levels were significantly increased in ReNcell-derived neurons following treatment with suramin (∼1.3-fold; p < 0.001), unchanged with chloroquine, and significantly decreased with rapamycin (∼15.9-fold; p < 0.001) relative to control. In contrast, p-mTOR levels were unchanged by suramin, but significantly decreased by both chloroquine (∼1.7-fold, p < 0.001) and rapamycin (∼3.4-fold, p < 0.001) relative to control.

Consistent with the present results, 24 h rapamycin treatment has repeatedly been observed to significantly decrease p-mTOR and p-p70S6K in cortical neurons and neural stem cells [83, 84]. We also observed a significant increase in p-Akt with rapamycin treatment (Supplementary Table 2), with literature variably reporting both increases and decreases in this marker in response to rapamycin exposure [83 , 85–87]. Consistent with chloroquine acting primarily downstream of the Akt/mTOR pathway, chloroquine had no significant effect on the phosphorylation state of the Akt/mTOR markers examined aside from p-mTOR (Fig. 1C). Reduction in p-mTOR following chloroquine treatment, as observed herein, has been previously reported in literature, although this effect appears to be context dependent [82 , 88–90]. In terms of suramin and the Akt/mTOR pathway, relevant literature is limited primarily to a single in vivo work [25]. In this study an increase in p-p706SK and no change in p-Akt or in p-mTOR, was reported in brain homogenates of Fragile X mice following intravenous suramin treatment relative to untreated controls, consistent with results herein (Supplementary Table 2).

Late onset sAD forebrain neuronal expression of canonical disease markers

After confirming that the cell viability and Akt/mTOR effects observed for ReNcell derived neurons for 24 h suramin treatment were retained for 72 h suramin treatment (Supplementary Figure 3 and Supplementary Table 4), the effects of suramin on forebrain neurons derived from iPSCs from a late-onset sAD patient were also examined. An extended (72 h) treatment time was selected for these sAD-derived neuron studies to allow greater time for neurons to respond to treatment and to potentially allow for observable changes in synaptic structure [91]. Towards this end, iPSCs from a sAD donor and a healthy matched control were first differentiated into forebrain neurons per established protocols. For both sAD and control groups, the resulting cells displayed a > 95% FOXG1⁺, NeuN⁺, MAP-2⁺, and S100β⁻ status based on ICC (Supplementary Figure 4). Mixed populations of excitatory/inhibitory forebrain neurons were present in both sAD-derived and control groups based on staining for GluA2/3/4 (excitatory) and VGAT (inhibitory) (Supplementary Figure 5). Furthermore, the ratio of purinergic receptors P2X7/P2Y2 (Supplementary Figure 5) was significantly higher in sAD forebrain neurons relative to healthy controls, consistent with literature on AD and other neuroinflammatory disorders [35].

In comparing healthy- to sAD-derived neurons, we examined: 1) retained cell viability in the suramin treated sAD group, 2) Aβ dysregulation (through Aβ_42/40), 3) tau dysregulation (through p-Tau/Tau), and 4) oxidative/nitrosative stress (through the ratio of GSH to oxNO). Net cell proliferation and loss in the suramin treated sAD neurons was statistically indistinguishable from the untreated sAD group (Supplementary Figure 6), consistent with the ReNcell data. Significant elevations were observed in untreated sAD-derived forebrain neurons for both the ratio of p-Tau/Tau (∼2.3-fold; p < 0.001) as well as the ratio GSH/oxNO (∼1.9-fold; p = 0.010) relative to healthy forebrain neuron controls (Fig. 2). While an increase in the ratio of Aβ_42/40 (∼1.3-fold) for untreated sAD-derived neurons relative to controls was also observed, this trend was not statistically significant.

Fig. 2

Suramin partially remediates AD-related markers in sAD-derived forebrain neurons. Expression of AD-associated markers in iPSC-differentiated forebrain neurons were measured by ELISA (Aβ_42/40, p-Tau(S199)/Tau) and by fluorometric/colorimetric detection assays (GSH and oxNO, respectively). Samples numbers per group: Control, n = 8–10; sAD, n = 5; sAD + suramin, n = 5. ^*,# denote a significant difference from the healthy-derived control (black) and untreated sAD-derived (light grey) groups, respectively (p < 0.05). Error bars represent SEM. White dots represent individual sample values.

Notably, these results are consistent with a separate study utilizing forebrain neurons derived from the same sAD and healthy control clones [92]. Following 1 year of forebrain neuron maturation, this study noted a 4.4-fold increase in p-Tau/Tau and a non-significant 1.6-fold increase in Aβ_42/40 in sAD-derived neurons relative to control-derived neurons [92]. A separate study examining neurons derived from iPSC lines from 4 late-onset sAD iPSC donors relative to 3 matched control donors found a significant 1.7-3.0-fold increase in p-Tau/Tau and no significant increase in Aβ_42/40 in sAD-derived forebrain neurons relative to controls following 10 weeks of forebrain neuron maturation past the NPC state [18]. It should further be noted that the lack of a statistically significant increase in Aβ_42/40 found herein is consistent with a growing number of publications reporting that this specific disease marker is more consistently found in familial AD, not sAD, lines [18]. Thus, the degree of disease features noted in our study are consistent with current iPSC literature, including studies with significantly more extended maturation times.

Effect of suramin treatment on AD marker expression in sAD forebrain neurons

Following 72 h treatment with low-dose 100μM suramin, the treated sAD-derived neurons were compared to untreated sAD and healthy controls (Fig. 2). Although suramin did not significantly impact the Aβ_42/40 ratio (p = 0.327), suramin treatment returned sAD-derived neurons to GSH/oxNO levels consistent with that of healthy controls, with GSH/oxNO decreasing ∼2.0-fold (p = 0.024) relative to untreated sAD-derived neurons. Furthermore, suramin decreased p-Tau/Tau by ∼1.5-fold (p = 0.015) in sAD-derived neurons, although p-Tau/Tau levels in the suramin treated sAD group remained elevated relative to healthy controls (Fig. 2).

To investigate the potential impact of suramin on synaptic structures, ICC and subsequent image processing was conducted for Syp, an abundantly expressed protein associated with pre-synaptic vesicles [93]. These semi-quantitative analyses revealed significantly lower overall Syp expression (∼1.3-fold; p = 0.001) and Syp⁺ synaptic density (∼2.4-fold; p < 0.001) in sAD-derived forebrain neurons relative to healthy controls (Fig. 3), consistent with literature [94]. However, treatment of sAD-derived forebrain neurons with 100μM suramin did not result in a significant shift in either Syp expression or Syp⁺ puncta density relative to untreated sAD neurons. Given that substantial alterations in synaptic structures often occur over more extended timescales [91], the suramin treatment period implemented herein (72 h) may have been insufficient to produce observable differences in Syp expression and synaptic arrangement.

