IGFBP1 facilitates tumor proliferation, invasion and migration by modulating the MAPK-ERK signaling pathway in lung adenocarcinoma

Abstract

Objective

Lung adenocarcinoma (LUAD) is the most prevalent histological subtype of lung cancer, and lymph node metastasis serves as a significant prognostic risk factor. The identification of molecular biomarkers associated with lymph node metastasis holds paramount importance in the prevention and treatment strategies for this condition.

Methods

Through bioinformatic analysis, we identified insulin-like growth factor binding protein 1 (IGFBP1) as the biomarker most strongly associated with lymph node metastasis. Subsequently, we performed proliferation, invasion, and migration assays using IGFBP1-overexpressing and knockdown cell lines. The single-cell-RNA-sequence data indicated that IGFBP1 facilitates the progression of LUAD cells through the MAPK signaling pathway. Subsequently, western blot analysis was performed to validate these findings, while the ERK inhibitor U0126 was employed for cellular experiments and in vivo verification to elucidate the precise biological function of IGFBP1.

Results

IGFBP1 was identified as the most significant biomarker for lymph node metastasis risk through immunohistochemical and Cox regression analyses. Genetic overexpression and knock-down of IGFBP1 demonstrated its critical role in regulating proliferation, migration, and invasion in vitro. These effects were mediated through ERK pathway activation, as confirmed by both Western blot analysis and pharmacological inhibition with U0126 in a subcutaneous tumor model.

Conclusions

IGFBP1 indicates the promotion of lymph node metastasis in LUAD by facilitating tumor proliferation, invasion, and migration through modulation of the MAPK-ERK signaling pathway. Our findings highlight IGFBP1 as a potential prognostic biomarker, though its therapeutic utility requires validation in preclinical models.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12885-025-15188-4.

Introduction

Lung cancer is a prevailing cause of morbidity and mortality worldwide [1]. In recent years, lung adenocarcinoma has emerged as the predominant histological subtype of lung cancer [2], characterized by its remarkable tumor heterogeneity [3]. Metastasis remains the primary driver of death in patients with lung adenocarcinoma (LUAD) [4]. Therefore, investigating the molecular mechanisms underlying LUAD metastasis holds paramount significance for effective prevention and treatment strategies.

Lymphatic metastases are consistently associated with a significant reduction in survival rates among patients diagnosed with LUAD [5]. For the majority of malignancies, lymphatic dissemination represents the predominant pathway for tumor cell metastasis [6–8]. Therefore, the presence of lymphatic metastases serves as a crucial indicator of tumor progression and signifies an exacerbation in tumor staging. Appropriate biomarkers should not only monitor disease progression and therapeutic response but also identify patients at high risk of recurrence. Currently, there is a scarcity of genetic markers associated with lymphatic metastases. Hence, it is imperative to urgently explore novel biomarkers for LUAD.

The gene encoding insulin-like growth factor binding protein 1 (IGFBP1) belongs to the IGFBP family and encodes a protein consisting of an N-terminal domain characteristic of IGFBPs and a thyroglobulin type-I domain [9–12]. This protein undergoes rapid clearance from the circulation, with its expression primarily regulated by insulin. The IGFBP1 family exerts significant control over the activity of insulin-like growth factors (IGFs) through strong binding interactions [13]. The IGFs serve as a multifunctional regulator of cell proliferation, promoting cellular differentiation and proliferation, facilitating individual growth and development, while also exerting an inhibitory effect on apoptosis [14]. The expression level of IGF-1 was significantly elevated in STAD, thereby facilitating the proliferation and invasion of tumor cells through activation of the JAK1/STAT3 signaling pathway [15, 16]. Recent in vivo studies have demonstrated that IGFBP1 functions as a pivotal regulator of IGF signaling, effectively modulating energy allocation from growth and development to prioritize metabolic processes crucial for survival during stressful conditions [17].

In this study, we conducted an analysis of TCGA-LUAD RNA-seq and scRNA-seq data to identify genes associated with lymph node metastasis in LUAD. Differential gene screening was performed between samples with and without lymph node metastasis, followed by the identification of prognostic genes using a LASSO regression model. Subsequently, IGFBP1 was identified as the gene exhibiting the highest risk for LUAD lymph node metastasis. Functional experiments in cells and in vivo experiments demonstrated that high expression of IGFBP1 plays a crucial role in promoting malignant progression of LUAD.

Materials and methods

Ethical statement

This study was conducted in accordance with the Declaration of Helsinki. Our study received approval from our Ethic Committee (Approval number: bc2022232). All animal procedures were approved by the IACUC at Tianjin Medical University Cancer Institute and Hospital, adhering to legal mandates and national guidelines for laboratory animal care and maintenance (Approval number: 2023078).

