Glutaminase inhibition with telaglenastat (CB-839) improves treatment response in combination with ionizing radiation in head and neck squamous cell carcinoma models

Abstract

The efficacy of ionizing radiation (IR) for head and neck cancer squamous cell carcinoma (HNSCC) is limited by poorly understood mechanisms of adaptive radioresistance. Elevated glutaminase gene expression is linked to significantly reduced survival (p < 0.03). The glutaminase inhibitor, telaglenastat (CB-839), has been tested in Phase I/II cancer trials and is well tolerated by patients. This study investigated if telaglenastat enhances the cellular response to IR in HNSCC models. Using three human HNSCC cell lines and two xenograft mouse models, we examined telaglenastat’s effects on radiation sensitivity. IR and telaglenastat combinatorial treatment reduced cell survival (p ≤ 0.05), spheroid size (p ≤ 0.0001) and tumor growth in CAL-27 xenograft bearing mice relative to vehicle (p ≤ 0.01), telaglenastat (p ≤ 0.05) or IR (p ≤ 0.01) monotherapy. Telaglenastat significantly reduced the Oxygen Consumption Rate/Extracellular Acidification Rate ratio in CAL-27 and HN5 cells in the presence of glucose and glutamine (p ≤ 0.0001). Telaglenastat increased oxidative stress and DNA damage in irradiated CAL-27 cells. These data suggest that combination treatment with IR and telaglenastat leads to an enhanced anti-tumor response. This pre-clinical data, combined with the established safety of telaglenastat justifies further investigation for the combination in HNSCC patients.

1. Introduction

Survival remains poor for head and neck cancer (HNC) patients. Although, 40% of HNC patients receive curative ionizing radiation (IR) therapy, failure within the radiation field approaches 50% with resultant morbidity and mortality [1,2]. Multimodality treatment including the use of concurrent chemotherapy with radiation is used to treat many HNCs. However, the use of systemic therapy lacks a clear molecular basis and is often administered based on empiric evidence, contributing to limited efficacy and increased toxicity [3–5]. Radiotherapy dose escalation is limited by the need to protect surrounding healthy tissues and current options for FDA-approved radiosensitizers are limited. Even though an increasing variety of intelligently designed, gene-targeted drugs are in or are entering clinical use, many are transient in their activity with the majority of patients recurring after a short period of benefit [6]. Therefore, there is increased need to understand and over-come mechanisms of radiation resistance to improve patient survival.

Radiation induces programmed cell survival and metabolic responses within the cell [7,8], representing potential mechanisms of adaptive resistance that reduce radiation efficacy and are consequently interesting targets for combination therapy. Preliminary Reverse Phase Protein Array analysis by our laboratory identified activation of metabolic pathways upon radiation treatment (data not shown). Based on these findings as well as increasing evidence that glutamine plays an important role in DNA damage repair, and the availability of a targeted inhibitor already in Phase I/II trials, we explore the role of elevated glutaminase as a potential mechanism of adaptive resistance to radiation therapy [9–11].

Glutaminase is the enzyme responsible for the conversion of glutamine to glutamate. Glutamate is subsequently converted to α-ketoglutarate, a key component of the Krebs cycle. Notably, glutaminase overexpression has been linked to increased metastasis [12]. If increased glutaminase is independently associated with worse tumor control, and radiation induces glutaminase levels or activity, then we hypothesize that inhibition of glutaminase reduces substrate availability for the Krebs cycle, decreases aerobic respiration, and potentially reduces cellular proliferation.

Inhibition of glutaminase by telaglenastat (Calithera Biosciences, Inc.) has been associated with reduced growth of lymphomas, breast, renal and pancreatic cancers in pre-clinical studies [12]. In Phase II clinical trials for renal cell and triple-negative breast cancer in combination with chemotherapy, telaglenastat exhibited increased bio-stability and bioavailability as compared to other glutaminase inhibitors, and was well tolerated by patients [12,13]. Notably, neither the role of glutamine in HNC pathogenesis nor the efficacy of telaglenastat in combination with radiation has been previously evaluated.

