hsa-let-7b-5p-associated BUB1/TMPO-AS1 ceRNA axis identified as a potential biomarker in lung adenocarcinoma

Abstract

Objective

BUB1, a key mitotic checkpoint kinase, is often dysregulated in cancer, yet its regulatory mechanism remains unclear. This study investigates the BUB1-centered miRNA–lncRNA–mRNA (ceRNA) network, its role in cell cycle regulation and immune modulation.

Methods

Prognostic significance and expression profiles were assessed using TCGA-based databases such as KM Plotter, UALCAN, OncoDB, ENCORI, GEPIA2, Lung Cancer Explorer, and TCGAnalyzeR. Transcription factors were identified via Enrichr, and a ceRNA network was constructed using miRNet. Binding affinity and folding energy between BUB1, miRNA, and lncRNA were predicted using miRWalk and RNA22v2. Molecular docking evaluated interactions with natural compounds, chemotherapeutics, and inhibitor. Immune subpopulations were visualized using the SPRING viewer and correlation analysis with the immune cells was conducted using the GSCA and TIMER2.0 databases.

Results

BUB1 overexpression correlated with poor LUAD prognosis, especially in smokers (HR = 1.76), with transcriptomic analysis showing a 2.46 log2-fold increase in BUB1 transcript levels in tumor. TF-E2F1 and lncRNA-TMPO-AS1 were positively correlated with BUB1 (R = 0.664 and R = 0.632, respectively), while miRNA hsa-let-7b-5p showed a negative correlation (R = − 0.366). TMPO-AS1 exhibited an inverse association with hsa-let-7b-5p, suggesting a molecular sponge formation, repressing its tumor-suppressive activity. Docking revealed strong binding affinity of hesperidin (− 9.4 kcal/mol) with BUB1. Additionally, BUB1 expression negatively correlated with CD4⁺ T cells, suggesting an immunosuppressive role.

Conclusion

This study identifies the BUB1/E2F1/TMPO-AS1/hsa-let-7b-5p axis as a potential prognostic biomarker and therapeutic target in LUAD. Targeting hsa-let-7b-5p may modulate this network, offering opportunities for both diagnostic and prognostic interventions.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s13008-025-00173-x.

Highlights

• BUB1 is highly overexpressed in LUAD and linked to poor survival, especially in smokers.

• A novel ceRNA network (E2F1, TMPO-AS1, hsa-let-7b-5p) regulates BUB1 in LUAD.

• BUB1 is enriched in the mitotic phase and associated with DNA repair, EMT, and metastasis.

• Positive correlation between BUB1, E2F1, TMPO-AS1, and cell cycle regulators, while hsa-let-7b-5p showed an inverse relationship.

• BUB1 overexpression reduces CD4⁺ T cell infiltration, indicating immune evasion.

• Hesperidin binds strongly to BUB1, showing promise as a low-toxicity therapeutic option.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13008-025-00173-x.

Introduction

Lung cancer is the most prevalent cancer worldwide, accounting for 12.4% of new cases and resulting in 18.7% of deaths annually, with global incidence and mortality projected to increase by over 85% and 95%, respectively, by 2050 [1, 2]. The high incidence and mortality rates are largely attributed to the disease’s heterogeneity and late-stage diagnosis, with persistent tobacco use being a significant contributing factor, particularly in Asia [3–5]. Notably, the prognosis varies markedly with disease stage; the 5 year survival rate exceeds 90% for stage IA patients but falls below 10% for those diagnosed at stage IV, highlighting the urgent need for early detection [6]. Histologically, lung cancer is classified into two major types: Non-small cell lung cancer (NSCLC) and Small cell lung cancer (SCLC) [7, 8]. NSCLC, the predominant subtype, arises from airway epithelial or mucus-secreting cells. Nearly 30% of patients are diagnosed at an early stage, with five-year recurrence rates ranging from 45% (Stage IB) to 76% (Stage III). The majority present with advanced disease, where five-year survival rates drop to ~ 35% for locally advanced and ~ 7% for metastatic cases [9]. NSCLC primarily comprises Lung adenocarcinoma (LUAD) and Squamous cell lung carcinoma (LUSC) as its predominant subtypes [10]. LUAD originates in the mucus-secreting cells deep within the lungs, while LUSC develops from the cells lining the airways [11]. Certain LUAD variants exhibit high invasiveness and metastatic potential [12, 13]. Currently available protein- and mutation-based biomarkers for LUAD, such as EGFR, KRAS, and ALK, offer limited prognostic value and are often ineffective in capturing the tumor’s molecular heterogeneity and dynamic regulatory networks [14]. This heterogeneous nature of lung cancer underscores the need for novel gene expression-based biomarkers that facilitate molecular classification, early detection, and targeted therapies [15, 16].

A crucial aspect of lung cancer progression is the dysregulation of the cell cycle, which is tightly controlled by checkpoint kinases [17]. Among these, the Budding Uninhibited by Benzimidazole (BUB) family members, particularly BUB1, play a vital role in maintaining chromosomal stability by ensuring accurate chromosome segregation and preventing genomic instability [18]. Proper attachment and alignment of microtubules are essential for maintaining mitotic fidelity, and errors in this process can drive tumorigenesis [19]. According to Foley and Kapoor, bipolar attachments to spindle microtubules are crucial for chromosome segregation as well as for forming and maintaining these connections [20].

Recent studies have begun to uncover the role of non-coding RNAs in modulating key cell cycle regulators, linking mitotic control with transcriptional and post-transcriptional regulation [21]. Given the association between BUB1 dysregulation and poor prognosis in several cancers, understanding the molecular mechanisms that regulate its expression is essential [22–24]. Emerging evidence indicates that competitive endogenous RNA (ceRNA) network molecules, including non-coding RNAs, especially microRNAs (miRNAs) and long non-coding RNAs (lncRNAs), are critical regulators of BUB1 expression, where miRNAs typically bind to the BUB1 mRNA to suppress translation, while lncRNAs act as molecular sponges or scaffolds that modulate miRNA availability and stabilize BUB1 transcripts [22, 25, 26]. However, the specific regulatory interactions remain unclear in the cases of lung cancer.

In this study, we investigated the prognostic significance of BUB1-network expression in LUAD patients and explored its regulation by key non-coding RNAs. Our findings revealed that BUB1 is overexpressed in LUAD and is regulated by the E2F1 transcription factor (TF), hsa-let-7b-5p, and TMPO-AS1. Moreover, it was observed that elevated expression levels of BUB1, E2F1, and TMPO-AS1, along with decreased levels of hsa-let-7b-5p, correlate with poor prognosis in LUAD patients. These findings indicate that BUB1 and its regulatory ceRNA network could serve as potential therapeutic targets for LUAD management. We hypothesize that the dysregulation of the TMPO-AS1/hsa-let-7b-5p/E2F1/BUB1 axis promotes LUAD progression by enhancing BUB1 expression, contributing to chromosomal instability (CIN) and immune evasion, particularly in smoker patients. Moreover, this regulatory network may serve as a clinically relevant biomarker and therapeutic target. We also evaluated hesperidin, a natural flavonoid, for its potential inhibitory interaction with BUB1, given the growing interest in plant-derived compounds as targeted anticancer agents. While the present study is entirely computational, it lays a foundation for future experimental validation to confirm these regulatory interactions and therapeutic possibilities.

Results

High BUB1 expression correlates with poor survivability in LUAD patients

BUB1, a key mitotic checkpoint kinase, has gained attention for its oncogenic potential and has been implicated in cancer progression through its dysregulated expression. Previous studies in other cancers have demonstrated that BUB1 overexpression is associated with tumor aggressiveness and poor prognosis. For example, BUB1 has been reported as a prognostic biomarker in liver hepatocellular carcinoma, and its aberrant overexpression promotes mitotic segregation errors and CIN in multiple myeloma [27, 28]. Building on these observations, we initially used the Kaplan–Meier Plotter (KMP) database to examine BUB1 expression in relation to the survivability of lung cancer patients with different clinical parameters. This included univariate and multivariate analyses with low and high expression cohorts as mentioned in Table 1. It was found that BUB1’s overexpression was significantly correlated with poor OS (HR = 1.7, CI 1.38–2.1, P = 3.5e−07, low expression cohort = 93, high expression cohort = 52), and FP (HR = 1.79, CI 1.39–2.3, P = 4.2e−06, low expression cohort = 30, high expression cohort = 13), whereas was insignificant in PPS (HR = 1.3, CI 0.97–1.72, P = 0.074, low expression cohort = 18.2, high expression cohort = 10.81), as illustrated in Fig. 1A-C, respectively. Our study next found that overexpression of BUB1 significantly correlates with poor OS in LUAD patients (HR = 1.86, CI 1.56–2.22, P = 2.6e−12, low expression cohort = 107, high expression cohort = 52), while no significant association was found in LUSC patients, as shown in Fig. 1D–E. Further, multivariate analysis indicated that elevated BUB1 levels were associated with poor OS, especially in early-stage LUAD patients, stage 1, with a hazard ratio score (HR) of 2.16 (CI 1.42–3.28, P = 0.00023, low expression cohort = 76, high expression cohort = 42.07) (Fig. 1F). However, BUB1 levels did not show a significant association with stage 2 or stage 3 LUAD patients, as shown in Supplementary Fig. 1A and B. Further, the HR for OS in LUAD male and female patients were found to be 1.6 and 2.02, respectively, both with significant p-values as shown in Fig. 1G–H.

