Dithiolethiones D3T and ACDT Protect Against Iron Overload-Induced Cytotoxicity and Serve as Ferroptosis Inhibitors in U-87 MG Cells

Iron overload-induced oxidative stress is implicated in various neurodegenerative disorders. Given the numerous adverse effects associated with current iron chelators, natural antioxidants are being explored as alternative therapeutic options. Dithiolethiones found in cruciferous vegetables have emerged as promising candidates against a wide range of toxicants owing to their lipophilic and cytoprotective properties. Here, we test the dithiolethiones 3H-1,2-dithiole-3-thione (D3T) and 5-amino-3-thioxo-3H-(1,2) dithiole-4-carboxylic acid ethyl ester (ACDT) against ferric ammonium citrate (FAC)-induced toxicity in U-87 MG astrocytoma cells. Exposure to 15 mM FAC for 24 h resulted in 54% cell death. A 24-h pretreatment with 50 μM D3T and ACDT prevented this cytotoxicity. Both dithiolethiones exhibited antioxidant effects by activating the nuclear factor erythroid 2-related factor-2 (Nrf2) transcription factor and upregulating levels of intracellular glutathione (GSH). This resulted in the successful inhibition of FAC-induced reactive oxygen species, lipid peroxidation, and cell death. Additionally, D3T and ACDT upregulated expression of the Nrf2-mediated iron storage protein ferritin which consequently reduced the total labile iron pool. A 24-h pretreatment with D3T and ACDT also prevented cell death induced by the ferroptosis inducer erastin by upregulating the transmembrane cystine/glutamate antiporter (xCT) expression. The resulting increase in intracellular GSH and alleviation of lipid peroxidation was comparable to that caused by ferrostatin-1, a specific ferroptosis inhibitor. Collectively, our findings demonstrate that dithiolethiones may show promise as potential therapeutic options for the treatment of iron overload disorders.


Introduction
Iron is a double-edged sword wherein both iron deficiency and iron overload conditions are deleterious to human health [1]. Although iron is essential for biological functions, the accumulation of this redox metal leads to the generation of reactive oxygen species (ROS) via the Fenton reaction [2]. ROS react with the polyunsaturated fatty acids in cell membranes leading to lipid peroxidation resulting in oxidative stress and eventual cell death. This underlying mechanism of iron-induced toxicity has also been implicated in neurodegenerative diseases such as Alzheimer's disease (AD), Parkinson's disease (PD), Huntington's disease, and Friedrich's ataxia [3][4][5].
Astrocytes have gained a spotlight as models of ironinduced oxidative stress [6]. Anatomically, astrocytes are located immediately beyond the blood-brain barrier (BBB) serving as an additional protective layer to the neurons beyond. Physiologically, these cells serve as iron stores and gatekeepers for the transport of iron and other toxic substances into the brain [7]. They have been reported to scavenge free iron in the extracellular space of the brain and accumulate and store large amounts of iron, thereby protecting neuronal cells from iron-mediated oxidative stress [8]. Astrocytes, when cultured together with neurons, attenuate the cytotoxic effects of iron [9]. Therefore, any toxic damage to these protective astrocytes is also associated with neurodegeneration [10].
Dithiolethiones are organosulfur compounds that possess chemotherapeutic, anti-inflammatory, and antioxidative properties [11]. They are highly lipophilic and readily cross the BBB [12]. They are known to induce the phase-II antioxidant system via activation of the nuclear factor erythroid 2-related factor-2-Kelch-like ECH-associated protein-1 (Nrf2-Keap1) pathway. The disulfide group in dithiolethiones reacts with molecular oxygen to form hydrogen peroxide. This generates a small amount of oxidative stress disrupting the Nrf2-Keap1 complex and causing the free Nrf2 to translocate to the nucleus, bind to the antioxidant response element (ARE), and upregulate various antioxidant target genes [13].
3H-1,2-Dithiole-3-thione (D3T) and 5-amino-3-thioxo-3H-(1,2) dithiole-4-carboxylic acid ethyl ester (ACDT) are dithiolethiones that have previously shown neuroprotective effects both in vitro and in vivo. For example, D3T, the most potent unsubstituted parent compound in this series, has displayed neuroprotective activity in a murine model of AD [14]. Additionally, D3T has displayed protection in models of neuroinflammation [15], stroke [16], and amyloid beta-induced toxicity [17]. ACDT, a disubstituted dithiolethione with pharmacokinetically favorable functional groups [12] has exhibited protection against neuroinflammation [18], a 6-hydroxydopamine model of PD [19], and manganese-induced neurotoxicity [20]. Overlap in the toxicity profiles of manganese and iron as well as the strong association of iron with neurodegenerative diseases including PD and AD led us to explore the protective role of these dithiolethiones against iron-induced toxicity in human astrocytoma (U-87 MG) cells. We hypothesize that activation of the Nrf2 pathway by dithiolethiones will protect U-87 MG cells against ferric ammonium citrate (FAC)induced oxidative stress and cell death.

