Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Apr 1.
Published in final edited form as: Cancer Res. 2024 Oct 1;84(19):3223–3234. doi: 10.1158/0008-5472.CAN-24-0254

Therapeutic Targeting of the GLS1-c-Myc Positive Feedback Loop Suppresses Glutaminolysis and Inhibits Progression of Head and Neck Cancer

Jianqiang Yang 1,2, Fanghui Chen 1,2, Liwei Lang 3, Fan Yang 1,2, Zhenzhen Fu 1,2, Juan Martinez 4, Amber Cho 1, Nabil F Saba 1,2, Yong Teng 1,2,3,5,*
PMCID: PMC11444885  NIHMSID: NIHMS2011829  PMID: 39024547

Abstract

Head and neck squamous cell carcinoma (HNSCC) is addicted to glutaminolysis. Targeting this metabolic dependency has emerged as a potential therapeutic approach for HNSCC. Here, we conducted a bioinformatic analysis of the TCGA HNSCC cohort that revealed a robust correlation between expression of MYC (encoding the protein c-Myc) and GLS1, which catalyzes the first step in glutaminolysis. Intriguingly, disruption of GLS1 signaling in HNSCC cells by genetic depletion or CB-839 treatment resulted in a reduction in c-Myc protein stability via a USP1-dependent ubiquitin-proteasome pathway. On the other hand, c-Myc directly binds to the promoter region of GLS1 and upregulates its transcription. Notably, the GLS1-c-Myc pathway enhanced ACC-dependent SLUG acetylation, prompting cancer cell invasion and metastasis. Thus, the GLS1-c-Myc axis emerged as a positive feedback loop critical for driving the aggressiveness of HNSCC. Therapeutically, combining CB-839 with the c-Myc inhibitor MYCi975 strongly suppressed GLS1-c-Myc signaling, resulting in a superior antitumor effect compared to either single agent in an orthotopic mouse model of HNSCC. These findings hold promise for the development of effective therapies for HNSCC patients, addressing an urgent need arising from the significant incidence and high metastatic rate of the disease.

Keywords: GLS1, c-Myc, head and neck cancer, cancer metabolism, positive feedback loop, combination treatment

Introduction

Alterations in cellular metabolism are a hallmark of cancer cells, and targeting dysregulated metabolism in cancer offers a novel strategy for developing more selective therapeutic interventions (1,2). Glutaminolysis is the process in which glutamine, an amino acid, is metabolized to produce glutamate, contributing to various cellular functions such as energy production and biosynthesis of key molecules. Cancer cells often exhibit upregulated glutaminolysis and use this pathway to meet their increased energy and biosynthetic needs. For example, increased glutaminolysis helps maintain cellular redox balance by contributing to the production of reducing equivalents, such as NADPH (3,4). This is critical for protecting cancer cells from oxidative stress and supporting their survival. We have previously shown that head and neck squamous cell carcinoma (HNSCC) cells have a strong addiction to glutamine (5); thus, targeting this pathway has emerged as a promising therapeutic strategy for this disease.

Glutaminase 1 (GLS1) is an amidohydrolase that catalyzes the first step in the conversion of glutamine to glutamate during the process of glutaminolysis (6). GLS1 is located on chromosome 2 and encodes two isoforms, kidney-type glutaminase (KGA) and glutaminase C (GAC), due to differential RNA splicing (7). GAC is more enzymatically efficient and expressed at a higher level than KGA in cancer (8,9). Elevated GLS1 expression has been associated with increased glutaminolysis and advanced stages of cancer and poor prognosis in some cases, suggesting its involvement in malignant progression (10,11). Over the past decade, GLS1 has gained attention as a potential therapeutic target. Inhibitors of GLS1, such as CB-839, have been explored as a strategy to interfere with abnormal glutaminolysis in various types of cancers (5,12,13).

Accumulating evidence has shown that GLS1 expression is influenced by several key oncogenes (e.g., MYC) and tumor suppressors (e.g., Rb) (14,15). c-Myc, encoded by the MYC gene, is a transcription factor that regulates the expression of a wide range of genes involved in cell proliferation, apoptosis, and metabolism (16,17). It is frequently overexpressed or amplified in a variety of human cancers. Overexpression of MYC is associated with an aggressive phenotype, poor prognosis, and resistance to therapy. Dysregulation of c-Myc alters cellular metabolism, leading to increased glycolytic and glutaminolytic activity (18). Recent studies have reported that c-Myc upregulates the expression of GLS1, promoting glutamine metabolism to support malignant progression (19,20). However, whether GLS1 has the potential to control c-Myc signaling in cancer cells and the precise regulatory mechanisms governing the functional interaction between c-Myc and GLS1 remain to be fully elucidated. Additionally, whether c-Myc-GLS1 signaling confers cancer metastasis is largely unknown. In this study, we examined the interaction between c-Myc and GLS1 and the role of this signaling axis in HNSCC progression.

Materials and Methods

Cell lines and culture

HN4, HN6, HN12, HN30, and UMSCC1 (SCC1) cells were a kind gift from Dr. Andrew Yeudall and cultured in DMEM/F-12 medium containing 10% fetal bovine serum (FBS) at 37°C in a humidified incubator supplied with 5% CO2 (21,22). HEK293 cells were purchased from American Type Culture Collection (ATCC) and cultured in DMEM containing 10% FBS. All cell lines were not genetically authenticated but were routinely screened for mycoplasma contamination by MycoAlert Mycoplasma Detection Kit (Lonza) and used for experiments before passage 10.

Reagents, antibodies, constructs, and standard assays

Cycloheximide (CHX), MG132, and 5-(tetradecyloxy)-2-furoic acid (TOFA) were purchased from Sigma-Aldrich (St Louis, MO). Glutamine/Glutamate-Glo Assay Kit was obtained from Promega (Madison, WI). CB-839 was purchased from Selleckchem (Houston, TX) and MYCi975 from Axon Medchem (Reston, VA). pLKO.1-puro TRC control shRNA targeting the gene encoding green fluorescent protein (shGFP) and specific shRNAs targeting the GLS1 gene (shGLS1–1 and shGLS1–2), USP1 gene (shUSP1–1 and shUSP1–2), or MYC gene (shMYC-1 and shMYC-2) were purchased from Horizon Discovery (Waterbeach, UK). ViraPower Lentiviral Packaging Mix containing an optimized mixture of the three packaging plasmids (pLP1, pLP2, and pLP/VSVG), was obtained from Invitrogen (Carlsbad, CA). Full-length human GLS1 and Flag-tagged human Slug (Slug-Flag) expression plasmids were obtained from Sinobiological Inc (Beijing, China). MYC overexpression plasmid and full-length HA-tagged human ubiquitin expression plasmid (HA-Ub) were provided by Dr. Shi-Yong Sun at Emory University. Full-length HA-tagged human USP1 expression plasmid (HA-USP1) was a gift from Dr. Jianmin Zhang at Roswell Park Comprehensive Cancer Center. Full-length human USP1 (USP1-WT) and its enzymatic dead mutant (USP1-C90S) plasmids were a gift from Dr. Tony T. Huang at New York University. All antibodies used are listed in Supplementary Table S1. Plasmid transfection and lentiviral infection, Western blot, glutamine consumption assay, wound healing assay, CHX chase assay, immunoprecipitation (IP), and cellular ubiquitination assays were carried out as we previously described (5,23).

