The MAD2L1–FOXM1 axis negatively regulates CD4⁺ T cell and macrophages infiltration in lung adenocarcinoma

Background

Lung adenocarcinoma (LUAD) exhibits high genetic heterogeneity and an immunosuppressive tumor microenvironment (TME), often characterized by reduced immune cell infiltration. Tumor-intrinsic regulators of cell cycle and genomic stability, such as MAD2L1, may also influence immune evasion, but their mechanistic role in LUAD remains unclear.

Methods

Co-expression profiling was carried out to identify genes positively and negatively associated with MAD2L1 by using Correlation AnalayzeR and was corroborated using ENCORI, GEPIA2 and TIMER2.0 databases. Subsequent prognostic evaluation for all the genes was performed by using KM Plotter. The Spring Viewer transcriptomic database was used to assess the immune cell infiltration patterns correlating with these genes in NSCLC and the association was evaluated using the GSCA and TIMER2.0 databases. Additionally, the expression dynamics across pathological stages was determined by using GSCA.

Results

FOXM1 was previously identified as a key upstream regulator of MAD2L1, with co-expression analysis revealing strong positive associations with cell cycle regulators (BIRC5, BUB1B, TTK, CCNA2, DLGAP5) and negative associations with immune-related genes (IL3RA, MPEG1, NCKAP1L, GIMAP8, MS4A14). High MAD2L1 and its positive gene panel correlated with reduced CD4⁺ T cell infiltration, while the negative panel correlated with macrophage-rich TMEs. Stage-wise analysis showed an increase in the positive panel and a decrease in the negative panel from stage I to IV.

Conclusion

The MAD2L1 axis emerges as a potential driver of immune exclusion in LUAD by suppressing CD4⁺ T cell infiltration, linking tumor proliferation to immune evasion. Targeting this axis may offer a strategy to restore anti-tumor immunity.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-025-04360-8.

Highlights

• MAD2L1 co-expresses with cell cycle genes and negatively associates with the immune-related genes.

• High MAD2L1 and its positively associated gene panel expression reduces CD4⁺ T cell infiltration in the tumor microenvironment.

• High MAD2L1 associated negative gene panel expression increases macrophage infiltration in the tumor microenvironment.

• MAD2L1’s positive regulators increase, and negative regulators decrease with advancing LUAD stage.

• The FOXM1–MAD2L1 axis could be proposed as a driver of immune exclusion in LUAD

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-025-04360-8.

Introduction

According to the World Cancer Research Fund 2022, lung cancer accounted for 2,480,675 new cases globally [1]. China reported the highest number of new cases (1,060,584), followed by the United States (226,033) and Japan (136,723) [2]. India ranked fourth with 81,748 new cases, while Russia (70,362), Germany (62,025), the United Kingdom (50,700), France (49,613), Brazil (44,213), and Italy (43,808) were also among the top ten countries with the highest incidence [2]. Lung adenocarcinoma (LUAD), the most common subtype of non-small cell lung cancer (NSCLC), remains a leading cause of cancer-related mortality worldwide [3]. This unfavourable prognosis reflects not only the high genetic heterogeneity of LUAD but also the complexity of its tumor microenvironment (TME), which actively promotes tumor growth and facilitates immune evasion [4, 5]. A key feature of this immunosuppressive landscape is the reduction or exclusion of CD4⁺ T cells, essential modulators of anti-tumor immunity whose presence strongly correlates with improved patient outcomes [6]. Although immune checkpoint inhibitors (ICIs) targeting PD-1, PD-L1, and CTLA-4 have transformed LUAD therapy, clinical benefit is often limited by primary or acquired resistance [7, 8]. Low T cell infiltration, impaired antigen presentation, and an immunosuppressive TME are recognized contributors to therapeutic failure [9]. Similarly, while chimeric antigen receptor (CAR) T cell therapy has shown remarkable efficacy in haematological malignancies, its application to LUAD has been hindered by poor trafficking, antigen heterogeneity, and T cell exhaustion [10]. These limitations underscore the need to identify tumor-intrinsic factors that shape immune cell infiltration and function. Recent studies suggest that mitotic checkpoint regulators, traditionally associated with genomic stability, may also influence the immune landscape of tumors. The Mitotic Arrest Deficient- Like protein MAD2L1, a key spindle assembly checkpoint component, ensures accurate chromosome segregation during mitosis [11]. Its dysregulation leads to chromosomal instability, a hallmark of aggressive cancers [12]. While MAD2L1’s role in tumor proliferation is well established, its potential immunomodulatory effects in LUAD remain largely unexplored.