Using The Cancer Genome Atlas’s (TCGA) transcriptome database, we identified that increased glutaminase gene expression was associated with reduced survival in HNSCC patients. As this association supports glutaminase as an important drug target in the treatment of HNSCC, we examined if the combination of glutaminase inhibitor, telaglenastat, and IR is more effective than monotherapy. Clonogenic assays revealed that combinatorial treatment decreased cell survival in CAL-27 and HN5 cell lines. Using 2 heterotopic HNSCC xenograft models, we identified that the combination of telaglenastat and IR reduced tumor volume relative to monotherapy. Telaglenastat also increased IR induced oxidative stress and DNA damage. In summary, our finding that the addition of telaglenastat significantly improves radiation treatment response in HNSCC provides preclinical data in support of future clinical trials.

2. Materials and methods

2.1. Cell culture

FaDu (pharynx) cells were grown in Gibco high-glucose MEM (10% FBS, 6 mM glutamine, 1% penicillin/streptomycin, 1% Gibco MEM amino acids, 1% sodium pyruvate). HN5 (tongue) and CAL-27 (tongue) were grown in DMEM (10% FBS, 6 mM l-glutamine, 1% penicillin/ streptomycin, 1% Gibco MEM amino acids and 1% sodium pyruvate). Cell lines were tested for mycoplasma every 3 months and authenticity was validated by STR.

2.2. Glutaminase activity assay

CAL-27 cells were treated with 0 or 10 Gy using a GammaCell cesium irradiator. Cell lysates were collected after 5 min and glutaminase activity was assessed using the PicoProbe^™ Glutaminase Activity Assay (BioVision, K455).

2.3. Survival analysis using TCGA transcriptome data

Messenger RNA (mRNA) expression data for HNSCC primary tumor tissues were obtained from TCGA using the illuminahiseq_rnaseqv2-RSEM_genes_normalized data set through TCGA biolinks package in R statistical software as published previously [14,15]. Data from 515 HNSCC tumor samples and associated, de-identified clinical data were used to assess survival over 5 years (Table S1). Survival was compared between patients with high (top quartile, n = 129) and low (bottom quartile, n = 129) glutaminase mRNA expression.

2.4. Proliferation assay

CAL-27, HN5, and FaDu cells were plated into 96 well plates to adhere overnight. Media was then changed to Gibco phenol-free DMEM (5 mM glucose, 1% penicillin/streptomycin, 1% Gibco MEM amino acids, 1% sodium pyruvate, 5% FBS) with either 0 or 0.5 mM glutamine. Cells were grown for 48 h. Proliferation was measured using the Dojindo CCK-8 kit (Dojindo, CK04) following the manufacturer’s instructions. Proliferation was compared using Student’s t-test.

2.5. Clonogenic assay to determine LD50 of telaglenastat and ionizing radiation

Clonogenic assay, colony counting, and survival fraction was assessed as previously described to determine the LD50 of IR and telaglenastat [16]. Cells were seeded into tissue culture plates. Cells were irradiated using a GammaCell cesium cell irradiator at 0–10 Gy in a single fraction or cells were grown in the presence of 0–1000 nM telaglenastat (CB-839 (telaglenastat), Cayman Chemical, 22038). Colonies were grown for 11–14 days until they could be easily observed without microscopy.

2.6. Three-dimensional (3D) spheroid generation and measurement

Ice-cold 24 well plates were coated with Corning RGF Matrigel and allowed to set at 37 °C, 5% CO₂ for 15–20 min. DMEM or MEM media containing 5% FBS, 1% penicillin/streptomycin, 1% sodium pyruvate, 1% GIBCO MEM amino acids and 1% Matrigel was prepared cold and warmed to room temperature. Cells were resuspended into this media and plated onto the Matrigel coated wells. Immediately after plating, vehicle or drug was added. Cells were grown at 37 °C, 5% CO₂.

Spheroids were imaged at 4× magnification after 48 h. Using Figi ImageJ, images were converted to binary and particle analysis was performed to measure spheroid area. Spheroid size was compared using boxplot analysis and Student’s t-test. For IF, spheroids were fixed after 8 days and stained with DAPI.

