MiR-181a-5p/CDK6 axis attenuates cell proliferation, migration and glycolytic reprogramming in oral squamous cell carcinoma

Abstract

Background

Metabolic reprogramming is a hallmark of rapid tumor cell progression. Cellular cyclin-dependent kinase 6 (CDK6) has been reported to exhibit abnormal expression in oral squamous cell carcinoma (OSCC). However, the detailed functional role of CDK6 in OSCC glycolytic reprogramming is still unknown and requires further investigation. Studies have demonstrated that miR-181a-5p plays a significant role in the malignant progression of OSCC. Therefore, this study aimed to elucidate the molecular mechanisms underlying the role of the miR-181a-5p/CDK6 axis during glycolysis in OSCC.

Methods

The expression and functional testing of miR-181a-5p and CDK6 were conducted both in vivo and in vitro, encompassing cell proliferation, migration, and glycolysis-related metabolites. The association between miR-181a-5p expression and clinicopathological factors as well as the overall survival was analyzed. A dual-luciferase reporter assay was used to confirm the relationship between miR-181a-5p and CDK6. Quantitative real-time polymerase chain reaction(qRT-PCR) and western blotting were performed to detect gene and protein expression, respectively.

Results

CDK6 knockdown significantly inhibited OSCC cell proliferation, migration, glucose consumption, lactate, ATP production, and downregulated the expression or catalytic activities of HK2, PFKM, PKM2, LDHA, PHGDH, PSAT1, but not G6PD. Overexpressed miR-181a-5p increased overall survival (OS). CDK6 was a direct target of miR-181a-5p. CDK6 overexpression phenotypically rescued the inhibitory effect of miR-181a-5p on the proliferation, migration, and glycolytic reprogramming of OSCC cells.

Conclusions

These findings underscore the critical role of CDK6 in the OSCC progression and provide the first evidence that miR-181a-5p regulates CDK6, thereby suppressing glycolysis and the serine-glycine biosynthesis pathway (SSP) in OSCC. However, miR-181a-5p attenuates the pentose phosphate pathway (PPP) via CDK6-independent targets.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12935-026-04171-z.

Background

Oral squamous cell carcinoma (OSCC), a prevalent form of head and neck cancer, is the sixth most common malignant tumor worldwide and the eighth leading cause of cancer-related mortality. This poses a substantial global health challenge in several regions [1]. Although significant advances have been made in diagnosis, surgery, radiation, and chemotherapy in recent years, the mortality rate associated with OSCC remains high [2]. This is primarily due to the absence of early detection markers, as well as issues with recurrence and metastasis, which have impeded the achievement of satisfactory outcomes. This comprehensive perspective is essential for developing targeted therapies and improving patient outcomes, recent advancements in research have illuminated the OIP5-AS1 is a critical player by influencing tumor dynamics and mechanisms of resistance in the progression of OSCC [3], Therefore, there is an urgent need to gain a deeper understanding of the molecular mechanisms that drive OSCC progression. These insights may pave the way for the development of effective therapeutic strategies..

Aerobic glycolysis, also known as the "Warburg effect,” is a pivotal component of metabolic reprogramming and plays a crucial role in the onset and progression of tumors. This process is the primary pathway for glucose utilization by tumor cells that do not undergo oxidative phosphorylation. Instead, the generated pyruvate is converted to lactate [4]. Accumulating evidence has demonstrated that tumor cells rely on aerobic glycolysis for their growth, invasion, and angiogenesis [5–8]. Evidence suggests that aerobic glycolysis is widespread in OSCC and is closely related to tumorigenesis. For example, the upregulation of protein disulfide isomerase family 6 (PDIA6) promotes aerobic glycolysis, thereby contributing to the aggressive behavior of OSCC [9]. Enolase 2 (ENO2) silencing inhibits glycolysis and attenuates OSCC cells [10]. These findings suggest that targeting aerobic glycolysis represents a promising therapeutic approach for the treatment of OSCC. Therefore, by interfering with the metabolic reprogramming that sustains tumor growth and progression, it is feasible to significantly decelerate or potentially arrest disease development. Achieving this goal requires a comprehensive understanding of the molecular mechanisms underlying aerobic glycolysis in OSCC cells.

CDK6 plays a vital role in regulating cell cycle progression and is indispensable for cell progression to the G1 phase [11]. Recently, the involvement of CDK6 in metabolic regulation has attracted increasing attention. Eukemic cells with kinase-inactive CDK6 show an increased need for aerobic glycolysis [12]. Inhibition of the cyclin D3-CDK6 complex in tumor cells has been shown to diminish flux through PPP and SSP, leading to the depletion of antioxidants such as nicotinamide adenine dinucleotide phosphate (NADPH) and glutathione [13]. Furthermore, Liu et al. suggested that CDK6 enhances glycolysis in colorectal cancer (CRC), with hexokinase 2 (HK2) potentially serving as a downstream target [14]. These findings indicate that the function of CDK6 extends beyond cell cycle regulation, encompassing metabolic modulation in cancer cells. Some studies have highlighted the elevated expression of CDK6 in OSCC tissues compared to that in normal mucosa, underscoring its crucial role in oncogenesis and positioning it as a potential therapeutic target[11–15]. However, the specific role of CDK6 in the regulation of OSCC cell metabolism remains unclear and requires further investigation. Uncovering the intricate interplay between CDK6 and metabolic pathways in OSCC may reveal novel therapeutic avenues for targeting this aggressive malignancy.

Collectively, these findings underscore the pivotal role of CDK6 in OSCC development, positioning it as a promising antitumor target. Previous results from our study also demonstrated that miR-181a-5p inhibits OSCC cell proliferation, colony formation, migration, invasion, and cell cycle progression, while promoting apoptosis by modulating its downstream targets [16]. This miRNA has emerged as a key player in cancer development and progression, functioning as an oncomiR or tumor suppressor in diverse cancer types and impacting multiple tumor hallmarks [17]. Using bioinformatics tools, such as TargetScan and miRTarBase, we predicted a targeted binding relationship between miR-181a-5p and CDK6. However, the precise mechanism through which miR-181a-5p influences glycolysis in OSCC by regulating CDK6 expression remains unclear.

In this study, we investigated the function of CDK6 in OSCC cells, validated the targeted binding between miR-181a-5p and CDK6, and elucidated the mechanism through which miR-181a-5p/CDK6-mediated aerobic glycolysis underlies OSCC tumorigenesis. Our findings provide invaluable insights, strengthening the understanding of the complex mechanisms involved in OSCC pathogenesis, and potentially paving the way for novel therapeutic strategies.

