The clinical data and gene expression profile within the TCGA-LUAD dataset were retrieved from the UCSC Xena database to investigate the expression pattern of NCAPD2 and its potential role in LUAD. The analysis included 513 tumor samples and 59 normal samples after excluding those with incomplete clinical information. Furthermore, the GSE30219 dataset from the GEO database was obtained to assess OS. All samples were divided into two groups based on the expression level of NCAPD2, either high or low.

The levels of NCAPD2 expression in various types of cancer were obtained from the TIMER database [29]. Comparative assessment of NCAPD2 expression within the TCGA-LUAD dataset followed, including the difference between normal and tumor specimens, and distinctions among various T, N, M, and pathologic stages. The plot R package was utilized to visualize these comparisons. Receiver operating characteristic (ROC) curves and area under the curve (AUC) were generated by the pROC R package to evaluate the diagnostic utility of NCAPD2. The ROC curve can assess the performance of classification models and the AUC was utilized as a metric to measure the overall performance of the model [30]. Typically, an AUC value exceeding 0.5 signifies a certain degree of classification ability in the model, with a higher value closer to 1 indicating superior performance.

For evaluating the correlation between NCAPD2 expression and OS, Kaplan-Meier survival curves were constructed utilizing the survival R package. The chi-squared test was used analyze the disparity in clinical features between the NCAPD2-high and NCAPD2-low groups. Survival outcomes were explored using univariate Cox regression analysis as well as multivariate Cox regression analysis to assess the impact of NCAPD2 expression levels [31].

The analysis of NCAPD2-related DEGs was conducted using the gene expression profile from the TCGA-LUAD dataset. This profile was derived from RNA-seq data obtained from frozen primary untreated tumors collected from patients with LUAD. The identification of DEGs associated with NCAPD2 was performed using the limma R package, with significance thresholds set at adjusted p < 0.05 and an absolute log fold change (logFC) ≥ 1. Volcano plots and heatmaps were generated to represent the DEGs. We subsequently applied the clusterProfiler R package to conduct Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses for NCAPD2-related DEGs. Terms with adjusted p values below 0.05 were considered statistically significant, and presented using the ggplot2 package [32]. To perform a more comprehensive investigation of the potential biological function of NCAPD2, we employe the clusterProfiler R package for gene set enrichment analysis (GSEA). This analysis utilized the c2.cp.reactome.v7.4.symbols.gmt dataset obtained from the Molecular Signatures Database [33].

To investigate the impact of NCAPD2 expression on immune infiltration in LUAD, 22 distinct types of infiltrating immune cells were scored for each sample according to the CIBERSORT algorithm [34] and the infiltration levels of immune cells were compared between the NCAPD2-high and NCAPD2-low groups. Moreover, the Spearman correlation test was employed to assess the correlation between the expression of NCAPD2 and the level of immune cell infiltration. To further substantiate the link between NCAPD2 expression and immune infiltration in LUAD, the TIMER database was explored.

A549 and H1299 cell lines of LUAD were acquired from the American Type Culture Collection (ATCC) (Manassas, VA, USA) and cultured in RPMI 1640 medium (Life Technology, CA, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco, Grand Island, USA) as per the provided guidelines [35].

To suppress NCAPD2 expression, siRNA transfection was performed. Following the manufacturer’s guidelines, we seeded 2 × 10⁵ cells into separate wells of a 6-well plate. After 24 h, siRNA molecules specifically targeting NCAPD2 were introduced into the cells using Lipofectamine™2000 transfection reagent (Invitrogen, MA, USA). All siRNAs, including three distinct siRNAs (si-1, si-2, and si-3) targeting NCAPD2, and a siRNA negative control (NC) were procured from ObiO Technology (Shanghai, China). There are the siRNA sequences below:

NCAPD2 si-1 sense: 5′-GTAUGUUGUGCAAGAGGUACU-3′

NCAPD2 si-1 antisense: 5′-AGUACCUCUUGCACAACAUAC-3′

NCAPD2 si-2 sense: 5′-CAAAGAAGAUACUCUGCAAUU-3′

NCAPD2 si-2 antisense: 5′-AAUUGCAGAGUAUCUUCUUUG-3′

NCAPD2 si-3 sense: 5′-GGCAGACAAGUCAGUGCUAGU-3′

NCAPD2 si-3 antisense: 5′-ACUAGCACUGACUUGUCUGC-3′.

