3.1 Clinical data of patients
The mean age of the participants was 60 (26–78) years, and 53% of patients were aged > 60 years. Moreover, there was a high proportion of male participants (80%).The participants were followed-up from 2005 to 2013. By the end of the follow-up period, 84 patients had distant metastases and tumor recurrence, and 91 died.
3.2 Expression of NCAPG in NSCLC tissues
Immunohistochemistry was used to detect NCAPG expression in 140 pairs of NSCLC tissues, with highly phenotyped renal papillary cell carcinoma serving as a strong positive control and low-expressing paraneoplastic lung tissue as a negative control (Fig.1a-d). In addition, NCAPG expression in NSCLC tissues was analysed based on TCGA data. The results showed that NCAPG expression was upregulated in NSCLC tissue samples. We found that the NCAPG expression was higher in NSCLC, LUAD, and LUSC tissue samples than in adjacent normal tissue samples (p < 0.001, Fig. 2a–c).
3.3 Diagnostic value of NCAPG expression in NSCLC
We assessed the diagnostic value of NCAPG expression according to the receiver operating characteristic curves, with an area under the curve of 0. 973 (95% confidence interval: 0.963–0.984; p < 0.001; Fig. 2d). Results revealed that NCAPG has good specificity and sensitivity for diagnosing NSCLC.
3.4 Association between a high NCAPG expression and poor prognosis
We used the median expression level as a classification criterion. Next, the patients were divided based on the expression of NCAPG and clinicopathological parameters. Moreover, the clinicopathological characteristics were compared. As shown in Table 1, there was no statistically significant correlation between NCAPG and age (p = 0.866), sex (p = 0.833), tumor site (p = 0.865), and other clinicopathological features. However, NCAPG expression was significantly correlated with tumour stage (p < 0.001) (Fig. 2e) and os events (p =0.007) (Fig. 2f).
Cox regression analysis was performed to explore the factors affecting the survival time of patients with NSCLC. As shown in Table 2, based on the univariate analysis, NSCLC stage and NCAPG expression were significantly associated with survival time (p < 0.05). Multifactorial analysis revealed that NCAPG expression was a prognostic factor for survival outcome in NSCLC (p = 0.008), independent of the clinical factors. Therefore, a high NCAPG expression was associated with poor prognosis in patients with NSCLC. The Kaplan- Meier method was used to obtain survival curves to evaluate the correlation between NCAPG and NSCLC prognosis based on long-term follow-up data. Results showed that patients with NSCLC with a high NCAPG expression had shorter survival than those with a low NCAPG expression (p < 0.001, Fig. 2g). Further, the survival of patients with LUSC and LUAD who presented with a high NCAPG expression was shorter than that of patients with a low NCAPG expression (p < 0.05, Figs. 2h, i). Taken together, NCAPG could be an independent prognostic biomarker for poor prognosis in NSCLC.
3.5 Relationship between NCAPG expression and immune cell infiltration
NCAPG is a critical factor influencing the immune status of NSCLS. Next, we analyzed the correlation between NCAPG expression and various immune cells in NSCLC (Fig. 3a). Data from CIBERSORT were used to analyze the effect of NCAPG on the extent of immune-related cell infiltration. In total, 22 immune cells were analyzed in the study. The difference in the proportion of 22 NSCLC species in the tumor tissues between the high and low NCAPG expression groups was presented as the violin plot (Fig. 3b). Subsequent analysis revealed that T cells, CD8T cells, dendritic cells, macrophages, mast cells, NK cells, and resting NK cells were negatively correlated with NCAPG expression (p < 0.001). Therefore, NCAPG expression was significantly correlated with the immune status of TME in NSCLC.
3.6 Analysis of the correlation between NCAPG expression and six types of infiltrating immune cells
In patients with LUSC and LUAD, we analyzed the relationship between NCAPG expression and six types of infiltrating immune cells, which were as follows: B cells, CD8 T cells, CD4 memory T cells, neutrophils, macrophages, and dendritic cells.
Data showed that NCAPG expression was correlated with tumor-filtering immune cells, B cells (p < 0.0001), CD8 T cells (p < 0.0001), CD4 memory T cells (p < 0.0001), neutrophils (p < 0.0001), macrophages (p < 0.0001), and activated-phase dendritic cells (p < 0.0001) (Fig. 4a). Therefore, NCAPG and its related genes were important in immune cell infiltration in tumor pathology.
We used the timer database, Kaplan–Meier evaluation of survival curves, and analyzed by log-rank test. Results showed that NCAPG expression affected the prognosis of LUSC and LUAD via immune cell infiltration. Dendritic cell (p < 0.05) and B-cell (p < 0.001) infiltration was significantly associated with NCAPG survival in LUAD (Fig. 4b). However, immune cell infiltration (CD4 T cells, B cells, dendritic cells, and macrophages) was not significantly correlated with NCAPG survival in LUSC. Therefore, B-cell and dendritic cell infiltration affected the survival outcomes of patients with LUAD. Thus, NCAPG was involved in the regulation of immune cell infiltration in LUAD.
3.7 Enrichment of gene sets in NCAPG expression phenotypes
GSEA was performed to explore the mechanisms of NCAPG in NSCLC. GSEA-based NCAPG-related signaling pathways were analyzed using TCGA datasets in the low and high NCAPG groups. We performed GO enrichment analysis and GSEA of NSCLC based on data from TCGA, which was significantly enriched in GO terms, including the components of mitotic nuclear division, mitotic spindle, microtubulin binding, cell cycle, humoral response, multivesicular body (Fig. 5a). Genes that were significantly associated with NCAPG are shown (Fig. 5b,c). In samples with a low NCAPG expression, B-cell and T-cell receptor signaling pathways were enriched, and this mechanism was associated with cancer progression (Fig. 5d-i). GSEA showed that cell cycle, proliferation, and adhesion-linked gene sets were enriched in samples with high NCAPG expression (Fig. 5j-o). Further, several cell-associated genomes were enriched. Hence, NCAPG might also play a role in promoting lung cancer cell proliferation. Therefore, NACPG might promote NSCLC progression via immunosuppression, which was in accordance with previous studies.
3.8 Prognostic model
For a risk score, we further constructed a column line plot combining two independent prognostic factors, tumor stage, and NCAPG. We summed each score to obtain a final score, with a higher total score indicating a worse prognosis for the patient (Fig. 6a). We also plotted the calibration graphs, which showed that the predictions at years, 3 years, and 5 years were closer to the actual situation, indicating that the NCAPG prognostic model has good accuracy in predicting the prognosis of NSCLC patients (Fig. 6b).