Methods

Cell culture

Two human OSCC cell lines, CAL27 and SCC9, and 293 T cells were obtained from Boster Biological Technology Co., Ltd. (Hubei, China). The normal epithelial keratinocyte line (HOK) was donated by the Key Laboratory of the School of Stomatology, Shanxi Medical University.

Cell transfection

siRNAs used to downregulate CDK6 (si-CDK6) and their negative controls (si-NC) were designed and synthesized. The primer sequences used to target CDK6 with si-CDK6 (1), si-CDK6 (2), si-CDK6 (3), and negative control siRNA (si-NC) are shown in Table S1. miR-181a-3p mimics (5′-AACAUUCAACGCUGUCGGUGAGU-3′and 5′-UCACCGACAGCGUUGAAUGUUUU-3′) and corresponding negative controls (NCs) (5′-UUCUCCGAACGUGUCACGUTT-3′and 5′ACGUGACACGUUCGGAGAATT-3′) were synthesized by GenePharma (Shanghai, China). CDK6 overexpression (oe-CDK6) plasmid and negative control overexpression were purchased from GenePharma (Shanghai, China). si-CDK6, si-NC, miR-181a-5p mimics and negative controls, and oe-CDK6 plasmid and negative control were introduced into cells through cell transfection using the TransIntro® EL reagent. (TransGen Biotech, China). Infected cells were incubated for 4 h before medium exchange.

RNA extraction and quantitative real-time polymerase chain reaction (qRT-PCR)

Total RNA from the isolated cells or tissues was extracted via the TRIzol method (TaKaRa, Japan) and then analyzed for both concentration and purity via spectrophotometry. miRNA cDNA was generated using the Mir-X™ miRNA First Strand Synthesis Kit (Takara, Japan), and the cDNA from mRNA was synthesized using the Prime Script™ RT Master Mix kit (Takara, Japan). Real-time PCR was performed using SYBR Green PCR Master Mix (Takara Bio Inc., Japan). U6 and β-actin were used as internal controls. The relative expression of target genes was represented via the 2^−ΔΔCt method. The primer sequences used in this study are listed in Supplementary Table S1.

Western blot

Whole cell extracts were lysed with RIPA lysis buffer containing protease inhibitors. Protein concentration was determined using a BCA assay kit (TaKaRa, China). A 10 μg protein sample obtained from each group was separated via 10% sodium dodecyl sulfate‒polyacrylamide gel electrophoresis (SDS‒PAGE), followed by transformation into polyvinylidene difluoride (PVDF) membranes. The membranes were blocked and probed overnight at 4℃ with specific primary antibodies (1:1000 dilution), including anti-CDK6 (ABclonal, A1545), anti-HK2 (ABclonal, A0466), anti-6-phosphofructokinase 1 muscle isoform (PFKM) (ABclonal, A5477), anti-pyruvate kinase M2 (PKM2) (ABclonal, A0268), anti-lactate dehydrogenase A (LDHA) (ABclonal, A1146), anti-glucose-6-phosphate dehydrogenase (G6PD) (ABclonal, A1537), anti-phosphoglycerate dehydrogenase (PHGDH) (ABclonal, A10461), anti-phosphoserine aminotransferase 1 (PSAT1) (BOSTER, PB10086), and anti-β-actin (Servicebio, K101527P) antibodies. After that, the membranes were probed with secondary antibodies (1:5000 dilution, Boster) for 1 h at 37℃. Protein signals were detected using GBOX-CHEMI-XX9-E (Syngene, UK) after incubation with ECL (Boster).

Cell proliferation and migration experiments

In this study, a CCK-8 kit (BOSTER, China) was used to detect the effect of CDK6 on the proliferation of OSCC cells. Five replicate wells were used for each group. Cells were cultured for 0, 24, 48, and 72 h after transfection, followed by the addition of 10% CCK-8 solution. The cells were incubated for 1.5 h at 37 °C. The optical density (OD) of the cells was measured at 450 nm using a microplate reader (Biotek, USA).

To investigate the effects of CDK6 on the migratory ability of OSCC cells, a scratch wound assay was performed lateral migration characteristics. Replicate wells were used for each group. OSCC cells were seeded in 6-well plates. After transfection for 4 h, wounds were generated via the tip of a 10 µL pipette, and the cells were washed to remove cell debris. The cells were subsequently placed in serum-free DMEM and further cultured in an incubator at 37 °C. The area between the wounds was measured at different time points (0, 24, and 48 h), and the relative migration rate was calculated using the ImageJ software.

Data sources used for analysis

RNA sequencing (RNA-seq) and clinical data of head and neck squamous cell carcinoma (HNSC) were obtained from the TCGA databases (https://portal.gdc.cancer.gov). Among these cases, clinical samples from the oral cavity (oral tongue, floor of the mouth, base of the tongue, hard palate, alveolar ridge, oral cavity, and buccal mucosa) comprised 341 OSCC and 32 matched normal oral mucosal epithelial tissues. The raw count data was used for the differential expression analysis of miR-181a-5p. Moreover, The miR-181a-5p mRNA expression and clinical data in the GSE45238 dataset (containing 40 pairs of OSCC matched samples) from the Gene Expression Omnibus (GEO) databases was downloaded. The datasets were analyzed using R software (version 4.2). The correlation between the expression of the glycolytic genes and CDK6 was assessed using Pearson’s correlation. The association between miR-181a-5p expression and clinicopathological features in patients with OSCC was analyzed using Fisher’s exact test. The p-value, OS(overall survival), hazard ratio (HR) and 95% confidence interval was determined by Cox regression analysis.

Glycolysis-related metabolite assays

OSCC cells (2 × 10⁶/well) were seeded in 6-well plates. After transfection, cells and culture supernatants were collected. A glucose kit (A154-1–1; Nanjing Jiancheng, China) was used to detect glucose uptake. A lactic acid assay kit (A019-2–2; Nanjing Jiancheng, China) was used to measure lactate production. ATP levels were measured using an ATP assay kit (A095-1–1; Nanjing Jiancheng, China). A hexokinase (HK) activity assay kit (BC0745; Solarbio, China) was used to analyze cellular HK activity. A phosphofructokinase (PFK) activity assay kit (BC0535; Solarbio, China) was used to analyze cellular PFK activity. A pyruvate kinase (PK) activity assay kit (BC0545, Solarbio, China) was used to analyze the cellular PK activity. A pyruvate acid (PA) content assay kit (BC2205, Solarbio) was used to analyze cellular PA levels, and a coenzyme II NADP (H) content assay kit (BC1105, Solarbio) was used to detect cellular NADPH levels. All results were normalized to the total protein concentration.

