Introduction

Head and neck cancer is the seventh most common cancer worldwide, with over 90% of cases classified as squamous cell carcinoma (HNSCC)1. Increasing rates of human papillomavirus (HPV) infection, particularly in the oropharynx, contribute significantly to the rising prevalence of head and neck cancer. HPV-positive HNSCC affects younger patients and demonstrates significantly improved survival compared to HPV-negative cases as evidenced by an up to 58% reduction in mortality, likely due to distinct biological and immunological characteristics2,3. However, standard therapies such as chemoradiation and surgical interventions frequently result in severe long-term morbidity, including dysarthria, dysphagia, dependence on tracheostomy or gastrostomy tubes, and disfiguring outcomes4. Consequently, identifying actionable metabolic vulnerabilities specific to HPV-positive HNSCC is crucial for developing targeted, less debilitating therapeutic strategies.

HPV promotes HNSCC tumorigenesis primarily through the viral oncoproteins E6 and E7, which target and inactivate the key tumor suppressor proteins p53 and retinoblastoma protein (Rb), respectively5. Because p53 and Rb regulate essential pathways in DNA repair, cell cycle control, and apoptosis, their disruption leads directly to uncontrolled cell proliferation and altered cellular metabolism6. A hallmark of such metabolic alteration is termed aerobic glycolysis—known as the Warburg effect—whereby cancer cells preferentially utilize glycolysis for energy even in the presence of ample oxygen, fueling rapid growth and biomass accumulation7. Indeed, HPV-positive cervical cancer exhibits distinct metabolic reprogramming, involving increased glycolysis, purine salvage, and glutathione metabolism8. Given the common viral mechanisms used by HPV to infect mucosal epithelial tissues in both cervix and oropharynx, it is necessary to determine whether similar metabolic vulnerabilities exist in HPV-positive HNSCC.

Previous metabolomic studies of head and neck cancer showed at least 321 metabolites that differ significantly from patients with oral squamous cell carcinoma (OSCC) versus controls9. Identification of pathways related to enriched metabolites demonstrated vast metabolic upregulation, including glucose, glutamine, serine, and arginine metabolism in pathologies such as pancreatic adenocarcinoma and triple-negative breast cancer (TNBC)9,10,11. Similarly, metabolomic analysis from OSCC patient-derived serum and saliva have demonstrated increases in the same metabolites among many others, such as sphingolipids, when compared to controls12,13,14. Therapeutic targets against glutamine are of emerging interest due to its central role in crucial metabolic pathways, including the tricarboxylic acid (TCA) cycle and oxidative phosphorylation (OXPHOS)15.

The metabolic landscape of HNSCC is likely shaped by both nutrient availability and HPV status. Although glutamine is vital for maximal proliferation, HNSCC cell lines have shown greater rates of cell death when deprived of glucose compared to glutamine16. Further, studies on HPV presence and absence in HNSCC have linked increased dependence on glucose with HPV presence but found no significant metabolomic differences between the two17. Additionally, while some studies link HPV-positive growth to OXPHOS and associate HPV-negative HNSCC with hypoxia-inducible factors (HIFs), others report the reverse metabolic pattern17,18,19. Knockdown of HPV oncoproteins E6 and E7 has been shown to reduce aerobic glycolysis in HPV-positive cervical cancer cells, indicating that HPV-associated cancers should demonstrate increased Warburg metabolism20. While several studies have investigated metabolic differences between HPV-positive and HPV-negative HNSCC, findings have been heterogeneous. Thus, a comprehensive metabolomic characterization is essential to clarify the detailed contributions of HPV to metabolic reprogramming.

In this study, we demonstrate for the first time to our knowledge that HPV-positive HNSCC exhibits specific metabolic alterations in multiple pathways including glycolysis, nucleotide biosynthesis, nicotinamide metabolism, and glutathione metabolism. By identifying and characterizing these metabolic differences, we highlight novel vulnerabilities that could be exploited therapeutically to develop targeted interventions aimed at treatment de-intensification for HPV-positive HNSCC and improved patient quality of life.

