Introduction

Lung cancer remains the most common cause of cancer death in China and worldwide despite widespread application of low-dose spiral computed tomography in lung cancer screening as well as continuous progress in chemotherapy, immunotherapy, targeted therapy, surgery and precision radiotherapy1. In recent decades, the incidence of lung adenocarcinoma (LUAD) has been gradually increasing, making it the most common pathological subtype of lung cancer, surpassing lung squamous carcinoma. Previous studies have identified some LUAD -associated molecular alterations, such as epidermal growth factor receptor (EGFR) gene mutations2,3 anaplastic lymphoma kinase (ALK)4, c-ros oncogene 1 receptor tyrosine kinase (ROS1)5 and proto-oncogene tyrosine-protein kinase receptor (RET)6 gene fusions. EGFR is a transmembrane glycoprotein that belongs to the protein kinase superfamily, serving as a central role for many pathways and functioning as a receptor for members of the epidermal growth factor family. Upon ligand binding, EGFR forms homodimers or heterodimers, leading to autophosphorylation and cross-phosphorylation at various tyrosine kinase residues. This process allows for the docking of multiple proteins, including Grb2, SHc-transforming protein and PLC-γ7, which contain the Src homology 2 (SH2) domains. EGFR transduces essential growth factor signalling from the extracellular environment to the cell via multiple processes. Since over 60% of non-small cell lung cancers (NSCLCs) express EGFR, it has emerged as a significant therapeutic target for treating LUAD. The classical activating mutations, including exon 19 deletions and the L858R point mutation, make up the majority of EGFR mutations and are well-established as strong predictors of a positive clinical response to EGFR-tyrosine kinase inhibitors (TKIs).

The development of high-throughput gene expression analysis platforms, such as gene microarrays, has provided a foundation for comprehensive mining, exploration and research of LUAD through bioinformatics. This approach is beneficial for molecular diagnosis, classification and prognosis prediction, as well as for the identification of new therapeutic targets and the prediction of treatment response in LUAD. With the help of sequencing technology, tumour heterogeneity in LUAD can be understood more clearly. Several gene expression profiling studies on LUAD carcinogenesis have been performed in recent years using microarray technology and showed hundreds of differentially expressed genes (DEGs) involved in different pathways, biological processes and molecular functions8. The application of microarray technology and bioinformatics analysis has enabled the analysis of biomarker profile in the development of LUAD9. Furthermore, the interactions among DEGs, particularly the pathways in the interaction network, require further clarification. In this study, we identified that caveolin-1 (Cav-1) facilitated the proliferation and invasion of LUAD by promoting EGFR degradation and regulating the phosphorylation of the downstream AKT/STAT3 pathway. As a result, Cav-1 may serve as a promising therapeutic target for the treatment of LUAD patients.

Methods

Datasets for the bioinformatics analysis

We obtained gene expression data of both LUAD and normal adjacent lung tissues from GSE130779, GSE32863 and GSE85841 datasets. GSE130779 dataset included mRNA expression data of the tumours and non-tumour adjacent tissues from eight patients with LUAD. GSE32863 dataset included mRNA expression data from 58 LUAD and 58 adjacent non-tumour lung fresh frozen tissues. For GSE85841 include 8 lung cancer and adjacent non-tumour tissues.

Data processing of DEGs

DEGs between the LUAD and the normal lung specimens were identified using the GEO2R web tool with |logFC| > 2 and an adjusted P value < 0.0510. We identified the common DEGs among the three datasets by Venn software.

Gene ontology and pathway enrichment analysis

The effective clustering of functional genes is mainly based on gene ontology (GO)11 and KEGG pathways12. The Database for Annotation, Visualisation and Integrated Discovery (DAVID)13 was used to analyse the DEGs enrichment of biological processes (BP), molecular functions (MF), cellular components (CC) and pathways.

Protein-protein interaction (PPI) network and module analysis

The goal of the STRING database is to gather and merge all publicly available sources of PPI information14. Cytoscape software was applied to examine the potential correlation between these DEGs (maximum number of interactors = 0 and confidence score ≥ 0.415). Furthermore, the Molecular Complex Detection (MCODE) plugin was employed to identify hub modules within the PPI network, using degree cutoff = 2, max depth = 100, k-core = 2 and node score cutoff = 0.2.