Fig. 3

Deficits in synaptic network formation in sAD-derived neurons relative to healthy controls are not significantly improved by 72 h suramin treatment. A) Left column: Syp (green) imaging of healthy (Control), diseased (sAD), and suramin treated, diseased (sAD + Suramin) iPSC-differentiated forebrain neurons. Scale bar = 100μm applies to all images. Right column: ImageJ processing to identify Syp⁺ pre-synaptic puncta (identified puncta highlighted in yellow). Magnified regions are a 3x zoom; B) Semi-quantitative assessment of per cell Syp expression levels; C) Calculated relative per cell Syp⁺ puncta expression. ^*denotes a significant difference from the “Control” group (p < 0.05). Error bars represent SEM. White dots represent individual samples.

Multivariate analysis of AD pathology and suramin treatment via PLS-DA

To more clearly delineate net phenotypic shifts due to suramin in terms of Aβ_42/40, p-Tau/Tau, and GSH/oxNO levels, PLS-DA analysis was applied to the iPSC-derived neuron data. In this reduced dimensional space, the cluster representing the untreated sAD-derived neurons (maroon cluster) could be clearly and significantly (p = 0.020) distinguished from the cluster representing the healthy-derived controls (blue cluster) (Fig. 4A). sAD neurons exposed to 100μM suramin (gold cluster) were shifted phenotypically closer to the control group within the same latent space (Fig. 4A). Quantitatively, suramin application reduced the degree of phenotypic separation, as quantified by the Mahalanobis distance metric, between the control and sAD-derived cluster centroids by 3.3-fold (Supplementary Table 5).

Fig. 4

Global PLS-DA visualization of the impact of suramin treatment on sAD-derived neuron expression of AD markers and Akt/mTOR pathway phosphorylation state. A) PLS-DA analysis of aggregated Aβ_42/40, p-Tau/Tau, and GSH/oxNO data; B) PLS-DA of Akt/mTOR multiplex protein assay results from iPSC-derived neurons. In both (A) and (B), sample numbers are: iPSC CTL, n = 9; iPSC sAD: n = 5; iPSC sAD Sura, n = 5. Healthy-derived (iPSC CTL), sAD-derived (iPSC sAD), and suramin-treated sAD-derived (iPSC sAD Sura) neuron groups are shown in blue, maroon, and gold, respectively. Cluster centroids (means) are shown as squares. The elliptical space around each centroid represents the space correlating to 2 standard deviations from the centroid mean. Calculated Mahalanobis distances and Hotelling’s t-squared statistics, and factor weightings for each PLS-DA diagram are given in Supplementary Tables 5 and 7, respectively. C) Relative expression of phosphorylated Akt/mTOR pathway markers identified as significantly contributing (>20%) to differences among groups along the Latent X2 axis in (B). p-PTEN is also included for comparison with ReNcell data. ^* and # denote a significant difference from the iPSC CTL and iPSC sAD groups, respectively (p < 0.05). Error bars represent SEM. White dots represent individual sample values.

Effect of suramin treatment on Akt/mTOR signaling in sAD forebrain neurons

Suramin treated sAD-derived neurons were similarly evaluated for shifts in the Akt/mTOR pathway relative to untreated sAD and healthy controls. Full immunoassay data can be found in Supplementary Table 6. As shown in Fig. 4B, PLS-DA of consolidated Akt/mTOR multiplex immunoassay data revealed that the untreated sAD group (maroon cluster) and the treated sAD group (gold cluster) were primarily distinguished by placement along the Latent X2 axis. Of the markers assessed, p-p70S6K and p-mTOR dominated the Latent X2, as evidenced by their factor loadings (λ) (Supplementary Table 7). Univariate examination shows that suramin treatment significantly increased sAD-derived neuron levels of p-p70S6K (∼1.3-fold, p = 0.012), while significantly decreasing levels of p-mTOR (∼1.3-fold, p < 0.001) (Fig. 4C). In contrast to the observations with ReNcells, p-PTEN was not significantly impacted by suramin treatment (p = 0.356). The cumulative sAD and ReNcell data, combined with prior literature, indicate that suramin consistently impacts the neuronal Akt/mTOR pathway through elevation of p-p70S6K [25].

DISCUSSION

The role of autophagy and associated proteostasis abnormalities in AD has been extensively documented, as have disruptions in purinergic signaling [3, 95]. In this study, we evaluated the potential impact of suramin, a non-canonical autophagy modulator, in addressing several AD pathological hallmarks at the neuronal level. Given that the effects of suramin on neuronal autophagy are relatively poorly understood, we first sought to contextualize the autophagic effects of suramin treatment by characterizing its impact on the phosphorylation state of the Akt/mTOR pathway. In these initial studies, suramin was compared to known autophagy modulators using forebrain neurons differentiated from the well-characterized ReNcell NPC line [96 –98]. Based on PLS-DA analyses, the differential effects of suramin on the Akt/mTOR pathway appeared to be dominated by a consistent increase in p-p70S6K, whereas rapamycin decreased p-p70S6K as well as p-mTOR.

p70S6K is a downstream mediator of the mTOR autophagy pathway activated via phosphorylation at Thr412 by unphosphorylated mTOR, which itself inhibits autophagy [99]. Thus, increased p-p70S6K may indicate increased unphosphorylated mTOR and therefore increased autophagic inhibition. Indeed, a study of post-mortem cortical samples from AD patients indicates a significant increase in p-p70S6K relative to age-matched healthy controls [100], and elevation in neuronal p-p70S6K has been suggested to precede neurofibrillary tangle formation [101]. In contrast, recent research into muscle metabolism indicates that p-p70S6K may induce autophagy: it is now understood that p-p70S6K directly inhibits mTOR activity (activates autophagy) [102, 103] in a negative feedback loop manner [104]. Moreover, p-p70S6K has been shown to phosphorylate and thereby inhibit IRS-1, a protein that suppresses autophagy by enhancing mTOR activity [105]. In the current study, suramin treatment was associated with a reduction in p-Tau/Tau and in GSH/oxNO, both of which would be consistent with autophagic induction.

Beyond increased p-p70S6K effects, the observed reduction in p-Tau/Tau between suramin treated and untreated sAD-derived neurons could potentially also be related to the increased p-GSK3α/β (inactive) (Supplementary Table 6) observed in suramin treated sAD-derived neurons. While various tau phosphorylation sites are known to be differentially dysregulated in AD, active (non-phosphorylated) GSK3β, an integral component of the Akt/mTOR pathway, specifically phosphorylates tau at Ser199, the tau phosphorylation site assessed in this study. This phosphorylation event is implicated in the pathological localization of tau to dendritic spines, and contributes to the initiation of synaptic dysfunction, a key feature of early AD progression [106].

