Glutamine deprivation counteracts hypoxia-induced chemoresistance

Abstract

The microenvironment of solid tumors is a key determinant of therapy efficacy. The co-occurrence of oxygen and nutrient deprivation is a common phenomenon of the tumor microenvironment and associated with treatment resistance. Cholangiocarcinoma (CCA) is characterized by a very poor prognosis and pronounced chemoresistance. A better understanding of the underlying molecular mechanisms is urgently needed to improve therapy strategies against CCA. We sought to investigate the importance of the conditionally essential amino acid glutamine, a centrally important nutrient for a variety of solid tumors, for CCA. Glutamine levels were strongly decreased in CCA samples and the growth of established human CCA cell lines was highly dependent on glutamine. Using gradual reduction of external glutamine, we generated derivatives of CCA cell lines which were able to grow without external glutamine (termed glutamine-depleted (GD)). To analyze the effects of coincident oxygen and glutamine deprivation, GD cells were treated with cisplatin or gemcitabine under normoxia and hypoxia. Strikingly, the well-established phenomenon of hypoxia-induced chemoresistance was completely reversed in GD cells. In order to better understand the underlying mechanisms, we focused on the oncogene c-Myc. The combination of cisplatin and hypoxia led to sustained c-Myc protein expression in wildtype cells. In contrast, c-Myc expression was reduced in response to the combinatorial treatment in GD cells, suggesting a functional importance of c-Myc in the process of hypoxia-induced chemoresistance. In summary, these findings indicate that the mechanisms driving adaption to tumor microenvironmental changes and their relevance for the response to therapy are more complex than expected.

Figure 1

Human CCA patients display glutamine-poor regions inside the tumor. Specimen were analyzed for the spatial distribution of glutamine (Gln [M + TAHS]x + : (323.1714 m/z) in tumor and adjacent benign tissue by using high-resolution MALDI-FTICR-MSI and were hematoxylin and eosin (H&E)-stained. The color scale represents the absolute intensity of detected ions. A patient with a portal vein embolization (PVE) was used as control specimen (Ctrl).

Figure 2

eCCA cells can overcome glutamine addiction by gradual reduction of external glutamine concentrations. A) Timeline of gradual glutamine depletion of EGI-1 and TFK-1 cells. The concentration of available glutamine in growth medium was monthly decreased by 50% to a final concentration of 0 mM. Cell lines obtained in this selection process are referred to as EGI-1 GD and TFK-1 GD and were standardly cultivated in glutamine-deprived growth medium. B) EGI-1 GD and TFK-1 GD show proliferation close to their parental counterparts. eCCA cells were seeded in regular glutamine-supplied growth medium and medium lacking glutamine. eCCA GD cells were seeded in medium lacking glutamine. Cells were cultivated for 7d and cell numbers were counted at the indicated time points using Trypan Blue staining. Color-coded significances placed above the time points represent the statistical comparison of growth curves to the reference (eCCA cells in glutamine-deprived medium). C) eCCA GD variants show morphological traits close to their parental counterparts. eCCA GD cell lines were seeded as described in Figure 2B. Bright field microscopy images were recorded after 72 h cultivation (200x magnification). Data represent mean ± SD of three individual experiments performed in triplicate.

Figure 3

Glutamine availability broadly affects mitochondrial respiration and proliferative capacity in TFK-1 cell line variants. A) Bioenergetic profiles of eCCA GD cell lines and their parental counterparts were analyzed depending on glutamine availability by using an extracellular flux analyzer. Cells were seeded overnight in their regular growth medium. One hour prior to measurement the growth medium was changed to the indicated assay medium variant. Mitochondrial stress test was used to measure oxygen consumption rate (OCR, left). Data are presented as mean + SD of six replicates per group. B) Quantification of respiration parameters is presented as mean ± SD of six replicates per group. C) Re-supplementation of glutamine increases proliferation of TFK-1 GD cells. eCCA GD cells were cultivated in glutamine-deprived or -supplemented growth medium for 96 h. Live cells were determined via Trypan Blue exclusion. Data are presented as mean ± SD of two individual experiments performed in triplicate. * p < 0.05, ** p < 0.01 and *** p < 0.001. One-Way ANOVA with Bonferroni post hoc test.