Table 1.

Survival Analysis of BUB1

S. no	Gene	Index		Patient Number	Hazard Ratio	CI	Log(P)	Low Expression Cohort (Months)	High Expression Cohort (Months)
1	BUB1	OS		2166	1.7	1.38–2.1	3.5e−07	93	52
		FP		1252	1.79	1.39–2.3	4.2e−06	30	13
		PPS		477	1.3	0.97–1.72	0.074	18.2	10.81
Histology
2	BUB1	Adenocarcinoma		1161	1.86	1.56–2.22	2.6e−12	107	52
		Squamous cell Carcinoma		780	0.97	0.8–1.17	0.73	56	54
OS + LUAD + Stages
3	BUB1	Stage1		370	2.16	1.42–3.28	0.00023	76	42.07
		Stage2		136	1.12	0.7–1.82	0.63	63	61.37
		Stage3		24	0.78	0.28–2.17	0.63	33.83	40.77
OS + LUAD + Gender
4	BUB1	Male		566	1.6	1.26–2.03	8.7e−05	96	46
		Female		537	2.02	1.52–2.69	7.2e−07	110.27	69.93
OS + LUAD + Smoking History
5	BUB1	Smoker		546	1.76	1.35–2.29	2.5e−05	95	59
		Non-Smoker		192	1.61	0.88–2.92	0.12	76	56.5
OS + LUAD + Smoking History + Gender
6	BUB1	Smoker	Male	319	1.6	1.13–2.28	0.0081	80	74
			Female	227	1.87	1.25–2.79	0.0021	96	52
		Non-smokers	Male	31	2.86	0.7–11.66	0.13	NA	NA
			Female	161	1.1	0.57–2.14	0.78	72	75.43

Fig. 1.

Prognostic significance of mRNA BUB1 expression in lung cancer patients. Kaplan–Meier survival curves were plotted for A Overall Survival (n = 833), B First Progression Survival (n = 584), C Post Progression Survival (n = 299), D OS + LUAD (n = 1161), E OS + LUSC (n = 780), F OS + LUAD + Stage1 (n = 370), G OS + LUAD + Male (n = 566), H OS + LUAD + Female (n = 537), I OS + LUAD + Smoker (n = 546), J OS + LUAD + Male + Smoker (n = 319), K OS + LUAD + Female + Smoker (n = 227)

These results suggest that BUB1 may serve as a predictive prognostic biomarker for LUAD patients regardless of gender and holds potential as an early diagnostic indicator. Notably, a significant link was found between high BUB1 expression and decreased survival in smokers (HR = 1.76, CI 1.35–2.29, P = 2.5e−05), as shown in Fig. 1I. While no significant association was observed in non-smokers (HR = 1.61, CI 0.88–2.92, P = 0.12), as shown in Supplementary Fig. 1C. Further, smoking-related survival analysis by gender revealed poorer prognosis in both male (HR = 1.6, CI 1.13–2.28, P = 0.0081) and female (HR = 1.87, CI = 1.25–2.79, P = 0.0021) smokers, as shown in Fig. 1J–K. In contrast, no significant results were observed in the cases of non-smoker male and female patients as depicted in Supplementary Fig. 1D–E. Hence, BUB1 may be crucial for LUAD prognosis, with overexpression linked to unfavorable outcomes, especially in smokers, but lacks prognostic significance in LUSC patients, suggesting its role as a molecular classifier.

Overexpression of BUB1 gene and protein in lung adenocarcinoma compared to normal tissue

To analyze the differential gene expression, UALCAN and OncoDB databases were employed, the results of which showed a significant upregulation of BUB1 in LUAD tumor samples compared to normal tissues. Specifically, BUB1 overexpression was statistically significant with P-values of < 1e−12 (UALCAN) and 8.1e−74 (OncoDB), as illustrated in Supplementary Fig. 2A–B. Additional databases, including ENCORI, Lung Cancer Explorer, and GEPIA2, supported these findings, as shown in Supplementary Fig. 2C–E. Further, BUB1’s expression data was extracted from the TCGA-LUAD repository using R-based packages, with sample sizes normalized across both groups (Normal = 59, Tumor = 59), and the resulting box plot (Fig. 2A) demonstrated consistent significant overexpression of BUB1 in LUAD tumor tissues compared to normal tissues (P = 2.63e−16). Additionally, data from the TCGAnalyzeR database revealed a 2.46 log₂-fold increase in BUB1’s transcript level in LUAD cells compared to normal cells, as indicated in Fig. 2B.

Fig. 2.

Differential expression of mRNA in lung cancer. A BUB1 expression was examined in normal lung tissue and primary tumors by using R-based package (normal n = 59, tumor n = 59), and Transcriptome Analysis in LUAD using B TCGAnalyzeR. Analysis of BUB1 expression in LUAD as compared to normal tissues by using UALCAN based on C Cancer Stages, D Patient’s gender, E Smoking habits, and F Nodal metastasis status. BUB1 expression in normal, tumor and metastatic state by using G TNMplot database. BUB1 expression in TP53 mutant and non-mutant cases by using H UALCAN, and I TIMER2.0 databases. Protein Expression of BUB1 was analyzed in LUAD J Tumor vs Normal K Correlation between BUB1 gene and protein using CancerProteome.

The UALCAN database was then utilized for expression analysis of different clinicopathological statuses. Results showed a significant upregulation of the BUB1 gene across advanced cancer stages, in both genders, but with higher expression in males, as shown in Fig. 2C–D. The analysis also revealed a marked upregulation of BUB1 in smoker patients relative to non-smokers, with expression levels further rising in cases with advanced nodal metastasis (Fig. 2E–F). These findings were further corroborated by the TNMplot database, which revealed significantly upregulated BUB1 expression in LUAD tumor and metastatic tissues compared to normal (P = 3.82e-106), underscoring its oncogenic role and potential involvement in metastasis (Fig. 2G). Additionally, as well known, tobacco-derived carcinogens, Nicotine and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), are among the principal stimulants associated with smoking-induced lung carcinogenesis [29, 30], and therefore to further understand their possible molecular influence on mitotic regulators, docking analysis with BUB1 revealed binding affinities of − 4.7 kcal/mol for Nicotine and − 5.5 kcal/mol for NNK, suggesting potential interactions that may contribute to the dysregulation of cell-cycle signaling in smokers. These findings reinforce the mechanistic association between tobacco exposure and aberrant BUB1 activity in the pathogenesis of LUAD, as shown in Supplementary Fig. 2F–G and Supplementary Table 2. Furthermore, previous studies indicate that the TP53 gene, known for its frequent mutations in human cancers, leads to a stable mutant protein that characterizes cancer cells [31]. These mutant p53 proteins lose their tumor-suppressing abilities while acquiring new oncogenic functions, highlighting their potential as targets for innovative cancer treatments. Our research suggests that increased BUB1 expression was significantly linked to TP53 mutations in LUAD patients, as evidenced by data from the UALCAN (P =  < 1e−12) and TIMER2.0 (P =  < 2.2e−16) databases (Fig. 2H–I). Further, strengthening our data, we found a significant upregulation of BUB1 protein in tumor samples compared to normal patients, along with a strong positive correlation (R = 0.713) between RNA and protein levels of BUB1, by using the CancerProteome database, as shown in Fig. 2J–K, suggesting effective transcriptional regulation and translation. Hence, this highlights BUB1’s potential as a biomarker in LUAD.

