CDKN1A promotes paclitaxel resistance through mediating formation of polyploid giant cancer cells and enhancing neosis in non-small cell lung cancer

Introduction

Lung cancer is the cancer with the highest incidence rate and mortality rate in the world. In 2022, there were more than 2.5 million new cases, with more than 1.8 million deaths, accounting for one eighth (12.4%) of the world’s cancer patients. Among them, non-small cell lung cancer (NSCLC) accounts for over 85% (1). Although molecular targeted therapy and immunotherapy have significantly improved the prognosis of NSCLC in the past two decades, the overall survival rate of NSCLC is still low, with a 5-year survival rate of less than 20% for lung cancer patients in most countries (2). As well-known, NSCLC is a heterogeneous disease, and tumor metastasis and recurrence remain the main causes of death in advanced NSCLC patients.

Paclitaxel (PTX), as a classic anti-microtubule drug, can promote microtubule protein polymerization, inhibit depolymerization, maintain microtubule protein stability, and inhibit cell mitosis. It has been a first-line chemotherapy drug for advanced NSCLC patients for many years. The issue of PTX resistance in NSCLC chemotherapy remains prominent, as euchromatic histone lysine methyltransferase 2 (EHMT2)/nuclear transcription factor Y subunit alpha (NFYA)-aldehyde dehydrogenase 2 (ALDH2) family member signaling axis modules regulate the rapidly accelerated fibrosarcoma (RAF) pathway to affect PTX resistance in lung cancer (3). The mechanism of PTX resistance is complex, mainly including: (I) changes in microtubule protein interactions (microtubule associated proteins, stathmin, neutrophils, cilia, spindle associated proteins, and motor proteins); (II) changes in the expression and activity of ATP-binding cassette (ABC) superfamily multidrug effect transporters (including P-glycoprotein P-gp/ABCB1); (III) overexpression of anti-apoptotic proteins, apoptosis inhibitory proteins, and tumor suppressor proteins (4); (IV) the activity modification of cytokines, chemokines, and transcription factors related to signal transduction pathways (5-9). Although there have been many studies on PTX resistance, there is no effective method to address resistance currently.

It was reported that PTX could block cell mitosis, thereby inducing the formation of polyploid tumor giant cells (PGCCs), which further enhance tumor cell resistance to PTX (10). PGCCs are characterized with three or more unconventional nuclei in morphology and are recognized as a special subgroup in malignant tumor cells. PGCCs could further yield daughter cells through asymmetric division, which is closely related to the occurrence, drug resistance, and metastasis of tumors (10-12). PGCCs are observed in multiple cancers, such as colorectal cancer (11,13,14), ovarian cancer (15), breast cancer (16,17), prostate cancer (18), and nasopharyngeal cancer (14). However, there is limited research on PGCCs in NSCLC.

Neosis, a non-mitotic cell division, is observed during the process of PGCCs yielding monocytes (19). Neosis is not merely an alternative mode of division but also a potential escape mechanism from chemotherapy-induced senescence or mitotic catastrophe. These newly formed monocytes possess stem cell characteristics (20-22) and play an important role in tumor self-renewal (23-25) and tumor resistance (26,27). Studies on epigenesis have been focused on ovarian cancer, colorectal cancer, prostate cancer and breast cancer, while less attention is paid to PGCCs in lung cancer.

Cyclin dependent kinase (CDK) inhibitors 1A (CDKN1A/Waf1/Cip1/p21), a cell cycle regulatory factor, are known to induce cell cycle arrest by inhibiting the activity of CDKs, involving in cell apoptosis, DNA damage, and regulation of stem cells. CDKN1A can function as an oncogene or a tumor suppressor dependent on its context. Herein, we screened for differentially expressed gene (DEG) CDKN1A by performing RNA-seq analysis on primitive tumor cells, descendant PGCCs and daughter cells. CDKN1A was lowly expressed in tumor cells and highly expressed in PGCCs. When CDKN1A expression was suppressed, it could significantly reduce the formation of PGCCs. Low expression of CDKN1A might reduce the formation of PGCCs by regulating multiple signaling pathways and suppressing senescence. Therefore, targeting CDKN1A would contribute to reducing the formation of PGCCs, thereby lowering tumor recurrence. We present this article in accordance with the MDAR and ARRIVE reporting checklists (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0155/rc).

Methods

Cancer cell lines and cell culture

Human lung cancer cell lines A549 (ATCC^® CCL-185™) and H1299 (ATCC^® CRL-5803™) were originally obtained from the American Tissue Culture Collection (ATCC), and have been properly preserved in our laboratory. A549 cells were cultured in F-12K Medium (Cat No. 11765054, Thermo Fisher Scientific, Waltham, MA, USA), while H1299 cells were cultured in RPMI-1640 medium (Cat No. 11835030, Thermo Fisher Scientific, USA,) plus 10% fetal bovine serum (FBS) (Cat No. 10091148, Thermo Fisher Scientific, USA). All cells were routinely cultured at a humidified atmosphere of 5% CO₂ at 37 ℃.

Formation of PGCCs and budding daughter cells

When A549 and H1299 cells reached 80% confluency, PTX (Bristol Myers Squibb, Princeton, NJ, USA) was added at a final concentration of 60 ng/mL, and continued to culture for another 18 h. After removing the PTX-containing medium, phosphate-buffered saline (PBS) (Cat No. C0221A, Beyotime, Shanghai, China) was added to wash the cells and then replaced them with regular culture medium. The cells were then cultured for another 10 days. After the treatment of PTX, most of the tumor cells died, while only a small portion of cells formed PGCCs and survived. Cell observation and photography were performed using an inverted microscope (Primo Vert, Zeiss, Oberkochen, Germany). PGCCs began to bud on day 8 of culture, and the daughter cells increased with the extension of culture time.

Collection of primitive tumor cells, PGCCs and daughter cells

Primitive tumor cells were collected at culture confluency of 80%. PGCCs were collected on day 7 after PTX treatment, at which 40–50% culture confluency was reached, and the proportion of PGCCs was approximately 90%. Daughter cells were harvested on day 10. Specifically, single-cell suspensions were obtained by trypsin digestion, and filtered through a 10 µm sterile nylon mesh. The cells passing through the nylon mesh were collected as daughter cells, and the content was around 90%.

PGCCs content analysis with flow cytometry

Lung cancer cells A549 and H1299 were cultured in PTX-containing medium for 18 h, and PTX was then removed and washed once with PBS. The cells were further cultured in the conventional medium for 10 days, with the medium being replaced every 3 days. At 18 h, as well as on days 4, 8, and 10, the content of PGCCs was respectively detected using flow cytometry with the cell cycle staining Kit (Cat No. CCS012, Multi Sciences, Hangzhou, China). Briefly, the cells were digested with 0.25% trypsin (Cat No. 15050057, Thermo Fisher Scientific, USA), and 2×10⁵ cells were collected. Cells were washed with PBS and then centrifuged to remove the supernatant; 1 mL of DNA Staining solution and 10 µL of Permeabilization solution were added to the precipitant, and vortexed for 5–10 s to mix well. The cells were then incubated in the dark at room temperature for 30 min. Flow cytometry assay was then performed on a FACS Canto II flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA).