2.7. Seahorse MitoStress test assay

Cell lines were plated onto Agilent Seahorse tissue culture plates and grown overnight. The following morning, media was removed, and wells were washed with Agilent Seahorse DMEM assay buffer containing 1% sodium pyruvate, and 1% GIBCO MEM amino acids. Media was replaced with indicated variants. Cells were equilibrated for 1 h at 37 °C with no CO₂.

Drugs loaded into the cartridge were DMSO or 1 μM telaglenastat, oligomycin, 2-[2-[4-(trifluoromethoxy) phenyl]hydrazinylidene] propanedinitrile (FCCP), and antimycin/rotenone. Seahorse XF measurement cycles had a 2 min mix, 1 min wait and then a 2 min measurement period. Basal respiration measurements were repeated for 6 cycles before vehicle or telaglenastat injection. After drug injection, measurements were recorded for 16 cycles before oligomycin injection. After 6 cycles, FCCP was injected and another 6 cycles of measurements were recorded. Finally, antimycin and rotenone were injected, and a final 6 cycles of measurements were recorded. Oxygen Consumption Rate (OCR)/Extracellular Cellular Acidification Rate (ECAR) ratio was calculated based on Agilent Wave data and plotted using R statistical software.

2.8. In vivo experiments with mouse xenograft models

With institutional review board approval at the University of Cincinnati (IACUC 18-11-08-01) and the Cincinnati Veteran’s Affairs Medical Center (ACORP 18-11-08-01), two xenograft mouse models were used to examine the effect of radiation and telaglenastat on tumor growth. We used the established HN5 xenograft model using 4–6 week old athymic Swiss nude (Experimental Oncology Animal Colony, MD Anderson Cancer Center; total 40 mice) [17,18]. Following overt tumor development, mice were randomly divided into 4 treatment groups (n = 8–9/group): vehicle, radiation with vehicle, telaglenastat alone, and radiation with telaglenastat. Vehicle for telaglenastat was 25% (w/v) hydroxypropyl-b-cyclodextrin in 10 mmol/L citrate, pH 2. Telaglenastat was suspended at 20 mg/ml. Mice were administered vehicle or telaglenastat at a dose of 200 mg/kg telaglenastat per gavage [12]. Starting two days prior to irradiation, mice were gavaged twice daily (8 a.m. and 3 p.m.) with vehicle or telaglenastat. After 48 h, mice were treated with 0 or 2 Gy radiation to the involved flank for 5 consecutive days with an ortho-voltage irradiator, with fractionated schedules that parallel schedules used for patients with HNSCC. Gavages continued for 14 days. Mice were euthanized at day 27 due to multiple vehicle mice having to be sacrificed due to tumor growth. A CAL-27 xenograft model using 5 week old homozygous J:Nu athymic nude mice (Jackson Laboratories, 007850, RRID:IMSR_JAX:007850; total of 60 female mice) was used with the same treatment paradigm. Tumors were allowed to grow to sufficient size before being randomly divided into the 4 previously described treatment groups (n = 14–15/group). Two days prior to irradiation, mice were gavaged every 12 h with the same dosages used with the HN5 xenograft mice. Gavaging continued throughout the duration of the experiment. After 48 h, the tumor tissue was then exposed to fractionated radiation in 2 Gy fractions (1 Gy/min at 320 kVp), using x-rays from an Xstrahl XenX Series Biological Irradiator on 5 consecutive days for a total dose of 10 Gy. Tumors were measured with precision calipers several times a week and normalized volumes were calculated using the formula [volume=(width²*length)/2].

2.9. Immunofluorescence

CAL-27 cells were seeded into chamber slides and allowed to attach. Vehicle, 10 μM telaglenastat, and/or 10 μM N-acetylcysteine (NAC) were added. After a 2 h incubation, cells were irradiated at 8 Gy. After 30 min, cells were fixed and probed for γ-H2AX (Cell Signaling Technology, 9718S) and 8-oxoguanine (R&D Systems, 4354-MC-050). Representative images were obtained using confocal microscopy. Fluorescence was scored in a blinded fashion (scale of 0–3) in each cell and averaged.

