Lithium chloride suppresses ferroptosis of induced pluripotent stem cells with ApoE4/E4 from a sporadic Alzheimer's disease patient

TL;DR

Lithium chloride helps protect brain cells from a type of cell death linked to Alzheimer's disease in cells with a specific genetic risk factor.

Contribution

This study shows lithium chloride reverses ferroptosis in ApoE4/E4 iPSCs from a sporadic Alzheimer's patient.

Findings

Lithium treatment improved cell viability and reversed ferroptosis markers in ApoE4/E4 iPSCs.

Lithium normalized mitochondrial function and reduced ROS production in these cells.

Lithium inhibited Ca2+ signaling by reducing InsP3R-1 protein expression.

Abstract

Alzheimer's disease (AD), particularly its sporadic form (SAD, 95 % AD patients), is strongly associated with the apolipoprotein E4 (ApoE4) genotype and characterized by oxidative stress, iron dysregulation, and increased susceptibility to ferroptosis. Lithium, a well-established neuroprotective agent, has shown potential to mitigate several pathological mechanisms in AD, including ferroptosis. This study investigates the therapeutic potential of lithium chloride in human induced pluripotent stem cells (iPSCs) derived from a SAD patient with ApoE4/E4 genotype and compared effects with those of isogenic gene-edited ApoE3/E3 control. Lithium treatment significantly improved cell viability in ApoE4/E4 iPSCs. It also reversed key ferroptosis phenotypes, including elevated cytosolic Fe2+, increased expression of divalent metal transporter 1, reduced level of glutathione peroxidase 4, enhanced lipid peroxidation and excessive ROS production. Moreover, lithium (0.25 mM) normalized mitochondrial respiration and reduced proton leak, indicating preservation of mitochondrial function and protection against mitochondrial damage and cell death. Lithium also reduced the expression of type 1 InsP3 receptor (InsP3R-1) protein, a Ca2+ channel located on the endoplasmic reticulum (ER) membrane. Together, these findings highlight lithium's inhibition of ferroptosis through modulation of iron metabolism, antioxidant defenses and inhibition of disrupted Ca2+ signaling. Given the drug's demonstrated efficacy in reversing ApoE4-driven cellular vulnerabilities, lithium salt warrants further investigation for the treatment of AD.

Figures

Full text

Introduction

Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by cognitive decline, neuronal loss and the accumulation of amyloid-beta plaques together with hyperphosphorylated tau. Sporadic AD (SAD), the most generic form, is strongly influenced by the apolipoprotein E4 (ApoE4) allele, which exacerbates the AD pathology through multiple pathways including oxidative stress, mitochondrial dysfunction, iron dysregulation and Ca^2+^ homeostasis disruption [1,2]. Knockdown of ApoE4 with ApoE-specific antisense oligonucleotides or siRNA leads to significant alleviation of AD pathologies in mouse disease models [3]. Ferroptosis, an iron-dependent form of programmed cell death, has emerged as a critical contributor to AD associated neurodegeneration and cognitive dysfunction [4]. This process is driven by elevated intracellular iron (Fe^2+^) levels mediated by upregulation of divalent metal transporter 1 (DMT1), resulting in reactive oxygen species (ROS) accumulation, lipid peroxidation and cell death [5,6]. ApoE4-expressing cells exhibit heightened susceptibility to ferroptosis due to reduced glutathione peroxidase 4 (GPX4) activity, a key enzyme mitigating lipid peroxidation [7,8]. ApoE4 has been demonstrated to exacerbate ferroptosis in AD [7,8]. Moreover, both Ca^2+^ and Fe^2+^ dysregulations contribute to ferroptosis in AD [9,10].

Lithium, a well-established mood stabilizer, exhibits neuroprotective effects in preclinical models of neurodegenerative diseases, notably AD, via multiple mechanisms including modulating oxidative stress, excitotoxicity, Ca^2+^ dysregulation and cellular signaling pathways [[11], [12], [13]]. A recent study indicates that a deficiency of lithium in the brain increases the risk of developing AD [14]. It is believed that inhibition of glycogen synthase kinase-3 and inositol monophosphatase are two primary initial events by which lithium exerts neuroprotection [reviewed in 15]. Lithium enhances GPX4 expression and activity through inhibition of glycogen synthase kinase-3 and activation of the nuclear factor erythroid 2-related factor 2 pathway, bolstering antioxidant defenses [[15], [16], [17]]. Conversely, inhibition of inositol monophosphatase by lithium triggers the induction of autophagy, a process prominently involved in the removal of pathological proteins and frequently dysregulated in brain disorders including AD [18,19]. Additionally, lithium reduces iron dysregulation by downregulating DMT1 expression, thereby limiting Fe^2+^ accumulation and ferroptosis [5]. Furthermore, lithium ameliorates ApoE4-and N-methyl-d-aspartate (NMDA) receptor-mediated Ca^2+^ dysregulation by inhibiting NMDA receptor-mediated Ca^2+^ influx and Src tyrosine kinase activity, mitigating excitotoxicity and oxidative stress that exacerbates ferroptosis in AD [20,21]. Lithium provides synergistic inhibition of ferroptosis via modulation of tau phosphorylation and MAPK signaling in a cell model of AD [22]. On the other hand, lithium enhances ferroptosis in melanoma cells [23]. Induced pluripotent stem cells (iPSCs) derived from SAD patients with ApoE4/E4 genotype provide a robust model to study these mechanisms, as they recapitulate disease-specific phenotypes, including elevated Fe^2+^, ROS, mitochondrial dysfunction, and Ca^2+^ dysregulation [24,25].