Results

Carbohydrate metabolism and pentose phosphate pathway

HPV-positive cell lines SCC47 and SCC104 demonstrated a substantial preference for carbohydrate metabolism by exhibiting higher relative abundance of glycolytic intermediates when compared to HPV-negative cell lines (SCC9 and Detroit562). Key metabolites such as 3-phosphoglycerate (FC = 27.32, p = 0.018), phosphoenolpyruvate (FC = 20.60, p = 0.0046), pyruvate (FC = 15.41, p = 0.00054), and glycerate (FC = 14.08, p = 0.010) 

were significantly elevated in HPV-positive cells (Fig. 1A and B). Interestingly, lactate, a terminal product of anaerobic glycolysis, was only notably elevated in comparison of SCC104 to Detroit562 (FC = 2.00, p = 0.023) (Fig. 1A). SCC47 exhibited increased levels of the isomeric sugar alcohols mannitol and sorbitol (FC = 2.17–2.18, p = 0.024 (Fig. 1A). SCC47 was also found to have upregulated pentose phosphate pathway (PPP) intermediates such as sedoheptulose-7-phosphate (FC = 9.41, p = 0.024), ribulose (FC = 2.83, p = 0.039), and ribose 5-phosphate (FC = 5.98, p = 0.034) (Fig. 1A). A similar but more modest upregulation of PPP intermediates was seen in SCC104, particularly sedoheptulose-7-phosphate (FC = 11.78, p = 0.018) and ribose 5-phosphate (FC = 2.68, p = 0.0041). (Fig. 1A and B).

Fig. 1
figure 1

Carbohydrate metabolism. (A) The value in the cell denotes the fold change between the two cell lines listed on the x-axis, with green color indicating significantly increased abundance (p-value < 0.05) and red color indicating significant decreased abundance (p-value < 0.05). Narrowly missed significant values (0.05 < p-value < 0.10) are indicated by light green and light red colors. (B) Bar graphs show abundances (Ion count/mg protein content) of selected metabolites in each cell line, with data points representing each triplicate value. Significance bars with asterisks denote the following: * = p-value < 0.05, ** = p-value < 0.01, *** = p-value < 0.001.

Full size image

Purine metabolism

Metabolites from the PPP, such as ribose-5-phosphate, can be shunted into creation of purines and pyrimidines. Our results showed this process to be upregulated in HPV-positive cell lines, especially when compared to Detroit562. There were increased measures of intermediates such as hypoxanthine (FC = 4.24, p = 0.035), xanthine (FC = 1.95, p = 0.042), and guanosine (FC = 5.33, p = 0.040 (Fig. 2A and B). SCC47 exhibited a slight increase in uric acid (FC = 2.43, p = 0.039), the terminal product of purine breakdown, when compared to Detroit562 (Fig. 2A).

Fig. 2
figure 2

Nucleotide metabolism. (A) Purine Metabolism: see Fig. 1 legend for color key. (B) Bar graphs show abundances (Ion count/mg protein content) of selected metabolites of purine metabolism in each cell line, with data points representing each triplicate value. See Fig. 1 legend for significance key. (C) Pyrimidine metabolism: see Fig. 1 legend for color key. (D) Bar graphs show abundances (Ion count/mg protein content) of selected metabolites of pyrimidine metabolism in each cell line, with data points representing each triplicate value. See Fig. 1 legend for significance key.

Full size image

Pyrimidine metabolism

Similar to purine metabolism, there was a modest upregulation of pyrimidine metabolism intermediates, such as uracil (FC = 2.00, p = 0.046), beta-alanine (FC = 2.95, p = 0.024), cytidine diphosphate (CDP) (FC = 4.82, p = 0.017), and cytidine triphosphate (CTP) (FC = 3.97, p = 0.023) in HPV-positive cell lines (Fig. 2C and D). Pantothenate showed upregulation across two comparisons up to FC = 2.56 (p = 0.034) (Fig. 2C). Uridine monophosphate (UMP) measured uniquely among the four cell lines with Detroit562 showing the greatest average abundance (1,235,459 ion count/mg protein concentration [IC/mg]), followed by SCC104 (461,181 IC/mg), and the lowest abundance in SCC9 (8,625 IC/mg) (Fig. 2D).