E2F1 identified as a key transcriptional activator of BUB1 in LUAD

As it is well known that dysregulation of transcription factor activity leads to tumorigenesis. Building upon this evidence, the TFs associated with BUB1 were identified using the Enrichr database, which revealed a strong association with several TF targets, suggesting their potential regulatory role in BUB1 expression (Supplementary Table 3). Using the ENCORI database, the correlation values between BUB1 and TFs, i.e., E2F1, E2F3, E2F4, TP53, RBL2 (mouse), YBX1, TRP53 (mouse), MYC, TFDP1, and E2F4 (mouse), were obtained as shown in Supplementary Table 4. Positive correlations were observed between BUB1 and E2F1 (R = 0.664), E2F4 (R = 0.219), E2F3 (R = 0.548), YBX1 (R = 0.413), MYC (R = 0.201), and TFDP1 (R = 0.442), where E2F1 showed the strongest correlation with BUB1. Additionally, previous studies have shown that the TF-E2F1 promotes NSCLC progression by activating oncogenic pathways such as PI3K/AKT [32]. Therefore, to illustrate its putative regulatory role in LUAD, a schematic model was generated, depicting how E2F1 overexpression drives BUB1 transcription and subsequent upregulation of BUB1 mRNA in tumor tissues, as shown in Fig. 3A.

Fig. 3.

Transcription Factor Analysis. A Proposed mechanism of E2F1-mediated regulation of BUB1 in LUAD. Correlation analysis between BUB1 and E2F1 by using B ENCORI. Survival analysis using KM Plotter C OS, D OS + LUAD, E OS + LUAD + Smoker, F OS + LUAD + Smoker + Male, G OS + LUAD + Smoker + Female, and Differential gene expression analysis of E2F1 in normal vs. tumor tissues by using H R-based package (normal n = 59, tumor n = 59). Analysis of E2F1 expression in LUAD as compared to normal tissues by using UALCAN based on I Cancer Stages, J Patient’s gender, K Smoking habits, and L Nodal metastasis status. E2F1 expression in normal, tumor and metastatic state by using M TNMplot database

The above-mentioned results suggested a strong, significant correlation between BUB1 and E2F1, as depicted in Fig. 3B. The same results were corroborated using the OncoDB and GEPIA2 databases, as shown in Supplementary Fig. 3A–B. Furthermore, to elucidate potential transcriptional regulation of BUB1 by E2F1, we analyzed the 2 kb upstream and 100 bp downstream region of the BUB1 transcription start site (TSS; hg38) using the JASPAR 2022 E2F1 motif (MA0024.3). Analysis using R based package with a position-weight matrix threshold of ≥ 80% revealed multiple high-confidence E2F1 binding sites distributed across both the forward and reverse DNA strands within the BUB1 promoter (Supplementary Fig. 3C). The predicted motifs exhibited relative binding scores ranging from 0.80 to 0.92, with the strongest clusters located approximately 0.6 – 1.9 kb upstream of the TSS. This strong interaction supports the hypothesis that E2F1 directly binds and potentially regulates BUB1 activity, contributing to mitotic progression and oncogenic signalling in LUAD. The study evaluated E2F1's prognostic significance using the KMP database and the results showed E2F1 upregulation linked to poor OS outcomes in LUAD patients, suggesting it could be a potential biomarker alongside BUB1 for smokers. Figure 3C-G shows E2F1 overexpression associated with poor survivability in cases of OS (HR = 1.43, CI 1.27–1.61, P = 3.6e−09), OS + LUAD (HR = 1.63, CI 1.37–1.94, P = 2.2e−08), OS + LUAD + Smoker (HR = 1.49, CI 1.14–1.93, P = 0.0028), OS + LUAD + Smoker + Male (HR = 1.43, CI 1.01–2.02, P = 0.045), and OS + LUAD + Smoker + Female (HR = 1.56, CI 1.05–2.32, P = 0.028). Further, the differential E2F1 gene expression analysis using UALCAN and ENCORI databases showed significant overexpression in LUAD samples as compared to normal, as shown in Supplementary Fig. 3D–E. Additionally, R packages were employed to analyze E2F1 expression in normalized samples (Normal = 59, Tumor = 59). The analysis revealed a significant overexpression of E2F1 in LUAD tumor tissues compared to normal tissues (P = 5.05e−08), as depicted in Fig. 3H. Figure 3I–L presents a gene expression analysis of E2F1 in relation to clinical parameters such as cancer stages, smoking history, patient gender, and nodal metastasis using the UALCAN database. The analysis revealed that E2F1 expression significantly increases in advanced stages of LUAD, peaking in stage 2. Male patients exhibit higher E2F1 expression than females, and smokers show elevated levels compared to non-smokers. Additionally, the highest expression was observed in the advanced nodal metastatic state (N3). Further, the TNMplot database was utilized to evaluate the Differential Gene Expression (DGE) of E2F1 in normal, tumor, and metastatic states. The violin plot showed its significant elevated expression in LUAD tumor and metastatic tissues compared to normal tissues (P = 1.47e−25), highlighting its involvement in metastatic progression (Fig. 3M). Hence, it can be suggested that E2F1 can be considered as a key regulator of BUB1 in LUAD conditions, and dysregulation in E2F1 expression levels could promote tumor progression.

Regulatory ceRNA network analysis reveals the hsa-let-7b-5p/TMPO-AS1 axis modulating BUB1 expression

Emerging evidence suggests that the ceRNA network plays a crucial role in post-transcriptional regulation of oncogene activation and cancer progression by modulating miRNA availability [33]. Building upon this, our study aimed to understand the regulatory mechanisms underlying BUB1 expression in LUAD by constructing a comprehensive ceRNA network comprising miRNAs, mRNAs, and lncRNAs. Using the miRNet database, we identified 17 miRNAs in association with both BUB1 and E2F1, underscoring the interaction between them (Supplementary Fig. 3F). In addition, an analysis of differential miRNA expression using the UALCAN database revealed that 5 out of 17 miRNAs, including hsa-mir-126-3p, hsa-let-7b-5p, hsa-let-7g-5p, hsa-miR-98-5p, and hsa-mir-29c-3p, were significantly downregulated in LUAD as shown in Supplementary Table 5. Further, Supplementary Table 6 reveals that miRNAs hsa-let-7b-5p and hsa-miR-29c-3p exhibit the strongest negative correlations with BUB1 and E2F1, prompting their selection for further analysis. To visually represent this mechanism, a schematic illustration was generated, highlighting that the downregulation of miRNA reduces its ability to bind BUB1 transcripts, thereby promoting BUB1 overexpression in LUAD, as shown in Fig. 4A. Subsequent visualization using the ENCORI database revealed significant negative correlations between hsa-let-7b-5p and BUB1 (R = –0.366), as well as E2F1 (R = − 0.365), and between hsa-mir-29c-3p and BUB1 (R =− 0.426), along with E2F1 (R = – 0.347), as shown in Fig. 4B–E. The prognostic significance of these miRNAs was further assessed by KM Plotter survival analysis, where low expressions of hsa-let-7b-5p (HR = 0.71, CI 0.53–0.95, P = 0.021) and hsa-mir-29c-3p (HR = 0.54, CI 0.37–0.79, P = 0.0012) were significantly associated with poor OS in LUAD patients (Fig. 4F–G). The prognostic significance of biomarkers like miRNA was assessed using Area Under the Curve (AUC) and Receiver Operating Characteristic (ROC) curves. The ROC curve measures the miRNA's ability to categorize patients into prognostic groups, with a steeper curve indicating greater efficacy. AUC summarizes the ROC curve, providing a general indication of a biomarker's capacity. The CancerMIRNome database demonstrated high prognostic significance for miRNAs, with hsa-let-7b-5p showing high accuracy (AUC = 0.89, CI 0.85–0.92); however, hsa-mir-29c-3p showed moderate accuracy (AUC = 0.67, CI 0.58–0.77), as shown in Fig. 4H-I, respectively. Based on the AUC values, hsa-let-7b-5p was considered for further analysis. Supplementary Fig. 3G–H also revealed significant downregulation of hsa-let-7b-5p in LUAD samples compared to normal tissues by using the CancerMIRNome and UALCAN databases, respectively. Further expression analysis of hsa-let-7b-5p in normalized samples (Normal = 46 and Tumor = 46) was performed using R-based tools, confirming its significant downregulation in LUAD tumor tissues relative to normal tissues (P = 0.025), as depicted in Fig. 4J. Furthermore, the expression of hsa-let-7b-5p in LUAD was analyzed using the UALCAN database, revealing consistent downregulation across various clinical parameters, indicative of poor prognosis, including smoking history and advanced stages (Fig. 4K–L), along with males and nodal metastasis (Supplementary Fig. 3I–J). The findings indicate that lower levels of hsa-let-7b-5p expression are linked to more aggressive progression of LUAD, and that restoring its expression can restore its tumor suppressor function. To validate these findings, analysis of the binding and folding energy of hsa-let-7b-5p with BUB1 and E2F1 was performed using miRWalk [34] and RNA22v2 [35]. The results showed that hsa-let-7b-5p binds to BUB1 (binding energy = − 21.8 kcal/mol) and E2F1 (binding energy = − 22.4 kcal/mol), with corresponding folding energies of − 13.89 kcal/mol and − 13.90 kcal/mol, respectively, as shown in Table 2. These values indicate a strong interaction between hsa-let-7b-5p and its target genes, reinforcing its role in post-transcriptional regulation. Similar to our findings, previous studies have highlighted the tumor-suppressive role of hsa-let-7b-5p, showing that it inhibits cancer cell growth and metastasis by repressing key metabolic pathways, such as hexokinase 2-mediated aerobic glycolysis in breast cancer [36]. This establishes its significance in post-transcriptional regulation and tumor progression.