This study investigates lithium's role in inhibiting ferroptosis in iPSCs derived from a SAD patient with parental ApoE4/E4 genotype, in comparison to the gene-edited isogenic ApoE3/E3, which is neutral in its association with AD occurrences. This is an excellent translational cell model of SAD to evaluate a new drug's therapeutic efficacy and underlying mechanisms, considering that ApoE4 is one of the primary risk factors for SAD. We hypothesize that lithium exerts cytoprotective effects in these iPSCs, at least in part, by inhibition of ferroptosis. In support of this hypothesis, we show here that lithium reduces DMT1-mediated Fe^2+^ accumulation and ameliorates mitochondrial dysfunction and other detrimental cellular events to prevent cell death. These findings provide insight into lithium's therapeutic potential as a cytoprotective agent against ferroptosis in AD.

Materials and Methods

Cell culture

Cell models featuring the primary susceptibility factor for SAD, namely ApoE4, and iPSCs were procured from the laboratory of Dr. Li-Huei Tsai [1], Massachusetts Institute of Technology. ApoE3/E3 cells were derived from CRISPR/Cas9 editing of SAD iPSCs homozygous for ApoE4/E4 (AG10788 cell line from an 87-year-old familial female AD patient). The expression levels of ApoE in the iPSCs in this study were not determined, but the ApoE protein level in astrocytes differentiated from the iPSCs of a patient with parental ApoE3/E3 was decreased in gene edited isogenic ApoE4/E4 astrocytes [1]. The human iPSCs were cultivated on Matrigel-coated plates from BD Biosciences, USA, utilizing mTeSR™ Plus medium (catalog No. 100–0276, Stem Cell Technologies, Canada), and were incubated in a 5 % CO_2_ humidified atmosphere at 37 °C. The culture medium underwent daily replacement. Before experiments, a regular assessment of cell health was conducted and any unhealthy cells, such as nonadherent cells, were systematically eliminated. The cells were randomly allocated into two distinct experimental groups.

Cell viability assessment

To evaluate cell viability in distinct wells of 96-well plates, we employed the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; Sigma–Aldrich, USA) reduction assay at a 24-h time point, using a previously established protocol [26,27]. Following a wash with phosphate-buffered saline, the samples underwent incubation with a fresh culture medium containing MTT (0.5 mg/ml in the medium) at 37 °C for 4 h in the absence of light. Subsequently, the medium was aspirated, and formazan reaction products were solubilized with dimethyl sulfoxide. The resulting absorbance was measured at 570 nm using a plate reader (Synergy H1 microplate reader, BioTek, CA, USA).

Measurement of mitochondrial function using a Seahorse XFp extracellular flux analyzer

In the pursuit of measuring oxygen consumption rate (OCR) and proton leak in iPSCs, a sophisticated methodology was employed, utilizing Seahorse XFp miniplates, the cell energy phenotype test and the Seahorse XFp Analyzer software. iPSCs were meticulously cultured and plated in Seahorse XFp miniplates to ensure optimal cell adherence under conditions that sustain cellular health. After cell adherence, the Seahorse XFp Analyzer was utilized to conduct the cell energy phenotype test, a dynamic assay providing real-time insights into critical parameters such as OCR and proton leak, pivotal indicators of cellular bioenergetics. The Seahorse XFp Analyzer diligently monitors iPSC OCR, reflecting the rate at which cells consume oxygen during oxidative phosphorylation. Measurements were conducted under basal conditions, and the reactions were augmented by the sequential addition of mitochondrial modulators, including oligomycin, carbonyl cyanide-4 (trifluoromethoxy) phenylhydrazone (FCCP) and antimycin A/rotenone. These additions facilitated the assessment of key mitochondrial functions, encompassing ATP production, maximal respiration and spare respiratory capacity. Proton leak, a pivotal facet of mitochondrial function, underwent specific evaluation, denoting the passive leakage of protons across the inner mitochondrial membrane, contributing to basal OCR. This provided insight into mitochondrial uncoupling and energy inefficiencies. The Seahorse XFp Analyzer, coupled with appropriate inhibitors, facilitated the precise quantification of proton leak in iPSCs. Data analysis was performed via web-based Seahorse software and results were normalized based on cell viability upon completion of the assay.

Intracellular MDA concentration assay for lipid peroxidation

Lipid peroxidation levels were evaluated by quantifying the cellular content of malondialdehyde (MDA) using the colorimetric Lipid Peroxidation Assay kit (ab118970, Abcam, Cambridge, UK), according to the manufacturer's instructions. In this procedural framework, MDA within the sample underwent a reaction with thiobarbituric acid (TBA), producing an MDA-TBA adduct, the colorimetric quantification of which was subsequently expressed as MDA levels. Specifically, following a 24-h drug treatment, iPSCs were collected and homogenized in 300 μL of MDA lysis solution. To generate the MDA-TBA adduct, 600 μL of TBA reagent was added to 200 μL of the sample. The resultant mixture was incubated at 95 °C for 60 min and then cooled on ice. The absorbance of the resulting supernatant, containing the MDA-TBA adduct, was measured at 532 nm using a microplate reader, and MDA levels (μM) were calculated by comparing the absorbance of each sample to a standard.