Data access

The bulk RNA-seq data of LUAD used in this study from The Cancer Genome Atlas (TCGA) database. The single-cell RNA-seq data of LUAD collected from Bischoff etc. research (10.24433/CO.0121060.v1.) [18]. We identified 522 samples of bulk RNA-seq data in this study, including 331 samples without lymphatic metastases (N0) and 187 samples with lymphatic metastases (N1, N2, N3). The single-cell dataset contained 8 samples, including 7 without lymphatic metastases (N0) and 1 with lymphatic metastases (N1). The specific methods and processes of this study are shown in Fig. 1.

Fig. 1.

Flowchart of the study

Differential expression genes (DEGs) with lymphatic metastases

The LUAD samples from the TCGA database were categorized into two groups based on the presence or absence of lymphatic metastasis. Subsequently, we employed the Limma package (version: 4.2.3) in R software to investigate differential mRNA expression. The differentially expressed genes (DEGs) between samples with and without lymphatic metastasis were identified using the Wilcoxon signed-rank test, considering an adjusted P value threshold < 0.05 and log2 (fold change) > 1 or log2 (fold change) < −1.

Protein-protein interaction (PPI) network construction

The PPI networks of genes associated with lymphatic metastasis were analyzed using the STRING database (string-db.org), which provides functional protein association networks. The identification of hub genes was performed using the node degree algorithm in Cytoscape software (version 3.9.1).

Enrichment analysis

The screened lymphatic metastasis-associated genes and IGFBP1-related DEGs were subjected to Gene Ontology (GO) enrichment analysis, which included the categories of biological process (BP), cellular component (CC), and molecular function (MF). Additionally, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis and GSEA software version 4.2.2 (www.gsea-msigdb.org/gsea/index.jsp) was employed to explore whether differentially regulated genes between these two groups were enriched in cancer-related biological pathways. Statistical significance was defined as a P value < 0.05.

Prognostic modeling

The LASSO regression method was employed to identify the most relevant genes among the twenty hub genes significantly associated with lymphatic metastasis [19]. The study identified seven genes associated with prognosis and subsequently constructed a seven-gene signature. Subsequently, the risk score for each patient was calculated based on the regression coefficients of the genes in the signature and their corresponding expression values. Risk scores were calculated using the following formula: Risk score = expression of Gene 1 ∗ β1 + expression of Gene 2 ∗ β2 + expression of Gene n ∗ βn.

Patients were stratified into high-risk and low-risk groups based on the median risk score. The overall survival (OS) analysis was performed using Kaplan-Meier estimation, followed by log-rank tests to assess statistical significance. The sensitivity and accuracy of the signature were validated through receiver operating characteristic (ROC) curves by the SurvivalROC package. Univariate and multivariate COX regression analyses were conducted separately, with corresponding results visually presented using forest plots.

Single-cell processing

To visualize the processed scRNA-seq data, we employed Seurat suite version 3.1 for cell clustering analysis. Initially, the filtered gene-barcode matrix obtained from Cell Ranger Count was imported into Seurat. Following quality control and data harmonization steps, cells were clustered using Uniform Manifold Approximation and Projection (UMAP). Differential expression analysis between two cell groups was performed using a Wilcoxon Rank Sum test implemented in the FindMarkers function of Seurat. For manual annotation of distinct clusters, we utilized the ‘singleR’ package.

Human samples

Human lung cancer tissues were collected from the our center, with approval obtained from the Institutional Review Board (IRB) of the same institute. Informed consent was acquired from all patients, and samples were anonymized prior to analysis.

Immunohistochemistry

The study encompassed 128 tumor specimens, comprising of 64 patients without lymphatic metastases (N0) and 64 samples with lymphatic metastases (N1-N3). Immunohistochemistry (IHC) was conducted using anti-IGFBP1 antibody (CST, #31025, 1:500) to examine the expression of IGFBP1. Staining intensity was assessed on a three-grade scale: negative or weakly positive (0), moderately positive (1), and strongly positive (2). Staining extent was categorized into four grades based on the percentage of stained tumor cells: grade 1 for < 25%, grade 2 for 25% to < 50%, grade 3 for 50% to < 75%, and grade 4 for ≥ 75%. The final score was determined by summing up the staining intensity and extent scores. For the statistical analyses, composite scores of ≥ 3 were designated as indicative of high expression, while scores < 3 were considered to represent low expression levels. Staining assessments were performed by two independent observers who remained blinded to the patient data.