In terms of oxidative/nitrosative stress, we observed a significant increase in the GSH/oxNO ratio in the sAD-derived forebrain neurons relative to healthy controls that was fully remediated by suramin treatment. GSH is a key mediator of oxidative stress that acts directly and indirectly on free radicals such as reactive oxygen species as well as NO [107, 108]. Recent literature postulates that the capacity of the body to effectively neutralize oxidative/nitrosative stress diminishes with age, resulting in cellular and mitochondrial damage and dysfunction. This aging-associated decline seems to be exacerbated in AD, with patients displaying decreased levels of cortical GSH relative to age-matched healthy controls [109]. As GSH levels decrease, oxidative/nitrosative stresses increase, resulting in neuronal loss, and subsequently, decreased levels of oxNO as AD advances [110, 111]. In addition to the observed increase in the GSH/oxNO ratio in sAD neurons relative to control, a breakdown of this ratio into its constituents revealed a 10% decrease in GSH and a 40% decrease in oxNO relative to control, aligning with advanced AD progression.

On examining Syp expression and Syp⁺ puncta density, we also found a decrease in both per neuron Syp levels and Syp⁺ synaptic puncta density in the sAD group relative to healthy controls, consistent with literature [112]. Specifically, Syp expression has been negatively correlated with AD severity, neurofibrillary tangle formation, and declining synaptic network maintenance [113 –116]. Indeed, Syp mRNA in postmortem human neocortical tissues has been shown to be significantly reduced in AD compared to age- and sex-matched healthy controls, showing a negative correlation with cognitive deficits and AD severity [116]. In a recent study of 171 individuals with varied levels of cognitive impairment, with and without AD, Syp levels decreased significantly as dementia progressed [117].

Interestingly, while suramin treatment ameliorated the dysregulated p-Tau/Tau and GSH/oxNO observed in the sAD-derived neurons relative to controls, it did not lead to a significant improvement in either Syp expression or Syp⁺ puncta density. A possible explanation for this observation may be the temporal dynamics of synaptic modifications in response to treatment [91]. Future studies extending the duration of suramin treatment may reveal further therapeutic effects on synaptic puncta density over longer periods. It should be noted that the potentially therapeutic impact of suramin observed herein contradicts the findings of a study by Dong et al., which found that 24 h suramin treatment (∼230 mM) augmented the detrimental effects of Aβ oligomer stimulation in neonatal rat neurons by increasing senile plaque formation and upregulating oxidative stress [49]. Given that the suramin concentration implemented by Dong et al. was over 1000-fold greater than that evaluated in this study, direct comparison is challenging. That said, we anticipate that suramin effects will display a potentially complex dose dependence and that higher suramin levels will be associated with increased off-target effects.

While the current findings are intriguing, there are several limitations that warrant further exploration and validation. A particular constraint of the present study is its reliance on a single iPSC donor, and a single clone per donor, for both the sAD and healthy control cohorts, which limits the generalizability of our results. Furthermore, only short-term suramin application at a single concentration (considered a relatively low concentration) [25] was examined. Future research will therefore incorporate a more extensive set of iPSC lines and a broader range of suramin treatment regimens to substantiate and elaborate upon the promising initial results of this study. Moreover, while suramin has demonstrated positive effects in the iPSC-derived forebrain neurons employed herein, it should be noted that suramin is a broad spectrum P2 receptor antagonist and SIRT inhibitor. Inhibition of distinct P2 receptors can result in differential and sometimes contradictory effects on neuronal health [38 , 119]. As such, the balance of P2 receptors present is anticipated to have substantial effect on the outcome of suramin treatment. This is particularly true as P2Y4, P2Y6, and P2Y12, for which inhibition is considered to negatively impact neuronal health, are highly expressed by microglia [120]. Thus, more comprehensive evaluation of suramin effects requires the use of more developed neural models incorporating glial cells and potentially longer maturation periods as well as characterization of primary P2 receptors and SIRTs. Such studies, combined with targeted evaluation of various suramin analogues/derivatives, may allow more selective inhibition and greater improvements to be achieved.

To our knowledge, this is the first investigation that supports the potential remedial impact of suramin in the context of AD-derived neurons. These data suggest that short-term application of low-dose suramin may reduce sAD-derived forebrain neuron dysregulation in part through differential modulation of Akt/mTOR signaling. Future work will utilize a more extensive array of iPSC donors and clones per donor to provide a more comprehensive understanding of disease pathology and contribute to the robustness of the observed phenotypic differences. In addition, future studies will incorporate broader assessment of suramin targets (P2 receptors and SIRTs) as well as its downstream effects [3 , 122]. In summary, this study highlights the potential of suramin or putatively its derivatives/analogues as novel therapeutic agents for AD, providing a foundation for further research to elucidate its mechanisms of action.

AUTHOR CONTRIBUTIONS

Mariah S. Hahn (Conceptualization; Funding acquisition; Project administration; Supervision; Writing – review & editing); Robert A. Culibrk (Conceptualization; Data curation; Methodology; Writing – original draft; Writing – review & editing); Katherine A. Ebbert (Data curation; Validation; Writing – review & editing); Daniel J. Yeisley (Data curation; Writing – review & editing); Rui Chen (Data curation; Validation; Writing – review & editing); Fatir A. Qureshi (Methodology; Software); Juergen Hahn (Methodology; Software; Supervision).

Footnotes

ACKNOWLEDGMENTS

Use of the Leica TCS SP8 STED Microscope was made possible by the Microscopy Core at Rensselaer Polytechnic Institute (Center for Biotechnology and Interdisciplinary Studies), led by Dr. Sergey Pryshchep.

FUNDING

This research was supported, in part, by the National Institute of Health R03AG067140 and R03AG067970 to MSH and by the National Science Foundation Graduate Research Fellowship (RAC) under Grant No. DGE 1744655.

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

All data generated or analyzed during this study are included in this article and its supplementary information. Any additional data that support the findings of this study are available from the corresponding author upon reasonable request. MATLAB code for the PLS-DA is included in Supplementary Materials.

The supplementary material is available in the electronic version of this article: .

References

Picanco

LCDS

, Ozela

, de Fatima de Brito Brito

, Pinheiro

, Padilha

, Braga

, de Paula da Silva

CHT

, dos Santos

CBR

, Rosa

JMC

, da Silva Hage-Melim

(2018) Alzheimer’s disease: A review from the pathophysiology to diagnosis, new perspectives for pharmacological treatment. Curr Med Chem 25, 3141–3159.

van Dyck

, Swanson

, Aisen

, Bateman

, Chen

, Gee

, Kanekiyo

, Li

, Reyderman

, Cohen

, Froelich

, Katayama

, Sabbagh

, Vellas

, Watson

, Dhadda

, Irizarry

, Kramer

, Iwatsubo

(2023) Lecanemab in early Alzheimer’s disease. N Engl J Med 388, 9–21.

