Metabolic stress–mediated cell death and autophagy in human lung cancer cells

Abstract

Lung cancer remains one of the major causes of cancer-related mortality. Thus, newer therapeutic approaches are urgently needed. Because cancer is a metabolic disease, lung cancer cells have also rewired their metabolism to gain growth advantage and support survival. Therefore, the use of metabolic stress–inducing agents as a therapeutic strategy for lung cancer is an attractive idea. In this study, we have investigated the anticancer potential of CB-839 and metformin. CB-839, a selective glutaminase-1 inhibitor, creates glutamine-deficient conditions, and metformin is an antidiabetic drug. We report that CB-839 and metformin induce metabolic stress and inhibit growth of human lung cancer cells. Of note, lung cancer cells that harbor mutant K-Ras are more sensitive to these agents compared to cells with wild-type K-Ras status. In the K-Ras mutant cells, these agents induce cell death partly, via death receptor 5 (DR5)–dependent extrinsic pathway. However, in the lung cancer cells harboring wild-type K-Ras, these agents activate autophagy without significant effect on DR5 regulation. Pretreatment of K-Ras wild-type cells with autophagy inhibitor improves the anticancer potential of these agents coupled with activation of DR5-dependent pathway. Our results further show that the growth inhibitory effects of these agents appear to be linked to the mutant K-Ras status because pan-K-Ras inhibitor that inhibits the mutant K-Ras proteins blunted the growth inhibitory effects of these agents in cells harboring mutant K-Ras. Collectively, our results provide valuable new insights into exploiting the metabolic rewiring of lung cancer cells by using metabolic stress–inducing drugs as an important therapeutic approach.

Significance Statement

Anticancer potential of CB-839 and metformin is investigated in lung cancer. These agents induce cell death partly, via death receptor 5–dependent pathway, and a relationship with K-Ras status of lung cancer cells is noted. Lung cancer cells with mutant K-Ras are more sensitive compared to cells with wild-type K-Ras. Autophagy inhibition of K-Ras wild-type cells improves the anticancer potential. This study provides new insights into exploiting the metabolic rewiring of lung cancer cells as an important therapeutic strategy.

1. Introduction

It is now well established that generalized loss of growth control is a hallmark of cancers, predominantly due to genetic abnormalities including deletions and mutations. Inspired by the work of Otto Warburg (Warburg effect), several lines of recent evidence indicate that cancer is also a metabolic disease, as cancer cells reprogram their energy metabolism and evade immune checkpoints (Hanahan and Weinberg, 2011; Wang et al, 2018). Some of the key features of the metabolic rewiring in cancer cells are believed to include alterations in glucose and amino acid uptake, consumption of tricarboxylic acid (TCA) cycle intermediates for biosynthesis and generation of NADPH, enhanced requirement for nitrogen, and changes in metabolite-mediated regulation of genes (Pavlova and Thompson, 2016). Thus, investigations are ongoing to exploit alterations in cancer metabolism to develop newer anticancer therapies. Accordingly, small molecules are being developed that target glucose and glutamine metabolisms as well as fatty acid and nucleotide syntheses (Stine et al, 2022).

Several recent studies have shown that cancer cells rely heavily on glutamine metabolism for their proliferation and survival (Jin et al, 2023). Glutamine, a nonessential amino acid, is an important carbon source for the citric acid cycle, and also a nitrogen source for the synthesis of amino acids and nucleotides (Li et al, 2021). It plays a key role in regulation of multiple cellular pathways. Glutamine is hydrolyzed to glutamate in a reaction catalyzed by the enzyme glutaminase (GLS). The GLS I isoform of this enzyme is overexpressed in several cancer types (Masisi et al, 2021). It has been reported that cancer cells, owing to their metabolic reprogramming, can generate glutamate via 2 alternative sources including the N-acetyl-aspartyl-glutamate pathway and the GLS II pathway (Li et al, 2021). Because of the importance of glutamine in cancer cell metabolism, glutaminolysis has become an important target to develop cancer therapeutics. Currently, 2 GLS I inhibitors namely, bis-2-(5-phenylacetamido-1,2,4-thiadiazol-2-yl)ethyl sulfide and CB-839 are being investigated for their anticancer potentials (Li et al, 2021; Jin et al, 2023). CB-839 (also known as telaglenastat) is a selective GLS-1 inhibitor, which inhibits the conversion of glutamine to glutamate for alpha-ketoglutarate formation (Song et al, 2018). Thus, CB-839 inhibits the production of the intermediate metabolites eventually preventing metabolism and energy (ATP) production. CB-839 is currently in clinical trials, being evaluated for treatment of various solid malignancies (https://clinicaltrials.gov).