Immunofluorescence microscopy

Cells were seeded onto glass slides in 12-well plates and treated according to the experimental methods. Cells were fixed with 4% paraformaldehyde (Cat No. P0099, Beyotime, China) and blocked with 5% goat serum (Cat No. C0265, Beyotime, China). Subsequently, the fixed cells were incubated with the primary antibody overnight. Then, the secondary antibody was added and incubated at 37 ℃ for 1 h. Finally, DAPI (Cat No. P0131, Beyotime, China) was used to stain the cell nuclei. The image was observed and captured under a laser confocal microscope (LSM710, Zeiss, Germany). Quantitative analyses were carried out using ImageJ software.

RNA sequencing (RNA-seq) and bioinformatics analysis

Primitive tumor cells of A549, descendant PGCCs and daughter cells were extracted for total RNA using TRIzol reagent (Cat No. R0011, Beyotime, China) according to the instructions. RNA purity and quantification were detected using a Nano Drop 2000 spectrophotometer (Thermo Fisher Scientific, USA), and RNA integrity was evaluated using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Then a transcriptome library was constructed using the VAHTS Universal V10 RNA-seq Library Prep kit (Premixed Version) (Vazyme, Nanjing, China) according to the instructions. Transcriptome sequencing and analysis were conducted by Shanghai OE Biotechnology Co., Ltd. (Shanghai, China).

Briefly, the library was sequenced using Illumina Novaseq X Plus sequencing platform, and 150 bp double ended reads were generated to obtain raw reads of the sample. Raw reads in fastq format were processed using fastp software to remove low quality reads and obtain clean reads for subsequent data analysis. Genome alignment and gene expression level [Fragments Per Kilobase of transcript per Million mapped reads (FPKM)] calculation were performed using HISAT2 software, and the read counts of each gene were obtained through HTSeq-count. Gene counts were analyzed and plotted using principal component analysis (PCA) with R (v 3.2.0) to evaluate sample biological replicates. After RNA-seq, differential expression gene analysis was performed using DESeq2 software, and genes that met the threshold of q-value <0.05 and fold change >2 or fold change <0.5 were defined as DEGs. DEGs were subjected to hierarchical clustering analysis using R (v 3.2.0) to demonstrate the expression patterns of genes in different groups and samples. The heatmaps and volcano maps of DEGs from different samples were processed using the OEBiotech website (https://cloud.oebiotech.com). Subsequently, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was performed on DEGs based on the hypergeometric distribution algorithm to screen for significantly enriched functional entries. The genes related to significantly enriched functional items combined with DEGs (available online: https://cdn.amegroups.cn/static/public/tlcr-2026-1-0155-1.xlsx) to create a Jvenn diagram to identify intersecting genes (the Jvenn website, https://jvenn.toulouse.inrae.fr).

shRNA-mediated gene knockdown and lentiviral infection

The shRNA oligos of CDKN1A gene (Table S1) were designed and synthesized by Sangon Biotech (Shanghai) Co., Ltd. to construct the pLKO.1 shRNA plasmid. Briefly, 293T cells in logarithmic growth phase were seeded and cultured overnight in a 6-well plate. The cell culture medium was removed and washed gently with PBS once. Then, 1.5 mL of serum-free and antibiotic free Dulbecco’s Modified Eagle Medium (DMEM) basic medium (Cat No. 21068028, Thermo Fisher Scientific, USA) was added to the centrifuged cell pellet. Another 250 µL DMEM basic medium was added with 10 µL Lipofectamine 3000 (Cat No. L3000015, Thermo Fisher Scientific) and kept at room temperature for 5 min. Then, 250 µL of 4 µg lentiviral vector mixture (cloning plasmid, packaging plasmid psPAX2, and shell protein particle pMD2G, molar ratio 4:2:1) was added and kept at room temperature for 20 min. Finally, the mixture was slowly added to the cell culture dishes and cultured for another 5 h. The cell culture medium was replaced with DMEM complete medium containing 10% FBS and continued to culture for 72 h.

The cell culture medium was collected in a centrifuge tube and centrifuged at 1,500 rpm. A549 and H1299 cells in logarithmic growth phase were seeded in a 6-well plate ensure cell fusion reached 30% the next day. The cells in the 6-well plate were infected overnight with virus solution and polybrene (Cat No. C0351, Beyotime, China). The cell culture medium was replaced with complete medium containing 10% FBS and continued to culture for 72 h. Appropriate concentrations of specific antibiotics were added to the cell culture medium to screen positive cells two or three times. Cells were passaged and cultured for subsequent experiments. The mock pLKO was used as a negative control.

Cell senescence detection with senescence-associated beta-galactosidase (SA-β-gal) staining

Cells (2×10⁵) in logarithmic growth phase were seeded onto a 6-well plate and cultured overnight. Cells were cultured in PTX-containing medium for 18 h, and PTX was then removed and washed once with PBS. The cells were further cultured in the conventional medium for 10 days, with the medium being replaced every 3 days. At 18 h as well as on day 10, cell senescence was detected by Senescence β-Galactosidase Staining Kit (Cat No. C0602, Beyotime, China). Briefly, the cell culture medium was removed and washed gently with PBS once; 1 mL of β-galactosidase staining fixative was added to the cells and fixed at room temperature for 15 min. Cells were washed three times with PBS for 3 min each time, then 1 mL of staining solution was added to each well and incubated overnight at 37 ℃. Cell observation and photography were performed using a microscope (DMi1, Leica, Wetzlar, Germany).

Gene expression detection with quantitative real-time polymerase chain reaction (PCR)

Total RNA in cells was extracted using the RNA simple Total RNA Extraction Kit (Cat No. DP419, TIANGEN, Beijing, China). The primers were purchased from Sangon Biotech (Shanghai, China) Co., Ltd. The qPCR primer sequence was shown in Table S2. Briefly, cDNA was obtained and reverse transcribed using ABScript II RT Master Mix (Cat No. RK20402, ABclonal, Wuhan, China). PowerUp™ SYBR™ Green Master Mix (Cat No. A25741, Thermo Fisher Scientific, USA) was used to perform real-time PCR. The experiment was carried out on ABI viia7 (Thermo Fisher Scientific, USA) to amplify downstream gene specific sequences. The expression level of CDKN1A was normalized by glyceraldehyde-3-phosphate dehydrogenase (GAPDH).

The Ct values of the target gene and the reference gene were obtained from the exported data, and the Ct value of the reference gene was subtracted from the Ct value of the target gene to obtain the ΔCt value. The ΔCt value of the control group was subtracted from the ΔCt value of the test sample to obtain the ΔΔCt value. The gene expression fold was calculated as 2^−ΔΔCt, and the gene expression levels were plotted in GraphPad Prism 8 (GraphPad Software Inc., San Diego, CA, USA).

Western blot analysis

Western blot was carried out using the conventional method. In short, cell culture was placed on ice, then the supernatant was discarded and washed once with PBS. Radio Immunoprecipitation Assay (RIPA) lysis buffer (Cat No. P0013B, Beyotime, China) containing a mixture of protease and phosphatase inhibitors (Cat No. P1045, Beyotime, China) was added to lyse the cells. The cell lysate was collected in a 1.5 mL Eppendorf (EP) tube, and the protein concentration was measured by bicinchoninic acid (BCA) protein assay (Cat No. P0010, Beyotime, China). The loading buffer was added and heated to 100 ℃ for 10 min. The membrane was separated by electrophoresis on SDS polyacrylamide gel and electrotransferred to 0.22 µm polyvinylidene fluoride (PVDF) membrane. The membrane was immersed in 5% skimmed milk powder (Cat No. P0216, Beyotime, China) and sealed for 30 min, then the primary antibody was incubated over night at 4 ℃, and the secondary antibody was incubated at room temperature for 2 h. Enhanced chemiluminescence (ECL) plus (Cat No. P0018, Beyotime, China) was used to measure protein expression. GAPDH and Vinculin was used as a protein-loading control. Quantitative analyses were performed on the triplicate protein bands using ImageJ software. The antibodies used in this study were listed in Table S3.