Figure 4

eCCA GD variants show varying sensitivity to chemotherapy. Dose-response curves of cisplatin and gemcitabine were prepared for eCCA and eCCA GD cell lines under normoxic and hypoxic conditions. Medium was changed to drug-supplemented medium 24 h after seeding. After 72 h drug exposure, viable cells were counted using Trypan Blue staining. Data are presented as mean of two replicates per time point, error bars were excluded for the purpose of clarity.

Figure 5

The adaption process to glutamine deprivation eliminates hypoxia-induced chemoresistance to cisplatin. A) Cisplatin-induced DNA damage is reduced under hypoxic conditions in eCCA cell lines but increased in eCCA GD cell lines. Cells were cultivated overnight either under normoxic (N) or hypoxic (H, 1% O₂) conditions and were treated with cisplatin [500 nM] or gemcitabine [10 nM] for 72 h. Data are shown as representative immunoblots of two or three independent experiments with detection of ATM, γH2AX and Tubulin as loading control. B) Quantification of relative ATM and γH2AX protein levels in cisplatin-treated eCCA cell line variants presented in Figure 5A. Data are shown as mean ± SD of two or three independent experiments. C) Hypoxia leads to a rescue of proliferation in cisplatin-treated eCCA cell lines but not eCCA GD variants. Cells were treated as described in A. Viable cells were determined via Trypan Blue exclusion. Data are presented as mean ± SD of two individual experiments performed in triplicate. * p < 0.05, ** p < 0.01 and *** p < 0.001. One-Way ANOVA with Bonferroni post hoc test.

Figure 6

The adaption process to glutamine withdrawal affects the response to hypoxia. A) eCCA GD cell lines show a slightly reduced sensitivity to hypoxia when compared to their parental counterparts. Cells were cultivated in their regular growth medium under normoxia or hypoxia (1% O₂) for 96 h. Cell numbers were counted daily using Trypan Blue staining. Results are shown as means ± SD of two individual experiments performed in triplicate. B) eCCA GD cell lines show increased hypoxia-induced HIF-2α stabilization in response to cytostatic drug treatment. Cells were treated as described in Figure 4A. Data are shown as representative immunoblots of two individual experiments with detection of HIF-1α, HIF-2α and YY1 as loading control. *** p < 0.001. One-Way ANOVA with Bonferroni post hoc test.

Figure 7

The adaption process to glutamine withdrawal determines the cellular regulation of c-Myc. A) c-Myc protein levels are decreased in response to glutamine deprivation in eCCA cell lines. EGI-1 and TFK-1 cells lines were cultivated with different glutamine concentrations for 24 h. Data are shown as immunoblots with detection of c-Myc and Tubulin as loading control. B) c-Myc levels of eCCA GD variants are adjusted to the basal level of their nourished parental counterparts. Densitometric analysis of relative c-Myc levels revealed no difference between the groups. Cells were cultivated under control conditions for 24 h. Data are shown as representative immunoblots of two individual experiments with detection of c-Myc and Tubulin as loading control. C) Cisplatin rescues c-Myc protein levels under hypoxia in eCCA cell lines but not in eCCA GD cell line variants. Cells were cultivated overnight either under normoxic or hypoxic (1% O₂) conditions and were treated with the established IC50 dose for cisplatin or gemcitabine for 72 h. Data are shown as representative immunoblots of two individual experiments with detection of c-Myc and Tubulin as loading control.

Figure 8

c-Myc potentially contributes to survival in parental CCA cell lines by sustaining cell growth and proliferation. Schematic summarization of the interaction between glutamine (Gln) withdrawal, chemotherapy and hypoxia, and their effect on c-Myc protein expression. HIC = hypoxia-induced chemoresistance.