The mechanism by which KAT2A increases the stability of CDC25A through acetylation to regulate glycolysis and mediate lung adenocarcinoma immune escape

Abstract

Background

Immune escape is a defining feature of malignant tumor initiation and progression. CDC25A is an oncogenic gene, highly expressed in various cancers, yet the molecular mechanisms behind its upregulation in lung adenocarcinoma (LUAD) and its role in immune evasion remain incompletely understood.

Methods

We assessed CDC25A and lysine acetyltransferase 2 A (KAT2A) expression in LUAD patients, leveraging the TCGA-LUAD database, and verified their mRNA and protein levels in LUAD cell lines and tissues with qRT-PCR and WB. Kaplan-Meier analysis was used to explore the prognostic relevance of CDC25A expression in LUAD, while Pearson correlation was performed to analyze the relationship between CDC25A and KAT2A. Cell viability and proliferation were determined using CCK-8 and colony formation assays. A cell metabolic analyzer was used to assess metabolic rates (ECAR and OCR). Glucose and lactate levels in the supernatant were measured using respective assay kits. CD8⁺T cells were co-cultured with LUAD cells, with their activation monitored by flow cytometry. The killing capacity of CD8⁺T cells towards LUAD cells was evaluated with lactate dehydrogenase (LDH) and ELISA kits. The interaction between CDC25A and hexokinase 2 (HK2) and their cytoplasmic co-localization were confirmed by CO-immunoprecipitation (CO-IP) and immunofluorescence. Chromatin Immunoprecipitation (ChIP) verified the binding relationship between KAT2A and CDC25A. In vivo validation was conducted in a mouse allograft tumor model.

Results

CDC25A was upregulated in LUAD tissues and cells, with an identified link to poorer patient survival. Knocking down CDC25A expression hindered LUAD cell activation and proliferation, lowered aerobic glycolysis, elevated the ratio of CD8⁺IFN-γ⁺ T cells, and boosted the cytotoxicity of CD8⁺T cells against tumor cells. Moreover, CDC25A could interact with HK2 to impact its protein expression, thereby modulating aerobic glycolysis in LUAD cells. KAT2A, highly expressed in LUAD tissues and cells, was positively correlated with CDC25A levels. KAT2A promoted the acetylation of CDC25A, promoting immune evasion in LUAD cells. The in vivo and in vitro studies had similar results.

Conclusion

This research reveals that KAT2A promotes the expression of CDC25A via acetylation. Subsequently, CDC25A interacts with HK2 to control aerobic glycolysis, thereby driving the immune evasion process in LUAD.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12931-026-03544-2.

Introduction

Lung cancer is one of the most widespread malignant tumors. According to the American Cancer Society, around 350 people die from lung cancer daily, making it the leading cause of cancer mortality [1]. Its predominant variant, lung adenocarcinoma (LUAD), has been continuously increasing in both incidence and mortality, causing a serious social burden and economic loss [2]. A large number of LUAD patients are diagnosed at later stages (III/IV), often missing the optimal window of treatment, which greatly affects their quality of life and survival chances, with a five-year survival rate of less than 20% [3, 4]. Compared to standard cancer treatments such as surgical intervention, radiation, chemotherapy, and targeted drugs, immune checkpoint blockade, exemplified by PD-1/PD-L1 inhibitors, has been proven to be a safer and more effective alternative. This therapeutic approach functions by targeting PD-1 on T cells or PD-L1 on tumor cells, thereby blocking signals that facilitate immune evasion. This process reactivates T cells to combat cancer cells and restore their immune responses [5]. However, the effectiveness of this therapy, as measured by clinical response rates, is relatively low in certain LUAD patients [6]. Thus, delving deeper into the molecular mechanisms of tumor immune evasion in LUAD holds significance for advancing clinical benefits and for the management of LUAD.

Metabolic reprogramming, one of the fundamental characteristics of cancer, is closely linked to tumor genesis and progression. Even in oxygen-abundant conditions, cancer cells are more likely to metabolize glucose by the glycolysis pathway, not oxidative phosphorylation, resulting in lactate accumulation. This lactate promotes tumor growth and metastasis through various biological mechanisms [7]. Burgeoning literature has validated the indispensable role of glycolysis in the proliferation, angiogenesis, metastasis, and immune evasion of tumors [8]. The robust glycolysis in tumor cells can impede the functionality of immune cells and the secretion of immune co-stimulatory factors, resulting in tumor immune escape. On the contrary, inhibiting this glycolytic activity can greatly enhance the effectiveness of cancer immunotherapy [9]. Specifically, cancer cells with high glycolytic rates have a stronger glucose uptake, affecting the glucose availability in the tumor immune microenvironment. This glucose scarcity can weaken the cytotoxic capability, promoting the immune evasion of tumor cells [10, 11]. Cascone et al. [12] found that heightened glycolytic levels could impede T-cell cytotoxicity against melanoma, and a negative association between the infiltration of T cells in melanoma and non-small cell lung cancer, and the expression of glycolytic genes ALDOA, ENO2, and GAPDH using TCGA. Moreover, clinical studies have shown that the level of aerobic glycolysis in tumors is negatively connected to the antitumor immune response in the host and the efficacy of cancer immunotherapy. For instance, the upregulation of glycolytic enzymes HK2, PFK1, and PMK2 in liver cancer is linked to immune evasion by modulating various signaling pathways, including PI3K/Akt and AMPK [13]. Hexokinase 2 (HK2), as a glucose sensor in glycolysis, can enhance the expression of PD-L1 in tumor cells, thereby facilitating tumor immune evasion [14]. To sum up, targeting aerobic glycolysis is effective in inhibiting tumor cells from evading the immune system. However, current understanding of the complex mechanisms by which aerobic glycolysis promotes tumor immune evasion is limited. A deeper investigation into the interplay between glycolysis and tumor immune evasion mechanisms is essential, as is the identification of novel biomarkers.