Fig. 4.

ceRNA network analysis-miRNA. A Visual representation of hsa-let-7b-5p mediated BUB1 gene expression regulation mechanism. Correlation graphs between B hsa-let-7b-5p and BUB1, C hsa-let-7b-5p and E2F1, D hsa-miR-29c-3p and BUB1, E hsa-miR-29c-3p and E2F1 using ENCORI, and survival analysis using KM Plotter: Overall Survival F hsa-let-7b-5p and G hsa-miR-29c-3p, AUC curve analysis for finding prognostic significance of miRNA H hsa-let-7b-5p and I hsa-miR-29c-3p using CancerMIRNome, and Differential gene expression analysis of hsa-let-7b-5p in normal vs. tumor tissues by using J R-based package (normal n = 46, tumor n = 46), and Expression Analysis of miRNA hsa-let-7b-5p across various clinical parameters using UALCAN K Smoking history and L Cancer stages

Table 2.

Binding and Folding energy of BUB1 with hsa-let-7b-5p

miRNA	Transcript	Binding Energy (miRWalk)	Folding energy (RNA22v2) (in -Kcal/mol)	Heteroduplex
hsa-let-7b-5p MIMAT0000063	BUB1 NM_001278617.2	− 21.8	− 13.89
	E2F1 NM_005225.3	− 22.4	− 13.90
	TMPO-AS1 NR_027157.1	–	−12.40

The study next analyzed the molecular sponge of miRNAs, i.e., lncRNAs, which are typically over 200 nucleotides in length and play key roles in regulating gene expression and tumor progression. Our study identified lncRNAs within the BUB1 ceRNA network in LUAD using the Enrichr database and found top 10 lncRNAs associated with BUB1 (Supplementary Table 7). Five out of the ten lncRNAs were significantly upregulated after being filtered out based on differential lncRNA expressions in LUAD samples using the UALCAN database (Supplementary Table 8). Further, Supplementary Table 9 lists the correlations between upregulated lncRNAs and hsa-let-7b-5p, E2F1, and BUB1, where TMPO-AS1 showed the most significant result with the molecules.

To provide a mechanistic overview, we generated a schematic illustration showing that upregulated TMPO-AS1 acts as a molecular sponge for hsa-let-7b-5p, decreasing its availability to bind BUB1 transcripts, thereby leading to BUB1 overexpression, as depicted in Fig. 5A. This ceRNA-mediated interaction supports the oncogenic potential of the TMPO-AS1/hsa-let-7b-5p/BUB1 axis. As shown in Fig. 5B–D, TMPO-AS1 showed a strong positive correlation with BUB1 (R = 0.632) and E2F1 (R = 0.581) and a negative correlation with hsa-let-7b-5p (R = -0.277).

Fig. 5.

ceRNA network analysis-lncRNA. A Schematic representation of mechanism of TMPO-AS1–hsa-let-7b-5p–BUB1 axis in LUAD. Correlation graphs between B TMPO-AS1 and BUB1, C TMPO-AS1 and E2F1, and D hsa-let-7b-5p and TMPO-AS1 using ENCORI. Survival analysis of TMPO-AS1 using KM Plotter: E OS, F FP, G PPS, H OS + LUAD, I OS + LUAD + Smokers, J OS + LUAD + Male + Smokers, and Differential gene expression analysis of TMPO-AS1 in normal vs. tumor tissues by using K UALCAN, and L R-based package (normal n = 59, tumor n = 59), and Expression Analysis of lncRNA TMPO-AS1 across various clinical parameters using UALCAN M Patient’s gender, N Smoking history and O Cancer stages

Further, the prognostic significance of TMPO-AS1 in lung cancer was evaluated using the KM Plotter database, incorporating both univariate and multivariate analyses based on low and high expression cohorts (Supplementary Table 10). Univariate analysis revealed that high TMPO-AS1 expression was significantly associated with poor survival outcomes, including OS (HR = 1.5, CI 1.29–1.74, P = 7.2e−08), FP (HR = 1.9, CI 1.52–2.38, P = 8.1e−09), PPS (HR = 1.77, CI = 1.32–2.39, P = 0.00013), and OS + LUAD (HR = 2.16, CI 1.69–2.76, P = 4.1e−10), as shown in Fig. 5E-H. Additionally, OS + LUAD + Stage 1 (HR = 2.65, CI 1.72–4.09, P = 4.4e-06), OS + LUAD + Stage 2 (HR = 2.1, CI 1.22–3.61, P = 0.0061), OS + LUAD + Male (HR = 1.95, CI 1.39–2.73, P = 9.1e−05), and OS + LUAD + Female (HR = 2.83, CI 1.84–4.35, P = 7.2e−07) are shown in Supplementary Fig. 4A–D. Further, Fig. 5I-J shows OS + LUAD + smoker (HR = 2.34, CI 1.41–3.88, P = 0.00066) and OS + LUAD + male smoker (HR = 2.33, CI 1.28–4.25, P = 0.0046), along with OS + LUAD + female smoker (HR = 3.13, CI 1.18–8.29, P = 0.016)(Supplementary Fig. 4E). Subsequently, survival plots revealed that overexpression of TMPO-AS1 was linked to poor survival in LUAD smokers.

The DGE analysis revealed significant upregulation of TMPO-AS1 in LUAD tissues compared to normal tissues, using UALCAN database and R-based packages were further employed for expression analysis in normalized samples, showing the significant overexpression of TMPO-AS1 in LUAD tumor tissues relative to normal tissues (P = 1.05e-13), as shown in Fig. 5K–L, respectively. To further validate these findings, OncoDB, ENCORI, and Lung Cancer Explorer databases were employed, as shown in Supplementary Fig. 4F–H. Moreover, analysis using the UALCAN database indicated that TMPO-AS1 expression was significantly higher in male patients compared to females, in smokers, and advanced cancer stages (Fig. 5M-O). Additionally, the miRNet database supports the interaction between the tumor suppressor hsa-let-7b-5p and the lncRNA TMPO-AS1, which regulates the BUB1 and E2F1 genes (Supplementary Fig. 4I). Further, to assess the interaction potential between hsa-let-7b-5p and TMPO-AS1, a folding energy analysis was performed using RNA22v2, revealing a folding energy of − 12.40 kcal/mol, indicating a stable heteroduplex structure, as shown in Table 2. The negative correlation between TMPO-AS1 and hsa-let-7b-5p indicates TMPO-AS1 acts as a molecular sponge, decreasing its accessibility to target BUB1 and causing its upregulation in LUAD smokers. Further supporting this, the targetgram analysis from the TNMplot database illustrates differential expression patterns of BUB1, E2F1, and TMPO-AS1 between normal and tumor lung tissues (Supplementary Fig. 4 J). The three concentric circles denote increasing expression levels, with outer rings representing higher gene expression. In tumor tissues, both BUB1 and E2F1 are markedly upregulated, while TMPO-AS1 shows a moderate increase, reinforcing its potential oncogenic role. These findings underscore a dysregulated ceRNA axis in LUAD, where elevated TMPO-AS1 expression may facilitate oncogene activation by sequestering tumor-suppressive miRNAs like hsa-let-7b-5p.