Cell Viability Assay
The FAC salt was used to induce iron toxicity owing to its readily water-soluble property and widespread use in past studies [21][22][23]. For a baseline toxicity evaluation, U-87 MG cells were treated with FAC (0-20 mM) for 24 h. To examine the protective ability of dithiolethiones against iron toxicity, the cells were pretreated with 50 µM D3T or ACDT for 24 h before exposure to 15 mM of FAC for an additional 24 h. To evaluate if D3T and ACDT act as ferroptosis inhibitors, cells were pretreated with dithiolethiones or with ferrostatin-1 (positive control), followed by subsequent exposure to the ferroptosis-inducer erastin. As a general setup, pretreated compounds were always aspirated before the addition of the toxicant. The MTS assay was used to determine cell viability as previously described [24]. The principle of this assay involves the reduction of MTS tetrazolium compound by live cells in the cell culture to generate a soluble, colored formazan dye, the absorbance of which is directly proportional to the number of viable cells. Absorbance was measured at a wavelength of 490 nm using the Synergy HT multimode microplate reader (Biotek Instruments, Winooski, VT).

Microscopy
Using an exposure setup similar to the cell viability assay, U-87 MG cells at a density of 1 × 10 6 per group were treated with D3T, ACDT, and FAC alone and in combination. Cell morphology changes were captured at a phasecontrast mode using the Keyence All-in-One Fluorescence Microscope (BZ-X800, Woburn, MA).

Measurement of Intracellular ROS
ROS levels were determined using the DCFDA dye following the manufacturer's protocol. Briefly, 5 × 10 4 cells/well were seeded in 96-well black microplates (3916, Corn-ing®) and were incubated with a 10 µM DCFDA fluorescent probe in phenol red-free media for 45 min. After removal of the dye, cells were treated with compounds and fluorescence signals directly proportional to the levels of intracellular ROS levels were measured at excitation/ emission wavelengths of 495/529 nm.

Measurement of Intracellular Glutathione (GSH) Levels
U-87 MG cells were pretreated with 50 µM D3T or ACDT for 24 h followed by exposure to 15 mM FAC for an additional 24 h. For ferroptosis experiments, 20 µM erastin was used as a ferroptosis inducer. The GSH-Glo glutathione assay kit was used to determine GSH concentration according to the manufacturer's instructions.

Lipid Peroxidation Assay
Treated U-87 MG cells were lysed, and the TBARS assay was performed according to the manufacturer's instructions.
The absorbance values were measured at 540 nm and malondialdehyde (MDA) levels normalized to a standard curve were reported as micromolar concentrations.

Calcein-AM Assay
Cells at a density of 5 × 10 4 cells/well in a clear-bottom black 96-well plate were pretreated with D3T or ACDT before exposure to FAC for 2 h. After aspiration of media, the cells were incubated with 3 µM calcein-AM fluorescent dye for 45 min. The dye was aspirated, and the cells were incubated for another 15 min with the serum-free media. The fluorescence was measured at excitation/emission wavelengths of 485/528 nm using the Synergy HT Multi-Mode Microplate Reader.

