Glyoxalase I inhibition by TLSC702 sensitizes glyoxalase I-dependent cancer cells to doxorubicin-induced cell death

Abstract

Glyoxalase I (GLO I) detoxifies the reactive glycolytic byproduct methylglyoxal (MG) and is frequently overexpressed in cancer cells. High GLO I expression is closely associated with resistance to anticancer drugs, including doxorubicin (DOX). However, the functional relationship between GLO I activity and DOX-induced cell death remains unclear. In this study, we investigated whether inhibition of GLO I enhances cellular sensitivity to DOX. Using non-small cell lung cancer cell lines with low (NCI-H460 cells) or high (NCI-H522 cells) GLO I expression, we found that NCI-H522 cells, which we previously reported to be highly sensitive to the GLO I inhibitor TLSC702 and to accumulate MG following GLO I inhibition, were more resistant to DOX than NCI-H460 cells. Cotreatment with exogenous MG significantly potentiated DOX-induced cytotoxicity through multiple cell death pathways, supporting a role for MG-related processes in modulating DOX responses. While simultaneous treatment with TLSC702 and DOX showed limited enhancement, pretreatment with TLSC702 followed by DOX markedly increased cell death in GLO I-high cells, suggesting that prior disruption of MG detoxification may sensitize cells to subsequent DOX exposure. This effect was accompanied by enhanced poly(ADP-ribose) polymerase (PARP) cleavage and was partially suppressed by caspase and autophagy inhibitors, indicating enhanced apoptosis with contributions from additional cell death pathways. Collectively, these results demonstrate that GLO I inhibition primes cancer cells for enhanced DOX-induced apoptosis, providing mechanistic insights into GLO I-associated chemoresistance.

INTRODUCTION

Cancer cells must efficiently secure large amounts of energy as well as biosynthetic materials, such as nucleic acids and lipids, to sustain rapid proliferation. To meet these demands, cancer cells preferentially utilize glycolysis even under oxygen-rich conditions, a phenomenon known as the Warburg effect (Warburg, 1956; Vander Heiden et al., 2009). Although this abnormal upregulation of glycolysis provides cancer cells with growth and survival advantages, it also imposes severe metabolic risks. One such risk is the accumulation of methylglyoxal (MG), a highly reactive α-dicarbonyl compound generated as a byproduct of glycolysis. MG readily reacts with amino groups in proteins as well as with DNA and RNA, leading to the formation of irreversible glycation products known as advanced glycation end-products. The resulting dicarbonyl stress can lead to protein dysfunction, genomic instability, and cell death, including apoptosis. Therefore, strict regulation of intracellular MG levels is critical for cancer cell survival (Kang et al., 1996).

Cells possess a defense system that degrades and detoxifies MG. Glyoxalase I (GLO I) is the rate-limiting enzyme responsible for this process. GLO I converts MG to S-D-lactoylglutathione through a glutathione (GSH)-dependent hemithioacetal intermediate, which is subsequently metabolized to D-lactate by glyoxalase II (GLO II). Thus, GLO I plays a critical role in maintaining intracellular homeostasis by preventing the accumulation of cytotoxic MG (Thornalley, 1990). Notably, numerous studies have reported that GLO I expression and activity are markedly elevated in a wide variety of cancers, including lung cancer (Sakamoto et al., 2001), colon cancer (Ranganathan and Tew, 1993), pancreatic cancer (Wang et al., 2012), gastric cancer (Cheng et al., 2012), prostate cancer (Burdelski et al., 2017), breast cancer (Tamori et al., 2018), melanoma (Bair et al., 2010), and anticancer drug-resistant human leukemia cells (Sakamoto et al., 2000), when compared with normal cells. Because cancer cells constitutively generate large amounts of MG through enhanced glycolysis, they are thought to evade MG-induced cytotoxicity by overexpressing GLO I. Therefore, pharmacological inhibition of GLO I is expected to induce selective accumulation of MG in cancer cells, thereby triggering cancer-selective cell death while minimizing cytotoxic effects in normal cells. Consequently, GLO I inhibitors have attracted considerable attention as promising candidates for anticancer therapeutics that target cancer-specific metabolic vulnerabilities.

In the early stages of GLO I inhibitor development, studies focused on inhibitors based on structural analogs of the cofactor GSH such as S-p-bromobenzylglutathione cyclopentyl diester. These compounds could suppress GLO I activity and induce apoptosis by increasing intracellular MG levels (Miller et al., 2006). However, GSH analogs are readily degraded by γ-glutamyltranspeptidase, which is widely expressed in the body, and carry a high risk of off-target inhibition of other GSH-dependent enzymes. In addition, their peptide-like features limit membrane permeability, complicating clinical translation (Vince et al., 1999; Jin et al., 2020). These limitations suggest that alternative, non-GSH-based scaffolds may offer advantages for the development of small-molecule GLO I inhibitors. Accordingly, exploration of novel non-GSH-based scaffolds for GLO I inhibition could open new avenues for anticancer drug discovery. From this perspective, our laboratory conducted a search for low-molecular-weight compounds targeting the MG-binding site within the active center of GLO I, aiming to identify molecules more suitable for drug development. We discovered TLSC702, a novel small-molecule inhibitor of human GLO I (Takasawa et al., 2011). TLSC702 more strongly inhibits the proliferation of human lung cancer NCI-H522 cells, which show higher GLO I expression than low GLO I–expressing NCI-H460 cells (Takasawa et al., 2016). In TLSC702-treated NCI-H522 cells, we observed accumulation of argpyrimidine adducts, a representative MG-derived advanced glycation end-product (Takasawa et al., 2016). Crystallographic analysis of the TLSC702–human GLO I complex defined the coordination of TLSC702 to the catalytic Zn²⁺ in the MG-binding site of human GLO I (Usami et al., 2022). Therefore, TLSC702 was employed as a pharmacological tool to investigate the biological consequences of GLO I inhibition in the present study.

A clinically important aspect of GLO I dysregulation is that high GLO I expression is closely associated with resistance to anticancer drugs, including doxorubicin (DOX) and etoposide (Sakamoto et al., 2000). In particular, DOX, an anthracycline anticancer agent, is a key chemotherapeutic drug with potent efficacy against a wide range of cancers, including bladder, breast, gastric, lung, ovarian, and thyroid cancers, as well as soft tissue sarcomas, multiple myeloma, and Hodgkin lymphoma (Rivankar, 2014). However, the greatest limitation in the clinical use of DOX is its dose-dependent and irreversible cardiotoxicity (Swain et al., 2003). Cardiomyocytes are rich in mitochondria and prone to generating reactive oxygen species (ROS), yet their defense mechanisms against oxidative and dicarbonyl stress are relatively weak compared with those of other tissues (Peng et al., 2022). Consequently, increasing the DOX dose to achieve sufficient therapeutic efficacy markedly elevates the risk of severe heart failure, necessitating strict limitations on cumulative lifetime dosage. If coadministration of a GLO I inhibitor could enhance the sensitivity of cancer cells to DOX, it might become possible to achieve adequate therapeutic effects with lower DOX doses. Such a strategy may offer a means to improve the therapeutic efficacy of DOX while potentially reducing its dose-limiting toxicities, particularly cardiotoxicity. However, it remains unclear whether inhibition of GLO I can sensitize cancer cells to DOX and which cell death pathways are involved in this process.