Culibrk

, Hahn

(2020) The role of chronic inflammatory bone and joint disorders in the pathogenesis and progression of Alzheimer’s disease. Front Aging Neurosci 12, 583884.

Hamano

, Enomoto

, Shirafuji

, Ikawa

, Yamamura

, Yen

S-H

, Nakamoto

(2021) Autophagy and tau protein. Int J Mol Sci 22, 7475.

Suzuki

, Terry

(1967) Fine structural localization of acid phosphatase in senile plaques in Alzheimer’s presenile dementia. Acta Neuropathol (Berl) 8, 276–284.

Silva

, Nandi

, Tentarelli

, Gurrell

, Jamier

, Lucente

, Dickerson

, Brown

, Brandon

, Haggarty

(2020) Prolonged tau clearance and stress vulnerability rescue by pharmacological activation of autophagy in tauopathy neurons. Nat Commun 11, 3258.

Park

, Kim

, Shin

, Lee

(2021) Intersection between redox homeostasis and autophagy: Valuable insights into neurodegeneration. Antioxidants 10, 694.

Filomeni

, De Zio

, Cecconi

(2015) Oxidative stress and autophagy: The clash between damage and metabolic needs. Cell Death Differ 22, 377–388.

Xie

, Chen

, Yan

, He

, Chen

, Li

(2014) Identification of a role for the PI3K/AKT/mTOR signaling pathway in innate immune cells. PLoS One 9, e94496.

10.

Cacabelos

, Carril

, Cacabelos

, Kazantsev

, Vostrov

, Corzo

, Cacabelos

, Goldgaber

(2019) Sirtuins in Alzheimer’s disease: SIRT2-related genophenotypes and implications for pharmacoepigenetics. Int J Mol Sci 20, 1249.

11.

Mehramiz

, Porter

, Laws

, Rainey-Smith

(2022) Sleep, sirtuin 1 and Alzheimer’s disease: A review. Aging Brain 2, 100050.

12.

Dai

, Sinclair

, Ellis

, Steegborn

(2018) Sirtuin activators and inhibitors: Promises, achievements, and challenges. Pharmacol Ther 188, 140–154.

13.

Trapp

, Meier

, Hongwiset

, Kassack

, Sippl

, Jung

(2007) Structure–activity studies on suramin analogues as inhibitors of NAD⁺-dependent histone deacetylases (sirtuins). Chem Med Chem 2, 1419–1431.

14.

Polletta

, Vernucci

, Carnevale

, Arcangeli

, Rotili

, Palmerio

, Steegborn

, Nowak

, Schutkowski

, Pellegrini

, Sansone

, Villanova

, Runci

, Pucci

, Morgante

, Fini

, Mai

, Russo

, Tafani

(2015) SIRT5 regulation of ammonia-induced autophagy and mitophagy. Autophagy 11, 253–270.

15.

Heras-Sandoval

, Pérez-Rojas

, Hernández-Damián

, Pedraza-Chaverri

(2014) The role of PI3K/AKT/mTOR pathway in the modulation of autophagy and the clearance of protein aggregates in neurodegeneration. Cell Signal 26, 2694–2701.

16.

Zhang

, Ma

, Wang

, Li

, Liu

, Zhao

(2020) The PI3K/AKT/mTOR pathway regulates autophagy to induce apoptosis of alveolar epithelial cells in chronic obstructive pulmonary disease caused by PM2.5 particulate matter. J Int Med Res 48, 0300060520927919.

17.

Jia

, Qiu

, Lin

, Zhang

, Li

, Jin

(2022) Inhibition of PI3K/AKT/mTOR signalling pathway activates autophagy and suppresses peritoneal fibrosis in the process of peritoneal dialysis. Front Physiol 13, 778479.

18.

Ochalek

, Mihalik

, Avci

, Chandrasekaran

, Téglási

, Bock

, Giudice

, Táncos

, Molnár

, László

, Nielsen

, Holst

, Freude

, Hyttel

, Kobolák

, Dinnyés

(2017) Neurons derived from sporadic Alzheimer’s disease iPSCs reveal elevated TAU hyperphosphorylation, increased amyloid levels, and GSK3B activation. Alzheimers Res Ther 9, 90.

19.

Chen

, Long

, Li

, Luo

, He

(2019) Alteration of the Wnt/GSK3β/β-catenin signalling pathway by rapamycin ameliorates pathology in an Alzheimer’s disease model. Int J Mol Med 44, 313–323.

20.

Cuchillo-Ibañez

, Lennol

, Escamilla

, Mata-Balaguer

, Valverde-Vozmediano

, Lopez-Font

, Ferrer

, Sáez-Valero

(2021) The apolipoprotein receptor LRP3 compromises APP levels. Alzheimers Res Ther 13, 181.

21.

Laili

, Nasir

MHM

, Jufri

, Ibrahim

, Hamid

(2023) Lysosomal dysfunction induced cytosolic vacuolation and increased intracellular amyloid-beta 42 (Aβ42) in human brain endothelial cells (HBEC-5i). Biomed Pharmacother 161, 114501.

22.

Iyaswamy

, Krishnamoorthi

, Zhang

, Sreenivasmurthy

, Zhu

, Liu

, Su

C-F

, Guan

X-J

, Wang

Z-Y

, Cheung

K-H

, Song

J-X

, Durairajan

SSK

, Li

(2021) Qingyangshen mitigates amyloid-β and Tau aggregate defects involving PPARα-TFEB activation in transgenic mice of Alzheimer’s disease. Phytomedicine 91, 153648.

23.

Majumder

, Richardson

, Strong

, Oddo

(2011) Inducing autophagy by rapamycin before, but not after, the formation of plaques and tangles ameliorates cognitive deficits. PLoS One 6, e25416.

24.

Tung

, Wang

, Hu

, Hsu

, Lee

, Huang

, Liao

(2012) Autophagy: A double-edged sword in Alzheimer’s disease. J Biosci 37, 157–165.

25.

Naviaux

, Wang

, Li

, Bright

, Alaynick

, Williams

, Powell

, Naviaux

(2015) Antipurinergic therapy corrects the autism-like features in the Fragile X (Fmr1 knockout) mouse model. Mol Autism 6, 1–20.

26.

Korolenko

, Johnston

, Vetvicka

(2020) Lysosomotropic features and autophagy modulators among medical drugs: Evaluation of their role in pathologies. Molecules 25, 5052.

27.

Silva

RCMC

, Ribeiro

, da Silva

GPD

, da Costa

, Travassos

(2022) Autophagy modulators in coronavirus diseases: A double strike in viral burden and inflammation. Front Cell Infect Microbiol 12, 845368.