Tricarboxylic acid cycle

There were minimal differences of TCA cycle intermediates between the four cell lines. However, cis-Aconitate (FC = 1.82, p = 0.026), oxalosuccinate (FC = 3.32, p = 0.042), and acetate (FC = 2.78, p = 0.018) were modestly upregulated in HPV-positive cell lines (Fig. 3A and B).

Fig. 3
figure 3

Tricarboxylic acid cycle and nicotinamide metabolism. (A) TCA Cycle: see Fig. 1 legend for color key. (B) Bar graphs show abundances (Ion count/mg protein content) of selected metabolites of TCA cycle metabolism in each cell line, with data points representing each triplicate value. See Fig. 1 legend for significance key. (C) Nicotinamide metabolism: see Fig. 1 legend for color key. (D) Bar graphs show abundances (Ion count/mg protein content) of selected metabolites of nicotinamide metabolism in each cell line, with data points representing each triplicate value. See Fig. 1 legend for significance key.

Full size image

Nicotinamide metabolism

Nicotinamide metabolism demonstrated a robust upregulation in HPV-positive cell lines as evidenced by the striking elevation in nicotinamide mononucleotide (NMN) across all four comparisons (FC = 37.04–55.97, p  = 0.017–0.020) (Fig. 3C and D). Nicotinamide (FC = 2.88, p = 0.025), nicotinamide riboside (FC = 3.10, p = 0.040), and nicotinate ribonucleoside (FC = 3.09, p = 0.041) were also upregulated (Fig. 3C and D). In general, compared to either HPV-negative cell lines, SCC47 demonstrated more upregulation of the nicotinamide intermediates than SCC104 (Fig. 3C).

Glutathione and methionine metabolism

HPV-positive cell lines had increased N-Formylmethionine (FC = 2.13, p = 0.024), S-adenosylhomocysteine (SAH) (FC = 4.83, p = 0.018), and cysteine (FC = 2.80, p = 0.028) (Fig. 4A and B). Cystine was markedly increased in SCC47 (FC = 7.90, p = 0.018) (Fig. 4A and B). The GSH:GSSG ratio, which serves as a marker for oxidative stress, calculated by metabolite abundance normalized to protein concentration was as follows: SCC9 = 219.1, Detroit562 = 241.8, SCC47 = 43.1, SCC104 = 225.2 (Fig. 4B).

Fig. 4
figure 4

Glutathione and methionine metabolism. (A) Glutathione and methionine metabolism: see Fig. 1 legend for color key. (B) Bar graphs show abundances (Ion count/mg protein content) of selected metabolites of glutathione and methionine metabolism in each cell line, with data points representing each triplicate value. See Fig. 1 legend for significance key.

Full size image

Amino acids

HPV-positive cell lines showed upregulation of several of the twenty standard amino acids, particularly when compared to Detroit562 (Fig. 5A). Those with significantly elevated abundance include valine (FC = 1.96, p  = 0.0026), leucine (FC = 2.51, p = 0.000029), lysine (FC = 3.04, p = 0.011), and arginine (FC = 2.64, p = 0.012) (Fig. 5A and B). Aspartate had the largest abundance in SCC47 with dramatic fold change when compared to SCC9 (FC = 12.67, p = 0.021) and Detroit 562 (FC = 19.73, p = 0.017) (Fig. 5B).

Fig. 5
figure 5

Amino acid metabolism. (A) Amino Acids: see Fig. 1 legend for color key. (B) Bar graphs show abundances (Ion count/mg protein content) of select amino acids in each cell line, with data points representing each triplicate value. See Fig. 1 legend for significance key.

