Skip to main content
NCBI home page
The EMBO Journal logoLink to The EMBO Journal
. 2018 Jun 6;37(14):e98597. doi: 10.15252/embj.201798597

Metabolic vulnerability of cisplatin‐resistant cancers

Florine Obrist 1,2,3,4,5,6, Judith Michels 1,2,3,4,5,6,7, Sylvere Durand 2,3,4,5,6, Alexis Chery 2,3,4,5,6, Jonathan Pol 2,3,4,5,6, Sarah Levesque 1,2,3,4,5,6, Adrien Joseph 2,3,4,5,6, Valentina Astesana 2,3,4,5,6,8, Federico Pietrocola 2,3,4,5,6, Gen Sheng Wu 9, Maria Castedo 2,3,4,5,6,†,✉, Guido Kroemer 2,3,4,5,6,10,11,†,✉
PMCID: PMC6043854  PMID: 29875130

Abstract

Cisplatin is the most widely used chemotherapeutic agent, and resistance of neoplastic cells against this cytoxicant poses a major problem in clinical oncology. Here, we explored potential metabolic vulnerabilities of cisplatin‐resistant non‐small human cell lung cancer and ovarian cancer cell lines. Cisplatin‐resistant clones were more sensitive to killing by nutrient deprivation in vitro and in vivo than their parental cisplatin‐sensitive controls. The susceptibility of cisplatin‐resistant cells to starvation could be explained by a particularly strong dependence on glutamine. Glutamine depletion was sufficient to restore cisplatin responses of initially cisplatin‐resistant clones, and glutamine supplementation rescued cisplatin‐resistant clones from starvation‐induced death. Mass spectrometric metabolomics and specific interventions on glutamine metabolism revealed that, in cisplatin‐resistant cells, glutamine is mostly required for nucleotide biosynthesis rather than for anaplerotic, bioenergetic or redox reactions. As a result, cisplatin‐resistant cancers became exquisitely sensitive to treatment with antimetabolites that target nucleoside metabolism.

Keywords: antimetabolites, cell metabolism, chemotherapy, glutamine, nucleotide

Subject Categories: Autophagy & Cell Death, Cancer, Metabolism

Introduction

The platinum derivative cis‐diamminedichloroplatinum(II) (CDDP), best known as cisplatin, is used for the antineoplastic treatment of patients affected by bladder, head and neck, lung, ovarian, uterine, cervical and germ cell cancers (Kelland, 2007). Intravenous injection of CDDP is associated with high rates of clinical responses. However, with the notable exception of germ cell tumors (Winter & Albers, 2011), neoplastic cells exposed to CDDP ineluctably acquire resistance to the cytostatic and cytotoxic effects of the drug, and eventually resume proliferation, thus causing fatal relapse (Galluzzi et al, 2014). Hence, chemoresistance (be it intrinsic or acquired) constitutes the most prominent obstacle against the clinical use of CDDP.

As a result of the reduced cytoplasmic concentration of chloride (as opposed to sodium) ions, intracellular CDDP is rapidly “aquated”, hence acquiring electrophilic reactivity. Aquated CDDP binds with high affinity to nuclear DNA, in particular to nucleophilic N7 sites on purines, thereby activating the DNA damage response (Wang & Lippard, 2005). Originally, the anticancer effects of CDDP were fully explained by its capacity to induce unrepairable DNA lesions, thereby either triggering an irreversible proliferative arrest known as cellular senescence (which causes cytostasis) or igniting the mitochondrial pathway of apoptosis (which leads to cytotoxicity; Galluzzi et al, 2012a). However, CDDP also physically interacts with cytoplasmic nucleophiles, including mitochondrial DNA (mtDNA) and multiple proteins, thereby (i) stimulating oxidative and reticular stress responses (Martins et al, 2011); (ii) igniting a lethal signaling pathway that involves the pro‐apoptotic BCL‐2 family members BAK1 and BAX, as well as the mitochondrion‐sessile voltage‐dependent anion channel 1 (VDAC1; Tajeddine et al, 2008); and (iii) activating the cytoplasmic pool of the tumor suppressor protein TP53 (Erster et al, 2004). The relative contribution of these cytoplasmic and nuclear pathways may be context‐dependent.

Cells selected by CDDP‐based chemotherapies in vivo or by constant exposure to low CDDP concentrations in vitro activate a variety of resistance mechanisms. Such alterations can (i) affect steps preceding the binding of CDDP to DNA (pre‐target resistance; Hall et al, 2008; Ishida et al, 2010; Karekla et al, 2017), (ii) be directly related to DNA‐CDDP adducts and their repair (on‐target resistance; Ray Chaudhuri et al, 2016; Sourisseau et al, 2016), (iii) invalidate the lethal signaling pathway(s) ignited by CDDP‐induced DNA damage (post‐target resistance; Li et al, 2016), or (iv) affect molecular circuitries that are not directly linked to CDDP‐elicited signals (off‐target resistance; Huang et al, 2016; Leung et al, 2016). In addition, CDDP‐resistant cells undergo a major rewiring in their metabolism, as revealed by changes in the ratio between the enzymes that generate active vitamin B6, pyridoxine kinase (PDXK), or destroy vitamin B6, pyridoxine phosphatase (PDXP; Galluzzi et al, 2012b) and an overactivation of the enzymatic activity of poly(ADP‐ribose; PAR) polymerase (PARP; Galluzzi et al, 2014; Michels et al, 2013). Decreased PDXK expression and high PARP activation are not uniformly associated with CDDP resistance (meaning that they do not occur in all resistant clones), yet constitute clinically useful biomarkers that predict poor prognosis in non‐small cell lung cancer (NSCLC; Galluzzi et al, 2012b; Michels et al, 2015).

The exact mechanisms that account for metabolic rewiring in CDDP resistance have not been elucidated. However, driven by the notion that such a reprogramming process occurs, we decided to explore the metabolic vulnerabilities of CDDP‐resistant cells in a systematic fashion. Here, we reveal the fact that CDDP‐resistant cells are particularly vulnerable to starvation‐induced cell death, due to their particular dependency on glutamine. In CDDP‐resistant cells, glutamine is mostly required for nucleoside biosynthesis rather than for bioenergetic metabolism. As a result, CDDP‐resistant cancer cells become sensitive to chemotherapeutic antimetabolites that poison nucleotide metabolism. Hence, CDDP‐resistant cancers become exquisitely susceptible to treatment by periodic fasting or specific antimetabolites.

Results

Cisplatin‐resistant cancer cells are sensitive to starvation

To identify potential metabolic vulnerabilities linked to cisplatin (CDDP) resistance, we comparatively assessed cell death induction (indicated by a DiOC6(3)‐detectable loss of the mitochondrial transmembrane potential, ΔΨm, alone or accompanied by a propidium iodide [PI]‐detectable loss of plasma membrane integrity) in human A549 non‐small cell lung cancer (NSCLC) cells that were either wild type (WT; i.e., parental) or CDDP‐resistant (clones R2 and R4). These CDDP‐resistant cells had been derived from WT cells by continuous culture in CDDP for several months (Michels et al, 2013). WT, R2 and R4 cells were exposed to a variety of microtubule and metabolic inhibitors. The largest differential susceptibility was observed when the cells were cultured in nutrient‐free conditions (NF), that is, Earle's balanced salt solution (EBSS), which contains no nutrients with the exception of a minimal glucose level of 5.6 mM (Fig 1A). CDDP‐resistant cells died in EBSS much more than CDDP‐sensitive counterparts did (Fig 1A and B). Moreover, CDDP‐resistant cells were more susceptible to cell death induction by microtubule inhibitors (paclitaxel, nocodazole, rotenone), caloric restriction mimetics (C646, spermidine, salicylate), the absence of glucose (or the inhibition of glucose phosphorylation by 2‐deoxyglucose), and lipid‐lowering medication (by means of lipid synthesis inhibitor 5‐(tetradecyloxy)‐2‐furoic acid (TOFA) or the statin simvastatin; Fig 1A). No differences were found for inhibitors of oxidative phosphorylation (antimycin A, metformin, oligomycin), suggesting that the selective susceptibility to rotenone was related to its capacity to inhibit microtubule assembly rather than its inhibitory effects on respiratory chain complex 1 (Meisner & Sorensen, 1966). The susceptibility of cancer cells to starvation‐induced cell death correlated with the accumulation of the PARP product PAR. Thus, CDDP‐resistant clones with high PAR levels (R2, R3, R4, R5) were particularly sensitive to culture in EBSS, while the clone with low PAR level (R1) and parental WT cells were relatively resistant (Fig 1C and D). Inhibition of the enzymatic activity of PARP with three distinct agents (PJ‐34, BMN‐673 and ABT‐888; Penning et al, 2009; Shen et al, 2013), leading to a strong reduction in intracellular PAR levels (Fig EV1A), failed to reverse the killing of R2 and R4 cells by culture in EBSS (Fig EV1B–D). This excludes the possibility that the hyperactivation of PARP would directly cause the selective vulnerability of such cells to starvation‐induced death. Of note, knockdown of pro‐apoptotic proteins from the BCL2 family (BAK, BAX, PUMA) reduced killing of R2 and R4 cells by EBSS, while knockdown of MCL1, which is anti‐apoptotic (Kozopas et al, 1993; Michels et al, 2014b), accelerated killing by EBSS (Fig EV1E and F). These results suggest the involvement of the mitochondrial cell death pathway in starvation‐induced cell death of CDDP‐resistant cancer cells.

Figure 1. Starvation preferentially kills CDDP‐resistant human A549 cancer cells.