Statistical Analysis
Data are reported as mean ± SD. Comparisons between multiple groups were performed using a one-way analysis of variance (ANOVA) followed by an appropriate post hoc test. Results with a p-value ≤ 0.05 were considered statistically significant. Each experiment was performed at least three times (n = 3). GraphPad Prism version 8.0 for Windows (La Jolla, CA) was utilized for statistical analyses and graphical representations of the data.

Impact of Dithiolethiones on FAC-Induced Cell Death
A 24-h exposure of U-87 MG cells to FAC (0-20 mM) resulted in concentration-dependent cytotoxicity (Fig. 1a). As the cell viability decreased to 54% (F 9,20 = 31.62; p < 0.0001) with 15 mM FAC, this concentration was selected for subsequent experiments. Pretreatment with 50 µM of D3T or ACDT significantly increased the cell viability to 94% and 90%, respectively in comparison to the FAC alone-treated group (Fig. 1b) (F 7,16 = 31.05; p < 0.0001). The dithiolethiones alone had no impact on cell viability. Cells treated with 15 mM FAC alone were visually sparse in number and appeared rounded in comparison to those in the control group, while cells pretreated with the dithiolethiones appeared morphologically healthier (Fig. 1c).

Impact of Dithiolethiones on Intracellular Iron Levels
While exposure to FAC alone significantly upregulated (2.4-fold) the expression of the iron storage protein ferritin, pretreatment with D3T and ACDT increased this level even further (4.1-and 3.1-fold), respectively (F 7,16 = 20.24; p < 0.0001, Fig. 3a). There was also a visible trend with D3T alone (1.3-fold) and as a pretreatment to FAC (0.6 to 0.9-fold) upregulating the expression of the iron exporter ferroportin (FPN) (F 5,12 = 3.22; ns, Fig. 3b). A significant decrease in the relative calcein-AM fluorescence on exposure to FAC alone for 2 h (0.5-fold vs control) indicated an increased intracellular labile iron pool, which was successfully countered by a 24-h pretreatment with D3T and ACDT to 0.7-fold each. Notably, treatment with D3T alone also significantly reduced free iron levels (F 5,12 = 131.80; p < 0.0001, Fig. 3c).