28.

Franke

, Sauer

, Rudolph

, Krügel

, Hengstler

, Illes

(2009) P2 receptor-mediated stimulation of the PI3-K/Akt-pathway in vivo. Glia 57, 1031–1045.

29.

Schuetz

, Min

, Antoshenko

, Wang

C-L

, Allali-Hassani

, Dong

, Loppnau

, Vedadi

, Bochkarev

, Sternglanz

, Plotnikov

(2007) Structural basis of inhibition of the human NAD+-dependent deacetylase SIRT5 by suramin. Structure 15, 377–389.

30.

Burnstock

(2008) Purinergic signalling and disorders of the central nervous system. Nat Rev Drug Discov 7, 575–590.

31.

Beamer

, Gölöncsér

, Horváth

, Beko

, Otrokocsi

, Koványi

, Sperlágh

(2016) Purinergic mechanisms in neuroinflammation: An update from molecules to behavior. Neuropharmacology 104, 94–104.

32.

Di Virgilio

, Sarti

, Coutinho-Silva

(2020) Purinergic signaling, DAMPs, and inflammation. Am J Physiol-Cell Physiol 318, C832–C835.

33.

Bressan

, Pecora

, Gagnon

, Snapyan

, Labrecque

, De Koninck

, Parent

, Saghatelyan

(2020) The dynamic interplay between ATP/ADP levels and autophagy sustain neuronal migration in vivo. eLife 9, e56006.

34.

Di Virgilio

, Sarti

, Falzoni

, De Marchi

, Adinolfi

(2018) Extracellular ATP and P2 purinergic signalling in the tumour microenvironment. Nat Rev Cancer 18, 601–618.

35.

Glaser

, Oliveira-Giacomelli

, Petiz

, Ribeiro

, Andrejew

, Ulrich

(2021) Antagonistic roles of P2X7 and P2Y2 receptors in neurodegenerative diseases. Front Pharmacol 12, 659097.

36.

León-Otegui

, Gómez-Villafuertes

, Díaz-Hernández

, Miras-Portugal

, Gualix

(2011) Opposite effects of P2X7 and P2Y2 nucleotide receptors on α-secretase-dependent APP processing in Neuro-2a cells. FEBS Lett 585, 2255–2262.

37.

Bian

, Sun

, Bai

, Zhang

, Li

, Enjyoji

, Junger

, Robson

, Wu

(2013) P2X7 integrates PI3K/AKT and AMPK-PRAS40-mTOR signaling pathways to mediate tumor cell death. PLoS One 8, e60184.

38.

Merighi

, Poloni

, Terrazzan

, Moretti

, Gessi

, Ferrari

(2021) Alzheimer and purinergic signaling: Just a matter of inflammation? . Cells 10, 1267.

39.

de Diego García

, Sebastián-Serrano

, Hernández

, Pintor

, Lucas

, Díaz-Hernández

(2018) The regulation of proteostasis in glial cells by nucleotide receptors is key in acute neuroinflammation. FASEB J 32, 3020–3032.

40.

Fernando

KKM

, Wijayasinghe

(2021) Sirtuins as potential therapeutic targets for mitigating neuroinflammation associated with Alzheimer’s disease. Front Cell Neurosci 15, 746631.

41.

Harting

, Knöll

(2010) SIRT2-mediated protein deacetylation: An emerging key regulator in brain physiology and pathology. Eur J Cell Biol 89, 262–269.

42.

de Oliveira

, Vicente Miranda

, Francelle

, Pinho

, Szegö

ÉM

, Martinho

, Munari

, Lázaro

, Moniot

, Guerreiro

, Fonseca

, Marijanovic

, Antas

, Gerhardt

, Enguita

, Fauvet

, Penque

, Pais

, Tong

, Becker

, Kügler

, Lashuel

, Steegborn

, Zweckstetter

, Outeiro

(2017) The mechanism of sirtuin 2–mediated exacerbation of alpha-synuclein toxicity in models of Parkinson disease. PLoS Biol 15, e2000374.

43.

, Lee

, Huang

, Messina-Graham

, Broxmeyer

(2014) SIRT1 positively regulates autophagy and mitochondria function in embryonic stem cells under oxidative stress. Stem Cells 32, 1183–1194.

44.

Lee

(2019) Mechanisms and disease implications of sirtuin-mediated autophagic regulation. Exp Mol Med 51, 1–11.

45.

El-kott

, Al-kahtani

, Shati

(2019) Calycosin induces apoptosis in adenocarcinoma HT 29 cells by inducing cytotoxic autophagy mediated by SIRT 1/AMPK -induced inhibition of Akt/mTOR. Clin Exp Pharmacol Physiol 46, 944–954.

46.

Pillai

, Sundaresan

, Gupta

(2014) Regulation of Akt signaling by sirtuins: Its implication in cardiac hypertrophy and aging. Circ Res 114, 368–378.

47.

Ghosh

, McBurney

, Robbins

(2010) SIRT1 negatively regulates the mammalian target of rapamycin. PLoS One 5, e9199.

48.

Fan

, He

, Wu

, Chen

, Cheng

, Zhan

, An

, Chen

, Wang

(2023) Sirt3 activates autophagy to prevent DOX-induced senescence by inactivating PI3K/AKT/mTOR pathway in A549 cells. Biochim Biophys Acta Mol Cell Res 1870, 119411.

49.

Dong

Y-T

, Cao

, Tan

L-C

, Wang

X-L

, Qi

X-L

, Xiao

, Guan

Z-Z

(2018) Stimulation of SIRT1 attenuates the level of oxidative stress in the brains of APP/PS1 double transgenic mice and in primary neurons exposed to oligomers of the amyloid-β peptide. J Alzheimers Dis 63, 283–301.

50.

Reichenbach

, Delekate

, Breithausen

, Keppler

, Poll

, Schulte

, Peter

, Plescher

, Hansen

, Blank

, Keller

, Fuhrmann

, Henneberger

, Halle

, Petzold

(2018) P2Y1 receptor blockade normalizes network dysfunction and cognition in an Alzheimer’s disease model. J Exp Med 215, 1649–1663.

51.

Chiang

C-H

, Cheng

C-Y

, Lien

Y-T

, Huang

K-C

, Lin

W-W

(2022) P2X7 activation enhances lipid accumulation during adipocytes differentiation through suppressing the expression of sirtuin-3, sirtuin-5, and browning genes. Front Pharmacol 13, 852858.

52.

Gao

, Wang

W-Y

, Mao

Y-W

, Gräff

, Guan

J-S

, Pan

, Mak

, Kim

, Su

, Tsai

L-H

(2010) A novel pathway regulates memory and plasticity via SIRT1 and miR-134. Nature 466, 1105–1109.