Cells and animals

H1975, H1299, PC9 and A549 cells, obtained from ATCC within the past decade, were cultured in media recommended by ATCC supplemented with 10% FBS, 100U/mL penicillin, and 100 µg/mL streptomycin. All experiments were conducted within one month of thawing early-passage cells. Six-week-old female BALB/c nude mice were procured from Beijing Vital River Laboratory Animal Technology Co., Ltd.

We used euthanasia methods to ensure the animals were unconscious by applying 1–3% mixture of isoflurane and oxygen (Veterinary drug production number: 15125). During induction, the concentration was increased to 5%. After confirming the mice were completely anesthetized, then we conducted cervical dislocation to obtain the tumors.

Western blotting

The proteins were separated using 10% Bis-Tris gels and transferred onto PVDF membranes. Blots were blocked in TBS containing 5% milk or bovine serum albumin (BSA) and incubated overnight with antibodies against IGFBP1 (CST, #31025, 1:1000), Phospho-Erk1/2 (Thr202/Tyr204) (CST, #4370), Erk1/2 (CST, 137F5), and β-actin (ab8227, abcam). Subsequently, the blots were washed three times with TBS containing 1% Tween and exposed to the secondary antibody at room temperature for one hour. After three additional washes, the blots were incubated with a luminol-based substrate for HRP-catalyzed detection. Luminescence signals were captured on an Amersham Imager 600.

IGFBP1 overexpression and knockdown

The primers for overexpressing IGFBP1 were designed as follows: (F)-ATGTCAGAGGTCCCCGTTGCTC and (R)-TCAGTTTTGTACATTAAAATAT. Human IGFBP1 cDNA was amplified from Hep-G2 cells, and the resulting PCR fragments were ligated into the lentiviral shuttle vector pCCL.PPT.hPGK.IRE-S.eGFP/pre containing GFP. Lentivirus production was achieved by transfecting Phoenix-293 cells with packaging plasmids pMD2.BSBG, pMDLg/pRRE, and pRSV-REV. Lentivirus-infected cancer cells were directly utilized in various assays including proliferation assay, matrigel invasion assay, migration assay, and animal experiments.

Lentiviral packaging and construction of stable knockdown cell lines

To establish stable IGFBP1-knockdown cell lines, 293 T cells were used to package recombinant lentiviruses containing IGFBP1-targeting short hairpin RNAs (shRNAs) (IGFBP1-sh1, IGFBP1-sh2, IGFBP1-sh3), with a lentiviral vector carrying a non-specific sequence serving as the control. Target cells were incubated in serum-rich medium supplemented with viral particles and polybrene, followed by selection to obtain stably transfected cell clones, The target sequences used in the experiments are as follows: IGFBP1-SH1: 5′-CAGGAGAAGAAATTTCCAAAT-3′; IGFBP1-SH2: 5′-CCACCATTCACATTTGATGTA-3′; IGFBP1-SH3: 5′-CCTGCCAAACTGCAACAAGAA-3′”.

Brdu cell proliferation assays

For the cell proliferation assays, 1 × 10⁴ cells were cultured in 96-well plates for 12 h. Subsequently, 5-Bromo-2′-deoxyuridine was added following the manufacturer’s instructions of the Cell Proliferation ELISA kit (11647229001, Roche) and incubated for 8 h. The absorbance at 450 nm was measured using an enzyme-labeled instrument.

Cell invasion assay

The cell plate was supplemented with 500µL of culture medium prior to the transfer of Transwell inserts containing an 8 μm membrane (BD Biosciences) coated with 20% Matrigel (Corning). A total of 2 × 10⁵ cells suspended in serum-free medium were introduced into the upper chamber of the Transwell inserts. Following incubation at 37 °C for 24 h, the cells that had invaded through the membrane onto the lower surface were fixed using a solution of 4% paraformaldehyde and subsequently stained with a solution of 0.05% crystal violet. The invaded cells were then photographed and quantified.

Cell migration assay

A 500 µl volume of culture medium containing 10% fetal bovine serum (FBS) was added to each well prior to transferring the Transwell inserts. A total of 2 × 10⁵ cells suspended in RPMI 1640 medium containing 0.1% FBS were seeded into the upper chamber of the Transwell inserts. Following incubation at 37 °C and 5% CO2 for 24 h, the migrated cells on the lower surface of the membrane were fixed with a solution of 4% paraformaldehyde for 15 min, stained with a solution of crystal violet (0.05%) for 10 min, and subsequently washed three times with phosphate-buffered saline (PBS). Images were captured using an inverted microscope.