The CDC25 family of dual-specificity phosphatases (DSP), integral to cell cycle regulation, has three isoforms: CDC25A, CDC25B, and CDC25C. The CDC25A protein, with 524 amino acid residues, is structured with an N-terminal regulatory domain and a C-terminal catalytic domain [15]. CDC25A is involved in a cascade of biological processes, such as cell division, cell proliferation, and DNA replication. As an oncogene, it is engaged in the occurrence and progression of tumors, upregulated in many malignant tumors, including lung cancer, breast cancer, esophageal cancer, and related to the malignancy and prognosis of the tumor [16, 17]. The overexpression of CDC25A in cancer is driven by complex mechanisms that act at the transcriptional, translational, and post-translational levels, with more studies focusing on post-translational modifications. The deubiquitinating enzyme USP29, for instance, can stabilize the CDC25A protein by preventing its ubiquitin-mediated degradation, thus contributing to the onset and progression of cancer [18]. Similarly, Biswas et al. [19]. identified USP7, another deubiquitinating enzyme associated with CDC25A, in their study on MCF7 breast cancer cells. They found that silencing brain and reproductive organ expressed (BRE) led to decreased CDC25A protein expression by engaging USP7. A recent study has unveiled a regulatory mechanism involving the acetylation and deacetylation of CDC25A, which plays a role in modulating CDC25A activity and governing the cell cycle. Post-DNA damage, CDC25A is acetylated by ARD1, an acetyltransferase, which confers enhanced stability to the protein. In the absence of cellular stress, acetylated CDC25A is targeted by the deacetylase HDAC11, with the equilibrium of CDC25A acetylation being regulated by the antagonism of ARD1 and HDAC11 [20]. Accumulating studies have revealed RNA modifications as potential therapeutic targets in cancer, linked to the aggressive biological characteristics of numerous cancers [21, 22]. Lysine acetyltransferase 2 A (KAT2A) mainly mediates H3K27 acetylation to promote transcription of downstream target genes [23, 24]. Therefore, this study investigates whether KAT2A modifies CDC25A through H3K27 acetylation.

Our study detected CDC25A overexpression in LUAD, where it interacts with HK2 to increase the level of aerobic glycolysis in cancer cells, thus reducing the effectiveness of CD8⁺T cells in killing tumor cells. In addition, our data showed that KAT2A is an upstream modulator of CDC25A, enabling the upregulation of CDC25A expression via acetylation and thus promoting immune evasion in LUAD.

Materials and methods

Bioinformatics

The TCGA database provided LUAD mRNA expression data from 59 normal and 539 tumor samples, which we analyzed for differential expression using the “edgeR” R package (|logFC|>1.0, FDR < 0.05), to obtain a list of differentially expressed mRNAs (DEmRNAs). CDC25A and its upstream regulatory gene KAT2A were selected after a literature review, and their expression profiles in normal and tumor samples were investigated, along with a Pearson correlation analysis. For survival analysis, the Kaplan-Meier database (https://kmplot.com) was consulted to predict the relationship between CDC25A expression and the 5-year survival rate of LUAD patients. PDB structures for CDC25A and HK2 were retrieved from UniProt for subsequent docking simulations on InterEvDock2 (https://bioserv.rpbs.univ-paris-diderot.fr/services/InterEvDock2/). EPIC algorithm of TCGA-LUAD was downloaded from Timer (https://compbio.cn/), with the correlation between gene expression and immune infiltration analyzed.

Patient samples

Between April 2024 and December 2024, LUAD and adjacent tissue samples were collected from 12 patients at the Affiliated Hospital of Jiaxing University (The First Hospital of Jiaxing) who were undergoing surgery for LUAD removal. Informed consent was obtained from all participants, and the study was performed with approval from the Ethics and Research Committee at the Affiliated Hospital of Jiaxing University (The First Hospital of Jiaxing), IRB approval number: 2025-LP-313. All samples were collected intraoperatively, instantly frozen in liquid nitrogen, and finally stored at −80 ℃ in an ultra-low temperature refrigerator.

Cell cultivation

BeNa Culture Collection (BNCC, China) supplied human lung epithelial BEAS-2B cells (BNCC359274) and LUAD cell lines, including A549 (BNCC337696), H1299 (BNCC100268), and Calu-3 (BNCC359757). CD8⁺T cells (PCS-800-017™) were from ATCC (USA), and mouse LUAD cells LLC (SNL-119) from SUNNCELL (China). All cell lines were STR profiled and periodically checked for mycoplasma contamination. DMEM-H complete medium (BNCC338068) was used for LLC and BEAS-2B cells, F-12 K complete medium (BNCC338550) for A549, RPMI-1640 complete medium (BNCC338360) for H1299, and MEM complete medium (BNCC338137) for Calu-3, all media from BNCC (China). CD8⁺T cells were cultured in a specific medium. Cultures were treated with 1% penicillin-streptomycin (Sigma, USA) and incubated at 37 ℃ with 5% CO₂.

The medium for CD8⁺T cell cultivation was enhanced with 20 ng/mL IL-2 (200-02, Peprotech, USA) and CD3/CD28 T cell activation agents (11161D, Gibco, USA) for two weeks. Following activation, the CD8⁺T cells were mixed with tumor cells at a 10:1 ratio and co-cultured. After a 24-h incubation, the supernatant was collected for subsequent experimental use.

Cell transfection

Overexpression plasmids for CDC25A, HK2, and KAT2A were constructed using the pcDNA3.1 vector, with the empty pcDNA3.1 vector as a control (oe-NC). The pGPH1 vector was employed for shRNA constructs against CDC25A and KAT2A in mouse and human, with the empty pGPH1 as a control (sh-NC). All plasmids were obtained from Shanghai GenePharma (China). The cells were transfected with these plasmids using Lipofectamine 2000 (Invitrogen, USA)’, and harvested 48 h later for further analysis.

qRT-PCR

Trizol reagent (Sangon, China) was employed for RNA extraction from tissues and cells, and the RNA concentration and quality were evaluated with a Thermo Fisher Scientific quantometer (USA). The ReverTra Ace qPCR RT Kit (Toyobo, Japan) facilitated the conversion of RNA to cDNA, which was then subjected to qPCR with SYBR Green Master Mix (Toyobo, Japan), normalizing to β-actin. The 2^−△△CT method was applied to determine the relative mRNA expression levels of the target genes. The qRT-PCR primers were all procured from YKang Biological (China), with sequences detailed in Table 1.