Full size image

Cell line doubling times

Cell doubling times were measured for SCC47 and SCC9. Values for SCC104 and Detroit562 were obtained from prior literature. Mean values were calculated from three independent replicates. The resulting doubling times were: SCC47, 31.4 h; SCC104, 48 h21; SCC9, 36.2 h; and Detroit562, 88.5 ± 34.7 h22 (Fig. 6B).

Fig. 6
figure 6

qPCR and doubling times. (A) Relative expression of G6PD, NAMPT, NAPRT, and NADK normalized to Detroit562 = 1 for each four cell lines. Points represent individual replicate values; bars indicate mean ± SEM. G6PD = Glucose-6-phosphate dehydrogenase, NAMPT = nicotinamide phosphoribosyltransferase, NAPRT = Nicotinate phosphoribosyltransferase, NADK = NAD + Kinase. * = p-value < 0.05, ** = p-value < 0.01, *** = p-value < 0.001. (B) Doubling times of SCC47 and SCC9 plated at 15k cells per well. Points represent mean cell count; bars indicate mean ± SD.

Full size image

Quantitative real-time PCR (qPCR)

We performed qPCR analysis of selected nicotinamide and glycolysis metabolism genes (G6PD, NAMPT, NAPRT, and NADK) across the four cell lines (Fig. 6A). HPV-positive cell lines generally exhibited lower expression of NAMPT and G6PD compared to HPV-negative counterparts, while differences in NAPRT and NADK varied across lines. NAMPT expression was lowest in SCC47, followed by SCC104, and highest in Detroit562. NAMPT expression was significantly lower in both HPV-positive lines compared with HPV-negative controls (0.00053 < p < 0.0050). SCC47 demonstrates higher levels of NAD + Kinase compared to Detroit562 (p = 0.038) and SCC9 (p = 0.0056).

Nicotinamide phosphoribosyltransferase (NAMPT) inhibition

All four cell lines were treated with increasing concentrations of the NAMPT inhibitor FK866. No reduction in cell viability was observed at concentrations up to 80 nM, and HPV-positive lines did not differ in sensitivity compared to HPV-negative lines (data not shown).

Discussion

Metabolic reprogramming, characterized by altered utilization of energy pathways, has emerged as a hallmark of cancer cells, enabling sustained proliferation and tumor growth23. HPV infection specifically influences cellular metabolism by selectively activating or suppressing metabolic pathways to meet increased energy demands and biosynthetic requirements associated with tumorigenesis. A prominent example of this metabolic shift is the Warburg effect, also known as aerobic glycolysis, a phenomenon that is well-characterized across several cancer types that supports not only tumor growth, but also immune modulation, angiogenesis, and metastasis24,25. Our results reveal that aerobic glycolysis is notably magnified in HPV-positive HNSCC cells, consistent with the Warburg effect. This enhanced glycolytic phenotype aligns closely with metabolic patterns observed in HPV-positive cervical cancer8. Given the parallels in metabolic signatures among HPV-driven cancers, our findings highlight a potential opportunity to adapt existing therapies targeting glycolysis in other HPV-associated malignancies for the treatment of HPV-positive HNSCC.

Glycolysis and phosphate pentose pathway

Glucose metabolized via glycolysis produces relatively little ATP, prompting cancer cells to preferentially upregulate glycolysis to generate biosynthetic intermediates that support rapid tumor growth—such as precursors for lactate production, nucleotide synthesis via the pentose phosphate pathway (PPP), and glycogen synthesis24,26. Our results demonstrate elevation of glycolytic intermediates in HPV-positive HNSCC cells (Fig. 1A), particularly SCC47, aligning with the enhanced aerobic glycolysis characteristic of the Warburg effect observed in other solid cancers24,25. Overexpression of glucose transporters, especially GLUT-3, have been correlated with worse prognosis, tumor differentiation, size, invasion, and recurrence in laryngeal SCC27. Interestingly, the fold change of lactate did not parallel the increase in glycolytic intermediates (Fig. 1A); however, this may be mediated by the cell’s natural response of exporting lactate to maintain favorable intracellular pH, with some cancer cells shown to upregulate these transporters28.