Molecular docking and visualization analysis reveals strong binding affinity of BUB1 with selective inhibitor, natural compounds and chemotherapeutic agents

To investigate the therapeutic potential of targeting BUB1, molecular docking and visualization analyses were performed to assess the binding interactions between BUB1 and known potential inhibitory compounds. BAY-1816032, a well-established, direct, and selective inhibitor of BUB1, specifically designed to block its kinase activity with high target precision [37] was docked, alongside natural compounds, Hesperidin and Quercetin selected based on their reported anti-cancer, anti-inflammatory, and kinase-inhibitory properties, combined with their favorable safety and bioavailability profiles [38, 39]. These were further compared with standard chemotherapeutics Paclitaxel and Docetaxel, which, although widely used, often lack kinase selectivity and are associated with systemic toxicity [40]. Molecular docking revealed binding affinities of − 8.0 kcal/mol for BAY-1816032, − 9.4 kcal/mol for Hesperidin, − 7.4 kcal/mol for Quercetin, − 8.2 kcal/mol for Paclitaxel, and − 7.6 kcal/mol for Docetaxel, as shown in Table 3. Notably, Hesperidin exhibited the strongest affinity, surpassing both standard chemotherapeutics, suggesting its potential as a highly effective BUB1-targeting agent. BAY-1816032, the reference inhibitor, also showed a strong and specific interaction pattern consistent with its known inhibitory mechanism, thereby validating the docking approach and reinforcing the reliability of the observed binding trends. Quercetin demonstrated comparable binding strength to Docetaxel, indicating its viability as a therapeutic alternative. To further analyze these interactions, we visualized the molecular docking results using Discovery Studio Visualizer (DSV), generating 2D interaction diagrams, revealing key molecular interactions, including hydrogen bonding, van der Waals forces, and hydrophobic interactions, which help explain the binding stability of these compounds. The docking results in Fig. 6A–E revealed that Hesperidin formed multiple conventional hydrogen bonds and van der Waals interactions with crucial residues within the BUB1 active site, resulting in a diverse and stable interaction network. In contrast, BAY-1816032, primarily forms π-alkyl interactions and hydrogen bonds, Hesperidin establishes a more distinct interaction network, suggesting stable and favorable binding as compared to chemotherapeutic agents such as Paclitaxel and Docetaxel. The presence of multiple hydrogen bond donors and acceptors in Hesperidin’s structure enhances its affinity, potentially conferring a more selective inhibitory effect on BUB1. This indicates that Hesperidin may serve as a promising natural inhibitor of BUB1. Overall, these findings indicate that natural compounds, particularly Hesperidin and, to an extent, Quercetin demonstrate strong and specific binding to BUB1, underscoring their potential as promising, low-toxicity alternatives to conventional chemotherapeutic agents.

Table 3.

Binding Affinity of BUB1 with its Selective inhibitor, Natural compounds, and Chemotherapeutic drugs

Molecule	Binding Affinity
	BAY-1816032	Hesperidin	Quercetin	Paclitaxel	Docetaxel
BUB1	− 8.0 kcal/mol	− 9.4 kcal/mol	− 7.4 kcal/mol	− 8.2 kcal/mol	− 7.6 kcal/mol

Fig. 6:

2D interaction diagrams of molecular docking between BUB1 protein and the ligands. A BUB1-BAY-1816032, B BUB1-Hesperidin, C BUB1-Quercetin, D BUB1-Paclitaxel, and E BUB1-Docetaxel

BUB1 is functionally enriched in cell cycle regulation

As is well known, G2/M checkpoint control is crucial for preventing genomic instability, and its disruption drives tumor progression. Centrosome-associated regulators, such as spindle checkpoint kinases, play a key role in this process and are emerging as promising therapeutic targets [41]. Consistent with this role, we investigated BUB1 to assess its G2/M-specific regulation and potential oncogenic role. The Cyclebase 3.0 database revealed high BUB1 expression during the M phase of the cell cycle, highlighting its crucial role in cell division and tumor progression, with BUB1 ranking 15th in expression levels (Fig. 7A). Further analysis of BUB1 gene expression across various functional states was performed using the CancerSEA database (Fig. 7B). The results showed significant association between BUB1 expression and several key biological processes, including DNA damage (R = 0.55), cell cycle regulation (R = 0.53), proliferation (R = 0.47), DNA repair (R = 0.42), invasion (R = 0.40), and epithelial-mesenchymal transition (EMT) (R = 0.37) (Fig. 7C-H). These findings suggest that BUB1 plays a pivotal role in regulating multiple pathways critical for tumor progression, particularly those related to proliferation, genomic instability, and metastasis. Additionally, gene enrichment analysis using the GENI database revealed insights into the biological processes, cellular components, and molecular functions of BUB1, E2F1, and TMPO-AS1 (Supplementary Fig. 5–7, respectively).

Fig. 7.

Cell Cycle and Biological Process Analysis. A Cell cycle analysis using the Cyclebase 3.0 database, B Association of BUB1 expression with various biological processes using the CancerSEA database, C–H Correlation of BUB1 expression with DNA Damage, Cell Cycle, Proliferation, DNA repair, Invasion, and EMT using CancerSEA

Next, the study analyzed the correlation of BUB1 with cyclins and CDKs, revealing a strong correlation between these genes and various cell cycle processes. Cyclin-dependent kinases (CDKs) and cyclins are complexes that work together to regulate the cell cycle and ensure cell functionality. CDK and cyclin molecules consist of four members each, with Cyclin D involved during the G1/early S-phase, Cyclin E involved during the late G1/S-phase, Cyclin A involved during the late S/G2-phase, and Cyclin B involved during the late G2/M phase. These interactions are crucial for maintaining cell functionality. Further, the correlation between CDK and cyclin molecules with the BUB1 gene, E2F1, hsa-let-7b-5p, and TMPO-AS1 were analyzed using the ENCORI database (Supplementary Table 11). The results showed strong correlations with CDK2: BUB1 (R = 0.72), E2F1 (R = 0.649), hsa-let-7b-5p (R = − 0.285), and TMPO-AS1 (R = 0.646); CDK1: BUB1 (R = 0.88), E2F1 (R = 0.631), hsa-let-7b-5p (R = − 0.319), and TMPO-AS1 (R = 0.626); CCNE1: BUB1 (R = 0.72), E2F1 (R = 0.681), hsa-let-7b-5p (R = − 0.315), and TMPO-AS1 (R = 0.522); and CCNB1: BUB1 (R = 0.84), E2F1 (R = 0.649), hsa-let-7b-5p (R = − 0.285), and TMPO-AS1 (R = 0.646). The data suggests that BUB1, E2F1, and TMPO-AS1 are positively correlated, whereas hsa-let-7b-5p is negatively correlated with CDK2/cyclin E and CDK1/cyclin B complexes that are S and M phase checkpoints and are crucial for cell cycle progression.

BUB1 shows progressive upregulation across LUAD stages

Disease stage remains one of the most decisive prognostic factors in LUAD, with 5-year survival rates exceeding 90% in stage IA patients but dropping to below 10% in those diagnosed at stage IV [6]. This stark decline underscores the critical need for biomarkers that can aid in both early detection and progression monitoring. To investigate the stage-specific behavior of potential biomarkers, we used the GSCA database to examine the expression trends of key biomarkers across pathological stages in LUAD. The trend plot showed the increasing expression levels of four genes: E2F1, MKI67, BUB1, and TMPO-AS1. E2F1 expression increased from Stage I to Stage III, MKI67 expression increased from Stage I to Stage II, with a dip in Stage III, whereas, BUB1 expression showed a steady increase across all pathological stages without fluctuations, and TMPO-AS1 expression showed a progressive increase, suggesting a potential role in tumor progression (Supplementary Fig. 8A). GSVA scores, compared in tumor and normal samples in LUAD conditions, showed higher scores in tumor samples, indicating more activation of all four genes in tumor tissues than normal lung tissues. This suggests BUB1 could be a valuable biomarker for prognosis in LUAD patients (Supplementary Fig. 8B).

The GSVA score reveals an upward trend in gene set activity as the disease progresses, indicating that the selected gene set becomes increasingly enriched or active in advanced stages of LUAD (Supplementary Fig. 8C–D). The trend plot for all four genes, including the tumor suppressor TP53, shows no significant change during the early stages (Supplementary Fig. 8E). However, a noticeable peak in expression was observed during intermediate Stage III, followed by a subsequent downregulation in Stage IV. This suggests evasion of TP53-mediated tumor suppression mechanisms, contributing to disease progression in advanced stages. These findings imply that these genes may play a critical role in LUAD development and progression.