U0126 treatment

Twenty 6-week-old female BALB/c nude mice were randomly assigned to each treatment group prior to the initiation of dosing, with 5 mice per group. Subsequently, a mixture of 1× 10⁶ H1975 cells expressing either empty vector or IGFBP1 in Matrigel (at a volume ratio of 1:1; BD Biosciences) was injected subcutaneously into the flank of the mice. One week after tumor cell inoculation, ERK inhibitor U0126 was administered to the mice. Mice were treated with vehicle (DMSO) or U0126 (25µmol/kg) twice weekly via intraperitoneal injection until the end of the experiment. After six weeks from inoculation, euthanasia was performed on the mice and subcutaneous tumors were dissected, fixed in 4% paraformaldehyde, and embedded in paraffin. Paraffin blocks were sectioned into 5 μm slices for Hematoxylin-Eosin staining.

Statistical analysis

The means of two populations were compared using an unpaired 2-tailed Student’s t-test. Multiple comparisons were conducted with a 1-way ANOVA test. Kaplan-Meier analysis was employed to compare DFS and OS, and differences in survival were assessed using the log-rank test. A significance level of less than 0.05 was considered statistically significant for all tests.

Results

Revealing the transcriptomic disparities between lymph node metastasis and non-lymph node metastasis

Based on the TCGA database, we conducted a comprehensive analysis of genes exhibiting differential expression in LUAD with and without lymphatic metastasis. Our findings revealed 204 differentially expressed genes in LUAD, encompassing 86 upregulated genes and 118 downregulated genes, as visually depicted in the generated volcano plot (Fig. 2A). Furthermore, the differential gene expression heat map effectively illustrates the top 24 genes displaying significant alterations (Fig. 2B).

Fig. 2.

Identification of differentially expressed genes associated with lymph node metastasis, construction of a protein-protein interaction (PPI) hub gene network, and functional enrichment analysis were performed. A The volcano plot displays the differential expression profile of genes in the TCGA-LUAD dataset, revealing 86 upregulated genes and 118 downregulated genes that are linked to lymph node metastasis. B The heatmap illustrates the top 24 genes exhibiting significant differential alterations. C A PPI network was constructed using the identified 86 upregulated and 118 downregulated genes, resulting in the identification of twenty key hub genes. D Expression patterns of twenty hub genes were compared between patients with or without lymphatic metastasis. Functional enrichment analysis was conducted for Gene Ontology (E) and Kyoto Encyclopedia of Genes and Genomes pathways (F), based on these twenty hub genes

Construction of the PPI network and analysis of the expression patterns of the top twenty hub genes

After analyzing the differentially expressed genes in LUAD with and without lymphatic metastasis, we constructed a PPI network using 86 upregulated and 118 downregulated genes (Figure S1). Subsequently, twenty hub genes were identified (Fig. 2C): AGT, ALB, IL1A, CALB2, S100A8, S100A9, S100A12, CACNA1B, PCSK2, SYP, AREG, KRT1, KRT16, PLP1, SYNPR, IGFBP1, TPH1, CRHR2, TGFBI and COL11A1. We further examined the expression of these twenty genes in both groups with and without lymphatic metastasis. The results revealed significant differential expression between the two groups: IL1A, CALB2, S100A8, COL11A1, KRT16, AGT, AREG, IGFBP1, S100A9, S100A12 and TGFBI were upregulated in the group with lymphatic metastasis; while ALB, KRTI, CRHR2, PCSK2, CACNA1B PLP1, SYNPR, SYP and TPH1 were downregulated in the group without lymphatic metastasis (Fig. 2D).

Conducting GO and KEGG analyses

To elucidate the molecular mechanisms underlying lymphatic metastasis-related genes, we conducted an enrichment analysis on this set of twenty genes. The results of the gene ontology (GO) analysis (Fig. 2E) revealed their crucial involvement in regulating inflammatory responses, cell growth, neutrophil migration, apoptotic signaling pathways, wound healing, and MAP kinase activity regulation. Additionally, the KEGG analysis demonstrated their participation in key pathways such as MAPK signaling pathway, IL-17 signaling pathway, and AGE-RAGE signaling pathway (Fig. 2F).

Reconstruction and validation of a prognostic model

We conducted a LASSO logistic regression analysis (Fig. 3A-B) on the expression levels of twenty hub genes from the TCGA dataset. Subsequently, we identified seven key genes (KRT1, CALB2, CRHR2, IGFBP1, S100A12, TGFBI and SYP) to construct a prognostic model for patients with or without lymph node metastasis. By calculating genetic scores based on these seven genes, patients were categorized into high-risk and low-risk groups. The expression heat map in Fig. 3C illustrates the differential gene expression patterns between these two risk groups. Furthermore, univariate Cox regression analysis revealed significant associations between all seven predictive genes and patient prognosis. Notably, IGFBP1 emerged as the most significant gene associated with poor prognosis (HR = 1.246, P < 0.001), as depicted in Fig. 3D. Kaplan-Meier survival curves (Fig. 3E) demonstrated a significant disparity in survival status between high-risk patients and those in the low-risk group (P < 0.001). Additionally, the ROC curves of this model exhibited AUC values of 0.761, 0.743, and 0.733 at one, three, and five years respectively (Fig. 3F). Furthermore, multivariate Cox regression analysis confirmed that the riskScore served as an independent prognostic factor (Fig. 3G).