Kaplan-Meier survival analysis

The prognostic significance of the genes in LUAD was evaluated using Kaplan-Meier plotter, an online database that includes gene expression data and clinical survival data16. To investigate the expression profiles of these DEGs between cancer patients and healthy cohorts, the GEPIA webserver was used to analyse data from thousands of samples included in the TCGA and GTEx projects17.

Identification and enrichment analysis of CAV1-related DEGs in GSE32863

In the GSE32863 dataset, we used the median expression value of CAV1 as a threshold to divide the patients into high expression and low expression groups. To perform functional enrichment analysis and visualisation, we used several R packages including ‘clusterProfiler’, ‘org.hs.eg.db’, ‘GOplot’, ‘ReactomePA’ and ‘enrichplot’. Pearson correlation analysis was applied to investigate the correlation between the expression of CAV1 and MKI67.

Immunohistochemistry (IHC)

Surgically excised LUADs were fixed in 4% paraformaldehyde, dehydrated and embedded in paraffin. After deparaffinisation and rehydration, the slides were retrieved for 10 min at 100 °C, followed by a hydrogen peroxide and goat serum blocking procedure. Primary antibodies against CAV1 and EGFR were purchased from Cell Signalling Technology, and a MaxVisionTM HRP-polymer anti-mouse IHC kit was applied as a secondary antibody. The dilution ratios of the primary antibodies are as follows: caveolin-1 monoclonal antibody (mAb) at 1:1000 and EGFR mAb at 1:50. In the chip experiment, the dilution ratios were adjusted to caveolin-1 mAb at 1:2000 and EGFR mAb at 1:100. For the subcutaneous tumor formation experiment in nude mice, the dilution ratio for caveolin-1 mAb was set to 1:1600. Additionally, antibodies targeting EGFR and Ki-67 were selected from the Maixin immunohistochemical antibody series. The signal was then detected using a DAB kit. Image J software was used to quantify the IHC results. The staining intensity of each protein by IHC was scored as 1 (negative), 2 (weak positive), 3 (positive) or 4 (strong positive) and the average optical density values were measured.

Ethical approval

The protocol for this study was approved by the Ethics Review Committee of the First Affiliated Hospital of Anhui Medical University (PJ2023-08-60) according to approaches to ethical review of life sciences and medical research involving humans.

CAV1 and EGFR expression in LUAD and paracancerous tissue

Ten patients with LUAD were randomly selected from Thoracic Surgery Department of the First Affiliated Hospital of Anhui Medical University. The general information of patients, tumor size, lymph node metastasis, postoperative stage of LUAD and other clinical information were collected. The postoperative specimens were collected and divided into cancer tissues and adjacent tissues, and stored in the environment of − 80 °C. The tissue was ground using a sample grinder (Jingxin Technology). The supernatants containing proteins were collected using 4 °C centrifuge at 13,300 rpm for 10 min. Then, western blotting was performed to detect CAV1 and EGFR.

Tissue microarray

Tissue microarray was used to verify the expression of CAV1 and EGFR in LUAD and adjacent tissues, and to analyze the clinical prognosis of cohort study. LUAD chips (HLugA180Su11-52944 and 52945) were obtained from Shanghai Xinchao Biotechnology Co., LTD. The clinical data included age, gender, TNM stage, pathological results and grade, etc. An 8-year survival follow-up and cohort study were conducted to collect the survival data of the patients. The immunohistochemical staining intensity score were interpreted as follows: 0 (negative), 1 (weakly positive), 2 (positive), and 3 (strongly positive). The scores of positive staining area were listed as follows: 0 (negative), 1 (1–25%), 2 (26-50%), 3 (51–75%), 4 (76-100%). The sum of staining intensity score and staining positive rate score was used as the total score.

Cell culture

Human LUAD cell lines, including A549, H1299, H1975 and PC-9, were cultured in RPMI 1640 medium supplemented with L-glutamine (CORNING), 10% FBS (Excell Bio), and 1% penicillin/streptomycin (HyClone). The cells were incubated in a humidified atmosphere containing 5% CO2 at 37 °C.

SiRNA interference

The CAV1 siRNA and negative control (NC) siRNA were pre-designed and purchased from Genepharma. The siRNAs were transfected using Lipofectamine 2000 transfection reagent according to the manufacturer’s instructions.