Total RNA was extracted from both untransfected and transfected cells using EZ-press RNA Purification Kit (EZBioscience, Roseville, USA). Then, HiScript III RT SuperMix for qPCR (Vazyme, Nanjing, China) was used to convert the RNA into cDNA. After that, we measured NCAPD2 expression levels using qRT-PCR, with β-actin serving as the internal reference. The formula RQ = 2^−ΔΔCt was used to analyze data from three independent experiments. The primer sequences are provided below:

NCAPD2-forward: 5′-TGGAGGGGTGAATCAGTATGT -3′;

NCAPD2-reverse: 5′-GCGGGATACCACTTTTATCAGG-3′

β-actin-forward: 5′-CGGGAAATCGTGCGTGAC-3′

β-actin- reverse: 5′-CAGGAAGGAAGGCTGGAAG-3′

Total protein was extracted from the cells using RIPA lysis buffer (Beyotime, Shanghai, China). After the protein concentrations were quantified with the Enhanced BCA Protein Assay Kit (Beyotime, Shanghai, China), equal quantities of protein were subsequently separated via SDS-PAGE, transferred to a nitrocellulose (NC) membrane and blocked at room temperature with 5% BSA. The membranes were then incubated with primary antibodies overnight at 4°C. Antibodies against NCAPD2 and β-actin were sourced from Abcam (Cambridge, UK); antibodies for vimentin, E-cadherin, CDK2, CDK4, CDK6, CyclinD1 and CyclinA2 were obtained from Cell Signaling Technology (Beverly, MA, USA) and c-Myc antibody was procured from ABclonal (Wuhan, China). Following the washing of the membrane with TBST, the application of suitable secondary antibodies for a 1-h incubation period was carried out, followed by another round of TBST washing prior to detection. All the experiments were independently conducted three times.

Untransfected A549 and H1299 cells, as well as transfected cells, were seeded in a 96-well plate at a density of 3000 cells per well. A 10 µl volume of CCK8 reagent (Beyotime, Shanghai, China) was introduced to each well after 0, 24, 48, and 72 h, after which the cells were incubated for 2.5 h at 37°C in the absence of light. Cell viability was evaluated by measuring the optical density (OD) at a wavelength of 450 nm.

Cells from the siNCAPD2 group and the NC group were evenly distributed into a 6-well plate respectively with a seeding density of 1000 cells per well. Subsequently, the cells were placed in a CO₂ incubator at 37°C for 10 days, after which the medium was changed as required. Following the incubation period, the supernatant was aspirated, and the cells were fixed with a 4% paraformaldehyde solution (Beyotime, Shanghai, China) for 30 min. Afterwards, the cells were dyed with a 0.1% crystal violet solution (Beyotime, Shanghai, China), and cell quantification was conducted using ImageJ software (NIH, USA).

For the cell migration assay, the upper compartment of the transwell chambers was initially filled with 200 μL of serum-free medium containing 2 × 10⁴ cells. The cells included untransfected A549 and H1299 cells, as well as transfected cells. The lower compartment was supplemented with 600 μl of medium containing 10% FBS. After a 48-h incubation period, the lower surface of the upper chamber was fixed with a 4% paraformaldehyde solution for 30 min, and then stained with a 0.1% crystal violet solution. Eventually, any cells that had not migrated from the upper compartment were scratched thoroughly. The results were counted by microscope. For the invasion assay, first, a diluted Matrigel solution (1:19) (BD Biosciences, NJ, USA) was introduced into the upper chamber, and the chambers were preincubated at 37°C for 4 h. The same protocol used for the migration assay was then followed.

To conduct the cell cycle analysis, the cells were initially subjected to trypsin treatment, washed with phosphate-buffered saline (PBS), and subsequently fixed with 70% ethanol at a temperature of 4°C for 12 h. The staining procedure adhered to the protocol outlined in the cell cycle and apoptosis detection kit (Beyotime, Shanghai, China). Detection and analysis were carried out using the flow cytometer (BD Bioscience, San Jose, CA).