Table 1.

Primer sequences

Name	Primer sequence (5’→3’)
CDC25A	F: AGAAGCTGTTGGGATGTAGTC R: GCCACGAGATACAGGTCTTAC
HK2	F: GCAAGGAGATGGAGAAAGGG R: AGCACACGGAAGTTGGTC
KAT2A	F: CTCTGCCTTAACTACTGGAAGC R: GCCATCTGGTGTAATTGACCTTG
β-actin	F: ACCTTCTACAATGAGCTGCG R: CCTGGATAGCAACGTACATGG

CCK-8

Processed cells were seeded into a 96-well plate at a density of 5 × 10³ cells per well, with five replicate wells for each experimental condition. The plate was cultured in an incubator at 37 ℃ with 5% CO₂. At the end of the culture period, 10 µL of CCK-8 reagent (MCE, USA) was added to each well for another 2-h incubation, with the absorbance at 450 nm determined with a microplate reader.

Colony formation assay

A 6-well plate received an inoculation of 1 × 10³ cells per well and was cultured for one week under standard conditions. When colonies were visibly formed, the plate was taken out and the culture medium discarded. Each well was fixed with 2 mL of 4% paraformaldehyde for 30 min. Subsequently, the paraformaldehyde was removed, and 2 mL of 0.1% crystal violet was introduced to each well for staining for 15 min. Finally, the crystal violet was rinsed away, and the colonies were enumerated by a computer.

Western blot (WB)

Cellular proteins were harvested using NP40 lysis buffer (Beyotime, China) supplemented with protease inhibitors (Beyotime, China), and protein concentrations were measured using a BCA protein assay kit (Beyotime, China). Following denaturation at 100 ℃ for 10 min, the proteins were separated via SDS-PAGE and transferred to a 0.45 μm PVDF membrane (Millipore, USA). The membrane was blocked with 5% skim milk for 1 h and TBST-washed three times for 5 min each. Primary antibodies against CDC25A (ab989), KAT2A (ab208097), and β-actin (ab227387), all at a 1:1000 dilution, from Abcam(UK), were incubated with the membrane overnight on ice. After washing, the membrane was incubated with HRP-conjugated secondary antibodies (goat anti-rabbit (ab205718) or goat anti-mouse (ab205719), 1:5000, Abcam, UK) for 1 h at room temperature. The proteins were then visualized using an ECL detection kit (ABclonal, China) after another three washes.

Flow cytometry

Trypsin was utilized to detach the transfected cells, which were then suspended in PBS to prepare a single-cell suspension. The cells were incubated with the PE-conjugated IFN-γ antibody (E-AB-F1196D, Elabscience, China) for 15 min. Afterward, they were washed with PBS, resuspended, and the data were acquired with flow cytometry, followed by data analysis.

Lactate dehydrogenase (LDH)and ELISA

The cytotoxicity of T cells on LUAD cells was determined with an LDH kit from YEASEN (China). Human ELISA kits (E-EL-H0108, E-EL-H0109, E-EL-H1617, E-EL-H1123) were used to quantify IFN-γ, TNF-α, Granzyme B, and Perforin in the supernatant of co-cultures. Mouse ELISA kits were applied for cytokine detection in mouse peripheral blood (E-EL-M0048 for IFN-γ, E-EL-M0594 for Granzyme B, E-EL-M3063 for TNF-α, E-EL-M0890 for Perforin).’

Extracellular acidification rate (ECAR) and oxygen consumption rate (OCR)

A Seahorse XFp analyzer (Agilent, USA) was used to measure ECAR and OCR as indicators of aerobic glycolysis. Post-transfection, cells were grown in Seahorse XF cell culture microplates for 24 h and then in XF Medium (DMEM supplemented with 25 mM glucose and 2 mM glutamine, non-buffered) for 2 h. For ECAR, reagents were added in a sequence outlined in the XF Glycolysis Stress Test Profile (Agilent, USA): 10 mM glucose, 3.5 µM oligomycin, and 100 mM 2-deoxy-D-glucose at precise time intervals. For OCR assessment, the XF Cell Mito Stress Test Profile (Agilent, USA) guided the addition of 3.5 mM oligomycin, 2 mM FCCP, and 1 mM rotenone/antimycin A. The Seahorse XFp Wave software was employed for subsequent data analysis.

Detection of glucose content and lactate levels

Glucose uptake and lactate production in the supernatants were analyzed with the Glucose Uptake Assay kit (YEASEN, China) and Lactate Assay kit (Elabscience, China), respectively.

CO-immunoprecipitation (CO-IP)

Post-treatment cell culture for 24 h was followed by triple washing with pre-cooled PBS. Cells were lysed on ice for 30 min in NP-40 buffer with added protease inhibitors. The lysed material was centrifuged at 4 ℃ and 14,000 g for 15 min, and the supernatant was moved to a fresh EP tube. A 100 µL portion of the supernatant was taken as the Input, with another 100 µL portion for IP and IgG samples. The IP sample was incubated with antibodies against CDC25A (ab989) and HK2 (ab208097) (Abcam, UK), while the IgG sample was incubated with IgG antibody (AC005, ABClonal, China), all at 4℃ for 1.5 h. A/G agarose beads (Proteintech Group, China) were added to each sample and incubated overnight at 4 ℃. After centrifugation at 2,000 g for 3 min and three washes with lysis buffer at 4 ℃, the samples were subjected to WB analysis.

Immunofluorescence (IF)

After treatment, cells were seeded onto 6-well plates with cell slides for a 24-h incubation period. The slides were fixed in 4% paraformaldehyde and rinsed twice with PBS. Cells were permeabilized with 0.5% Triton X-100 (Sigma, USA) for 10 min at room temperature, then blocked with 5% BSA for 1 h. Primary antibodies specific for CDC25A (A1173, ABClonal, China) and HK2 (ab209847, Abcam, USA) were applied and incubated overnight. Two PBS washes preceded a 1-h incubation with secondary antibodies Goat Anti-Mouse IgG H&L/FITC (bs-0296G-FITC) or Goat Anti-Rabbit IgG H&L/AF555 (bs-0295G-AF555) (BIOSS, China) in the dark. DAPI was applied for 5 min to counterstain cell nuclei. The slides were imaged under a fluorescence microscope from Keyence (Japan).