Discussion
Iron overload is a serious problem that is often overlooked because of the non-specificity and gradual development of symptoms. Examples of iron-overload induced toxicity are widespread. Hemochromatosis, a common genetic disorder, results in iron accumulation in the liver, heart, and skin [25]. Accidental iron poisoning and death in children are also common because of the candy-like appearance of iron preparations used in treating anemia [26]. Iron overload and accumulation are linked to various diseases such as PD, AD, and neurodegeneration associated with brain iron accumulation [27]. The current therapeutic approach for treating iron toxicity is the use of iron chelators such as deferoxamine, deferiprone, and deferasirox [28]. Given the wide range of adverse effects such as gastrointestinal disorders, nephrotoxicity, and agranulocytosis associated with these iron chelators [29], exogenous antioxidants such as N-acetylcysteine, flavonoids, carotenoids, and diterpenes have also been explored as potential therapeutic options [30,31]. Based on the promising antioxidant activity and favorable pharmacokinetic and safety profiles of dithiolethiones [32], we investigated their effects against iron-induced cytotoxicity in U-87 MG astrocytoma cells. While D3T and ACDT have previously been shown to be cytoprotective against various toxicants, this is the first study to test their effects against iron-induced toxicity. We chose 50 μM as the test concentration for both dithiolethiones based on data from past studies [19,33]. The color of FAC did not interfere with any of the assays used in our study, as readings from wells (with no cells) containing FAC alone did not differ from wells containing media alone. A 50% cytotoxicity with 10 mM FAC at 24 h [34] and with 5 mM FAC at 48 h [35] were previously reported in SH-SY5Y human neuroblastoma cells. The relatively high (15 mM) concentration of FAC required in our study is likely due to the cell line itself, as astrocytes can store large amounts of iron and are more resistant to toxicity compared to neurons [36]. The role of astrocytes as gatekeepers to neurons has been described previously, suggesting that protecting astrocytes could indirectly protect neurons from toxic insult [37,38].
Oxidative stress is believed to be the principal mechanism of iron-induced toxicity. Iron is a transition metal that catalyzes the formation of ROS, depletion of intracellular antioxidants, and overall cellular oxidative damage [39]. By causing the nuclear translocation of Nrf2, dithiolethiones activate ARE and initiate the transcription of various antioxidant genes [11] such as catalytic subunit of glutamate cysteine ligase GCLC. The enzyme GCLC catalyzes the ratelimiting step in GSH synthesis, a natural cellular antioxidant. The increase in Nrf2 levels typically occurs early (between Fig. 4 Impact of dithiolethiones on erastin-induced ferroptosis. The impact of a 24 h pretreatment with D3T (50 µM), ACDT (50 µM), or ferrostatin-1 (10 µM) prior to erastin (20 µM, 24 h) exposure on a xCT protein expression, b total GSH levels, c MDA levels, and d cell viability. Data represented as mean ± SEM (n = 3), were analyzed by one-way ANOVA using Tukey's post hoc test. **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001 vs. control group; # p ≤ 0.05, ## p ≤ 0.01, ### p ≤ 0.001, #### p ≤ 0.0001 vs erastin alone group; and $ p ≤ 0.05 vs. D3T + erastin treated group 2-6 h) and is a transient event [19,40]. Therefore, while we were able to measure the changes in Nrf2 levels induced by a 4 h treatment with dithiolethiones alone, the impact of these compounds as a 24 h pretreatment prior to a 24 h FAC exposure could not be assessed due to the extended timeline. Similar to previous results, we identified a dithiolethioneinduced increase in Nrf2 levels to explain the observed trends in GCLC and GSH levels [19,40,41]. This impact was much more prominent in cells pretreated with dithiolethiones, and exhibited downstream by the significant decrease in ROS and lipid peroxidation levels in these cells when consequently exposed to FAC. D3T and ACDT have been previously shown to increase GSH levels in neuronal cells [19,42,43]. It is noteworthy that basal GSH levels in neurons are lower than those in astrocytes and reportedly increase when neurons are co-incubated with astrocytes [44]. A dithiolethione-mediated increase in GSH levels in astrocytic cells could therefore be hypothesized to benefit adjacent neurons as well.
The labile iron pool is strictly regulated by various mechanisms including the Nrf2-mediated transcriptional upregulation of ferritin and ferroportin (FPN1) expression [45]. Both proteins reduce cellular iron levels. Ferritin is an intracellular iron storage protein that stores excess iron in a redox inactive form [46], while FPN1 is the iron exporter that exports iron from the cell into the plasma [47]. Incubation with ferric iron has previously been reported to increase ferritin expression in Caco-2 cells [48] and astrocytes [23] as a potential cellular defense mechanism against iron overload. In our study, pretreatment with D3T and ACDT further increased ferritin expression. This is in line with a previous study reporting dithiolethione-mediated activation of the Nrf2-ferritin heavy chain (Fth1)-ARE pathway induced ferritin expression beyond normal in response to oxidative stress [49]. Similar to results from Tangudu