Cell viability detection by cell counting kit-8 (CCK-8) assay

Lung cancer cells (A549, A549/CDKN1A-sh1, and A549/CDKN1A-sh2) with 80% confluence were harvested and counted. Briefly, 6×10³ cells were seeded into 96-well plates and cultured overnight. 60 µg/mL PTX was used to treat cells for 18 h. Fresh culture medium was replaced, and cells were cultured until day 7. CCK-8 solution (Cat No. C0039, Beyotime, China) was added to wells, and the plates were incubated at 37 ℃ for 1 h. Bio-Tek microplate reader (EON, Vermont, USA) was used to measure absorbance based on the optical density (OD) values at 450 nm. The cell viability was detected using the following formula: cell viability rate (%) = [(As-Ab) / (Ac-Ab)] × 100% (As, experimental well; Ac, control well; Ab, blank well). CCK-8 assays were conducted from days 0 to 7.

Animal procedures

All animal experiments were conducted strictly in accordance with the approved protocol by the Ethics Committee of Shanghai Chest Hospital (No. SYXK2023-0020). Briefly, 6-week-old BALB/c nude mice were purchased from Gempharmatech (Jiangsu, China). Primitive tumor cells of A549 cells (3×10⁶) were seeded in cell culture dishes. Then descendant PGCCs and daughter cells were induced by PTX. A549, descendant PGCCs and daughter cells were suspended in 100 µL PBS and inoculated subcutaneously into BALB/c nude mice. Tumor formation and tumor volume were observed and measured. Core formula for tumor calculation: V= (L × W²)/2, V: tumor volume; L: the length of the tumor, i.e., the maximum length; W: the short diameter of the tumor, which is the maximum width perpendicular to the long diameter. Tumor burden did not exceed 2,000 mm³, and all mice were euthanized humanely. The harvested tumor tissues were fixed in 4% formalin neutral fixative (Cat No. G1101, Servicebio, Wuhan, China) for subsequent experiments.

Hematoxylin and eosin (H&E) staining

HE staining is used to display the morphological and structural characteristics of different components in tissues or cells. Tissues samples of NSCLC were obtained from the BioBank of Shanghai Chest Hospital, between 2023 to 2024. Fresh patient and mouse tissue samples were soaked in 4% formalin neutral fixative 24 h and then embedded in paraffin, cut into 4 mm sections, baked overnight at 60 ℃, dewaxed with xylene, hydrated with alcohol, and counterstained with hematoxylin for 1 min and eosin for 1 min. Then the slices were dehydrated with alcohol and made transparent with xylene. The gum was dropped onto the already transparent slices and covered with glass slides. The image was observed and captured under a microscope (Axio Lab A1, Zeiss, Germany). The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethical Review Board of Shanghai Chest Hospital (No. KS25031), and the patient information was kept confidential. Informed consent was obtained from the patients.

Immunohistochemistry assay

Immunohistochemistry staining technology utilizes the principle of antigen antibody reaction to locate and qualitatively detect antigens within tissues or cells by labeling antibodies. Fresh patient and mouse tissue samples were soaked in 4% formalin neutral fixative for 24 h and then embedded in paraffin, cut into 4 mm sections, baked overnight at 60 ℃, dewaxed with xylene, hydrated with alcohol, and antigen repaired; 5% body surface area (BSA; Cat No. bs114, Biosharp, Hefei, China) was dropped onto the slides and incubated at room temperature for 1 h, followed by the addition of primary antibody and overnight incubation at 4 ℃. Then the slides were washed out, the secondary antibody was dropped onto the slides and incubated at room temperature for 2 h. DAB solution (Cat No. SB-D6077, Share-bio, Shanghai, China) was used for display reaction, followed by counterstaining, dehydration, and transparency. The gum was dropped onto the already transparent slices and covered with glass slides. The image was observed and captured under a microscope (Axio Lab A1, Zeiss, Germany). The antibodies used in this study were listed in Table S3.

Statistical analysis

All statistical analyses were carried out using GraphPad Prism 8 (GraphPad Software Inc., San Diego, CA, USA). All data represent the mean ± standard deviation (SD) of at least three independent experiments. The Student’s t-test was used to assess statistical significance between two groups. A two-way repeated-measures analysis of variance (ANOVA) followed by Sidak’s test was used for the statistical analysis of the animal experiment data. P value <0.05 was considered to have statistical significance. *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001; NS, not significant.

Results

PTX induces the formation of PGCCs in NSCLC

We first tested whether PTX could induce the formation of PGCCs in NSCLC cells. Lung cancer cell lines A549 and H1299 were cultured, and treated with 60 ng/mL PTX. The treatment scheme of cancer cells with PTX was shown in Figure 1A. After PTX treatment, it was observed under the microscope that most of the cells were killed, while a small portion of cells turned into multinucleated giant cells (Figure 1A, black arrows) on day 4, which had three or more nuclei. On day 8, polyploid giant cells came into budding (Figure 1A, green arrows), and gradually gave rise to smaller daughter cells (Figure 1A, red arrows). We performed flow cytometry to detect the changes in PGCCs content in cells treated with PTX at different time points. The results showed that after PTX treatment, PGCCs in A549 and H1299 reached the peak on the fourth day, which were 94.8% or 66.9% respectively. With the extension of time, the content of PGCCs gradually decreased (Figure 1B,1C). We applied laser confocal microscopy to detect the formation of PGCCs in cells after treatment with PTX, as well as the expression of F-actin and α-tubulin during the process from PGCCs budding to generation of daughter cells (Figure 1D). The cell surface area was statistically analyzed according to the images, and it was shown that PGCCs were significantly larger than primitive tumor cells, while daughter cells were the smallest (Figure 1E).

Figure 1 PTX induces the formation of PGCCs in NSCLC. (A) The process of PTX induced PGCCs in lung cancer cells and the experimental design and morphological changes of daughter cells by phase contrast microscope. PGCCs (black arrows) appeared on day 4 after PTX treatment, budded on day 8 (green arrows), and formed a large number of daughter cells on day 10 (red arrows). (B) Flow cytometry assays were used to detect the changes in PGCCs content at different time points during PTX treatment. (C) The ratios of PGCCs were shown. (D) The process of PTX induced PGCCs in lung cancer cells and the production of daughter cells were observed under laser confocal microscopy. Scale bar, 10 μm. (E) Quantitative analyses of cell surface area for A549 vs. A549-P vs. A549-D. Data were presented as mean ± SD; statistical significance was assessed using a Student’s t-test. ***, P<0.001; ****, P<0.0001. A549-P, A549-PGCCs; A549-D, A549-daughter cells. DAPI, 4',6-diamidino-2-phenylindole; NSCLC, non-small cell lung cancer; PTX, paclitaxel; PGCC, polyploid giant cancer cell; SD, standard deviation.