Dual-luciferase reporter gene detection

Putative miR-181a-5p targets were predicted using several bioinformatics tools, including TargetScan (http://www.targetscan.org/) and miRTarBase (http://mirtarbase.mbc.nctu.edu.tw). The 3'UTR fragment of the CDK6 gene containing the miR-181a-5p binding site was cloned and inserted into the Xhol and Notl sites of the dual-luciferase vector, and the plasmid was verified to be constructed correctly via double zymography and sequencing. The obtained sequences were aligned with the wild-type target sequence to confirm the successful construction. The target fragment was ligated to the vector using the HB infusion™(HANBIO, China) one-step cloning ligation system, and the plasmid was extracted via an Endo-free Plasmid Mini Kit II (Omega, USA) after ligation product transformation. The target plasmid was transfected into 293 T cells purchased from Boster Biological Technology Co., Ltd. (Hubei, China). After transfection for 48 h, a luciferase reporter assay was performed using a Dual-Luciferase Reporter Assay System.

Construction of the stable cell line of miR-181a-5p overexpression

Lentivirus vectors carrying overexpress miR-181a-5p (called OE), and their negative controls (NC) entrusted by Shanghai GeneChem Co., Ltd (Shanghai, China). Following lentivirus particle packaging, the optimal multiplicity of infection (MOI) was set at 10, according to well established protocols reported in previous studies. Meanwhile, a systematic puromycin concentration gradient screening assay was conducted to determine 5 μg/mL as the optimal selection concentration for the target cell model. To establish transfected LV-miR-181a-5p and LV-NC cell lines, CAL-27 cells were cultured in medium supplemented with 5 μg/mL puromycin for resistance selection. Finally, Transfection efficiency was comprehensively validated through two complementary approaches: evaluating the proportion of fluorescence-positive cells using laser scanning confocal microscopy and quantifying miR-181a-5p expression levels via qRT-PCR (Fig. S3A-C).

In vivo animal assays

Female NOD-PrkdcscidIl2rgem1/Smo (M-NSG) mice (4 weeks old, 16–20 g) were divided into miR-181a-5p OE (n = 6) and NC (n = 6) groups. The cell suspension was disaggregated with a 1 mL syringe, and 12 mice were subcutaneously injected with transfected CAL-27 cells. The weights of the mice and tumor volumes were calculated every 2 days beginning 6 days after the inoculation of the transfected cells into the mice. Then, the mice were sacrificed, the xenogeneic tumors were harvested and weighed, and some tissues were used for histopathological analyses and gene expression assays. The animal experimental protocol was approved by the Animal Care and Use Committee of Shanxi Medical University(SYDL2023017).

Hematoxylin and eosin (H&E) staining

The tumors and liver, and lung organs were fixed in 4% paraformaldehyde, dehydrated, cleared, and embedded in paraffin. The tissues were cut into 4 μm-thick sections. Sections were dewaxed in xylene, dehydrated by gradient elution in ethanol, and stained with hematoxylin and eosin. Finally, the sections were observed under a microscope.

Immunohistochemistry staining

Paraffin-embedded tissue sections of tumor tissues were used for IHC staining to detect the expression levels of Ki-67 and caspase-3. After the tissue samples were deparaffinized and rehydrated with ethanol, antigen retrieval was performed by heating in citrate buffer (pH 6.0) for 5 min. Endogenous peroxidase activity was blocked by incubation with 3% H₂O₂ for 10 min. The sections were incubated with diluted primary antibodies against Ki-67 (GB151499, Servicebio, China, 1:500) and caspase-3 (BS-0081R, ThermoFisher, US, 1:300) overnight at 4 ℃ in a humidified chamber. After that, the sections were incubated with a horseradish peroxidase (HRP)-conjugated secondary antibody at room temperature for 1 h. The images were quantitatively analyzed and automatically scored using the IHC profiler of Image J.

Statistical analysis

The experiments were independently repeated at least three times. SPSS Statistics 22.0 software (IBM) was used for data analysis. All the quantitative data are presented as the means ± standard deviations (SDs). Statistical differences between the two groups were analyzed using a two-tailed unpaired Student’s t-test; for multiple groups, a one-way analysis of variance (ANOVA) with Tukey’s multiple comparisons tests were performed. The Pearson’s correlation was used to assess the correlation between the expression of the glycolytic genes and CDK6. Chi-square tests were used to analyze the relationship between miR-181a-5p expression levels and clinical characteristics. A Cox proportional hazards model was used to estimate the HR of each clinicopathological variable for OS. P values were considered statistically significant at P < 0.05.

Results

CDK6 was up-regulated in OSCC

To elucidate the influence of CDK6 on OSCC, we first conducted a search in the GEPIA database (http://gepia.cancer-pku.cn). After comparing the expression levels of CDK6 in 519 head and neck squamous cancer (HNSC) patients and 44 normal individuals, we found that the CDK6 expression in HNSC tissues was significantly higher than that in normal tissues. (Fig. 1A). Furthermore, CDK6 expression levels were correlated with patient prognosis, with high levels of CDK6 expression generally associated with a poorer prognosis (Fig. 1B). A previous study showed that the mRNA (P < 0.0001) and protein (CAL27, P = 0.0133; SCC9, P = 0.0097) expression levels of CDK6 were significantly higher in CAL27 and SCC9 cells than in HOK cells (Fig. 1C, D). These results suggest that CDK6 may undergo aberrant activation during the course of OSCC and play a crucial role in the development of OSCC.

Fig. 1.

Interference with CDK6 inhibits the proliferation and migration of OSCC cells. (A) CDK6 expression levels in 519 HNSC patients and 44 normal individuals. (B) CDK6 expression levels are correlated with patient prognosis. (C) Relative mRNA expression of CDK6 in SCC9 and CAL27 cells. (D) Relative protein expression of CDK6 in SCC9 and CAL27 cells. (E, F) The effects of CDK6 interference on in vitro proliferation (E) and migration (F) in SCC9 and CAL27 cells. The data are representative of three independent experiments. The data represent the mean ± SD. ns: not significant (P ≥ 0.05), *P < 0.05, **P < 0.01, ***P < 0.001

Interference with CDK6 significantly inhibited the proliferation and migration of OSCC cells

We validated the biological role of CDK6 in OSCC. We used siRNA to inhibit the expression of CDK6 and selected si-CDK6(2) for subsequent experiments (P < 0.0001) (Fig. S1A, B). The cell proliferation assay showed that knockdown of CDK6 expression significantly inhibited the proliferation of OSCC cells compared to the negative control and blank control groups at 48 h and 72 h (Fig. 1E). In addition, knockdown of CDK6 expression significantly inhibited OSCC cell migration (Fig. 1F). These results suggest that CDK6 is involved in the regulation of cell biological behaviors in OSCC cells, and that interfering with this gene can significantly inhibit the proliferation and migration of OSCC cells.