Immune cell infiltration analysis reveals BUB1 axis–mediated CD4⁺ T cell suppression in LUAD

In LUAD, immune evasion is often driven by checkpoint pathways such as PD-1/PD-L1 and CTLA-4, which suppress T cell activation. Approved antibodies like pembrolizumab, nivolumab, atezolizumab, durvalumab (anti-PD-1/PD-L1), and ipilimumab (anti-CTLA-4) improve survival but are costly and have immune-related toxicities [42]. Tumor-infiltrating immune cells are critical determinants of the tumor microenvironment (TME), influencing tumor growth, immune escape, and therapeutic response. To understand the immune-related role of the BUB1 regulatory axis in LUAD, we visualized its expression across immune subpopulations using the trajectory-based SPRING viewer database. The immune clustering of NSCLC cells revealed distinct compartments wherein BUB1 was broadly expressed across several immune subsets (Supplementary Fig. 9A–C), along with moderate expression of E2F1 and TMPO-AS1 (Supplementary Fig. 9D-E). Subsequently, to statistically assess the immune association, we evaluated the correlation between the expression of BUB1, its TF-E2F1, and associated lncRNA TMPO-AS1 with immune cell infiltration in LUAD using the GSCA database. Among all immune types, CD4⁺ T cells demonstrated the strongest negative correlation with all three components (Supplementary Fig. 9F-G), indicating a potential suppressive influence of this axis on adaptive immunity. The combined GSVA score derived from the BUB1–E2F1–TMPO-AS1 signature also showed a robust inverse correlation with CD4⁺ T cell infiltration (R = − 0.56, FDR = 4.2e−42), highlighting their immunomodulatory potential (Supplementary Fig. 10A).

To extend these findings, we used TIMER2.0 to assess correlations with CD4⁺ T cell subtypes. Consistently, BUB1 and E2F1 were negatively correlated with CD4⁺ T cell subsets across multiple estimation algorithms (Supplementary Fig. 10B–C), supporting a mechanistic link between the overexpression of this axis and CD4⁺ T cell suppression in LUAD. Collectively, these results suggest that the overexpression of the BUB1–E2F1–TMPO-AS1 axis may contribute to an immunosuppressive tumor microenvironment by actively hindering CD4⁺ T cell recruitment and infiltration.

Discussion

LUAD, the most common and aggressive subtype of NSCLC, is marked by high CIN and immune evasion, arising from dysregulation of cell cycle and transcriptional control pathways [43–45]. There are many traditional treatments for cancer, including surgery, chemotherapy, and radiation therapy, but their limitations have led to the advances in targeted treatments like EGFR and ALK inhibitors, along with immune checkpoint inhibitors, which further lead to better outcomes for patients [46–48]. Besides this, challenges like therapy resistance, late-stage diagnosis, and tumor heterogeneity persist [49]. This underscores the critical need to identify robust prognostic biomarkers and novel therapeutic targets that can guide precision oncology efforts based on gene expression profiling. Cell cycle dysregulation, a hallmark of cancer, is often associated with disruptions in spindle assembly checkpoint (SAC), further disrupting mitotic checkpoint complex formation (MCC) [50]. BUB1, a key component of this SAC, is a serine/threonine kinase that ensures precise chromosome segregation during cell division by being recruited to kinetochores to monitor chromosome attachment to the mitotic spindle, delaying the transition to anaphase if chromosomes are not attached properly [49]. BUB1 stabilizes the Anaphase Promoting Complex/Cyclosome’s (APC/C’s) inhibitor complex to prevent premature separation of sister chromatids and interacts with several molecules to regulate the SAC and ensure chromosomal stability during cell division [51, 52].

Recent studies have shown that mutations or dysregulation of BUB1 can cause CIN, leading to genetic variability in tumor progression [52], while overexpression is linked to poor prognosis and therapeutic resistance in various cancers, including neuroblastoma [53], bladder [54], endometrial [55], gastric [56], pancreatic [57], adrenocortical [58], breast [37, 38], colorectal [61], and hepatocellular carcinomas [62]. Additionally, the knockdown of BUB1 in breast cancer cells leads to the restoration of chromosomal stability and a reduction in tumor growth, highlighting its potential for therapeutic applications [63]. In conjunction with the previous literature, our study also demonstrates that BUB1 is significantly overexpressed in LUAD compared to normal lung tissues, with increased expression correlating with poor OS, FP, and PPS. BUB1 overexpression is associated with significant prognostic implications in smokers, exhibiting a high hazard ratio correlated with reduced survivability in both male and female populations. This observation is consistent with previous research showing that smoking-induced genomic instability increases the abnormal expression of mitotic checkpoint regulators such as BUB1 [64]. Further, our study identified a transcription factor, E2F1, as a significant regulator of BUB1 expression and observed a significant positive correlation between BUB1 and E2F1, with E2F1 overexpression associated with advanced LUAD stages and smoking history. Previous studies have consistently shown that E2F1 plays a crucial role in regulating cell cycle genes and is often dysregulated in multiple cancer types, such as lung cancer [65]. The novelty of the study was to identify a ceRNA network essential for the regulation of BUB1 expression in LUAD, mediated by interactions among miRNAs, lncRNAs, and transcription factors. Notably, out of the 17 miRNAs and 10 lncRNAs, hsa-let-7b-5p and TMPO-AS1 were found to be strongly correlated with both BUB1 and E2F1. Interestingly, the correlation analysis showed a negative correlation between TMPO-AS1 and hsa-let-7b-5p, pointing towards the hypothesis that TMPO-AS1 acts as a molecular sponge that binds to the miRNA, hsa-let-7b-5p, thereby diminishing its capacity to suppress BUB1, resulting in its overexpression. Furthermore, TMPO-AS1 showed a positive correlation with both E2F1 and BUB1. These findings align with earlier research indicating that ceRNA networks, especially those involving lncRNA-mediated miRNA sponging, facilitate oncogenic pathways in cancers through the modulation of critical regulators [66, 67]. This network offers a potential strategy for therapeutic intervention, wherein targeting lncRNAs (inhibiting) or restoring miRNA (mimicking) levels may inhibit BUB1 overexpression and its associated oncogenic effects. Recently, in colorectal cancer, BUB1’s interaction with similar transcriptional and post-transcriptional regulators, including E2F1 and miRNAs, has been well-documented, mirroring the mechanisms observed in LUAD [68]. Researchers have tried to develop non-invasive diagnostic techniques for cancer based on the circulating miRNAs, such as in a study conducted by Abdipourbozorgbaghi et. al. which found a panel of miRNAs for LUAD and LUSC diagnosis that served as independent prognostic markers for survival [69]. Similarly, point-of-care (POC) testing devices have been designed to detect miR-21 in urine samples, aiding in prostate cancer diagnosis [70]. Further, to explore the extracellular expression potential of hsa-let-7b-5p, we analyzed its abundance across various human body fluids by using the EVAtlas and EVmiRNA 2.0 databases [71, 72]. The analysis revealed detectable levels of hsa-let-7b-5p across a broad range of human biofluids, including serum, plasma, breast milk, amniotic fluid, and cord blood, indicating its strong detectability and relevance for liquid biopsy-based diagnostics (Supplementary Fig. 10D–E). Based on these results, it can be stated that small molecules, such as miRNA mimics or lncRNA disruptors targeting these regulatory pathways, could offer a more precise therapeutic approach by restoring natural gene suppression mechanisms. The study suggests increasing hsa-let-7b-5p levels to inhibit BUB1 expression, consistent with strategies used in other cancers utilizing miRNA mimics to modify gene expression [73]. Further, targeting the E2F1-BUB1 axis using small molecule inhibitors could be a viable strategy to disrupt LUAD progression, reducing systemic toxicity and combating resistance more effectively than traditional inhibitors. Additionally, our finding reveals that BUB1, E2F1, hsa-let-7b-5p, and TMPO-AS1 play a crucial role in the M phase of the cell cycle, especially through interaction with the CDK1/Cyclin B complex. BUB1 overexpression is also associated with mutations in TP53 and other pathways critical to cancer progression [55]. Given that TP53 loss disrupts the p53/Rb regulatory axis, leading to unchecked activation of E2F1, it is notable that TP53 mutation acts as an upstream event driving the concurrent overexpression of E2F1 and its downstream effector BUB1 [74]. This deregulated p53–E2F1–BUB1 axis could represent a unifying mechanism contributing to chromosomal instability and uncontrolled proliferation in LUAD. Moreover, it was observed that BUB1 plays a complex role in tumor biology by interacting with vital processes like DNA damage repair and epithelial-to-mesenchymal transition (EMT), crucial for metastasis, influencing the aggressive nature of LUAD in advanced stages [75, 76].