Figure 1

Open in a new tab
  • A, B
    Parental A549 cells (WT) and two CDDP‐resistant derivatives (R2 and R4) were maintained in control condition (CTL) or treated with CDDP (30 μM), MEDICA 16 (200 μM), antimycin A (100 μM), pyridoxine (2 mM), pyridoxal (2 mM), metformin (10 mM), oligomycin (10 μM), 3‐bromopyruvate (200 μM), FK866 (500 nM), simvastatin (50 μM), mevastatin (20 μM), nocodazole (200 nM), paclitaxel (100 nM), C646 (50 μM), spermidine (1 mM), perhexiline (10 μM), 6‐aminonicotinamide (100 μM), 2‐deoxygucose (80 μM), salicylate (20 mM), TOFA (80 μM), rotenone (1 μM) or cultured in glucose‐free or EBSS media (nutrient‐free, NF) for 24–48 h. Thereafter, the cells were subjected to the flow cytometry‐assisted measurement of cell death parameters upon co‐staining with the vital dye propidium iodide (PI) and the mitochondrial membrane potential (Δψm)‐sensing dye DiOC6(3) (mean ± SEM; three independent experiments). *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test), in comparison with equally treated WT cells. Representative dot plots of cells cultured in nutrient‐free (NF) conditions are shown in (B) (numbers refer to the percentage of cells found in each quadrant).
  • C
    Parental WT A549 cell line and five CDDP‐resistant (R1–R5) derivatives were cultured in normal growth medium and processed for the immunoblotting‐based assessment of PAR‐containing proteins. Actin levels were monitored to ensure equal loading of lanes. The densitometric analysis of PARylated proteins/actin ratio (upper panel; mean ± SEM, n = 3) and a representative immunoblot (lower panel) are shown. *P < 0.05 (Student's t‐test), as compared to WT cells.
  • D–G
    A549 (D), H460 (E), H1650 (F), and TC‐1 (G) WT and R cells were cultured in normal growth medium (CTL) or nutrient‐free medium (NF) for 24 h (D) or for 36 h (E, F). Thereafter, the cells were subjected to the flow cytometry‐assisted measurement of cell death parameters upon co‐staining with the vital dye propidium iodide (PI) and the mitochondrial membrane potential (Δψm)‐sensing dye DiOC6(3). Data represent mean ± SEM of n independent experiments (n = 3 in D, 4 in E, 5 in F, and 4 in G). *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test), as compared to equally treated WT cells.

Source data are available online for this figure.

Figure EV1. Knockdown of pro‐apoptotic BCL2‐like proteins but not inhibition of the enzymatic activity of PARP reduces starvation‐induced cell death.

Figure EV1

Open in a new tab
  • A
    CDDP‐resistant (R2 and R4) A549 cells were cultured in normal growth medium in the absence or in the presence of three PARP inhibitors (PJ‐34, BMN‐673, ABT‐888) at the indicated concentrations for 15 h and processed for the immunoblotting‐based assessment of PAR‐containing proteins and actin to control the loading of each line.
  • B–D
    Treatment of WT, R2, and R4 A549 cells with 15 μM PJ‐34 (B), 30 nM BMN‐673 (C), and 100 nM ABT‐888 (D) for 24 h failed to significantly reverse cell death parameters induced by culture in EBSS medium (nutrient‐free, NF). Data are represented either as mean ± SEM (n = 3, in B) or mean ± SD (n = 2, in C and D).
  • E, F
    CDDP‐resistant R2 and R4 A549 cells were transfected with control unrelated siRNA (siUNR) or with siRNAs specific for Bcl‐2 family members for 36 h. Thereafter, cells were cultured for 24 h either in the complete medium (CTL) or in EBSS (nutrient‐free, NF) prior to the cytofluorometric assessment of apoptosis‐related variables (mean ± SEM; n = 3). *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test), as compared to cells of the same type transfected with siUNR.

Source data are available online for this figure.

Importantly, the selective susceptibility of CDDP‐resistant cells to EBSS was observed for other pairs of sensitive versus resistant human NSCLC lines such as H460, H1650 and mouse lung cancer line TC‐1 (Fig 1E–G). We also evaluated the possibility to combine CDDP treatment with nutrient depletion. This combination exhibited an additive cell‐killing potential, when applied to CDDP‐sensitive A549 cells, yet failed to yield such effects on CDDP‐resistant cells. Such cells succumbed to nutrient depletion alone, and this starvation‐induced killing was not further enhanced by CDDP (Fig EV2A and B).

Figure EV2. Nutrient depletion and CDDP resistance.

Figure EV2

Open in a new tab
  • A, B
    A549 WT and R2 cells were cultured in complete medium (CTL) or EBSS medium (NF) and exposed for 30 h to the indicated concentrations of CDDP. Thereafter, cells were subjected to flow cytometry‐assisted measurement of cell death parameters. Values represent the percentage of dying DiOC6(3)lowPI− plus dead PI+ cells. Data represent mean ± SEM of three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test), as compared to cells in the same medium, without CDDP.

Considering the effect of nutrient depletion on CDDP‐resistant cells in vitro, we tested whether the selective susceptibility of CDDP‐resistant cells to nutrient depletion could be observed in vivo as well. Indeed, A549 R4 tumors developing in immunodeficient mice reduced their growth in response to periodic starvation (24 h of fasting twice per week), while parental A549 tumors were not affected by this regimen (Fig 2A and B). Accordingly, periodic starvation was able to prolong the survival of mice bearing xenografted CDDP‐resistant but not parental NSCLC (Fig 2C and D).

Figure 2. Therapeutic effects of starvation on CDDP‐resistant xenografts in vivo .

Figure 2

Open in a new tab
  • A, B
    WT A549 cell line (A) and its CDDP‐resistant derivative R4 (B) were subcutaneously xenografted into athymic nu/nu mice (12 mice in WT CTL, 11 mice in WT NF, 8 mice in R4 CTL and R4 NF). When tumors became palpable, mice were fed ad libitum or underwent cycles of starvation (24 h, two times a week). Tumor growth was routinely monitored with a standard caliper and is reported as means ± SEM. *P < 0.05 (Wald test, type 2 ANOVA), as compared to mice fed ad libitum.
  • C, D
    Kaplan–Meier survival curves of nude mice xenografted with A549 WT or CDDP‐resistant R4 cells, and fed ad libitum or starved 24 h, two times a week (12 mice in WT CTL, 11 mice in WT NF, 8 mice in R4 CTL, and 8 mice in R4 NF). Starvation significantly prolongs survival of mice xenografted with CDDP‐resistant R4 A549 cancer cells (log‐rank test).

Glutamine dependency of cisplatin‐resistant cancer cells

Next, we attempted to determine which specific nutrients might rescue CDDP‐resistant cancer cells from death occurring in EBSS. Glutamine (GLN) turned out to be the most effective agent to close‐to fully suppress the death of R2 or R4 cells in EBSS (Fig 3A and B). Glutamate (GLU) had a smaller but still significant effect, while the cell‐permeable α‐ketoglutarate precursor, dimethyl α‐ketoglutarate, exhibited rather partial effects. In contrast, glucose, amino acids, the cell‐permeable pyruvate derivative, 3‐methyl pyruvate, polyamines and glutathione‐replenishing agents (glutathione ester or N‐acetylcysteine) failed to reverse the lethal effects of EBSS (Fig 3A). The effects of GLN were obtained at relatively low doses (20 μM) at which GLU had no effects (Fig 3C). The rescue by GLN was observed in multiple CDDP‐resistant human and mouse cancer cell lines (Fig 3C–F). These results underline the key role of glutamine in cell survival in the context of nutrient depletion. Of note, fasting of mice for 24 h (which reduced the growth of CDDP‐resistant tumors, see above, Fig 2A and B) also led to a reduction in plasma GLN levels (Fig EV3A).

Figure 3. Glutamine and glutamate sustain the survival of CDDP‐resistant cells during starvation.

Figure 3

Open in a new tab
  • A, B
    CDDP‐resistant A549 R2 and R4 cells were cultured for 24 h in EBSS in the absence or presence of the indicated nutrients (l‐glutamine (2 mM), glutamate (2 mM), dimethyl α‐ketoglutarate (1 mM), citrate (1 mM), d‐glutamine (2 mM), histidine (0.15 mM), pyridoxal phosphate (2 mM), leucine (0.1 mM), valine (0.45 mM), glucose (5.56 mM), nicotinamide (100 mM), pyridoxine (2 mM), glutathion ester (5 mM), N‐acetylcysteine (10 mM), spermidine (30 μM), putrescine (100 uM), asparagine (0.05 mM), arginine (0.7 mM), aspartate (0.5 mM)), then processed for the cytofluorometric determination of cell death‐related parameters. *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test; n = 3), compared with cells of the same type exposed to EBSS alone. Rescue = (% of cells death in EBSS − % of cell death in EBSS supplemented with nutrient)/(% of cells death in EBSS) × 100. Representative dot plots of cells cultured in EBSS medium (nutrient‐free, NF) in the absence or in the presence of 2 mM glutamine (GLN) are shown in (B). Numbers refer to the percentage of cells found in each quadrant.
  • C
    A549 WT and R cells were cultured in normal growth medium (CTL) or EBSS medium (NF) and exposed to increasing concentrations (0.02, 0.2, and 2 mM) of glutamine (GLN) or glutamate (GLU) before the evaluation of the cell death‐associated parameters. White and black columns depict the percentage of dying and dead cells, respectively (mean ± SEM; n = 3). *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test), in comparison with cells of the same type in EBSS alone.
  • D–F
    H460 (D), H1650 (E), and TC‐1 (F) WT and R cells were cultured in EBSS alone or in combination to 2 mM glutamine (GLN) for 36 h. PI+ = dead cells; DiOC6(3)low PI− = dying cells. Data represent mean ± SEM of n independent experiments (n = 5 in D, n = 4 in E, and n = 3 in F). *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test), as compared to cells of the same type exposed to EBSS alone. # P < 0.05, ## P < 0.01 (Student's t‐test), as compared to equally treated WT cells.

Figure EV3. Relationship of GLN and CDDP sensitivity.