Associations between CDK6 and glycolytic/SSP/PPP gene expression in OSCC cohorts

The correlation between the expression of the glycolytic genes and CDK6 was assessed using Pearson’s correlation. CDK6 showed a significant positive correlation with the expression of several genes, including HK2, PFKM, PKM, and LDHA (P < 0.001) (Fig. 2A–D), no significant correlation with PSAT1 and G6PD (P > 0.05) (Fig. 2F, G), and a significant negative correlation with PHGDH(P < 0.001) (Fig. 2E).

Fig. 2.

Glycolytic genes expression and correlated with the expression of CDK6 in OSCC. (A–G) Spearman correlation analyses between CDK6 and HK2, PFKM, PKM, LDHA, PHGDH, PSAT1or G6PD at the mRNA expression levels were conducted through TCGA dataset mining. (H) Relative protein expression levels of HK2, PKM2, PFKM, and LDHA in SCC9 and CAL27 cells following knockdown of CDK6. (I) Relative protein expression levels of PHGDH and PSAT1 in SCC9 and CAL27 cells following knockdown of CDK6. (J) Relative protein expression of G6PD in SCC9 and CAL27 cells following CDK6 knockdown. Data are representative of three independent experiments. The data represent the mean ± SD.ns: not significant (P ≥ 0.05),^*P < 0.05, ^**P < 0.01, ^***P < 0.001

Interfering with CDK6 disrupts the glycolysis and SSP in OSCC cells, yet it does not affect the PPP

The effect of CDK6 on glycolysis was also investigated. We measured the concentrations of key metabolites of the glycolytic pathway, including glucose (CAL27, P < 0.0001; SCC9, P = 0.0009), pyruvate (CAL27, P < 0.0001; SCC9, P < 0.0001), lactate (CAL27, P < 0.0001; SCC9, P < 0.0001), and ATP (CAL27, P = 0.0082; SCC9, P = 0.0002). The results revealed that interfering with CDK6 expression significantly increased the extracellular glucose content in CAL27 and SCC9 cells (Fig. 3A), and inhibited the production of PA, lactate, and ATP in CAL27 and SCC9 cells (Fig. 3B–D). In addition, knockdown of CDK6 expression significantly downregulated the protein expression levels of key enzymes involved in the glycolytic pathway, including HK2 (CAL27, P = 0.0051; SCC9, P = 0.0007), PFKM (CAL27, P = 0.0054; SCC9, P = 0.0068), and LDHA (CAL27, P = 0.0025 and SCC9, P = 0.0455) (Fig. 2H). Furthermore, interfering with CDK6 expression significantly decreased the enzymatic activities of HK, PFK, and PK (Fig. 3E–G). Collectively, interference with CDK6 expression suppresses the glycolytic pathway in OSCC.

Fig. 3.

Contents of glycolysis-related metabolites and enzyme activities with the expression of CDK6 in OSCC. (A–D) Glucose concentration (A), pyruvate concentration (B), lactic acid concentration (C) and ATP concentration (D) in SCC9 and CAL27 cells after knockdown of CDK6. (E–H) Activities of PK, PEK, HK, and NADPH in SCC9 and CAL27 cells after knockdown of CDK6. The data are representative of three independent experiments. The data represent the mean ± SD. ns: not significant (P ≥ 0.05),^*P < 0.05, ^**P < 0.01, ^***P < 0.001

The SSP is a glycolytic branching pathway that consumes glucose carbons for serine makeup and α-ketoglutarate generation [18]. To interpret the influence of CDK6 on the SSP of OSCC cells, the concentration of NADPH was measured after CDK6 expression was changed. The results revealed that interference with CDK6 expression significantly reduced NADPH production in OSCC cells (CAL27, P < 0.0001; SCC9, P < 0.0001) (Fig. 3H). We also explored the effects of CDK6 on PHGDH and PSAT1, key enzymes involved in the serine synthesis pathway. The results showed that interference with CDK6 expression significantly downregulated the protein expression levels of PHGDH (CAL27, P = 0.0010; SCC9, P = 0.0016) and PSAT1(CAL27, P = 0.0009; SCC9, P < 0.0001) in OSCC cells compared to those in the si-NC group (Fig. 2I). These results demonstrate that interference with CDK6 expression inhibits the serine synthesis pathway in OSCC. PPP, which branches from glycolysis, is required for the synthesis of ribonucleotides and is a major source of NADPH [19]. G6PD is the only rate-limiting enzyme in the PPP. Western blotting experiments revealed that interference with CDK6 expression did not significantly decrease the protein expression level of G6PD in OSCC cells compared to that in the si-NC group (CAL27, P = 0.5011; SCC9, P = 0.0735) (Fig. 2J). Collectively, these data verified that CDK6 downregulation did not affect the PPP in OSCC.

The expression of miR-181a-5p and relevant prognosis in oral cancer

To verify the expression of miR-181a-5p, previous results have also shown that miR-181a-5p was downregulated in human OSCC cell lines and human OSCC specimens.[16]. Analysis of the published dataset (GSE45238) confirmed that miR-181a was significantly downregulated in OSCC tissues compared with normal tissues (Fig. S2 A). The experimental results also revealed that miR-181a-5p was significantly downregulated in OSCC cells compared to HOK cells (CAL27 and SCC9, P < 0.0001) (Fig. 4A). To assess the association between miR-181a-5p expression and prognosis in oral cancer, a survival analysis revealed a significantly reduced OS in oral cancer patients with low miR-181a-5p expression (P = 0.009) (Fig. S2 B). Furthermore, to evaluate the correlation between miR-181a-5p expression and different clinicopathological features in patients with oral cancer, we investigated the association between miR-181a-5p expression and clinicopathological parameters in oral cancer patients. As Supplementary Table 2 revealed, the expression of miR-181a-5p was significantly correlated with the Pathologic T stage (P = 0.005; χ²-test). miR-181a-5p has a lower expression level at the pathological stage T3. By contrast, there was no relationship between the miR-181a-5p expression and patients’ age, gender, smoking status, PNM stage, clinical TNM stage, and tumor size. Consequently, these results indicated that low miR-181a-5p expression was likely to be associated with later-stage local tumor infiltration in oral patients.

Fig. 4.