Forkhead Box M1 (FOXM1), a transcription factor frequently overexpressed in solid tumors, is a critical upstream regulator of MAD2L1 and other G2/M phase genes [13]. FOXM1 drives tumor proliferation, metastasis, and therapy resistance, and emerging evidence suggests it may influence immune-related gene expression [14]. This raises the possibility that a FOXM1–MAD2L1 transcriptional axis could serve as a mechanistic bridge between cell cycle dysregulation and immune suppression in LUAD. We hypothesize that aberrant activation of the FOXM1–MAD2L1 axis contributes to immune evasion by restricting CD4⁺ T cell infiltration into the tumor microenvironment. This study integrates multi-database co-expression profiling, immune infiltration analysis, and clinical outcome correlations to elucidate the role of this axis in LUAD progression. By defining the transcriptional network surrounding MAD2L1 and mapping its immune associations, we aim to provide mechanistic insight into LUAD’s immunosuppressive phenotype and identify potential targets for therapeutic intervention.

Methodology

After establishing the MAD2L1’s upregulation in LUAD and identifying FOXM1 as its transcriptional regulator in our previous study [15]. We aimed to explore the heterogeneity associated with MAD2L1 and its related microenvironment. In this context, we have used several databases, for which the selection criteria that we have set is mentioned in Supplementary Table 1 and the workflow of our study is shown in Supplementary Fig. 4. For the same, MAD2L1-associated genes were identified using the Correlation AnalyzeR platform https://correlationanalyzer.bishop-lab.com/ [16]. Pearson correlation coefficients (R) were obtained to determine the top 10 positively and negatively associated genes, and to corroborate these co-expression relationships, ENCORI https://rnasysu.com/encori/ [17], GEPIA2 (Gene Expression Profiling Interactive Analysis 2) http://gepia2.cancer-pku.cn/#index [18], and TIMER2.0 (Tumor Immune Estimation Resource) http://timer.cistrome.org/ [19] databases were employed. Furthermore, to determine the clinical significance of MAD2L1’s positively and negatively associated genes, overall survival (OS) analysis was performed in LUAD using the Kaplan–Meier Plotter database https://kmplot.com/analysis/ [20]. Genes with significant Hazard Ratios (HR) and log-rank p-values were selected for further analysis. The gene symbol and Affymetrix IDs were as follows: CCNA2 (Affy: 213226_at), CDK1 (Affy: 203213_at), AURKA (Affy: 208079_s_at), DLGAP5 (Affy: 210227_at), BIRC5 (Affy: 202094_at), NUF2 (Affy: 223381_at), BUB1B (Affy: 203755_at), TTK (Affy: 204822_at), CHEK1 (Affy: 238075_at), CCNB1 (Affy: 228729_at), CAPN3 (Affy: 210944_s_at), IL3RA (Affy: 206148_at), MPEG1 (Affy: 226841_at), NCKAP1L (Affy: 209734_at), ITGAX (Affy: 210184_at), PDES1B (Affy: 206444_at), GIMAP8 (Affy: 235306_at) and MS4A14 (Affy: 229510_at).

Immune microenvironment exploration of MAD2L1 associated genes

Additionally, to visualize immune cell–specific expression patterns, the SPRING Viewer database https://kleintools.hms.harvard.edu/tools/springViewer_1_6_dev.html?datasets/Zilionis2019/human/NSCLC_all_cells [21] (a tool of the Harvard institute), a transcriptomic-based projection, was employed. Gene expression profiles were overlaid on reference immune cell cluster maps (including CD4⁺ T cells, B cells, NK cells, macrophages, dendritic cells, and neutrophils). Additionally, quantitative immune correlation analysis was conducted using GSCA (Gene Set Cancer Analysis) https://guolab.wchscu.cn/GSCA/#/ [22] to calculate Spearman correlation between gene expression and immune infiltration scores for selected immune cell types, which were further validated by using TIMER2.0. Furthermore, two gene panels were defined: a positive panel (MAD2L1, FOXM1, and top positively correlated genes) and a negative panel (top negatively correlated genes), for which pathological stage-associated expression patterns were analyzed using Kruskal–Wallis testing to detect significant trends across stages I–IV by using the GSCA database.