2.10. Statistical analysis

Statistical analysis was performed in R v.3.6.1 or Graphpad Prism 9 software. Kaplan-Meier survival curves were generated using the survminer and ggplot2 packages [19,20]. Significance was determined as log-rank for survival analysis. In vitro experiments were performed ≥3 independent times and with technical replicates. The LD50 for telaglenastat was determined using 4-parameter log regression with drc and ggplot2 packages [19,21]. The LD50 for IR was determined using glm with MASS, tibble and ggplot2 packages as radiation doses increased in a linear fashion [22–24]. Differences in tumor volume sizes were assessed using the Kruskal-Wallis statistical test. Comparisons were performed between treatment groups for a given day. Statistical significance was determined as p ≤ 0.05.

3. Results

To establish the role of glutaminase in HNSCC, we first queried the transcriptome database from TCGA. High glutaminase expression (upper quartile) versus low glutaminase expression (lower quartile) was associated with reduced patient survival (p = 0.03) (Fig. 1).

Fig. 1.

Patients with increased glutaminase mRNA levels in their tumor had significantly reduced survival as compared to patients with lower levels of glutaminase. Kaplan Meier survival analysis was used to compare survival between low mRNA expression (grey dotted line) defined as the bottom quartile and high mRNA expression (black solid line) defined as the top quartile values. Log-rank analysis was used to determine statistically significance. Data was obtained from TCGA transcriptome database.

As HNSCC represents 90% of HNC cases, this study utilized 3 established human-derived HNSCC lines (CAL-27, HN5, and FaDu) to explore methods for reducing radiation resistance and response to combinations of radiation and glutaminase inhibition. All 3 lines were derived from p16 negative tumors as these non-virally driven tumors are relatively radioresistant compared to tumors that are associated with human papillomavirus [25,26]. As glutaminase expression is associated with significantly reduced survival of HNSCC patients and IR is a standard treatment of HNCs, we first examined if glutaminase activity was induced in HNSCC cells treated with IR. Glutaminase activity was assessed in CAL-27 cells treated with or without IR. Within 5 min of CAL-27 cells being treated with 10 Gy IR, glutaminase activity significantly increased with an average increase of 13.2% relative to mock irradiated samples (p ≤ 0.01) (Fig. S1).

We next established the role of glutamine-driven cellular proliferation in HNSCC. CAL-27, HN5 and FaDu cells had significantly higher proliferation in glutamine containing media versus media with no glutamine (p ≤ 0.0001)(Fig. 2A), suggesting that all 3 cell lines are at least partially dependent on exogenous glutamine for proliferation.

Fig. 2.

(A) Glutamine dependence: Proliferation of CAL-27, and HN5 cell lines is glutamine dependent. Cell lines were grown with or without 0.5 mM glutamine. Error bars represent SD. (B) Efficacy of telaglenastat in inhibiting spheroid growth: Telaglenastat significantly reduced the size of CAL-27 and HN5 spheroids after 48 h. (C) Representative images of CAL-27, HN5 and FaDu spheroids at 8 days. HNC cells were grown in 3D culture and treated with vehicle or 1 μM CB-839 for 8 days. Scale bars represent 100 μM. Error bars represent SE (***p ≤ 0.001, ****p ≤ 0.0001).

While versatile, 2D monolayers do not mimic the tumor microenvironment that is represented in 3D culture systems [27–29]. As such, CAL-27, HN5, and FaDu were treated with vehicle or telaglenastat and 3D tumor spheroids sizes were measured after 48 h. Average CAL-27 (p ≤ 0.0001) and HN5 (p ≤ 0.001) spheroid size was reduced in the presence of telaglenastat. FaDu spheroids were slightly larger in the presence of telaglenastat (p ≤ 0.05) (Fig. 2B and C).