RNA isolation, real-time quantitative PCR (RT-qPCR) and RNA-sequencing (RNA-seq)

Total RNA was extracted from HNSCC cells with RNeasyMini Kit (Qiagen) following the manufacturer’s instructions. cDNA synthesis was carried out using reverse transcriptase and oligo(dT) primer. RT-qPCR was performed on the StepOne Plus Real-Time PCR System (Applied Biosystems) using Applied Biosystems™ Power SYBR™ Green PCR Master Mix (ThermoFisher Scientific). Primers used are listed in Supplementary Table S2. For RNA-seq, Illumina sequencing was carried out by Novogene for each sample. The bowtie parameter mismatch was 2. Clean data were mapped back onto the assembled transcriptome, and a read count for each gene was obtained from the mapping results. Differentially expressed genes (LogFC ≥ |±0.25|, p < 0.05) were analyzed using Gene Ontology (GO) and HALLMARK pathways through Gene Set Enrichment Analysis (GSEA). Detailed RNA-seq information is available in GSE240956 deposited in the NIH Gene Expression Omnibus (GEO) database.

Chromatin immunoprecipitation (ChIP) assay

ChIP assay was performed using a ChIP Assay Kit (ThermoFisher Scientific). Briefly, cells were crosslinked with 1% formaldehyde for 10 minutes, followed by quenching with glycine for 10 minutes and lysed to generate soluble chromatin, which was sonicated with a Sonifier F-100 (Fisher Scientific, USA; 20-second sonication at 30% amplitude and 20-second pulse). Soluble chromatin was centrifuged at 13,800 rpm for 10 minutes, and the supernatant was collected. Immunoprecipitations of soluble chromatin were performed with anti-c-Myc antibody (1:100 dilution) and 20 μL protein A/G beads. The beads were placed in a rotating wheel overnight at 4°C. The immunoprecipitates were washed 3 times and eluted from the beads with elution buffer, and reverse crosslinked with 0.5M NaCl for 4 hours at 65°C. Crosslinked chromatin was treated with RNAse A for 1 hour at 37°C and phenol-extracted, and the eluate was run on an agarose gel to visualize specific DNA-protein complexes. The relative enrichment of specific DNA sequences was measured on the StepOne Plus Real-Time PCR System. Primers used are listed in Supplementary Table S2.

Cell invasion assays

For transwell invasion assay, cells cultivated with DMEM/F12 medium without FBS were put in the upper chamber of 24-well plates coated with Matrigel. DMEM/F12 medium with 20% FBS was put into the lower chamber. After 24 hours of incubation, cells were removed from the upper chamber. The number of cells migrating to the lower chamber was calculated by inverted microscopy after crystal violet staining. The number and mean number of cells in each field represented the invasive ability of the cells. Hanging drop aggregation assay was performed according to our previously published protocol (24). Cell spheroids were observed under an inverted microscope and NIH ImageJ software was used to measure the perimeter of the spheroids in different groups. For 3D spheroid invasion assay, 3 ×103 cells were seeded in each well of low attachment 96-well plates (Corning) and incubated for 3–4 days to form spheroids. Spheroids were embedded in Matrigel and cultured in a dish. After Matrigel solidified, culture medium was added for 2 days. Microscopic images were taken at day 0 and day 2, and the invasion area was quantified by outlining the spheroid perimeter using NIH ImageJ software. The change in invasion area was determined by subtracting the day 0 spheroid area from the day 2 spheroid area.

Assessment of drug synergism

The combination index (CI) was calculated by CompuSyn software using CI equation algorithms. CI<1 indicates additive effect, CI =1 indicates synergistic effect, and CI >1 indicates antagonistic effect.

Primary tissue specimens and immunohistochemistry (IHC)

Formalin-fixed, paraffin-embedded (FFPE) tissues from HNSCC patients were obtained from the Head and Neck Satellite Tissue Bank of Emory University. All clinical specimens were obtained with written informed consent from the patients. The studies were conducted in accordance with recognized ethical guidelines (Declaration of Helsinki), and that the studies were approved by Emory Institutional Review Board. Tissue sections were deparaffinized with xylene, rehydrated through a graded alcohol series, and incubated in 3% hydrogen peroxide. Sections were placed in 10 mM sodium citrate buffer (pH 6.0) at sub-boiling temperatures for 10 min and incubated with 10% normal goat serum, followed by incubation with the indicated primary antibodies. Immunoreactivity was visualized using the DAB Detection kit (Vector laboratories, Burlingame, CA) and counterstained with hematoxylin. Slides were dehydrated, mounted, and scanned using the Olympus Nanozoomer whole slide scanner (Olympus, Center Valley, PA). The German semi-quantitative scoring method was used to examine the final immunoreactivity score as we described previously (25). Briefly, each specimen was scored for intensity (no staining = 0; weak staining = 1; moderate staining = 2; strong staining = 3) and for extent of stained cells (0% = 0; 1–24% = 1; 25–49% = 2; 50–74% = 3; 75–100% = 4). Signal Index (SI) was the product of the intensity score multiplied by the extent score. For other studies, IHC staining was quantified using Image pro-Plus6.0 software (Media Cybernetics, Silver Springs, MD) and presented as integrated optical density (IOD).

Acetyl-CoA measurement

HNSCC cells with specified gene modifications or treatments were rinsed with cold PBS, scraped, and sonicated. Cellular proteins were precipitated using perchloric acid. The resulting supernatant was collected after centrifugation and neutralized by the addition of potassium bicarbonate until the sample pH fell in the range of 6–8. Precipitates were removed by centrifugation, and the supernatant was used to measure acetyl-CoA concentration in duplicate using the Acetyl-CoA assay kit (Sigma).

Animal studies

Six-week-old male NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) mice were purchased from the Jackson Laboratory (Bar Harbor, ME). All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Emory University and Augusta University. To generate an orthotopic tongue tumor model to evaluate the role of GLS1, 1×105 GLS1 knockdown or control HN12 cells were suspended in 50 μl of PBS/Matrigel (3:1) and injected into the anterior ~ 1/3 tongue of NSG mice under anesthesia (5,26). To generate an orthotopic buccal mucosal tumor model to evaluate the antitumor efficacy of CB-839 and/or MYCi975, 1×106 HN12 cells were implanted into NSG mice by intramucosal injection (5,26). Ten days after cell inoculation, mice were randomized to receive treatment with vehicle, CB-839, MYCi975, or the drug combination. Both MYCi975 and CB-839 were dissolved in 5% DMSO in corn oil. MYCi975 was given by intraperitoneal injection at 100 mg/kg once daily for 15 days, and CB-839 was given by oral gavage at 200 mg/kg once daily for 15 days. Tumor dimensions were measured twice a week with electronic calipers, and tumor volume was calculated by the formula of V = length × width2 × 1/2. The body weight and physical activity of each animal were followed as markers of toxicity. Mice were sacrificed at the experimental endpoint, and tumors and lymph nodes were excised for histopathological and IHC analysis.

Statistical analysis

The mRNA expression of genes from the Cancer Genome Atlas (TCGA) HNSCC cohort was obtained from the GDC Data Portal. To determine the influence of MYC and GLS1 expression on overall survival of HNSCC patients, Kaplan-Meier survival analysis was performed based on patient groups stratified by high and low gene expression in TCGA HNSCC cohort. Data were analyzed using GraphPad Prism 9 (San Diego, CA). Experimental values are expressed as mean ± standard error of the mean (SEM). For comparison between two groups, statistical analysis was performed using unpaired Student’s t-test. One-way analysis of variance (ANOVA) followed by Tukey’s multiple testing correction was used for comparison of more than two groups. p values less than 0.05 were considered statistically significant.

Data availability

The RNA-seq data generated in this study is available in GEO at GSE240956. Public data included in TCGA HNSCC cohort analysis were downloaded from GDC Data Portal (https://portal.gdc.cancer.gov/projects/TCGA-HNSC). All other raw data generated in this study are available upon request from the corresponding author.