The observed increase in glycolytic intermediates in HPV-positive cells without an equivalent increase in TCA cycle metabolites (Figs. 1A and 3A)—is consistent with activation of HIF, which blocks pyruvate from entering mitochondria and decreases mitochondrial reactive oxygen species (ROS) generation29. However, additional studies are required to directly investigate HIF activity in HPV-positive HNSCC. Furthermore, we observed increased PPP activity in HPV-positive cells, especially SCC47, which has been linked to overexpression of specific transcription factors that enhance cell proliferation and aerobic glycolysis in colon and gastric cancers30. The PPP provides essential substrates for nucleotide synthesis and produces NADPH, critical for maintaining redox homeostasis under oxidative stress conditions frequently encountered by rapidly proliferating tumor cells31. Targeting glycolysis and the PPP, potentially through inhibitors of glycolytic enzymes or glucose transporters, thus represents a promising therapeutic approach specifically for HPV-positive HNSCC.

Purine and pyrimidine metabolism

HPV-positive HNSCC cells displayed moderate upregulation of purine and pyrimidine metabolic pathways (Fig. 2A and C), reflecting increased nucleotide synthesis required to sustain DNA and RNA synthesis. Furthermore, elevated uric acid observed in SCC47 compared to Detroit562 indicates active purine catabolism, suggesting differences in proliferation rates or preferential recycling of purines back to guanosine monophosphate (GMP) and inosine monophosphate (IMP)32. UMP is an intermediate produced from glutamine, uridine, and uracil breakdown, and is necessary for synthesis of cytidine triphosphate (CTP) and thymidine monophosphate (dTMP), both essential for DNA replication and repair33. High UMP concentrations observed particularly in Detroit562 and SCC104 may reflect increased glutamine catabolism, a hallmark of proliferating cancer cells34. Although SCC47 does not reflect similar levels of UMP, SCC47 showed high levels of CTP, which functions similarly to ATP as a high-energy molecule and has been correlated with increased DNA synthesis and tumor growth in pancreatic cancer35.

Therapeutically, purine and pyrimidine metabolism pathways are attractive targets for anticancer treatments. Established chemotherapy drugs, such as methotrexate, directly inhibit key enzymes but often cause significant systemic adverse effects36,37. Emerging therapies for cancers characterized by high nucleotide metabolic rates, including pancreatic and hematologic malignancies, target selective enzymes using monoclonal antibodies, showing promise for reduced systemic toxicity and enhanced therapeutic specificity38. Given our findings, future metabolomic and transcriptomic studies could further evaluate enzyme expression and nucleotide transport mechanisms as potential therapeutic vulnerabilities within HPV-positive HNSCC.

Tricarboxylic acid cycle

The TCA cycle is central to cellular metabolism, producing electron carriers and metabolic intermediates essential for lipid, nucleotide, amino acid, and cofactor biosynthesis38. The TCA cycle in cancer cells is primarily fueled through glutaminolysis, which breaks glutamine down into intermediates that can be fed into the cycle. To support this process, cancer cells have been shown to upregulate glutamine transporters and retain an excess abundance of glutamine39. SCC47 reflected significant fold increase of glutamine (Fig. 5A and B) as well as the major breakdown product of glutaminolysis, aspartate (Fig. 5A), indicating that glutaminolysis may be upregulated in HPV-positive cell lines40. Additionally, aspartate has been implicated in regulation of redox balance, nucleotide biosynthesis, and the tumor microenvironment41.

While SCC47 did show several narrowly missed significant TCA cycle intermediates (Fig. 3A), most metabolites were similar between HPV-positive and HPV-negative cell lines. Thus, it is suspected that cancer cells with HPV still retain the same activity level of TCA cycle for energy but additionally upregulate glycolysis to further support their proliferation.