Results

MAD2L1–FOXM1 axis in LUAD female smokers

In our prior study, we comprehensively demonstrated that MAD2L1 is significantly overexpressed in LUAD, and was strongly associated with poor prognosis in smoker females, as evidenced by survival analysis that highlighted MAD2L1 as a potential prognostic biomarker. Furthermore, transcription factor prediction and expression correlation analysis confirmed FOXM1 as a key upstream regulator of MAD2L1 [15]. Collectively, these findings establish the FOXM1–MAD2L1 axis as a critical pathway contributing to the aggressiveness and unfavourable outcomes observed in LUAD among female smokers.

MAD2L1 is co-regulated with cell cycle genes in LUAD

To explore the heterogenetic landscape associated with MAD2L1 in LUAD, we first identified the top 100 positively and negatively associated genes using Correlation AnalyzeR and found their correlation with MAD2L1 as mentioned in Supplementary Table 2 and 3. Only top 10 positively and negatively genes were considered for further analysis based on their strongest correlation coefficients and highly significant p-values. The positively correlated genes were predominantly cell cycle regulators and showed strong correlation with MAD2L1, including CCNA2 (R = 0.871), CDK1 (R = 0.870), AURKA (R = 0.850), DLGAP5 (R = 0.849), BIRC5 (R = 0.846), NUF2 (R = 0.845), BUB1B (R = 0.841), TTK (R = 0.838), CHEK1 (R = 0.833), and CCNB1 (R = 0.832) as listed in Table 1. We subsequently validated their correlation with MAD2L1 by using the ENCORI, GEPIA2, and TIMER2.0 databases. This consistent co-expression pattern between MAD2L1 and cell cycle genes across independent databases highlights the integration of MAD2L1 into mitotic progression and genomic stability networks, reinforcing its role in driving LUAD proliferation.

Table 1.

Positively associated genes with MAD2L1

S. No	Gene	Correlation Value (R-value)
		Correlation AnalyzeR	ENCORI	GEPIA2	TIMER2.0
1.	CCNA2	0.871	0.940	0.910	0.945
2.	CDK1	0.870	0.884	0.70	0.886
3.	AURKA	0.850	0.793	0.680	0.790
4.	DLGAP5	0.849	0.873	0.770	0.881
5.	BIRC5	0.846	0.883	0.820	0.855
6.	NUF2	0.845	0.841	0.730	0.835
7.	BUB1B	0.841	0.840	0.750	0.834
8.	TTK	0.838	0.858	0.760	0.862
9.	CHEK1	0.833	0.783	0.730	0.811
10.	CCNB1	0.832	0.871	0.820	0.874

Conversely, CAPN3, IL3RA, MPEG1, CD163, NCKAP1L, MS4A6A, ITGAX, PDE1B, IMAP8 and MS4A14 were among the top 10 genes exhibiting inverse expression patterns with MAD2L1. The correlation values observed were CAPN3 (R = -0.657), IL3RA (R = -0.646), MPEG1 (R = -0.643), CD163 (R = -0.637), NCKAP1L (R = -0.632), MS4A6A (R = -0.629), ITGAX (R = -0.628), PDE1B (R = -0.623), GIMAP8 (R = -0.614) and MS4A14 (R = -0.614), as determined by Correlation AnalyzeR. The correlation values were consistent when validated by using ENCORI, GEPIA2, and TIMER2.0 databases for all the genes except for CD163 and MS4A6A (Table 2). Hence, they were not considered for further analysis.

Table 2.

Negatively associated genes with MAD2L1

S. No	Gene	Correlation Value (R-value)
		Correlation AnalyzeR	ENCORI	GEPIA2	TIMER2.0
1.	CAPN3	-0.657	-0.518	-0.150	-0.513
2.	IL3RA	-0.646	-0.229	-0.270	-0.129
3.	MPEG1	-0.643	-0.111	-0.150	-0.056
4.	CD163	-0.637	0.085	-0.073	0.123
5.	NCKAP1L	-0.632	-0.103	-0.200	-0.041
6.	MS4A6A	-0.629	0.008	-0.032	-0.034
7.	ITGAX	-0.628	-0.103	-0.160	-0.059
8.	PDE1B	-0.623	-0.231	-0.190	-0.232
9.	GIMAP8	-0.614	-0.235	-0.270	-0.157
10.	MS4A14	-0.614	-0.208	-0.240	-0.156