CDKN1A is enriched in PGCCs

After PTX induction, tumor cells form PGCCs, which then regenerated into daughter cells, leading to tumor recurrence. Distinctly, primitive tumor cells, descendant PGCCs and daughter cells should exert different biological characteristics. Therefore, to analyze the differential gene expression in primitive tumor cells, descendant PGCCs (accounted for approximately 90%) and daughter cells (accounted for approximately 90%), RNA-seq analysis were performed on the three types of cells, and compared the DEGs between A549-P vs. A549, A549-D vs. A549-P, and A549-D vs. A549, respectively (Figure 2A,2B). The results showed that there were 583 low expression genes and 119 high expression genes among the P<0.05 DEGs between A549-P vs. A549. Compared with A549-P, there were 88 low expression genes and 575 high expression genes among the DEGs with P<0.05. Compared with A549, A549-D had 106 low expression genes and 168 high expression genes among the P<0.05 DEGs. We analyzed the KEGG signaling pathways associated with DEGs and identified KEGG enrichment top 20 (Figure 2C). The results showed significant differences in the P53 signaling pathway, cell cycle, cell senescence, and other related signaling pathways. We screened for significantly DEGs (Figure 2D), including ATM, ATR CDKN1A, etc.

Figure 2 CDKN1A is enriched in PGCCs. (A) DEGs were identified using RNA-seq experiments. (B) Volcano plot showing the log₂FC vs. −log₁₀ (P value) of differential mRNA expression. (C) KEGG enrichment top 20 were shown. (D) DEGs were used to screen out the top 14 genes with significantly differentially expressed genes, and the (E) Venn diagram analysis intersected with genes involved in signaling pathways such as cell senescence, cell cycle, P53 signaling pathway, and oxidative stress. (F) TCGA analysis of the relationship between CDKN1A and patient survival. (G) Real-time qPCR and Western blot were used to detect the expression level of CDKN1A. DEGs, differentially expressed genes; FC, fold change; KEGG, Kyoto Encyclopedia of Genes and Genomes; PGCC, polyploid giant cancer cell; qPCR, quantitative polymerase chain reaction; TCGA, The Cancer Genome Atlas.

We intersected A549-D vs. A549-P, A549-P vs. A549 with genes from KEGG enrichment top 20 related to cell senescence, cell cycle, P53 signaling pathway, oxidative stress to screen for a common DEG, CDKN1A (Figure 2E). Is the expression of CDKN1A related to patient survival? To address this question, we conducted a correlation analysis between the 5-year survival and CDKN1A expression of patients with lung adenocarcinoma and lung squamous cell carcinoma in The Cancer Genome Atlas (TCGA) database (Figure 2F). The results showed that patients with high expression of CDKN1A had a shorter survival period. This indicated that high expression of CDKN1A was an important factor in poor prognosis in lung adenocarcinoma and squamous cell carcinoma. Then, we detected the mRNA and protein expression of CDKN1A in primitive tumor cells of A549 and H1299, descendant PGCCs and daughter cells. The results showed that CDKN1A was highly expressed in PGCCs (Figure 2G). These results suggested there was a correlation between the expression of CDKN1A and the formation of PGCCs.

CDKN1A mediates the formation of PGCCs induced by PTX

To investigate whether the content of PGCCs is affected by the expression of CDKN1A, we constructed CDKN1A low expression cell lines of A549 and H1299 and tested the knockdown effect. The results showed that at the mRNA and protein levels, the CDKN1A of the interference group cell lines was significantly reduced compared to the control group (Figure 3A,3B). In order to investigate whether the expression of CDKN1A is related to the content of PGCCs induced by PTX treatment of cells, we treated the cells with PTX and performed flow cytometry at 0, 18 h, and day 7 post-treatment. The results showed that compared with the control group, the content of PGCCs in CDKN1A low expression cell lines was significantly lower than that in the normal control group at 18 h and day 7 post-treatment (Figure 3C). These results suggested that low expression of CDKN1A could indeed reduce the content of PGCCs induced by PTX.

Figure 3 CDKN1A mediates the formation of PGCCs induced by paclitaxel. (A,B) Interference efficiency of CDKN1A was detected by qPCR and Western blot in A549 (A) and H1299 (B). (C) The ratios of PGCCs were statistically shown. Each bar represents the mean ± SD of three independent experiments. Student’s t-test was used to determine statistical significance. (D) Western blot was used to detect the expression changes of CDKN1A under different concentrations of UC2288 inhibitor treatment. (E) Quantification of PGCCs number after UC2288 treatment. (F) The morphological changes induced by PTX in lung cancer cell PGCCs and daughter cells at different time points after treatment with UC2288 were observed by phase contrast microscope. (G) The morphological changes induced by PTX in PGCCs and daughter cells of lung cancer cells were observed under laser confocal microscopy after treatment with UC2288 at different time points. Scale bar, 10 μm. (H) Quantification analysis of tumor cell surface area following PTX and UC2288 treatment over time. **, P<0.01; ***, P<0.001. DAPI, 4',6-diamidino-2-phenylindole; PGCC, polyploid giant cancer cell; PTX, paclitaxel; qPCR, quantitative polymerase chain reaction; SD, standard deviation.

To further verify that CDKN1A can indeed reduce the formation of PGCCs induced by PTX, we prepared the CDKN1A inhibitor UC2288 and tested whether inhibiting CDKN1A could reduce the content of PGCCs induced by PTX. We treated A549 cells with UC2288, a novel, cell-permeable, and selective active p21 attenuator, at 0, 1.25, 2.5, 5, and 10 µM for 24 h, and detected the expression of CDKN1A in these cells. At 5 µM, UC2288 could significantly inhibit the protein expression of CDKN1A (Figure 3D). A549 cells were treated with PTX at 5 and 10 µM, and the content of PGCCs was detected by flow cytometry after treated with PTX for 18 h, and 4, 7, and 10 days. The results showed that after adding 5 µM UC2288 inhibitor, the cells could reduce the formation of PGCCs at 18 h, and on days 4 and 7 (Figure 3E). We also found that on day 10, tumor cells treated with PTX alone were able to generate a large number of daughter cells, while cells treated with UC2288 inhibitor did not generate daughter cells on day 10, with only a small amount of PGCCs still in the initial budding state (Figure 3F). This indicated that when CDKN1A was inhibited, it could not only reduce the formation of PGCCs, but also inhibit the budding of PGCCs into daughter cells. Meanwhile, we examined the expression of F-actin and α-tubulin in cells at different time points after the addition of inhibitors (Figure 3G). After treated with UC2288, the cell surface area also decreased (Figure 3H). On day 10, tumor cells treated with PTX alone were able to bud and generate a large number of daughter cells, while cells treated with UC2288 inhibitor remained in PGCCs state on day 10, with only few cells budding and not yet forming daughter cells. This indicated that the expression of CDKN1A played an important role in the formation of PGCCs induced by PTX and neosis.

CDKN1A is associated with AKT, ERK, autophagy, stemness and EMT markers during the formation of PGCCs induced by PTX and neosis

After treatment with PTX, primitive tumor cells were induced to produce descendant PGCCs, which then budded to generate daughter cells. This is a complex biological process involving multiple mechanisms. Primitive tumor cells, descendant PGCCs and daughter cells exhibited different biological characteristics. Low expression of CDKN1A not only reduced the formation of PGCCs but also inhibited the production of daughter cells. To investigate the roles of CDKN1A in this process, we performed western blot assays to detect the related signaling pathways and correlative proteins. We detected the AKT signaling pathway, ERK signaling pathway, signal transducer and activator of transcription 3 (STAT 3) signaling pathway, and examined cell autophagy, proliferation, and stemness related proteins. The results showed that when PGCCs were induced after treatment with PTX on day 7, the expression of p-AKT increased. On day 10 (daughter cells were purified to approximately 90%), the expression of p-AKT decreased. In cells with CDKN1A knockdown, the overall expression of p-AKT decreased. After interference with CDKN1A, p-AKT was found to be lowly expressed on days 7 and 10. This indicated that the expression of CDKN1A played a role in the formation of PGCCs induced by PTX and neosis, possibly by regulating the expression changes of p-AKT (Figure 4A).