To examine the functional effects of telaglenastat treatment on the Krebs cycle, the Seahorse MitoStress test was used to assess inhibition of aerobic metabolism. CAL-27, HN5, FaDu cells were grown in media containing both 5 mM glucose and 0.5 mM glutamine, 5 mM glucose alone or 0.5 mM glutamine alone. During the MitoStress test, either vehicle or 1000 nM telaglenastat was injected after basal measurements were taken. The oxygen consumption ratio (OCR), an index of aerobic respiration, and the extracellular acidification rate (ECAR), an index of anaerobic respiration were evaluated. The OCR/ECAR during maximal respiration was calculated for each cell line and treatment condition to compare changes to the basal cell energy state. Telaglenastat significantly reduced the OCR/ECAR ratio relative to vehicle in CAL-27 and HN5 cells (p ≤ 0.0001) in the presence of both glucose and glutamine (Fig. 3). The reduction in OCR/ECAR ratio demonstrates that glutaminase inhibition with telaglenastat shifts the cells away from aerobic respiration involving the Krebs cycle. In control samples involving media containing glucose or glutamine alone, there was no significant difference between vehicle or telaglenastat. In FaDu cells, there was no significant difference in the OCR/ECAR ratio observed in either media containing both glucose and glutamine, glucose alone, or glutamine alone (Fig. S2).

Fig. 3.

Telaglenastat decreases OCR/ECAR ratio in HN5 and CAL-27 cells. OCR/ECAR ratios were calculated from Seahorse MitoStress test analysis in CAL-27, HN5 and FaDu cells. Vehicle or 1000 nM telaglenastat were injected during the MitoStress test into media containing 5 mM glucose and 0.5 mM glutamine. Error bars represent SE (****p ≤ 0.0001).

To ensure sufficient cell population remained to identify combined efficacy of telaglenastat and IR, sub-lethal doses of the independent treatment of telaglenastat and IR were determined using clonogenic cell survival assays. The LD50 for telaglenastat was 10.9 nM in CAL-27, 13.7 nM in HN5 and 7.4 nM in FaDu. For further experiments, 10 nM telaglenastat was selected to treat all cell lines unless otherwise noted. The LD50 for IR was 4.7 Gy in CAL-27, 4.2 Gy HN5 and 4.4 Gy in FaDu cells. For further experiments, 4 Gy was selected for in vitro experiments (Fig. S3).

Sub-lethal doses of IR and telaglenastat were set at their respective LD50s. Clonogenic cell survival assays were used to examine if combination of IR and telaglenastat was more effective in killing HNSCC cell lines relative to monotherapy. CAL-27, HN5 and FaDu cells were treated with vehicle, IR, telaglenastat, or IR with telaglenastat. Treatment groups were normalized to the mock irradiated vehicle controls. For CAL-27, (Fig. 4), combination treatment significantly reduced cell survival relative to vehicle (p ≤ 0.001), telaglenastat alone (p ≤ 0.0001), and radiation alone (p ≤ 0.0001). For HN5 cells, combinatorial treatment in HN5 cells significantly reduced survival relative to vehicle (p ≤0.001), telaglenastat alone (p ≤ 0.05), radiation alone (p ≤ 0.0001). In FaDu cells, combinatorial treatment as compared to telaglenastat led to significantly reduced survival (p ≤ 0.01). No significant difference was observed between radiation alone and combination treatment in FaDu cells.

Fig. 4.

Combination of telaglenastat and IR significantly reduces survival of HNSCC cell lines as compared to independent treatment with either telaglenastat or IR. (A) CAL-27 (Kruskal-Wallis: p =5.1e-07), (B) HN5 (Kruskal-Wallis: p =1.8e-06) or (C) FaDu (Kruskal-Wallis: p =1.4e-04) cells were treated with either 0 or 4 Gy IR and grown in the presence of vehicle or 10 nM telaglenastat. Error bars represent SD (*p ≤0.05,**p ≤0.01, ***p ≤0.001, ****p ≤0.0001).