Nicotinamide metabolism

Nicotinamide metabolism is essential for the biosynthesis of NAD + and NADP + from vitamin B3 (niacin), facilitating the production of NADH and NADPH—critical electron carriers involved in key cellular processes, including the TCA cycle and fatty acid synthesis42. NADPH plays a particularly important role in maintaining cellular redox balance, protecting proliferative cells against the detrimental effects of ROS. Excess ROS, rapidly generated in highly proliferative cancer cells, can induce significant cellular and DNA damage43. Notably, nicotinamide mononucleotide (NMN), a direct NAD + precursor, was markedly elevated in HPV-positive HNSCC cells43. The substantial increase in NMN without a parallel rise in NAD + levels might indicate rapid NAD + consumption driven by NAD+-dependent enzymes involved in DNA repair, glycolysis, and lactate production, or alternatively, accelerated NMN biosynthesis.

One important biosynthetic route of NMN is the conjugation of nicotinamide with phosphoribosyl pyrophosphate (PRPP), catalyzed by nicotinamide phosphoribosyltransferase (NAMPT)44. qPCR of the four cell lines interestingly demonstrated lower expression of NAMPT in SCC47 and SCC104 compared to the HPV-negative cell lines, indicating that increased levels of NMN are potentially due to reduced NMN consumption or post-translational modifications that stabilize or enhance NAMPT activity (Fig. 6A). As a functional validation, NAMPT inhibitor FK866 was applied to the four cell lines. Cell viabilities of all four cell lines were unchanged even at the highest FK866 concentrations, supporting the conclusion that HPV-associated alterations in nicotinamide metabolism are not primarily mediated by NAMPT transcriptional regulation. Current therapeutic strategies exploring NAMPT inhibitors in combination with drugs such as lactate dehydrogenase inhibitors and cisplatin have shown efficacy in reducing growth of lymphoma and cholangiocarcinoma45,46. These raise the possibility that NAMPT inhibitors, while not sufficient as single agents, may augment the activity of other drugs in HPV-positive HNSCC. A thorough characterization of nicotinamide metabolism is needed to determine whether cells compensate by upregulating alternative pathways, such as those utilizing nicotinic acid or tryptophan, to evade NAMPT inhibition43.

SCC47 exhibited increased expression of NADK compared to other cell lines, an enzyme necessary to phosphorylate NAD + to NADP+, reflected by its increased fold change in NADP + and NADPH, although narrowly significant (FC = 4.62–6.59, p = 0.078–0.093) (Figs. 3B and 6A). NADK is crucial for downstream regulation of NADPH to regulate oxidative stress. Knockdown of NADK in colon carcinoma cells inhibits cell proliferation both in vitro and in vivo, highlighting a potential therapeutic vulnerability that could also be exploited in HPV-positive cancer cells47. Further studies are warranted to assess the role of NADK in HPV-positive HNSCC metabolism and cell proliferation.

Glutathione and methionine metabolism

Methionine metabolism plays a crucial role in cellular antioxidant capacity, primarily through oxidation of glutathione (GSH) to glutathione disulfide (GSSG). The GSH:GSSG ratio serves as an indicator of oxidative stress, with increased GSSG levels reflecting elevated oxidative pressure commonly associated with rapid cell proliferation48. Among the tested cell lines, SCC47 exhibited the lowest GSH:GSSG ratio, indicating heightened oxidative stress and rapid glutathione consumption to counteract ROS generation. In contrast, the other three cell lines displayed relatively higher GSH:GSSG ratios, suggesting more effective antioxidant buffering48.

Interestingly, SCC47 demonstrated uniquely elevated levels of cystine, a dimeric form of the amino acid cysteine (Fig. 4B). Cancer cells preferentially import extracellular cystine over cysteine via the cystine-glutamate antiporter System XC (SLC7A11/xCT). Cystine is then converted to cysteine using NADPH, fueling glutathione synthesis to manage intracellular oxidative stress49,50. Inhibition of System XC can induce lethal oxidative stress in cancer cells reliant on glutathione synthesis, highlighting this pathway’s viability as a therapeutic target50. The elevated cystine uptake observed in SCC47 might represent a compensatory mechanism to maintain sufficient glutathione levels and prevent oxidative stress-induced cell death, a vulnerability previously identified in cancer cells, though future validation on SLC7A11 gene expression in HNSCC is needed50. Given these findings, targeting cystine uptake via System XC, NADPH regeneration, or intracellular cysteine availability represents potential therapeutic strategies against aggressively proliferating HPV-positive cancer cells.