Based on this background, we hypothesized that, in cancer cells with high GLO I expression, treatment with the GLO I inhibitor TLSC702 would impair GLO I-mediated protection against MG-associated stress and thereby enhance cellular sensitivity to DOX-induced cytotoxic stress. If effective, this strategy could enable the use of lower DOX doses while maintaining therapeutic efficacy and reducing treatment-associated toxicity. Therefore, in the present study, we aimed to evaluate the effects of TLSC702 on DOX-induced growth inhibition and cell death in cancer cells with high GLO I expression. In addition, we sought to characterize the mode of cell death induced by this treatment and to investigate the underlying molecular mechanisms.

MATERIALS AND METHODS

Materials

TLSC702 and MG were purchased from Namiki Shoji Co., Ltd. (Tokyo, Japan). DOX and an HRP-conjugated anti-GAPDH antibody were obtained from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). Z-VAD-FMK (Z-VAD) was purchased from Peptide Institute, Inc. (Osaka, Japan). Ferrostatin-1 (Fer-1) and 3-Methyladenine (3-MA) were purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). IM-54 (IM) was purchased from Cayman Chemical (Ann Arbor, MI, USA). Necrostatin-1 (Nec-1) was purchased from Santa Cruz Biotechnology (Dallas, TX, USA). Necrostatin-1 Inactive Control (Nec-1i) was purchased from Merck Millipore (Darmstadt, Germany). The primary antibody against PARP was obtained from Proteintech Group, Inc. (Rosemont, IL, USA). An HRP-conjugated anti-rabbit IgG secondary antibody was purchased from Cell Signaling Technology (Danvers, MA, USA).

Cell culture

The human non-small cell lung cancer cell lines NCI-H522 and NCI-H460 were maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 µg/mL streptomycin. Cells were cultured at 37°C in a humidified atmosphere containing 5% CO₂.

Measurements of cell viability

Cells were seeded in 96-well plates at densities of 1 × 10⁴ and 5 × 10³ cells/well for experiments with treatment durations of 24 and 48 hr, respectively, and cultured for 24 hr before drug treatment. To compare sensitivity to DOX between NCI-H522 and NCI-H460 cells, the cells were treated with 0, 0.003, 0.01, 0.03, or 0.1 µM DOX for 48 hr. To evaluate the combined effect of DOX and MG, NCI-H522 cells were treated with 0, 0.125, 0.25, 0.5, 1, or 2 µM DOX in the presence or absence of 400 µM MG for 24 hr. Cell viability was assessed using the Cell Counting Kit-8 assay according to the manufacturer's instructions (Dojindo Laboratories, Kumamoto, Japan). The absorbance of WST-8-derived formazan was measured at 450 nm using a SpectraMax microplate reader (Molecular Devices, San Jose, CA, USA).

Cell counting using trypan blue staining

Cells were seeded in 12-well plates at densities of 5 × 10⁴, 2.5 × 10⁴, and 1.25 × 10⁴ cells/well for experiments with treatment durations of 24, 48, and 72 hr, respectively, and cultured for 24 hr before drug treatment. In experiments involving simultaneous cotreatment with TLSC702 and DOX, NCI-H522 cells were incubated with 0, 150, or 200 µM TLSC702 and 0, 0.01, or 0.03 µM DOX for 48 or 72 hr. In experiments involving simultaneous MG–DOX cotreatment, NCI-H522 cells were incubated with 0 or 400 µM MG and 0, 1, or 2 µM DOX for 24 hr. To examine the types of cell death induced by MG–DOX cotreatment, NCI-H522 cells were pretreated for 1 hr with an apoptosis inhibitor (Z-VAD), a ferroptosis inhibitor (Fer-1), an oxidative stress-induced necrosis inhibitor (IM), a necroptosis inhibitor (Nec-1), or an inactive analog of Nec-1 (Nec-1i), or were pretreated for 3 hr with an autophagy inhibitor (3-MA), and were subsequently cotreated with MG and DOX for 24 hr. In experiments involving TLSC702 pretreatment followed by DOX exposure, cells were pretreated with 0 or 400 µM TLSC702 for 24 hr, after which 0 or 2 µM DOX was added and cotreatment with TLSC702 and DOX was continued for an additional 24 hr. To examine the types of cell death induced under these conditions, cells were pretreated for 24 hr with TLSC702 together with one of the following inhibitors: Z-VAD, Fer-1, IM-54, Nec-1, Nec-1i, or 3-MA; DOX was then added, and incubation was continued for an additional 24 hr in the presence of TLSC702 and the respective inhibitor. After drug treatment, the cells were detached with 0.1% trypsin and collected. The resulting cell suspension was mixed with 0.4% trypan blue (Gibco, New York, USA) at a 1:1 ratio, loaded into a Countess cell counting chamber slide (Invitrogen, Thermo Fisher Scientific, Massachusetts, USA). The numbers of viable (trypan blue-negative) and nonviable (trypan blue-positive) cells were counted using a Countess II FL automated cell counter (Invitrogen, Thermo Fisher Scientific, Massachusetts, USA).

Evaluation of drug combinatorial effects

The combinatorial effects of MG and DOX or TLSC702 and DOX were evaluated using the EOBA model (Borisy et al., 2003; Rotin et al., 2016). The model is defined by the following formula:

EOBA = C − [A + B − (A × B)],

where C is the fractional inhibition of both drugs simultaneously, A is the fractional inhibition of drug A, B is the fractional inhibition of drug B, and fractional inhibition was calculated as 1.0 minus cell viability (expressed as a value from 0.0 to 1.0). Positive (> 0.1), negative (< −0.1), and near-zero (−0.1 to 0.1) EOBA values reflect synergistic, antagonistic, and additive drug combinations, respectively.

Western blotting

Cells were seeded in 6-well plates at densities of 6 × 10⁵ and 3 × 10⁵ cells/well for experiments with treatment durations of 24 and 48 hr, respectively, and cultured for 24 hr before treatment. In experiments with simultaneous MG–DOX cotreatment, NCI-H522 cells were incubated with 0 or 400 µM MG and 0, 1, or 2 µM DOX for 24 hr. In experiments involving TLSC702 pretreatment and combined DOX treatment, NCI-H522 cells were pretreated with 0 or 400 µM TLSC702 for 24 hr, followed by the addition of 0 or 2 µM DOX and incubation for an additional 24 hr. Cells treated as indicated were homogenized in Cell Lysis Buffer (Cell Signaling Technology, Danvers, MA, USA). Total cellular proteins were separated by SDS-PAGE and transferred to nitrocellulose membranes. Membranes were blocked with 5% skim milk in Tris-buffered saline containing 0.1% Tween 20 (TBST) for 1 hr at room temperature and then incubated overnight at 4°C with a primary antibody against PARP or an HRP-conjugated anti-GAPDH antibody. After washing with TBST, membranes probed with the anti-PARP antibody were incubated with an HRP-conjugated anti-rabbit IgG secondary antibody for 1 hr at room temperature. Immunoreactive bands were visualized using ImmunoStar LD (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) and imaged using an iBright Western Blot Imaging System (Thermo Fisher Scientific K.K., Tokyo, Japan). GAPDH was used as a loading control.