Plasmid transfection

The CAV1-overexpressing Pcmv-3flag plasmids were successfully constructed. Cells were cultured in 6-well plates and transfected with either control vector plasmids or CAV1 overexpression plasmids using HighGene transfection reagent according to the manufacturer’s instructions.

Western blotting

Cells were lysed in a lysis buffer containing protease inhibitors (Sparkjade), followed by centrifugation at 13,300 rpm for 10 min at 4 °C. The total protein was collected and stored at − 80 °C until use. Protein concentrations were determined using the BCA Protein Assay Kit (Beyotime Biotechnology). Equal amounts of proteins were separated by SDS-PAGE and then transferred to PVDF membranes. The primary antibodies used for immunoblotting were anti-Cav-1, anti-EGFR, anti-AKT, anti-p-AKT (Ser473), anti-STAT3, anti-p-STAT3 (Tyr705), anti-Bax, anti-Caspase-3, anti-Bcl-2 (Cell Signaling TECHNOLOGY), anti-Flag (Sigma) and anti-β-actin (TransGen Biotech).

Detection of EGFR degradation

The degradation of EGFR protein in LUAD cell lines was determined by western blotting after treatment with cycloheximide (CHX) at successive time points. Both transfected and non-transfected cells were treated with 50 µg/mL CHX. The time points examined in this study were 0 h, 0.5 h, 1 h, 2 h, 3 h and 4 h.

qRT-PCR

Total RNA from cells was isolated using the RNA Easy Fast Tissue/Cell Kit (TIANGEN). cDNA was made using TransScript First-Strand cDNA Synthesis SuperMix. Real-time qPCR was performed utilising the LightCycler 96 Real-Time PCR System. The primers were listed as follows: CAV1 forward, 5’-ATTGACTTTGAAGATGTGATTGCAG-3’ and reverse, 5’- AAGAGAGAATGGCGAAGTAAATGCC-3’; EGFR forward, 5’-CCAGTGACTGCTGCCACAACCAGTG-3’ and reverse, 5’-TACGGGGACACTTCTTCACGCAG-3’.

CCK8 assay

Following transfection with siRNA or plasmid, cells were seeded into 96-well plates and cultured for 24 h, 48 h, 72 h, 96 h and 120 h. Next, 10µL of CCK8 solution (Biosharp® Life Sciences) was added to each well, and the cells were incubated for 1.5 h. Absorbance was measured at 450 nm using the SpectraMax® Absorbance Reader CMax Plus.

Colony formation assay

After transfecting with siRNA or plasmid, cells were seeded in 6-well plates at an equal amount of 500 and cultured for 10–14 days until colony formation. Following methanol fixation, crystal violet staining and washing, the colonies were photographed and counted.

Wound healing assay

The cells were cultured in 6-well plates and allowed to grow to confluency. Once the cells reached almost 100% confluency, the monolayer of cells was scratched using a small pipette tip, followed by culturing with serum-free medium. The scratched area was then photographed at 0, 12, 24 and 48 h, respectively.

Transwell assay

The transwell assay was performed using Matrigel matrix and chambers from Corning. The transfected cells were seeded in the upper chamber using medium without serum, while the lower chamber was filled with medium containing 10% FBS. After an overnight incubation period, the non-invading cells were removed. The cells that had successfully invaded through the Matrigel matrix membrane in the chambers were then stained and photographed.

Apoptosis assay

The transfected cells were cultured in a 6-well plate for 48 h. Subsequently, the cells were harvested, washed with PBS and then resuspended in 500 µL of binding buffer. Apoptosis of the cells was measured by staining with 5 µL of propidium iodide (PI) and 5 µL of FITC-anti-Annexin V antibody (BD Biosciences), followed by analysis using a Beckman Coulter flow cytometer.