Results

Blocking GLS1 results in c-Myc downregulation in HNSCC cells

The Myc family consists of three related human genes: c-Myc (MYC), l-Myc (MYCL), and n-Myc (MYCN) (27). To explore the potential functional interaction between GLS1 and c-Myc in HNSCC cells, we initially assessed their co-expression using gene expression data from TCGA HNSCC cohort. This analysis revealed a strong correlation between MYC and GLS1 expression (R = 0.35, p = 2.6e_13) in this cohort (Fig. 1A). Among 30 HNSCC patient tumor samples collected at Emory University, IHC data showed that tumor cells with high levels of c-Myc highly expressed GLS1, demonstrating their significant positive association in primary tumor samples (R = 0.34, p = 0.02) (Fig. 1B). In mammalian cells, GLS1 encodes two isoforms: KGA and GAC (7). Western blot showed that the GAC isoform exhibited predominant expression across HEK293 and all HNSCC cell lines examined. Conversely, the KGA isoform was barely detected in HN6 and HN12 cells (Supplementary Fig. S1). To focus on GAC, we selected HN6 and HN12 as the primary cell models for subsequent investigations.

Figure 1. Blockade of GLS1 signaling downregulates c-Myc in HNSCC cells.

Figure 1.

Open in a new tab

(A) Positive correlation between GLS1 and MYC expression in TCGA HNSCC cohort. (B) Correlation between GLS1 and c-Myc levels in primary HNSCC tumor samples determined by IHC. SI = positive staining × intensity score. Representative images of co-expression of GLS1 and c-Myc in the same tumor area and quantitative data (n=30) are shown in the left and right panels, respectively. (C) GO HALLMARK enrichment analysis of DEGs resulting from GLS1 knockdown based on RNA-seq data. (D) Downregulated enrichment plot for a priori gene sets for HALLMARK_MYC_TARGET_V2 in GLS knockdown HN6 cells vs control cells expressing shGFP. (E) Effect of GLS1 knockdown on c-Myc protein levels in HN6 and HN12 cells determined by Western blot. (F) Effect of GLS1 overexpression on c-Myc protein levels in HN6 and HN12 cells. (G) Immunostaining of c-Myc in tumor tissues derived from GLS1 knockdown or control HN6 cells. (H) Protein levels of GLS1 and c-Myc treated with various concentrations of CB-839. (I) Immunostaining of c-Myc in HN6 tumor tissues treated with vehicle or CB-839. 200 mg/kg CB-839 was administered by oral gavage twice a day for two weeks. In (G) and (I), representative IHC images and quantitative IOD data (n = 5 mice/group) are shown in the left and right panels, respectively. (J) Protein levels of GLS1 and c-Myc in HN6 and HN12 cells upon glutamine (Gln) deprivation at different time points. EV: empty vector; GLS1 O/E: GLS1 overexpression vector; shGFP: shRNA against the GFP gene as a negative control; shGLS1–1 and shGLS1–2: two specific shRNAs against the GLS1 gene. **p<0.01.

We previously found that GLS1 knockdown led to suppression of cellular glutamine uptake and cell proliferation in HN6 cells (5). RNA-seq analysis revealed that 1669 genes were upregulated and 1909 were downregulated (LogFC ≥ |±0.25|, p < 0.05) in HN6 cells when GLS1 was knocked down (Supplementary Fig. S2). HALLMARK pathway analysis via GSEA for significantly downregulated genes (shGLS1 vs shGFP) revealed robust correlation of GLS1 with Myc-related HALLMARK pathways, including HALLMARK_MYC_TARGETS_V1 (p = 0.08) and HALLMARK_MYC_TARGETS_V2 (p < 0.001) (Fig. 1C and 1D). Western blot confirmed a reduction in c-Myc levels in GLS1 knockdown HN6 and HN12 cells (Fig. 1E). Accordingly, GLS1 overexpression increased c-Myc levels in HN6 and HN12 cells (Fig. 1F). We examined c-Myc levels in tumors derived from GLS1 knockdown and control HN6 cells (5) and found remarkably reduced c-Myc levels in GLS1 knockdown tumors compared with control tumors (Fig. 1G). Moreover, the selective GLS1 inhibitor CB-839 dose-dependently inhibited c-Myc protein expression in HN6 and HN12 cells (Fig. 1H). Significantly decreased c-Myc levels were also observed in HN6 tumors upon CB-839 treatment (Fig. 1I). Western blot further showed that glutamate deprivation impaired GLS1 levels in HN6 and HN12 cells, accompanied by correspondingly decreased c-Myc levels (Fig. 1J), suggesting the regulation of c-Myc by GLS1 in HNSCC cells.

GLS1 prevents c-Myc degradation through USP1-dependent deubiquitination in HNSCC cells

GLS1 knockdown had no noticeable effect on MYC expression at the mRNA level (Supplementary Fig. S3), raising the possibility that GLS1 may regulate c-Myc at the post-translational level. To block de novo protein synthesis, we treated HN6 and HN12 cells with CHX, in the presence or absence of CB-839, and collected cell lysates at different timepoints for Western blot. CB-839 treatment significantly shortened the half-life of c-Myc proteins compared with control cells (Fig. 2A). Moreover, the proteasome inhibitor, MG132, largely rescued the reduction of c-Myc in the presence of CB-839 (Fig. 2B). To determine whether blocking GLS1 signaling could alter the ubiquitination of c-Myc, HN6 and HN12 cells transfected with HA-Ub expression vectors were pretreated with MG132 and subsequently treated with CB-839. IP of anti-c-Myc antibody followed by HA immunoblotting revealed that inhibition of GLS1 increased the fraction of ubiquitinated c-Myc (Fig. 2C). These results support the notion that GLS1 loss destabilizes c-Myc proteins by prompting its proteasomal degradation.​

Figure 2. Loss of GLS1 leads to destabilization of c-Myc proteins in HNSCC cells through proteasomal degradation by suppression of USP1.

Figure 2.

Open in a new tab

(A) Effect of CB839 on the half-life of c-Myc protein in HN6 and HN12 cells determined by CHX chase assays. HN6 and HN12 cells were treated with 50 μg/ml CHX for indicated hours in the presence or absence of 2 μM CB-839. Quantitative data of c-Myc protein levels (n = 3) are shown in the right panel. (B) Effect of MG132 on CB-839-mediated c-Myc protein reduction in HN6 and HN12 cells. HN6 and HN12 cells were treated with 10 μM proteasome inhibitor MG132 for 4 hours and then treated with 2 μM CB-839 for 12 hours. (C) Effect of CB-839 on ubiquitination of c-Myc protein in HN6 and HN12 cells. HN6 and HN12 cells transfected with the HA-Ub plasmid were pretreated with 10 μM MG132 for 4 hours before treatment with 2 μM CB-839 for 12 hours. Cell lysates were immunoprecipitated using anti-c-Myc antibody and immunoblotted with anti-HA antibody. (D) Fold changes in the three most downregulated ubiquitination-related genes (UBE2S, SIAH2, and USP1) in GLS1 knockdown HN6 cells vs control HN6 cells based on RNA-seq data. (E) Expression alterations in UBE2S, SIAH2, and USP1 genes in GLS1 knockdown and control HN12 cells validated by RT-qPCR. (F) Effect of GLS1 knockdown on USP1 protein levels in HN6 and HN12 cells. (G) Effect of USP1 knockdown on c-Myc protein levels in HN6 and HN12 cells. (H) Effect of restoring USP1 expression on GLS1 knockdown-mediated reduction of c-Myc protein levels in HN6 and HN12 cells. (I) Interaction between USP1 and c-Myc in HN6 and HN12 cells transfected with the HA-USP1 plasmid. Cell lysates were collected for IP with anti-HA antibody, followed by Western blot with anti-c-Myc antibody. Preimmune IgG was used as a control. (J) Effect of USP1 knockdown on ubiquitination of c-Myc protein in HN6 and HN12 cells. HN6 and HN12 cells transfected with the HA-Ub plasmid were pretreated with 10 μM MG132 for 4 hours before cell lysates were collected for IP with anti-c-Myc antibody and immunoblotted with anti-HA antibody. (K) Effect of overexpression of USP1 and USP1 enzymatic dead mutant (C90S) on c-Myc protein levels in HN6 and HN12 cells. (L) Effect of overexpression of USP1 and USP1 enzymatic dead mutant (C90S) on ubiquitination of c-Myc protein in HN6 and HN12 cells. (M) Effect of restoring USP1 expression on GLS1 knockdown-mediated ubiquitination of c-Myc protein in HN6 and HN12 cells. In (L) and (M), HN6 and HN12 cells transfected with HA-Ub were pretreated with 10 μM MG132 for 4 hours before cell lysates were collected for IP with anti-c-Myc antibody and immunoblotted with anti-HA antibody. In (C), (J), (L) and (M), β-Actin immunoblotting on total lysates is shown to normalize the input. shGFP: shRNA against the GFP gene as a negative control; shGLS1: shRNA against the GLS1 gene; shUSP1–1 and shUSP1–2: two specific shRNAs against the USP1 gene; USP1-HA: USP1 overexpression vector. *p<0.05; **p<0.01; ***p<0.001.