Overexpression of miR-181a-5p suppresses proliferation and migration of OSCC cells by targeting CDK6. (A) Relative mRNA expression of miR-181a-5p in SCC9 and CAL27 cells. (B) Relative luciferase activities of plasmids carrying WT-CDK6-3’UTR or Mut-CDK6-3’UTR cotransfected with miR-181a-p mimics. (C, D) Relative mRNA and protein expression of CDK6 after transfection with the miR-181a-5p mimic and the corresponding CDK6-plasmid. (E, F) Effects on proliferation (E) and migration (F) of SCC9 and CAL27 cells after transfection with the miR-181a-5p mimic and the corresponding CDK6-plasmid. Data are representative of three independent experiments. The data represent the mean ± SD. ns: not significant (P ≥ 0.05),^*P < 0.05, ^**P < 0.01, ^***P < 0.001

Univariate Cos regression analysis showed that Pathologic T3 (HR, 4.194; 95% CI, 1.772–9.930; p = 0.0011) and T4 stage (HR, 4.081; 95% CI, 1.753–9.500; p = 0.0011), Pathologic N2 & N3 stage (HR, 2.344; 95% CI, 1.592–3.452; p = 0.00), and Clinical N2 & N3 stage (HR, 1.548; 95% CI, 1.073–2.234; p = 0.0196) were risk factors for OS in OSCC patients. Multivariate Cox regression analysis showed the Age (HR, 1.655; 95% CI, 1.138–2.406, p = 0.0084) was an independent risk factor for patient prognosis (Table 1).

Table 1.

Univariate and multivariate analysis of various prognostic parameters in patients with OSCC for OS

	Univariate analysis			Multivariate analysis
	P-value	Hazard Ratio (HR)	95% confidence interval	P-value	Hazard Ratio (HR)	95% confidence interval
Pathologic T stage
T1
T2	0.1232	1.977	0.831–4.702	0.3834	2.493	0.320–19.446
T3	0.0011	4.194	1.772–9.930	0.2054	3.726	0.487–28.524
T4	0.0011	4.081	1.753–9.500	0.1678	4.267	0.543–33.543
Pathologic N stage
N0
N1	0.2962	0.717	0.384–1.338	0.1767	0.587	0.271–1.272
N2&N3	1.6e-05	2.344	1.592–3.452
Clinical T stage
T1
T2	0.8484	1.081	0.486–2.403
T3	0.2459	1.606	0.721–3.576
T4	0.3988	1.401	0.640–3.068
Clinical N stage
N0
N1	0.8784	1.034	0.671–1.595	0.7873	0.930	0.547–1.579
N2&N3	0.0196	1.548	1.073–2.234	0.9454	0.983	0.601–1.607
Clinical M stage
M0
M1	0.3274	2.679	0.373–19.247
Gender
Female
Male	0.4695	0.885	0.636–1.232
Age
< = 60
> 60	0.0579	1.364	0.990–1.880	0.0084	1.655	1.138–2.406
Smoker
No
Yes	0.1740	1.306	0.889–1.920
hsa-miR-181a-5p
Low
High	0.1796	0.804	0.585–1.106

CDK6 is a direct target of miR-181a-5p in OSCC cells

Based on the dual-luciferase reporter assay, The experimental results demonstrated that overexpression of miR-181a-5p significantly suppressed the luciferase activity of the reporter gene with the wild type construct but not with the mutant CDK6 3’-UTR construct (Fig. 4B). CAL27 and SCC9 cells were treated with miR-81a-5p mimic, miR-81a-5p mimic + oe-CDK6, or the corresponding NC. The expression of CDK6 was significantly reduced in both the miR-81a-5p-mimic and miR-81a-5p-mimic + oe-CDK6 groups, whereas the expression of CDK6 was significantly elevated in the miR-81a-5p-mimic + oe-CDK6 group compared with that in the miR-81a-5p-mimic group (Fig. 4C and D). These results suggest that miR-181a-5p targets CDK6 in OSCC cells.

Overexpression of miR-181a-5p affects biological behavior of OSCC cells by targeting CDK6

We investigated whether miR-181a-5p affects the biological behavior of OSCC cells by targeting CDK6. The results revealed that, compared with the control and si-NC groups, the overexpression of miR-181a-5p significantly inhibited the proliferation of OSCC cells, and restoring CDK6 expression restored cell proliferation (Fig. 4E). Scratch assays revealed that, compared with the control and si-NC groups, the overexpression of miR-181a-5p significantly inhibited the migration ability of OSCC cells, especially after 48 h of scratch treatment. Based on the results of CDK6 supplementation, the migration ability of OSCC cells after 24 and 48 h was improved compared with that of the miR-181a-5p-overexpressing group but was still lower than that of the control and NC + CDK6 NC groups (Fig. 4F). These results indicated that miR-181a-5p inhibited the proliferation and migration of OSCC cells by targeting CDK6.

Overexpression of miR-181a-5p inhibits the glycolysis by targeting CDK6 in OSCC cells

To elucidate the effect of miR-181a-5p on the glycolytic pathway by targeting CDK6 in OSCC cells, we examined the levels of several key metabolites involved in the glycolytic pathway. The results revealed that the overexpression of miR-181a-5p decreased glucose consumption and inhibited PA, lactate, and ATP production in OSCC cell lines, whereas supplementation with CDK6 indicated an opposite phenotype (Fig. 5A–D). We further confirmed that the overexpression of miR-181a-5p significantly suppressed the mRNA and protein expression of HK2, PKM2, and PFKM. Supplementation with CDK6 significantly increased the mRNA and protein expression levels of these enzymes but did not significantly change LDHA (Fig. 6A–C). Finally, an enzyme activity assay revealed that overexpression of miR-181a-5p decreased the viability of PK, PFK, and HK OSCC cells, whereas complementary CDK6 expression increased their viability (Fig. 5E–G). Collectively, these findings indicate that miR-181a-5p suppresses the glycolytic pathway in OSCC cells by targeting CDK6.

Fig. 5.

MiR-181a-5p regulations contents of glycolysis-related metabolites and enzyme activities in OSCC cells by targeting CDK6. (A–D) Glucose concentration (A), pyruvate concentration (B), lactic acid concentration (C), and ATP concentration (D) in SCC9 and CAL27 cells after transfection with the miR-181a-5p mimic and the corresponding CDK6-plasmid. (E–G) Activity of PK (E), PEK (F), and HK (G) in SCC9 and CAL27 cells after transfection with the miR-181a-5p mimic and the corresponding CDK6-plasmid. Data are representative of three independent experiments. The data represent the mean ± SD. ns not significant (P ≥ 0.05), *P < 0.05, **P < 0.01, ***P < 0.001

Fig. 6.