Fig. 3.

Construction of the prognostic risk model was performed. A LASSO analysis was conducted to identify relevant genes; (B) Risk scores were calculated and correlated with survival time and status using data from the TCGA dataset. The top panel displays a scatterplot illustrating the distribution of risk scores ranging from low to high, while the bottom panel presents a scatterplot depicting the relationship between survival time, survival status, and different risk score values for each sample. C A heat map visualizing gene expression patterns in the prognostic model. D Univariate Cox regression analysis revealed that IGFBP1 was identified as the most significant prognostic gene associated with high risk. E Kaplan-Meier survival curves were plotted to compare the outcomes of patients classified as high-risk and low-risk groups. F Receiver operating characteristic (ROC) curves were constructed to evaluate the predictive performance of this risk model for one year (AUC = 0.761), three years (AUC = 0.743), and five years (AUC = 0.733). G Multivariate Cox regression analysis demonstrated that the riskScore derived from this model independently served as a prognostic factor

Expression profiling of a prognostic model comprising seven genes in single-cell data

Based on the scRNA-seq data of Bischoff P et al., we obtained gene expression profiles of 26,643 cells from 8 primary LUAD samples for subsequent analysis. The cells were grouped based on their expression profiles using uniform manifold approximation and projection (UMAP) analysis, leading to the identification of 17 transcriptionally distinct cell clusters (Fig. 4A). Subsequently, the cell identity of each cluster was annotated by referencing a dataset from the Human Primary Cell Atlas (Fig. 4B). Furthermore, we conducted a detailed analysis of specific cell populations with (meta_yes) and without (meta_no) lymph node metastasis, revealing that cells exhibiting lymph node metastasis were predominantly localized within the epithelial cell cluster (Fig. 4C). Subsequently, we examined the expression patterns of the seven genes comprising the prognostic model at single-cell resolution. IGFBP1 and CALB2 exhibited predominant expression in epithelial cells (Fig. 4D-E), while S100A12 showed preferential expression in monocytes (Fig. 4F). TGFBI and SYP displayed widespread expression across almost all cell types (Fig. 4G-H). KRT1 and CRHR2 exhibited relatively lower levels of expression in cells (Fig. 4I-J).

Fig. 4.

The validation of hub genes was performed in a single-cell RNA-seq dataset. A UMAP plots were generated to visualize the transcriptomic clusters in 26,643 high-quality cells. B UMAP plots were used for cell type classification in the same set of cells. C UMAP plots demonstrated the presence or absence of lymph node metastasis in these cells, with meta_yes and meta_no groups shown separately. D-J Expression levels of IGFBP1, CALB2, S100A12, TGFBI, SYP, KRT1, and CRHR2 were analyzed using single-cell RNA-seq data. K-L Functional enrichment analysis (GO and KEGG, respectively) revealed that IGFBP1 expression is associated with distinct pathways in tumor samples

GO and Kyoto KEGG enrichment analysis was performed to investigate the expression of IGFBP1 in single-cell data

Subsequently, we conducted a comprehensive examination of the distribution patterns exhibited by the seven genes comprising our prognostic model at the single-cell level. Notably, IGFBP1 emerged as the most prominent risk-associated gene within this model. To gain insights into the unique molecular characteristics of IGFBP1-expressing cells in tumor samples, distinct marker genes were identified using FindAllMarkers. Subsequent enrichment analysis revealed that these marker genes were significantly associated with various biological functions including regulation of endothelial cell migration, control of cell growth, positive modulation of wound healing processes, as well as involvement in both PI3K-Akt and MAPK signaling pathways (Fig. 4K).

The expression of IGFBP1 in clinical specimens and its impact on prognosis

To validate the expression of IGFBP1 in tumor tissues from patients with and without lymph node metastasis, we collected tumor samples from 128 patients with primary LUAD at our institution for immunohistochemical staining of IGFBP1 (Fig. 5A-B). The results indicate a significantly higher expression of IGFBP1 in patients with lymph node metastasis compared to those without (Fig. 5D). Additionally, there was a significant increase in IGFBP1 expression levels with increasing T stage and TNM stage (Fig. 5C, E). Kaplan-Meier survival curves demonstrate that high expression of IGFBP1 is associated with worse disease-free survival (HR = 1.562, P = 0.016) and overall survival (HR = 2.230, P = 0.004) rates among patients (Fig. 5F-G). Univariate and multivariate Cox regression analyses revealed that IGFBP1 expression serves as an independent prognostic factor for both disease-free survival and overall survival outcomes (Fig. 5H-I). Further subgroup analyses focusing on lymph node metastasis confirmed IGFBP1 as a critical determinant of metastatic propensity and an independent predictor of lymphatic dissemination.