53.

McCain

, Wu

, Nickel

, Kassack

, Kreimeyer

, Gagliardi

, Collins

, Zhang

Z-Y

(2004) Suramin derivatives as inhibitors and activators of protein-tyrosine phosphatases. J Biol Chem 279, 14713–14725.

54.

Syed

, Kennedy

(2012) Pharmacology of P2X receptors. Wiley Interdiscip Rev Membr Transp Signal 1, 16–30.

55.

Illes

, Müller

, Jacobson

, Grutter

, Nicke

, Fountain

, Kennedy

, Schmalzing

, Jarvis

, Stojilkovic

, King

, Di Virgilio

(2021) Update of P2X receptor properties and their pharmacology: IUPHAR Review 30. Br J Pharmacol 178, 489–514.

56.

Jacobson

, Delicado

, Gachet

, Kennedy

, Von Kügelgen

, Li

, Miras-Portugal

, Novak

, Schöneberg

, Perez-Sen

, Thor

, Wu

, Yang

, Müller

(2020) Update of P2Y receptor pharmacology: IUPHAR Review 27. Br J Pharmacol 177, 2413–2433.

57.

Gill

, Windebank

(1998) Suramin induced ceramide accumulation leads to apoptotic cell death in dorsal root ganglion neurons. Cell Death Differ 5, 876–883.

58.

Von Kügelgen

(2006) Pharmacological profiles of cloned mammalian P2Y-receptor subtypes. Pharmacol Ther 110, 415–432.

59.

, Jing

, Lin

(2014) Sirtuin inhibitors as anticancer agents. Future Med Chem 6, 945–966.

60.

Takada

, Aláez-Verson

, Burgute

, Nitrini

, Sosa

, Castilhos

, Chaves

, Longoria

E-M

, Carrillo-Sánchez

, Brucki

SMD

, Flores-Lagunes

, Molina

, Olivares

, Ziegemeier

, Petranek

, Goate

, Cruchaga

, Renton

, Fernández

, Day

, McDade

, Bateman

, Karch

, Llibre-Guerra

, for the Dominantly Inherited Alzheimer Network (2022) Discovery and validation of dominantly inherited Alzheimer’s disease mutations in populations from Latin America. Alzheimers Res Ther 14, 108.

61.

Yang

, Arseni

, Zhang

, Huang

, Lövestam

, Schweighauser

, Kotecha

, Murzin

, Peak-Chew

, Macdonald

, Lavenir

, Garringer

, Gelpi

, Newell

, Kovacs

, Vidal

, Ghetti

, Ryskeldi-Falcon

, Scheres

SHW

, Goedert

(2022) Cryo-EM structures of amyloid-β 42 filaments from human brains. Science 375, 167–172.

62.

Redmann

, Benavides

, Berryhill

, Wani

, Ouyang

, Johnson

, Ravi

, Barnes

, Darley-Usmar

, Zhang

(2016) Inhibition of autophagy with bafilomycin and chloroquine decreases mitochondrial quality and bioenergetic function in primary neurons. Redox Biol 11, 73–81.

63.

Dai

, Ciccotosto

, Cappai

, Wang

, Tang

, Hoyer

, Schneider

, Velkov

, Xiao

(2018) Rapamycin confers neuroprotection against colistin-induced oxidative stress, mitochondria dysfunction, and apoptosis through the activation of autophagy and mTOR/Akt/CREB signaling pathways. ACS Chem Neurosci 9, 824–837.

64.

Zimmerman

, Biggers

, Li

(2018) Rapamycin treatment increases hippocampal cell viability in an mTOR-independent manner during exposure to hypoxia mimetic, cobalt chloride. BMC Neurosci 19, 82.

65.

Israel

, Yuan

, Bardy

, Reyna

, Mu

, Herrera

, Hefferan

, Van Gorp

, Nazor

, Boscolo

, Carson

, Laurent

, Marsala

, Gage

, Remes

, Koo

, Goldstein

LSB

(2012) Probing sporadic and familial Alzheimer’s disease using induced pluripotent stem cells. Nature 482, 216–220.

66.

Almeida

, Zhang

, Coppola

, Mao

, Futai

, Karydas

, Geschwind

, Tartaglia

, Gao

, Gianni

, Sena-Esteves

, Geschwind

, Miller

, Farese

, Gao

F-B

(2012) Induced pluripotent stem cell models of progranulin-deficient frontotemporal dementia uncover specific reversible neuronal defects. Cell Rep 2, 789–798.

67.

Autar

, Guo

, Rumsey

, Long

, Akanda

, Jackson

, Narasimhan

, Caneus

, Morgan

, Hickman

(2021) A functional hiPSC-cortical neuron differentiation and maturation model and its application to neurological disorders. Stem Cell Rep 17, 96–109.

68.

Herrera

, Morcuende

, Talaverón

, Benítez-Temiño

, Pastor

, Matarredona

(2021) Purinergic receptor blockade with suramin increases survival of postnatal neural progenitor cells. Int J Mol Sci 22, 713.

69.

Malheiro

, Morgan

, Baker

, Moroni

, Wieringa

(2020) A three-dimensional biomimetic peripheral nerve model for drug testing and disease modelling. Biomaterials 257, 120230.

70.

Rana

, Luerman

, Hess

, Rubitski

, Adkins

, Somps

(2017) Utilization of iPSC-derived human neurons for high-throughput drug-induced peripheral neuropathy screening. Toxicol In Vitro 45, 111–118.

71.

Wang

, Wang

, Ranefall

, Broussard

, Wang

, Shi

, Lyu

, Wu

C-T

, Wang

, Tian

, Yu

(2020) SynQuant: An automatic tool to quantify synapses from microscopy images. Bioinformatics 36, 1599–1606.

72.

Zeballos

, Bernstein

, Thompson

, Forfia

, Seyedi

, Shen

, Kaminski

, Wolin

, Hintze

(1995) Pharmacodynamics of plasma nitrate/nitrite as an indication of nitric oxide formation in conscious dogs. Circulation 91, 2982–2988.

73.

Ståhle

, Wold

(1987) Partial least squares analysis with cross-validation for the two-class problem: A Monte Carlo study. J Chemom 1, 185–196.

74.

Barker

, Rayens

(2003) Partial least squares for discrimination. J Chemom 17, 166–173.

75.

Ruiz-Perez

, Guan

, Madhivanan

, Mathee

, Narasimhan

(2020) So you think you can PLS-DA? . BMC Bioinformatics 21, 2.

76.

Bao

, Ledderose

, Graf

, Brix

, Birsak

, Lee

, Zhang

, Junger

(2015) mTOR and differential activation of mitochondria orchestrate neutrophil chemotaxis. J Cell Biol 210, 1153–1164.

77.