Metformin is another drug that is linked to cellular metabolism. It is widely used for the treatment of type 2 diabetes. Metformin is also being tested in clinical trials for its anticancer potential. The mechanism by which metformin mediates its effects is complex and multifaceted in nature. It is believed to inhibit complex I of the electron transport chain in mitochondria (Sugden and Holness, 2012), which causes decreases in the NADH oxidation and TCA flux, thus, leading to low levels of TCA metabolites (Sugden and Holness, 2012). Metformin also activates AMPK (AMP-activated kinase) and has been shown to inhibit the growth of various cancer cells in culture (Dowling et al, 2007; Storozhuk et al, 2013; Saraei et al, 2019).

Lung cancer continues to be the most frequently diagnosed cancer in the United States and rest of the world (Bray et al, 2024; Siegel et al, 2023, 2024). Lung cancer is broadly classified as small cell lung carcinoma, which accounts for 15% of all lung cancers, and non–small cell lung carcinoma (NSCLC), which accounts for 85% of all lung cancers (Rodriguez-Canales et al, 2016). In general, lung cancer is diagnosed late when it has already metastasized. Because of high morbidity and mortality associated with lung cancer, newer therapeutics are urgently needed to manage this malignancy. In the present study, we have investigated the anticancer potentials of CB-839 and metformin against human lung cancer cells. We report that treatment with CB-839 and metformin, alone and in combination, induced metabolic stress that led to growth inhibition of human lung cancer cells. These agents predominantly induced apoptotic cell death in A549 and H460 lung cancer cells, at least in part, via death receptor 5 (DR5)–dependent extrinsic pathway. However, these agents led to activation of autophagy in another lung cancer cell line, namely, H1299 cells. We also noted that inhibition of autophagy in H1299 lung cancer cells increased sensitivity to metabolic stress–inducing agents; this increased sensitivity appears to occur, partly, via activation of DR5-dependent extrinsic pathway of cell death. Our results also indicate that the growth inhibitory effects of these agents in A549 and H460 lung cancer cells are linked to K-Ras mutant status.

2. Materials and methods

2.1. Antibodies and reagents

Antibodies anti-Vinculin (#13901), anti–β-actin (#4970, #3700), anti-DR5 (#8074), anti–cleaved caspase-3 (#9661), anti–cleaved caspase-8 (#9496), anti-PARP (#4304), anti-ULK1 (#8054), anti-LC3A/B (#4108), and the horseradish peroxidase–conjugated anti-rabbit (#7074) and anti-mouse (#7076) were procured from Cell Signaling Technologies. Deuterium water, 3-(trimethylsilyl)-1-propanesulfonic acid sodium salt for NMR analyses, and 3-methyl adenine (3MA, autophagy inhibitor) were procured from Sigma-Aldrich. Propidium iodide and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) reagent were purchased from Invitrogen. Pan K-RAS inhibitor (BI-2865) was purchased from Selleck Chem.

2.2. Cells and culture conditions

The following human lung cancer cell lines were used in this study: A549 (K-Ras mutant, from the National Institutes of Health), H1299 (K-Ras wild type, p53 null; from the National Institutes of Health) and H460 (K-Ras mutant, from American Type Culture Collection kindly provided by Dr. Shi Yong Sun, Emory University). Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Corning Life Sciences) containing 10% fetal bovine serum (Gemini Bio-Products Inc). For pyruvate deprivation experiments, cells were washed 3 times with 1X PBS and drug treatment was given in DMEM without glucose, glutamine and pyruvate (Gibco). Cells were treated with CB-839 and/or metformin dissolved in DMSO (100% molecular biology). DMSO was used as vehicle-only for control treatments.

2.3. MTT assay

MTT-based assays for cell growth inhibition were performed as we have described previously (An et al, 2012). Briefly, cells seeded in 96-well plates were treated with vehicle (DMSO) or CB-839 and/or metformin for 24 or 48 hours in pyruvate-deficient media, after which those were incubated with MTT (1 mg/ml) for approximately 3–4 hours. The formazan precipitate formed after MTT incubation were dissolved into DMSO (100% molecular biology). A Bio-Rad Smart-Spec 3000 spectrophotometer was used to determine the sample absorbances at 570 and 690 nm. For quantitative analysis, the A690 value was subtracted from A570 value to remove the background (from all samples and respective controls). The relative cell viability for each sample was then calculated considering the absorbance for their respective controls as 100%. The 24-hour treatments had their respective vehicle control. The 48-hour treatments had their respective control prepared separately from the control for 24-hour treatments.

2.4. NMR for metabolic profiling

NMR for metabolic profiling was performed at SUNY ESF using an 800 Hz Bruker instrument. The obtained spectra were then analyzed using Chenomx software and GraphPad Prism.

2.5. Cell cycle analysis using flow cytometry

Cell cycle analyses were performed using LSRII in Upstate Flow Core facility using protocols as previously described (Rong et al, 2007). Briefly, cells were treated for 48 hours, after which cells were harvested and stained with Propidium iodide to determine cell cycle stages.

