From GST inhibitors to bithionol: dual-targeting mitochondria to overload ROS in gastric cancer

Why earlier GST inhibitors fell short

Glutathione S-transferases are overexpressed in many cancers, where they conjugate glutathione to electrophilic substrates and detoxify chemotherapeutics, contributing to drug resistance [1]. The rationale for GST inhibition as a chemosensitization strategy has been pursued for decades, with ethacrynic acid entering phase I testing in combination with thiotepa in 1991 [3]. However, clinical translation has been limited by modest efficacy, and off-target toxicity remains a concern because both glutathione and thioredoxin systems play essential roles in normal cells [5]. Ethacrynic acid inhibits purified GSTM1 similarly to bithionol but produces only a mild ROS increase in A549 and MGC-803 cells, suggesting that GST inhibition alone is insufficient to generate lethal oxidative stress [1]. This gap between target engagement and cellular consequence defines the frontier that bithionol was tested against.

The mechanistic limitation of single-target GST inhibitors is that cancer cells retain parallel antioxidant defenses. GSTM2, for example, protects malignant rhabdoid tumor of the kidney cells against ferroptosis independently of GPX4, and its knockdown sensitizes cells to erastin while reducing PRDX1 expression [2]. This demonstrates that GST isoforms can buffer oxidative stress through pathways that do not require the canonical SLC7A11/GPX4 axis [2]. A single GST inhibitor may therefore be bypassed by compensatory antioxidant mechanisms, explaining why ethacrynic acid and related compounds have not delivered robust clinical responses.

Bithionol's H-site occupancy and mitochondrial second hit

The anchor paper establishes bithionol as a multi-GST inhibitor through co-crystal structures of GSTM1 in binary complex with bithionol alone (2.50 Å resolution) and in ternary complex with both GSH and bithionol (2.15 Å resolution) [1]. Bithionol occupies the xenobiotic-binding H-site, inserting one ring into a hydrophobic pocket formed by F209, M109, H108, G12, Y7, and I10, while the other ring engages M113, Y116, and M212 through van der Waals interactions [1]. This binding mode explains why bithionol reduces Vmax rather than affecting GSH Km, as it does not occupy the GSH-binding site [1]. Derivative analysis showed that removal or methylation of the Y116-interacting hydroxyl abolishes inhibitory activity, and mutagenesis confirmed that I10V and M113S mutations reduce bithionol potency while H108R enhances it [1].

Beyond GST inhibition, bithionol independently impairs mitochondria, causing rapid loss of mitochondrial membrane potential within 1 hour, followed by progressive ROS accumulation, ATP depletion, GSH reduction, and PARP cleavage over 24 hours [1]. This temporal sequence indicates that mitochondrial dysfunction is an early event, not a secondary consequence of cell death or GST inhibition [1]. The lipophilic weak acid nature of bithionol likely contributes to its protonophore-like effects on mitochondrial membranes, though the authors note that further investigation is needed to determine whether it also affects the electron transport chain directly [1]. The dual action creates what the authors term a two-hit mechanism: direct mitochondrial ROS production combined with inhibition of GST-mediated detoxification, generating a self-amplifying ROS crisis [1].

Mixed apoptosis and ferroptosis from ROS overload

Bithionol-induced cell death involves both apoptosis and ferroptosis, as shown by Annexin V/PI flow cytometry, PARP1 cleavage, GPX4 downregulation, and elevated malondialdehyde levels [1]. The pan-caspase inhibitor Q-VD-Oph significantly attenuated bithionol-induced PARP1 cleavage and partially rescued cell viability, while the ferroptosis inhibitor ferrostatin-1 suppressed lipid peroxidation but did not affect PARP1 cleavage [1]. This mixed death phenotype distinguishes bithionol from canonical GST inhibitors, which primarily trigger apoptosis [1]. The profound GSH depletion caused by bithionol's dual action disables the GPX4 antioxidant system, creating a permissive environment for lethal lipid peroxidation [1].