Statistical analysis

Data are presented as the mean ± standard deviation (SD) of three independent experiments. Statistical analyses were performed using Microsoft Excel or EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a graphical user interface for R (R Foundation for Statistical Computing, Vienna, Austria). EZR is a modified version of R Commander designed to add statistical functions frequently used in biostatistics (Kanda, 2013). Differences between two groups were evaluated using Student’s t-test, and differences among more than two groups were evaluated using one-way ANOVA followed by Tukey’s post hoc test. A p value <0.05 was considered statistically significant.

RESULTS

Effects of DOX monotherapy on the survival of cancer cell lines with high or low GLO I expression

We previously reported that TLSC702 more strongly inhibited the proliferation of human lung cancer NCI-H522 cells, which exhibit high GLO I expression, than that of NCI-H460 cells with low GLO I expression (Takasawa et al., 2016). In this study, we evaluated the sensitivity of NCI-H522 and NCI-H460 cells to DOX using a WST assay. In contrast to the effect of TLSC702, treatment with DOX for 48 hr more strongly reduced the viability of NCI-H460 cells than that of NCI-H522 cells (Fig. 1). These findings indicate that NCI-H522 cells, which are more sensitive to the GLO I inhibitor TLSC702, were more resistant to DOX than NCI-H460 cells.

Fig. 1

The dose-dependent effect of DOX on cell viability in NCI-H522 and NCI-H460 cells. NCI-H522 and NCI-H460 cells were treated with 0, 0.003, 0.01, 0.03, or 0.1 µM DOX for 48 hr. Cell viability (% of control) was measured using a WST assay. Results are the averages of triplicate experiments (n = 3), and bars show standard deviations. Statistical significance between NCI-H522 and NCI-H460 cells at each DOX concentration was evaluated using Student's t-test. Asterisks (* and ***) indicate p values smaller than 0.05 (p < 0.05) and 0.001 (p < 0.001), respectively.

Evaluation of cell death induction by combined treatment with TLSC702 and DOX

To examine the effect of TLSC702 and DOX cotreatment on cell death induction in NCI-H522 cells, cell numbers were assessed by trypan blue staining. The percentage of dead cells and the number of viable cells were calculated. In addition, excess-over-Bliss additivism (EOBA) values were determined to evaluate potential synergistic effects of the combination treatment. No synergistic effect was detected after either 48 (Fig. 2A and B) or 72 hr (Fig. 2C and D) of cotreatment. At both time points, DOX induced cell death, as evidenced by an increased percentage of dead cells and a decreased number of viable cells in a concentration-dependent manner. However, cotreatment with TLSC702 did not further increase the dead cell population compared with DOX treatment alone. Although TLSC702 reduced the number of viable cells, this effect was not accompanied by a dose-dependent increase in cell death. These findings suggest that TLSC702 primarily suppresses cell proliferation rather than inducing cell death and does not enhance DOX-induced cytotoxicity under the present experimental conditions.

Fig. 2

Effects of TLSC702–DOX cotreatment on cell death and viability in NCI-H522 cells. NCI-H522 cells were treated with combinations of TLSC702 (0, 150, or 200 µM) and DOX (0, 0.01, or 0.03 µM) for 48 hr (A, B) or 72 hr (C, D). Cell death and viable cell numbers were assessed by trypan blue exclusion and cell counting. Results are the averages of triplicate experiments (n = 3), and bars show standard deviations. EOBA values were calculated from the same datasets and are shown in the in-panel tables.

Effects of MG and DOX on proliferation and cell death induction in cultured cancer cells

To investigate the effect of combined treatment with MG and DOX on cell viability in NCI-H522 cells, cells were treated with MG and DOX for 24 hr and cell viability was evaluated using a WST assay. EOBA values were then calculated from the measured cell viability data to assess potential synergistic effects of the combination treatment. Compared with treatment with DOX alone, cotreatment with MG significantly enhanced the reduction in cell viability. Calculated EOBA values revealed that the combination of 400 µM MG with 0.25–2 µM DOX exerted synergistic effects in reducing cell viability (Fig. 3A). Next, to determine whether increased cell death contributed to the reduction in cell viability induced by MG–DOX cotreatment, cells were treated for 24 hr and cell numbers were counted using the trypan blue exclusion assay. The percentage of dead cells (Fig. 3B) and the number of viable cells (Fig. 3C) were calculated. The percentage of dead cells was significantly increased by the combination treatment compared with single-agent treatments. EOBA values calculated from both the percentage of dead cells and the number of viable cells demonstrated synergistic enhancement of cell death induction and reduction of viable cell number.

Fig. 3

MG–DOX cotreatment synergistically increases cell death and reduces the number of viable cells in NCI-H522 cells. (A) NCI-H522 cells were treated with 0, 0.125, 0.25, 0.5, 1, or 2 µM DOX with or without 400 µM MG for 24 hr. Cell viability (% of control) was measured using the WST assay. Results are the averages of triplicate experiments (n = 3), and bars show standard deviations. Statistical significance between the DOX and DOX + MG groups at each DOX concentration was evaluated using Student's t-test. Asterisks (* and ***) indicate p values smaller than 0.05 (p < 0.05) and 0.001 (p < 0.001), respectively. (B), (C) NCI-H522 cells were treated with combinations of MG (0 or 400 µM) and DOX (0, 1, or 2 µM) for 24 hr. Cell death and viable cell numbers were assessed by trypan blue exclusion and cell counting. Results are the averages of triplicate experiments (n = 3), and bars show standard deviations. Statistical significance between the DOX and DOX + MG groups at each DOX concentration was evaluated using Student's t-test. Asterisks (* and **) indicate p values smaller than 0.05 (p < 0.05) and 0.01 (p < 0.01), respectively. EOBA values were calculated from the same datasets and are shown in the in-panel tables.

Role of apoptosis in MG–DOX cotreatment-induced cell death

To examine the involvement of apoptosis in MG–DOX cotreatment-induced death in NCI-H522 cells, cleavage of PARP, a biochemical marker of apoptosis, was evaluated by western blotting. The results showed no significant difference in PARP cleavage between the single-agent treatments and the cotreatment for 24 hr (Fig. 4A and B). Next, to further assess the contribution of apoptosis to MG–DOX cotreatment-induced cell death, we examined whether the pan-caspase inhibitor Z-VAD could suppress MG–DOX-induced cell death. Cell numbers were quantified using the trypan blue exclusion assay, and the percentage of dead cells (Fig. 4C) and the number of viable cells (Fig. 4D) were calculated. Cell counting revealed that treatment with Z-VAD significantly attenuated MG–DOX cotreatment-induced cell death, as evidenced by a decreased percentage of dead cells and an increased number of viable cells compared with the MG–DOX cotreatment group. These results indicate that MG–DOX-induced cell death is at least partially dependent on caspase activity. However, the protective effect of Z-VAD was incomplete, and cell death was not completely abolished. Taken together, these results suggest that apoptosis contributes to MG–DOX cotreatment-induced cell death, although additional non-apoptotic cell death pathways are also involved.

Fig. 4

Role of apoptosis in MG–DOX cotreatment-induced cell death. (A), (B) NCI-H522 cells were treated with combinations of MG (0 or 400 µM) and DOX (0, 1, or 2 µM) for 24 hr. Cleavage of PARP in NCI-H522 cells was analyzed using western blotting. Results are the averages of triplicate experiments (n = 3), and bars show standard deviations. (C), (D) NCI-H522 cells were pretreated for 1 hr with 40 µM Z-VAD or vehicle. Subsequently, the cells were treated for 24 hr with MG (0 or 400 µM), DOX (0 or 2 µM), and Z-VAD (0 or 20 µM). Cell death and viable cell numbers were assessed by trypan blue exclusion and cell counting. Results are the averages of triplicate experiments (n = 3), and bars show standard deviations. Statistical significance between cells treated in the absence and presence of Z-VAD under each treatment condition was evaluated using Student's t-test. Asterisks (*) indicate p values smaller than 0.05 (p < 0.05).