Next, we sought to determine the molecular mechanism by which GLS1 mediates deubiquitination of c-Myc. GO analysis of our RNA-seq data showed significant alterations in the molecular function (MF) associated with “Ubiquitin-like protein ligase binding”, “Ubiquitin protein ligase binding” and “Unfolded protein binding” (Supplementary Fig. S4). RNA-seq data further indicated that USP1, SIAH2 and UBE2S genes were the three most downregulated genes encoding ubiquitin ligases upon GLS1 knockdown in HN6 cells (Fig. 2D). Among these three genes, USP1 exhibited the greatest reduction in expression in GLS1 knockdown HN6 cells (Fig. 2E). USP1, one of the best-characterized human deubiquitinases (DUBs) that play a vital role in a cell cellular response to DNA damage (28), was consistently decreased at the protein level in GLS1 knockdown HN6 and HN12 cells compared with their corresponding control cells (Fig. 2F).

Knockdown of USP1 by shRNAs in HN6 and HN12 cells led to a remarkable decrease in c-Myc (Fig. 2G) and restoring USP1 expression in GLS1 knockdown cells by overexpression of USP1-HA rescued c-Myc levels (Fig. 2H). Intriguingly, c-Myc was detected in both endogenous and exogenous USP1 immunocomplexes (Fig. 2I and Supplementary Fig. S5), and an increased fraction of ubiquitinated c-Myc was seen in HA-Ub expressing HN6 and HN12 cells when USP1 was depleted (Fig. 2J). Overexpression of USP1 resulted in elevated c-Myc protein levels, while overexpression of USP1 enzymatic dead mutant led to a decrease in c-Myc levels in both HN6 and HN12 cells (Fig. 2K). Consistently, we detected increased ubiquitination of c-Myc in cells overexpressing USP1 and decreased ubiquitination of c-Myc in cells overexpressing USP1 enzymatic dead mutant (Fig. 2L). Elevated levels of ubiquitinated c-Myc were also observed in GLS1 knockdown HN6 and HN12 cells; however, restoration of USP1 expression reduced the fraction of ubiquitinated c-Myc (Fig. 2M), suggesting that GLS1 inhibits c-Myc protein degradation through USP1-mediated deubiquitination in HNSCC cells.

GLS1 loss suppresses HNSCC cell invasion and metastasis through downregulation of the c-Myc-Slug signaling axis.

Our previous study demonstrated that GLS1 knockdown inhibited HNSCC cell growth (5). Here, we found that GLS1 loss also resulted in a significant reduction in HNSCC cell migration and invasion in both 2D and 3D cultures (Fig. 3A–3C and Supplementary Fig. S6). HN12 cells had higher metastatic potential compared with HN6 cells; thus, GLS1 knockdown and control HN12 cells were individually injected into the anterior tongue of NSG mice to generate an orthotopic tongue tumor model. Four weeks after cell inoculation, mice implanted with GLS1 knockdown HN12 cells exhibited a marked decrease in tumor volume (Fig. 3D) and decreased potential for lymph node metastasis, as evidenced by reductions in lymph node weight and intensity of pan-keratin immunostaining in lymph nodes (Fig. 3E and 3F), compared with mice implanted with knockdown control cells. Since epithelial-mesenchymal transition (EMT) contributes to tumor progression and metastasis, we screened EMT markers by Western blot in HN6 and HN12 cells with or without GLS1 knockdown. We observed increased E-cadherin with decreased N-cadherin and Slug in GLS1 knockdown cells (Fig. 3G), but no noticeable changes in EMT markers TWIST1, ZEB1 and Snail (Fig. 3G). Downregulation of Slug was more pronounced than changes in other EMT markers in GLS1 knockdown cells (Fig. 3G). We thus focused more on this molecule. Decreased Slug levels were also seen in GLS1 knockdown HN12 tumors compared with control tumors (Fig. 3H). We also observed decreased intensity of Slug immunostaining in HN6 tumors treated with CB-839 versus vehicle (Supplementary Fig. S7). Moreover, MYC knockdown reduced Slug protein levels (Fig. 3I) and restoration of MYC expression in GLS1 knockdown cells rescued Slug levels (Fig. 3J), suggesting the involvement of c-Myc in GLS1-mediated Slug downregulation. Most importantly, restoring either MYC or Slug expression in GLS1 knockdown cells led to a partial rescue of decreased cell migration and invasion (Fig. 3K and 3L). In addition, the reintroduction of Slug expression in GLS1 knockdown HN12 cells did not impact tumor growth (Fig. 3M), yet it did restore the metastatic potential of tumors to lymph nodes (Fig. 3N and 3O). These findings strongly suggest that the c-Myc-Slug signaling axis plays a role in GLS1-mediated invasion and metastasis of HNSCC cells.

Figure 3. GLS1 knockdown suppresses HNSCC cell invasion and metastasis through downregulation of the c-Myc-Slug axis.

Figure 3.

Open in a new tab

(A-C) Effect of GLS1 knockdown on cell invasion of HN6 and HN12 cells determined by Transwell invasion assay (A), hanging drop aggregation assay (B) and 3D invasion assay (C), respectively. Representative images and quantitative data (n = 3) are shown in the left and right panels, respectively. Scale bar in (B) and (C) is 200 μm. (D) Representative tongue tumor images and growth curve of tumors derived from GLS1 knockdown or control HN12 cells 20 days after cell inoculation (n = 5 mice/group). (E-F) Effect of GLS1 knockdown on lymph node metastasis determined by lymph node weight (E) and IHC with anti-pan-keratin antibody (F). In (E), lymph node images and quantitative data (n = 5 mice/group) are shown in the left and right panels, respectively. (G) Effect of GLS1 knockdown on the protein levels of EMT molecular markers in HN6 and HN12 cells. (H) Immunostaining of Slug in tumor tissues derived from GLS1 knockdown or control HN12 cells. Representative IHC images and quantitative IOD data (n = 5 mice/group) are shown in the left and right panels, respectively. (I) Effect of MYC knockdown on Slug protein levels in HN6 and HN12 cells. (J) Effect of restoring MYC expression on GLS1 knockdown-mediated Slug expression in HN6 and HN12 cells. (K-L) Effect of restoring MYC or Slug expression on GLS1 knockdown-mediated suppression of cell migration and invasion in HN6 and HN12 cells. shGFP: shRNA against the GFP gene as a negative control; shGLS1–1 and shGLS1–2: two specific shRNAs against the GLS1 gene. (M) Effect of Slug expression restoration in GLS1 knockdown on tumor growth. (N-O) Effect of Slug expression restoration in GLS1 knockdown on lymph node metastasis determined by lymph node weight (N) and IHC with anti-pan-keratin antibody (O). MYC O/E: MYC overexpression vector; Slug O/E: Slug-Flag overexpression vector. *p<0.05; **p<0.01.