Figure 4 CDKN1A is associated with AKT, ERK, autophagy, stemness and EMT markers during the formation of PGCCs induced by paclitaxel and neosis. (A-D) Western blot analysis of protein expression in pLKO and CDKN1A-interfered A549 cells. p-AKT and p-ERK (A), LC3 (B), P-STAT3 and PCNA (C), p-histone H2AX and CD133, VIMENTIN (D). (E) Gray value analysis of Western blot. All data represent the mean ± SD of at least three independent experiments. P values were calculated using t-tests. **, P<0.01; ***, P<0.001; ns, not significant. (F) β-gal detection of cells at different time points. PGCC, polyploid giant cancer cell; SD, standard deviation.

In primitive tumor cells, after treated with PTX, p-ERK1/2 was highly expressed in PGCCs and daughter cells, and its expression level decreased overall after CDKN1A interference. It was shown that autophagy was involved in the formation of PGCCs, we then detected the expression of autophagy related protein microtubule-associated protein 1A/1B-light chain 3 (LC-3) during this process. It was found that autophagy related LC-3 was highly expressed in PGCCs, and its expression level decreased overall after CDKN1A interference (Figure 4B). P-STAT3 was also lowly expressed in PGCCs and daughter cells of day 7 after CDKN1A interference (Figure 4C). The vimentin related to epithelial mesenchymal transition also underwent changes, it was lowly expressed in PGCCs and daughter cells of day 7 after CDKN1A interference, and the CD133 related to stem cells was also observed to decrease on days 4, 7 and 10 after CDKN1A interference (Figure 4D), gray value analyses of protein bands were performed using Image J software (Figure 4E). These findings indicated that the formation of PGCCs by PTX induced primitive tumor cells and subsequent budding into daughter cells was a complex biological process associating with AKT, ERK, autophagy markers stemness and EMT-related proteins, but did not establish direct mechanistic involvement in each pathway. The causal hierarchy among these pathways remained unresolved.

Therefore, we analyzed the KEGG signaling pathways associated with DEGs and identified the top 20 differentially expressed KEGG enrichment genes. The results showed significant differences in the P53 signaling pathway, cell cycle, cell senescence, and other related signaling pathways. Once senescence occurs, PGCCs regulates the tumor microenvironment, which may promote chemotherapy resistance in cells (28). Senescence PGCCs formed due to cell cycle failure induced by PTX can lead to genome rearrangement and reprogramming, resulting in the emergence of new drug-resistant cancer cells (12). It was shown that rapid cell division of senescent tumor cells was the main reason for the ineffectiveness of chemotherapy drugs, and the escape senescence of PGCCs can generate daughter cells, leading to malignant cell transformation (29). Hence, we conducted β-gal detection on these cells. The results showed that A549 cells exhibited senescence both at 18 h and on day 10 post PTX-treatment. And we observed that there were a large number of PGCCs budded daughter cells on day 10 (Figure 4F). In A549 cells interfered with CDKN1A, there were fewer senescent cells (Figure 4F). These findings suggested that the level of CDKN1A in A549 cells could affect cell senescence, further affect the formation of PGCCs induced by PTX and neosis.

CDKN1A mediated tumor formation in mice

We further explored whether CDKN1A interference could suppress tumorigenesis via inhibition of PGCCs formation and neosis in vivo. So, A549, A549 silenced by CDKN1A-sh1, A549 treated with UC2288 inhibitor, and descendant PGCCs as well as daughter cells were subcutaneously injected into the axilla of mice to observe tumor formation (Figure 5A). The results showed that the tumorigenic ability of tumors decreased significantly when CDKN1A was interfered or inhibited (Figure 5B). Further, mouse tumor tissue was used for hematoxylin-eosin staining and immunohistochemical detection. Individual PGCCs could be observed in mouse tumors formed by primitive tumor cells (indicated by green arrows), while obvious PGCCs were observed in tumors formed by descending PGCCs. However, PGCCs were not significant in tumor cells formed by CDKN1A interfered or inhibited groups (Figure 5C). These results indicated that the content of PGCCs was related to the expression level of CDKN1A, and low expression of CDKN1A might reduce the generation of PGCCs.

Figure 5 CDKN1A mediated tumor formation in mice. (A) The size of tumors formed by inoculating mice (n=5) with tumor cells, PGCCs, daughter cells. (B) Statistical analysis of tumor volumes. All data represent the mean ± SD of at least three independent experiments. A two-way repeated-measures ANOVA followed by Sidak’s test was employed for the statistical analysis of the animal experiment data. *, P<0.05; **, P<0.01. (C) H&E staining of mice tumor tissues. Scale bar, 50 μm. The green arrows indicate PGCCs. (D) Immunohistochemical detection of p-ERK in mice tumor tissues. Scale bar, 50 μm. ANOVA, analysis of variance; H&E, hematoxylin and eosin; PGCC, polyploid giant cancer cell; SD, standard deviation.

Immunohistochemical analysis showed that p-ERK was highly expressed in the mouse tumors formed by descendant PGCCs and daughter cells, and when CDKN1A was lowly expressed or inhibited, p-ERK expression decreased in mouse tumors formed by descendant PGCCs and daughter cells (Figure 5D). These findings revealed that the ERK signaling pathway was involved in the biological process of PGCCs budding and generation of daughter in primitive cells, and its signaling pathway was influenced by the expression level of CDKN1A.

Clinical correlation of CDKN1A and PGCCs content

Then, we explored whether CDKN1A expression level was correlated with the content of PGCCs. We conducted an open label, observational retrospective study on 6 patients between 2023 and 2024, whose were newly diagnosed patients with stage II and III NSCLC (Table S4). Thereupon, clinical tissue samples of NSCLC were collected, and the content of PGCCs was observed by hematoxylin-eosin staining. These samples were divided into PGCC^high group (PGCCs content ≥3), and PGCC^low group (PGCCs content <3) (Figure 6A). The expression of CDKN1A in the tissue samples was determined by immunohistochemical detection method, and the correlation between CDKN1A expression and PGCCs content was accordingly analyzed in proportion. The results showed that in the group with high PGCCs content, the expression of CDKN1A was high, while in the group with low PGCCs content, the expression of CDKN1A was low (Figure 6B). This result suggested that the content of PGCCs in clinical tissue samples might be related to the expression level of CDKN1A.

Figure 6 Clinical correlation of CDKN1A and PGCCs content. (A) H&E staining of clinical samples from NSCLC patients with low or high levels of PGCC. The black arrows indicate PGCCs. (B) Immunohistochemical detection of CDKN1A expression in human tissues from NSCLC patients with different PGCCs content. Scale bar, 50 μm. H&E, hematoxylin and eosin; NSCLC, non-small cell lung cancer; PGCC, polyploid giant cancer cell.