Prognostic significance of MAD2L1 associated genes in LUAD

To further elucidate the clinical significance of MAD2L1 associated genes, a comprehensive Kaplan–Meier survival analysis was conducted to evaluate the prognostic impact of the positively and negatively correlated genes identified through correlation analysis. Among the genes positively correlated with MAD2L1 expression, BIRC5 (HR = 1.96, P = 1.5e-14), CCNA2 (HR = 1.36, P = 0.00049), CCNB1 (HR = 1.36, P = 0.012), AURKA (HR = 1.56, P = 4.2e-07), CHEK1 (HR = 1.77, P = 3e-06), NUF2 (HR = 1.53, P = 0.00049), BUB1B (HR = 1.67, P = 5.3e-09), TTK (HR = 1.68, P = 5.3e-09), and DLGAP5 (HR = 1.48, P = 7e-06) demonstrated a statistically significant association with poor overall survival in LUAD patients, indicating their role as oncogenes, as shown in Fig. 1A-J and Table 3. This is consistent with previous evidence highlighting their oncogenic potential. BUB1B, a critical mitotic checkpoint kinase, was also linked to adverse prognosis, reinforcing its role in chromosomal instability and tumor proliferation [23]. TTK, another spindle assembly checkpoint kinase, is overexpressed in NSCLC and its elevated levels predict advanced stage and poorer prognosis [24]. Similarly, DLGAP5, a regulator of spindle microtubule dynamics, is frequently upregulated in LUAD and linked to early-stage tumor detection [25]. CCNA2 and CCNB1, core regulators of the G2/M transition, as well as AURKA and CHEK1, which govern mitotic entry and DNA damage responses, have also been implicated in tumor aggressiveness and genomic instability [26–28]. BIRC5, an inhibitor of apoptosis, and NUF2, a kinetochore component essential for chromosome segregation, further reinforce this pro-tumorigenic network [29, 30]. Collectively, these findings indicate that co-expression of MAD2L1 with other mitotic and G2/M transition-related genes exacerbates malignant potential and contributes to adverse clinical outcomes in LUAD.

Fig. 1.

Prognostic significance of MAD2L1 positively correlated genes in LUAD of (A) BIRC5, (B) CCNA2/ CCNA, (C) CCNB1, (D) AURKA, (E) CDC2/ CDK1, (F) CHEK1, (G) TTK/ CT96, (H) NUF2, (I) BUB1B/ SSK1, and (J) DLGAP5

Table 3.

Survival analysis of MAD2L1 positively associated genes

S.NO	Gene	Index	Patient Number	Hazard Ratio	CI	Log(P)	Low Expression Cohort	High Expression Cohort
1.	BIRC5	OS + LUAD	1161	1.96	1.65–2.34	1.5e-14	112.67	52
2.	CCNA2	OS + LUAD	1161	1.36	1.14–1.61	0.00049	96	71
3.	CCNB1	OS + LUAD	672	1.36	1.07–1.72	0.012	119.87	88.7
4.	AURKA	OS + LUAD	1161	1.56	1.31–1.85	4.2e-07	106	63.4
5.	CDC2	OS + LUAD	1161	1.1	0.92–1.3	0.29	80	84
6.	CHEK1	OS + LUAD	672	1.77	1.39–2.26	3e-06	125.77	73.3
7.	TTK/ CT96	OS + LUAD	1161	1.68	1.41–2	5.3e-09	96	57
8.	NUF2	OS + LUAD	672	1.53	1.2–1.96	0.00049	125.77	86
9.	SSK1/ BUB1B	OS + LUAD	1161	1.67	1.4–1.99	5.3e-09	107	61.3
10.	DLGAP2	OS + LUAD	1161	1.48	1.25–1.76	7e-06	107	69