To extend the in vitro findings, two heterotopic in vivo xenograft mouse models were used to test the combination treatment of telaglenastat and radiation: HN5 and CAL-27 xenografts models in athymic nude mice. Animals were treated with 200 mg/kg twice a day, as this regimen was shown to fully inhibit glutaminase and inhibit tumor growth in other heterotopic models [30]. In the HN5 xenograft model, the greatest differences in tumor volume were observed on day 25. Radiation alone, telaglenastat alone and combinatorial treated tumors were 74.3%, 94.9% and 61.7% the size of vehicle treated tumors (p =0.19, 0.021, 0.28; respectively). Combination treatment significantly reduced tumor volume compared to telaglenastat alone (p = 0.01) (Fig. S4). There were no significant differences between the combination treatment with vehicle or radiation treated mice. In the CAL-27 xenograft model, mice receiving the combinatorial treatment averaged tumors 1/3 the size of tumors in vehicle treated mice (p = 0.0011, day 9) (Fig. 5A). Combinatorial treatment was significantly more effective than radiation or telaglenastat alone (p = 0.00019, p = 0.0331; respectively). By day 19, there was no significant tumor volume differences between radiation and vehicle treated mice with radiation treated mice having tumors only 60% the size of vehicle treated mice (Fig. 5B). There were also no significant differences in tumor sizes between telaglenastat and vehicle treated mice, where telaglenastat mice tumors were 91% the size of vehicle treated mice. Mice treated with combinatorial therapy had significantly reduced tumor volume relative to vehicle, radiation alone, and telaglenastat alone (p = 0.0041, p = 0.01, p = 0.043, day 19).

Fig. 5.

Combination of IR and telaglenstat resulted in significantly reduced tumor volume relative to vehicle or independent treatment. Striped and solid black arrows indicate the start and stop of telaglenstat administration, respectively. (A) Representative images of CAL-27 xenograft tumors (B) Normalized volumes of CAL-27 xenograft tumors. P values represent significant different between combinatorial treatment versus independent treatment or vehicle. P values were from comparing groups to combinatorial treatment mice on a given day. Error bars represent SE (*p < 0.05,**p < 0.01,***p < 0.001,****p < 0.0001).

To examine possible mechanisms for reduced tumor cell viability and tumor growth, we examined the role of telaglenastat in increasing oxidative stress (assessed via 8-oxoguanine) and DNA damage (assessed by γ-H2AX) (Fig. 6A). In CAL-27 cells, oxidative stress was significantly increased when telaglenastat was combined with radiation relative to telaglenastat or radiation monotherapy (p ≤ 0.01, p ≤ 0.05; respectively) (Fig. 6B). Addition of antioxidant, NAC, reduced 8-oxoguanine to levels observed in monotherapy treated cells. Combination of telaglenastat and radiation resulted in significantly higher levels of γ-H2AX relative to telaglenastat monotherapy (p ≤ 0.0001)(Fig. 6C).

Fig. 6.

(A) Immunofluorescent images of CAL-27 treated with vehicle, telaglenastat and radiation monotherapy, combination of telaglenastat and radiation, and combination treatment with NAC; 8-oxoguanine (green), γ-H2AX (red), DAPI (blue): nuclei. (B) Comparison of 8-oxoguanine and (C) γ-H2AX levels. Scale bars equal 50 μM. Error bars represent 95% confidence intervals (*p ≤0.05, **p ≤0.01, ***p ≤0.001, ****p ≤0.0001). Quantification results are from a representative biological replicate.

4. Discussion

The primary goal of this study was to examine the role of glutaminase inhibition as a means to address adaptive resistance to IR in head and neck cancer. The combination of glutaminase inhibition with telaglenastat and IR was compared to the effects of independent treatment alone in clonogenic assays and in two xenograft models. We first demonstrate that radiation increased glutaminase activity in glutamine dependent cells within 5 min (Fig. S1). This increase could be used as a mean to exploit therapeutic vulnerabilities of HNSCC cells in combination with IR. Importantly for translation to the clinic, combination of telaglenastat and IR significantly reduced cell survival in CAL-27, and HN5 relative to other treatment groups. In the CAL-27 xenograft mouse models, combination of telaglenastat and IR significantly reduced tumor volume as compared to vehicle or independent treatment alone (Fig. 5B). Together, these results support further investigation into the clinical utility of telaglenastat with concurrent radiation in the treatment of HNSCC.