Intracellular ferrous iron concentration assay

The Cell Ferrous Iron Colorimetric Assay Kit (EBC-K881-M, Elabscience, Houston, Texas, USA) was employed for the quantification of intracellular ferrous iron levels. Approximately 1 × 10^6^ cells were extracted and homogenized on ice for 10 min with 200 μL of lysis buffer, followed by centrifugation at 15,000×g for 10 min to collect the supernatant. Subsequently, 80 μL of the supernatant underwent treatment with the iron probe or the control reagent for 10 min at 37 °C. A Multiscan FC plate reader (Synergy H1 microplate reader, BioTek, USA) was utilized to measure the absorbance at 593 nm. The relative ferrous iron levels were determined as the difference between the ferrous iron contents of the experimental and control groups. A standard curve for cell ferrous iron was employed to calculate the cellular ferrous iron content.

Intracellular ROS assay

To assess intracellular levels of reactive oxygen species (ROS), we conducted the 2′,7′-Dichlorofluorescein diacetate, Diacetyldichlorofluorescein (DCFDA) assay following a meticulous microplate protocol. Initially, iPSCs were seeded and allowed to be attached in a 96-well plate. A washing step with an appropriate buffer was implemented to eliminate extraneous substances that could potentially interfere with the assay. Subsequently, the iPSCs were subjected to staining with DCFDA, a fluorogenic dye sensitive to ROS. The staining duration was 45 min to ensure optimal dye penetration. Post-staining, thorough washing was performed to remove excess, unincorporated dye, ensuring a specific and accurate assessment of intracellular ROS levels. The stained iPSC samples remained in the 96-well plate, distributed evenly across the wells. Employing a plate reader (Synergy H1 microplate reader, BioTek, USA), fluorescence measurements were then taken at the specified excitation and emission wavelengths (Ex/Em = 485/535 nm) for DCFDA. The recorded fluorescence readings for each well were subsequently analyzed, providing a quantitative measure of intracellular ROS levels in the iPSC population.

Western blotting analysis

The Western blotting procedures adhered to established standards. Total protein extracts from iPSCs were acquired by lysing cells in cold RIPA buffer (9806S, Cell Technology, CA, USA) supplemented with protease inhibitor cocktails (P8340 Roche). Following centrifugation, the supernatant was collected, and total protein quantification was performed using a bicinchoninic acid protein assay kit (Thermo Scientific, MA, USA). Equal quantities of protein for each lane were loaded and subjected to separation on a 4–20 % TGX gel (BioRad, CA, U.S.A.) through electrophoresis. Post-electrophoresis, the proteins were transferred onto a polyvinylidene fluoride membrane. Subsequently, the membranes were obstructed with 5 % bovine serum albumin dissolved in phosphate-buffered saline-T for 1 h at room temperature and then probed with the primary antibody (GPX4, Cell Signaling Technology, Dilution 1:1000), DMT-1 (20507-1-AP, Dilution, 1:2000), and InsP_3_R-1 (Catalog # PA1-901, Invitrogen, CA, U.S.A, Dilution, 1:1000) at 4 °C overnight. After thorough washing with phosphate-buffered saline-T, the membranes underwent incubation with secondary antibodies (horseradish peroxidase–conjugated anti-rabbit and anti-mouse IgG) at 1:10,000 dilutions, with GAPDH serving as a loading control. Signals were detected through an enhanced chemiluminescence detection system (Millipore, MA, USA) and quantified using scanning densitometry.

Data analysis and statistics

Prior to the study, no formal statistical power analysis was conducted, and the determination of the sample size was based on our previous experience with a similar experimental design. The normality of data distribution was evaluated using the Tukey test and Fisher test, which guided the decision on whether to employ parametric or nonparametric statistical analyses. Variables meeting the assumptions for parametric analysis were presented as mean ± SD and subjected to either one-way or two-way analysis of variance, followed by Sidak's post hoc analysis. Consideration was given to factors such as the nature of factors (e.g., repeated measures) and the grouping of these factors in the analysis of variance. Variables meeting the assumptions for nonparametric analysis were scrutinized using the Tukey multiple comparisons test and Fisher's Exact test. Statistical analyses and graphical representations were conducted using GraphPad Prism software (GraphPad Software, Inc., MA, USA). A significance threshold of P < 0.05 was employed to determine statistical significance.