Live/death cell staining

Treated cells were collected and inoculated into 6-well plates placed with cell culture slides. After 24 h of incubation, the cell culture slides were fixed using 4% paraformaldehyde. After fixation was completed, the cell culture slides were washed twice with PBS buffer, followed by washing the cell culture slides fully with 1×Assay Buffer for two to three times. Next, the cell suspension was prepared with 1× Assay Buffer. 100 µL of Calcein-AM/PI Live/Dead Cell Kit (Solarbio, China) working solution was taken and added to 200 µL of cell suspension, which was mixed thoroughly and incubated at 37 ℃ for 15 min. Finally, the cells were observed, photographed, and recorded using a fluorescence microscope system (KEYENCE, Japan) to distinguish between live and dead cells.

Chromatin Immunoprecipitation (ChIP) assay

Cells were treated with 1% formaldehyde to induce covalent cross-linking between proteins and DNA, thereby fixing the protein-DNA complexes. After cross-linking, the cells were lysed using the lysis buffer from the ChIP kit (Thermo Scientific, USA). The cross-linked chromatin was fragmented into small pieces by digestion with micrococcal nuclease. KAT2A antibody (ab217876, Abcam, UK), H3K27ac (ab4729), or IgG antibody (ab172730, Abcam, UK) was added to allow the antibodies to bind to the protein-DNA complexes. The antibody-protein-DNA complexes were then captured using Protein A magnetic beads. After multiple washes to remove non-specifically bound components, protease K was added to reverse the cross-linking. The enrichment of DNA was detected by qPCR. The ChIP-qPCR primer sequences were as follows: CDC25A F: CCCACACTCTGAGGCAATGT, CDC25A R: CGGAAGAAAGGGGTCCACAA.

RNA stability assay

The treated cells were collected, and when the cell confluence reached 70–80%, they were treated with 5 mg/mL of actinomycin D (MCE, USA) for 0, 2, 4, and 6 h to block RNA transcription, respectively. At different time points, RNA in the cells was extracted by lysis with Trizol reagent (Sangon, China), and the remaining experimental steps were consistent with qRT-PCR experiments.

Animal experiment

All animal experiments were approved by the Ethics Committee of the Affiliated Hospital of Jiaxing University (The First Hospital of Jiaxing), Approval number JXYY2025-028, and were conducted in accordance with the institution’s guidelines for the care and use of animals. Eighteen C57BL/6 mice (4–5 weeks old) from Beijing Vital River Laboratory Animal Technology (China) were randomly assigned to three groups for the experiment. LLC cells were transfected with either sh-NC + oe-NC, sh-KAT2A + oe-NC, or sh-KAT2A + oe-CDC25A plasmids and then subcutaneously injected into the right flank of the mice to build allograft tumor models. Tumor growth was tracked by measuring the length and width with calipers every 6 days, and tumor volume was computed using the formula V=(length×width²)/2. On day 30, mice were euthanized under isoflurane anesthesia, and tumor tissues were collected and weighed.

Immunohistochemistry (IHC)

Fixed in 10% formalin, tumor tissues were sectioned and prepared for histological examination. Sections were deparaffinized with xylene, rehydrated through gradient ethanol concentrations, and heated in a pressure cooker with EDTA antigen retrieval solution (pH = 8.0, Sangon, China). After a 10-min incubation with peroxidase blocking solution, non-specific antibody binding was blocked with 5% goat serum for 30 min. Antibodies against CDC25A (ab203618, Abcam, UK), KAT2A (A2224, ABclonal, China), and HK2 (ab209847, Abcam, UK) were used in incubation at 4 ℃ overnight. A subsequent incubation with a goat anti-rabbit secondary antibody (ab205718, Abcam, UK) was finished at 37 ℃ for 30 min. The slides were developed with DAB, counterstained with hematoxylin, and visualized under an inverted microscope.

IF for tissue sections

Slide preparation and processing steps were the same as those of the IHC assay. After an overnight incubation at 4 ℃ with the primary CD8 antibody (ab316778, Abcam, UK), slides were incubated with Alexa Fluor 555-labeled goat anti-rabbit IgG (H + L) secondary antibody (ab150078, Abcam, UK) at 37 ℃ for 30 min. DAPI was applied for 5 min to counterstain cell nuclei, followed by visualization using a fluorescence microscope from Olympus (Japan).

Statistical analysis

All experiments were carried out in triplicate, and results are shown as mean ± standard deviation. GraphPad Prism 8.0 (USA) was utilized for statistical analysis. The Mann-Whitney test was applied for comparisons between two groups, and one-way ANOVA was used for multiple-group comparisons, with statistical significance defined as P < 0.05.

Results

CDC25A is highly expressed in LUAD and associated with prognosis

Data from the TCGA-LUAD database indicated that CDC25A was highly expressed in LUAD patients (Fig. 1A). Kaplan-Meier survival analysis revealed a link between high CDC25A expression and shorter survival (Fig. 1B). Using qRT-PCR, we then compared CDC25A mRNA expression between 12 sets of LUAD tissues and their adjacent normal counterparts, observing a notable increase in the tumor samples (Fig. 1C). Examination of CDC25A mRNA and protein in the LUAD cell lines A549, H1299, and Calu-3, relative to the BEAS-2B cells, showed a marked overexpression in the cancer cells (Fig. 1D-E). In summary, CDC25A overexpression in LUAD correlates with an unfavorable outcome.

Fig. 1.