Discussion

The reactivation of dormant tumor cells is an important factor in causing tumor recurrence and metastasis (30). During dormancy, tumor cells enter a temporary state of mitotic arrest (31,32). It was shown that PGCCs could produce offspring with self-renewal ability, which in turn led to tumor recurrence (12,33).

PTX, a broad-spectrum drug, was found to block meiosis, leading to the formation of PGCCs. Nevertheless, PGCCs generally underwent cell death due to meiotic failure. However, there was still data indicating that PGCCs could undergo neosis to form small monocytes (19). These small monocytes generated through neosis might exhibit stronger proliferation, invasion and migration abilities, and thus play a vital role in tumor recurrence (25,34). We wondered whether it was possible to reduce the generation of PGCCs, thereby reducing the generation of daughter cells and further reducing tumor recurrence.

Primitive tumor cells, descendant PGCCs and daughter cells possess different biological characteristics. To analyze the differential gene expression in primitive tumor cells, descendant PGCCs and daughter cells, we conducted RNA-seq analysis on the three types of cells and screened for DEGs. We identified a common DEG CDKN1A. The results showed that CDKN1A was highly expressed in PGCCs, and it was lowly expressed in primitive tumor cells and daughter cells. It suggested there was a correlation between the expression of CDKN1A and the formation of PGCCs.

The expression of CDKN1A varied in different types of tumors. Compared with normal tissues, CDKN1A was highly expressed in renal cancer and prostate cancer, while in colorectal cancer and lung cancer, CDKN1A was lowly expressed (35). Actually, the role of CDKN1A in various types of tumors has been extensively studied. It was reported that lamin B2 (LMNB2) could promote the proliferation of colon cancer cells by silencing CDKN1A expression (36), while cytoplasmic polyadenylation element binding protein 4 could induce G1 phase cell cycle arrest and inhibit the proliferation of renal cancer cells by increasing the stability of CDKN1A mRNA (37). It was also shown that targeting CDKN1A⁺ senescent stromal populations could enhance immunotherapy outcomes in advanced prostate cancer (38). However, there was limited researches focusing on the function of CDKN1A in the formation of PGCCs in lung cancer.

To investigate whether CDKN1A affected the formation of PGCCs induced by PTX, flow cytometry assay was conducted to analyze the content of PGCCs in cells with low expression of CDKN1A. It was shown that the content of PGCCs in CDKN1A low expression cell lines was significantly lower than that in the control group at 18 h and on day 7 post PTX-treatment. It indicated low expression of CDKN1A could indeed reduce the content of PGCCs induced by PTX. CCK-8 assay was performed to detected cell viability in A549 and CDKN1A knockdown cells after PTX treatment. Substantial decline in cell viability was shown in CDKN1A low expression cell lines from days 1 to 7. These results indicated that the increased death of these cells might cause the content of PGCCs decrease in CDKN1A low-expression cells (Figure S1).

To further verify that CDKN1A could indeed reduce the formation of PGCCs induced by PTX, we investigated the role of CDKN1A with its inhibitor UC2288 in vitro and in vivo. The results showed that after adding 5 µM UC2288 inhibitor, primitive tumor cells could reduce the formation of PGCCs. We also found that tumor cells treated with PTX alone were able to generate a large number of daughter cells on day 10, while those treated with UC2288 inhibitor did not generate daughter cells, in which there was only a small amount of PGCCs in the initial budding state. It revealed that when CDKN1A was inhibited, it could not only reduce the formation of PGCCs, but also inhibit the budding of PGCCs into daughter cells. These findings indicated that the expression of CDKN1A played an important role in the formation of PGCCs induced by PTX and neosis.

As mentioned above, it is a complex biological process that, after treatment with PTX, primitive tumor cells were induced to produce descendant PGCCs, which then budded to generate daughter cells, and multiple mechanisms were involved. To further investigate the roles of CDKN1A during PGCCs formation and neosis, we performed western blot assay to detect the related signaling pathways and proteins. AKT signaling pathway, ERK signaling pathway, STAT 3 signaling pathway, and autophagy, cell proliferation, as well as cell stemness related proteins were all identified. After treatment with PTX, the expression of p-AKT increased on day 7, while it decreased on day 10. As known, the PI3K/AKT signaling is crucial for the survival, proliferation, and metabolism of cancer cells. It was shown that enhanced activation of the PI3K/AKT/mTOR signaling pathway could promote PGCCs’ resistance to chemotherapy, thereby increasing cell survival and leading to tumor recurrence (39,40). As mentioned above, PGCC were formed on day 4 after PTX treatment, and they began to bud on day 8 to generate daughter cells. It might be due that cells in PGCC state need to accumulate energy for budding. Therefore, on day 7, the activation of the PI3K/AKT signaling pathway was enhanced, thereby improving cell survival and chemotherapy resistance. On day 10, the cells were in the state of daughter cells, and the expression of p-AKT in daughter cells was slightly downregulated.

After interference with CDKN1A, the overall expression of p-AKT decreased. p-AKT was found to be lowly expressed on days 7 and 10. It indicated that the expression of CDKN1A played a vital role in the formation of PGCCs induced by PTX and neosis. P-ERK1/2 was highly expressed in PGCCs and daughter cells, and its expression level decreased overall after CDKN1A interference. Autophagy related LC3 is highly expressed in PGCCs, and its expression level decreased overall after CDKN1A interference. P-STAT3 was lowly expressed in day 7 PGCCs and daughter cells after CDKN1A interference. Vimentin and CD133 also decreased in days 4, 7 and 10 after CDKN1A interference. These findings indicated that the formation of PGCCs by PTX induced primitive tumor cells and subsequent budding into daughter cells was a complex biological process associating with AKT, ERK, autophagy markers stemness and EMT-related proteins, but did not establish direct mechanistic involvement in each pathway. The causal hierarchy among these pathways are needed to determine.

We analyzed the KEGG signaling pathways associated with DEGs and identified KEGG enrichment top. The results showed significant differences in the P53 signaling pathway, cell cycle, cell senescence, and other related signaling pathways. Once senescence occurs, PGCCs regulates the tumor microenvironment, which may promote chemotherapy resistance in cells (28). It was found that rapid cell division of senescent tumor cells was the main reason for the ineffectiveness of chemotherapy drugs. The escape senescence of PGCCs could generate daughter cells, leading to malignant cell transformation (29). CDKN1A played an important role in cell cycle inhibition, cell senescence induction, tumor suppression, apoptosis, differentiation, and reprogramming of pluripotent stem cells (41).

Our research found that senescent cells decreased when CDKN1A was lowly expressed. And on day 10, the control group A549 generated many daughter cells, while the CDKN1A interference group had not yet generated daughter cells. It indicated that when CDKN1A was interfered, it could reduce cell senescence, thereby reducing PGCCs budding. Our study revealed that CDKN1A affected the formation of PGCCs induced by PTX and neosis through cell senescence.

When we subcutaneously injected primitive tumor cells of A549, A549 interfered with CDKN1A, A549 treated with UC2288 inhibitor, as well as descendant PGCCs and daughter cells into the axilla of mice, we found that tumorigenic ability of tumors decreased significantly when CDKN1A was interfered or inhibited. Individual PGCCs could be observed in mouse tumors formed by primitive tumor cells, while obvious PGCCs were observed in tumors formed by descending PGCCs. However, PGCCs were not significant in tumor cells formed by CDKN1A interfered or inhibited groups. These results indicated that the content of PGCCs in mice was related to the expression level of CDKN1A, and low expression of CDKN1A might reduce the generation of PGCCs. p-ERK was highly expressed in the mouse tumors formed by descendant PGCCs and daughter cells, and when CDKN1A was lowly expressed or inhibited, p-ERK expression decreased in mouse tumors formed by descendant PGCCs and daughter cells. It suggested that the ERK signaling pathway was involved in the biological process of PGCCs budding and generation of daughter in primitive cells, and its signaling pathway was influenced by the expression level of CDKN1A.