Results

Lithium enhances viability and restores intracellular Ca2+ homeostasis in ApoE4/E4 iPSCs

To evaluate the cytoprotective potential of lithium in SAD-relevant cell lines, cell viability was assessed in parental ApoE4/E4 and isogenic gene-modified ApoE3/E3 iPSCs using the MTT assay, which measures mitochondrial dehydrogenase activity. ApoE4/E4 cells demonstrated significantly reduced viability in comparison to ApoE3/E3 controls, consistent with genotype-associated cellular vulnerability (Fig. 1A). Treatment with a low-dose (0.25 mM) or high-dose (1.5 mM) lithium chloride for 24 h resulted in a marked improvement in ApoE4/E4 cell viability (Fig. 1A). Thus, lithium confers protective effects against ApoE4-mediated cytotoxicity. To determine if elevation of cytosolic Ca^2+^ concentration plays a role in ApoE4 mediated iPSCs damage, we used BAPTA-AM to ameliorate the increase of cytosolic Ca^2+^, which dose-dependently promoted ApoE4/E4 iPSC survival (Fig. 1B). On the other hand, NMDA, which selectively activates the NMDA subtype of glutamate receptors on plasma membrane and increases cytosolic Ca^2+^ concentration, induced more potent and robust cell damage in ApoE4/E4 than in ApoE3/E3 iPSCs (Fig. 1C), and these NMDA-induced cytotoxic effects were abolished by lithium treatment (Fig. 1D).Fig. 1Treatment with lithium chloride inhibits ApoE4/NMDA-mediated mitochondrial and cell damage. iPSCs with parental ApoE4/E4 (ApoE4/E4) from a SAD patient or its isogenic gene-edited ApoE3/E3 (ApoE3/E3) were treated for 24 h with 0.25 or 1.5 mM lithium chloride (Li) (A) or indicated concentration of an intracellular Ca^2+^ chelator BAPTA-AM to reduce levels of cytosolic Ca^2+^ (B). Cells were also exposed to various concentrations (C) or 1 μM (D) of NMDA for 24 h in the absence or presence of pretreatment with Li (0.25 mM) for 1 h followed by continuous Li treatment during NMDA exposure. Cell viability was then determined by MTT assay. Lower MTT value corresponds to a greater level of mitochondrial and cell damage. Data are presented as mean ± SD from 6 repeats of 4–8 separate experiments (N = 24–32 repeats). Statistical significance was analyzed by two-way ANOVA followed by Tukey multiple comparison tests. ∗∗∗∗P < 0.0001 (A, D). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001 compared to control in same type of cells (B, C). ^#^P < 0.05, ^##^P < 0.01, ^###^P < 0.001, ^####^P < 0.0001 compared to ApoE3 **(B, C).**Fig. 1

Lithium suppresses ApoE4/NMDA-mediated increases of type 1 InsP3R proteins

To determine whether lithium modulates Ca^2+^ signaling through regulation of type 1 inositol 1,4,5-trisphosphate receptor (InsP_3_R-1) expression, protein levels of InsP_3_R-1 were assessed using Western blotting (Fig. 2). ApoE4/E4 cells exhibited a three to four-fold increase in InsP_3_R-1 protein levels compared to ApoE3/E3 controls. Lithium treatment completely blocked the InsP_3_R-1 protein upregulation in ApoE4/E4 cells, supporting that lithium suppresses ApoE4-associated upregulation of this intracellular Ca^2+^ channel. This finding suggests that lithium inhibits InsP_3_R-1 expression driven by ApoE4 genotype. This result highlights a potential mechanism by which lithium mitigates Ca^2+^ dysregulation in ApoE4-mediated cell damage.Fig. 2Lithium blocks the robust elevation of type 1 InsP_3_R (InsP_3_R1) protein induced by ApoE4. iPSCs with parental ApoE4/E4 (ApoE4/E4; indicated by “+” sign above the lane) from a SAD patient or its isogenic gene-edited ApoE3/E3 (ApoE3/E3; indicated by “+” sign above the lane) were treated with lithium chloride (0.25 mM) or vehicle (H_2_O) or without any treatment (Tx) for 24 h. Western Blotting analyses of the receptor protein were then performed (A). Quantified results are mean ± SD from three independent experiments (B). GAPDH protein level was used as a loading control. Statistical significance was analyzed by two-way ANOVA followed by Tukey's multiple comparison test. ∗∗∗∗P < 0.0001.Fig. 2

Lithium reduces intracellular Fe2+ accumulation in ApoE4/E4 cells

To investigate lithium's influence on iron homeostasis, a major facilitator of cell death by ferroptosis, cytosolic ferrous iron (Fe^2+^) levels were measured using a colorimetric iron assay kit. ApoE4/E4 cells exhibited significantly elevated cytosolic Fe^2+^ concentrations relative to ApoE3/E3 controls (Fig. 3A). Pretreatment with lithium normalized Fe^2+^ levels in ApoE4/E4 cells, restoring them to near-control levels (Fig. 3A). Furthermore, lithium treatment completely suppressed the additional increase in cytosolic Fe^2+^ concentration induced by exposure to NMDA (Fig. 3B) and RSL-3, a GPX4 inhibitor and ferroptosis inducer (Fig. 3C), in ApoE4/E4 iPSCs. NMDA/RSL-3 treatment produced only a trend to increase intracellular iron concentrations, which was ameliorated by LiCl, but the difference did not reach statistical significance.Fig. 3Lithium completely suppresses an increase in cytosolic Fe^2+^ concentration induced by ApoE4, NMDA or a GPX4 inhibitor RSL-3. iPSCs with parental ApoE4/E4 (ApoE4/E4) from a SAD patient or its isogenic gene-edited ApoE3/E3 (ApoE3/E3) were treated with lithium chloride (0.25 mM) for 1 h followed by 24-h exposure to NMDA (1 μM) or RSL-3 (5 μM), as indicated (A, B and C). Data are presented as the mean ± SD from three independent experiments with a total of 9 repeats. Statistical analysis using two-way analysis of variance followed by Tukey multiple comparison tests. ∗P < 0.05, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. RSL-3, Ras selective lethal 3.Fig. 3