Role of ferroptosis, necrosis, necroptosis, and autophagy in MG–DOX cotreatment-induced cell death

Because Z-VAD only partially suppressed MG–DOX cotreatment-induced cell death, we next investigated which non-apoptotic cell death pathways contribute to this process. To this end, inhibitors of ferroptosis, necrosis, necroptosis, and autophagy were used to evaluate their involvement in MG–DOX cotreatment-induced cell death in NCI-H522 cells.

To investigate the involvement of ferroptosis, necrosis, necroptosis, and autophagy in MG–DOX cotreatment-induced cell death in NCI-H522 cells, we examined whether the corresponding inhibitors could suppress cell death. Cells were pretreated with Fer-1 (a ferroptosis inhibitor), IM (an inhibitor of oxidative stress-induced necrosis), Nec-1 (a necroptosis inhibitor), Nec-1i (an inactive analog of Nec-1), or 3-MA (an autophagy inhibitor), followed by 24-hr treatment with MG and DOX. Cell numbers were quantified using the trypan blue exclusion assay, and the percentage of dead cells and the number of viable cells were calculated (Fig. 5A–J). Treatment with Fer-1 did not significantly alter either the percentage of dead cells or the number of viable cells in response to MG–DOX cotreatment (Fig. 5A and B), suggesting that ferroptosis is unlikely to contribute substantially to MG–DOX cotreatment-induced cell death. In contrast, treatment with IM significantly reduced the percentage of dead cells induced by MG–DOX cotreatment (Fig. 5C and D), indicating the involvement of necrotic cell death. To investigate the contribution of necroptosis, both Nec-1 and its inactive analog Nec-1i were examined. Although treatment with Nec-1 significantly reduced the percentage of dead cells induced by MG–DOX cotreatment, a similar reduction was observed with Nec-1i (Fig. 5E–H). Because Nec-1i lacks RIPK1 inhibitory activity, these findings suggest that the observed effect is unlikely to be attributable to specific inhibition of RIPK1-dependent necroptosis. Treatment with 3-MA significantly reduced the percentage of dead cells and increased the number of viable cells in MG–DOX-cotreated cells (Fig. 5I and J), indicating that autophagy contributes to MG–DOX cotreatment-induced cell death. However, none of the inhibitors completely abolished cell death. Taken together, these findings suggest that necrosis and autophagy contribute, at least in part, to MG–DOX cotreatment-induced cell death, whereas ferroptosis and RIPK1-dependent necroptosis are unlikely to play major roles.

Fig. 5

Role of ferroptosis, necrosis, necroptosis, and autophagy in MG–DOX cotreatment-induced cell death. (A, B) NCI-H522 cells were pretreated for 1 hr with 4 µM Fer-1 or vehicle. Subsequently, the cells were treated for 24 hr with MG (0 or 400 µM), DOX (0 or 2 µM), and Fer-1 (0 or 2 µM). (C, D) NCI-H522 cells were pretreated for 1 hr with 20 µM IM or vehicle and then treated for 24 hr with MG (0 or 400 µM), DOX (0 or 2 µM), and IM (0 or 10 µM). (E, F) NCI-H522 cells were pretreated for 1 hr with 20 µM Nec-1 or vehicle and then treated for 24 hr with MG (0 or 400 µM), DOX (0 or 2 µM), and Nec-1 (0 or 10 µM). (G, H) NCI-H522 cells were pretreated for 1 hr with 20 µM Nec-1i or vehicle and then treated for 24 hr with MG (0 or 400 µM), DOX (0 or 2 µM), and Nec-1i (0 or 10 µM). (I, J) NCI-H522 cells were pretreated for 3 hr with 10 mM 3-MA or vehicle and then treated for 24 hr with MG (0 or 400 µM), DOX (0 or 2 µM), and 3-MA (0 or 5 mM). Cell death and viable cell numbers were assessed by trypan blue exclusion and cell counting. Results are the averages of triplicate experiments (n = 3), and bars show standard deviations. Statistical significance between cells treated in the absence and presence of the indicated inhibitor under each treatment condition was evaluated using Student's t-test. Asterisks (* and **) indicate p values smaller than 0.05 (p < 0.05) and 0.01 (p < 0.01), respectively.

Evaluation of cell death induced by TLSC702 pretreatment followed by DOX cotreatment

Because MG cotreatment enhanced DOX-induced cell death whereas simultaneous treatment with TLSC702 and DOX did not, we next examined whether pre-exposure to TLSC702 could sensitize NCI-H522 cells to DOX-induced cell death. In our previous study, treatment of NCI-H522 cells with 300 μM TLSC702 for 24 hr induced accumulation of argpyrimidine adducts, a representative MG-derived advanced glycation end-product (Takasawa et al., 2016). Because lower concentrations and shorter treatment periods were not evaluated in that study, cells were pretreated with 400 μM TLSC702 for 24 hr in the present study to promote intracellular MG accumulation, which may be reflected by the accumulation of argpyrimidine adducts, before DOX exposure. TLSC702 pretreatment significantly enhanced DOX-induced cell death, as evidenced by an increase in the percentage of dead cells and a decrease in the number of viable cells compared with single-agent treatments (Fig. 6A and B).

Fig. 6

TLSC702 pretreatment followed by cotreatment with DOX synergistically increases cell death and reduces the number of viable cells in NCI-H522 cells. (A), (B) NCI-H522 cells pretreated with TLSC702 for 24 hr were cotreated with DOX for an additional 24 hr. Cell death and viable cell numbers were assessed by trypan blue exclusion and cell counting. Results are the averages of triplicate experiments (n = 3), and bars show standard deviations. Statistical significance was evaluated using one-way ANOVA followed by Tukey's post hoc test. Asterisks (* and ***) indicate p values smaller than 0.05 (p < 0.05) and 0.001 (p < 0.001), respectively. Hash symbols (###) indicate p values smaller than 0.001 (p < 0.001) versus the control group. EOBA values were calculated from the same datasets and are shown in the in-panel tables.

Role of apoptosis in TLSC702 pretreatment + DOX cotreatment-induced cell death

To examine the involvement of apoptosis in TLSC702 pretreatment + DOX cotreatment-induced cell death in NCI-H522 cells, cleavage of PARP was evaluated by western blotting. The results showed that PARP cleavage was significantly increased in the TLSC702-pretreatment + DOX cotreatment group compared with the single-agent treatment groups (Fig. 7A and B). Next, to further assess the contribution of apoptosis to TLSC702 pretreatment + DOX cotreatment-induced cell death, we examined whether the pan-caspase inhibitor Z-VAD could suppress cell death. Cell numbers were quantified using the trypan blue exclusion assay, and the percentage of dead cells (Fig. 7C) and the number of viable cells (Fig. 7D) were calculated. Cell counting revealed that treatment with Z-VAD significantly attenuated TLSC702 pretreatment + DOX cotreatment-induced cell death, as evidenced by a decreased percentage of dead cells and an increased number of viable cells compared with the TLSC702-pretreatment + DOX cotreatment group. However, the protective effect of Z-VAD was incomplete, and cell death was not completely abolished. Taken together, these results suggest that apoptosis contributes to TLSC702 pretreatment + DOX cotreatment-induced cell death, although additional non-apoptotic cell death pathways are also involved.