The expression of CDKN1A in the tissue samples from NSCLC patients was detected by immunohistochemistry, and the correlation between CDKN1A expression and PGCCs content was analyzed. The results showed that in the group with high PGCCs content, the expression of CDKN1A was high, while in the group with low PGCCs content, the expression of CDKN1A was low. It suggested that the content of PGCCs in clinical tissue samples might be related to the expression of CDKN1A.

To explore whether the subcellular localization of CDKN1A would change during the budding process of PGCCs, immunofluorescence microscopy was performed. Our data showed that CDKN1A was mainly located in the nucleus of tumor cells. In contrast, CDKN1A significantly transferred to the cytoplasm in PGCCs. In daughter cells, CDKN1A relocated to the nucleus (Figure S2). These findings suggested that the enrichment of CDKN1A in the cytoplasm might be associated with PGCC-mediated chemotherapy resistance and neosis. Much better than this, we should further explore the mechanism of CDKN1A in PGCCs formation in ongoing investigations, such as using CDKN1A phosphorylation mutants (such as T145A/S146A) to explore the relationship between AKT mediated phosphorylation and cytoplasmic localization, and in combination CDKN1A overexpression cells, functional mutants, and with specific inhibitors of AKT, ERK, STAT3 and autophagy to clarify the key pathway in PGCCs formation. In that, it would be possible to reduce PGCC-mediated chemotherapy resistance by targeting this pathway.

Moreover, we should address senescence of these cells to our future investigations. For instance, it would be necessary to clarify whether senescence is a prerequisite state that enables neosis, and explore whether the elevated CDKN1A levels maintain a senescent state that subsequently facilitates emergence of daughter cells.

Besides, some more limitations of this study should be acknowledged. Firstly, the presented signaling pathways were primarily based on correlation analysis, which could not establish direct mechanistic causality. There was a lack of evidences that CDKN1A directly bound or interacted with the regulatory factors of the identified pathway. Secondly, the effect of UC2288 inhibitor on CDKN1A is an important aspect of this study. However, the specificity and off-target effects of the inhibitor still need to be taken into consideration. Subsequent experiments are required to investigate the overexpression of inhibitor resistant CDKN1A variants or constitutive activation of downstream effectors (such as AKT), so as to clarify whether the observed phenotypic changes are specifically mediated by CDKN1A inhibition. Thirdly, compared to our conventional animal experiments, more precise genetic models can be applied to conduct more in-depth research. Finally, due to the limited sample size in the clinical correlation analysis, validation in larger independent cohorts is necessary to determine the clinical significance of this study.

Hereinto, we basically demonstrated that targeting CDKN1A might help to reduce the formation of PGCCs, thereby reducing neosis and reversing chemoresistance of PTX. Currently, direct pharmacological inhibitors of CDKN1A have not been used clinically, yet alternative approaches could be considered. For example, in combination PTX with HDAC inhibitors or AKT inhibitors would be a creditable attempt to indirectly downregulate CDKN1A. However, more rigorous research and optimization of the combined treatment regimens are still needed. Hopefully, this approach might provide a new strategy for overcoming PTX resistance and improving the prognosis of cancer patients.

Conclusions

Our research revealed that CDKN1A was highly expressed in PGCCs induced by PTX, it mediated the formation of PGCCs and enhancing neosis. CDKN1A was associated with p-AKT, p-ERK, stemness-associated genes, and autophagy signaling pathways in this process. p-AKT and p-ERK were found to be lowly expressed in PGCCs and daughter cells after CDKN1A was interfered with. Low expression of CDKN1A could reduce the senescence of PGCCs too. Our results indicated that CDKN1A knockdown was associated with dysregulation of AKT, ERK, autophagy markers and stemness/EMT-related proteins, and reduced cell senescence, thereby reducing neosis of PGCCs. A critical role for CDKN1A in PTX-induced PGCCs formation and subsequent neosis in NSCLC was identified. CDKN1A inhibition reduced PGCCs formation, suppressed neosis and enhanced tumor sensitivity in vivo, thereby supporting its potential as a therapeutic target to overcome PTX resistance.

Therefore, CDKN1A might be a potent target for the treatment of PTX chemotherapy resistance.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the MDAR and ARRIVE reporting checklists. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0155/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0155/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0155/prf