SPSS 22.0 (IBM, USA) and GraphPad Prism 9.0 (La Jolla, CA, USA) were used for statistical analysis. t-test or Wilcoxon test was applied for group comparisons, and the Log-rank test assessed OS differences. Spearman correlation analysis investigated gene relationships, and a p value below 0.05 indicated statistical significance.

Initially, TIMER 2.0 was used to assess the expression of NCAPD2 in diverse malignant tumors. In comparison to that in corresponding normal tissues, NCAPD2 demonstrated significantly elevated expression in a range of tumors, encompassing cervical squamous cell carcinoma, breast cancer, bladder cancer, bile duct cancer, head and neck squamous cell carcinoma, esophageal cancer, colon cancer, glioblastoma, lung squamous cell carcinoma, hepatocellular carcinoma, renal papillary cell carcinoma, lung adenocarcinoma, rectal adenocarcinoma, gastric cancer, thyroid cancer, and endometrial cancer (Fig. 1A). In contrast to that in normal tissues, the TCGA-LUAD dataset revealed a notable increase in NCAPD2 expression in tumor tissues (p < 0.001) (Fig. 1B). Moreover, to evaluate the diagnostic utility of NCAPD2 for discriminating between LUAD tissues and normal lung tissue, a ROC curve was generated, yielding an AUC of 0.85 (Fig. 1C). Additionally, compared to those in adjacent normal tissues in paired samples from TCGA-LUAD dataset, the expression levels of NCAPD2 in LUAD tissues were elevated (p < 0.01) (Fig. 1D). Furthermore, a significant increase in NCAPD2 expression was observed in the T2&T3&T4 stage group compared with T1 stage group. (p < 0.001), and both the stage II group and the stage III & stage IV group exhibited higher levels of NCAPD2 expression than did the stage I group (p < 0.05) (Figs. 1E–1H). These findings provide evidence of an association between elevated NCAPD2 expression and advanced T stage as well as pathologic stage in LUAD patients.

We gathered the expression profiles and clinical data of 513 cases in the TCGA-LUAD dataset and 83 cases in the GSE30219 dataset to determine whether NCAPD2 expression is associated with prognosis in LUAD. Analysis via the Kaplan-Meier method showed that increased NCAPD2 expression was linked to worse OS in both the TCGA-LUAD dataset (median OS: NCAPD2-high vs. NCAPD2-low = 1280 days vs. 1600 days, HR = 1.37, 95% CI 1.02–1.85, p < 0.05) and the GSE30219 dataset (median OS: NCAPD2-high vs. NCAPD2-low = 49 months vs. 137 months, HR = 1.99, 95% CI 1.07–3.73, p < 0.05) (Figs. 2A and 2B). Furthermore, strong associations were observed between high NCAPD2 expression and gender (p < 0.01), pathologic stage (p < 0.01), and smoking years (p < 0.01) (Table 1). The findings of the univariate Cox regression analysis demonstrated significant associations between advanced M stage (p = 0.005, HR = 2.15), N stage (p < 0.001, HR = 2.58), T stage (p = 0.003, HR = 1.69), pathologic stage (p < 0.001, HR = 2.94), and high NCAPD2 expression level (p = 0.035, HR = 1.37) and poorer prognosis. However, only N stage (p = 0.041, HR = 1.85) and T stage (p = 0.013, HR = 1.77) remained independent prognostic factors for LUAD according to multivariate Cox regression analysis (Table 2).

The analysis of DEGs identified 247 NCAPD2-related DEGs, consisting of 156 upregulated genes and 91 downregulated genes (Figs. 3A and 3B). To elucidate the biological function of the NCAPD2-related DEGs, we conducted GO function and KEGG pathway enrichment analyses, and the outcomes of the GO function enrichment analysis can be divided into three groups: biological processes (BP), cellular components (CC), and molecular functions (MF). Within the BP category, the NCAPD2-relted DEGs were significantly enriched in nuclear division and organelle fission. In CC, these genes were primarily associated with microtubule binding and tubulin binding. For MF, the enriched terms were predominantly linked to spindle and chromosomal regions (Fig. 3C). The KEGG pathway enrichment analysis primarily showed enrichment in pathways such as cell cycle (Fig. 3D). Furthermore, GSEA was employed to further explore the signaling pathways related to NCAPD2, revealing enrichment of pathways such as cell cycle, cell cycle mitotic, cell cycle checkpoints, and mitotic prometaphase (Fig. 3E).