In contrast, among the genes negatively correlated with MAD2L1, few genes demonstrated potential tumor-suppressive behaviour based on their favourable prognostic associations. Notably, upregulation of MS4A14 (HR = 0.77, P = 0.039), MPEG1 (HR = 0.61, P = 5.1e-05), CAPN3 (HR = 0.75, P = 0.0014), IL3RA (HR = 0.79, P = 0.009), and GIMAP8 (HR = 0.57, P = 9.7e-06) were significantly linked to prolonged survival, suggesting that higher expression of these genes may suppress oncogenic signalling pathways as shown in Fig. 2A-H and Table 4. MS4A14, a member of the membrane-spanning 4A family, has been implicated in immune regulation and may enhance anti-tumor immune surveillance [31]. MPEG1 encodes perforin-2, a pore-forming protein involved in innate immunity, which facilitates the clearance of intracellular pathogens and may also contribute to anti-cancer immunity [32]. CAPN3, a skeletal muscle–specific calpain, has emerging evidence of acting as a tumor suppressor by modulating cytoskeletal dynamics and inhibiting metastasis [33]. IL3RA, the interleukin-3 receptor alpha chain, is primarily involved in haematopoietic cell signaling, and its expression in the tumor microenvironment may support immune cell activation against malignant cells [34]. GIMAP8, a GTPase of the immunity-associated protein family, plays a role in lymphocyte survival and homeostasis, potentially strengthening adaptive immune responses [35]. Collectively, the elevated expression of these genes may reflect enhanced immune-mediated tumor control and a less aggressive disease phenotype in LUAD.

Fig. 2.

Prognostic significance of MAD2L1 negatively correlated genes in LUAD of (A) MS4A14, (B) MPEG1, (C) CAPN3, (D) IL3RA, (E) NCKP1L, (F) GIMAP8, (G) ITGAX, (H) PDES1B

Table 4.

Survival analysis of MAD2L1 negatively associated genes

S.NO	Gene	Index	Patient Number	Hazard Ratio	CI	Log(P)	Low Expression Cohort	High Expression Cohort
1.	MS4A14	OS + LUAD	672	0.77	0.61–0.99	0.039	90	110.27
2.	MPEG1	OS + LUAD	672	0.61	0.48–0.77	5.1e-05	74	110.27
3.	CAPN3	OS + LUAD	1161	0.75	0.63–0.9	0.0014	74	95
4.	IL3RA	OS + LUAD	1161	0.79	0.67–0.94	0.009	73	96
5.	NCKP1L	OS + LUAD	1161	1.67	1.4–1.99	4.8e-09
6.	GIMAP8	OS + LUAD	672	0.57	0.45–0.74	9.7e-06	75.43	175
7.	ITGAX	OS + LUAD	1161	1.19	1–1.41	0.044	96	74
8.	PDES1B	OS + LUAD	1161	1.48	1.25–1.76	7.3e-06	108.97	68

Collectively, this analysis indicates that the MAD2L1 expression landscape is enriched for both oncogenic and immune-regulatory gene signatures.

Mapping MAD2L1-associated genes to the immune landscape reveals CD4⁺ T cell and macrophage enrichment

To understand the immune contexture associated with MAD2L1 and its network of positively and negatively co-expressed genes, we first employed SPRING-based transcriptomic projection to visualize their expression across diverse immune cell populations, as shown in Fig. 3A-L and Fig. 4A-E, respectively. Each gene was overlaid on the reference SPRING map that demarcates immune cell clusters such as CD4 + T cells, B cells, NK cells, macrophages, dendritic cells, and neutrophils. Notably, multiple MAD2L1-associated genes, including FOXM1, BIRC5, CCNA2, BUB1B, TTK, and DLGAP1, showed visible enrichment in clusters corresponding to T cells and macrophages, suggesting a potential link between these mitotic regulators and specific immune phenotypes.

Fig. 3.

SPRING-based mapping of MAD2L1 and its positively correlated genes across immune cell populations. (A) Reference SPRING map depicting distinct immune cell clusters, including T cells, B cells, NK cells, macrophages, dendritic cells, neutrophils, and plasma cells. (B–L) Expression projection of MAD2L1 and its positively correlated genes: (B) MAD2L1, (C) FOXM1, (D) BIRC5, (E) CCNA2, (F) CCNB1, (G) AURKA, (H) DLGAP5, (I) CHEK1, (J) BUB1B, (K) NUF2, and (L) TTK

Fig. 4.