Figure EV3

Open in a new tab
  • A
    Mass spectrometric assessment of plasma GLN levels in mice before and after 24 h of starvation. Mice bearing parental (WT) A549 or CDDP‐resistant R4 tumors (as in Fig 2) were subjected to one cycle of 24‐h fasting (with ad libitum access to water), and blood was drawn before and immediately after the fasting cycle. Values represent means ± SEM (n = 4 for WT, n = 6 for R4). **P < 0.01 (paired Student's t‐test) as compared to the fed state.
  • B, C
    Sensitization of CDDP‐resistant cells to cell death induction by GLN depletion. Parental (B) and CDDP‐resistant (C) TOV 112D cells were cultured in complete medium (CTL) or glutamine (GLN)‐free medium and exposed for 48 h to the indicated concentrations of CDDP. Thereafter, the cells were subjected to the flow cytometry‐assisted measurement of cell death parameters. Values represent the percentage of dying DiOC6(3)lowPI− plus dead PI+ cells (mean ± SEM; n = 3). *P < 0.05, **P < 0.01 (Student's t‐test), as compared to cells of the same type in complete medium.

In the subsequent step, we determined whether GLN withdrawal from the medium would be sufficient to kill CDDP‐resistant cells. While this was not the case, GLN depletion was sufficient to reestablish CDDP‐induced killing of a priori CDDP‐resistant cells. Thus, A549 R2 and R4 clones, as well as other CDDP‐resistant cells (such as the NSCLC H460 R cell line, the NSCLC H1650 R cell line and the ovarian carcinoma TOV 112D R cell line), became susceptible to CDDP‐induced cell death when they were cultured in the absence of GLN (Figs 4A–J, and EV3B and C). In conclusion, it appears that the abundance of GLN has a major impact on the cytotoxicity of CDDP, in particular in cells that have been selected for CDDP resistance.

Figure 4. Glutamine starvation sensitizes human cancer cells to CDDP .

Figure 4

Open in a new tab
  • A–J
    A549 (A–F), H460 (G, H), and H1650 (I, J) WT and R cells were cultured in complete medium (CTL) or glutamine (GLN)‐free medium, and exposed for 24 h (A–C) or 48 h (D–J) to the indicated concentrations of CDDP. Thereafter, cells were subjected to flow cytometry‐assisted measurement of cell death parameters. Values represent the percentage of dying DiOC6(3)lowPI− plus dead PI+ cells. Data represent mean ± SEM of three independent experiments except for (E) (n = 4). *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test), as compared to cells of the same type in CTL medium.

Glutamine‐fueled nucleoside synthesis in cisplatin resistance

To understand the mechanism through which GLN rescues CDDP‐resistant cells from starvation‐induced death, we resorted to mass spectrometric metabolomics. We compared the levels of metabolites detectable in EBSS (i.e., in conditions of starvation, also referred as nutrient‐free condition, NF) with those found in complete medium (control, CTL) or in EBSS supplemented with 2 mM GLN (NF + GLN). As expected (Zhang et al, 2017), GLN was particularly efficient in replenishing its amino acid derivatives alanine, asparagine and GLU, the GLU metabolite α‐ketoglutarate, some intermediates of the Krebs cycle (fumarate, malate) and glutathione (written in red in Fig 5A). Of note, in normal culture conditions, resistant clones were characterized by a relative depletion of Krebs cycle intermediates (α‐ketoglutarate, fumarate, malate, citrate/isocitrate, oxaloacetate/pyruvate) when compared to parental A549 cells (Figs 5B and EV4A). Driven by these observations, we explored the mechanisms through which GLN rescues CDDP‐resistant cells. To fuel the Krebs cycle, intracellular GLN must be converted to GLU (which is the precursor of the anaplerotic substrate α‐ketoglutarate). This amidohydrolase reaction is catalyzed by glutaminase (GLS; Fig 5C). We therefore expected that GLS inhibition by bis‐2‐(5‐phenylacetamido‐1,3,4‐thiadiazol‐2‐yl)ethyl sulfide (BPTES) would abolish the rescue effect of GLN. In stark contrast, however, BPTES failed to counteract the pro‐survival action of GLN on CDDP‐resistant cells cultured in EBSS. Rather, BPTES reduced the mortality of R2 and R4 cells in EBSS as it reduced the intracellular GLU concentrations (in WT, R2 and R4 cells), while it tended to augment GLN (in WT cells; Fig 5D and E). Similarly, another pharmacological GLS inhibitor compound 968 (C968) reduced the killing of R2 and R4 cells by starvation (Fig EV4B). Finally, knockdown of GLS with two distinct, non‐overlapping siRNAs (Fig EV4C) partially rescued R2 and R4 cells from the cytotoxic consequences of starvation (Figs 5F and EV4D).

Figure 5. Inhibition of glutaminase (GLS) extends the survival of nutrient‐starved CDDP‐resistant cancer cells.

Figure 5

Open in a new tab
  1. WT, R2, and R4 A549 cells were cultured for 10 h in complete medium (CTL) or nutrient‐deprived medium (NF), in the absence or presence of 2 mM glutamine (NF + GLN). Heatmap represents the amount of each metabolite (log2 scale) in nutrient‐deprived medium (NF), shown as a black (high) and white (low) gradient. Metabolite differences between CTL and NF, or NF + GLN and NF are shown as a color gradient (log2 scale). Five replicates per condition. Both metabolites (rows) and conditions (columns) were clustered by means of the Ward method on the Euclidean distance matrix.
  2. Heatmap indicates the level of Krebs cycle‐related intermediates in WT, R2, and R4 A549 cells maintained in complete medium. For all metabolites, except for fumarate, differences between parental (WT), and CDDP‐resistant (R2 and R4) cells were significant (P < 0.001, Student's t‐test). These data were extracted from Fig EV4A.
  3. Schematic representation of the main pathways of glutamine (GLN) metabolism.
  4. WT, R2, and R4 A549 cells were cultured in complete medium (CTL) or EBSS in the absence or presence of the GLS inhibitor BPTES (5 μM) for 24 h and then assayed for cell death parameters. DiOC6(3)low PI− = dying cells, PI+ = dead cells (mean ± SEM; n = 3). *P < 0.05 (Student's t‐test) as compared to cells of the same type cultured in the same medium, but in the absence of BPTES.
  5. Levels of GLU and GLN in parental WT and the two CDDP‐resistant R2 and R4 cancer cells cultured for 10 h in EBSS supplemented or not with 5 μM BPTES. Data are shown as area of the metabolite peak, normalized to the metabolite peak of WT cells cultured without BPTES. Means ± SEM of five replicates. *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test), as compared to cells of the same type cultured in the absence of BPTES.
  6. Parental (WT) and CDDP‐resistant (R) A549 cells were transfected with control siRNA (siUNR) or with siRNAs specific for glutaminase (siGLSA) for 48 h. Thereafter, cells were cultured for 24 h either in the complete medium (CTL) or in EBSS prior to the cytofluorometric assessment of apoptosis‐related variables. DiOC6(3)low PI− = dying cells, PI+ = dead cells (mean ± SEM; n = 3 independent experiments). *P < 0.05, **P < 0.01 (Student's t‐test), as compared to cells of the same type transfected with siUNR.

Figure EV4. Inhibition of glutaminase activity during starvation increases cell survival and nucleotide levels in CDDP‐resistant cells.

Figure EV4

Open in a new tab
  • A
    Heatmap visualization of metabolic profiling in WT, R2, and R4 A549 cells cultured for 10 h in complete medium (CTL). Both metabolites (rows) and conditions (columns) were clustered by means of the Ward method on the euclidean distance matrix. Expression level of each metabolite in each type of cells maintained in EBSS is represented as a gradient from blue (low) to red (high) in a log2 scale (five replicates per condition). In red, metabolites represented in Fig 5B.
  • B
    WT and R cells were cultured in complete medium (CTL) or EBSS in the absence or in the presence of the GLS inhibitor C968 (1 μM) for 24 h and then assayed for cell death parameters (mean ± SEM; n = 3). *P < 0.05 (Student's t‐test) as compared to cells of the same type cultured in the absence of C968.
  • C, D
    Parental A549 cells (WT) and two CDDP‐resistant derivatives (R2 and R4) were transfected either with control siRNA (UNR) or with two distinct glutaminase‐specific siRNAs (si GLSA or si GLSB). Representative immunoblots confirming downregulation of GLS in CDDP‐resistant R4 cells 36 h after transfection are shown in (C). Thirty‐six hours after transfection with control or GLS specific siRNA, cells were cultured in complete medium (CTL) or in EBSS for additional 24 h and then subjected to the cytofluorometric assessment of cell death parameters (D). DiOC6(3)lowPI− = dying cells, PI+ = dead cells (mean ± SEM; n = 3). *P < 0.05, ***P < 0.001 (Student's t‐test), as compared to cells of the same type transfected with UNR siRNA.
  • E
    Levels of AMP and UMP in parental WT and CDDP‐resistant (R2 and R4) cancer cells cultured for 10 h in EBSS in the absence or in the presence of 5 μM BPTES. Data are shown as area of the metabolite peak. Means ± SEM of five replicates. *P < 0.05 (Student's t‐test), as compared to cells of the same type cultured in EBSS.

Source data are available online for this figure.

Based on the aforementioned results, we speculated that GLN‐fueled nucleoside biosynthesis (which does not require the action of GLS and actually would be favored by GLS inhibition, Fig 5C) might account for its rescue effect on starved CDDP‐resistant cells. Indeed, GLN was able to normalize the intracellular concentration of succinyl adenosine (a precursor of AMP), adenosine monophosphate (AMP), adenosine diphosphate (ADP) and adenosine triphosphate (ATP) in CDDP‐resistant cells cultured in nutrient‐free conditions (Fig 6A–D). Similarly, GLS inhibition by BPTES resulted in a significant elevation of AMP and uridine monophosphate (UMP) in R2 and R4 cells (Fig EV4E). Direct addition of nucleosides (and in particular a mixture of all four ribonucleosides: adenosine, guanosine, uridine and cytidine [AGUC]) rescued all tested CDDP‐resistant NSCLC lines from starvation‐induced killing. Ribonucleosides were more efficient than their desoxyribonucleoside derivatives (Fig 6E and F).