MiR-181a-5p regulations glycolytic genes expression in OSCC cells by targeting CDK6 (A–C) Relative mRNA and protein expression of HK2, PKM2, PFKM, and LDHA in SCC9 and CAL27 cells after transfection with the miR-181a-5p mimic and the corresponding CDK6-plasmid. Data are representative of three independent experiments. The data represent the mean ± SD. ns: not significant (P ≥ 0.05), *P < 0.05, **P < 0.01, ***P < 0.001.

To further investigate the effect of the overexpression of miR-181a-5p on the SSP, CDK6 was targeted. We examined the amount of NADPH and found that NADPH production in OSCC cells was reduced after the overexpression of miR-181a-5p and increased after expression of the complementary CDK6 gene (Fig. 7A). We further explored the mRNA and protein expression levels of PHGDH and PSAT1. The results revealed that overexpression of miR-181a-5p significantly inhibited the mRNA and protein expression levels of PHGDH and PSAT1, and the addition of CDK6 significantly increased the expression of both enzymes (Fig. 7B and C). Collectively, these data strongly suggest that the overexpression of miR-181a-5p inhibits SSP by targeting CDK6.

Fig. 7.

The Overexpression of miR-181a-5p inhibits SSP and PPP in OSCC cells by targeting CDK6. (A) NADPH concentrations in SCC9 and CAL27 cells after transfection with the miR-181a-5p mimic and the corresponding CDK6-plasmid. (B, C) Relative mRNA and protein expression of PHGDH and PSAT1 in SCC9 and CAL27 cells after transfection with the miR-181a-5p mimic and the corresponding CDK6-plasmid. (D, E) Relative expression of G6PD mRNA and protein in SCC9 and CAL27 cells after transfection with the miR-181a-5p mimic and the corresponding CDK6-plasmid. Data are representative of three independent experiments. The data represent the mean ± SD. ns: not significant (P ≥ 0.05),^*P < 0.05, ^**P < 0.01, ^***P < 0.001

In addition, we assessed the effect of the overexpression of miR-181a-5p on the PPP by targeting CDK6 in OSCC cells. The mRNA and protein expression levels of G6PD were significantly lower in the miR-181a-5p mimic group than in the NC group. Notably, this alteration in expression was hindered by the upregulation of CDK6 (Fig. 7D and E). However, a previous study found that knockdown of CDK6 did not affect the expression level of G6PD. These findings suggest that overexpression of miR-181a-5p represses PPP via CDK6-independent targets in OSCC cells.

Overexpression of miR-181a-5p inhibits tumor growth and the glycolytic pathways, SSP, and PPP of OSCC

Rigorous in vivo assessments were performed to substantiate the in vitro observations. M-NSG mice were subcutaneously inoculated with CAL27 cells in which miR-181a-5p was stably overexpressed. We first confirmed that the expression level of miR-181a-5p was elevated in cancer tissues of the OE group (P = 0.0026) (Fig. 8E). The weight of the OE group gradually increased compared to that of the NC group 20 days after transfection (Fig. 8B). We observed notable decreases in both growth velocity and tumor mass in the OE group (Fig. 8A and C). H&E staining revealed that the tumor tissues were less malignant in the OE group than in the NC group, and that there were metastatic foci in the liver and lungs (Fig. 8D). IHC staining showed that overexpression of miR-181a-5p significantly reduced Ki-67 (a marker of proliferation) expression (P = 0.0026), but increased caspase-3 (an apoptosis marker) expression (P = 0.005) in tumor tissues (Fig. 8N). Moreover, a consequent attenuation in the expression levels of CDK6 (P < 0.0001), HK2 (P < 0.0001), PFKM (P = 0.0008), PKM2 (P < 0.0001), PHGDH (P < 0.0001), PSAT1 (P = 0.0003), and G6PD (P < 0.0001) was observed following miR-181a-5p overexpression (Fig. 8F–M). These in vivo findings support the hypothesis that miR-181a-5p plays a crucial role in attenuating OSCC tumorigenesis.

Fig. 8.

Regulatory role of miR-181a-5p in inhibiting tumor growth and glycolysis, SSP, and PPP in vivo. (A) Images of tumor-bearing mice and excised tumors. (B) Body weight curves of the mice and weight comparison scatter diagram of NC and OE group mice at 24 days post-inoculation. (C) Tumor growth curves and tumor weight comparison scatter diagram of NC and OE group mice at 24 days post-inoculation. (D) Pathological changes in the H&E-stained tumor tissues, lungs, and livers of mice in each group. (E) Relative mRNA expression of miR-181a-5p in tumor tissues. (F, G) Relative mRNA and proteinexpression of CDK6 in tumor tissues. (H, I) Relative mRNA and protein expression of HK2, PKM2, PFKM, and LDHA in tumor tissue. (J, K) Relative mRNA and protein expression of PHGDH and PSAT1 in tumor tissues. (L, M) Relative mRNA and protein expression of G6PD in tumor tissues. (N) Representative IHC images for Ki-67 and caspase-3 in xenograft tissues. (O) The quantitative analysis results for Ki-67 and caspase-3 (P) Summary model illustrating that the miR-181a-5p/CDK6 axis attenuates glycolytic reprogramming in OSCC cells. Data are representative of three independent experiments. The data represent the mean ± SD. ns: not significant (P ≥ 0.05),*P < 0.05, **P < 0.01, ***P < 0.001

Discussion

As cancer cells proliferate in an uncontrolled manner, their metabolism must be adjusted according to their increased need for energy [12]. As a ubiquitous metabolic adaptation, aerobic glycolysis supplies abundant intermediates for nucleotide, lipid, and nonessential amino acid synthesis, and produces ATP to provide energy for tumor cells, thereby fueling rapid tumor cell proliferation and invasion [20, 21]. However, the precise roles and mechanisms that govern aerobic glycolysis in OSCC remain unclear. Recent studies have highlighted the involvement of PDIA6 and DEP domain-containing 1 (DEPDC1) in enhancing aerobic glycolysis and the aggressiveness of OSCC cells. Specifically, elevated PDIA6 expression was found to augment glucose consumption, lactate production, and ATP levels in OSCC [9], whereas DEPDC1 promoted aerobic glycolysis, migration, and invasion via the WNT/β-catenin pathway [22]. Furthermore, HOXC6-mediated transcriptional activation of ENO2 promoted OSCC progression via the Warburg effect [10]. Our study focused on CDK6, a protein with abnormal expression in patients with HNSC from the TCGA dataset (Fig. 1A), which was confirmed to be aberrantly overexpressed in OSCC cells (Fig. 1C, D). Furthermore, patients with high CDK6 expression had significantly lower overall survival rates than those with low CDK6 expression (Fig. 1B), which is consistent with observations in stomach cancer [23], prostate cancer [24], and multiple myeloma [25]. Collectively, these findings highlight the potential role of CDK6 in cancer development.