To investigate the role of NCAPD2 in tumor immunity, the CIBERSORT algorithm was utilized to analyze the relationship between NCAPD2 expression and immune cell infiltration. The findings revealed significant differences in the infiltration levels of M1 macrophages, activated memory CD4+ T cells, naïve CD4+ T cells and memory B cells between the NCAPD2-high and NCAPD2-low groups. Furthermore, the infiltration levels of M1 macrophages, naïve CD4+ T cells and activated memory CD4+ T cells exhibited positive correlation with the NCAPD2 expression, while the infiltration level of memory B cells showed a negative correlation with NCAPD2 expression (Figs. 4A and 4B). Moreover, according to the TIMER database, the infiltration levels of CD8+ T cells (r = 0.185, p < 0.001), neutrophils (r = 0.268, p < 0.001), and macrophages (r = 0.216, p < 0.001) were positively correlated with NCAPD2 expression, whereas the infiltration level of B cells (r = −0.185, p < 0.001) was negative correlated with NCAPD2 expression (Figs. 4C–4F).

Our previous research illustrated a correlation between elevated NCAPD2 expression and an unfavorable prognosis, thus, the function of NCAPD2 was tested in LUAD cells in vitro. To assess the efficiency of NCAPD2 knockdown, the mRNA and protein expression levels of NCAPD2 in cells were measured after siRNA transfection. The results indicated that si-3 had the highest knockdown efficiency (Figs. 5A–5F), consequently, si-3 was selected for subsequent experiments. Assays for CCK8 and colony formation were performed to evaluate cell proliferation capacity. Overall, the proliferation of A549 and H1299 cells was strongly inhibited by the knockdown of NCAPD2, in contrast to that in the NC group (Figs. 5G–5J). The migration assay primarily evaluates the locomotive ability of cells within a two-dimensional plane, while the invasion assay specifically explores the invasive potential of cells by assessing their capacity to penetrate a supportive matrix that mimics the intricate extracellular environment encountered in vivo. The migratory and invasive capabilities of A549 and H1299 cells were significantly suppressed following the knockdown of NCAPD2 (Figs. 6A–6D). The effect of NCAPD2 on EMT was further evaluated in the context of tumor metastasis. Compared with the NC group, the siNCAPD2 group exhibited higher levels of the E-cadherin protein and lower levels of the vimentin protein both in A549 cells (Figs. 6E, 6F) and H1299 cells (Figs. 6G, 6H), indicating a shift toward an EMT-inhibiting phenotype.

The influence of NCAPD2 on the cell cycle progression of LUAD cells was investigated using flow cytometry. Compared to that in the NC group, knocking down NCAPD2 increased the proportion of A549 and H1299 cells in the G0/G1 phase and reduced the proportion of cells in the S phase. The percentage of A549 cells in the G2/M phase decreased slightly, while the percentage of H1299 cells in the G2/M phase did not significantly differ (Figs. 7A and 7B). The Western blot results indicated that NCAPD2 knockdown led to decreased protein levels of CyclinA2, CDK2 and CDK6 in both cell types, but there was no significant difference in CyclinD1 or CDK4 levels. Previous research has indicated that NCAPD2 co-expressed genes are enriched in the MYC target pathway in various tumors [23]. Additionally, c-Myc and its induced genes play crucial roles in cell cycle control and cell growth [36]. Thus, we analyzed the correlation between NCAPD2 expression and c-Myc expression in TCGA-LUAD cohort. Our analysis revealed a positive correlation between the expression levels of NCAPD2 and c-Myc (R = 0.47, p < 0.001) (Fig. 7C). Moreover, following NCAPD2 knockdown, both the mRNA and protein levels of c-Myc exhibited the most significant decreases (Figs. 7D–7I). Based on these results, knockdown of NCAPD2 induces cell cycle arrest in the G0/G1 and S phases and may do so by regulating c-Myc expression levels.