Furthermore, BAY-1816032, a selective BUB1 inhibitor, has shown potential in sensitizing lung cancer cells to chemotherapy and radiotherapy, enhancing DNA damage responses, and promoting apoptosis [77]. In preclinical models, BAY-1816032 significantly improved the efficacy of chemoradiation therapies, positioning it as a promising agent for integration into current lung cancer treatment protocols [78]. However, these inhibitors often face limitations like off-target effects, toxicity, and resistance [79]. In this context, given the growing interest in plant-derived compounds as targeted anticancer agents, our findings demonstrate that Hesperidin, a bioflavanoid, exhibits the strongest binding affinity to BUB1 compared to conventional chemotherapeutic agents like Paclitaxel and Docetaxel in docking analyses. This interaction is attributed to Hesperidin’s ability to establish multiple hydrogen bonds, van der Waals interactions, and π-π stacking with key residues in the BUB1 active site, ensuring stable and specific binding. Additionally, Hesperidin’s natural origin and reported low toxicity offer a promising therapeutic advantage, reducing the adverse effects typically associated with traditional chemotherapy [39]. Given these findings, Hesperidin may emerge as a compelling candidate for BUB1-targeted therapy in LUAD. Our findings suggest that BUB1 not only plays a critical role in LUAD cell proliferation but also influences the tumor immune microenvironment. Tumor-infiltrating immune cells (TIICs) are key regulators of immune evasion, and we observed a negative correlation between activated CD4 + T cells and BUB1 expression. Moreover, recent original research underscores the complexity of the TME in NSCLC and its impact on immunotherapy response. For example, Hu et al. (2023) revealed substantial remodelling of immune, myeloid and stromal cells following neoadjuvant PD-1 blockade plus chemotherapy, where responders exhibited increased antigen-presentation and reduced suppressive myeloid populations [80]. Furthermore, Desai et al. (2025) demonstrated that spatial dysfunction of T-cells and enriched immunoregulatory macrophages in the NSCLC TME are associated with impaired antitumour immunity [81]. In addition, a study targeting CAFs via biomimetic nanovesicles (2025) showed how cytokines such as IL-6 and CCL2 secreted by stromal components modulate tumour growth and immune-suppression [82]. Taken together, these findings support our observation of BUB1’s negative correlation with activated CD4 + T cells and suggest that overexpression of mitotic-checkpoint regulators may not only drive proliferation but also contribute to establishing an immunosuppressive TME, thereby influencing immunotherapy efficacy.

While our research offers a fresh perspective on the early diagnosis and prognosis of lung cancer, certain limitations remain. Our analyses are based on data from publicly accessible databases that have been validated; however, it is crucial to acknowledge the necessity for further prospective studies with larger clinical cohorts, and future research should include experimental validation to strengthen the study’s clinical significance.

Conclusion

Our study identifies the BUB1/E2F1/hsa-let-7b-5p/TMPO-AS1 axis as a key regulatory ceRNA network contributing to LUAD progression, particularly in smoker patients. We have shown that BUB1 is significantly overexpressed in LUAD and is transcriptionally regulated by E2F1, while its post-transcriptional control is modulated by hsa-let-7b-5p and TMPO-AS1 through a ceRNA mechanism. The miRNA hsa-let-7b-5p is downregulated, while lncRNA TMPO-AS1 is upregulated, facilitating BUB1’s overexpression. This dysregulation is associated with poor prognosis and reduced immune infiltration, particularly of CD4⁺ T cells, suggesting a role in immune evasion. Furthermore, Hesperidin demonstrated strong binding affinity to BUB1, highlighting its potential as a targeted therapeutic agent with lower toxicity. Overall, this regulatory network offers promise as a prognostic biomarker and therapeutic target in LUAD, especially for smoker subgroups, and lays the groundwork for further experimental validation and translational research.

Methods

Survival analysis

The KMP database was utilized to evaluate the prognostic significance of BUB1 in lung cancer, assessing its Overall survival (OS), First progression survival (FP), Post progression Survival (PPS), subtypes (LUAD and LUSC), and clinical parameters (stages, gender, and smoking history) [83]. “Gene symbol and Affy ID used: BUB1, 209642_s_at”. Patients were stratified into low-expression and high-expression groups based on median expression level of gene to minimize biasness and outlier effects. Inclusion and exclusion criteria were established for further analysis, including univariate and multivariate analysis, as mentioned in Supplementary Table 1.

BUB1 expression profiling and transcription factor analysis

To evaluate the expression pattern of the BUB1 gene in normal vs. LUAD samples, multi-omics databases such as UALCAN [84], ENCORI [85], Lung Cancer Explorer [86], OncoDB [87], and GEPIA2 [88] were employed. Expression profile of BUB1 in LUAD was extracted from TCGA-LUAD dataset using R-based packages, RNA-Seq (STAR–Counts) data were retrieved using the TCGAbiolinks. Data handling, filtering, and visualization were performed in R (v4.5.1) with the libraries tidyverse, dplyr, biomaRt, ggplot2, and ggpubr. Expression values were normalized to Transcripts per million (TPM), and the Ensembl ID for BUB1 (ENSG00000169679) was identified using biomaRt. To ensure balanced group comparison, primary tumor samples were randomly downsampled to match the number of solid tissue normal samples (n = 59 each). Statistical comparison of BUB1 expression was performed using the Wilcoxon rank-sum test, and results were visualized as box plot with annotated p-values. Additionally, the transcriptomic expression of BUB1 transcripts in LUAD was assessed using single-cell RNA sequencing data retrieved from the TCGAnalyzeR [89] database. Further, the BUB1’s differential gene expression was analyzed in normal, tumor and metastatic state by using TNMplot database [90]. Next, BUB1’s expression analysis was conducted based on clinical parameters, and TP53 mutants versus normal by utilizing UALCAN and TIMER2.0 databases [91], while BUB1 gene and its protein expression correlation were analyzed using the Cancer Proteome database [92]. Furthermore, to explore the potential influence of tobacco-derived carcinogens on BUB1 activity, molecular docking and visualization analyses were performed using NNK (PubChem CID: 47,289) and Nicotine (PubChem CID: 89,594) as ligands. The BUB1 protein structure was docked with these compounds to evaluate their binding affinities and interaction profiles, thereby assessing whether these carcinogens could directly modulate BUB1’s kinase function or contribute to its dysregulation in LUAD.

The Enrichr dataset (TRRUST) [93] was used to find the transcription factor regulating BUB1 expression, with correlation analysis conducted by using the ENCORI, OncoDB, and GEPIA2 databases. Additionally, the 2 kb upstream and 100 bp downstream region of the BUB1 transcription start site (TSS; GRCh38/hg38) was retrieved using the TxDb.Hsapiens.UCSC.hg38.refGene and BSgenome.Hsapiens.UCSC.hg38 packages in R. The genomic sequence corresponding to this region was extracted using the getSeq() function. To identify potential transcription factor binding motifs, the E2F1 position frequency matrix (PFM; ID: MA0024.3) was obtained from the JASPAR 2022 database via the TFBSTools package. The PFM was converted into a position weight matrix (PWM) and scanned across the BUB1 promoter sequence using the searchSeq() function, with a minimum score threshold of 80% relative to the maximum possible PWM score. Predicted binding sites with relative scores ≥ 0.80 were considered high-confidence motifs. The identified binding site positions, orientation, and scores were visualized using ggplot2 in R, marking the transcription start site (TSS) as reference (0 bp) and plotting the relative positions on both forward and reverse DNA strands. Further, the survival analysis of the transcription factor was observed using the KMP database (E2F1, 204947_at), while differential gene expression and clinical parameters were analyzed through the UALCAN and ENCORI databases, with further validation performed on normalized samples using R packages (Ensembl ID: ENSG00000101412). Additionally, the E2F1’s differential gene expression was analyzed in normal, tumor, and metastatic states by using the TNMplot database.