ShRNA lentivirus infection and construction of stable strain

The sequence of shNC shRNA was as follows: 5’-TTCTCCGAACGTGTCACGT-3’. The sequence of shCAV1-1 shRNA was as follows: 5’-CCTTCACTGTGACGAAATA-3’. The sequence of shCAV1-2 shRNA was 5’-AGACTCGGAGGGACATCTCTA-3’. The lentiviral shRNA plasmid was supplied by Genepharma. PC-9 cell lines stably transformed with resistant GFP plasmid were chosen by puromycin gradient concentration selection. PC-9 cells were seeded at the number of 1.9 × 106 cells per well. After 24 h, 900 µL of complete medium mixed with 100 µL of lentivirus solution carrying the corresponding shRNA GFP plasmid and 2 µL of Polybrene were added respectively. The cells were cultured in medium containing 600ng/mL puromycin.

Subcutaneous tumor formation experiment in nude mice

In vivo experiments were performed according to the protocol approved by the Experimental Animal Ethics Committee of Anhui Medical University (LLSC20232228). The study was reported in accordance with ARRIVE guidelines (https://arriveguidelines.org). Four-week-old SPF nude mice were obtained from Beijing Weitong Lihua Laboratory Animal Technology Co., LTD. The mice were randomly divided into 3 groups (n = 10 for each group). PC-9 cells carrying shNC, shCAV1-1 and shCAV-2 were cultured to the required number (2 × 106 cells per mouse). The cell suspensions were mixed with matrigel at a 2:1 volume ratio, and inoculated subcutaneously into the back of nude mice in corresponding groups. Mice were weighed every 2–3 days and tumor volume was estimated by (long diameter * square of short diameter) /2 in 4 weeks. After weighed, the tumors were fixed in 4% paraformaldehyde for 24 h. The tumor tissues were dehydrated, embedded and sectioned, then HE staining and immunohistochemical staining were performed. For detection of EGFR and Ki-67, Maixin antibody was used. Two pathologists scored and interpreted the results. The percentage of positive cells was scored as follows: < 10% received 0 points, 10–24% received 1 point, 25–49% received 2 points, and 50–100% received 3 points. Dyeing intensity: unstained is 0 points, light yellow is 1 point, light brown is 2 points, light tan is 3 points, and dark brown is 4 points. The final scoring system was determined by summing the scores for both parameters: a total score of 0 was considered negative, 1–2 was considered weakly positive, 3–4 was considered moderately positive, and 5–7 was considered strongly positive. Pathologists assessed the Ki-67 positivity of tissue by counting 500–2000 tumor cells by examining the area of the tissue with the highest expression.

Statistical analysis

Mean ± standard deviation was used to represent normally distributed data. Student’s t-test was employed to compare data between two groups, while analysis of variance was used to compare data among multiple groups. In the cohort study, the chi-square test was used to compare the count data, categorical variables, and ordinal data between groups. Kaplan-Meier survival analysis and cox proportional hazard model regression analysis were performed combined with multiple grouping variables. SPSS 26.0 and Graphpad Prism 9 software were used for statistical analysis. Statistical significance was considered to be indicated by P < 0.05.

Results

Mapping of prognostic core genes in LUAD

In the present study, 74 LUAD tissues and 74 non-tumour adjacent tissues were included. We identified 2122, 188 and 2178 DEGs from GSE130779, GSE32863 and GSE85841, respectively. Subsequently, we used online Venn diagram software to identify commonly DEGs in the three datasets, resulting in the detection of 89 common DEGs, including 86 down-regulated genes (logFC < 0) and 3 up-regulated genes (logFC > 0) in LUAD tissues (Table 1; Fig. 1 A and B). All the 86 down-regulated DEGs were analysed by DAVID software and the results of GO analysis indicated that (1) for biological processes (BP), down-regulated DEGs were enriched in hydrogen peroxide catabolic process, response to hypoxia, ethanol oxidation, heart trabecula formation, negative regulation of angiogenesis, negative regulation of transforming growth factor beta receptor signalling pathway; (2) for molecular function (MF), down-regulated DEGs were particularly enriched in heparin binding, peroxidase activity, carbohydrate binding, calcium ion binding, haptoglobin binding, oxygen binding; (3) for cell component (CC), down-regulated DEGs were significantly enriched in extracellular space and matrix, proteinaceous extracellular matrix, membrane raft, endocytic vesicle membrane (Table 2). The KEGG analysis results are listed in Table 3, which showed that the down-regulated DEGs were significantly enriched in the malaria pathway and proteoglycans in the cancer pathway. CAV2, CAV1, DCN and TIMP3 were significantly enriched in the sulfate/ sulfate/dermatan sulfate proteoglycan pathway (Fig. 4A). CAV1 and CAV2 may inhibit tumour growth by degrading EGFR. A total of 86 down-regulated and 3 up-regulated DEGs were entered into the DEGs PPI network complex, which consisted of 89 nodes and 77 edges (Fig. 1C). Upon performing module analysis using Cytoscape software and MCODE (with a degree cutoff of 2, node score cutoff of 0.2, k-core of 2 and a maximum depth of 100), 18 central nodes were identified (namely, SCGB1A1, CLDN18, SFTPA1, SFTPC, PDK4, LPL, CD36, FABP4, ALDH1A1, SDPR, ADH1B, ADH1A, CAV1, CAV2, EDNRB, CXCL2, HBEGF and GNG11).