Figure 6. Glutamine promotes nucleotide synthesis in starved CDDP‐resistant cells.

Figure 6

Open in a new tab
  • A–D
    Levels of succinyl adenosine (A), AMP (B), ADP (C), and ATP (D) in parental WT and CDDP‐resistant R2 and R4 cancer cells cultured for 10 h in complete medium (CTL) or in EBSS medium (nutrient‐free: NF) supplemented or not with 2 mM GLN (five replicates per condition). Data are shown as area of the metabolite peak. Means ± SEM of five replicates. # P < 0.05, ## P < 0.01, ### P < 0.001 (Student's t‐test), as cells in EBSS compared to cells of the same type in complete medium; *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test), as compared to cells of the same type cultured in EBSS alone.
  • E, F
    CDDP‐resistant cells were incubated in EBSS with or without adenosine (A), guanosine (G), uridine (U), cytidine (C), deoxyadenosine (dA), deoxyguanosine (dG), thymidine (dT), deoxycytidine (dT) each at 0.1 mM or in combination during 24 h (A549 cells) or 36 h (H1650 cells), then processed for the cytofluorometric determination of cell death‐related parameters upon co‐staining with the vital dye propidium iodide (PI) and the mitochondrial membrane potential (Δψm)‐sensing dye DiOC6(3). Heatmaps in (E) represent the percentage of cell rescue by nucleosides. Rescue = (% of cells death in EBSS − % of cell death in EBSS supplemented with nucleosides)/(% of cells death in EBSS) × 100. *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test; n = 3), as compared to cell death in the absence of nucleosides. Representative dot plots of A549 cells are shown in (F). Numbers refer to the percentage of cells found in each quadrant.

Altogether, these results suggest that GLN rescues CDDP‐resistant cells from starvation‐induced death by elevating the intracellular concentrations of nucleosides rather than by fueling anaplerotic reactions.

Selective susceptibility of cisplatin‐resistant cells to antimetabolites targeting nucleotide biosynthesis

We next explored the possibility that CDDP‐resistant cells might be particularly vulnerable to chemotherapeutic agents that target nucleotide‐related pathways such as 5‐fluorouracil (5‐FU, an inhibitor of thymidylate synthase; Chon et al, 2017), clofarabine and gemcitabine (CLO and GCB, two inhibitors of ribonucleotide reductase; Aye et al, 2015). Upon short‐term exposure (24 h), such antimetabolites failed to kill A549 R2 and R4 cells on their own, yet counteracted the rescue effect of GLN on starved cells (Fig 7A and B). Upon long‐term exposure (48 h), 5‐FU, CLO, cytarabine (CTB) and cladribine (2CdA), two other antimetabolites antagonizing nucleotide metabolism, killed R2 and R4 cells more efficiently than their parental equivalent (Figs 7C and D, and EV5A and B). Similarly, 5‐FU killed other CDDP‐resistant NSCLC cell lines (H1650, H460) more efficiently than their CDDP‐susceptible precursors (Fig EV5C and D). Moreover, CDDP‐resistant A549 tumors significantly reduced their growth upon treatment with 5‐FU in vivo, in immunodeficient mice, contrasting with wild‐type tumors that barely responded to this chemotherapeutic regimen (Fig 7E and F). Accordingly, 5‐FU was able to prolong survival of mice bearing xenografted CDDP‐resistant but not parental NSCLC (Fig 7G and H). In conclusion, it appears that CDDP‐resistant tumors are endowed with an exquisite sensitivity to antimetabolites targeting nucleotide biosynthesis.

Figure 7. Inhibition of nucleotide biosynthesis preferentially kills CDDP‐resistant cancer cells.

Figure 7

Open in a new tab
  • A, B
    Cytofluorometric assessment of cell death in A549 CDDP‐resistant R2 (A) and R4 (B) cancer cells cultured in complete medium (CTL), EBSS (NF), or EBSS containing 0.02 mM glutamine (NF + GLN), in the absence or in the presence of 5‐fluorouracil (5‐FU; 60 μM), clofarabine (CLO; 2 μM), or gemcitabine (GCB; 2 μM) for 24 h. DiOC6(3)low PI− = dying cells, PI+ = dead cells (mean ± SEM; n = 3). *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test), as compared to cells in the same culture medium in the absence of nucleotide antagonists.
  • C, D
    A549 parental (WT) and CDDP‐resistant (R2 and R4) cancer cells were cultured in complete medium, either untreated or exposed to the indicated concentrations of 5‐fluorouracil (5‐FU; in C) or clofarabine (CLO; in D). After 48 h of incubation, the cells were subjected to the flow cytometry‐assisted measurement of cell death parameters. Values represent the percentage of dying DiOC6(3)lowPI− plus dead PI+ cells (mean ± SEM; n = 4 in C and 3 in D). *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test), as compared to equally treated WT cells.
  • E, F
    WT cell line (E) and its CDDP‐resistant R4 derivative (F) were subcutaneously xenografted into athymic nu/nu mice. When tumors became palpable, animals were randomized and treated with 5‐fluorouracil (5‐FU; i.p. injection) or an equivalent volume of vehicle (CTL), three times per week for 12 weeks. Tumor growth is reported as means ± SEM (WT CTL and WT 5‐FU, 10 mice; R4 CTL, 13 mice; R4 5‐FU, 11 mice). **P < 0.01 (Wald test, type 2 ANOVA), as compared to CTL.
  • G, H
    Kaplan–Meier survival curves of nude mice xenografted with A549 WT (G) or CDDP‐resistant R4 cancer cells (H), and treated with 5‐FU or an equivalent volume of vehicle (CTL). Treatment with 5‐FU significantly prolongs survival of mice xenografted with CDDP‐resistant R4 A549 cancer cells (log‐rank test).

Figure EV5. Selective susceptibility of cisplatin‐resistant cells to antimetabolites.

Figure EV5

Open in a new tab
  • A, B
    A549 parental (WT) and CDDP‐resistant (R2 and R4) cancer cells were cultured in complete medium, either untreated or exposed to the indicated concentrations of cytarabine (CTB; in A) or cladribine (2CdA; in B). After 48 h of incubation, the cells were subjected to the flow cytometry‐assisted measurement of cell death parameters. Values represent the percentage of dying DiOC6(3)low PI− plus dead PI+ cells (mean ± SEM; n = 3 in a and 4 in b). *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test), as compared to equally treated WT cells.
  • C, D
    WT and CDDP‐resistant H1650 (C) or H460 (D) human NSCLC cells cultured in normal growth medium were either untreated or exposed to the indicated concentrations of 5‐fluorouracil (5‐FU) during 48 h. Then, the cells were subjected to the flow cytometry‐assisted measurement of cell death parameters. Values represent the percentage of dying DiOC6(3)low PI− plus dead PI+ cells (mean ± SEM; n = 3). *P < 0.05, **P < 0.01, ***P < 0.001 (Student's t‐test), as compared to equally treated WT cells.

Discussion

In spite of the surge of targeted anticancer treatments and immunotherapies, CDDP is still the most widely used anticancer agent and CDDP resistance continues to pose a major problem in the clinical management of malignant diseases. Here, we investigated the metabolic vulnerabilities of CDDP‐resistant NSCLC and ovarian cancers. We found that CDDP‐resistant cells became sensitive to nutrient depletion, meaning that they died in vitro upon culture in nutrient‐free medium. Moreover, CDDP‐resistant cancers drastically reduced their growth in mice that were subjected to repeated fasting cycles, contrasting with their CDDP‐sensitive parental cancers that were not affected by fasting. Although the biochemical consequences of starvation of cells in vitro (by removal of multiple or individual nutrients from the medium) and starvation of mice in vivo (by removal of the food supply) admittedly could be quite distinct, the selective susceptibility of CDDP‐resistant cells to both types of starvations (in vitro and in vivo) appears coherent in its pattern. Notably, it has been previously shown that fasting cycles can sensitize cancer cells to chemotherapy (Lee et al, 2012), but our data demonstrate for the first time the specific vulnerability of CDDP‐resistant tumors to fasting cycles. This could have promising applications in clinical settings, as fasting cycles could be proposed to patients, after CDDP treatment, to target emerging CDDP‐resistant cells. Stimulated by this encouraging result, we investigated the particular metabolic needs of CDDP‐resistant cells and identified glutamine as a factor that can suppress starvation‐induced cell death and whose depletion alone would be sufficient to re‐sensitize normally CDDP‐resistant cells to CDDP.