Fig. 7

Role of apoptosis in TLSC702 pretreatment + DOX cotreatment-induced cell death. (A), (B) NCI-H522 cells pretreated with TLSC702 for 24 hr were cotreated with DOX for an additional 24 hr. Cleavage of PARP in NCI-H522 cells was analyzed using western blotting. Results are the averages of triplicate experiments (n = 3), and bars show standard deviations. Statistical significance was evaluated using one-way ANOVA followed by Tukey's post hoc test, as appropriate. Asterisks (* and ***) indicate p values smaller than 0.05 (p < 0.05) and 0.001 (p < 0.001), respectively. Hash symbols (# and ###) indicate p values smaller than 0.05 (p < 0.05) and 0.001 (p < 0.001), respectively, versus the control group. (C), (D) NCI-H522 cells were pretreated with 0 or 20 µM Z-VAD and 0 or 400 µM TLSC702 for 24 hr and subsequently cotreated with 0 or 2 µM DOX for an additional 24 hr. Cell death and viable cell numbers were assessed by trypan blue exclusion and cell counting. Results are the averages of triplicate experiments (n = 3), and bars show standard deviations. Statistical significance between cells treated in the absence and presence of Z-VAD under each treatment condition was evaluated using Student's t-test. Asterisks (* and ***) indicate p values smaller than 0.05 (p < 0.05) and 0.001 (p < 0.001), respectively.

Role of ferroptosis, necrosis, necroptosis, and autophagy in TLSC702 pretreatment + DOX cotreatment-induced cell death

To investigate the involvement of ferroptosis, necrosis, necroptosis, and autophagy in TLSC702 pretreatment + DOX cotreatment-induced cell death in NCI-H522 cells, we examined whether the corresponding inhibitors could suppress cell death. Cells were pretreated with TLSC702 in the presence or absence of Fer-1, IM, Nec-1, Nec-1i, or 3-MA for 24 hr, followed by an additional 24-hr treatment with DOX. Cell numbers were quantified using the trypan blue exclusion assay, and the percentage of dead cells and the number of viable cells were calculated (Fig. 8A–J). Treatment with Fer-1 did not significantly alter either the percentage of dead cells or the number of viable cells in response to TLSC702 pretreatment + DOX cotreatment (Fig. 8A and B), suggesting that ferroptosis is unlikely to contribute substantially to this form of cell death. Likewise, neither Nec-1 nor Nec-1i significantly affected TLSC702 pretreatment + DOX cotreatment-induced changes in dead cell percentage or viable cell number (Fig. 8E–H), suggesting that RIPK1-dependent necroptosis is unlikely to play a major role. Treatment with IM also failed to suppress TLSC702 pretreatment + DOX cotreatment-induced cell death (Fig. 8C and D). Notably, IM significantly increased DOX-induced cell death under the present experimental conditions. In contrast, treatment with 3-MA significantly reduced the percentage of dead cells induced by TLSC702 pretreatment + DOX cotreatment and increased the number of viable cells (Fig. 8I and J), although cell death was not completely abolished. Taken together, these findings suggest that autophagy contributes, at least in part, to TLSC702 pretreatment + DOX cotreatment-induced cell death, whereas ferroptosis and RIPK1-dependent necroptosis are unlikely to play major roles.

Fig. 8

Role of ferroptosis, necrosis, necroptosis, and autophagy in TLSC702 pretreatment + DOX cotreatment-induced cell death. (A, B) NCI-H522 cells were pretreated for 24 hr with TLSC702 (0 or 400 µM) in the presence or absence of 2 µM Fer-1, followed by an additional 24-hr treatment with DOX (0 or 2 µM). (C, D) NCI-H522 cells were pretreated for 24 hr with TLSC702 (0 or 400 µM) in the presence or absence of 10 µM IM, followed by an additional 24-hr treatment with DOX (0 or 2 µM). (E, F) NCI-H522 cells were pretreated for 24 hr with TLSC702 (0 or 400 µM) in the presence or absence of 10 µM Nec-1, followed by an additional 24-hr treatment with DOX (0 or 2 µM). (G, H) NCI-H522 cells were pretreated for 24 hr with TLSC702 (0 or 400 µM) in the presence or absence of 10 µM Nec-1i, followed by an additional 24-hr treatment with DOX (0 or 2 µM). (I, J) NCI-H522 cells were pretreated for 24 hr with TLSC702 (0 or 400 µM) in the presence or absence of 5 mM 3-MA, followed by an additional 24-hr treatment with DOX (0 or 2 µM). Cell death and viable cell numbers were assessed by trypan blue exclusion and cell counting. Results are the averages of triplicate experiments (n = 3), and bars show standard deviations. Statistical significance between cells treated in the absence and presence of the indicated inhibitor under each treatment condition was evaluated using Student's t-test. Asterisks (*, **, and ***) indicate statistically significant differences at p < 0.05, p < 0.01, and p < 0.001, respectively.

DISCUSSION

We first examined the sensitivity of a GLO I-high-expressing cell line (NCI-H522) and a GLO I-low-expressing cell line (NCI-H460) to DOX-induced cell death. The results suggested that NCI-H522 cells, which are more sensitive to the GLO I inhibitor TLSC702 than NCI-H460 cells, exhibited greater resistance to DOX than NCI-H460 cells. Based on these findings, we analyzed the cell-death–inducing effects of TLSC702–DOX cotreatment in NCI-H522 cells, which show high dependency on GLO I. When cells were cotreated with TLSC702 and DOX for 48 hr, no synergistic effects were observed in either the decrease in the number of viable cells or the increase in the percentage of dead cells. In addition, the percentage of dead cells did not increase in a TLSC702 dose-dependent manner, and only a reduction in viable cell number was observed. These results suggest that 48-hr treatment with TLSC702 acts primarily in a cytostatic manner. Notably, knockout or overexpression of GLO I in hepatocellular carcinoma cells mainly affects the number of cells in the G2 and S phases of the cell cycle (Zhang et al., 2024). Therefore, the growth-inhibitory effect of TLSC702 observed in this study may be mediated through GLO I inhibition-driven alterations in cell-cycle regulation. Because abnormalities in cell cycle regulation often ultimately lead to DNA damage and apoptosis (Xia et al., 2019), we hypothesized that longer exposure to TLSC702 might induce cell death. Accordingly, we extended the treatment duration from 48 to 72 hr. However, even with 72-hr treatment, no synergistic effects of the combination were observed in either the number of viable cells or the percentage of dead cells, similar to the results at 48 hr. Collectively, these results indicate that, at the concentrations tested, TLSC702 primarily induces growth inhibition rather than cell death and does not enhance DOX-induced cytotoxicity when administered simultaneously.

Previous reports have demonstrated a positive correlation between high GLO I expression and multidrug resistance to anticancer agents, including DOX (Alhujaily et al., 2021). However, it remains unclear whether GLO I–mediated detoxification of MG directly contributes to DOX resistance. A previous study showed that simultaneous treatment with MG and DOX enhanced cell death in human breast cancer cell lines (MDA-MB-231 and MCF-7) (Roy et al., 2018). Notably, the type of cell death induced by the MG–DOX combination differed between cell lines. Apoptosis was observed in MDA-MB-231 cells, whereas necroptosis was induced in MCF-7 cells. Therefore, in this study, we examined whether synergistic effects could also be observed in the human non-small cell lung cancer cell line NCI-H522 by simultaneously treating cells with MG and DOX as a model of intracellular MG accumulation.