Fig. 5.

Immunohistochemistry verified the expression of IGFBP1 in LUAD and its relationship with prognosis. A-B IGFBP1 is highly expressed in LUAD with lymph node metastasis (A) and low expressed in tissues without lymph node metastasis (B). C-E Expression of IGFBP1 in T staging (C), N staging (D) and TNM staging (E). F-G Kaplan-Meier survival curves of IGFBP1 expression, DFS (F) and OS (G). H-I Univariate and multivariate Cox regression analysis of DFS (H) and OS (I). *P < 0.5; **P < 0.01; ***P < 0.001; ****P < 0.0001

The expression of IGFBP1 exerted a significant impact on the malignant progression in LUAD cells

Subsequently, we ectopically expressed IGFBP1 in human lung adenocarcinoma H1975 and H1299 cells to investigate the impact of IGFBP1 (Fig. 6A, E). Our findings revealed that overexpression of IGFBP1 significantly enhanced cell proliferation, as assessed by bromodeoxyuridine (BrdU) incorporation analysis in both H1975 and H1299 cells (Fig. 6B, F). Furthermore, elevated expression of IGFBP1 exhibited a substantial increase in cell invasion and migration capabilities in both H1975 and H1299 cells (Fig. 6C, G-H).

Fig. 6.

Biological function of IGFBP1 in LUAD cells. A Western blot of IGFBP1 over expression in H1975 cell. B Cell proliferation assay of H1975 cell. C Cell invasion assay of H1975 cell. D Cell migration assay of H1975 cell. E Western blot of IGFBP1 over expression in H1299 cell. F Cell proliferation assay of H1299 cell. G Cell invasion assay of H1299 cell. H Cell migration assay of H1299 cell. I KEGG analysis revealed that IGFBP1 expression is involved in MAPK signaling pathways in the tumor samples. J GSEA analysis revealed that IGFBP1 expression is involved in cell proliferation and migration in the tumor samples. K Western blot revealed that IGFBP1 overexpression could significantly increase p-ERK expression level in H1299 and H1975 cells. ****P < 0.0001

Based on IGFBP1 expression profiling across lung adenocarcinoma cell lines, we selected PC9 and A549 for subsequent knockdown experiments (Figure S2A). Quantitative densitometry demonstrated comparable knockdown efficiencies in both cell models, with residual IGFBP1 protein levels reduced to 31% (sh2) and 37% (sh3) in PC9 cells, and 26% (sh2) and 47% (sh3) in A549 cells relative to controls (Figure S2B). IGFBP1 knockdown significantly reduced p-ERK expression in both PC9 and A549 cell lines (Figure S2C) and markedly suppressed BrdU incorporation (Figure S2D). Furthermore, IGFBP1 knockdown significantly inhibited the invasion and migration of PC9 and A549 cells (Figure S2E-H). Collectively, these results provide compelling evidence supporting the role of IGFBP1 in promoting tumor progression in LUAD cells.

IGFBP1 induces activation of the MAPK-ERK signaling pathway, thereby influencing malignancy in LUAD cells

In order to elucidate the impact of IGFBP1 on the malignant progression of LUAD cells, we conducted KEGG and GSEA enrichment analyses using TCGA-LUAD RNA-seq data to investigate the expression of IGFBP1. The results of our analysis revealed that IGFBP1 expression primarily influenced the MAPK signaling pathway, positively regulated wound healing, and promoted cell growth (Fig. 6I), which is consistent with the biological characteristics observed in previous scRNA-seq data (Fig. 4K). Western blot analysis further confirmed that overexpression of IGFBP1 in H1299 and H1975 cells significantly upregulated p-ERK expression levels (Fig. 6K). Concomitantly, the knockdown of IGFBP1 in PC9 and A549 cell lines demonstrated a marked downregulation in phosphorylated ERK (p-ERK) expression levels (Figure S2F).

The role of U0126, an ERK inhibitor, in the malignant progression of LUAD cells under IGFBP1 expression was investigated through inhibition experiments using H1975 and H1299 cells. In comparison to the control group, U0126 significantly suppressed p-ERK levels during high expression of IGFBP1 (Fig. 7A). Cell proliferation assays demonstrated that U0126 effectively inhibited the proliferation of H19751 and H1299 cells following IGFBP1 overexpression, while no significant changes were observed in the control group (Fig. 7B). Furthermore, cell invasion and migration experiments revealed that U0126 markedly hindered invasion and migration after IGFBP1 overexpression, whereas no notable alterations were observed in the control group (Fig. 7D). These findings suggest that elevated expression of IGFBP1 can influence the malignant progression of LUAD cells via modulation of the MAPK-ERK signaling pathway.