The results confirmed that CDK6 silencing notably inhibited the proliferation and migration of OSCC cells (Fig. 1E, F). Given that tumor cells rely heavily on the aerobic glycolytic pathway to fulfill their increased bioenergetic and biosynthetic demands for growth, metastasis, and invasion [9, 26–28], aerobic glycolysis is characterized by the breakdown of glucose into pyruvate via enzymes such as HK, PFK, and PK [26, 29, 30]. To further confirm the effect of CDK6 on glycolysis, we investigated whether the key enzymes involved in glycolysis, such as PFK, PKM2, and HK2, were affected by the loss of CDK6 [13, 14]. Our findings revealed that CDK6 silencing led to a marked reduction in the expression levels of both HK2 and PFKM (Fig. 2H), along with a decrease in their enzymatic activities (Fig. 3F, G). CDK6 regulates different metabolic enzymes and plays distinct roles in glycolysis in various cancer types. There are three different phosphofructokinase 1(PFK1) isoforms: muscle, liver, and platelets, which differ depending on the tissue of residence [31]. Research has shown that CDK6 phosphorylates PFKP at serine 679 in cancer cells, resulting in a dimeric transition of PFKP from its tetrameric form [13]. Another study on breast cancer revealed that lncRNA LINC00538 promotes the association between CDK6 and cyclin D3, leading to enhanced phosphorylation of PFKFB3 and STK38. PFKFB3 catalyzes F-2,6BP, which allosterically modulates the enzymatic activity of PFK1, leading to enhanced glycolysis [32]. HK2 is regarded as a crucial regulator of glucose metabolism, facilitating the shift of glycolysis from oxidative phosphorylation to aerobic glycolysiss [33], It has been proposed that CDK6 increases the expression of mTOR, thereby leading to the activation of mTORC1 and consequent upregulation of HIF-1α, which leads to increation of HK2 and glycolysis [34]. Another study on colorectal cancer discovered that CDK6 enhances glycolysis and that HK2 might be a potential downstream target of CDK6 [13]. Additionally, we observed inhibition of PKM2 activity (Fig. 3E), while PKM2 expression was not significantly affected (Fig. 2H), underscoring the multifaceted regulation of key glycolytic enzymes by CDK6. Wang et al. reported that CDK6 was involved in the phosphorylation of PKM2, and the hyperphosphorylation of PKM2 would lead to the decreased PK activity [13], Yang et al. also verified this notion [35]. Notably, as the final enzyme in the glycolytic cascade, LDHA is indispensable for converting pyruvate into lactate, which is a pivotal step in tumor metabolism [36, 37]. In the present study, CDK6 silencing significantly suppressed LDHA expression (Fig. 2E), further emphasizing the critical role of CDK6 in glycolytic regulation. Consistent with these findings, we observed a notable decrease in lactate and pyruvate concentrations in the CDK6-silenced group, whereas glucose levels in the culture medium increased (Fig. 3A–C). These results unequivocally demonstrated that CDK6 silencing effectively impedes glycolytic activity in OSCC cells, shedding light on the pivotal role of CDK6 in sustaining the Warburg effect and driving tumor progression.

Glucose taken up by cells can be channeled catabolically into a bifurcating ATP‐generating glycolysis pathway, leading to the formation of pyruvate, or shunted to the NADPH-producing SSP and PPP [30]. Our results revealed that CDK6 silencing decreased NADPH production in the OSCC cell lines (Fig. 3H). SSP, an integral part of glycolysis, involves critical enzymes such as PSAT1, PSPH, and PHGDH [38]. Notably, PHGDH is a pivotal enzyme in this pathway, serving as the primary rate-limiting factor that catalyzes the conversion of 3-phosphoglycerate to 3-phosphohydroxypyruvate, which is a crucial step in serine biosynthesis [39, 40]. Subsequently, PSAT1, which participates in the second step of this pathway, facilitates the transformation of 3-phosphohydroxypyruvate into l-phosphoserine via a glutamate-linked transamination reaction [41, 42]. CDK6 silencing markedly reduced the protein levels of PHGDH and PSAT1 (Fig. 2I). These findings suggest that CDK6 exerts a regulatory influence on SSP in OSCC cells. This discovery underscores the potential of CDK6 as a therapeutic target for modulating serine‒glycine metabolism and disrupting tumor progression.

The rapid division of cells relies heavily on PPP for essential metabolites that fuel ribonucleotide synthesis and maintain a crucial intracellular redox balance [43]. PPP operates as a bypass route from glucose-6-phosphate, a glycolysis intermediate, through two stages of oxidation and group transfer, regenerating ribose 5-phosphate and NADPH [19]. The core of this pathway is G6PD, the sole rate-limiting enzyme that governs its flux [44]. Our current investigation revealed a novel finding that G6PD expression was not influenced by the silencing of CDK6 (Fig. 2J). However, some studies have shown that CDK6 inhibitors reduce the PPP by inhibiting G6PD activity [45]. Other studies have found that the inhibition of CDK4 and CDK6 disrupts the balance between the oxidative and non-oxidative branches of the PPP [46]. These differences may be due to differences in the tumor microenvironment. The importance of CDK6 in pentose phosphate pathway regulation in OSCC requires further investigation.

The precise mechanisms that CDK6 directly regulate transcription of glucose metabolism enzymes remain insufficiently explored. Some studies demonstrate that HIF-1α-dependent suppression of NF-κB transcriptional activity requires CDK6 [47], CDK6 and p21 have both been shown to induce NF-κB activity [48], Another study discovered that CDK4/6 knockdown causes upregulation of MYC and downregulation of HIF-1α [49]. HK2 is highly expressed in human cancers and is a direct target of both c-Myc and HIF-1. c-Myc and HIF-1 cooperate to enhance glucose metabolism by inducing HK2 [50]. PFKP has been reported to function upstream and downstream of c-Myc. PFKP increases ERK-mediated stability of c-Myc and stimulates PFKP expression at the transcriptional level in neck squamous cell carcinoma [51]. It has been well reported that HIF-1α binds to the promoter regions of PHGDH, PSAT1, and PSPH, respectively, and enhances the transcriptional expression of SSP genes in response to S/G deprivation [52]. Another study discovered that HIF-1α-elevated HMGA2 regulates PHGDH transcription by binding to the PHGDH promoter, resulting in metabolic reprogramming from oxidative phosphorylation to glycolysis [53]. It has been shown that CDK4/6 inhibition increased phosphorylation of ATK1S1, which may trigger the activation of mTORC1 signaling [13, 53], On the other hand, CDK4/6 inhibition drove mTOR pathway activation [54] Activating transcription factor 4 (ATF4) has been reported to function both upstream and downstream of mTOR in different studies. mTORC1 activated ATF4 and stimulated MTHFD2 expression [55, 56]. ATF4, as a crucial glycolytic activator, directly binds to the promoter of HK2 and interacts with HIF-1α to maintain HIF-1α stability by reducing the ubiquitination levels [57]. There is evidence that ATF4 transcriptionally regulates PHGDH expression during ER stress-triggered Cd-induced autophagy [58]. Another study discovered that tumor microenvironmental cues induce ATF4-mediated PHGDH expression in ECs, triggering a redox-dependent mechanism that regulates endothelial glycolysis and leads to EC overgrowth [59]. Therefore, based on the above studies, we speculated that CDK6 could regulate glucose metabolic enzymes via different metabolic transcription factors, thus influencing metabolism in OSCC. However, further experiments are required to test this hypothesis.