Synergistic effects were observed in both decreased cell viability (WST assay) and decreased viable cell number together with increased percentage of dead cells (trypan blue exclusion assay). We therefore aimed to identify the type of cell death induced by MG–DOX cotreatment. First, to examine the involvement of apoptosis, which has been reported to be induced by MG accumulation, we analyzed PARP cleavage by western blotting. DOX treatment induced PARP cleavage in a dose-dependent manner. However, no synergistic enhancement of PARP cleavage was observed with the addition of MG. Furthermore, when the involvement of apoptosis was examined using the apoptosis inhibitor Z-VAD, the cell death induced synergistically by MG–DOX cotreatment was suppressed by Z-VAD, although the suppression was incomplete. Because previous studies reported that the MG–DOX-induced mode of cell death differs among cell types (Roy et al., 2018), we hypothesize that additional types of cell death also contribute to the effects observed in NCI-H522 cells. To test this possibility, we analyzed the involvement of various cell death pathways using specific inhibitors. Treatment with IM or 3-MA partially suppressed the cell death induced by MG–DOX cotreatment, although complete suppression was not achieved. These findings suggest that in addition to apoptosis, necrosis and autophagy contribute, at least in part, to the effects of MG–DOX cotreatment.

Taken together, these results demonstrate that MG enhances DOX-induced cell death. This finding suggests that to achieve synergistic effects with GLO I inhibitors, intracellular accumulation of MG must be established prior to DOX administration. Based on this hypothesis, we examined whether pretreatment with TLSC702 for 24 hr to induce MG accumulation, followed by 24 hr cotreatment with DOX, could enhance cell death. Cell counting using trypan blue staining revealed a synergistic effect of the combination on the percentage of dead cells. Next, to determine whether the mode of cell death induced by TLSC702 pretreatment + DOX cotreatment was similar to that induced by MG–DOX cotreatment, PARP cleavage was analyzed by western blotting. In contrast to MG–DOX cotreatment, TLSC702 pretreatment + DOX cotreatment showed a clear synergistic increase in PARP cleavage. One possible explanation is that the intracellular accumulation of MG through GLO I inhibition may promote apoptosis more effectively than extracellular addition of MG. Indeed, a previous study using rat aortic smooth muscle cells demonstrated that only 1.8% of total added MG (¹⁴C-labeled) was taken up into cells (Che et al., 1997), indicating that most extracellularly added MG remains in the culture medium. Moreover, extracellularly added MG readily reacts with proteins and amino acids in the culture medium (Lo et al., 1994), suggesting that MG generated intracellularly by GLO I inhibition may behave differently from exogenously added MG. However, the present study did not directly evaluate intracellular MG levels. Although accumulation of argpyrimidine adducts, a representative MG-derived advanced glycation end-product, has previously been demonstrated in TLSC702-treated NCI-H522 cells (Takasawa et al., 2016), intracellular MG levels themselves were not directly measured under the experimental conditions used in the present study. Therefore, the proposed involvement of intracellular MG accumulation in enhanced sensitivity to DOX remains to be confirmed. In addition, the studies that demonstrated limited cellular uptake of exogenously added MG and its rapid reactions with proteins and amino acids in the culture medium (Che et al., 1997; Lo et al., 1994) were performed using different cell types and experimental conditions from those used in the present study. Furthermore, to our knowledge, there are no studies directly comparing the biological effects of extracellularly applied MG with those of intracellularly generated MG. Therefore, the above explanation remains speculative and requires further investigation. Alternatively, the difference in cell death pathways observed between MG co-treatment and GLO I inhibition may reflect differences in the mode or duration of MG exposure. While exogenously added MG may be partially consumed through reactions in the culture medium before entering cells, GLO I inhibition increases intracellular MG levels by suppressing its detoxification. Consequently, the cellular effects induced by these two approaches may not be identical, which could contribute to the observed differences in cell death pathways. Further studies are needed to clarify these possibilities. Direct quantification of intracellular MG levels following exogenous MG treatment or GLO I inhibition may help determine whether differences in intracellular MG exposure contribute to the distinct cell death pathways observed in this study. In addition, comparative analyses of MG-derived protein modifications and cell death-related signaling pathways, as well as comparisons between the effects of intracellularly delivered MG and extracellularly added MG, may provide further insight into the mechanisms underlying these differences. Characterization of the time course of intracellular MG accumulation and MG-derived modifications following TLSC702 treatment will also be important for determining the pretreatment conditions required for enhancement of DOX sensitivity. In the present study, a 24-hr pretreatment period was used based on our previous findings. The observed enhancement of DOX-induced cell death following TLSC702 pretreatment is consistent with the possibility that GLO I inhibition and the resulting MG-related modifications contribute to increased sensitivity to DOX. However, the optimal duration of pretreatment remains to be determined. Furthermore, TLSC702 was used at a relatively high concentration (400 μM) in the present study. Therefore, the possibility of off-target effects cannot be completely excluded. Future studies employing direct measurements of intracellular MG levels, detailed time-course analyses, and additional approaches to validate target specificity will help clarify the contribution of GLO I inhibition to DOX sensitization and establish optimal treatment conditions for this strategy.

Although enhanced apoptosis was clearly observed in NCI-H522 cells subjected to TLSC702 pretreatment + DOX cotreatment, results from MG–DOX cotreatment suggested the involvement of cell death pathways other than apoptosis, such as necrosis and autophagy. Therefore, we considered that multiple death pathways might also be engaged by TLSC702 pretreatment + DOX cotreatment. To investigate this possibility, we analyzed the involvement of each type of cell death using specific inhibitors. Treatment with Z-VAD and 3-MA suppressed the synergistic cell death induced by TLSC702 pretreatment + DOX cotreatment, although the suppression was incomplete, supporting an apoptotic contribution together with autophagy. The potential contribution of autophagy showed a tendency similar to that observed with MG–DOX cotreatment. In contrast, IM did not suppress cell death under TLSC702 pretreatment + DOX cotreatment conditions. Moreover, IM significantly enhanced DOX-induced cell death under the present experimental conditions, precluding a clear evaluation of the contribution of necrosis. Taken together, these findings suggest that apoptosis and autophagy contribute substantially to TLSC702 pretreatment + DOX cotreatment-induced cell death, although additional cell death pathways may also be involved.

Apoptosis and autophagy frequently occur within the same cells, and it is well established that, in many cases, autophagy precedes apoptosis (Wirawan et al., 2010). In addition, autophagy and apoptosis induced by endoplasmic reticulum stress share common upstream signaling pathways (Ogata et al., 2006). Furthermore, autophagy or autophagy-related proteins can facilitate the induction of apoptosis (Song et al., 2017). Based on these findings, we hypothesize that cell death induced by MG–DOX cotreatment and by TLSC702 pretreatment + DOX cotreatment involves closely related biological mechanisms, particularly apoptosis and autophagy.