Fig. 7.

IGFBP1 enhances the proliferation, invasion, and migration of LUAD cells by modulating the MAPK-ERK signaling pathway. A Western blot analysis was performed by treating H1975 and H1299 cells with an ERK inhibitor (U0126). B-C 8\+The Brdu uptake assay revealed that U0126 significantly suppressed the proliferation of IGFBP1 high-expressing tumors. D-E The invasion and migration assay demonstrated that U0126 significantly suppressed the invasive and migratory abilities in high-IGFBP1 group. F Subcutaneous tumor formation in nude mice revealed that U0126 significantly inhibited the cell proliferation of IGFBP1 high expression tumors in vivo. G Weight histogram of subcutaneous tumor. H HE staining of subcutaneous tumors revealed that U0126 could significantly inhibit the invasion of IGFBP1 high expression tumors in vivo. **P < 0.01; ***P < 0.001; ****P < 0.0001. Bars, 100 μm

Vivo experiments substantiated the role of IGFBP1 in promoting the progression of LUAD cells through activation of the ERK pathway

The results of the H1975 subcutaneous tumor inoculation experiment in nude mice indicated that the overexpression of IGFBP1 led to a significantly larger growth of subcutaneous tumors compared to the control group. Additionally, treatment with U0126 effectively suppressed the proliferation of subcutaneous tumors in IGFBP1-overexpressing mice, while no significant change was observed in tumor volume within the control group (Fig. 7F-G). Histological examination using HE staining demonstrated that U0126 could notably inhibit tumor invasion (Fig. 7H). These findings, combined with in vitro cell inhibition assays, suggest that IGFBP1 promotes lung adenocarcinoma progression by activating the MAPK-ERK pathway.

Discussion

The metastasis of lung adenocarcinoma typically involves hematogenous and lymphatic dissemination, which is indicative of an advanced stage and unfavorable prognosis [20–22]. During the process of metastasis, the regional lymph nodes in close proximity to the primary tumor commonly serve as the predominant site for dissemination of cancer cells [23]. In addition, the presence of lymphatic metastasis is directly correlated with distant recurrence and overall survival [24, 25]. Therefore, the identification of tumor markers associated with lymph node metastasis holds great significance for predicting the prognosis of patients with LUAD.

In this study, we conducted a comprehensive analysis of the TCGA-LUAD dataset to identify differential expression genes associated with lymph node metastasis. Through LASSO regression, we identified a set of 7 prognostic model genes that exhibited the highest correlation with lymph node metastasis, including IGFBP1, S100A12, CALB2, TGFBI, KRT1, CRHR2 and SYP. Previous studies have demonstrated that IGFBP1 facilitates cell survival during confined migration by augmenting the detoxification of mitochondrial reactive oxygen species (ROS), thereby facilitating tumor metastasis [26]. The proteins S100A12 and CALB2 have been identified as potential biomarkers for metastatic prostate cancer with a particularly unfavorable prognosis [27]. Additionally, CALB2 has been found to promote tumor progression by influencing tumor cell adhesion [28]. Analysis of extracellular matrix (ECM) changes during neoplastic transformation reveals the involvement of macrophage-secreted transforming growth factor beta-induced (TGFBI) protein in early ovarian cancer immunosuppression [29]. Silencing of keratin 1 (KRT1) enhances myocardial ischemia and reperfusion injury (MIRI) by activating the Notch signaling pathway in mice [30] while targeted inhibition of KRT1 activates the Notch pathway, thereby protecting against coronary atherosclerosis [31]. Reduced expression of corticotropin-releasing hormone receptor 2 (CRHR2) promotes tumor proliferation and metastasis through IL-6/Stat3 signaling in colon cancer [32]. Synaptophysin (SYP) plays a crucial role in neuroendocrine tumors [33]. These seven hub genes play a pivotal role in the development of prognostic models. The robustness of the model is confirmed by the ROC curve, demonstrating its substantial predictive value.

The expression of hub genes in single-cell data suggests their significant role in tumor development. Integration of single-cell data expression and TCGA-LUAD RNA-seq data, followed by GO and KEGG enrichment analysis, indicates that these hub genes primarily contribute to cell proliferation, invasion, migration, and are closely associated with the MAPK signaling pathway. These findings underscore the robust predictive value of the hub genes in tumor metastasis. Moreover, we conducted in vivo and in vitro experiments targeting IGFBP1, which exhibited the highest risk factor among all screened genes, to elucidate its specific involvement in LUAD malignant progression particularly related to metastasis.