2.6. Western blotting

Western blotting was done by standard protocols as previously described (Sun et al, 2013). For all experiments, 20–30 μg/sample protein (for each sample) was loaded and by separated via 10% SDS-PAGE, followed by transfer to a nitrocellulose membrane. Membranes were then blocked using a blocking buffer, followed by overnight incubation at 4 °C with the primary antibodies on a shaker. After incubation with primary antibodies, the membrane was washed 3 times with 1X tri-buffered saline with tween-20, followed by incubation with secondary antibodies. The membrane was washed 3 times again after incubation with secondary antibodies. SuperSignal Pico and Femto West chemiluminescent substrates (Thermo Fisher Scientific) were used to develop the signals and detected using Bio-Rad ChemiDoc XRS+ imaging system. For quantitative analyses, the band intensities were measured using Image Lab v4.1 software (Bio-Rad).

2.7. Cycloheximide chase assay

Cycloheximide chase assay was performed according to protocol previously described (Khayam et al, 2020). Cells were treated in pyruvate-deficient media for 48 hours, after which the media was replaced with DMEM containing 25 μg/ml cycloheximide. Cells were harvested at various time points including 0, 2, 4, and 8 hours; the samples were prepared and analyzed for DR5 and vinculin levels by Western blotting.

2.8. Statistical analyses

Experiments were performed at least 3 times using cells from different passages and on different dates for each cell line. The data were analyzed for statistical significance using 2-way ANOVA in GraphPad Prism (version 10). For all experiments, the groups treated with CB-839 and/or metformin were compared to their respective vehicle (DMSO)-treated controls using Dunnett’s test. The graphs were plotted using Microsoft Excel and GraphPad Prism. Plotted values represent mean ± SEM; P values less than .05 were considered statistically significant.

3. Results

We used H460, A549, and H1299 NSCLC cell lines in our study. H460 cells contain K-Ras (Q61H) mutation, A549 cells harbor K-Ras (G12S) mutation, whereas H1299 cells have wild-type K-Ras. These cells were treated with increasing doses of CB-839 or metformin for 24 hours and 48 hours in pyruvate-deficient cell culture media. As shown in Fig. 1, CB-839 and metformin inhibited the relative cell proliferation of these cells. We noted some differences in these cell lines’ response to treatment with CB-839 and metformin. In the case of H460 cells, both agents inhibited their growth in a dose- and time-dependent manner (Fig. 1A). Both agents also inhibited the growth of A549 cells. However, between the 2 agents, A549 cells were more sensitive to metformin when compared to CB-839 (Fig. 1B). H1299 cells also showed some growth inhibition when treated with these agents (Fig. 1C) but were not as sensitive as H460 cells or A549 cells. We also investigated the effects of the combination of these 2 agents on the growth of these cells. Cells were treated with various combinations of doses for 48 hours. As shown in Fig. 1D for H460 and A549 cells, potentiation of growth inhibitory effects was noted when both agents were combined. We also noted some potentiation of growth inhibitory effects by the combination of these 2 agents in H1299 cells, but the effects were not as pronounced as noted for H460 cells or A549 cells. Because 10 mM metformin alone and in combination with 4 μM CB-849 showed more pronounced growth inhibition in these lung cancer cells, we used these 2 doses and combination in the next series of experiments.

Fig. 1.

CB-839 and metformin inhibit growth of human lung cancer cells. Cells were treated with the indicated doses for indicated time and relative cell viabilities were determined by MTT assay. Results with single agent treatments as indicated: (A) H640 cells, (B) A549 cells, and (C) H1299 cells. (D) Combination treatments in all 3 cell lines. The experiments were performed 3 independent times on different plates using cells from different passages at different times. Plotted values correspond to mean ± SEM of 3-independent experiments; ∗∗P < .01, ∗∗∗P < .001, and ∗∗∗∗P < .0001 in comparison with vehicle-treated controls.

Both CB-839 and metformin affect cellular metabolism, and thus, can be designated as metabolic stress–inducing agents. Next, we sought to investigate the effects of their treatments on cellular metabolic profiles. To that end, we performed NMR metabolomic analysis using A549 cells and H1299 cells. These cells were treated with CB-839 and metformin, alone and in combination, in pyruvate-deficient cell culture media, and samples were prepared for NMR metabolomic analyses. Figure 2A shows the heat maps indicating alterations in the metabolites of the citric acid cycle, amino acids and some other metabolites in A549 cells and H1299 cells. Quantitative results in Fig. 2B show that treatment with CB-839 and metformin, alone and in combination, resulted in a significant decrease in the metabolites of the citric acid cycle in A549 cells when compared to vehicle-treated controls. Alterations in the levels of various amino acids and other metabolites in A549 cells were also noted (Fig. 2C). In the case of H1299 cells, decreases in the components of the citric acid cycle were not as pronounced as noted for A549 cells (Fig. 2D). We did note some alterations in the levels of various amino acids and other metabolites in H1299 cells following treatment with these drugs (Fig. 2E). These results indicate that indeed, these agents induce metabolic stress as is evidenced by their effects on cellular metabolic profiles. Furthermore, the differences in the metabolic profiles noted in these 2 cell lines following treatment with these agents are in line with the differences noted in their growth behavior in response to these agents.