In conclusion, this study demonstrated that MG enhances DOX-induced cell death in cultured cancer cells with high GLO I expression. In addition, DOX-induced cell death is similarly enhanced when DOX is added to cells in which MG has accumulated through treatment with a GLO I inhibitor. Furthermore, it was revealed that apoptosis and autophagy contribute to the enhanced cell death induced by MG–DOX cotreatment and TLSC702 pretreatment + DOX cotreatment. The underlying premise of this study is the hypothesis that DOX-associated adverse effects could be reduced by lowering its dosage while enhancing cancer cell-specific cytotoxic effects. In this study, we clarified the mechanism by which GLO I inhibition overcomes DOX resistance in vitro. However, the effects of TLSC702 pretreatment on cellular sensitivity to lower concentrations of DOX were not examined in the present study. Therefore, it remains unclear whether GLO I inhibition can shift the effective concentration range or reduce the IC₅₀ of DOX. Evaluation of TLSC702-mediated sensitization across a broader range of DOX concentrations will be important for determining whether this strategy may enable dose reduction of DOX while maintaining anticancer efficacy. In future studies, we aim to investigate two key issues in vivo: whether the synergistic effects of a GLO I inhibitor and DOX can be observed and whether GLO I inhibitors can mitigate the serious side effects of DOX. It should also be noted that the safety profile of GLO I inhibitors themselves, including potential cardiac toxicity, remains to be determined. Therefore, comprehensive toxicological evaluation will be necessary to assess the overall therapeutic benefit of combining GLO I inhibitors with DOX. If both synergistic anticancer effects and attenuation of DOX-induced adverse effects can be confirmed in vivo, it is expected that this strategy could lead to the establishment of a novel combination chemotherapy for cancer treatment.

Funding

This work was supported by a Grant-in-Aid for Scientific Research (C) from the Japan Society for the Promotion of Science to Ryoko Takasawa (JSPS KAKENHI Grant Number JP23K05154). We would like to thank Dr. Kazunori Akimoto for discussion.

Conflict of interest

The authors declare that there is no conflict of interest.