SPRING-based mapping of MAD2L1 negatively correlated genes across immune cell populations: (A) CAPN3, (B) IL3RA, (C) MPEG1, (D) GIMAP8, and (E) MS4A14

To validate and quantify these observations, we next evaluated the correlation between the expression of MAD2L1 and its co-regulated genes with the above mentioned immune cell types using the GSCA database. Out of all the cells (CD4_T cells, B cells, macrophages, NK cells), strong inverse correlation patterns emerged for several MAD2L1 positively associated genes with CD4 + T cells (MAD2L1: R = -0.46, FOXM1: R = -0.39, CCNA2: R = -0.44, AURKA: R = -0.44, DLGAP5: R = -0.41, BIRC5: R = -0.44, NUF2: R = -0.43, BUB1B: R = -0.40, TTK: R = -0.43, CHEK1: R = -0.39, CCNB1: R = -0.43 as listed in Table 5. Figure 5A illustrates the enrichment of these genes in CD4_T cell tumor microenvironments, whereas scatter plots obtained from GSCA confirmed these negative associations, with a clear downward trend between all genes’ expression levels and CD4⁺ T cell infiltration scores except for DLGAP5 as shown in Figs. 5B–L. Hence, it was excluded from further analysis. Same correlations were validated for these genes with CD4_T cells by using the TIMER2.0 database (Supplementary Fig. 1A-J). On the contrary, MAD2L1 negatively associated genes (IL3RA: R = 0.41, MPEG1: R = 0.58, GIMAP8: R = 0.38 and MS4A14: R = 0.48) displayed strong positive correlations with macrophages except for CAPN3 (R = 0.08), as listed in Table 6. Figure 6A illustrates the enrichment of these genes in macrophage-rich tumor microenvironments, while scatter plots in Fig. 6B–E depict a consistent upward correlation trend between gene expression and macrophage infiltration scores. To confirm the immune cell correlations observed through SPRING Viewer and GSCA analyses, we validated our findings using the TIMER2.0 database. Consistent with previous results, MAD2L1’s negatively correlated genes showed significant positive correlations with macrophage infiltration levels (Supplementary Fig. 2A-D).

Table 5.

Correlation MAD2L1 and its positively associated genes with Immune cells

S. No	Gene Name	CD4_T	B cell	Macrophage	NK	Infiltration Score
1.	MAD2L1	-0.46	0.18	-0.06	-0.24	-0.21
2.	FOXM1	-0.39	0.20	-0.13	-0.27	-0.24
3.	CCNA2	-0.44	0.19	-0.05	-0.24	-0.19
4.	AURKA	-0.44	0.22	-0.14	-0.25	-0.28
5.	DLGAP5	-0.41	-0.05	-0.04	-0.25	-0.23
6.	BIRC5	-0.44	0.21	-0.09	-0.25	-0.22
7.	NUF2	-0.43	0.25	-0.15	-0.25	-0.27
8.	BUB1B	-0.40	0.22	-0.14	-0.28	-0.27
9.	TTK	-0.43	0.24	-0.07	-0.26	-0.22
10.	CHEK1	-0.39	0.22	-0.14	-0.28	-0.23
11.	CCNB1	-0.43	0.24	-0.08	-0.27	-0.25

Fig. 5.

Correlation of MAD2L1 and its positively co-expressed genes with immune cell infiltration in LUAD. (A) Heatmap showing Spearman correlation coefficients between MAD2L1 positively associated genes and various immune cell types in LUAD, as obtained from GSCA database. (B–L) Scatter plots depicting the negative correlations between CD4⁺ T cell infiltration scores and expression levels of (B) MAD2L1, (C) FOXM1, (D) CCNA2, (E) AURKA, (F) DLGAP5, (G) BIRC5, (H) NUF2, (I) BUB1B, (J) TTK, (K) CHEK1, and (L) CCNB1

Table 6.

Correlation MAD2L1 negatively associated genes with Immune cells

S. No	Gene Name	CD4_T	B cell	Macrophage	NK	Infiltration Score
1.	CAPN3	0.58	-0.06	0.08	0.26	0.32
2.	IL3RA	0.19	-0.17	0.41	0.50	0.61
3.	MPEG1	0.23	0.02	0.58	0.45	0.69
4.	GIMAP8	0.26	-0.07	0.38	0.56	0.55
5.	MS4A14	0.32	-0.12	0.48	0.32	0.62

Fig. 6.