Fig. 2.

Metabolic stress–inducing agents alter the metabolic profiles of human lung cancer cells. (A) Heatmaps showing changes in metabolic profiles of A549 cells (left panel) and H1299 cells (right panel) after treatment with CB-839 and metformin. Controls correspond to samples from vehicle (DMSO)-treated cells. (B) Quantitative results show the effect of CB-839 and metformin, alone and in combination, on the metabolites of the citric acid cycle in A549 cells when compared to vehicle-treated controls. (C) Quantitative results show the effect of CB-839 and metformin, alone and in combination, on amino acids and other metabolites in A549 cells when compared to vehicle-treated controls. (D) Quantitative results show the effect of CB-839 and metformin, alone and in combination, on the metabolites of the citric acid cycle in H1299 cells when compared to vehicle-treated controls. (E) Quantitative results show the effect of CB-839 and metformin, alone and in combination, on amino acids and other metabolites in H1299 cells when compared to vehicle-treated controls. The experiments were performed 3 independent times using cells from different passages at different times for each cell line. Plotted values correspond to mean ± SEM of 3-independent experiments; ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, and ∗∗∗∗P < .0001 in comparison with vehicle-treated controls. Bars without ∗ symbol indicate no significant changes.

Our preceding results indicate that CB-839 and metformin mediate growth inhibition in lung cancer cells and that is coupled with induction of metabolic stress. Next, we sought to determine the mechanism underlying their growth inhibition. It was possible that these agents induced growth arrest and/or cell death. To explore these alternatives, we first investigated the effects of CB-839 and metformin on cell cycle progression in the cell lines used. A549 cells and H1299 cells were treated with CB-839 and metformin, either agent alone and in combination, and cell cycle profiles were determined. Supplemental Figure 1, A and C, shows that A549 cells exhibited some alterations in cell cycle profiles following treatment with these agents. For example, metformin alone and in combination with CB-839 modestly increased the G0/G1 fraction. Metformin alone also exhibited a modest increase in the S phase fraction that was coupled with a concomitant decrease in the G2/M fraction (Supplemental Fig. 1, A and C). CB-839 did not have appreciable effects on cell cycle regulation in A549 cells (Supplemental Fig. 1, A and C). In H1299 lung cancer cells, neither CB-839 nor metformin treatment, alone or in combination, appreciably affected cell cycle regulation (Supplemental Fig. 1, B and D). Together, these results indicate that although metformin has a modest effect on cell cycle regulation in A549 cells, in general, these agents do not strongly affect cell cycle regulation.

Because these agents did not appreciably affect cell cycle regulation in the lung cancer cells used, we investigated their effects on cell death regulation. In general, cell death is mediated via the intrinsic or the extrinsic pathway. The extrinsic pathway is controlled by cell surface death receptors and is also regulated by extracellular stress conditions. DR5 is one of the key death receptors and is regulated in response to cellular stresses. Because CB-839 and metformin induce metabolic stress, therefore, we investigated their effects on DR5 regulation. H460, A549, and H1299 lung cancer cells were treated with CB-839 and metformin, alone and in combination, under similar culture conditions as for the other experiments; Western blot analyses were performed to determine the effect on DR5 protein levels. Figure 3, A and B, shows that CB-839 and metformin treatment increased the DR5 levels in H460 cells and A549 cells, but not appreciably in H1299 cells. DR5 is known to be activated by its ligand Apo2L/TRAIL or if the receptor is overexpressed. Once activated, DR5 then promotes the activation of initiator caspases such as caspase-8, which further activates the effector caspases such as caspase-3. We reasoned that since these agents upregulated DR5 levels, and increased DR5 levels can result in its activation, therefore, initiator caspase 8 and the effector caspase 3 should also be activated. Our results (Fig. 3, A and B) indicate that to be the case. As shown in Fig. 3, A and B, treatment with these agents increased the levels of cleaved caspase 3 and caspase 8 in H460 and A549 cells but not in H1299 cells that also did not show DR5 upregulation (Fig. 3, A and B). These agents also increased PARP cleavage in H460 and A549 cells but not in H1299 cells (Fig. 3, A and B). DR5 upregulation by CB-839 and metformin in H460 and A549 cells, that is coupled with activation of caspases 8 and 3, suggests metabolic stress-induces apoptosis in these cells and engages the extrinsic pathway of cell death, at least in part via, DR5. It is of note that H1299 cells are not very sensitive to the growth inhibitory effects of these agents and accordingly, the DR5 pathway was not appreciably engaged in H1299 cells.

Fig. 3.