Amino acids

Amino acids are widely used in a diverse range of reactions, including their immediate role as a substrate for protein synthesis, and reprogramming of their metabolism is crucial for continued survival of cancer cells. Pancreatic ductal adenocarcinoma cells (PDAC) have been shown to exhibit a process known as macropinocytosis, in which the cell consumes a portion of the microenvironment and degrades the proteins into free amino acids for further use40. Given the broad roles of amino acids and their generalized increased abundance in HPV-positive cells, it is of interest to determine whether HPV-driven enhancement of amino acid acquisition utilizes a similar method, potentially exploring inhibitory mechanisms that can simultaneously modulate multiple metabolic pathways.

Limitations

While our study identified distinct metabolic signatures distinguishing HPV-positive from HPV-negative HNSCC, several limitations warrant consideration. First, our analytical methods were limited in detecting and quantifying larger biomolecules, restricting comprehensive assessments of certain pathways, particularly lipid metabolism. Additionally, although 2D cell culture models provide clear advantages in reproducibility and accessibility, they do not fully represent the complexity of the tumor microenvironment observed in vivo. In vitro conditions allow cancer cells unrestricted access to nutrients and oxygen, contrasting sharply with the spatial constraints, nutrient gradients, hypoxia, and cell-cell interactions encountered within tumors in vivo51. Cell-cell interactions, particularly involving cancer-associated fibroblasts (CAFs), significantly influence tumor biology, signaling, and metabolic regulation. Thus, future studies employing three-dimensional culture systems and co-culture models with CAFs specific to HPV-positive tumors would enhance biological relevance and translational potential52. Furthermore, extending our metabolic analyses to patient-derived biofluids such as saliva or serum will be essential for clinically validating these findings and further elucidating direct metabolic markers of HPV-driven carcinogenesis12. Lastly, our measured doubling time for SCC47 (31.2 h) was faster than reported in the literature (52.8 h)53, raising the possibility that cell line–specific proliferation rates, rather than HPV status alone, may contribute to some of the observed metabolic differences.

Conclusion

In summary, our findings reveal distinct metabolic characteristics differentiating HPV-positive from HPV-negative HNSCC cell lines. The HPV-positive cells, particularly SCC47, showed significant elevations in glycolytic intermediates, nucleotide synthesis, nicotinamide metabolism, and antioxidant demand, reflective of their rapid proliferation and increased oxidative stress. SCC104, while exhibiting similar trends, displayed these metabolic changes to a lesser degree, consistent with its comparatively slower growth rate. In contrast, HPV-negative cell lines primarily utilize the TCA cycle, suggesting reliance on oxidative metabolism and lipid synthesis. These metabolic differences represent promising therapeutic vulnerabilities and may guide the development of targeted treatments tailored specifically to HPV-driven HNSCC.

Methods

Materials, cell lines, sample preparation

HPV-positive cell lines (SCC47 and SCC104) and HPV-negative cell lines (Detroit562 and SCC9) were obtained from Millipore Sigma-Aldrich (St. Louis, MO, USA) and cultured in biological triplicates21,53,54,55. The cells were cultured in media recommended by previous studies. All cell lines were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 1% penicillin-streptomycin, and 1% nonessential amino acids. Cells were cultured as adherent monolayers at 37 °C in a humidified atmosphere with 5% CO₂. The cells were passaged and plated on 6-well plates in triplicates at a density of 300,000 cells per well. After 72 hours and at 70–90% confluency, the cells were harvested for metabolomics analysis.