Lithium blocks DMT1 overexpression in ApoE4/E4 cells

To determine the potential mechanism through which lithium affects cytosolic iron concentration, we measured the changes in protein level of the iron transporter divalent Metal Transporter 1 (DMT1). ApoE4/E4 cells exhibited a pronounced upregulation of DMT1 compared to ApoE3/E3 controls (Fig. 4). Importantly, lithium treatment markedly reduced the DMT1 protein upregulation in ApoE4/E4 cells to the level in ApoE3/E3 iPSCs. Considering the report that DMT1 can be ubiquitinated and degraded via E3 ubiquitin ligases in conjunction with Ndfip1 in human neurons [28], one surmises that DMT1 downregulation by lithium may involve lithium activation of E3 ligases caused by the drug's interaction with its direct targets such as glycogen synthase kinase-3 and/or inositol monophosphatase.Fig. 4Lithium abolishes ApoE4-mediated pathological elevation of divalent metal transporter 1 (DMT1) protein. iPSCs with parental ApoE4/E4 (ApoE4/E4; indicated by “+” sign above the lane) or its isogenic gene-edited ApoE3/E3 (ApoE3/E3; indicated by “+” sign above the lane) were treated with lithium chloride (0.25 mM) or vehicle (H_2_O), or without any treatment (No Tx) for 24 h. Western blotting analyses of DMT1 were then performed (A). Quantified results are mean ± SD from three independent experiments (B). Statistical significance was analyzed by two-way ANOVA followed by multiple Tukey test. ∗∗∗P < 0.001.Fig. 4

Lithium restores antioxidant defense and reduces oxidative stress

Glutathione Peroxidase 4 (GPX4), an antioxidant enzyme protecting cells from oxidative stress, is typically downregulated in ferroptosis. GPX4 expression was significantly reduced in ApoE4/E4 cells compared to ApoE3/E3 controls (Fig. 5). Lithium treatment restored GPX4 expression to levels comparable to those in ApoE3/E3 cells (Fig. 5A), indicating a reconstitution of lipid peroxide repair capacity.Fig. 5Lithium treatment restores ApoE4-mediated pathological decrease of glutathione peroxidase 4 protein (GPX4) expression. iPSCs with parental ApoE4/E4 (ApoE4/E4; indicated by “+” sign above the lane) or its isogenic gene-edited ApoE3/E3 (ApoE3/E3; indicated by “+” sign above the lane) were treated with lithium chloride (0.25 mM) for 24 h. Cell lysates were then collected for Western blotting to determine levels of GPX4 protein. (A). Representative of Western blotting images of GPX4 proteins expression under different experimental conditions in these two types of cells. (B). Statistical analysis for Western blot data. Data are presented as mean ± SD, from 5 to 6 independent experiments. Two-way ANOVA followed by Tukey multiple comparison tests were used for statistical analyses. ∗P < 0.05; ∗∗P < 0.01.Fig. 5

To determine intracellular oxidative stress level, we measured change in malondialdehyde (MDA), a metabolite of lipid peroxidation and reactive oxygen species (ROS). Intracellular MDA concentration in ApoE4/E4 iPSCs was significantly higher than that in ApoE3/E3 iPSCs. Treatment with lithium chloride at 0.25 mM for 24 h restored the elevated intracellular MDA concentration in ApoE4/E4 iPSCs to the level in ApoE3/E3 iPSCs (Fig. 6A). Treatment with 1 μM NMDA further increased intracellular MDA concentration in ApoE4/E4 iPSCs, which was significantly inhibited by lithium treatment (Fig. 6B). The intracellular ROS levels in ApoE4/E4 iPSCs tended to be higher than those in ApoE3/E3 iPSCs. RSL-3, an inhibitor of GPX4 (Fig. 6C), and NMDA (Fig. 6D) significantly increased intracellular ROS levels in ApoE4/E4 but not ApoE3/E3 iPSCs. Pretreatment with lithium for 24 h effectively ameliorated the RSL-3 (Fig. 6C) or NMDA (Fig. 6D) -induced pathological elevation of intracellular ROS.Fig. 6Lithium inhibits elevations of intracellular ROS and MDA mediated by parental ApoE4 and treatment with NMDA or RSL-3. iPSCs with parental ApoE4/E4 (ApoE4/E4) or its isogenic gene-edited ApoE3/E3 (ApoE3/E3) were treated with lithium chloride (0.25 mM) for 24 h (A), with or without co-treatments of NMDA (1 μM) (B, D) or RSL-3 (5 μM) (C). Levels of intracellular malondialdehyde (MDA) and ROS were then quantified. Results are presented as mean ± SD of 16(A), 12(B), 16(C) and 16(D) repeats from four independent experiments. Data were analyzed by two-way ANOVA followed by Tukey multiple comparison test. **∗∗∗∗**P < 0.0001.Fig. 6

Lithium normalizes mitochondrial oxygen consumption and reduces proton leak in ApoE4/E4 cells