CB-839 and metformin-induced cell death is associated with upregulation of DR5 and activation of caspases 8 and 3 in human lung cancer cells. (A) Western blot analyses show that CB-839 and metformin upregulate DR5 coupled with caspases 8 and 3 activations and PARP cleavage in H460 and A549 cells but not in H1299 cells. Vinculin signals in the upper part of the figure serve as loading controls for DR5 and caspase 3. Vinculin signals in the lower part serve as loading controls for caspase 8 and PARP. (B) Bar graphs show quantitative levels of DR5, cleaved caspases-8 and-3, and PARP proteins in all 3 cell lines as detected by Western blotting. Statistical analyses show that CB-839 and metformin treatment do not significantly increase DR5 levels in H1299 cells. Metabolic stress–mediated PARP cleavage was not noted in H1299 cells. These experiments were performed 3 times using cells from different passages at different times for each cell line. Plotted values correspond to mean ± SEM of 3 independent experiments; ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001 in comparison with vehicle-treated controls. Bars without ∗ symbol indicate no significant changes.

We also performed time course of the drugs’ effect on DR5 levels. Because H460 cells were more sensitive to CB-839 than metformin (Fig. 1A), therefore, we used 2 different doses (2 μM and 4 μM) of CB-839 to do time course analyses of DR5 regulation. Our results indicate that CB-839–mediated upregulation of DR5 was noted within 12 hours and maximum induction was achieved by 24 hours (Fig. 4A). Metformin-mediated upregulation of DR5 was also noted within 12 hours of treatment in H460 cells and maximum induction was noted by 24 hours post-treatment (Fig. 4B). In A549, both CB-839 and metformin-mediated DR5 upregulation was noted within 12 hours after treatments (Fig. 4C). Next, we sought to investigate the mechanism by which CB-839 and metformin upregulated DR5 protein levels. One potential mechanism was the effect on DR5 protein stability. To investigate this possibility, we used protein synthesis inhibitor cycloheximide-based assays in H640 cells, and the effect of these agents on DR5 protein stability was determined. Our results demonstrated that both CB-839 and metformin clearly increased the DR5 protein stability in these cells (Fig. 5, A and B).

Fig. 4.

Metabolic stress–inducing agents regulate DR5 protein levels in a dose- and time-dependent manner in H460 and A549 lung cancer cells. (A) Western blot analyses show that CB-839 regulates DR5 in a dose- and time-dependent manner in H460 lung cancer cells. Samples were also probed for cleaved caspase-3 to confirm that drugs were functional. The panel on the right shows quantifications of Western blots signals for DR5 plotted as bar graphs. (B) Western blot analyses show that metformin regulates DR5 in a dose- and time-dependent manner in H460 lung cancer cells. Samples were also probed for cleaved caspase-3 to confirm that drugs were functional. The panel on the right shows quantifications of Western blots signals for DR5 plotted as bar graphs. (C) Western blot analyses show that CB-839 and Metformin regulate DR5 in a time-dependent manner in A549 lung cancer cells. Samples were also probed for cleaved caspase-3 to confirm that drugs were functional. The panel on the right shows quantifications of Western blots signals for DR5 plotted as bar graphs. These experiments were performed 3 independent times using cells from different passages at different times for each cell line. Plotted values correspond to mean ± SEM of 3 independent experiments; ∗P < .05, ∗∗∗P < .001, and ∗∗∗∗P < .0001 in comparison with vehicle-treated controls. Bars without ∗ symbol indicate no significant changes.

Fig. 5.

Metabolic stress–inducing agents increase DR5 protein stability in H460 lung cancer cells. Western blot analyses (A) and quantification of signals as plotted (B) show that CB-839 and metformin increase DR5 protein stability in H460 lung cancer cells. Plotted values represent mean ± SEM (n = 6 for vehicle-treated controls, n = 3 for CB-839-treated samples, n = 3 for metformin-treated samples). These experiments were performed indicated numbers of times independently using cells from different passages at different times for each cell line. Plotted values correspond to mean ± SEM of 3 independent experiments; ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, and ∗∗∗∗P < .0001 in comparison with vehicle-treated controls.