GLS1 loss inhibits Slug acetylation via the c-Myc-ACC signaling axis in HNSCC cells

Our RT-qPCR data did not show a significant change in Slug expression between GLS1 knockdown and control HN6 cells (Supplementary Fig. S8), suggesting that GLS1 may regulate Slug by another mechanism. Slug protein is subjected to rapid turnover, and its stability is highly controlled by post-translational acetylation on the epsilon-amine groups of lysine residues (29). Western blot with anti-acetylated lysine antibody showed decreased total levels of acetylated lysine in HN6 and HN12 cells upon GLS1 knockdown (Fig. 4A) or CB-839 treatment (Fig. 4B). Moreover, GLS1 knockdown cells exhibited decreased levels of acetyl-CoA (Fig. 4C). co-IP analysis further demonstrated acetylation of Slug in HNSCC cells (Fig. 4D) and knockdown of GLS1 reduced total and acetylated Slug (Fig. 4E). To determine whether GLS1 affects Slug acetylation, GLS1 knockdown and control HN6 and HN12 cells transfected with Slug-Flag were subjected to IP using anti-Flag antibody. Western blot revealed a significant reduction in histone H3 acetylation-associated Slug in both cell lines when GLS1 was depleted (Fig. 4F).

Figure 4. GLS1 knockdown inhibits Slug acetylation via the c-Myc-ACC signaling axis in HNSCC cells.

Figure 4.

Open in a new tab

(A) Effect of GLS1 knockdown on the levels of acetylated lysine in HN6 and HN12 cells. (B) Effect of CB-839 on the levels of acetylated lysine in HN6 and HN12 cells. HN6 and HN12 cells were treated with 2 μM CB-839 for 48 hours and cell lysates were collected for Western blot with anti-acetylated lysine antibody. (C) Effect of MYC on acetyl-CoA levels in HN6 and HN12 cells with or without GLS1 knockdown determined by acetyl-CoA assay. (D) Acetylation of Slug in HN6 and HN12 cells determined by either IP with anti-acetylated lysine antibody followed by immunoblotting with anti-Slug antibody or IP with anti-Slug antibody followed by immunoblotting with anti-acetylated lysine antibody. (E, F) Effect of GLS1 knockdown on acetylated Slug levels in HN6 and HN12 cells. In (E), cell lysates from GLS1 knockdown and control cells were collected for IP with anti-acetylated-lysine antibody, followed by Western blot with anti-Slug antibody. In (F), cell lysates from GLS1 knockdown and control cells transfected with Slug-Flag (Slug O/E) were collected for IP with anti-Flag antibody, followed by Western blot with antibodies against acetylated lysine and acetyl-histone H3. (G) Effect of GLS1 knockdown on ACC phosphorylation levels in HN6 and HN12 cells. (H) Effect of ACC inhibitor TOFA on Slug levels in HN6 and HN12 cells with or without GLS1 knockdown. (I) Effect of MYC on ACC-Slug signaling axis in HN6 and HN12 cells with or without GLS1 knockdown. shGFP: shRNA against the GFP gene as a negative control; shGLS1: specific shRNA against the GLS1 gene. MYC O/E: MYC overexpression vector; Slug O/E: Slug-Flag overexpression vector. *p<0.05; **p<0.01; ns, not significant.

GLS1-mediated glutaminolysis augments intracellular acetyl-CoA via the tricarboxylic acid cycle (TCA) (30). Since acetyl-CoA is regulated by ACC (31,32), we investigated changes in ACC levels in HNSCC cells following GLS1 knockdown. We found a significant decrease in phosphor-ACC levels, accompanied by an increase in total ACC levels, in GLS1 knockdown HN6 and HN12 cells when compared to their respective control cells (Fig. 4G). Increased Slug levels with enhanced ACC phosphorylation were observed in GLS1 knockdown cells in the presence of TOFA, an allosteric inhibitor of ACC (33) (Fig. 4H), suggesting that Slug suppressed by GLS1 knockdown is highly dependent on ACC in HNSCC cells. Given that c-Myc orchestrates fatty acid metabolism and controls the abundance of acetyl-CoA (14), we overexpressed MYC in GLS1 knockdown and control HN6 and HN12 cells and measured acetyl-CoA levels in these cells. Restoration of MYC expression reversed the decreased acetyl-CoA levels and ACC phosphorylation in GLS1 knockdown cells (Fig. 4C and 4I), supporting the notion that GLS1 elicits the induction of Slug acetylation through activating the c-Myc-ACC signaling axis in HNSCC cells.

MYC knockdown transcriptionally downregulates GLS1 expression and more potently suppresses Slug acetylation and HNSCC cell invasion in the presence of CB-839

c-Myc is a multifunctional transcription factor that exhibits robust coexpression with GLS1 in HNSCC (Fig. 1A and 1B). Consistent with another report (20), blocking c-Myc signaling with its inhibitor MYCi975 or by shRNAs significantly inhibited cellular glutamine consumption (Fig. 5A and 5B). MYCi975 treatment resulted in dose-dependent degradation of c-Myc, which was accompanied by decreased GLS1 levels in both HN6 and HN12 cells (Fig. 5C). The same tendency was observed in GLS1 levels in these cells upon MYC knockdown (Fig. 5D), which was confirmed by RT-qPCR analysis (Fig. 5E). Overexpression of MYC enhanced GLS1 levels in HN6 and HN12 cells (Fig. 5F), indicating the regulatory role of c-Myc in GLS1 expression.

Figure 5. MYC knockdown reduces GLS1 expression and is more potent in suppressing Slug acetylation and HNSCC cell invasion in the presence of CB-839.

Figure 5.

Open in a new tab

(A) Glutamine consumption of HN6 and HN12 cells treated with or without 5 μM MYCi975 for 48 hours. (B) Alterations in glutamine consumption in GLS1 knockdown HN6 and HN12 cells. (C) Dose-dependent effects of MYCi975 on the protein levels of GLS1 and c-Myc. (D-E) Effect of MYC knockdown on GLS1 gene expression and protein levels in HN6 and HN12 cells determined by Western blot and RT-qPCR, respectively. (F) Effect of MYC overexpression on GLS1 protein levels in HN6 and HN12 cells. (G) Schematic diagram of the binding site of c-Myc in the upstream promoter of the GLS1 gene. (H) Binding of c-Myc on the GLS1 gene promoter determined by ChIP assay in MYC overexpressing and control HN6 and HN12 cells. (I) Effect of MYC knockdown on Slug protein levels in the presence or absence of 2 μM CB-839. (J) Effect of MYC knockdown on acetylated Slug levels in the presence or absence of CB-839. MYC knockdown and control HN6 and HN12 cells were transfected with Slug-Flag and treated with or without 2 μM CB-839 for 24 hours. Cell lysates were then collected for IP with anti-Flag antibody and Western blot with anti-acetylated lysine antibody. (K) Quantitative data of cell invasion of MYC knockdown and control HN6 and HN12 cells in the presence or absence of 2 μM CB-839 for 24 hours. shGFP: shRNA against the GFP gene as a negative control; shMYC-1 and shMYC-2: two specific shRNAs against the MYC gene. EV: empty vector; MYC O/E: MYC overexpression vector. *p<0.05; **p<0.01; ***p<0.001.

c-Myc controls the expression of its target genes by binding to the E-box sequence on their promoters (34). Through promoter analysis, an E-box sequence spanning −1218 to −1229 was identified within the 2,000 bp upstream of the GLS1 gene promoter (Fig. 5G). To explore whether the binding of c-Myc to the GLS1 gene promoter was required for the transcriptional activation of GLS1, we performed ChIP-qPCR assay. Results revealed a distinct c-Myc occupancy on the GLS1 gene promoter in both HN6 and HN12 cells (Fig. 5H). Knockdown of MYC resulted in a dramatic decrease in Slug protein levels, which was more pronounced in the presence of CB-839 (Fig. 5I). Consistently, reduced levels of Slug acetylation were observed in Slug overexpressing MYC knockdown HN6 and HN12 cells compared to control cells, with the effect significantly enhanced in the presence of CB-839 (Fig. 5J). In addition, MYC knockdown was more potent in suppressing HNSCC cell invasion in the presence of CB-839 (Fig. 5K and Supplementary Fig. S9). These data suggest that reducing c-Myc levels at both mRNA and protein levels could more effectively block Slug signaling.