The main finding of this study is that CDK6 is a crucial regulator of cell glycolytic reprogramming in OSCC, thereby contributing to a deeper understanding of the metabolic underpinnings of this disease and potentially providing novel therapeutic strategies. Targeting CDK6 is a promising antitumor approach in OSCC. We hypothesized that CDK6 inhibitors suppress OSCC growth and metastasis by disrupting glycolysis. Mechanistically, miRNAs exert their regulatory effects by targeting specific mRNAs, either inhibiting their translation or inducing their degradation through complementary binding to the 3' UTR [60–62]. Therefore, studying the effects of miRNAs on tumor cell proliferation, invasion, and glycolysis has clinical value. For example, miR-34a-3p has been deemed as a potential therapeutic target for OSCC [63]. The previous study has also demonstrated that miR-181a-5p was downregulated in human OSCC cell lines and human OSCC specimens [16]. Moreover, analysis of the published dataset (GSE45238) confirmed that miR-181a was significantly downregulated in OSCC tissues compared with normal tissues. In this study, bioinformatic analysis predicted a potential interaction between miR-181a-5p and CDK6, and dual-fluorescein experiments validated the targeted binding of both, prompting us to investigate the impact of miR-181a-5p on glycolysis in OSCC by targeting CDK6 (Fig. 4B). Previous studies have revealed that miR-181a-5p targets NDRG2 to promote cell proliferation, invasion, and glycolysis by activating the PTEN/AKT pathway [64]. Our findings revealed that the overexpression of miR-181a-5p effectively inhibited the malignant behavior of OSCC cells, accompanied by the downregulation of CDK6 expression, confirming that CDK6 is a functional target of miR-181a-5p (Fig. 4C–F). Overexpression of miR-181a-5p led to significant inhibition of key glycolytic enzymes, including HK2, PFKM, and PKM2, as evidenced by reduced glucose uptake, and lactate and pyruvate production. Notably, this inhibitory effect on glycolysis was largely reversed by supplementation with CDK6, suggesting that miR-181a-5p suppresses glycolysis through CDK6 inhibition. However, miR-181a-5p overexpression did not significantly alter LDHA expression. Moreover, our data demonstrated that miR-181a-5p inhibits SSP in OSCC by targeting CDK6. Specifically, the overexpression of miR-181a-5p led to the downregulation of PHGDH and PSAT1 expression, which was reversed by CDK6 overexpression (Fig. 5B and C). Additionally, we observed that the miR-181a-5p mimic decreased the expression of G6PD. Importantly, CDK6 rescued the reduction in G6PD expression induced by miR-181a-5p (Fig. 5D and E). However, a previous study found that knockdown of CDK6 did not affect the expression level of G6PD. Therefore, we hypothesized that PPP suppression may involve CDK6-independent targets of miR-181a-5p. Collectively, our results provide the first evidence that miR-181a-5p negatively modulates CDK6, leading to the consequential inhibition of glycolysis and SSP, and attenuation of PPP via CDK6-independent targets in OSCC.

Unfortunately, this study has some limitations. First, we did not fully elucidate some mechanisms, such as systematically evaluating the dual regulatory effects of CDK6 on proliferation—metabolism and the possible CDK6-independent effects of miR-181a-5p. Second, due to limited time and funding, we could not collect clinical patient-derived models from OSCC to verify the expression profiles of miR-181a-5p and CDK6. Therefore, it is necessary to collect OSCC clinical samples for future studies. Finally, only two established OSCC cell lines were used for this study. More OSCC cell lines are utilized to validate the experiment. Moreover, the scratch-wound assay may reflect not only impaired migration but also slower cell growth. Migration assays that control for proliferation will be included in future studies. These issues will be addressed in future studies.

Conclusions

Our findings reveal a novel mechanism by which CDK6 promotes glycolysis in OSCC by activating crucial enzymes such as HK2, PFKM, PKM2, LDHA, PHGDH, and PSAT1. These findings establish a direct link between CDK6 and enhancement of glycolytic metabolism and the serine-glycine biosynthetic pathway in OSCC (Fig. 8P). Importantly, we firstly demonstrated that miR-181a-5p negatively modulates CDK6, leading to the consequential inhibition of glycolysis and SSP, and attenuation of PPP via CDK6-independent targets in OSCC. These groundbreaking findings regarding the regulatory mechanisms governing metabolic reprogramming in OSCC offer promising avenues for therapeutic interventions.

Author contributions

JPG, YYY and XQC contributed to data collection, interpretation of data, designing the study, and writing the original manuscript. XTW, ZRZ, SXS, YQL and JJS contributed in analysis of data. ZYC has critically revised the manuscript. RS and GHS contributed reagents, material and equipment, designing and review the manuscript. All authors gave their final approval and agree to be accountable for all aspects of the work.

Funding

This work was supported by the National Natural Science Foundation of China (31970513), the Central Government’s Guide to Local Science and Technology Development Fund of Shanxi Province (YDZJSX2022A060), the special fund for Science and Technology Innovation Teams of Shanxi Province (202204051002032), the Natural Science Foundation of Shanxi Province (20210302124093), the Shanxi Scholarship Council of China (2021–086), the Shanxi Province Higher Education “Billion Project” Science and Technology Guidance Project (BYJL016), and the Shanxi Bethune Hospital Talent Introduction Research Start-up Fund of China (2022RC13).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

No experiments involving human subjects were performed in this study. The animal research was arovpped by the Animal Experiment Ethics Committee of the Shanxi Medical University (SYDL2023017).

Consent for publication

All the authors agreed to be published.

Competing interests

The authors declare no competing interests.