Previous studies have dealt with changes in energy metabolism that are coupled to CDDP resistance. Thus, it has been reported for different CDDP‐resistant cell lines that they increase glycolysis (Qian et al, 2017), switch to oxidative phosphorylation (Galluzzi et al, 2014; Matassa et al, 2016; Wangpaichitr et al, 2017), and/or increase GLN metabolism by upregulating the GLN transporter ASCT2 and GLS (Hudson et al, 2016; Wangpaichitr et al, 2017). According to one study, the critical target explaining glutamine metabolism‐linked CDDP resistance was GLS, meaning that knocking down GLS was sufficient to re‐sensitize CDDP‐resistant ovarian cancers to CDDP, while its transgenic overexpression in CDDP‐sensitive cells could confer CDDP resistance (Hudson et al, 2016). However, we found that genetic or pharmacologic GLS inhibition did not reverse the capacity of GLN to rescue CDDP‐resistant cells from starvation‐induced death. Rather, GLS inhibition had a rescue effect on its own, perhaps because it prevented the conversion of dwindling sources of intracellular GLN to GLU, hence maintaining GLN at a level compatible with fueling nucleotide biosynthesis. In accord with this analysis, cell‐permeable dimethyl α‐ketoglutarate, a precursor of α‐ketoglutarate that readily penetrates into A549 cells (Marino et al, 2014), was unable to replace GLN and hence to rescue CDDP‐resistant A549 cells from starvation‐induced death. Moreover, high doses of the glutathione precursors glutathione ethyl ester or N‐acetyl cysteine also were unable to replace GLN in this rescue assay. Altogether, these results indicate that anaplerotic, Krebs cycle‐related, and redox reactions were not important for the GLN‐mediated rescue effect. In line with this idea, even rather small doses of GLN (in the range of 50 μM) were sufficient to rescue CDDP‐resistant A549 cells from death, underscoring the notion that subtle effects (such as nucleotide biosynthesis) rather than bioenergetically relevant reactions (that would involve the conversion of GLN into GLU and then into the anaplerotic substrate α‐ketoglutarate) requiring high GLN concentrations are involved in the rescue effect. In line with this notion, Tardito et al have shown that GLN‐starved glioblastoma cells were not rescued by TCA cycle replenishment (Tardito et al, 2015). Here, we report that supplementation with ribonucleosides could effectively suppress starvation‐induced cell death in CDDP‐resistant NSCLC cells. Of note, Brown et al (2017) have recently shown that genotoxic chemotherapeutic agents (including cisplatin) can induce an elevation of nucleotide synthesis, which is necessary for cell survival. It is tempting to speculate that such an adaptation in nucleotide metabolism occurs in response to DNA repair during CDDP treatment and then persists in CDDP‐resistant cells after CDDP removal thereby inducing metabolic vulnerabilities. Of note, our work does not clarify whether such metabolic rewiring is responsible for cisplatin resistance. Rather, it reveals a specific characteristic of cisplatin‐resistant cells that be taken advantage of to kill them.

More importantly from the therapeutic point of view, we observed that CDDP‐resistant cells acquired an exquisite susceptibility to several chemotherapeutic agents that inhibit nucleotide metabolism such as 5‐fluorouracil (5‐FU, an inhibitor of thymidylate synthase; Chon et al, 2017) or clofarabine (CLO, an inhibitor of ribonucleotide reductase; Aye et al, 2015). Interestingly, a recent siRNA‐based genetic screen also revealed that knockdown of ribonucleoside‐diphosphate reductase subunit M2 B can sensitize cancer cells to CDDP as well (Leung et al, 2016), pleading in favor of the specificity of the effects. Moreover, several combination chemotherapy trials have established the superiority of CDDP combined with the aforementioned chemotherapeutic antimetabolites over monotherapies (Decker et al, 1983; Heinemann et al, 2006; Comella et al, 2007). The present data may provide a rational explanation for this combination effect. In line with this notion, the combination of CDDP and raltitrexed, a chemotherapeutic agent that is a folic acid antagonist inhibiting the synthesis of nucleotides precursors, improves overall survival compared with CDDP alone in patients with malignant pleural mesothelioma (van Meerbeeck et al, 2005). Of note, the pretreatment with pemetrexed (Alimta®), another folate antimetabolite, re‐established in vitro CDDP‐induced killing of a CDDP‐resistant NSCLC cell population (Tieche et al, 2016).

Previous studies revealed that hematopoietic stem cells only undergo erythroid differentiation upon supplementation of extra GLN or nucleosides, exemplifying a physiological case of “GLN addiction” (Oburoglu et al, 2014). In the context of cancer, activation of the oncogenic transcription factor MYC is well known to induce GLN addiction (Yuneva et al, 2007, 2012; Altman et al, 2016). Moreover, autophagy‐deficient KRAS‐induced lung cancers reportedly rely on extra supply of GLN or nucleosides (Guo et al, 2016). Although MYC can cause CDDP resistance (Sklar & Prochownik, 1991), we found no signs of autophagy deficiency in the CDDP‐selected NSCLC cell lines characterized here (Michels et al, 2014a). Hence, the exact relationship between transcriptional effects and metabolic reprogramming with respect to GLN metabolism remains to be investigated.

In synthesis, CDDP resistance is coupled to major shifts in cellular metabolism that, in several NSCLC and ovarian cancer models, causes a relative GLN dependency. Metabolomic, genetic, and pharmacological studies indicate that GLN must fuel a nucleotide biosynthesis pathway in the context of CDDP resistance. Consequently, CDDP‐resistant cells become exclusively sensitive to fasting as well as to antimetabolites that target nucleotide synthesis.

Materials and Methods

Cell lines, culture conditions, and chemicals

Culture media and cell culture supplements were purchased from Life Technologies (Carlsbad, CA, USA) unless otherwise specifically mentioned. Non‐small cell lung cancer (NSCLC) cells and both parental (also known as wild type (WT)) and their CDDP‐resistant counterparts were maintained at 37°C under 5% CO2, in the following culture media: Glutamax‐containing Dulbecco's modified Eagle's medium/F12 medium supplemented with 10% fetal bovine serum (FBS), 10 mM HEPES buffer, 100 units/ml penicillin G sodium, and 100 mg/ml streptomycin sulfate for human NSCLC A549 cells; RPMI‐1640 medium supplemented as above for human NSCLC H460 and H1650 cells; A 1:1 mixture of MCDB 105/M199 medium (Sigma‐Aldrich, St Louis, MO, USA) supplemented as above and with 0.75 g/l sodium bicarbonate in addition, for TOV‐112D cells; RPMI‐1640 medium supplemented as above, and with non‐essential amino acids in addition, for murine TC1 cells. WT cells were purchased from American Type Culture Collection, and their CDDP‐resistant counterparts were obtained in vitro by prolonged culture of parental WT cells with sublethal CDDP concentrations as previously described (Michels et al, 2013). The following chemicals were purchased from Sigma‐Aldrich: acetic acid, acetonitrile, adenosine, 6‐aminonicotinamide, antimycin A, arginine, asparagine, aspartate, BPTES, 3‐bromopyruvate, C646, CDDP, chloroform, citrate, cladribine, cytidine, 2‐deoxyglucose, deoxyadenosine, deoxycytidine deoxyguanosine, dibutylamine acetate concentrate (DBAA), dimethyl α‐ketoglutarate, EBSS, FK866, 5‐fluorouracil, gemcitabine, glucose, glutamate, D‐glutamine, l‐glutamine, glutathione reduced ethyl ester, guanosine, histidine, leucine, MEDICA 16, methanol, methoxyamine, mevastatin, N‐acetyl‐l‐cysteine, necrostatin‐1, nicotinamide, N‐methyl‐N‐(trimethylsilyl)trifluoroacetamide (MSTFA), nocodazole, N‐tert‐butyldimethylsilyl‐N‐methyltrifluoroacetamid (MSTBFA), O‐ethylhydroxylamine hydrochloride, oligomycin, paclitaxel, perhexiline, putrescine, pydoxine, pyridoxal‐5‐phosphate, rotenone, salicylate, simvastatin, spermidine, thymidine, TOFA, uridine, and valine. 3‐methylpyruvate was purchased from FLUKA. Veliparib (ABT‐888), BMN 673, and clofarabine were purchased from Selleckchem. Compound 968 (C968) was purchased from Calbiochem‐Merck. Z‐val‐ala‐asp(Ome)‐fluoromethylketone (Z‐VAD‐fmk) was purchased from BACHEM.

RNA interference

The siRNA heteroduplexes specific for Bak (sense 5′‐CCGACGCUAUGACUCAGAGdTdT), Bax (sense 5′‐GGUGCCGGAACUGAUCAGAdTdT), Mcl‐1 (sense 5′‐GUGCCUUUGUGGCUAAACAdTdT), p53 (sense 5′‐GCAUGAACCGGAGGCCCAU dTdT‐3′; Martinez et al, 2002), PUMA (sense 5′‐GGAUGGCGGACGACCUCAAdTdT), Glutaminase (siGLS; sense 5′‐CUGAAUAUGUGCAUCGAUAdTdT), and one nontargeting siRNA (UNR, sense 5′‐GCCGGUAUGCCGGUUAAGUdTdT‐3′) were purchased from Sigma‐Proligo. A second siRNA specific for GLS (GLSB) was purchased from Qiagen (SI04243148 FlexiTubegen solution, Qiagen). A549 cells pre‐seeded in 12‐well plates at 20,000 cells per well were transfected with siRNAs after 30 h using Hiperfect transfecting agent (Qiagen). Cells were treated with EBSS 36 h after transfection, during 24 h.

Cytofluorometry

To measure apoptotic features, adherent and non‐adherent cells were collected and co‐stained for 30 min at 37°C with 40 nM 3,3′ dihexiloxalocarbocyanine iodide (DiOC6(3)), Molecular Probes‐Invitrogen), a mitochondrial transmembrane potential‐sensitive dye, and 1 μg/ml propidium iodide (PI), which only accumulates in dead cells exhibiting plasma membrane rupture. Cytofluorometric acquisitions were carried out on a Milteny cytofluorometer (MACSQuant® Analyzer 10), and statistical analyses were performed by using the FlowJo software (LLC, Oregon, USA) upon gating on events exhibiting normal forward scatter (FSC) and side scatter (SSC) parameters.