To evaluate the impact of lithium on mitochondrial bioenergetics, oxygen consumption rate (OCR) and proton leak were measured using Seahorse metabolic flux analysis. ApoE4/E4 iPSCs exhibited elevated basal OCR, which was significantly exacerbated by RSL-3 (Fig. 7 A, B, C, E). Pretreatment with lithium chloride at 0.25 mM for 24 h significantly inhibited RSL-3-induced elevation of basal OCR. Proton leak (an indication of mitochondrial and cell damage by apoptosis [29]) in ApoE4/E4 increased in comparison to ApoE3/E3 iPSCs, which was further significantly increased by RSL-3 (Fig. 7 A, B, D, F). Lithium chloride (0.25 mM) pretreatment for 24 h abolished RSL-3-induced proton leaks in ApoE4/E4 iPSCs (Fig. 7D–F).Fig. 7**Lithium inhibits RSL****-**3-induced pathological elevations of basal mitochondrial oxygen consumption rate (OCR) and proton leaks in ApoE4 cells. iPSCs with parental ApoE4/E4 (ApoE4/E4) or its isogenic gene-edited ApoE3/E3 (ApoE3/E3) were treated with lithium chloride (0.25 mM) for 24 h and then challenged with RSL-3 (5 μM) or vehicle for 1 h. Seahorse XF Mito Stress assay was performed to show OCR profiles in iPSCs with ApoE3 (A) or ApoE4 (B). Arrows indicate sequential addition of vehicle, oligomycin (Oligo; 1.5 μM), FCCP (0.5 μM), and rotenone together with antimycin (0.5 μM). Basal respiration OCR was determined under treatment with LiCl or vehicle (C), as well as with or without RSL-3 (E). Control cells received no treatment (No Tx). Mitochondrial proton leak was used as an apoptosis biomarker and calculated based on OCR in these two types of cells under treatment with LiCl or vehicle (D), as well as with or without treatment of RSL-3 (F). Control cells received no treatment (No Tx). Results are presented as mean ± SD from three independent experiments (C–F). Data were analyzed by two-way ANOVA followed by Tukey multiple comparison test. ∗P < 0.05 (F); ∗∗P < 0.01 (D, E), ∗∗∗P < 0.001.Fig. 7

Discussion

This study provides compelling evidence that lithium treatment protects ApoE4/E4 human iPSC-derived neurons from ferroptosis through multimodal mechanisms involving the regulation of iron, Ca^2+^ signaling and antioxidant defense. Specifically, lithium blocked the upregulation of type 1 InsP_3_R proteins, reduced cytosolic Fe^2+^ accumulation, suppressed the increased iron transporter DMT1, restored GPX4 expression, normalized the mitochondrial function, and mitigated lipid peroxidation and ROS generation in ApoE4/E4 cells, with or without glutamate excitotoxicity and additional oxidative stress. These effects were accompanied by a significant increase in cell viability in ApoE4/E4 cells, which are known to confer elevated susceptibility to oxidative and metabolic stressors [[30], [31], [32]].

These findings expand upon a growing body of literature demonstrating lithium's neuroprotective effects in various models of neurodegeneration. Lithium has long been shown to exert neurotrophic and anti-apoptotic effects through inhibition of glycogen synthase kinase-3, enhancement of the signaling of brain-derived neurotrophic factor (BDNF) and other trophic factors, as well as modulation of mitochondrial function [18,33]. In the context of AD, lithium has been associated with reduced amyloid pathology, tau hyperphosphorylation, and neuroinflammation in both preclinical and clinical settings [22,34,35]. However, its role in regulating ferroptosis, a distinct form of regulated cell death driven by iron accumulation and lipid peroxidation, is only beginning to be elucidated [22]. Recent studies have implicated ferroptosis in the pathophysiology of AD, particularly in the presence of the ApoE4 genotype, which has been shown to disrupt lipid metabolism, increase oxidative stress and promote iron accumulation [32,36,37]. Our data are consistent with this genotype-dependent vulnerability and further demonstrate that lithium treatment can reverse these ferroptosis phenotypes in ApoE4/E4 iPSC. This suggests that lithium's cytoprotection may extend beyond traditional anti-apoptotic pathways to include ferroptosis inhibition [22].

Consistent with previous studies in iPSC-derived astrocytes or neurons, ApoE4 increased mitochondrial respiration, suggesting a gain-of-function [38,39]. However, the increased mitochondrial respiration was associated with decreased ATP production and increased proton leak, suggesting futile mitochondrial respiration and mitochondrial stress or damage. In addition, ApoE4 impaired lipid metabolism and increased inflammation in iPSC-derived astrocytes [5]. Interestingly, the ApoE2 or ApoE knockout typically demonstrated the opposite response mediated by ApoE4 [38,40]. Together with the results from the current study that lithium inhibited ApoE4-mediated iron dysregulation and associated mitochondrial damage and oxidative stress, it seems reasonable to speculate that ApoE knockdown or knockout in conjunction with lithium may provide additive or even synergistic protection in ApoE4/E4 cells, especially at sub-optimal concentrations of this drug.