Correlation of MAD2L1 negatively associated genes with macrophage infiltration in LUAD by using GSCA. (A) Heatmap showing Spearman correlation between various immune cell types and the GSVA score of the negative gene panel in LUAD, (B) IL3RA, (C) MPEG1, (D) GIMAP8, and (E) MS4A14

By combining visual mapping using the SPRING Viewer database with quantitative correlation analysis from GSCA and TIMER2.0, we identified CD4⁺ T cells and macrophages as the two immune cell types most closely associated with the MAD2L1 gene network. These results suggest that MAD2L1 and its co-expressed genes may play a dual role in modulating distinct aspects of the tumor-immune microenvironment.

Expression trends across pathological stages

To further explore their clinical relevance, we analyzed expression patterns across pathological stages. The positive gene panel exhibited an upward trend in expression from stage I to stage IV, with consistent increases for MAD2L1, FOXM1, CCNA2, AURKA, BIRC5, NUF2, BUB1B, TTK, CHEK1, and CCNB1 (Supplementary Fig. 3A). Conversely, the negative gene panel consisting of MPEG1, MS4A14, GIMAP8 and IL3RA demonstrated a decreasing trend across advancing stages (Supplementary Fig. 3B). These opposite stage-associated patterns align with the infiltration correlations, reinforcing the potential role of the positive gene panel in immune evasion and disease progression and the negative gene panel in immune-supportive tumor microenvironments.

Discussion

The tumor immune microenvironment plays a decisive role in shaping the progression and therapeutic responsiveness of LUAD, the most common and aggressive subtype of lung cancer [36]. Despite the transformative impact of immune checkpoint inhibitors in subsets of patients, many LUAD tumors remain refractory to immunotherapy, often due to impaired T cell infiltration and an immunosuppressive niche [37]. In recent years, researchers have increasingly focused on gene expression profiles in LUAD to better understand tumor behavior and patient prognosis. Transcriptomic studies have revealed that dysregulation of cell cycle–related genes and suppression of immune-related gene programs are recurrent features in LUAD, and both patterns strongly influence disease progression and therapeutic response [38]. Within this context, our study identifies MAD2L1 as a key regulator not only of mitotic checkpoint control but also of immune modulation, revealing its close transcriptional association with FOXM1 and its integration into a broader network of cell cycle regulators. Beyond its canonical role in chromosome segregation, elevated MAD2L1 expression in LUAD was linked to altered immune cell composition, highlighting its dual role in driving proliferation and reshaping the immune microenvironment.

From an immunological perspective, LUAD represents a highly complex ecosystem where diverse immune cells exert both protective and pathogenic roles. CD8⁺ cytotoxic T cells act as the frontline anti-tumor effectors, but their function is frequently reduced by regulatory T cells (Tregs) and myeloid-derived suppressor cells (MDSCs) [39]. CD4⁺ helper T cells are crucial for sustaining immune activation, and their depletion, as suggested in MAD2L1-high tumors may compromise immune surveillance. Macrophages are particularly versatile, while M1 macrophages facilitate antigen presentation and anti-tumor activity, M2 macrophages promote angiogenesis, tissue remodeling, and immune suppression [40]. NK cells, though central to innate immunity, are often poorly infiltrated in LUAD tumors, further shifting the balance toward immune escape [41]. By revealing that MAD2L1 expression stratifies these immune interactions, our study highlights how mitotic checkpoint regulators may indirectly determine the equilibrium between pro- and anti-tumor immunity.

Recent advances in immunotherapy have reshaped the clinical management of LUAD, with immune checkpoint inhibitors (ICIs) achieving unprecedented survival benefits in subsets of patients. Trials such as KEYNOTE-024 demonstrated the efficacy of pembrolizumab, a PD-1 inhibitor, in improving overall survival in patients with high PD-L1 expression [42]. Similarly, CheckMate-057 showed that nivolumab significantly outperformed docetaxel in previously treated LUAD patients, while IMpower110 highlighted the benefit of atezolizumab as a first-line therapy [43, 44]. Despite these successes, resistance to ICIs remains common, often linked to poor T cell infiltration or immunosuppressive tumor microenvironments [45]. Our findings that MAD2L1 overexpression correlates with reduced CD4⁺ T cell infiltration raise the possibility that tumors with a high-MAD2L1 profile may exhibit intrinsic resistance to checkpoint blockade. Conversely, the enrichment of macrophage-related gene signatures in the MAD2L1-negative gene panel suggests that immune-competent contexts may be more favourable for ICI responsiveness. These insights support the rationale for exploring MAD2L1–FOXM1 axis inhibition in combination with ICIs and potentially with emerging modalities such as CAR-T cell therapy targeting tumor-associated antigens.