Immunoblotting

Cells were trypsinized, collected, washed twice with cold PBS, and lysed in a buffer containing 50 mM Tris–HCl pH 6.8, glycerol 10%, 2% SDS, 10 mM DTT, and 0.005% bromophenol blue. Subsequently, protein extracts (30 μg/lane) were separated on precast 4–12% SDS–PAGE gels (Invitrogen) followed by electrotransfer to nitrocellulose membranes (Biorad) and immunoblotting with primary antibodies targeting PAR (Clone 10H, mAb to Poly(ADP‐ribose Abcam, 1:1,000) or glutaminase (GLS; SAB2105954, Sigma‐Aldrich, 1:1,000). An antibody, which recognizes actin (mAb to beta actin, ab 49900, Abcam, 1:5,000), was used to monitor equal lane loading. Thereafter, membranes were incubated with appropriate horseradish peroxides‐conjugated secondary antibodies (Southern Biotech), followed by chemiluminescence detection with the ECLTM Prime Western Blotting Detection Reagent (GE Healthcare), before being revealed by the ImageQuantTM LAS 4000 Biomolecular Imager (GE Healthcare Life Sciences). Finally, protein expression was quantified by ImageJ software (NIH, USA).

Mouse housing and experiments

Mice were maintained in specific pathogen‐free conditions, at 25°C, with 12‐h light/12‐h dark cycles. All animals were used under an approved protocol by the local Ethics Committee (C2EA 26 no E‐94‐076‐11, protocol no 1113 and C2EA 05 no B‐75‐06‐12, protocol no 7810) under conditions in accordance with the EU Directive 63/2010. Eight‐week‐old female nude athymic (nu/nu) mice were purchased from Envigo France. Sample sizes were calculated to detect a statistically significant effect. For tumor growth experiments, 5 × 106 WT and CDDP‐resistant R4 A549 cells were injected subcutaneously. The estimation of the tumor surface (longest dimension × perpendicular dimension) was measured using a common caliper. When the tumor surface reached 30–40 mm2, mice were randomized into the different groups to be treated (by starvation or drugs). The investigator was blinded during the tumor size measurement.

Starvation regimen in vivo

After randomization, 8‐week‐old female nude athymic (nu/nu) mice were either kept in standard conditions (food and water ad libitum), or left for 24 h in the absence of nutrients (though with ad libitum access to drinking water) two times a week. Mice weight was routinely monitored, and nutrients absence was stopped if weight loss was superior to 20%.

Drug treatment in vivo

After randomization, mice were treated intraperitoneally either with 20 or 30 mg/kg 5‐fluorouracil (5‐FU) in a mix of 200 μl PBS containing 3% DMSO, or with 200 μl PBS containing 3% DMSO alone. Mice were sacrificed when tumor reached 2 cm2.

Sample preparation for metabolome analysis

WT, R2, and R4 A549 cells were seeded in 6‐well plates and cultured for 48 h in complete medium. Ten hours before extraction, medium was changed and cells were cultured either in complete medium (CTL) or nutrient‐deprived medium (NF), in the absence or presence of 2 mM glutamine (NF + GLN). Five replicates per condition. Subsequently, cells were washed six times with cold PBS and then scraped in 500 μl of methanol (90%)‐water (10%). After a centrifugation (10,000 g, 10 min, 4°C), 100 μl chloroform was added, and a second centrifugation was performed (10,000 g, 10 min, 4°C). The whole supernatant was evaporated at 40°C to obtain dried extracts. 300 μl of methanol was added on dried extract and split in two 150 μl fractions for GC‐MS and LC‐MS analyses, respectively. For GC‐MS assay, methanol solubilized aliquots were transferred to glass tubes and solvent was evaporated. 50 μl of methoxyamine (20 mg/ml in pyridine) was added on dried extracts and then stored at room temperature in dark, during 16 h. The day after, 80 μl of MSTFA was added and final derivatization occurred during 30 min at 40°C. Samples were then transferred to vials and directly injected into GC‐MS. After a second evaporation round, LC‐MS dried extracts were solubilized with 300 μl of MilliQ water, centrifuged (10 min at 15,000 g, 4°C), and aliquoted in three microcentrifuge tubes (100 μl). Aliquots were transferred in UHPLC vials and injected into the UHPLC/MS or kept at −80°C until injection.

Plasma preparation for metabolome analysis

A volume of 50 μl of plasma was mixed with 500 μl of a cold solvent mixture (MeOH/Water/Chloroform, 9/1/1, −20°C) and then vortexed and centrifuged (10 min at 15,000 g, 4°C). Then upper phase of the supernatant was split in two parts: 220 μl for the GC/MS experiment and 200 μl for the UHPLC/MS experimentations. Concerning the GC‐MS aliquots, 30 μl from each sample was pooled in a QC vial, and then, 150 μl of samples was transferred in vial injection and evaporated. 50 μl of methoxyamine (20 mg/ml in pyridine) was added on dried extracts and then stored at room temperature in dark, during 16 h. The day after, 80 μl of MSTFA was added and final derivatization occurred at 40°C during 30 min. Samples were then directly injected into GC‐MS. Concerning the LC‐MS aliquots, the collected supernatant was evaporated at 40°C in a pneumatically assisted concentrator (Techne DB3, Staffordshire, UK). The LC‐MS dried extracts were solubilized with 450 μl of MilliQ water. After picked up 60 μl from each microtubes to create pool of QC, samples were aliquoted (100 μl) for LC methods and backup. Biological samples and QC aliquots were kept at −80°C until injection or transferred in vials for direct analysis by UHPLC/MS.

Untargeted analysis of intracellular metabolites by ultra‐high‐performance liquid chromatography (UHPLC) coupled to a quadrupole–time of flight (QTOF) mass spectrometer

Profiling of intracellular metabolites was performed on a Liquid Chromatography (LC) 1260 System (Agilent Technologies, Waldbronn, Germany) coupled to a QTOF 6520 (Agilent Technologies) equipped with an electrospray source operating in both positive and negative mode and full scan mode from 50 to 1,000 Da. The gas temperature was set to 350°C with a gas flow of 12 l/min. The capillary voltage was set to 3.5 kV and the fragmentor at 120 V. Two reference masses were used to maintain the mass accuracy during analysis: m/z 121.050873 and m/z 922.009798 in positive mode and m/z 112.985587 and m/z 980.016375 in negative mode. 10 μl of sample was injected on a SB‐Aq column (100 × 2.1 mm particle size 1.8 μm) from Agilent Technologies, protected by a guard column XDB‐C18 (5 × 2.1 mm particle size 1.8 μm), and heated at 40°C. The gradient mobile phase consisted of water with 0.2% of acetic acid (A) and acetonitrile (B). The flow rate was set to 0.3 ml/min. Initial condition is 98% phase A and 2% phase B. Molecules were then eluted using a gradient from 2 to 95% phase B in 7 min. The column was washed using 95% mobile phase B for 3 min and equilibrated using 2% mobile phase B for 3 min. The autosampler was kept at 4°C. Data processing was performed using in‐house script to align molecular features found by the Agilent MassHunter qualitative software (B.07.00).

Targeted analysis of intracellular metabolites by ultra‐high‐performance liquid chromatography (UHPLC) coupled to a triple quadrupole (QQQ) mass spectrometer

Targeted analysis was performed on a LC 1260 System (Agilent Technologies, Waldbronn, Germany) coupled to a Triple Quadrupole 6410 (Agilent Technologies) equipped with an electrospray source operating in positive mode. The gas temperature was set to 350°C with a gas flow of 12 l/min. The capillary voltage was set to 3.5 kV. 10 μl of sample were injected on a Zorbax Eclipse Plus C18 column (100 × 2.1 mm particle size 1.8 μm) from Agilent technologies, protected by a guard column XDB‐C18 (5 × 2.1 mm particle size 1.8 μm), and heated at 40°C. The gradient mobile phase consisted of 2 mM of dibutylamine ammonium acetate (DBAA) in water (A) and acetonitrile (B). The flow rate was set to 0.2 ml/min, with the gradient as follows: initial condition was 90% phase A and 10% phase B, maintained during 4 min, from 10 to 95% phase B over 3 min, 95% mobile phase B for 3 min, and finally 10% mobile phase B for 3 min. The autosampler was kept at 4°C. Peak detection and integration were performed using the Agilent MassHunter quantitative software (B.07.01).

Targeted analysis of intracellular metabolites gas chromatography (GC) coupled to a triple quadrupole (QQQ) mass spectrometer

The GC‐MS/MS method was performed on a 7890A gas chromatography (Agilent Technologies, Waldbronn, Germany) coupled to a triple quadrupole 7000C (Agilent Technologies, Waldbronn, Germany) equipped with an electronic impact source (EI) operating in positive mode. The injection was performed in splitless mode with a front inlet temperature set to 250°C. The transfer line and the ion‐source temperature were, respectively, at 250 and 230°C. The septum purge flow was fixed at 3 ml/min. The purge flow set to split vent and operated at 80 ml/min during 1 min. Gas saver mode was set to 15 ml/min after 5 min. The helium gas flowed at 1 ml/min through the column (J&WScientificHP‐5MS, 30 m × 0.25 mm, i.d. 0.25 mm, d.f., Agilent Technologies Inc.). Column temperature was held at 60°C for 1 min, then raised to 210°C (10°C/min), followed by a step to 230°C (5°C/min), and reached 325°C (15°C/min), and be held at this temperature for 5 min. The collision gas was nitrogen. Peak detection and integration were performed using the Agilent MassHunter software (B.07.01). Data were presented in hitmaps generated with Gene E software, Broad Institute, Cambridge, USA.

Statistical procedures of in vitro experiments

Unless otherwise specified, all experiments were conducted in duplicate and independently repeated at least three times, yielding comparable results. No statistical methods were used to predetermine sample size. For in vitro studies, data were analyzed with Microsoft Excel (Microsoft Co.) and statistical significance was assessed by means of unpaired Student's t‐test except for Fig EV3A (paired). P‐values were considered significant when lower than 0.05. In the experiments in which the effect meets the criterion for significance in either direction, a two‐sided t‐test was used (Figs 1A, 3A, 5E, 6A–D, and EV3A). In all other experiments, as we expected the effect to be in a given direction, one‐sided t‐tests were applied.