The combination of CB-839 and MYCi975 suppresses tumor growth and metastasis in mice more potently than either treatment alone

Our analysis of TCGA HNSCC cohort showed that elevated levels of GLS1 or MYC expression were significantly associated with poorer overall survival in HNSCC patients (Fig. 6A). Given that GLS1 and c-Myc formed a double positive feedback loop in HNSCC cells, there was potential for enhanced antitumor activity by inhibiting GLS1 and c-Myc simultaneously. Administration of 2 μM CB-839 or 5 μM MYCi975 significantly inhibited cell proliferation, colony formation and glutamine consumption in HN6 and HN12 cells (Fig. 6B–6D and Supplementary Fig. S10). Treatment with both drugs suppressed these phenotypes more effectively than either drug alone (Fig. 6B–6D and Supplementary Fig. S10), with a CI of 0.48413 (Supplementary Fig. S11). More strikingly, this combination was also more effective than monotherapy in suppressing HN6 and HN12 cell invasion (Fig. 6E and 6F).

Figure 6. The combination of CB-839 and MYCi975 exhibits a synergistic antitumor effect in the orthotopic mouse model of HNSCC.

Figure 6.

Open in a new tab

(A) Correlation between the expression of GLS1 (on the left) or MYC (on the right) and the overall survival of patients in TCGA HNSCC cohort. (B) Effect of CB-839 and/or MYCi975 on cell proliferation of HN6 and HN12 cells. HN6 and HN12 cells were treated with 2 μM CB-839 and/or 5 μM MYCi975 for 72 hours. (C) Effect of CB-839 and/or MYCi975 on cell colony formation of HN6 and HN12 cells. HN6 and HN12 cells were treated with 2 μM CB-839 and/or 5 μM MYCi975 for 10 days. (D) Effect of CB-839 and/or MYCi975 on cellular glutamine consumption in HN6 and HN12 cells. HN6 and HN12 cells were treated with 2 μM CB-839 and/or 5 μM MYCi975 for 72 hours. (E) Effect of CB-839 and/or MYCi975 on 3D cellular aggregates of HN6 and HN12 cells determined by hanging drop aggregation assay. HN6 and HN12 cells were cultured on the inside lid of Petri dishes in a drop of culture media (30 μl) at 7500 cells/drop, followed by treatment with 2 μM CB-839 and/or 5 μM MYCi975 for 4 days. Representative bright-field images and quantitative data of the variation in time of perimeter (n = 3) are shown in the upper and lower panels, respectively. Scale bar: 200 μm. (F) Effect of CB-839 and/or MYCi975 on cell invasion in HN6 and HN12 cells determined by Transwell invasion assay. HN6 and HN12 cells were treated with 2 μM CB-839 and/or 5 μM MYCi975 for 24 hours. (G) Treatment scheme in the orthotopic mouse model. Ten days after intramucosal injection, HN12 tumor-bearing NSG mice were randomized to receive treatment with vehicle, CB-839, MYCi975, alone or in combination. MYCi975 was given by intraperitoneal injection at 100 mg/kg once daily consecutively for 15 days, and CB-839 was given by oral gavage at 300 mg/kg once daily consecutively for 15 days. The experiment was terminated on day 28 after cell inoculation. (H-J) Tumor growth curve, tumor weight, and body weight in different treatment groups. (K) Representative lymph node images and quantitative data of lymph node weight in different treatment groups. (L) Effect of CB-839 and/or MYCi975 on lymph node metastasis determined by HE and IHC with anti-pan-keratin antibody. Representative images and quantitative data (n = 5 mice/group) are shown in the left and right panels, respectively. (M) Mouse survival curves for groups receiving CB-839 and/or MYCi975 treatment (log-rank test). *p<0.05; **p<0.01; ***p<0.001.

Since tumors in the buccal mucosa can grow larger than those in the tongue without affecting the mice’s ability to eat and drink, an extended experimental window can be used to assess the therapeutic efficacy of the CB-839 and MYCi975 combination. In mice with orthotopic mucosal tumors, a significant reduction in tumor burden was observed following 15 days of treatment with either drug alone, as evidenced by smaller tumor size, slower tumor growth rate, and lower tumor weight (Fig. 6G–6J). The addition of CB-839 to MYCi975 suppressed tumor growth and tumoral c-Myc levels in mice more potently than either treatment alone (Fig. 6G–6J and Supplementary Fig. S12). No noticeable changes in body weight were observed across the treatment groups, indicating good tolerability to both single and combination drug treatments (Fig. 6J). In addition, the combination of CB-839 and MYCi975 almost completely suppressed lymph node metastasis, resulting in a more robust suppression of tumor metastasis than single arm treatment (Fig. 6K and 6L). As a result, mice receiving the combination treatment survived longer than mice receiving single treatment (Fig. 6M). These findings suggest that the combination of CB-839 and MYCi975 is a safe and more effective strategy for HNSCC treatment.

Discussion

HNSCC is a highly aggressive and heterogeneous cancer (26,35,36), in which metabolic reprogramming not only generates an energy source but also provides structural components, facilitating rapid proliferation and progression and reducing susceptibility to elimination (37). Our previous study demonstrated that HNSCC cells are highly addicted to GLS1-dependent glutaminolysis (5). While the importance of GLS1 in cancer cell glutamine metabolism and tumor growth has been well established, its role in tumor metastasis is less clear. Using HNSCC cell lines and animal models as research platforms, we have uncovered the critical role of GLS1 in preventing c-Myc degradation through USP1-dependent deubiquitination (Fig. 7). Concurrently, c-Myc activates transcription of GLS1 by binding to its gene promoter, establishing a dual positive regulatory loop. This loop increases acetyl-CoA levels in HNSCC cells via ACC, leading to an increase in acetylated Slug protein and ultimately promoting EMT-associated cancer invasion and metastasis (Fig. 7). Co-administration of CB-839 and MYCi975 disrupts the GLS1-c-Myc feedback loop in HNSCC cells more effectively than either drug alone, resulting in greater suppression of malignant progression (Fig. 7).

Figure 7. Schematic depicting the molecular mechanism of a dual positive feedback loop between GLS1 and c-Myc in HNSCC cells.

Figure 7.

Open in a new tab

In HNSCC cells, GLS1 plays a crucial role in preventing the degradation of c-Myc through USP1-dependent deubiquitination. Simultaneously, c-Myc transcriptionally activates GLS1 expression by binding to its gene promoter, establishing a double positive regulatory loop. This loop further promotes protein acetylation through ACC, resulting in an upregulation of acetylated Slug protein, ultimately contributing to EMT-associated invasion and metastasis of HNSCC cells. The combination of CB-839 and MYCi975 for co-treatment effectively inhibits the GLS1-c-Myc feedback loop, resulting in a more potent suppression of HNSCC metastasis compared with either drug administered alone.