Lin

, Han

, Weng

, Wang

, Chen

(2018) Rapamycin inhibits proliferation and induces autophagy in human neuroblastoma cells. Biosci Rep 38, BSR20181822.

78.

Wang

, Ma

, Zhang

, Liu

, Wang

(2017) Role of mammalian target of rapamycin signaling in autophagy and the neurodegenerative process using a senescence accelerated mouse-prone 8 model. Exp Ther Med 14, 1051–1057.

79.

Sotthibundhu

, McDonagh

, von Kriegsheim

, Garcia-Munoz

, Klawiter

, Thompson

, Chauhan

, Krawczyk

, McInerney

, Dockery

, Devine

, Kunath

, Barry

, O’Brien

, Shen

(2016) Rapamycin regulates autophagy and cell adhesion in induced pluripotent stem cells. Stem Cell Res Ther 7, 166.

80.

, Hao

, Guo

Y-Y

, Yang

G-L

, Mei

, Li

X-H

, Zhai

Q-X

(2018) Chloroquine inhibits autophagy and deteriorates the mitochondrial dysfunction and apoptosis in hypoxic rat neurons. Life Sci 202, 70–77.

81.

Tan

, Li

, Meng

, Song

, Fang

, Li

, Wang

, Miao

, Zhang

(2018) Autophagy activation aggravates hippocampal neuronal apoptosis caused by intermittent hypoxia. Oncotarget 9, s55–s70.

82.

Nalbandian

, Llewellyn

, Nguyen

, Yazdi

, Kimonis

(2015) Rapamycin and chloroquine: The and effects of autophagy-modifying drugs show promising results in valosin containing protein multisystem proteinopathy. PLoS One 10, e0122888.

83.

Fletcher

, Evans

, Watts

, Jimenez

, Digicaylioglu

(2013) Rapamycin treatment improves neuron viability in an model of stroke. PLoS One 8, e68281.

84.

Lee

, Lim

, Park

, Koh

(2016) S6K promotes dopaminergic neuronal differentiation through PI3K/Akt/mTOR-dependent signaling pathways in human neural stem cells. Mol Neurobiol 53, 3771–3782.

85.

Chen

, Liu

, Song

, Wang

, Dong

, Hu

, Lan

, Guan

, Zhou

, Xu

, Lei

, Ye

, Peng

, Du

, Zhuang

(2010) Rapamycin regulates Akt and ERK phosphorylation through mTORC1 and mTORC2 signaling pathways. Mol Carcinog 49, 603–610.

86.

Sarbassov

, Ali

, Sengupta

, Sheen

J-H

, Hsu

, Bagley

, Markhard

, Sabatini

(2006) Prolonged rapamycin treatment inhibits mTORC2 assembly and Akt/PKB. Mol Cell 22, 159–168.

87.

Ahn

, Jang

, Choi

, Lee

, Oh

S-H

, Lee

, Yoon

, Kim

(2014) GSK3β, but not GSK3α, inhibits the neuronal differentiation of neural progenitor cells as a downstream target of mammalian target of rapamycin complex1. Stem Cells Dev 23, 1121–1133.

88.

Wang

, Zhong

, Li

, Cui

, Huang

(2018) Autophagy of macrophages is regulated by PI3k/Akt/mTOR signalling in the development of diabetic encephalopathy. Aging 10, 2772–2782.

89.

Borack

, Dickinson

, Fry

, Reidy

, Markofski

, Deer

, Jennings

, Volpi

, Rasmussen

(2021) Effect of the lysosomotropic agent chloroquine on mTORC1 activation and protein synthesis in human skeletal muscle. Nutr Metab 18, 61.

90.

El-Gowily

, Loutfy

, Ali

EMM

, Mohamed

, Mansour

(2021) Tioconazole and chloroquine act synergistically to combat doxorubicin-induced toxicity via inactivation of PI3K/AKT/mTOR signaling mediated ROS-dependent apoptosis and autophagic flux inhibition in MCF-7 breast cancer cells. Pharmaceuticals 14, 254.

91.

Citri

, Malenka

(2008) Synaptic plasticity: Multiple forms, functions, and mechanisms. Neuropsychopharmacology 33, 18–41.

92.

Rouleau

, Cantley

, Liaudanskaya

, Berk

, Du

, Rusk

, Peirent

, Koester

, Nieland

TJF

, Kaplan

(2020) A long-living bioengineered neural tissue platform to study neurodegeneration. Macromol Biosci 20, e2000004.

93.

Wiedenmann

, Franke

(1985) Identification and localization of synaptophysin, an integral membrane glycoprotein of Mr 38,000 characteristic of presynaptic vesicles. Cell 41, 1017–1028.

94.

Colom-Cadena

, Spires-Jones

, Zetterberg

, Blennow

, Caggiano

, DeKosky

, Fillit

, Harrison

, Schneider

, Scheltens

, de Haan

, Grundman

, van Dyck

, Izzo

, Catalano

, the Synaptic Health Endpoints Working Group (2020) The clinical promise of biomarkers of synapse damage or loss in Alzheimer’s disease. Alzheimers Res Ther 12, 21.

95.

Cieślak

, Wojtczak

(2018) Role of purinergic receptors in the Alzheimer’s disease. Purinergic Signal 14, 331–344.

96.

Donato

, Miljan

, Hines

, Aouabdi

, Pollock

, Patel

, Edwards

, Sinden

(2007) Differential development of neuronal physiological responsiveness in two human neural stem cell lines. BMC Neurosci 8, 36.

97.

Song

, Subramanian

, Berberich

, Rodriguez

, Latorre

, Luria

, Everley

, Albers

, Mitchison

, Sorger

(2019) A dynamic view of the proteomic landscape during differentiation of ReNcell VM cells, an immortalized human neural progenitor line. Sci Data 6, 190016.

98.

Fröhlich

, Jaeger

, Weiss

, Kriehuber

(2016) Inhibition of BCL-2 leads to increased apoptosis and delayed neuronal differentiation in human ReNcell VM cells. Int J Dev Neurosci 48, 9–17.

99.

Sunami

, Byrne

, Diehl

, Funabashi

, Hall

, Ikuta

, Patel

, Shipman

, Smith

, Takahashi

, Zugay-Murphy

, Iwasawa

, Lumb

, Munshi

, Sharma

(2010) Structural basis of human p70 ribosomal S6 kinase-1 regulation by activation loop phosphorylation. J Biol Chem 285, 4587–4594.

100.

Oddo

(2012) The role of mTOR signaling in Alzheimer disease. Front Biosci 4, 941–952.

101.

W-L

, Cowburn

, Li

, Braak

, Alafuzoff

, Iqbal

I-G

, Winblad

, Pei

J-J

(2003) Up-regulation of phosphorylated/activated p70 S6 kinase and its relationship to neurofibrillary pathology in Alzheimer’s disease. Am J Pathol 163, 591–607.