Table 1 In total 89 DEGs were detected from three profile datasets, including 86 down-regulated genes and 3 up-regulated genes in LUAD tissues compared with adjacent lung tissues.
Full size table
Fig. 1
figure 1

Verification of 89 common DEGs in the three datasets (GSE130779, GSE32863 and GSE85841) through online Venn diagram software. (A) Three DEGs were up-regulated in the three datasets (logFC > 0). (B) Eighty-six DEGs were down-regulated in three datasets (logFC < 0). (C) Common DEGs PPI network constructed by STRING online database and Module analysis. There were 89 DEGs in the PPI network complex.

Full size image
Table 2 Gene ontology analysis of downregulated genes in LUAD.
Full size table
Table 3 KEGG pathway analysis of DEGs in LUAD.
Full size table

The Kaplan-Meier plotter (www.kmplot.com) was utilised to obtain prognostic information on the 18 core genes. Analysis revealed that out of the 18 core, 13 genes were significantly associated with better survival, while four genes showed no statistical significance (Fig. 2). The 13 genes were namely SCGB1A1, CLDN18, SFTPC, PDK4, LPL, ALDH1A1, ADH1B, CAV1, CAV2, EDNRB, CXCL2(GRO2), GNG11 and SDPR. Next, GEPIA was applied to exploit the 13 gene expression levels between LUAD and normal tissue. The results indicated that all 13 genes exhibited lower expression in LUAD samples compared with normal lung tissue samples (P < 0.05, Fig. 3).

Fig. 2
figure 2

The survival information of the core genes. Kaplan-Meier plotter online tools were utilized to identify the prognostic information of the 17 core genes. Thirteen of the 17 genes were verified to be related with a significantly better survival rate (P < 0.05).

Full size image
Fig. 3
figure 3

Expression of 13 genes in LUAD tissues and 74 non-tumour adjacent tissues via GEPIA online tool.

Full size image

DEGs and functional enrichment analysis between the CAV1 high expression group and the CAV1 low expression group

With the median CAV1 expression value (11.93697 FPKM) as the threshold, the GSE32863 dataset was divided into CAV1 high or low expression groups. Total 635 up-regulated DEGs and 399 down-regulated DEGs were identified between the two groups (Fig. 4B). Seventy-two signalling pathways with enrichment could be identified by GSEA, among which CAV1 and EGFR were involved in the endocytosis pathway (Fig. 4C). The enrichment analysis revealed several significant categories of BP; including growth factor binding and cargo receptor activity (Fig. 4D). Furthermore, the MF was significantly enriched for receptor-mediated endocytosis (Fig. 4E). CCs such as the endocytosis vesicle and endocytic vesicle membrane were significantly enriched (Fig. 4F). Enrichment analysis revealed several significant KEGG pathways, including phagosome, complement and coagulation cascades, Staphylococcus aureus infection, haematopoietic cell lineage, pertussis, rheumatoid arthritis, leishmania, extracellular matrix (ECM)-receptor interactions, malaria and asthma pathways (Fig. 4G).