GLS1 has two isoforms, KGA and GAC, which are generated by alternative splicing of the GLS1 gene and have distinct functions and cellular localization. GAC is mainly found in the mitochondria and is involved in the regulation of cellular bioenergetics, while KGA is mainly found in the cytoplasm and has been implicated in the regulation of cell proliferation and survival (11). The GAC splice variant exhibits a higher frequency of upregulation in cancer cells compared to KGA (7). Corroborating this finding, our study identified GAC as the predominant isoform of GLS1 in HNSCC cells. Notably, the presence of KGA is scarcely detectable in HN6 and HN12 cells, implying that GAC, but not KGA, plays a specific role in c-Myc-associated Slug acetylation. Acetylation likely plays a role in stabilizing Slug protein, and increased Slug acetylation has been correlated with promoting EMT-associated invasion and metastasis (29,38). Our study is the first to demonstrate that disrupting Slug acetylation, by blocking GLS1, can attenuate Slug-mediated HNSCC cell invasion and metastasis. Interestingly, we found a decrease in N-cadherin and an increase in E-cadherin in GLS1 knockdown HNSCC cells, accompanied by Slug downregulation. A recent study highlighted that Slug acetylation, facilitated by CBP at lysine 166 and 211, prolongs its half-life and stability and promotes breast cancer migration by suppressing E-cadherin expression while enhancing N-cadherin expression (38). The potential direct regulation of N-cadherin and E-cadherin expression by GLS1-mediated Slug acetylation will be a focus of investigation in our forthcoming studies.

In certain types of cancers, such as those driven by c-Myc, tumor cells appear to be dependent on glutamine (1,39). Our discovery proposes a novel mechanism by which GLS1, the key enzyme involved in glutamine metabolism, may play a role in metastasis through functional interactions with c-Myc. USP1 is a deubiquitinating enzyme regulating protein stability by removing ubiquitin moieties from target proteins (40). The involvement of USP1 in stabilizing proteins related to cell cycle control and DNA repair suggests that its aberrant activity could contribute to genomic instability and tumor development (41). In diffuse large B-cell lymphoma (DLBCL) cells, USP1 directly binds to MAX, a c-Myc binding protein, and maintains its protein stability, which facilitates the transcription of c-Myc target genes (42). Here, we identified that USP1 directly stabilizes c-Myc through deubiquitination in HNSCC cells, and this regulatory process is governed by GLS1.

Oncogenic alterations in metabolism create a dependency of cancer cells on specific nutrients, making metabolic pathways such as glutaminolysis a target for therapeutic intervention. MYCi975 is a c-Myc inhibitor that has shown significant antitumor activity in mice and promising pharmacokinetic properties, such as high plasma concentration, longer half-life, and improved tumor penetration (43). Although single-target drugs have shown success in preclinical studies, they have limited clinical efficacy in cancer. In a pioneering approach, we used CB-839 and MYCi975 simultaneously to target GLS1 and c-Myc in HNSCC cells. Encouragingly, we report here that this combination is potent in reducing head and neck cancer growth and metastasis, providing a hopeful avenue for therapeutic intervention by disrupting the double positive loop between GLS1 and c-Myc. Further investigation is imperative to determine whether the combination of CB-839 and MYCi975 has the potential to reconstitute antitumor immunity in syngeneic mouse models of HNSCC. In addition, thorough investigations are essential to assess the safety of treatment regimens based on the simultaneous inhibition of GLS1 and c-Myc in both preclinical and clinical settings.

Supplementary Material

1

Significance.

GLS1 and c-Myc form a positive feedback loop that promotes head and neck cancer metastasis and can be targeted as a promising therapeutic strategy for this disease.

Acknowledgements

This work was partially supported by R01 funding from NIH/NIDCR to YT (R01DE028351, R01DE033433 and R01DE033691). Additional support to YT was provided by the Georgia Cancer Research Fund, I3 Morningside Center Research Award, and I3 Nexus Research Award from Emory School of Medicine, a gift from Woodruff Fund Inc., and through the Georgia CTSA NIH award (UL1-TR002378). The study was also supported by Winship Invest$ Team Science Award and Winship Invest$ Pilot Award under award number P30CA138292. YT is the inaugural recipient of the Wally Award from Winship Cancer Institute. JM was supported by LGS-SOAR Internship at Emory University and American Cancer Society for Diversity in Cancer Research Internship. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. We would like to acknowledge Anthea Hammond for critical reading of this manuscript and the technical support from the Shared Resource at Winship for Cancer Tissue and Pathology.

Abbreviation list:

ACC

Acetyl-CoA carboxylase

Acetyl-CoA

Acetyl-coenzyme A

ANOVA

Analysis of variance

ChIP

Chromatin immunoprecipitation

CHX

Cycloheximide

DUBs

Deubiquitinating enzymes

EMT

Epithelial-mesenchymal transition

FBS

Fetal bovine serum

FFPE

Paraffin-embedded

GAC

Glutaminase C

GEO

Gene Expression Omnibus

GO

Gene Ontology

GSEA

Gene Set Enrichment Analysis

GLS1

Glutaminase 1

HNSCC

Head and neck squamous cell carcinoma

IHC

Immunohistochemistry

IP

Immunoprecipitation

MF

Molecular function

KGA

Kidney-type glutaminase

USP1

Ubiquitin-specific protease 1

RT-qPCR

Quantitative reverse transcription-polymerase chain reaction

SEM

Mean ± standard error of the mean

shRNAs

Short hairpin RNAs

SIAH2

Seven in absentia homolog 2

TCA

Tricarboxylic acid cycle

TOFA

5-(tetradecyloxy)-2-furoic acid

UBE2S

Ubiquitin conjugating enzyme E2S

Footnotes

Disclosure of Potential Conflict of Interest: The authors declare no competing financial interests