Our results (Fig. 1C) indicate that H1299 cells did show some growth inhibition by CB-839 and metformin, but these cells were clearly less sensitive to these agents when compared to their effects on H460 and A549 cells. Next, we sought to investigate the molecular basis of why H1299 cells were less sensitive to the metabolic stress–inducing agents. Accordingly, we investigated the effects of these agents on autophagy regulation. Our results indicated that indeed, these agents activated the autophagic pathway in H1299 cells as was evidenced by increased levels of Ulk1 and LC3A/B protein levels (Fig. 6A). Ulk1 and LC3A/B proteins are key components of the autophagic pathway; these proteins are the most frequently used markers for autophagy detection. We also investigated CB-839 and metformins' effects on the autophagic pathway in H460 and A549 cells, and our results indicated that these agents did not activate autophagy in H460 cells (Fig. 6A). In A549 cells, we noted a small increase in Ulk1 levels but no appreciable changes in the levels of LC3A/B (Fig. 6A). Autophagy is part of cell survival pathways; accordingly, recent studies have shown that autophagy can help tumor cells survive metabolic stress (Li et al, 2020). Because H1299 cells show less sensitivity to the growth inhibitory effects of metabolic stress–inducing agents such as CB-839 and metformin, it was possible that these cells activated autophagy to counter the effects of these agents. To further explore this possibility, we used autophagy inhibitor 3-Methyladenine (3MA) which has been reported to blunt autophagy activation (Wu et al, 2010; Chicote et al, 2020). We pretreated H1299 cells with autophagy inhibitor 3MA and then subjected them to CB-839 and metformin treatment. Our results (Fig. 6B) show that autophagy inhibitor 3MA increased the sensitivity of H1299 lung cancer cells to CB-839 and metformin. That the autophagy inhibitor 3MA was functional was confirmed by analyzing the effect on LC3A/B, and indeed metabolic stress–induced increase in LC3A/B levels was blunted by 3MA (Fig. 6C), thereby confirming partial autophagy inhibition as has been reported previously (Wu et al, 2010; Chicote et al, 2020). These results suggest that H1299 cells by activating autophagy appear to counter the growth inhibitory effects of metabolic stress–inducing drugs. We next performed experiments to investigate whether the partial inhibition of autophagy also impacted DR5 and cell death regulation in response to metabolic stress in H1299 cells. Our results in (Fig. 6D, i–iv) indicate that once autophagy was inhibited in H1299 cells, metabolic stress led to upregulation of the DR5 protein coupled with biochemical evidence of apoptosis as was ascertained by increasing levels of cleaved caspases 8 and 3 (Fig. 6D, i–iv). To summarize, these results indicate that H1299 cells appear to counter the apoptotic effects of metabolic stress, at least in part, by activating autophagy; once autophagy is inhibited, metabolic stress upregulates DR5 and consequently activates the DR5-dependent extrinsic pathway of cell death as is evidenced by activation of caspases 8 and 3.

Fig. 6.

CB-839 and metformin induce autophagy in H1299 lung cancer cells. (A) Western blot analyses show CB-839 and metformin increase autophagy markers Ulk1 and LC3A/B in H1299 cells but not in H460 and A549 cells. Vinculin is used as a loading control. The bar graphs show quantifications of the Western blots’ signals. (B, C) Pretreatment of H1299 lung cancer cells with 3MA followed by treatment with CB-839 and metformin leads to partial inhibition of autophagy. (B) Pretreatment of H1299 lung cancer cells with 3MA sensitizes H1299 cells to growth inhibitory effects of CB-839 and metformin. (C) Western blot analyses showing significant decrease in LC3A/B levels in H1299 after pretreatment with 3MA; Vinculin serves as a loading control. The panel on the right shows quantifications of Western blots signals as bar graphs (D) Autophagy inhibition in H1299 cells is coupled with DR5 upregulation and concomitant activation of caspases 8 and 3. (Di) Representative Western blots. (Dii) Quantifications of Western blots signals for DR5. (Diii) Quantifications of Western blots signals for cleaved caspase 8. (Div) Quantifications of Western blots signals for cleaved caspase 3. All experiments reported in this figure were performed 3 independent times using cells from different passages at different times for each cell line. Plotted values correspond to mean ± SEM of 3 independent experiments; ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, and ∗∗∗∗P < .0001 in comparison with vehicle-treated controls.

We have shown in the previous sections that H460 cells and A549 cells are clearly more sensitive to the growth inhibitory effects of metabolic stress. Our preceding results also indicate that activation of autophagy in H1299 cells appears to play a role in modulating response to metabolic stress. H460 cells and A549 cells have mutant K-Ras, whereas H1299 cells harbor wild-type K-Ras, it was possible that the status of K-Ras might also play a role in the sensitivity of these cells to metabolic stress. For example, the mutant K-Ras in H460 and A549 lung cancer cells could make these cells more susceptible to metabolic stress. We further investigated the molecular basis of the growth inhibitory effects of metabolic stress on H460 and A549 cells. For that purpose, we used pan-K-Ras inhibitor (BI-2865) in our next series of experiments. pan-K-Ras inhibitor, BI-2865, has been reported to inhibit various types of K-Ras mutants (Kim et al, 2023). We pretreated H460 and A549 lung cancer lines with the pan-K-Ras inhibitor (BI-2865) followed by treatment with CB-839 and metformin under the culture conditions used in our preceding experiments. Our results (Fig. 7) show that in the CB-839- and metformin-treated cells, the relative cell viability was increased in the presence of pan-K-Ras inhibitor. Thus, the pan-K-Ras inhibitor protected these lung cancer cells from the growth inhibitory effects of the metabolic stress induced by CB-839 and metformin. These results indicate that the mutant K-Ras appears to sensitize the H460 and A549 cells to the growth inhibitory effects of metabolic stress.