Data availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

Author contributions

Conceptualization: Ryoko Takasawa

Data curation: Ryoko Takasawa, Haruka Mori

Formal analysis: Ryoko Takasawa, Haruka Mori

Funding acquisition: Ryoko Takasawa

Investigation: Haruka Mori

Methodology: Ryoko Takasawa, Haruka Mori

Project administration: Ryoko Takasawa

Supervision: Ryoko Takasawa

Validation: Ryoko Takasawa, Haruka Mori

Visualization: Ryoko Takasawa, Haruka Mori

Writing—original draft: Ryoko Takasawa, Haruka Mori

Writing—review & editing: Ryoko Takasawa, Haruka Mori

Ethical approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

REFERENCES

•  Alhujaily,  M.,  Abbas,  H.,  Xue,  M.,  de la Fuente,  A.,  Rabbani,  N. and  Thornalley,  P.J. (2021): Studies of Glyoxalase 1-linked multidrug resistance reveal glycolysis-derived reactive metabolite, methylglyoxal, is a common contributor in cancer chemotherapy targeting the spliceosome. Front. Oncol., 11, 748698.
•  Bair,  W.B. 3rd,  Cabello,  C.M.,  Uchida,  K.,  Bause,  A.S. and  Wondrak,  G.T. (2010): GLO1 overexpression in human malignant melanoma. Melanoma Res., 20, 85-96.
•  Borisy,  A.A.,  Elliott,  P.J.,  Hurst,  N.W.,  Lee,  M.S.,  Lehar,  J.,  Price,  E.R.,  Serbedzija,  G.,  Zimmermann,  G.R.,  Foley,  M.A.,  Stockwell,  B.R. and  Keith,  C.T. (2003): Systematic discovery of multicomponent therapeutics. Proc. Natl. Acad. Sci. USA, 100, 7977-7982.
•  Burdelski,  C.,  Shihada,  R.,  Hinsch,  A.,  Angerer,  A.,  Göbel,  C.,  Friedrich,  E.,  Hube-Magg,  C.,  Burdak-Rothkamm,  S.,  Kluth,  M.,  Simon,  R.,  Möller-Koop,  C.,  Sauter,  G.,  Büscheck,  F.,  Wittmer,  C.,  Clauditz,  T.S.,  Krech,  T.,  Tsourlakis,  M.C.,  Minner,  S.,  Graefen,  M.,  Schlomm,  T.,  Wilczak,  W. and  Jacobsen,  F. (2017): High-Level Glyoxalase 1 (GLO1) expression is linked to poor prognosis in prostate cancer. Prostate, 77, 1528-1538.
•  Che,  W.,  Asahi,  M.,  Takahashi,  M.,  Kaneto,  H.,  Okado,  A.,  Higashiyama,  S. and  Taniguchi,  N. (1997): Selective induction of heparin-binding epidermal growth factor-like growth factor by methylglyoxal and 3-deoxyglucosone in rat aortic smooth muscle cells. The involvement of reactive oxygen species formation and a possible implication for atherogenesis in diabetes. J. Biol. Chem., 272, 18453-18459.
•  Cheng,  W.L.,  Tsai,  M.M.,  Tsai,  C.Y.,  Huang,  Y.H.,  Chen,  C.Y.,  Chi,  H.C.,  Tseng,  Y.H.,  Chao,  I.W.,  Lin,  W.C.,  Wu,  S.M.,  Liang,  Y.,  Liao,  C.J.,  Lin,  Y.H.,  Chung,  I.H.,  Chen,  W.J.,  Lin,  P.Y.,  Wang,  C.S. and  Lin,  K.H. (2012): Glyoxalase-I is a novel prognosis factor associated with gastric cancer progression. PLoS One, 7, e34352.
•  Jin,  T.,  Zhao,  L.,  Wang,  H.P.,  Huang,  M.L.,  Yue,  Y.,  Lu,  C. and  Zheng,  Z.B. (2020): Recent advances in the discovery and development of glyoxalase I inhibitors. Bioorg. Med. Chem., 28, 115243.
•  Kanda,  Y. (2013): Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant., 48, 452-458.
•  Kang,  Y.,  Edwards,  L.G. and  Thornalley,  P.J. (1996): Effect of methylglyoxal on human leukaemia 60 cell growth: modification of DNA G1 growth arrest and induction of apoptosis. Leuk. Res., 20, 397-405.
•  Lo,  T.W.,  Westwood,  M.E.,  McLellan,  A.C.,  Selwood,  T. and  Thornalley,  P.J. (1994): Binding and modification of proteins by methylglyoxal under physiological conditions. A kinetic and mechanistic study with N alpha-acetylarginine, N alpha-acetylcysteine, and N alpha-acetyllysine, and bovine serum albumin. J. Biol. Chem., 269, 32299-32305.
•  Miller,  A.G.,  Smith,  D.G.,  Bhat,  M. and  Nagaraj,  R.H. (2006): Glyoxalase I is critical for human retinal capillary pericyte survival under hyperglycemic conditions. J. Biol. Chem., 281, 11864-11871.
•  Ogata,  M.,  Hino,  S.,  Saito,  A.,  Morikawa,  K.,  Kondo,  S.,  Kanemoto,  S.,  Murakami,  T.,  Taniguchi,  M.,  Tanii,  I.,  Yoshinaga,  K.,  Shiosaka,  S.,  Hammarback,  J.A.,  Urano,  F. and  Imaizumi,  K. (2006): Autophagy is activated for cell survival after endoplasmic reticulum stress. Mol. Cell. Biol., 26, 9220-9231.
•  Peng,  M.L.,  Fu,  Y.,  Wu,  C.W.,  Zhang,  Y.,  Ren,  H. and  Zhou,  S.S. (2022): Signaling pathways related to oxidative stress in diabetic cardiomyopathy. Front. Endocrinol. (Lausanne), 13, 907757.
•  Ranganathan,  S. and  Tew,  K.D. (1993): Analysis of glyoxalase-I from normal and tumor tissue from human colon. Biochim. Biophys. Acta, 1182, 311-316.
•  Rivankar,  S. (2014): An overview of doxorubicin formulations in cancer therapy. J. Cancer Res. Ther., 10, 853-858.
•  Rotin,  L.E.,  Gronda,  M.,  MacLean,  N.,  Hurren,  R.,  Wang,  X.,  Lin,  F.H.,  Wrana,  J.,  Datti,  A.,  Barber,  D.L.,  Minden,  M.D.,  Slassi,  M. and  Schimmer,  A.D. (2016): Ibrutinib synergizes with poly(ADP-ribose) glycohydrolase inhibitors to induce cell death in AML cells via a BTK-independent mechanism. Oncotarget, 7, 2765-2779.
•  Roy,  A.,  Sarker,  S.,  Upadhyay,  P.,  Pal,  A.,  Adhikary,  A.,  Jana,  K. and  Ray,  M. (2018): Methylglyoxal at metronomic doses sensitizes breast cancer cells to doxorubicin and cisplatin causing synergistic induction of programmed cell death and inhibition of stemness. Biochem. Pharmacol., 156, 322-339.
•  Sakamoto,  H.,  Mashima,  T.,  Kizaki,  A.,  Dan,  S.,  Hashimoto,  Y.,  Naito,  M. and  Tsuruo,  T. (2000): Glyoxalase I is involved in resistance of human leukemia cells to antitumor agent-induced apoptosis. Blood, 95, 3214-3218.
•  Sakamoto,  H.,  Mashima,  T.,  Sato,  S.,  Hashimoto,  Y.,  Yamori,  T. and  Tsuruo,  T. (2001): Selective activation of apoptosis program by S-p-bromobenzylglutathione cyclopentyl diester in glyoxalase I-overexpressing human lung cancer cells. Clin. Cancer Res., 7, 2513-2518.
•  Song,  S.,  Tan,  J.,  Miao,  Y.,  Li,  M. and  Zhang,  Q. (2017): Crosstalk of autophagy and apoptosis: involvement of the dual role of autophagy under ER stress. J. Cell. Physiol., 232, 2977-2984.
•  Swain,  S.M.,  Whaley,  F.S. and  Ewer,  M.S. (2003): Congestive heart failure in patients treated with doxorubicin: a retrospective analysis of three trials. Cancer, 97, 2869-2879.
•  Takasawa,  R.,  Tao,  A.,  Saeki,  K.,  Shionozaki,  N.,  Tanaka,  R.,  Uchiro,  H.,  Takahashi,  S.,  Yoshimori,  A. and  Tanuma,  S. (2011): Discovery of a new type inhibitor of human glyoxalase I by myricetin-based 4-point pharmacophore. Bioorg. Med. Chem. Lett., 21, 4337-4342.
•  Takasawa,  R.,  Shimada,  N.,  Uchiro,  H.,  Takahashi,  S.,  Yoshimori,  A. and  Tanuma,  S. (2016): TLSC702, a novel inhibitor of human glyoxalase I, induces apoptosis in tumor cells. Biol. Pharm. Bull., 39, 869-873.
•  Tamori,  S.,  Nozaki,  Y.,  Motomura,  H.,  Nakane,  H.,  Katayama,  R.,  Onaga,  C.,  Kikuchi,  E.,  Shimada,  N.,  Suzuki,  Y.,  Noike,  M.,  Hara,  Y.,  Sato,  K.,  Sato,  T.,  Yamamoto,  K.,  Hanawa,  T.,  Imai,  M.,  Abe,  R.,  Yoshimori,  A.,  Takasawa,  R.,  Tanuma,  S.I. and  Akimoto,  K. (2018): Glyoxalase 1 gene is highly expressed in basal-like human breast cancers and contributes to survival of ALDH1-positive breast cancer stem cells. Oncotarget, 9, 36515-36529.
•  Thornalley,  P.J. (1990): The glyoxalase system: new developments towards functional characterization of a metabolic pathway fundamental to biological life. Biochem. J., 269, 1-11.
•  Usami,  M.,  Ando,  K.,  Shibuya,  A.,  Takasawa,  R. and  Yokoyama,  H. (2022): Crystal structures of human glyoxalase I and its complex with TLSC702 reveal inhibitor binding mode and substrate preference. FEBS Lett., 596, 1458-1467.
•  Vander Heiden,  M.G.,  Cantley,  L.C. and  Thompson,  C.B. (2009): Understanding the Warburg effect: the metabolic requirements of cell proliferation. Science, 324, 1029-1033.
•  Vince,  R.,  Brownell,  J. and  Akella,  L.B. (1999): Synthesis and activity of gamma-(L-gamma-azaglutamyl)-S-(p-bromobenzyl)-L-cysteinylglycine: a metabolically stable inhibitor of glyoxalase I. Bioorg. Med. Chem. Lett., 9, 853-856.
•  Wang,  Y.,  Kuramitsu,  Y.,  Ueno,  T.,  Suzuki,  N.,  Yoshino,  S.,  Iizuka,  N.,  Akada,  J.,  Kitagawa,  T.,  Oka,  M. and  Nakamura,  K. (2012): Glyoxalase I (GLO1) is up-regulated in pancreatic cancerous tissues compared with related non-cancerous tissues. Anticancer Res., 32, 3219-3222.
•  Warburg,  O. (1956): On the origin of cancer cells. Science, 123, 309-314.
•  Wirawan,  E.,  Vande Walle,  L.,  Kersse,  K.,  Cornelis,  S.,  Claerhout,  S.,  Vanoverberghe,  I.,  Roelandt,  R.,  De Rycke,  R.,  Verspurten,  J.,  Declercq,  W.,  Agostinis,  P.,  Vanden Berghe,  T.,  Lippens,  S. and  Vandenabeele,  P. (2010): Caspase-mediated cleavage of Beclin-1 inactivates Beclin-1-induced autophagy and enhances apoptosis by promoting the release of proapoptotic factors from mitochondria. Cell Death Dis., 1, e18.
•  Xia,  P.,  Liu,  Y.,  Chen,  J. and  Cheng,  Z. (2019): Cell cycle proteins as key regulators of postmitotic cell death. Yale J. Biol. Med., 92, 641-650.
•  Zhang,  Y.,  Tang,  X.,  Liu,  L.,  Cai,  D.,  Gou,  S.,  Hao,  S.,  Li,  Y.,  Shen,  J.,  Chen,  Y.,  Zhao,  Y.,  Wu,  X.,  Li,  M.,  Chen,  M.,  Li,  X.,  Sun,  Y.,  Gu,  L.,  Li,  W.,  Wang,  F.,  Zhang,  Z.,  Wang,  X.,  Deng,  S.,  Xiao,  Z.,  Yao,  L. and  Du,  F. (2024): GLO1 regulates hepatocellular carcinoma proliferation and migration through the cell cycle pathway. BMC Cancer, 24, 1297.