Upregulation of CDC25A in LUAD tissues and cells. A The TCGA-LUAD database provides an analysis of CDC25A expression, where blue corresponds to normal tissues and red to tumor tissues; (B) Kaplan-Meier analysis of CDC25A in LUAD patients; (C) qRT-PCR detection of CDC25A mRNA in human lung epithelial cells (BEAS-2B) and LUAD cell lines (A549, H1299, Calu-3); (D) qRT-PCR detection of CDC25A mRNA in 12 pairs of LUAD and adjacent normal tissues; (E) WB analysis of CDC25A protein expression levels in human lung epithelial cells (BEAS-2B) and LUAD cell lines (A549, H1299, Calu-3). * represents P < 0.05

CDC25A downregulation inhibits tumor immune escape

We transfected A549 cells with sh-NC and sh-CDC25A plasmids, followed by qRT-PCR to evaluate transfection efficiency. qRT-PCR and WB confirmed successful knockdown, showing a significant decrease in both CDC25A mRNA and protein expression after transfection with the sh-CDC25A plasmid (Fig. 2A-B). CCK-8 showed that CDC25A knockdown suppressed A549 cell viability (Fig. 2C). Live/death cell staining results showed an increase in cell death after knockdown of CDC25A (Fig. 2D). Furthermore, the colony formation assay indicated that low CDC25A expression inhibited the proliferation of LUAD cells (Fig. 2E). To investigate the correlation between CDC25A and immune cell infiltration, we combined the expression data downloaded from the TCGA database with the EPIC algorithm downloaded from the Timer database to analyze the Spearman correlation between CDC25A and CD8⁺T cell immune infiltration. The integrated data (a total of 539 LUAD tumor samples were obtained) showed a negative correlation between CDC25A and CD8⁺T cell immune infiltration levels (Fig. 2F). To examine the effect of CDC25A expression on LUAD immunology, we engaged sh-NC and sh-CDC25A cell groups in indirect co-cultures with activated CD8⁺T cells. CD8⁺T cell activation was measured by flow cytometry, showing a pronounced increase in IFN-γ fluorescence intensity with CDC25A knockdown (Fig. 2G). LDH assay detected that CD8⁺T cells enhanced cytotoxicity towards A549 cells when CDC25A expression was reduced (Fig. 2H). ELISA kit tests on the co-culture supernatant revealed that TNF-α, IFN-γ, Perforin, and Granzyme B levels were substantially higher in the sh-CDC25A group than in the sh-NC group (Fig. 2I). In conclusion, the reduction of CDC25A in LUAD cells augmented the cytotoxic capabilities of CD8⁺T cells against tumor cells.

Fig. 2.

CDC25A knockdown enhances the cytotoxicity of CD8⁺ T cells against LUAD cells. A-B qRT-PCR and WB analysis of transfection efficiency for sh-NC and sh-CDC25A plasmids; (C) CCK-8 test for cell viability; (D) Live/death cell staining assay for detecting cell live and death; (E) Colony formation assay for evaluating cell proliferation; (F) Analysis of the correlation between CDC25A and CD8⁺T cell immune infiltration by combining TCGA database and Timer database; (G) Flow cytometry detection of IFN-γ fluorescence intensity; (H) LDH detection of the cytotoxic effect of CD8⁺T cells on LUAD cells; (I) ELISA for quantifying TNF-α, IFN-γ, Perforin, and Granzyme B in the co-culture supernatants. * represents P < 0.05

CDC25A promotes immune escape of LUAD cells through aerobic glycolysis

The capability of CDC25A to regulate cell metabolism and stimulate aerobic glycolysis in certain cancer cells has been documented. To understand CDC25A’s role in LUAD cell glycolysis, we measured ECAR and OCR to assess how its knockdown affects A549 cell glycolysis. The findings indicated a marked decrease in glycolytic activity and levels upon CDC25A knockdown (Fig. 3A), coupled with an elevation in ATP synthesis and peak OCR (Fig. 3B). Following transfection with the sh-CDC25A plasmid, in the cell culture supernatant, there was a surge in residual glucose levels and a notable decline in lactate levels (Fig. 3C-D). These results suggested that downregulation of CDC25A could inhibit the proliferation of tumor cells. Thus, we postulated that CDC25A could enhance glycolysis in these cells, thereby inhibiting the cytotoxicity of CD8⁺T cells. To verify this, we transfected A549 cells with an oe-CDC25A plasmid and performed rescue experiments with the glycolysis inhibitor 2-DG, with cell groups set as oe-NC + PBS, oe-CDC25A + PBS, and oe-CDC25A + 2-DG. CCK-8 was applied to measure cell viability across groups, which showed that overexpression of CDC25A notably increased cell vitality, an increase that was reversed by the addition of 2-DG (Fig. 3E). Live/death cell staining results also showed that overexpression of CDC25A significantly inhibited cell death, while the results were reversed by further addition of 2-DG treatment (Fig. 3F). Colony formation assay was conducted to evaluate the proliferative capacity of the cells, demonstrating that high levels of CDC25A can enhance tumor cell proliferation, but this enhancement was reduced with 2-DG treatment (Fig. 3G). Next, the cells from the three groups were co-cultured with activated CD8⁺T cells, and the activation of CD8⁺T cells (CD8 + IFN-γ+) was evaluated using flow cytometry. The IFN-γ fluorescence intensity in the oe-CDC25A + PBS group was much lower than in the control group, while the IFN-γ fluorescence intensity in the oe-CDC25A + 2-DG treated group was markedly increased compared to the oe-CDC25A + PBS group (Figs. 3H-I). Thereafter, the supernatant from the co-culture was collected, and LDH activity was measured to evaluate the cytotoxic effect of CD8⁺T cells on tumor cells. Transfection with oe-CDC25A reduced LDH enzyme activity, which was then elevated to control levels after 2-DG treatment (Fig. 3J). ELISA tests on the co-culture supernatant for TNF-α, IFN-γ, Perforin, and Granzyme B levels corroborated these results (Fig. 3K). Overall, CDC25A facilitates immune escape by enhancing glycolysis in LUAD cells.

Fig. 3.