Ultra-performance liquid chromatography - mass spectrometry (UPLC-MS)

The harvested cells were washed twice with cold PBS and fixed in 80% methanol (Fisher Chemical, LC-MS grade) precooled in dry ice. Cells were detached by scraping and then transferred to a 2-mL screw cap tube and dried using a nitrogen evaporator. Dried metabolites were resuspended in acetonitrile/methanol/water (40:40:20). Before applying LC-MS, metabolites were mixed with acetonitrile containing 0.2% folic acid solution and spun down at 12,000 rpm for 10 min at 4 °C. The supernatant was injected into Agilent Accurate Mass 6230 time-of-flight (TOF) coupled with Agilent 1290 LC system. Detected ions were deemed metabolites on the basis of unique accurate mass-retention time identifiers for masses exhibiting the expected distribution of accompanying isotopologues56.

Data quantification, normalization, and extraction

Area under the curve was used to quantify the abundance for each metabolite. Abundances were then normalized to raw protein concentration and compounds were identified using comparison to curated metabolite libraries from the Kyoto Encyclopedia of Genes and Genomes (KEGG)57,58,59. The protein concentration of metabolite extracts was measured with a bicinchoninic acid (BCA) protein assay kit (Thermo Scientific, Waltham, MA, USA) to normalize samples to cell biomass60. Values that were missing were manually input as the minimal measured value. The abundance of metabolites was extracted using Agilent Qualitative Analysis B.07.00 and MassHunter Profinder B.06.00 software (Agilent Technologies, Santa Clara, CA, USA) with a mass tolerance of < 0.005 Da.

Statistical analysis

All statistics were completed using the natural log transformation of data normalized to raw protein concentration. This was done due to the well-known right skew of metabolomics data61. Cell line comparisons were conducted using Student’s two-tailed t-test. To acknowledge the false discovery rate, all p-values were then corrected using the Benjamini-Hochberg procedure with only readjusted p-values < 0.05 considered significant62. Additionally, fold change cut off was set to 1 ± 0.5 such that fold changes in the range 0.5–1.5 were not considered in analysis. Metabolites were considered significant if they met both criteria for fold change cutoff and readjusted p-value. Figures were generated using Prism by GraphPad Software (Boston, MA, USA).

Cell line doubling time

For each cell line, cells were seeded in 24-well plates at 1.5 × 103 cells per well in triplicates and maintained in standard culture medium. At 24-hour intervals over three consecutive days, cells were detached using trypsinization and counted by hemocytometer. Raw counts were adjusted for dilution factors to obtain total cell numbers. Mean cell counts at each time point were used to calculate population doubling times using the equation: [Time Duration * ln(2)]/[ln(Final Cell Count) - ln(Initial Cell Count)].

Quantitative real-time PCR (qPCR)

Total RNA was isolated using the RNeasy Plus Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol. cDNA was synthesized by reverse transcription via the LunaScript RT SuperMix Kit (New England Biolabs, Ipswitch, MA, USA). Quantitative PCR was performed with PowerUp SYBR Green Master Mix (Thermo Fisher Scientific, Waltham, MA, USA) on a Mastercycler nexus X2 (Eppendorf, Hamburg, Germany). Primers were obtained from Origene (Rockville, MD, USA), except ACTB, which was adopted from Racz et al.63. The sequences of all primers are provided in Supplementary Table 1. Relative gene expression levels were calculated using the ΔΔCt method, normalized to ACTB as the internal control. Relative transcript expression levels compared to Detroit562 were reported. Student’s two-tailed t-test and Benjamini-Hochberg correction were applied to achieve corrected p-values, with p-values < 0.05 considered significant.

NAMPT inhibition

All four cell lines were seeded into 96-well plates and allowed to adhere for 24 h. Cells were then treated with ten concentrations of FK866 (Millipore Sigma-Aldrich, St. Louis, MO, USA) in a two-fold dilution series (80, 40, 20, 10, 5, 2.5, 1.25, 0.625, 0.3, and 0 nM), prepared in 100 µL of culture medium. After 72 h of treatment, cell viability was assessed using the WST-1 assay. WST-1 reagent was diluted 1:10 in culture medium, and 100 µL of the mixture was added to each well. Plates were incubated at 37 °C for 1 hour, and absorbance was measured at 450 nm using a microplate reader (SpectraMax iD3, Molecular Devices, San Jose, CA, USA).