All 3 HNSCC cell lines (CAL-27, HN5 and FaDu) examined were found to be dependent on glutamine for proliferation (Fig. 2A). Moreover, inhibiting glutaminase with telaglenastat significantly reduced HNSCC spheroid growth (Fig. 2B) and functionally telaglenastat significantly reduced aerobic respiration (Fig. 3). In addition to direct DNA strand breakage, IR induces DNA damage and cell death through formation of ROS [31,32]. Telaglenastat in combination with radiation significantly increases this oxidative stress in CAL-27 cells and increases DNA damage (Fig. 6B and C). Thus, telaglenastat-mediated glutaminase inhibition increases ROS and further sensitizes cancer cells to IR. Based on this data, we posit that IR induces glutaminase activity in HNSCC cells and that telaglenastat impairs the formation of glutamate and subsequently slows cell metabolism and enhances oxidative stress, leading to increased death in combination with radiation (Fig. 7).

Fig. 7.

IR induces glutaminase activity in HNSCC and thereby increases production of glutamate for entry into the Krebs cycle. Glutaminase inhibitor, telaglenastat, reduces formation of glutamate, and thus impairs the Krebs cycle’s ability to create biomolecules necessary for rapid cellular proliferation.

Of note, in vitro assays indicated no significant decrease in spheroid growth, OCR/ECAR ratio, or cell survival with combinatorial treatment in FaDu cells. Interestingly, FaDu cells contain a defect in the Fanconi Anemia (FA) gene. This FA gene mutation is not present in either CAL-27, HN5 or many other cells [33]. Lymphoblast cells with the FA defect demonstrate impaired glutaminolysis [34]. Compared to HN5 and CAL-27, FaDu cells are less glutamine dependent and this may explain the differences observed in the response to telaglenastat. It is plausible that the FA defect in FaDu cells leads to impaired glutaminolysis, making telaglenastat less efficacious. An additional limitation of this study is that our cell lines demonstrated a partial dependence on extracellular glutamine, however it is unclear to what extent this dependence exists in vivo [35]. Future planned experiments using patient derived xenograft animal models, will elucidate this prior to moving into human clinical trials. Despite this, this study supports that telaglenastat significantly improves response to IR in CAL-27 and HN5 cells, which are more representative of HNSCC.

The efficacy of head and neck radiotherapy is plagued by treatment resistance, with 2 year local failure rates of 50% with primary radiation and 70–76% with re-irradiation [2,36]. Adaptive resistance to radiotherapy is a major contributor to treatment failure and yet, options for “radiosensitizers” are limited. Multiple groups have associated molecular changes to the administration of radiation itself, including upregulation of metabolic pathways [8,37].

Glutaminase is essential for the conversion of glutamine to glutamate, a substrate of the Krebs cycle. Pertinent to the pre-clinical studies herein, telaglenastat is already in Phase I/II trials with a favorable safety profile. Exploratory analyses using TCGA transcriptome data identified that increased glutaminase gene expression is associated with significantly reduced patient survival (Fig. 1). The reduced survival associated with increased glutaminase supports glutaminase as a potential drug treatment target. Clinical trials have been conducted in patients with solid tumors including renal cell and triple negative breast cancer using telaglenastat in combination with conventional chemotherapy [9,10]. However, there has been no examination of telaglenastat in HNSCC. Moreover, there are few studies examining the combined effects of telaglenastat and radiation, the latter of which is a standard treatment option for HNSCC.

Cisplatin is the mainstay of systemic therapy given concurrently with radiation therapy for HNSCC, however, many HNSCC patients are medically ineligible. With current trials including NRG HN-004 (NCT03258554), seeking to identify alternative systemic therapies, and the safety of telaglenastat already well-established, there is exciting potential to directly translate these findings into a phase II trial in the clinic, as these studies represent a true mechanism to counter adaptive radioresistance.