CDC25A enhances glycolysis to suppress the cytotoxicity of CD8⁺ T cells against LUAD cells. A-B ECAR and OCR levels of the two batches of culture from the same cell line were measured using the Seahorse XFe96 cell metabolic analyzer; (C-D) Glucose and lactate content were assessed with respective assay kits; (E) Cell viability was determined by CCK-8; (F) Live/death cell staining assay for detecting cell live and death; (G) Colony formation assay for cell proliferation capacity; (H-I) Flow cytometry for the activation level of CD8⁺T cells; (J) LDH assay for the cytotoxic effect of CD8⁺T cells on A549 cells; (K) ELISA for the content of TNF-α, IFN-γ, Perforin, and Granzyme B in co-culture supernatants. * represents P < 0.05

CDC25A regulates LUAD cell glycolysis via interaction with HK2

Aerobic glycolysis is governed by three key regulatory enzymes—HK2, PFK-1, and PKM2. To identify the precise mechanism by which CDC25A promotes aerobic glycolysis, we conducted molecular docking simulations with InterEvDock2, comparing CDC25A with these enzymes. The results showed promising binding affinities between CDC25A and each enzyme (Table S1). Considering the literature’s silence on CDC25A’s regulation of HK2, we selected HK2 as the target molecule for this study. Figure 4A displays the molecular docking simulation between CDC25A and HK2. The binding interaction between CDC25A and HK2 was further validated by CO-IP (Fig. 4B). IF experiments confirmed the colocalization of CDC25A and HK2 within the cytoplasm of LUAD cells (Fig. 4C). To explore the molecular mechanism by which CDC25A influences LUAD cell glycolysis, we established cell groups including sh-NC + oe-NC, sh-CDC25A + oe-NC, and sh-CDC25A + oe-HK2. WB analysis of HK2 expression in A549 cells indicated that CDC25A knockdown reduced HK2 protein levels, and overexpressing HK2 in the presence of CDC25A knockdown reversed this effect (Fig. 4D-E). Continuing our investigation, we evaluated glycolytic capacity through ECAR and OCR measurements in the cell groups. Transfection with sh-CDC25A resulted in lowered glycolytic activity and elevated ATP synthesis and peak OCR in these cells, and transfection with sh-CDC25A and oe-HK2 mitigated these effects (Fig. 4F-H). With specific assay kits, we found that residual glucose in the sh-CDC25A + oe-NC group was notably elevated, and lactate generation was notably reduced compared to the control group. The sh-CDC25A + oe-HK2 group exhibited a decline in residual glucose and an upsurge in lactate levels when contrasted with the sh-CDC25A + oe-NC group (Figs. 4I-J). In conclusion, CDC25A could cooperate with HK2 to stimulate aerobic glycolysis in LUAD cells.

Fig. 4.

CDC25A binds to HK2 to promote aerobic glycolysis in LUAD cells. A Molecular docking simulation of CDC25A and HK2 conducted with InterEvDock2, red: HK2, blue: CDC25A; (B) CO-IP detection of the interaction between CDC25A and HK2; (C) Cellular IF to assess the localization of CDC25A and HK2 within cells; (D-E) WB analysis of HK2 protein expression in three groups of cells; (F-H) Seahorse XFe96 cell metabolic analyzer used to measure ECAR and OCR levels in the three cell groups; (I-J) Glucose content and lactate production levels detected with respective assay kits. * represents P < 0.05

KAT2A enhances the stability and expression of CDC25A through acetylation

To gain a deeper comprehension of the mechanisms through which CDC25A promotes immune evasion in LUAD cells, we examined the TCGA-LUAD database and found that KAT2A was overexpressed in LUAD (Fig. 5A). Pearson correlation analysis also highlighted a positive correlation between CDC25A and KAT2A (Fig. 5B). qRT-PCR and WB analysis of KAT2A mRNA and protein in BEAS-2B, A549, H1299, and Calu-3 cells showed a pronounced upregulation in LUAD cell lines in contrast to human lung epithelial cells (Fig. 5C-E). Given the acetyltransferase activity of KAT2A, we hypothesized that KAT2A might regulate CDC25A expression via histone acetylation. The UCSC online website (https://genome.ucsc.edu/) was used to predict CDC25A’s histone acetylation peak (H3K27ac) (Figure S1), indicating that CDC25A may be acetylated to promote chromatin accessibility and activate gene expression. The binding relationship between KAT2A and CDC25A was examined through ChIP experiments. The results showed that the promoter region of CDC25A was significantly enriched in the KAT2A group compared to the IgG group (Fig. 5F). Additionally, KAT2A was knocked down in A549 cells, and qRT-PCR results confirmed successful transfection (Fig. 5G). The qRT-PCR results further verified that knocking down KAT2A reduced the mRNA level of CDC25A (Fig. 5H). However, WB analysis indicated that knocking down KAT2A significantly reduced the protein expression of both CDC25A and KAT2A in the cells (Fig. 5I). This further supported our hypothesis that KAT2A influenced CDC25A expression through the acetylation pathway. The CHIP experiment demonstrated that H3K27ac was significantly enriched in the promoter region of CDC25A compared to the IgG group. After knocking out KAT2A, the enrichment level of H3K27ac was significantly reduced, indicating that the acetylation modification level of H3K27ac in CDC25A was regulated by KAT2A expression (Fig. 5J). Additionally, to further explore the role of the KAT2A-mediated acetylation pathway in CDC25A expression, we overexpressed KAT2A and administered actinomycin D (Act D) to cells, creating cell groups of oe-NC+DMSO, oe-KAT2A+DMSO, and oe-KAT2A + Act D. qRT-PCR analysis showed that the overexpression of KAT2A elevated the mRNA levels of CDC25A, and the treatment with Act D 24 h following oe-KAT2A transfection led to a reversal of this effect on CDC25A mRNA expression (Fig. 5K). These findings demonstrated that H3K27ac acetylation, as mediated by KAT2A, plays an important role in the upregulation of CDC25A expression.

Fig. 5.