Reference

  • 1.Martínez-Reyes I, Chandel NS. Cancer metabolism: looking forward. Nat Rev Cancer 2021;21:669–80. [DOI] [PubMed] [Google Scholar]
  • 2.Bergers G, Fendt SM. The metabolism of cancer cells during metastasis. Nat Rev Cancer 2021;21:162–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Jin L, Alesi GN, Kang S. Glutaminolysis as a target for cancer therapy. Oncogene 2016;35:3619–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Yoo HC, Yu YC, Sung Y, Han JM. Glutamine reliance in cell metabolism. Exp Mol Med 2020;52:1496–516. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Lang L, Wang F, Ding Z, Zhao X, Loveless R, Xie J, et al. Blockade of glutamine-dependent cell survival augments antitumor efficacy of CPI-613 in head and neck cancer. J Exp Clin Cancer Res 2021;40:393. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Kodama M, Oshikawa K, Shimizu H, Yoshioka S, Takahashi M, Izumi Y, et al. A shift in glutamine nitrogen metabolism contributes to the malignant progression of cancer. Nat Commun 2020;11:1320. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.van den Heuvel AP, Jing J, Wooster RF, Bachman KE. Analysis of glutamine dependency in non-small cell lung cancer: GLS1 splice variant GAC is essential for cancer cell growth. Cancer Biol Ther 2012;13:1185–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Daemen A, Liu B, Song K, Kwong M, Gao M, Hong R, et al. Pan-Cancer Metabolic Signature Predicts Co-Dependency on Glutaminase and De Novo Glutathione Synthesis Linked to a High-Mesenchymal Cell State. Cell Metab 2018;28:383–99.e9. [DOI] [PubMed] [Google Scholar]
  • 9.Katt WP, Lukey MJ, Cerione RA. A tale of two glutaminases: homologous enzymes with distinct roles in tumorigenesis. Future Med Chem 2017;9:223–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Xiang L, Mou J, Shao B, Wei Y, Liang H, Takano N, et al. Glutaminase 1 expression in colorectal cancer cells is induced by hypoxia and required for tumor growth, invasion, and metastatic colonization. Cell Death Dis 2019;10:40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Li B, Cao Y, Meng G, Qian L, Xu T, Yan C, et al. Targeting glutaminase 1 attenuates stemness properties in hepatocellular carcinoma by increasing reactive oxygen species and suppressing Wnt/beta-catenin pathway. EBioMedicine 2019;39:239–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Wu S, Fukumoto T, Lin J, Nacarelli T, Wang Y, Ong D, et al. Targeting glutamine dependence through GLS1 inhibition suppresses ARID1A-inactivated clear cell ovarian carcinoma. Nat Cancer 2021;2:189–200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Stine ZE, Schug ZT, Salvino JM, Dang CV. Targeting cancer metabolism in the era of precision oncology. Nat Rev Drug Discov 2022;21:141–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Dong Y, Tu R, Liu H, Qing G. Regulation of cancer cell metabolism: oncogenic MYC in the driver’s seat. Signal Transduct Target Ther 2020;5:124. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Reynolds MR, Lane AN, Robertson B, Kemp S, Liu Y, Hill BG, et al. Control of glutamine metabolism by the tumor suppressor Rb. Oncogene 2014;33:556–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Stine ZE, Walton ZE, Altman BJ, Hsieh AL, Dang CV. MYC, Metabolism, and Cancer. Cancer Discov 2015;5:1024–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Ashrafizadeh M, Zarabi A, Hushmandi K, Moghadam ER, Hashemi F, Daneshi S, et al. C-Myc Signaling Pathway in Treatment and Prevention of Brain Tumors. Curr Cancer Drug Targets 2021;21:2–20. [DOI] [PubMed] [Google Scholar]
  • 18.Hirschey MD, DeBerardinis RJ, Diehl AME, Drew JE, Frezza C, Green MF, et al. Dysregulated metabolism contributes to oncogenesis. Semin Cancer Biol 2015;35 Suppl:S129–s50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Wu S, Yin X, Fang X, Zheng J, Li L, Liu X, Chu L. c-MYC responds to glucose deprivation in a cell-type-dependent manner. Cell Death Discov 2015;1:15057. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Bott AJ, Peng IC, Fan Y, Faubert B, Zhao L, Li J, et al. Oncogenic Myc Induces Expression of Glutamine Synthetase through Promoter Demethylation. Cell Metab 2015;22:1068–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Gao L, Zhao X, Lang L, Shay C, Andrew Yeudall W, Teng Y. Autophagy blockade sensitizes human head and neck squamous cell carcinoma towards CYT997 through enhancing excessively high reactive oxygen species-induced apoptosis. J Mol Med (Berl) 2018;96:929–38. [DOI] [PubMed] [Google Scholar]
  • 22.Lang L, Shay C, Xiong Y, Thakkar P, Chemmalakuzhy R, Wang X, Teng Y. Combating head and neck cancer metastases by targeting Src using multifunctional nanoparticle-based saracatinib. J Hematol Oncol 2018;11:85. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Lang L, Loveless R, Dou J, Lam T, Chen A, Wang F, et al. ATAD3A mediates activation of RAS-independent mitochondrial ERK1/2 signaling, favoring head and neck cancer development. J Exp Clin Cancer Res 2022;41:43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Teng Y Hanging Drop Aggregation Assay of Breast Cancer Cells. Bio-protocol 2015;5(3) [Google Scholar]
  • 25.Zhao X, Xu F, Dominguez NP, Xiong Y, Xiong Z, Peng H, et al. FGFR4 provides the conduit to facilitate FGF19 signaling in breast cancer progression. Mol Carcinog 2018;57:1616–25. [DOI] [PubMed] [Google Scholar]
  • 26.Gao L, Lang L, Zhao X, Shay C, Shull AY, Teng Y. FGF19 amplification reveals an oncogenic dependency upon autocrine FGF19/FGFR4 signaling in head and neck squamous cell carcinoma. Oncogene 2019;38:2394–404. [DOI] [PubMed] [Google Scholar]
  • 27.Dang CV. MYC on the path to cancer. Cell 2012;149:22–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.García-Santisteban I, Peters GJ, Giovannetti E, Rodríguez JA. USP1 deubiquitinase: cellular functions, regulatory mechanisms and emerging potential as target in cancer therapy. Mol Cancer 2013;12:91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Zhou W, Ni TK, Wronski A, Glass B, Skibinski A, Beck A, Kuperwasser C. The SIRT2 Deacetylase Stabilizes Slug to Control Malignancy of Basal-like Breast Cancer. Cell Rep 2016;17:1302–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Xia X, Cao G, Sun G, Zhu L, Tian Y, Song Y, et al. GLS1-mediated glutaminolysis unbridled by MALT1 protease promotes psoriasis pathogenesis. J Clin Invest 2020;130:5180–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Pougovkina O, te Brinke H, Ofman R, van Cruchten AG, Kulik W, Wanders RJ, et al. Mitochondrial protein acetylation is driven by acetyl-CoA from fatty acid oxidation. Hum Mol Genet 2014;23:3513–22. [DOI] [PubMed] [Google Scholar]
  • 32.Rios Garcia M, Steinbauer B, Srivastava K, Singhal M, Mattijssen F, Maida A, et al. Acetyl-CoA Carboxylase 1-Dependent Protein Acetylation Controls Breast Cancer Metastasis and Recurrence. Cell Metab 2017;26:842–55.e5. [DOI] [PubMed] [Google Scholar]
  • 33.Gao L, Xu Z, Huang Z, Tang Y, Yang D, Huang J, et al. CPI-613 rewires lipid metabolism to enhance pancreatic cancer apoptosis via the AMPK-ACC signaling. J Exp Clin Cancer Res 2020;39:73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Pelengaris S, Khan M, Evan G. c-MYC: more than just a matter of life and death. Nat Rev Cancer 2002;2:764–76. [DOI] [PubMed] [Google Scholar]
  • 35.Lalonde CS, Teng Y, Burtness BA, Ferris RL, Ahmed R, Saba NF. The Quest to Eradicate HPV-Related Oropharyngeal Carcinoma: An Opportunity Not to Miss. J Natl Cancer Inst 2022;114:1333–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Saba NF, Steuer CE, Ekpenyong A, McCook-Veal A, Magliocca K, Patel M, et al. Pembrolizumab and cabozantinib in recurrent metastatic head and neck squamous cell carcinoma: a phase 2 trial. Nat Med 2023;29:880–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Gao L, Yang F, Tang D, Xu Z, Tang Y, Yang D, et al. Mediation of PKM2-dependent glycolytic and non-glycolytic pathways by ENO2 in head and neck cancer development. J Exp Clin Cancer Res 2023;42:1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Dai X, Xin Y, Xu W, Tian X, Wei X, Zhang H. CBP-mediated Slug acetylation stabilizes Slug and promotes EMT and migration of breast cancer cells. Sci China Life Sci 2021;64:563–74. [DOI] [PubMed] [Google Scholar]
  • 39.Muir A, Danai LV, Gui DY, Waingarten CY, Lewis CA, Vander Heiden MG. Environmental cystine drives glutamine anaplerosis and sensitizes cancer cells to glutaminase inhibition. Elife 2017;6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Estavoyer B, Messmer C, Echbicheb M, Rudd CE, Milot E, Affar EB. Mechanisms orchestrating the enzymatic activity and cellular functions of deubiquitinases. J Biol Chem 2022;298:102198. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Coleman KE, Yin Y, Lui SKL, Keegan S, Fenyo D, Smith DJ, et al. USP1-trapping lesions as a source of DNA replication stress and genomic instability. Nat Commun 2022;13:1740. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Li XY, Wu JC, Liu P, Li ZJ, Wang Y, Chen BY, et al. Inhibition of USP1 reverses the chemotherapy resistance through destabilization of MAX in the relapsed/refractory B-cell lymphoma. Leukemia 2023;37:164–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Madden SK, de Araujo AD, Gerhardt M, Fairlie DP, Mason JM. Taking the Myc out of cancer: toward therapeutic strategies to directly inhibit c-Myc. Mol Cancer 2021;20:3. [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

1
NIHMS2011829-supplement-1.docx (2.1MB, docx)

Data Availability Statement

The RNA-seq data generated in this study is available in GEO at GSE240956. Public data included in TCGA HNSCC cohort analysis were downloaded from GDC Data Portal (https://portal.gdc.cancer.gov/projects/TCGA-HNSC). All other raw data generated in this study are available upon request from the corresponding author.

RESOURCES