Statistical procedures of in vivo experiments

Longitudinal analyses of tumor growth data were carried out by linear mixed‐effect modeling on tumor sizes. Wald tests were used to compute P‐values by testing jointly that both tumor growth slopes and intercepts were the same between treatment groups of interest. For graphing, tumor growth data are represented in group‐averaged tumor size alongside its SEM at each time point. Survival data are represented in Kaplan–Meier survival curves. Log‐rank test was used to compute P‐values.

Data availability

Raw data from metabolomic experiments were deposited on figshare (https://figshare.com/s/3994153f2df6e8a7f7f0).

Author contributions

FO, MC, SL, AJ, and VA performed the experiments. SD and AC performed mass spectrometry and data analysis. GSW provided TOV‐112D cell lines; FO, SD, and MC analyzed and interpreted the data; JM, JP, and FP reviewed and edited the initial draft; FO, MC and GK designed the study and wrote the paper.

Conflict of interest

The authors declare that they have no conflict of interest.

Supporting information

Expanded View Figures PDF

Click here for additional data file. (366.8KB, pdf)

Source Data for Expanded View

Click here for additional data file. (6.4MB, zip)

Review Process File

Click here for additional data file. (547.7KB, pdf)

Source Data for Figure 1C

Click here for additional data file. (2.7MB, pdf)

Acknowledgements

The authors would like to thank David Enot (Gustave Roussy Cancer Campus; Villejuif, France) for metabolomics and statistical analyses. GK is supported by the Ligue contre le Cancer (équipe labelisée); Agence National de la Recherche (ANR)—Projets blancs; ANR under the frame of E‐Rare‐2, the ERA‐Net for Research on Rare Diseases; Association pour la recherche sur le cancer (ARC); Cancéropôle Ile‐de‐France; Institut National du Cancer (INCa); Institut Universitaire de France; Fondation pour la Recherche Médicale (FDM20140630126 and FDM 40739); the European Commission (ArtForce); the European Research Council (ERC); the LeDucq Foundation; the LabEx Immuno‐Oncology; the RHU Torino Lumière, the SIRIC Stratified Oncology Cell DNA Repair and Tumor Immune Elimination (SOCRATE); the SIRIC Cancer Research and Personalized Medicine (CARPEM); and the Paris Alliance of Cancer Research Institutes (PACRI). GSW is supported by National Institute of Health (NIH Grant R01 CA174949).

The EMBO Journal (2018) 37: e98597

Contributor Information

Maria Castedo, Email: marie.castedo-delrieu@gustaveroussy.fr.

Guido Kroemer, Email: kroemer@orange.fr.