KAT2A mediates acetylation modification of H3K27ac to enhance CDC25A expression. A Analysis of KAT2A expression in LUAD tissues through the TCGA-LUAD database, with blue signifying normal tissue samples and red signifying tumor tissue samples; (B) Pearson correlation analysis between CDC25A and KAT2A; (C) qRT-PCR for KAT2A mRNA levels in human pulmonary epithelial cells and LUAD cell lines; (D-E) WB analysis of KAT2A protein expression levels in human pulmonary epithelial cells and LUAD cell lines; (F) ChIP experiment confirming the interaction between CDC25A and KAT2A; (G) qRT-PCR evaluation of transfection efficiency of sh-NC/sh-KAT2A plasmids; (H) qRT-PCR analysis of CDC25A mRNA expression levels after KAT2A knockdown; (I) WB analysis of KAT2A and CDC25A protein expression levels after KAT2A knockdown; (J) ChIP experiment verifying H3K27ac modification of CDC25A mRNA in A549 cells; (K) qRT-PCR detection of CDC25A mRNA expression following KAT2A knockdown. * represents P < 0.05

KAT2A interacts with CDC25A to promote LUAD cell immune escape

Building on the aforementioned research, we created cell groups of sh-NC + oe-NC, sh-KAT2A + oe-NC, and sh-KAT2A + oe-CDC25A for rescue studies. qRT-PCR was performed to detect the mRNA expression of CDC25A in cells. The results showed that knocking down KAT2A reduced the mRNA expression level of CDC25A, while knocking down KAT2A and overexpressing CDC25A upregulated the mRNA expression level of CDC25A (Fig. 6A). WB results demonstrated that knocking down KAT2A downregulated the protein expression level of CDC25A, while overexpressing CDC25A upregulated its protein expression level back to that of the control group (Fig. 6B). Cell viability, measured by the CCK-8 assay, showed a marked decrease in sh-KAT2A-transfected cells, which was countered by co-transfecting with oe-CDC25A (Fig. 6C). IF results indicated that knockdown of KAT2A promoted cell death, whereas simultaneous knockdown of KAT2A and overexpression of CDC25A suppressed the level of cell death (Fig. 6D). The colony formation assay for cell proliferation confirmed the trends observed in the CCK-8 assay (Fig. 6E). Using a cell metabolism analyzer, we quantified the ECAR and OCR levels in the cell groups. The sh-KAT2A + oe-NC group exhibited a steep decline in glycolysis activity and levels, and a substantial rise in ATP synthesis and peak OCR, when compared to the control group. The sh-KAT2A + oe-CDC25A group had much higher glycolysis levels and activity, and much lower ATP synthesis and peak OCR, compared to the sh-KAT2A + oe-NC group (Fig. 6F-G). Assay kits revealed that KAT2A knockdown led to higher residual glucose levels and reduced lactate production, which were reversed by CDC25A overexpression (Fig. 6H-I). To investigate how KAT2A influences LUAD cell immune evasion by mediating CDC25A, we co-cultured the three cell groups with activated CD8⁺T cells. Flow cytometry analysis of CD8⁺T cell activation revealed a significant increase in IFN-γ fluorescence intensity in the sh-KAT2A + oe-NC group compared to the control, and a decrease in the sh-KAT2A + oe-CDC25A group compared to the sh-KAT2A + oe-NC group (Fig. 6J). Subsequent analysis of LDH activity and levels of TNF-α, IFN-γ, Perforin, and Granzyme B in the co-culture supernatants confirmed that KAT2A, through its interaction with CDC25A, inhibited the cytotoxic ability of CD8⁺T cells towards tumor cells (Fig. 6K-L). Taken together, KAT2A promotes LUAD cell immune escape through interaction with CDC25A.

Fig. 6.

KAT2A inhibits the cytotoxic capacity of CD8⁺T cells against LUAD cells by interacting with CDC25A. A-B qRT-PCR and WB analysis of CDC25A mRNA expression levels in the three cell groups; (C) CCK-8 assay for cell viability assessment; (D) Live/death cell staining assay for detecting cell live and death; (E) Colony formation assay for cell proliferation capacity; (F-G) Seahorse XFe96 cell metabolic analyzer for ECAR and OCR levels in the three cell groups; (H-I) Glucose content and lactate production measured by respective assay kits; (J) Flow cytometry for CD8⁺T cell activation levels in the three groups; (K) LDH assay for the cytotoxic effect of CD8⁺T cells on A549 cells; (L) ELISA for TNF-α, IFN-γ, Perforin, and Granzyme B levels in co-culture supernatants. * represents P < 0.05

CDC25A downregulation enhances immunity and inhibits LUAD growth in mice

To examine the effects of CDC25A expression levels on tumor occurrence and growth, we constructed allograft tumor models by injecting mouse LUAD cells transfected with sh-NC + oe-NC, sh-KAT2A + oe-NC, and sh-KAT2A + oe-CDC25A into mice and monitored tumor development in each group. A pronounced reduction in both tumor volume and weight was observed in the sh-KAT2A + oe-NC group in comparison with the control group. Meanwhile, the sh-KAT2A + oe-CDC25A group displayed a substantial increase in tumor size and weight, surpassing the levels seen in the sh-KAT2A + oe-NC group (Figs. 7A-C). IHC analysis of the mouse tumor tissues revealed that the protein expression of KAT2A, CDC25A, and HK2 was downregulated in tumors with KAT2A knockdown, while in tumors with KAT2A knockdown and CDC25A overexpression, the expression of the proteins, except for KAT2A, returned to control levels (Fig. 7D). WB assay also showed that both KAT2A and CDC25A protein expression were downregulated in tumor tissues after knockdown of KAT2A, whereas CDC25A protein expression reversed to the control level after knockdown of KAT2A and overexpression of CDC25A, with no effect on KAT2A expression (Fig. 7E). IF analysis of tumor tissues revealed that the infiltration of CD8⁺T cells in the sh-KAT2A + oe-NC group was higher than that in the control group, and still higher when compared to the sh-KAT2A + oe-CDC25A group (Fig. 7F). Examination of peripheral blood for lactate and inflammatory factors indicated that sh-KAT2A transfection markedly elevated TNF-α, IFN-γ, Perforin, and Granzyme B, which were alleviated by co-transfection with sh-KAT2A + oe-CDC25A (Fig. 7G). Consequently, the KAT2A/CDC25A/HK2 axis affects the tumorigenicity of LLC cells and immune responses.

Fig. 7.