and group [50], we observed trends of an FACinduced decrease and D3T-mediated increase in FPN1 levels, but these were not statistically significant and may only have partly contributed to the overall decrease in intracellular free iron levels observed in this study. There is also a possibility that the time point chosen for FPN1 protein level analysis may have impacted our results. Nrf2 levels were found to increase with a 4 h exposure to these compounds, suggesting that an impact on FPN1 levels may have diminished by the time of testing. In line with other studies [51], we have reported measurement of labile iron with a shorter (2 h) exposure to FAC to account for the stability of the calcein-AM dye. Notably, D3T alone showed a strong trend towards increased ferritin and ferroportin levels eventually causing significant reduction in labile iron level. Even though a mild degree of anemia has previously been reported with high doses of D3T in animal studies [52], this dithiolethione is still considered relatively safe compared to the currently available iron chelators.
Recent research has expanded to other forms of ironinduced cell death such as ferroptosis. First coined in 2012, ferroptosis is an iron-dependent, non-apoptotic cell death characterized by increased lipid peroxidation and associated with reduced intracellular antioxidant capacity [53]. It is morphologically, biochemically, and genetically distinct from all other forms of cell death. Inhibition of intracellular antioxidants such as GSH and glutathione peroxidase (GPX4) by the known ferroptosis inducers erastin and RAS-selective lethal-3 (RSL3) leads to oxidative stress and lipid peroxidation which then drive ferroptosis [54]. Astrocytes have been found to be susceptible to ferroptosis induced by erastin and RSL3 [55]. Ferroptosis is inhibited by iron chelators and lipophilic antioxidants [56]. The Nrf2 transcription factor has been reported to upregulate the expression of various ferroptosis-related proteins like xCT and GPX4 thereby inhibiting ferroptosis [57]. A previous study has proposed that D3T could potentially offer protection against ferroptosis since it has the ability to activate Nrf2, but this has not been tested yet [58]. We investigated the effects of D3T and ACDT against the canonical ferroptosis inducer erastin using ferrostatin-1 as a standard comparator. Ferrostatin-1 is a lipophilic antioxidant that specifically inhibits ferroptosis by preventing lipid peroxidation without affecting any other forms of cell death [59]. Erastin causes ferroptosis by inhibiting the transmembrane cystine/ glutamate antiporter (xCT). This transporter normally allows the uptake of cystine into the cells which is then reduced to cysteine, a molecule indispensable for GSH synthesis. Therefore, inhibition of xCT leads to a decrease in cysteine levels and consequent intracellular GSH depletion, increased lipid peroxidation, and resulting ferroptosis [60]. Our results align with previous reports of erastin-induced downregulation of xCT expression and increase in lipid peroxidation in primary cortical neurons [61], as well as erastin-induced GSH depletion in human cervical adenocarcinoma cells [60] and HT-1080 human fibrosarcoma cells [62]. Moreover, an increase in the xCT expression by D3T and ACDT against erastin is one of our key findings which can at least partly be explained by the activation of the Nrf2-ARE pathway that controls the expression of this antiporter [63]. A corresponding increase in GSH levels was also observed with dithiolethione pretreatment, along with a reduction in lipid peroxidation and cell death. These effects were similar to those observed with ferrostatin-1 which is also reported to induce the Nrf2 transcription factor [64]. Even though the concentration of dithiolethiones (50 µM) used here is higher than that of ferrostatin-1 (10 µM), dithiolethiones offer the advantage of being of natural origin. Ferrostatin-1 is not available for therapeutic use. These results imply that dithiolethiones can successfully protect cells against the ferroptotic form of cell death.
A potential limitation of this study includes the lack of corresponding data on neuronal cells. As a future study, we suggest using a co-culture technique to test the cytoprotective effects of dithiolethiones on astrocytes and adjacent neurons. Further mechanistic studies to test the role of dithiolethiones against other ferroptosis inducers and in other cell lines including primary astrocytes are also warranted given the complex and evolving nature of this newly discovered form of cell death.
In conclusion, this study is the first to report that dithiolethiones D3T and ACDT exert significant cytoprotection against iron overload-induced cytotoxicity in U-87 MG cells. Using a Nrf2-dependent pathway, dithiolethiones prevent cell death by upregulating ferritin and GSH expression and reducing ROS production and lipid peroxidation. These dithiolethiones also successfully inhibit erastin-induced ferroptosis by increasing xCT expression and resulting GSH levels, alleviating lipid peroxidation, thereby protecting against ferroptotic cell death comparable to ferrostatin-1 (Fig. 5). Hence for the first time, we introduce dithiolethiones D3T and ACDT as ferroptosis inhibitors and provide compelling evidence to support their further investigation as potential candidates for the treatment of iron overload conditions.