102.

McMahon

, Choi

, Lin

T-A

, Abraham

, Lawrence

(2002) The rapamycin-binding domain governs substrate selectivity by the mammalian target of rapamycin. Mol Cell Biol 22, 7428–7438.

103.

Edinger

, Thompson

(2004) An activated mTOR mutant supports growth factor-independent, nutrient-dependent cell survival. Oncogene 23, 5654–5663.

104.

Figueiredo

, Markworth

, Cameron-Smith

(2017) Considerations on mTOR regulation at serine 2448: Implications for muscle metabolism studies. Cell Mol Life Sci 74, 2537–2545.

105.

Tremblay

, Marette

(2001) Amino acid and insulin signaling via the mTOR/p70 S6 kinase pathway. J Biol Chem 276, 38052–38060.

106.

Sen

, Saha

, Jiang

, Sen

(2020) Sulfhydration of AKT triggers Tau-phosphorylation by activating glycogen synthase kinase 3β in Alzheimer’s disease. Proc Natl Acad Sci U S A 117, 4418–4427.

107.

Aquilano

, Baldelli

, Ciriolo

(2014) Glutathione: New roles in redox signaling for an old antioxidant. Front Pharmacol 5, 196.

108.

Panday

, Talreja

, Kavdia

(2020) The role of glutathione and glutathione peroxidase in regulating cellular level of reactive oxygen and nitrogen species. Microvasc Res 131, 104010.

109.

Ansari

, Scheff

(2010) Oxidative stress in the progression of Alzheimer disease in the frontal cortex. J Neuropathol Exp Neurol 69, 155–167.

110.

Singh

, Kukreti

, Saso

, Kukreti

(2019) Oxidative stress: A key modulator in neurodegenerative diseases. Molecules 24, 1583.

111.

Tohgi

, Abe

, Yamazaki

, Murata

, Isobe

, Ishizaki

(1998) The cerebrospinal fluid oxidized NO metabolites, nitrite and nitrate, in Alzheimer’s disease and vascular dementia of Binswanger type and multiple small infarct type. J Neural Transm 105, 1283–1291.

112.

Hulshof

, van Nuijs

, Hol

, Middeldorp

(2022) The role of astrocytes in synapse loss in Alzheimer’s disease: A systematic review. Front Cell Neurosci 16, 899251.

113.

Calkins

, Reddy

(2011) Amyloid beta impairs mitochondrial anterograde transport and degenerates synapses in Alzheimer’s disease neurons. Biochim Biophys Acta 1812, 507–513.

114.

Callahan

, Vaules

, Coleman

(1999) Quantitative decrease in synaptophysin message expression and increase in cathepsin D message expression in Alzheimer disease neurons containing neurofibrillary tangles. J Neuropathol Exp Neurol 58, 275–287.

115.

Jang

, Liew

, Kim

Y-M

, Choi

, Kim

, Hyung Lee

, Ohshima

, Mikoshiba

, Suh

Y-H

(2013) p35 deficiency accelerates HMGB-1-mediated neuronal death in the early stages of an Alzheimer’s disease mouse model. Curr Alzheimer Res 10, 829–843.

116.

Heffernan

, Eastwood

, Nagy

, Sanders

, McDonald

, Harrison

(1998) Temporal cortex synaptophysin mRNA is reduced in Alzheimer’s disease and is negatively correlated with the severity of dementia. Exp Neurol 150, 235–239.

117.

Poirel

, Mella

, Videau

, Ramet

, Davoli

, Herzog

, Katsel

, Mechawar

, Haroutunian

, Epelbaum

, Daumas

, El Mestikawy

(2018) Moderate decline in select synaptic markers in the prefrontal cortex (BA9) of patients with Alzheimer’s disease at various cognitive stages. Sci Rep 8, 938.

118.

Puchałowicz

, Tarnowski

, Baranowska-Bosiacka

, Chlubek

, Dziedziejko

(2014) P2X and P2Y receptors—role in the pathophysiology of the nervous system. Int J Mol Sci 15, 23672–23704.

119.

Mut-Arbona

, Sperlágh

(2023) P2 receptor-mediated signaling in the physiological and pathological brain: From development to aging and disease. Neuropharmacology 233, 109541.

120.

Illes

, Rubini

, Ulrich

, Zhao

, Tang

(2020) Regulation of microglial functions by purinergic mechanisms in the healthy and diseased CNS. Cells 9, 1108.

121.

Sarkar

, Abujamra

, Loew

, Forman

, Perrine

, Faller

(2011) Histone deacetylase inhibitors reverse CpG methylation by regulating DNMT1 through ERK signaling. Anticancer Res 31, 2723–2732.

122.

Mandal

, Shukla

, Tripathi

, Ersland

(2019) Cognitive improvement with glutathione supplement in Alzheimer’s disease: A way forward. J Alzheimers Dis 68, 531–535.

Impact of Suramin on Key Pathological Features of Sporadic Alzheimer’s Disease-Derived Forebrain Neurons

Abstract

Background:

Objective:

Methods:

Results:

Conclusions:

Keywords

INTRODUCTION

METHODS

Comparative analysis of Akt/mTOR modulation by suramin, chloroquine, and rapamycin in ReNcell-derived forebrain neurons

ReNcell culture, differentiation, and treatment

Immunocytochemistry analysis of ReNcell differentiation

Assessment of cell density and autophagy modulation via Akt/mTOR signaling pathway

Analysis of suramin treatment on forebrain neurons derived from iPSCs from sporadic AD relative to healthy controls

iPSC culture, differentiation, and treatment

Immunostaining and semi-quantitative image analysis

Aβ42/40 and p-Tau(S199)/total tau enzyme-linked immunosorbent assays (ELISAs)

Assessment of oxidative/nitrosative stress via measurement of glutathione and oxidized nitric oxide

Assessment of cell density and autophagy modulation via Akt/mTOR signaling pathway

Data analysis

Univariate data analysis

Partial least squares discriminant analysis (PLS-DA)

RESULTS

Suramin effects relative to well-characterized modulators of autophagy

Late onset sAD forebrain neuronal expression of canonical disease markers

Effect of suramin treatment on AD marker expression in sAD forebrain neurons

Multivariate analysis of AD pathology and suramin treatment via PLS-DA

Effect of suramin treatment on Akt/mTOR signaling in sAD forebrain neurons

DISCUSSION

AUTHOR CONTRIBUTIONS

Footnotes

ACKNOWLEDGMENTS

FUNDING

CONFLICT OF INTEREST

DATA AVAILABILITY

References

Aβ_42/40 and p-Tau(S199)/total tau enzyme-linked immunosorbent assays (ELISAs)