In vivo validation of the regulatory role of the KAT2A/CDC25A/HK2 axis on tumor growth and immune status. A-C Documentation of tumor growth in murine models; (D) IHC analysis of CDC25A, KAT2A, and HK2 expression levels in tumor tissues; (E) WB analysis of CDC25A and KAT2A expression levels in tumor tissues; (F) IF analysis of CD8⁺T cell infiltration in tumor tissues; (G) ELISA for TNF-α, IFN-γ, Perforin, and Granzyme B levels in peripheral blood. * represents P < 0.05

Discussion

By integrating bioinformatics, clinical sample collection, and cellular experiments, we identified the elevated expression of CDC25A in LUAD. Functional assays confirmed that this upregulation promotes the proliferation and viability of LUAD cells. We discovered that CDC25A boosted aerobic glycolysis by binding to HK2, thus inhibiting the cytotoxic effect of CD8⁺T cells on LUAD cells. To explore the regulatory mechanisms of CDC25A overexpression, we found an upregulation of the histone acetyltransferase KAT2A through bioinformatics, which was positively associated with CDC25A levels. Further analysis unveiled that KAT2A facilitated CDC25A expression via an acetylation mechanism. Thus, we proposed CDC25A as a candidate molecular target for impeding the immunological evasion of LUAD by modulating aerobic glycolysis.

During the 1920 s, cancer cells were confirmed to demand an intake of glucose from their surroundings that far exceeds the needs of normal cells to produce energy, a phenomenon termed the Warburg effect [25]. This effect has garnered attention in tumor biology, with numerous molecules related to it being considered promising targets for the development of anti-cancer therapies. For instance, HK2, with its inhibitor 3-BrPA, can impede aerobic glycolysis, deplete GSH, and stimulate ROS production, suggesting its potential as an anti-cancer drug [26]. With the gradual clarification of the molecular mechanisms of tumor development, Hanahan and Weinberg [27] in 2011 identified immune evasion and metabolic reprogramming as additional defining characteristics of cancer cells. A growing body of evidence suggests aerobic glycolysis could be a positive regulator in the immune evasion of tumor cells, but the detailed molecular mechanisms behind this process are elusive [28]. The molecular nexus between these two cancer hallmarks has been under increasing scrutiny in recent years. Professor Zhimin Lu’s team first detected a pronounced upregulation of PD-L1 in the presence of high glucose in many tumor cells. Their further investigation identified HK2 as an essential factor and its downregulation to promote the penetration of CD8⁺T cells [14]. Mechanistically, HK2 exerts its kinase function to phosphorylate IκBα, triggering its degradation. This event subsequently activates the NF-κB signaling pathway, increasing PD-L1 expression levels. This discovery elucidates the mechanism of how glycolysis drives tumor immune evasion. In this study, we discovered that CDC25A is upregulated in LUAD, and its knockdown could suppress aerobic glycolysis, enhancing the cytotoxicity of CD8⁺T cells towards LUAD cells. By investigating its mechanism, we found that CDC25A could bind to HK2, and its activity is reliant on the metabolic functions of HK2. Our study preliminarily confirmed the molecular pathway through which CDC25A influences the immune evasion of LUAD cells via aerobic glycolysis.

H3K27ac is a well-established epigenetic marker associated with active enhancers and promoters [29]. CBP/p300 can induce H3K27ac in the promoter, enhancer, and super enhancer regions of target genes, activating gene transcription [30]. When studying the mechanism of high expression of CDC25A, we found a positive correlation between KAT2A and CDC25A. KAT2A, as a member of the KAT family, is widely involved in the acetylation processes of histones and non-histones [31]. KAT2A-dependent CRC cells are found to exhibit higher levels of H3K27ac enrichment at gene loci associated with intestinal cell differentiation [23]. The research of Zhen et al. [32] showed that Kat2a increases the expression of Tfrc and Hmox1 by increasing H3K27Ac and H3K9ac enrichment in the promoter region and promotes ferroptosis in diabetes cardiomyopathy. We therefore sought to determine whether KAT2A regulates H3K27ac at the CDC25A locus. We performed ChIP experiments to confirm this speculation, demonstrating an interaction between CDC25A and KAT2A. The ChIP assay showed that CDC25A promoter region harbors more H3K27ac, which was reduced by KAT2A knockdown, suggesting a direct impact of KAT2A expression on the H3K27ac acetylation levels of CDC25A. Furthermore, we found that KAT2A promotes immune evasion of LUAD cells by interacting with CDC25A, offering a valuable lead for elucidating the interplay between acetylation modifications and immune evasion in LUAD.

In essence, our study has presented compelling evidence from both in vivo and in vitro experiments, clarifying the carcinogenic function of CDC25A in the progression of LUAD and the underlying molecular mechanisms that govern the immune escape in this tumor. Specifically, KAT2A-driven H3K27ac acetylation stabilizes CDC25A expression, and CDC25A, through its interaction with HK2, facilitates aerobic glycolysis, thereby modulating the inhibition of CD8⁺T cell cytotoxicity towards LUAD cells. Therefore, the KAT2A/CDC25A/HK2 axis could be a promising target for therapies aimed at reducing immune evasion in LUAD.

Authors’ contributions

WY C and XL T contributed to the study design. Q Y conducted the literature search. CS B, XL T and YF X acquired the data. WY C, YF X and CS B performed data analysis and drafted. XL T and Q Y revised the article. All authors gave the final approval of the version to be submitted.

Funding

1. Scientific Technology Plan Program for Healthcare in Zhejiang Province (NO. 2023RC098, 2023KY1196, 2023KY329);

2. National Oncology Clinical Key Speciality (NO. 2023-GJZK-001);

3. Key Construction Disciplines of Provincial and Municipal Co construction of Zhejiang [NO. 2023-SSGJ-002];

4. Zhejiang Province Postdoctoral Research Project (NO. ZJ2024156);

5. Key Project of Jiaxing First Hospital (NO. 2025-ZD-002);

6. Jiaxing Science and Technology Plan Project (NO. 2025CGZ041);

7. Key project of Jiaxing Health Science and Technology (NO. JWKD-25001).

Data availability

The data and materials in the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Informed consent was obtained from all participants, and the study was performed under the aegis of the Ethics and Research Committee approval at the Affiliated Hospital of Jiaxing University (The First Hospital of Jiaxing), IRB approval number: 2025-LP-313. All animal experiments were approved by the Ethics Committee of the Affiliated Hospital of Jiaxing University (The First Hospital of Jiaxing), Approval number JXYY2025-028, and were conducted in accordance with the institution’s guidelines for the care and use of animals.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.