Crucially, the present study identifies a novel mechanism by which lithium inhibits ferroptosis. Previous work has shown that excessive Ca^2+^ influx, such as that triggered by glutamatergic excitotoxicity or NMDA receptor overstimulation, leads to mitochondrial dysfunction and suppression of antioxidant enzymes, including GPX4 [41,42]. Our findings suggest that glutamate excitotoxicity induced by NMDA receptor overactivation plays a pathological role in ApoE4-mediated cell damage. Additionally, lithium attenuates NMDA-induced ROS and lipid peroxidation and preserves GPX4 expression, pointing to a protective role against Ca^2+^-mediated ferroptosis initiation. Notably, lithium-induced inhibition of glycogen synthase kinase-3 activity has been linked to GPX4 overexpression [16]. Further, we demonstrated that lithium inhibits ApoE4-mediated InsP_3_R-1 upregulation. The robust effects of lithium to suppress InsP_3_R-1 upregulation may be contributed by lithium inhibition of inositol monophosphatase, given that changes in levels of inositol trisphosphate or calcium influenced by inositol monophosphatase activity affect InsP_3_R-1 expression [43]. This aligns with earlier reports that lithium can exert neuroprotective effects by modulating intracellular Ca^2+^ dynamics through inhibition of inositol monophosphatase and downregulation of NMDA receptor activity, in addition to blocking glycogen synthase kinase-3 activity to modulate gene transcription [15,20]. Furthermore, ApoE4 has been shown to increase Tau phosphorylation, by activation of glycogen synthase kinase-3 [44]. Accordingly, lithium as an inhibitor of this kinase may decrease tau pathology in AD [45].

Mitochondrial respiration data revealed that lithium normalized aberrant increases in OCR and proton leak triggered by NMDA/RSL-3 exposure in ApoE4/E4 cells, further supporting its role in maintaining mitochondrial bioenergetic homeostasis and reducing apoptosis-prone mitochondrial stress. In this context, lithium's inhibition of oxidative stress and lipid peroxidation contributes to preserved mitochondrial function and overall cell survival. A recent study has demonstrated that lithium deficiency in the brain of AD patients contributes to the onset of AD pathology and that lithium supplement suppresses the pathology and memory loss in AD transgenic mice [14]. The novel findings in our study demonstrate that lithium robustly downregulates DMT1 but upregulates GPX4, suggesting that lithium may protect against AD pathology, at least in part, via correcting the intracellular iron dysregulation and associated programmed cell death by ferroptosis [22]. A schematic illustration of the effects of lithium on the pathological events in ApoE4/E4 iPSCs and potential underlying mechanisms of lithium neuroprotection against ferroptosis in AD is shown and described in detail in Fig. 8.Fig. 8A proposed scheme for lithium-inhibition of mitochondrial dysfunction, oxidative stress and ferroptosis in ApoE4 iPSCs. In AD, excessive elevation of glutamate over-activates cell-surface NMDA receptors (NMDAR), leading to excessive Ca^2+^ influx into cytosol through the receptor channel. This can be aggravated by the SAD risk factor ApoE4. The number and function of InsP_3_ receptors (InsP_3_R) on the endoplasmic reticulum (ER) are pathologically increased in AD, leading to excessive Ca^2+^ release from the ER into the cytosol. Elevated [Ca^2+^]c increases the expression of divalent metal transporter 1 (DMT1), an iron transporter, leading to the pathological elevation of cytosolic Fe^2+^ and further increases of ROS, lipid peroxidation and ferroptosis. RSL-3 induces ferroptosis specifically by inhibiting glutathione peroxidase 4 (GPX4). Lithium provides robust protection by multiple mechanisms including suppression of AD-associated NMDAR over-activation, InsP_3_R upregulation, DMT1 overexpression and GPX4 downregulation. These combined actions of lithium result in normalization of critical Ca^2+^ and Fe^2+^ dysregulation and associated mitochondria dysfunction, oxidative stress and programmed cell death by apoptosis or ferroptosis as well as ultimate cognitive improvement.Fig. 8

Our recent study also demonstrated that lithium inhibits pathological inflammation and programmed cell death through pyroptosis by inhibiting upstream Ca^2+^ dysregulation in the brains of 5XFAD AD transgenic mice [11]. Taken together, lithium may be preventive or therapeutic in AD through its inhibition of neurodegeneration triggered by ferroptosis and/or pyroptosis. It has been proposed that effective pharmacological treatment of AD should inhibit the critical upstream pathology or inhibit multiple pathologies, and lithium fits these requirements. One of the problems of using lithium to treat bipolar disorder is its low therapeutic window (0.6–1.0 mM), rendering lithium treatment prone to side effects or organ toxicity such as thyroid or kidney damage. It has been reported that treatment with lithium at low clinical concentration (0.25–0.5 mEq) improved mild cognitive function in AD patients [46]. Our recent study demonstrated that intranasal administration of lithium chloride in a Ryanodex formulation, rather than in water, significantly enhanced the lithium concentration in the brain and the brain/blood lithium concentration ratio in an AD mouse model, thus potentially promoting lithium's therapeutic effects in the CNS while minimizing the peripheral side effects or organ toxicity.