The tumor microenvironment (TME) exerts a profound influence on the therapeutic efficacy of immune checkpoint blockade in LUAD by regulating immune cell function, cytokine signaling, and spatial organization of immune–tumor interactions. Beyond overall infiltration levels, the composition and activation state of specific T cell subsets, macrophages, and cytokine networks determine whether the TME supports immune clearance or tolerance. Recent single-cell and spatial multi-omics studies in LUAD have demonstrated that remodeling of T cell and myeloid compartments following immunotherapy strongly correlates with clinical outcome [46]. Spatial profiling has further revealed that the proximity of T cells to tumor nests and the polarization state of macrophages (M1 vs. M2) predict therapeutic response more accurately than bulk immune scores [47–49]. Emerging literature across diverse cancer models further highlights that TME behavior is influenced not only by immune–stromal communication but also by metabolic and stress-response programs that shape cytokine networks and immune cell crosstalk [50–54]. Collectively, these findings emphasize that the TME is not a passive background but an active determinant of immune efficacy, orchestrated through cytokine gradients, myeloid–lymphoid crosstalk, and stromal signaling. Integrating such multi-dimensional TME features with molecular regulators like the MAD2L1–FOXM1 axis could therefore refine predictive models of immunotherapy response in LUAD.

The integration of SPRING mapping with GSCA and TIMER2.0 data advances our understanding of the immunological dimensions of MAD2L1 biology. One of the major challenges in LUAD is its pronounced tumor heterogeneity, this heterogeneity shapes treatment outcomes, influences resistance to therapy, and complicates biomarker development [55]. To better capture this complexity, we extended our analysis beyond MAD2L1 itself and explored its positively and negatively associated gene networks, reasoning that these co-regulated partners may provide a broader view of how MAD2L1 interacts with diverse immune contexts. The consistent inverse association between MAD2L1 (and its positive gene network) with CD4⁺ T cell infiltration may reflect an immune-exclusion phenotype. In contrast, the macrophage enrichment observed for the negative gene panel may indicate compensatory immune activity, although the functional state of these macrophages remains to be clarified. These findings align with previous literature that mitotic checkpoint dysregulation can modulate immune surveillance, either by promoting genomic instability- derived neoantigens or by fostering an immunosuppressive microenvironment [56, 57].

Clinically, these results suggest that targeting the MAD2L1–FOXM1 axis could have dual benefits: suppressing tumor proliferation while potentially restoring immune competence. Given the increasing emphasis on immunotherapy in LUAD, the combination of cell cycle inhibition with immune checkpoint blockade warrants investigation, particularly in patients exhibiting the high-MAD2L1/low-CD4⁺ T cell phenotype identified here.

Limitations

While this study offers comprehensive insights by integrating multiple transcriptomic platforms and immune infiltration analyses, certain aspects require further exploration. Correlation-based approaches provide strong indications of functional associations but cannot fully establish direct mechanistic relationships. Experimental validation through in vitro and in vivo models will therefore be valuable to determine whether MAD2L1 directly governs immune cell recruitment or acts via its regulatory network. In addition, CD4⁺ T cells encompass diverse subsets, including effector, helper, and regulatory populations, while macrophages span pro-inflammatory and immunosuppressive phenotypes. Future studies that dissect these functional subpopulations in the context of MAD2L1 signaling could provide a more granular understanding of how this axis shapes the tumor immune microenvironment.

Conclusion

In conclusion, our work positions MAD2L1 as a central node linking mitotic regulation and immune modulation in LUAD. By elucidating both the proliferative and immunological dimensions of MAD2L1 biology, we provide a rationale for its inclusion in prognostic assessment and as a potential target in combinatorial treatment strategies.

Author contributions

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

Funding

Open access funding provided by Manipal University Jaipur.

Data availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. The complete list of databases and workflows used for data retrieval, preprocessing, statistical analyses, and visualization, is available in our GitHub repository: https://github.com/chainseesaini27/MAD2L1-VS-IMMUNE-CELLS-LUAD/tree/main.

Declarations

Ethics approval and consent to participate

This study does not require ethical approval.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.