The competitive endogenous RNA (ceRNA) network analysis

The study next aimed to identify regulatory molecules, specifically non-coding RNAs (miRNAs and lncRNAs) affecting BUB1 gene expression. The miRNet database [94] was utilized to establish a link between BUB1, TF, and their miRNAs. A list of miRNAs was downloaded and analyzed for differential expression using the UALCAN database. Further, downregulated miRNAs were selected, and their correlation with BUB1 and E2F1 was analyzed by using the ENCORI database. The prognostic significance of selected miRNAs were assessed using the KMP database, followed by the evaluation of area under the curve (AUC) values by using the CancerMIRNome database [95]. Differential expression of the selected miRNA was evaluated using CancerMIRNome, UALCAN, and R-based packages with normalized LUAD samples (n = 46), and its expression was correlated with clinical parameters using UALCAN. Further, to identify lncRNAs associated with BUB1, the Enrichr database was employed, followed by expression and correlation analyses using UALCAN and ENCORI. Survival analysis was performed using the KMP database. Additionally, differential expression of lncRNA in LUAD was assessed using UALCAN, OncoDB, ENCORI, and the Lung Cancer Explorer, with validation carried out using R-packages on normalized datasets. Furthermore, the miRNet database was used for network analysis (lncRNA/BUB1/TF/miRNA). Databases such as miRWalk [96] and RNA22v2 [35] were utilized to evaluate the binding affinities between miRNA-hsa-let-7b-5p and the BUB1 gene, the TF-E2F1, and lncRNA-TMPO-AS1. Radar plots were generated using the TNMplot database by querying BUB1, E2F1, and TMPO-AS1 expression in normal and tumor lung tissues. Expression values were derived from integrated RNA-seq datasets (TCGA, GTEx, GEO) and visualized as log2-transformed TPM values, with concentric circles indicating increasing expression levels.

Molecular docking and visualization analysis

BUB1’s crystal structure (PDB ID: 6F7B) was retrieved from the Protein Data Bank (PDB). Ligand structures for BAY-1816032 (PubChem ID: 118,958,833), paclitaxel (PubChem ID: 36,314), docetaxel (PubChem ID: 148,124), quercetin (PubChem ID: 5,280,343), and hesperidin (PubChem ID: 10,621) were acquired from PubChem, and minimization and protein cleaning were done using UCSF Chimera [97]. AutoDock Tools 1.5.7 [98] was employed to perform molecular docking, generating ten docking conformations for each ligand. The top-ranked conformation based on binding affinity was selected for analysis. Interaction profiling, including hydrogen bonding, van der Waals forces, and hydrophobic contacts, was conducted using Discovery Studio Visualizer (DSV) to assess binding stability.

Cell cycle and biological process analysis

To explore the functional roles of BUB1, E2F1, and TMPO-AS1 in lung cancer, a comprehensive analysis was conducted to assess their involvement in the cell cycle and key biological processes. Cyclebase 3.0 was used to visualize BUB1 expression levels in different cell cycle checkpoints [99], while the CancerSEA database revealed BUB1's involvement in various biological pathways [100]. Further, gene enrichment analysis for BUB1, E2F1, and TMPO-AS1 was performed via the GENI database [101], revealing their upregulation linked to biological processes, cellular components, and molecular functions. The ENCORI database further analyzed the correlation between the BUB1 gene and cell cycle regulators, Cyclins and CDKs.

Analysis of pathological stage expression trends

Next, by using MKI67 as a gold standard, the correlation between BUB1, E2F1, and TMPO-AS1 genes across various pathological stages was analyzed by utilizing the GSCA database [102]. Additionally, a trend plot was created for five genes, including the TP53 tumor suppressor gene, to analyze their expression trends across different stages. The study aimed to compare these genes for their prognostic significance. Furthermore, Gene Set Variation Analysis (GSVA) scores were computed to validate the findings by comparing the expression patterns of the five genes between normal and tumor samples.

Immune cell infiltration analysis

To evaluate the immune involvement of the BUB1 regulatory axis, we conducted a multi-step immune cell infiltration analysis. Initially, single-cell immune landscape mapping was performed using the SPRING viewer database [103], where BUB1 expression was visualized across immune compartments in NSCLC to understand its spatial localization. To further explore immunological associations, correlation analysis was conducted via the GSCA database using LUAD-specific datasets to assess the expression patterns of BUB1, its transcriptional regulator E2F1, and associated lncRNA TMPO-AS1 against various immune cell types. A GSVA score was generated for the combined gene set (BUB1–E2F1–TMPO-AS1), and its relationship with immune infiltration, particularly CD4⁺ T cells, was quantified. Finally, to validate these correlations, the TIMER2.0 platform was utilized to identify the association between gene expression and infiltration levels of CD4⁺ T cell subsets, including effector and memory phenotypes in LUAD.

Statistical analysis

The study investigated differences in gene expression between tumor and normal tissues by utilizing an online model (in silico gene set) to explore the relationship between gene expression and prognosis. Additionally, the raw RNA-seq data was log2-transformed, and statistical analysis was performed using the Wilcoxon rank-sum test, with a significance threshold of p < 0.05. The results were visualized using ggplot2 in R, highlighting the differential expression of BUB1, TF-E2F1, lncRNA-TMPO-AS1, and miRNA-hsa-let-7b-5p between tumor and normal tissues.

Abbreviations

BUB

Budding uninhibited by benzimidazoles

CancerSEA

Cancer single cell state Atlas

LUAD

Lung adenocarcinoma

LUSC

Lung squamous cell carcinoma

HR

Hazard ratio

CI

Confidence interval

TF

Transcription factor

NSCLC

Non- small cell lung adenocarcinoma

SCLC

Small cell lung adenocarcinoma

TP53

Tumor protein p53

TRRUST

Transcriptional regulatory relationships unraveled by sentence-based text mining

LBC

Lepidic-predominant adenocarcinoma

ceRNA

Competing endogenous RNA

DNA

Deoxyribonucleic acid

E2F1

Eukaryotic transcription factor1

ENCORI

Encyclopedia of RNA Interactomes

ONCODB

Oncology database

GEPIA2

Gene expression profiling interactive analysis

hsa-let-7b

Homosapiens MicroRNA family

miRNet

MicroRNA network

Enrichr

Enrichment analysis resource

KM PLOTTER

Kaplan–Meier Plotter

KMP

Kaplan–Meier Plotter

DGE

Differential gene expression

lncRNA

Long noncoding RNA

miRNA

Micro ribonucleic acid

MKI67

Marker of proliferation Ki-67

OS

Overall survival

FP

First progression

PPS

Post progression survival

GENI

Global environment for network innovations

RNA

Ribonucleic acid

TCGA

Portal-the cancer genomic atlas portal

TCGAnalyzerv1.0

The cancer genome Altas analyzer

TMPO-AS1

Thymopoietin antisense transcript1

TNM Plot

Tumor node metastasis plot

UALCAN

The University of Alabama at Birmingham Cancer Data Analysis Portal

YBX1-Y

Box binding protein 1

RBL2

Retinoblastoma- like protein 2

TFDP1

Transcription fcator DP1

MYC

Myelocytomatosis

AUC

Area under the curve

ROC

Receiver operating characteristic

EGF

Epidermal growth factor receptor

ALK

Anaplastic lymphoma kinase

APC

Anaphase promoting complex

EMT

Epithelial mesenchymal transition

GSCA

Gene set cancer analysis

GSVA

Gene set variation analysis

TNM Plot

Tumor, node, metastasis plot

TPM

Transcripts per million

TIMER2.0

Tumor immune estimation resource 2.0

Author contributions

RN: Conception, study design, critical reading, intellectual assessment of the manuscript, preparation of the manuscript and final approval. AK: Conception, study design, critical reading, intellectual assessment of the manuscript, preparation of the manuscript, and final approval. BB: Study design, and preparation of the manuscript, critical review, PV: Study design, and preparation of the manuscript, critical review, and CS: Study design, and preparation of the manuscript, critical review.

Funding

Open access funding provided by Manipal University Jaipur. RN would like to thank the funding support from Manipal University Jaipur for the Enhanced Seed Grant under the Endowment Fund (No. E3/2023–24/QE-04–05) and DST-FIST project (DST/2022/1012) from Govt. of India to Department of Biosciences, Manipal University Jaipur.

Availability of data and materials

The data that support the findings of this in silico analysis are available from the corresponding author upon request. The datasets and scripts used for the analyses in this study are publicly available in the following GitHub repository: https://github.com/Bhavika-2001/BUB1_LUAD.

Declarations

Ethics approval and consent to participate

This study does not require any ethical approval or consent of participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.