References

  1. Altman BJ, Stine ZE, Dang CV (2016) From Krebs to clinic: glutamine metabolism to cancer therapy. Nat Rev Cancer 16: 619–634 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Aye Y, Li M, Long MJ, Weiss RS (2015) Ribonucleotide reductase and cancer: biological mechanisms and targeted therapies. Oncogene 34: 2011–2021 [DOI] [PubMed] [Google Scholar]
  3. Brown KK, Spinelli JB, Asara JM, Toker A (2017) Adaptive reprogramming of de novo pyrimidine synthesis is a metabolic vulnerability in triple‐negative breast cancer. Cancer Discov 7: 391–399 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chon J, Stover PJ, Field MS (2017) Targeting nuclear thymidylate biosynthesis. Mol Aspects Med 53: 48–56 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Comella P, Filippelli G, De Cataldis G, Massidda B, Frasci G, Maiorino L, Putzu C, Mancarella S, Palmeri S, Cioffi R, Roselli M, Buzzi F, Milia V, Gambardella A, Natale D, Bianco M, Ghiani M, Masullo P, Southern Italy Cooperative Oncology Group (2007) Efficacy of the combination of cisplatin with either gemcitabine and vinorelbine or gemcitabine and paclitaxel in the treatment of locally advanced or metastatic non‐small‐cell lung cancer: a phase III randomised trial of the Southern Italy Cooperative Oncology Group (SICOG 0101). Ann Oncol 18: 324–330 [DOI] [PubMed] [Google Scholar]
  6. Decker DA, Drelichman A, Jacobs J, Hoschner J, Kinzie J, Loh JJ, Weaver A, Al‐Sarraf M (1983) Adjuvant chemotherapy with cis‐diamminodichloroplatinum II and 120‐hour infusion 5‐fluorouracil in Stage III and IV squamous cell carcinoma of the head and neck. Cancer 51: 1353–1355 [DOI] [PubMed] [Google Scholar]
  7. Erster S, Mihara M, Kim RH, Petrenko O, Moll UM (2004) In vivo mitochondrial p53 translocation triggers a rapid first wave of cell death in response to DNA damage that can precede p53 target gene activation. Mol Cell Biol 24: 6728–6741 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Galluzzi L, Senovilla L, Vitale I, Michels J, Martins I, Kepp O, Castedo M, Kroemer G (2012a) Molecular mechanisms of cisplatin resistance. Oncogene 31: 1869–1883 [DOI] [PubMed] [Google Scholar]
  9. Galluzzi L, Vitale I, Senovilla L, Olaussen KA, Pinna G, Eisenberg T, Goubar A, Martins I, Michels J, Kratassiouk G, Carmona‐Gutierrez D, Scoazec M, Vacchelli E, Schlemmer F, Kepp O, Shen S, Tailler M, Niso‐Santano M, Morselli E, Criollo A et al (2012b) Prognostic impact of vitamin B6 metabolism in lung cancer. Cell Rep 2: 257–269 [DOI] [PubMed] [Google Scholar]
  10. Galluzzi L, Vitale I, Michels J, Brenner C, Szabadkai G, Harel‐Bellan A, Castedo M, Kroemer G (2014) Systems biology of cisplatin resistance: past, present and future. Cell Death Dis 5: e1257 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Guo JY, Teng X, Laddha SV, Ma S, Van Nostrand SC, Yang Y, Khor S, Chan CS, Rabinowitz JD, White E (2016) Autophagy provides metabolic substrates to maintain energy charge and nucleotide pools in Ras‐driven lung cancer cells. Genes Dev 30: 1704–1717 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hall MD, Okabe M, Shen DW, Liang XJ, Gottesman MM (2008) The role of cellular accumulation in determining sensitivity to platinum‐based chemotherapy. Annu Rev Pharmacol Toxicol 48: 495–535 [DOI] [PubMed] [Google Scholar]
  13. Heinemann V, Quietzsch D, Gieseler F, Gonnermann M, Schonekas H, Rost A, Neuhaus H, Haag C, Clemens M, Heinrich B, Vehling‐Kaiser U, Fuchs M, Fleckenstein D, Gesierich W, Uthgenannt D, Einsele H, Holstege A, Hinke A, Schalhorn A, Wilkowski R (2006) Randomized phase III trial of gemcitabine plus cisplatin compared with gemcitabine alone in advanced pancreatic cancer. J Clin Oncol 24: 3946–3952 [DOI] [PubMed] [Google Scholar]
  14. Huang D, Duan H, Huang H, Tong X, Han Y, Ru G, Qu L, Shou C, Zhao Z (2016) Cisplatin resistance in gastric cancer cells is associated with HER2 upregulation‐induced epithelial‐mesenchymal transition. Sci Rep 6: 20502 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hudson CD, Savadelis A, Nagaraj AB, Joseph P, Avril S, DiFeo A, Avril N (2016) Altered glutamine metabolism in platinum resistant ovarian cancer. Oncotarget 7: 41637–41649 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Ishida S, McCormick F, Smith‐McCune K, Hanahan D (2010) Enhancing tumor‐specific uptake of the anticancer drug cisplatin with a copper chelator. Cancer Cell 17: 574–583 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Karekla E, Liao WJ, Sharp B, Pugh J, Reid H, Quesne JL, Moore D, Pritchard C, MacFarlane M, Pringle JH (2017) Ex vivo explant cultures of non‐small cell lung carcinoma enable evaluation of primary tumor responses to anticancer therapy. Cancer Res 77: 2029–2039 [DOI] [PubMed] [Google Scholar]
  18. Kelland L (2007) The resurgence of platinum‐based cancer chemotherapy. Nat Rev Cancer 7: 573–584 [DOI] [PubMed] [Google Scholar]
  19. Kozopas KM, Yang T, Buchan HL, Zhou P, Craig RW (1993) MCL1, a gene expressed in programmed myeloid cell differentiation, has sequence similarity to BCL2. Proc Natl Acad Sci USA 90: 3516–3520 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Lee C, Raffaghello L, Brandhorst S, Safdie FM, Bianchi G, Martin‐Montalvo A, Pistoia V, Wei M, Hwang S, Merlino A, Emionite L, de Cabo R, Longo VD (2012) Fasting cycles retard growth of tumors and sensitize a range of cancer cell types to chemotherapy. Sci Transl Med 4: 124ra27 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Leung AW, Hung SS, Backstrom I, Ricaurte D, Kwok B, Poon S, McKinney S, Segovia R, Rawji J, Qadir MA, Aparicio S, Stirling PC, Steidl C, Bally MB (2016) Combined use of gene expression modeling and siRNA screening identifies genes and pathways which enhance the activity of cisplatin when added at no effect levels to non‐small cell lung cancer cells in vitro . PLoS One 11: e0150675 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Li Q, Zhan M, Chen W, Zhao B, Yang K, Yang J, Yi J, Huang Q, Mohan M, Hou Z, Wang J (2016) Phenylethyl isothiocyanate reverses cisplatin resistance in biliary tract cancer cells via glutathionylation‐dependent degradation of Mcl‐1. Oncotarget 7: 10271–10282 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Marino G, Pietrocola F, Eisenberg T, Kong Y, Malik SA, Andryushkova A, Schroeder S, Pendl T, Harger A, Niso‐Santano M, Zamzami N, Scoazec M, Durand S, Enot DP, Fernandez AF, Martins I, Kepp O, Senovilla L, Bauvy C, Morselli E et al (2014) Regulation of autophagy by cytosolic acetyl‐coenzyme A. Mol Cell 53: 710–725 [DOI] [PubMed] [Google Scholar]
  24. Martinez LA, Naguibneva I, Lehrmann H, Vervisch A, Tchenio T, Lozano G, Harel‐Bellan A (2002) Synthetic small inhibiting RNAs: efficient tools to inactivate oncogenic mutations and restore p53 pathways. Proc Natl Acad Sci USA 99: 14849–14854 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Martins I, Kepp O, Schlemmer F, Adjemian S, Tailler M, Shen S, Michaud M, Menger L, Gdoura A, Tajeddine N, Tesniere A, Zitvogel L, Kroemer G (2011) Restoration of the immunogenicity of cisplatin‐induced cancer cell death by endoplasmic reticulum stress. Oncogene 30: 1147–1158 [DOI] [PubMed] [Google Scholar]
  26. Matassa DS, Amoroso MR, Lu H, Avolio R, Arzeni D, Procaccini C, Faicchia D, Maddalena F, Simeon V, Agliarulo I, Zanini E, Mazzoccoli C, Recchi C, Stronach E, Marone G, Gabra H, Matarese G, Landriscina M, Esposito F (2016) Oxidative metabolism drives inflammation‐induced platinum resistance in human ovarian cancer. Cell Death Differ 23: 1542–1554 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. van Meerbeeck JP, Gaafar R, Manegold C, Van Klaveren RJ, Van Marck EA, Vincent M, Legrand C, Bottomley A, Debruyne C, Giaccone G, European Organisation for Research and Treatment of Cancer Lung Cancer Group , National Cancer Institute of Canada (2005) Randomized phase III study of cisplatin with or without raltitrexed in patients with malignant pleural mesothelioma: an intergroup study of the European Organisation for Research and Treatment of Cancer Lung Cancer Group and the National Cancer Institute of Canada. J Clin Oncol 23: 6881–6889 [DOI] [PubMed] [Google Scholar]
  28. Meisner HM, Sorensen L (1966) Metaphase arrest of Chinese hamster cells with rotenone. Exp Cell Res 42: 291–295 [DOI] [PubMed] [Google Scholar]
  29. Michels J, Vitale I, Galluzzi L, Adam J, Olaussen KA, Kepp O, Senovilla L, Talhaoui I, Guegan J, Enot DP, Talbot M, Robin A, Girard P, Orear C, Lissa D, Sukkurwala AQ, Garcia P, Behnam‐Motlagh P, Kohno K, Wu GS et al (2013) Cisplatin resistance associated with PARP hyperactivation. Cancer Res 73: 2271–2280 [DOI] [PubMed] [Google Scholar]
  30. Michels J, Obrist F, Castedo M, Vitale I, Kroemer G (2014a) PARP and other prospective targets for poisoning cancer cell metabolism. Biochem Pharmacol 92: 164–171 [DOI] [PubMed] [Google Scholar]
  31. Michels J, Obrist F, Vitale I, Lissa D, Garcia P, Behnam‐Motlagh P, Kohno K, Wu GS, Brenner C, Castedo M, Kroemer G (2014b) MCL‐1 dependency of cisplatin‐resistant cancer cells. Biochem Pharmacol 92: 55–61 [DOI] [PubMed] [Google Scholar]
  32. Michels J, Adam J, Goubar A, Obrist F, Damotte D, Robin A, Alifano M, Vitale I, Olaussen KA, Girard P, Cremer I, Castedo M, Soria JC, Kroemer G (2015) Negative prognostic value of high levels of intracellular poly(ADP‐ribose) in non‐small cell lung cancer. Ann Oncol 26: 2470–2477 [DOI] [PubMed] [Google Scholar]
  33. Oburoglu L, Tardito S, Fritz V, de Barros SC, Merida P, Craveiro M, Mamede J, Cretenet G, Mongellaz C, An X, Klysz D, Touhami J, Boyer‐Clavel M, Battini JL, Dardalhon V, Zimmermann VS, Mohandas N, Gottlieb E, Sitbon M, Kinet S et al (2014) Glucose and glutamine metabolism regulate human hematopoietic stem cell lineage specification. Cell Stem Cell 15: 169–184 [DOI] [PubMed] [Google Scholar]
  34. Penning TD, Zhu GD, Gandhi VB, Gong J, Liu X, Shi Y, Klinghofer V, Johnson EF, Donawho CK, Frost DJ, Bontcheva‐Diaz V, Bouska JJ, Osterling DJ, Olson AM, Marsh KC, Luo Y, Giranda VL (2009) Discovery of the Poly(ADP‐ribose) polymerase (PARP) inhibitor 2‐[(R)‐2‐methylpyrrolidin‐2‐yl]‐1H‐benzimidazole‐4‐carboxamide (ABT‐888) for the treatment of cancer. J Med Chem 52: 514–523 [DOI] [PubMed] [Google Scholar]
  35. Qian X, Xu W, Xu J, Shi Q, Li J, Weng Y, Jiang Z, Feng L, Wang X, Zhou J, Jin H (2017) Enolase 1 stimulates glycolysis to promote chemoresistance in gastric cancer. Oncotarget 18: 47691–47708 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Ray Chaudhuri A, Callen E, Ding X, Gogola E, Duarte AA, Lee JE, Wong N, Lafarga V, Calvo JA, Panzarino NJ, John S, Day A, Crespo AV, Shen B, Starnes LM, de Ruiter JR, Daniel JA, Konstantinopoulos PA, Cortez D, Cantor SB et al (2016) Replication fork stability confers chemoresistance in BRCA‐deficient cells. Nature 535: 382–387 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Shen Y, Rehman FL, Feng Y, Boshuizen J, Bajrami I, Elliott R, Wang B, Lord CJ, Post LE, Ashworth A (2013) BMN 673, a novel and highly potent PARP1/2 inhibitor for the treatment of human cancers with DNA repair deficiency. Clin Cancer Res 19: 5003–5015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Sklar MD, Prochownik EV (1991) Modulation of cis‐platinum resistance in Friend erythroleukemia cells by c‐myc. Cancer Res 51: 2118–2123 [PubMed] [Google Scholar]
  39. Sourisseau T, Helissey C, Lefebvre C, Ponsonnailles F, Malka‐Mahieu H, Olaussen KA, Andre F, Vagner S, Soria JC (2016) Translational regulation of the mRNA encoding the ubiquitin peptidase USP1 involved in the DNA damage response as a determinant of Cisplatin resistance. Cell Cycle 15: 295–302 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Tajeddine N, Galluzzi L, Kepp O, Hangen E, Morselli E, Senovilla L, Araujo N, Pinna G, Larochette N, Zamzami N, Modjtahedi N, Harel‐Bellan A, Kroemer G (2008) Hierarchical involvement of Bak, VDAC1 and Bax in cisplatin‐induced cell death. Oncogene 27: 4221–4232 [DOI] [PubMed] [Google Scholar]
  41. Tardito S, Oudin A, Ahmed SU, Fack F, Keunen O, Zheng L, Miletic H, Sakariassen PO, Weinstock A, Wagner A, Lindsay SL, Hock AK, Barnett SC, Ruppin E, Morkve SH, Lund‐Johansen M, Chalmers AJ, Bjerkvig R, Niclou SP, Gottlieb E (2015) Glutamine synthetase activity fuels nucleotide biosynthesis and supports growth of glutamine‐restricted glioblastoma. Nat Cell Biol 17: 1556–1568 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Tieche CC, Peng RW, Dorn P, Froment L, Schmid RA, Marti TM (2016) Prolonged pemetrexed pretreatment augments persistence of cisplatin‐induced DNA damage and eliminates resistant lung cancer stem‐like cells associated with EMT. BMC Cancer 16: 125 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Wang D, Lippard SJ (2005) Cellular processing of platinum anticancer drugs. Nat Rev Drug Discov 4: 307–320 [DOI] [PubMed] [Google Scholar]
  44. Wangpaichitr M, Wu C, Li YY, Nguyen DJM, Kandemir H, Shah S, Chen S, Feun LG, Prince JS, Kuo MT, Savaraj N (2017) Exploiting ROS and metabolic differences to kill cisplatin resistant lung cancer. Oncotarget 8: 49275–49292 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Winter C, Albers P (2011) Testicular germ cell tumors: pathogenesis, diagnosis and treatment. Nat Rev Endocrinol 7: 43–53 [DOI] [PubMed] [Google Scholar]
  46. Yuneva M, Zamboni N, Oefner P, Sachidanandam R, Lazebnik Y (2007) Deficiency in glutamine but not glucose induces MYC‐dependent apoptosis in human cells. J Cell Biol 178: 93–105 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Yuneva MO, Fan TW, Allen TD, Higashi RM, Ferraris DV, Tsukamoto T, Mates JM, Alonso FJ, Wang C, Seo Y, Chen X, Bishop JM (2012) The metabolic profile of tumors depends on both the responsible genetic lesion and tissue type. Cell Metab 15: 157–170 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Zhang J, Pavlova NN, Thompson CB (2017) Cancer cell metabolism: the essential role of the nonessential amino acid, glutamine. EMBO J 36: 1302–1315 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Expanded View Figures PDF

Click here for additional data file. (366.8KB, pdf)

Source Data for Expanded View

Click here for additional data file. (6.4MB, zip)

Review Process File

Click here for additional data file. (547.7KB, pdf)

Source Data for Figure 1C

Click here for additional data file. (2.7MB, pdf)

Data Availability Statement

Raw data from metabolomic experiments were deposited on figshare (https://figshare.com/s/3994153f2df6e8a7f7f0).

Articles from The EMBO Journal are provided here courtesy of Nature Publishing Group

RESOURCES