Fig. 7.

Mutant K-Ras sensitizes H460 and A549 lung cancer cells to the growth inhibitory effects of metabolic stress–inducing agents. Relative cell viability showing the effects of 48-hour (left panel) and 72-hour (right panel) CB-839 and metformin treatments after mutant K-RAS inhibition in H460 and A549 lung cancer cells. Mutant K-Ras inhibition by pan-K-Ras inhibitor (KRASi) prior to treatment with CB-839 and metformin protects H460 cells (A) and A549 cells (B) from the growth inhibitory effects of these agents. These experiments were performed 3 independent times using cells from different passages at different times for each cell line. Plotted values correspond to mean ± SEM of 3 independent experiments; ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, and ∗∗∗∗P < .0001 in comparison with vehicle-treated controls.

4. Discussion

In this study, we report that CB-839 and metformin induce metabolic stress and affect cell viability in human lung cancer cells. Both drugs are clinically relevant, and in clinical trials that are exploring their anticancer potentials. CB-839 induces glutamine-deficient state by inhibiting GLS I, whereas metformin is an antidiabetic drug. However, the mechanisms underlying their anticancer effects remain to be fully elucidated. Here we report that both agents induce cell death by engaging the DR5-dependent extrinsic pathway. Both agents upregulated DR5 levels that was associated with activation of caspases 8 and 3, and PARP cleavage in H460 and A549 lung cancer cell lines. Thus, these agents by inducing metabolic stress mediate their anticancer effects, at least in part, via the DR5 pathway. We also investigated the mechanism by which metabolic stress upregulates DR5. Our results indicate that these agents upregulate DR5 protein levels, at least in part, via the post-translational mechanism by increasing the DR5 protein stability. To the best of our knowledge, DR5 regulation by these agents in this manner has never been reported before.

The lung cancer cell lines used in our study had a variable response to the metabolic stress induced by these agents. H460 and A549 cells were sensitive to the growth inhibitory effects of these agents although, between the 2 cell lines, H460 cells were more sensitive. H1299 cells on the other hand were not as sensitive as H460 or A549 cells. Consistent with these findings, the DR5 pathway was also not engaged in H1299 cells by these agents. We investigated the molecular basis for the differences in these cell lines’ response to treatment with CB-839 and metformin and noted that H1299 cells appear to activate autophagy to counter the negative growth effects of these agents. Autophagy is a complex process that involves multiple players. We used Ulk1 and LC3 as markers of autophagy in our study, as these markers are among the key markers that are linked to activation of autophagy as has been demonstrated by multiple studies. Recently, 604 genes linked to regulation of autophagy were identified (Bordi et al, 2021). Those genes were characterized into different categories; 1 of the categories being the autophagy core. The core group contained 20 markers including Ulk1 and LC3. It was concluded that cell-specific regulation of autophagy markers generally occurs because not all markers linked to autophagy are always regulated. However, the core group contained markers including Ulk1 and LC3 that were intimately linked to regulation of autophagy (Bordi et al, 2021). Bordi et al (2021) further reported that 15 out of these 20 markers were significantly upregulated when cells were subjected to nutrient deficiency, including Ulk1 and LC3. Numerous recent studies have shown that the role of autophagy as a tumor suppressor or promoter relies critically on the cell type and stage of the tumor (Debnath et al, 2023; Li et al, 2020). However, there is now evidence that irrespective of the trigger for autophagy induction, LC3 levels are regulated even if most other markers of autophagy core remain unchanged (Bordi et al, 2021).

Our findings indicate that inhibition of autophagy in H1299 cells sensitized these cells to the growth inhibitory effects of these agents and there was a concomitant upregulation of DR5. Thus, in our study it appears that: (i) H1299 cells activate autophagy to counter the effects of these agents, (ii) induction of autophagy also inhibits DR5 upregulation in response to metabolic stress, and (iii) inhibition of autophagy rescues the ability of these agents to upregulate DR5 coupled with growth inhibition. A recent study investigating the effects of antidepressant desipramine on lung cancer cells noted that autophagy inhibition by chloroquine potentiated the growth inhibitory effects of desipramine that was associated with increased DR5 levels (Zinnah et al, 2023). Thus, it is conceivable that autophagy suppression of DR5 regulation is a common mechanism, and autophagy inhibitors can be used in combination with metabolic stress–inducing agents to further sensitize the growth inhibitory effects of metabolic stress–inducing drugs on cancer cells.

Unlike apoptosis, autophagy is considered as a cell survival mechanism. It is activated in response to stress. Accordingly, it has been implicated in various cellular processes that are linked to cell growth and survival. It is of note that autophagy has been noticed in dying cells under certain cellular conditions. Autophagic cell death is a term that has been coined to describe certain type of cell death, but the exact role of autophagy remains unknown (Denton et al, 2012). Here, we report metabolic stress–induced cell death or autophagy in a cell-specific manner. A crosstalk between metabolic stress–induced apoptosis and autophagy has been reported (Nikoletopoulou et al, 2013). However, the interplays between components of apoptosis and autophagy are complex and warrant in-depth investigations.