Funding: This study was supported by the Foundation of Shanghai Key Laboratory of Thoracic Tumor Biotherapy (No. 2025SZ1009), the Nurture Projects for Basic Research of Shanghai Chest Hospital (No. 2022YNJCQ11), and the National Natural Science Foundation of China (No. 82073203).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-1-0155/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethical Review Board of Shanghai Chest Hospital (No. KS25031), and the patient information was kept confidential. Informed consent was obtained from the patients. All animal experiments were performed under a project license (No. SYXK2023-0020) granted by Ethics Committee of Shanghai Chest Hospital, in compliance with Shanghai Chest Hospital institutional guidelines for the care and use of animals.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024;74:229-63. [Crossref] [PubMed]
2. Wang M, Herbst RS, Boshoff C. Toward personalized treatment approaches for non-small-cell lung cancer. Nat Med 2021;27:1345-56. [Crossref] [PubMed]
3. Wang W, Wang J, Liu S, et al. An EHMT2/NFYA-ALDH2 signaling axis modulates the RAF pathway to regulate paclitaxel resistance in lung cancer. Mol Cancer 2022;21:106. [Crossref] [PubMed]
4. Wang F, Xu X, Guan B, et al. Regulation and reversal of paclitaxel resistance via the STAT1 mediated apoptotic pathway in ovarian cancer. Int J Oncol 2026;68:19. [Crossref] [PubMed]
5. Trnkova L, Burikova M, Soltysova A, et al. Molecular profiling of chemotherapy-resistant breast cancer reveals DNA methylation remodeling associated with the acquisition of paclitaxel resistance. Drug Resist Updat 2026;85:101350. [Crossref] [PubMed]
6. Liu T, Yang D, Wei Q, et al. The RNA-stability-independent role of the RNA m(6)A reader YTHDF2 in promoting protein translation to confer tumor chemotherapy resistance. Mol Cell 2025;85:2320-2336.e9. [Crossref] [PubMed]
7. Zhang J, Lv S, Peng X, et al. CircERC1 facilitates chemoresistance through inhibiting pyroptosis and remodeling extracellular matrix in pancreatic cancer. Mol Cancer 2025;24:185. [Crossref] [PubMed]
8. Rong D, Gao L, Chen Y, et al. Suppression of the LKB1-AMPK-SLC7A11-GSH signaling pathway sensitizes NSCLC to albumin-bound paclitaxel via oxidative stress. Redox Biol 2025;81:103567. [Crossref] [PubMed]
9. Lai H, Wang R, Li S, et al. LIN9 confers paclitaxel resistance in triple negative breast cancer cells by upregulating CCSAP. Sci China Life Sci 2020;63:419-28. [Crossref] [PubMed]
10. Amend SR, Torga G, Lin KC, et al. Polyploid giant cancer cells: Unrecognized actuators of tumorigenesis, metastasis, and resistance. Prostate 2019;79:1489-97. [Crossref] [PubMed]
11. Zhao Y, Lu T, Song Y, et al. Cancer Cells Enter an Adaptive Persistence to Survive Radiotherapy and Repopulate Tumor. Adv Sci (Weinh) 2023;10:e2204177. [Crossref] [PubMed]
12. Chen J, Niu N, Zhang J, et al. Polyploid Giant Cancer Cells (PGCCs): The Evil Roots of Cancer. Curr Cancer Drug Targets 2019;19:360-7. [Crossref] [PubMed]
13. Zhou X, Ning Y, Wang X, et al. Activated PLK1 promotes the migration and invasion of polyploid giant cancer cells with daughter cells via β-catenin/STAT3 signaling pathway. Hum Cell 2025;38:136. [Crossref] [PubMed]
14. You B, Xia T, Gu M, et al. AMPK-mTOR-Mediated Activation of Autophagy Promotes Formation of Dormant Polyploid Giant Cancer Cells. Cancer Res 2022;82:846-58. [Crossref] [PubMed]
15. Xu H, Zeng S, Wang M, et al. Cytoplasmic SIRT1 enhances the stemness of polyploid giant cancer cells by promoting β-catenin protein stability and nuclear accumulation in ovarian carcinoma upon neoadjuvant chemotherapy. Cancer Lett 2026;639:218193. [Crossref] [PubMed]
16. Saini G, Joshi S, Garlapati C, et al. Polyploid giant cancer cell characterization: New frontiers in predicting response to chemotherapy in breast cancer. Semin Cancer Biol 2022;81:220-31. [Crossref] [PubMed]
17. Ma Y, Shih CH, Cheng J, et al. High-Throughput Empirical and Virtual Screening To Discover Novel Inhibitors of Polyploid Giant Cancer Cells in Breast Cancer. Anal Chem 2025;97:5498-506. [Crossref] [PubMed]
18. Mannan R, Wang X, Bawa PS, et al. Polypoidal giant cancer cells in metastatic castration-resistant prostate cancer: observations from the Michigan Legacy Tissue Program. Med Oncol 2020;37:16. [Crossref] [PubMed]
19. Mahaddalkar T, Banerjee A, Ketkar M, et al. Aurora Kinase A and B inhibition abrogates ‘Neosis’, a non-mitotic cell division of GBM residual cells and prevents GBM recurrence. Oncogene 2025;44:2103-15. [Crossref] [PubMed]
20. Niu N, Mercado-Uribe I, Liu J. Dedifferentiation into blastomere-like cancer stem cells via formation of polyploid giant cancer cells. Oncogene 2017;36:4887-900. [Crossref] [PubMed]
21. White-Gilbertson S, Voelkel-Johnson C. Giants and monsters: Unexpected characters in the story of cancer recurrence. Adv Cancer Res 2020;148:201-32. [Crossref] [PubMed]
22. Zhang S, Mercado-Uribe I, Xing Z, et al. Generation of cancer stem-like cells through the formation of polyploid giant cancer cells. Oncogene 2014;33:116-28. [Crossref] [PubMed]
23. Rejili M, Hashemi F. Unveiling the role of PGCCs in tumor recurrence and therapeutic resistance: Hidden architects of cancer’s comeback. Pathol Res Pract 2026;278:156347. [Crossref] [PubMed]
24. Zhao R, Song Y, Xie J, et al. S100A4 facilitates radiation-induced tumor repopulation by driving polyploid giant cancer cells budding. Cancer Lett 2025;633:218017. [Crossref] [PubMed]
25. Zhang Z, Feng X, Deng Z, et al. Irradiation-induced polyploid giant cancer cells are involved in tumor cell repopulation via neosis. Mol Oncol 2021;15:2219-34. [Crossref] [PubMed]
26. Esmatabadi MJ, Bakhshinejad B, Motlagh FM, et al. Therapeutic resistance and cancer recurrence mechanisms: Unfolding the story of tumour coming back. J Biosci 2016;41:497-506. [Crossref] [PubMed]
27. Liu J, Liu K, Yan M, et al. Dormant and activated polyploid giant cancer cells: From chemoradiotherapy resistance to cancer progression. Cancer Lett 2026;636:218137. [Crossref] [PubMed]
28. Bharadwaj D, Mandal M. Senescence in polyploid giant cancer cells: A road that leads to chemoresistance. Cytokine Growth Factor Rev 2020;52:68-75. [Crossref] [PubMed]
29. Zhang S, Mercado-Uribe I, Sood A, et al. Coevolution of neoplastic epithelial cells and multilineage stroma via polyploid giant cells during immortalization and transformation of mullerian epithelial cells. Genes Cancer 2016;7:60-72. [Crossref] [PubMed]
30. Aguirre-Ghiso JA. How dormant cancer persists and reawakens. Science 2018;361:1314-5. [Crossref] [PubMed]
31. Recasens A, Munoz L. Targeting Cancer Cell Dormancy. Trends Pharmacol Sci 2019;40:128-41. [Crossref] [PubMed]
32. Marx V. How to pull the blanket off dormant cancer cells. Nat Methods 2018;15:249-52. [Crossref] [PubMed]
33. Moein S, Adibi R, da Silva Meirelles L, et al. Cancer regeneration: Polyploid cells are the key drivers of tumor progression. Biochim Biophys Acta Rev Cancer 2020;1874:188408. [Crossref] [PubMed]
34. Wang X, Zheng M, Fei F, et al. EMT-related protein expression in polyploid giant cancer cells and their daughter cells with different passages after triptolide treatment. Med Oncol 2019;36:82. [Crossref] [PubMed]
35. Bartha Á, Győrffy B. TNMplot.com: A Web Tool for the Comparison of Gene Expression in Normal, Tumor and Metastatic Tissues. Int J Mol Sci 2021;22:2622. [Crossref] [PubMed]
36. Dong CH, Jiang T, Yin H, et al. LMNB2 promotes the progression of colorectal cancer by silencing p21 expression. Cell Death Dis 2021;12:331. [Crossref] [PubMed]
37. Di J, Wang H, Zhao Z, et al. CPEB4 Inhibit Cell Proliferation via Upregulating p21 mRNA Stability in Renal Cell Carcinoma. Front Cell Dev Biol 2021;9:687253. [Crossref] [PubMed]
38. Zhou L, DeMarco KD, Murphy KC, et al. p21-Positive Senescent Stromal Cells Promote Prostate Cancer Immune Suppression and Progression That Can Be Reversed by Senolytic Therapy. Cancer Discov 2026;16:571-91. [Crossref] [PubMed]
39. Temaj G, Saha S, Chichiarelli S, et al. Polyploid giant cancer cells: Underlying mechanisms, signaling pathways, and therapeutic strategies. Crit Rev Oncol Hematol 2025;213:104802. [Crossref] [PubMed]
40. Glaviano A, Foo ASC, Lam HY, et al. PI3K/AKT/mTOR signaling transduction pathway and targeted therapies in cancer. Mol Cancer 2023;22:138. [Crossref] [PubMed]
41. Yan J, Chen S, Yi Z, et al. The role of p21 in cellular senescence and aging-related diseases. Mol Cells 2024;47:100113. [Crossref] [PubMed]