The involvement of ferroptosis aligns with emerging evidence that GST isoforms can suppress this death pathway. In malignant rhabdoid tumor of the kidney, GSTM2 protects cells against ferroptosis through a mechanism independent of GPX4 and SLC7A11, instead involving PRDX1 reduction [2]. This suggests that GST-mediated ferroptosis resistance may operate through multiple parallel pathways, and that bithionol's broad inhibition of several GST isoforms could circumvent compensatory mechanisms that single-isoform targeting would miss. The anchor paper's finding that wild-type GSTM1 overexpression reverses bithionol-induced ROS accumulation, ATP depletion, and cell death, while the binding-impaired M113S mutant confers only partial rescue, confirms on-target engagement within cells [1].

Context-dependent synergy with doxorubicin

Bithionol synergizes potently with doxorubicin in MGC-803 gastric cancer cells but shows only modest synergy in A549 lung cancer cells [1]. This differential response correlates with distinct antioxidant expression profiles: MGC-803 cells express higher levels of GSTA1, SOD1, SOD2, and GPX4 but lower levels of GSTM1 and GSTP1 compared to A549 cells [1]. In MGC-803 cells, doxorubicin alone fails to generate significant ROS, but bithionol co-treatment dismantles the antioxidant shield and unleashes a supra-additive ROS surge [1]. The magnitude of ROS augmentation, rather than basal ROS levels, appears to predict synergy, nominating it as a potential biomarker for bithionol-based combinations [1].

Earlier work on bithionol combinations in ovarian cancer showed that synergy with paclitaxel depends on drug concentration and sequence of addition, with simultaneous treatment producing synergy at lower bithionol concentrations but antagonism at higher concentrations [4]. That study also found that ascorbic acid did not restore viability when cells were pretreated with bithionol followed by paclitaxel, suggesting that ROS-independent mechanisms may contribute to some combination effects [4]. The anchor paper's finding that NAC or exogenous GSH attenuates the bithionol-doxorubicin interaction supports ROS overload as the primary driver of synergy in gastric cancer cells, but the context-dependence observed across cell lines and drug combinations indicates that the redox landscape shapes therapeutic outcomes in ways that require further characterization [1].

What remains uncertain before clinical translation

The conclusions rest on cell line experiments and co-crystal structures, with no animal model efficacy or pharmacokinetic data presented [1]. The authors explicitly state that further in vivo efficacy and pharmacokinetic studies are required to assess therapeutic potential [1]. Cell culture experiments used DMEM with 25 mM glucose, a supraphysiological concentration that may influence metabolic and oxidative stress readouts [1]. Primary mouse hepatocytes and cardiomyocytes showed IC50 values of 47.3 μM and 30.1 μM, respectively, compared to 11.9 μM and 15.5 μM in A549 and MGC-803 cancer cells, suggesting a therapeutic window but also indicating that normal cells are not unaffected [1].

The broader context of GST inhibition as a therapeutic strategy includes concerns about off-target toxicity, as both glutathione and thioredoxin systems play essential roles in normal tissues [5]. The observation that GSTP1 promoter methylation varies across breast cancer subtypes and correlates with expression levels suggests that GST isoform expression is heterogeneous across cancers, which could affect bithionol sensitivity [6]. Additionally, the role of GST isoforms in chemoresistance extends beyond direct detoxification: GST-π levels are regulated by lncRNA XIST in doxorubicin-resistant colorectal cancer, and GST-π expression is modulated by REG4 in ovarian cancer through PI3K/Akt/mTOR signaling [7,8]. These findings indicate that GST expression is embedded in complex regulatory networks, and bithionol's efficacy may depend on the specific molecular context of each tumor. The anchor paper provides a mechanistic blueprint for dual-targeting GST and mitochondria, but whether this approach can overcome the redundancy and adaptability of antioxidant systems in vivo remains an open question [1].

About These Sources

This research page is built on 6 peer-reviewed studies — published from 1991 to 2026, 3 from 2024 or later, collectively cited 235 times — selected as the most relevant from 10 studies that passed quality screening, drawn from 72 papers retrieved from a database of over 500 million.

Sources used in this answer