H460 and A549 cells were sensitive to the growth inhibitory effects of CB-839 and metformin that exhibited similar molecular response to metabolic stress in terms of DR5 regulation. Although both cell lines are not isogenic, these do harbor mutant K-Ras. Interestingly, inhibition of mutant K-Ras by pan-K-Ras inhibitor blunted the growth inhibitory effects of CB-839 and metformin. These findings suggest that mutant K-Ras also appears to play a role in sensitizing H460 and A549 cells to the growth inhibitory effects of metabolic stress–inducing drugs. This contention is further supported by recent studies reporting that NSCLC cells harboring mutant-K-Ras exhibited a greater dependence on glutamine metabolism (Galan-Cobo et al, 2019; Harding et al, 2021; Xia et al, 2021).

CB-839 is a first-in-class oral agent. The results of a phase I trial in advanced or metastatic solid tumors indicated that CB-839 was safe and had suitable pharmacokinetics and pharmacodynamics (Harding et al, 2021). A proof-of-concept phase II study investigating CB-839 in combination with everolimus in advanced renal cell carcinoma reported CB-839 to improve progression-free survival in patients with metastatic renal cell carcinoma who had been previously treated with tyrosine kinase inhibitors and checkpoint inhibitors. (Lee et al, 2022). Recently, results of a phase Ib/II clinical trial using CB-839 with azacytidine, a DNA methylation inhibitor, in patients with myelodysplastic syndromes indicated this combination to be well tolerated with an objective response rate of 70% (M. Konopleva et al, preprint, DOI: https://doi.org/10.21203/rs.3.rs-2518774/v1). A phase I trial using CB-839 in combination with mTOR kinase inhibitor sapanisertib in patients with advanced NSCLC is also ongoing (Riess et al, 2021).

A large body of evidence from preclinical studies also highlights the anticancer potential of metformin for various malignancies (Iliopoulos et al, 2011; Vancura et al, 2018; Hua et al, 2023). In many observational studies, metformin has shown anticancer potential (Iliopoulos et al, 2011; Vancura et al, 2018; Hua et al, 2023). Results from randomized clinical trials for various malignancies indicate that metformin may not exert anticancer effects in all but a subset of patients with cancer (Iliopoulos et al, 2011; Vancura et al, 2018; Hua et al, 2023). For example, in neoadjuvant setting, metformin was associated with increased pathological complete remission in patients with HER2-positive breast cancer harboring C allele of ATM rs11212617 (Cuyàs et al, 2019). A clinical trial of pediatric patients with brain tumor showed that in a subset of patients, metformin use was associated with better cognitive and neural recovery (Ayoub et al, 2020; Hua et al, 2023). According to a phase II multicenter trial, metformin use was linked to objective prostate-specific antigen response and stabilization of disease in chemotherapy-naïve patients with castration-resistant prostate cancer who did not have diabetes (Rothermundt et al, 2014; Upton et al, 2023). It is of note that, in most of the clinical trials undertaken to date, metformin was found to be safe and well tolerated by the trial participants. It is, therefore, important to identify a subset of patients with lung cancer who would respond to this safe drug. That mutant K-Ras status of lung cancer cells is linked to metformin sensitivity as reported here suggests that, in future clinical trials, patients with lung cancer can be stratified based on their K-Ras status to investigate the anticancer potential of this important and safe medicine.

A finding from our study also indicates that the lung cancer cells that harbor wild-type K-Ras appear to activate autophagy to counter the growth inhibitory effects of metabolic stress–inducing agents such as CB-839 and metformin. Activation of autophagy in cancer cells often serves as a protective mechanism allowing the components of the damaged cells to be utilized as essential nutrients to help in cancer cell survival (Vucicevic et al, 2011; Marino et al, 2012; Debnath et al, 2023). Therefore, in the cancers driven by mutant K-Ras, there appears to be an increased demand for cellular energy; accordingly, autophagy can serve as a gateway to help mitigate this demand (Vucicevic et al, 2011; Marino et al, 2012; Debnath et al, 2023).

NSCLC is the most common type of lung cancer (Rodriguez-Canales et al, 2016) that is also managed by immune checkpoint inhibitors (Lilenbaum, 2024). However, not all patients respond to immunotherapy and some who initially respond, do indeed relapse. Clearly, a lot remains to be done to improve the overall and disease-free survivals in patients with advanced metastatic NSCLC including those that do not respond or relapse on immunotherapy. The metabolic stress–inducing agents reported here can be dseveloped as a therapeutic strategy for such patients.

Conflict of interest

The authors have no financial conflicts to declare relevant to this study.