Insulin-like growth factor binding protein-1 (IGFBP-1), a secreted protein, is implicated in various cellular processes such as proliferation, migration, invasion, adhesion, and survival [10]. In this study, utilizing TCGA-LUAD data and single-cell data, we observed a predominant expression of IGFBP1 in tumor tissues with lymph node metastasis. Subsequent in vitro and in vivo experiments validated its overexpression as a promoter of LUAD cell proliferation, consistent with previous literature findings [34, 35]. However, a study on hepatocellular carcinoma demonstrated the inhibitory effect of IGFBP1 on tumor cell proliferation [36], suggesting that IGFBP1 may play an opposite role in tumor proliferation through different pathways. In LUAD cells and mouse subcutaneous tumor formation experiments, overexpression of IGFBP1 significantly enhances invasion and migration, suggesting a substantial impact on the promotion of malignant progression in LUAD. This finding aligns with the results obtained from the prognosis model constructed using lymph node-related genes, indicating that IGFBP1 can serve as a potential marker for metastasis in LUAD. Furthermore, we investigated the underlying pathways influencing IGFBP1 expression. Both TCGA-LUAD data and single-cell RNA-seq results revealed an association between IGFBP1 expression and the MAPK signaling pathway. Subsequently, through in vitro and in vivo experiments, we demonstrated that targeting the MAPK-ERK signaling pathway effectively suppresses proliferation, invasion, and migration of LUAD cells. But we observed that the inhibitory effect of U0126 on the MAPK-ERK pathway was not evident in wild-type H1299 and H1975 cell lines. Combined with Western blot results showing differential IGFBP1 expression levels across eight lung cancer cell lines, we hypothesize that the lack of pathway inhibition in these two cell lines may be attributed to their relatively low basal IGFBP1 expression, which likely prevents significant activation of the MAPK-ERK pathway. Consequently, U0126 exhibited no efficacy in wild-type cells. Additionally, previous literature indicates that U0126 treatment significantly downregulates IGFBP1 expression [37]. However, due to time constraints, the precise mechanistic interplay between U0126 and IGFBP1 remains to be fully elucidated in future studies. Collectively, these findings suggest that IGFBP1 promotes malignant progression of LUAD by modulating the MAPK-ERK signaling pathway.

However, it is important to acknowledge certain limitations in this study. While pharmacological inhibition of ERK attenuated IGFBP1-driven tumor progression, further studies such as rescue experiments with constitutively active ERK are needed to establish direct causality. The analysis was based on TCGA-LUAD bulk RNA-seq data limited in vivo or in vitro experiments. Furthermore, no further investigation was conducted regarding the impact of IGFBP1 on upstream and downstream genes in the mechanistic study. Therefore, additional experimental verification is warranted to complement these findings.

Conclusions

In conclusion, we demonstrate significant prognostic value in constructing a correlation model for lymph node metastasis. As a biomarker associated with metastasis, IGFBP1 promotes proliferation, invasion, and migration through the MAPK-ERK signaling pathway in LUAD. Targeting the MAPK-ERK signaling pathway effectively inhibits the proliferation, invasion, and migration of LUAD cells. These findings offer a novel therapeutic target for preventing and treating lymph node metastasis in LUAD.

Author contributions

Mengzhe Zhang, Yiying Di, Zhenfa Zhang designed the manuscript. Zengtuan Xiao and Mengzhe Zhang, Zuo Liu, Meng Yan performed the experiment, wrote and completed the manuscript. All authors completed the data download and preprocessing and data analysis. Zengtuan Xiao, Mengzhe Zhang provide experimental platforms and resources, and make corrections to manuscripts. All the authors approved the final manuscript.

Funding

No specific grants were received for this research from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

Dataset analyzed during the current study (TCGA-LUSC dataset) were previously generated and are available in the TCGA repository (https://portal.gdc.cancer.gov/). The scRNA-seq data are available in *Bischoff etc.* research ([10.24433/CO.0121060.v1])([10.24433/CO.0121060.v1]). The original contributions presented in the study are included in the article/Supplementary Materials. Data request please contact corresponding Zengtuan Xiao by emails: zlyyfbzlkxzt@tmu.edu.cn

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki. Our study received approval from the Ethic Committee of Tianjin Medical University Cancer Institute and Hospital (Approval number: bc2022232). All animal procedures were approved by the IACUC at Tianjin Medical University Cancer Institute and Hospital, adhering to legal mandates and national guidelines for laboratory animal care and maintenance (Approval number: 2023078).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.