One limitation of using lithium as a first-line drug to treat bipolar disease is its narrow therapeutic window [29]. The primary organ toxicity is its impairment of thyroid and kidney functions [47,48]. Our recent study in mice demonstrated that intranasal lithium chloride in Ryanodex formulation vehicle (RFV) [49,50] significantly increased lithium brain concentration, duration, and brain/blood lithium concentration ratio, compared to oral lithium chloride in RFV or intranasal lithium chloride dissolved in water [11]. Consistently, intranasal lithium chloride in RFV significantly inhibited memory loss and depressive behavior via inhibition of inflammatory pyroptosis in 5XFAD transgenic AD mice, without obvious adverse effects or organ toxicity toward the kidney or thyroid. In fact, intranasal lithium chloride in RFV protects kidney function in 5XFAD mice [11]. Together with the results demonstrating lithium protection against inflammatory pyroptosis in the current study, we propose that intranasal lithium in RFV may be an effective drug treatment for AD. As ApoE4 is a prominent risk factor for SAD, the capability of lithium to prevent mitochondrial and cell damage in ApoE4/E4 iPSCs from a SAD patient in this study suggests its potential to treat SAD patients, pending future clinical studies.

Overall, these results support a model in which lithium acts at multiple levels to prevent ferroptosis in ApoE4/E4 iPSCs from an SAD patient: (1) reducing iron uptake via DMT1 downregulation possibly through enhanced protein degradation, (2) preserving lipid detoxification capacity through upregulation of GPX4, and (3) inhibiting Ca^2+^- and iron dysregulation associated oxidative stress, a key upstream trigger of ferroptosis death [43,44,46]. This multimodal action highlights lithium as a particularly promising therapeutic candidate for targeting ferroptosis in AD, especially in genetically susceptible populations with the ApoE4/E4 genotype in SAD.

The regulatory influence of lithium on ApoE expression, particularly the ApoE4 isoform, has recently been elucidated as a key factor in maintaining microglial homeostasis. In a landmark 2025 study, Yankner and colleagues demonstrated that endogenous lithium levels serve as a critical regulator of ApoE expression within the brains of AD patients and AD mouse models [14]. High-resolution snRNA-seq analysis revealed that lithium deficiency triggers a significant increase in a specific population of microglia expressing high levels of ApoE in the mouse brain, suggesting that adequate lithium is necessary to suppress this potentially pathogenic state (Fig. 3d in reference 14). Conversely, the study showed that this process is reversible; treating mice with lithium orotate effectively downregulates ApoE expression. These findings align with the observations in our current study. By suppressing the over-expression of ApoE, particularly in the context of the pro-inflammatory ApoE4 allele, lithium treatment may mitigate the microglial dysfunction and neurodegeneration associated with AD.

Despite these promising results, several limitations warrant discussion. The findings are derived from in vitro models, particularly in iPSCs, not neurons, and further validation in animal models or human brain tissue is required to establish translational relevance. Moreover, while the data supports a role for Ca^2+^ in GPX4 regulation, the precise molecular intermediates involved, such as Ca^2+^-dependent transcription factors or signaling kinases, remain to be identified. Lithium-induced expression of GPX4 is regulated by a combination of transcriptional, translational and post-translational mechanisms. Future investigations should include the roles of antioxidants, lipids, oxygen tension and non-coding RNA in lithium-induced GPX4 upregulation and protection against ferroptosis. Lithium is known to activate the Wnt/β-catenin signaling pathway by inhibition of glycogen synthase kinase-3, and a deficit in Wnt signaling has been reported in the brains of AD patients and a mouse AD model [51]. We have not yet assessed whether lithium treatment is able to restore the loss of Wnt signaling and whether this action is critical for the lithium benefits shown in this report. Additionally, future studies should investigate whether ApoE knockdown will ameliorate ApoE4/E4 effects synergistically with lithium. Finally, the present study employed undifferentiated iPSCs with ApoE4/E4 from a single sAD patient; future studies employing additional lines of iPSCs with ApoE4/E4s and neurons, astrocytes and microglia derived from these undifferentiated iPSCs will gain further mechanistic insights. Of interest, ApoE4 impairs lipid metabolism, including lipid droplets, which may play a significant role in AD pathology [[52], [53], [54]] and lithium treatment may decrease lipid droplets. The effects of lithium on lipid metabolism and lipid droplets should be further investigated.

In conclusion, this study identifies lithium as a potent inhibitor of ferroptosis in human ApoE4/E4 iPSCs and provides mechanistic insight into its action via Ca^2+^-dependent regulation of GPX4, illustrating a molecular mechanism of ApoE4-mediated pathology in SAD patients. These findings suggest a neuroprotective mechanism for lithium and support its continued evaluation as a disease-modifying therapy in AD.

Author contributions

H.W. conceived and designed the study. Y.W., S.A., P.B., L.S., G.L. and H.W. conducted experiments and acquired the data. Y.W., S.A., P.B., L.S. and H.W. analyzed data. S.A. D.-M.C. and H.W. evaluated the experimental data and constructed the manuscript. All the authors reviewed and approved the final manuscript.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Huafeng Wei reports financial support was provided by grants from the National Institute on Aging. Huafeng Wei has patent pending to Huafeng Wei. Other authorsthey declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

NIH Disclaimer: The contributions of the NIH author De-Maw Chuang are considered Works of the United States Government. The findings and conclusions presnted in this paper are those of the author and do not necessarily reflect the views of the NIH or the U.S. Department of Health and Human Services.