Fig. 4
figure 4

CAV1 is involved in proteoglycans in cancer pathway, followed by identification and enrichment analysis of CAV1-related DEGs in GSE32863. (A) KEGG pathway enrichment showed that CAV2, CAV1, DCN and TIMP3 were significantly enriched in Chondroitin sulfate/ Dermatan sulfate proteoglycan pathway. (B) Volcano map of DEGs between high and low CAV1 expression groups in 58 LUAD patients of GSE32863 dataset. Red dots represented 635 genes raised more than 2 times, and blue dot represented 399 genes decreased more than 2 times (q < 0.01, |logFC|≥1). (C) GSEA functional enrichment between the high and low CAV1 expression groups. CAV1 and EGFR were involved in the endocytosis pathway. (D) Based on the functional analysis of DEGs, the important categories of BP determined by the gene ontology analysis were made into bubble map. Growth factor binding and cargo receptor activity were identified as significant categories of BP. (E) Enrichment analysis of important MF recognized receptor-mediated endocytosis importantly. (F) The significant categories of CC, such as endocytosis vesicle and endocytosis vesicle membrane were identified by enrichment analysis. (G) Enrichment analysis of KEGG pathways. The number of enriched genes was indicated by bubble size, and the significance of difference was indicated by bubble color and adjusted P value. The larger the bubbles, the more genes were enriched, and the degree of the red color represented the significance of difference.

Full size image

Inverse tissue localisation trends of caveolin-1 and EGFR in LUAD

Resected tumour tissues were obtained from 10 invasive LUAD patients who underwent surgery. We observed that CAV1 expression was significantly higher in paracancerous tissues when compared to LUAD tissues (Fig. 5A). There was no significant difference in the expression level of EGFR between cancer tissues and adjacent tissues (Fig. 5B). In the tissue localisation, CAV1 was mainly expressed in the adjacent tissues and the stroma of tumour tissues, while EGFR was mainly expressed in cancer cells, which indicated a negative correlation between them (Fig. 5C, D).

Fig. 5
figure 5

Comparison of CAV1 and EGFR expression in LUAD tissues and adjacent tissues. (A) Compared with LUAD tumor tissues according to IHC scores, CAV1 was highly expressed in adjacent tissues. (B) There was no significant difference in the expression level of EGFR between cancer tissues and adjacent tissues. (C) IHC staining of caveolin-1 and EGFR in LUAD tissues (200x magnification and 400x magnification). (D) IHC staining of caveolin-1 and EGFR in paracancerous tissues (200x magnification and 400x magnification). (E) Expression levels of caveolin-1 and EGFR were determined by western blotting in LUAD tumor tissues and adjacent tissues. (F) Relative expression levels of caveolin-1 in LUAD tissues and adjacent tissues. (G) Compared with LUAD tumor tissues according to western blotting, caveolin-1 was highly expressed in adjacent tissues. (H) Relative expression levels of EGFR in LUAD tissues and adjacent tissues. (I) There was no significant difference in the expression level of EGFR between cancer tissues and adjacent tissues. *Represents P < 0.05. **Represents P < 0.01.

Full size image

Resected tumour samples for western blot analysis were obtained from another 10 invasive LUAD patients who underwent surgery (Fig. 5E). Western blotting showed that compared with lung adenocarcinoma tissues, caveolin-1 was highly expressed in paracancerous tissues (P < 0.01, Fig. 5F, G). There was no significant difference in the expression level of EGFR between cancer tissues and adjacent tissues (Fig. 5H, I).

Immunohistochemical analysis of caveolin-1 and EGFR expression in LUAD tissue microarrays and cohort studies revealed distinct patterns of localization and clinical correlations

Immunohistochemical staining was conducted on tissue microarrays (Fig. 6A, D). Caveolin-1 and EGFR were detected in the cell membrane and cytoplasm. Specifically, caveolin-1 exhibited predominant expression in adjacent tissues and tumor stroma, with relatively low expression in cancer cells (Fig. 6B). Conversely, EGFR was primarily expressed in cancer cells, while its expression was minimal in adjacent tissues and tumor stroma (Fig. 6E). Compared to LUAD tissues, CAV1 demonstrated higher expression levels in paracancerous tissues (Fig. 6C), whereas EGFR expression was lower in adjacent tissues (Fig. 6F). A negative correlation was observed between the expression levels of caveolin-1 and EGFR and their respective tissue localizations (Fig. 6G).

Fig. 6
figure 6

Immunohistochemical results of caveolin-1 and EGFR in LUAD tissue microarray and cohort study. (A) Immunohistochemical detection of caveolin-1 expression in LUAD and paracancerous tissue microarray. (B) Representative sample of caveolin-1 expression in LUAD and paracancerous tissue. (C) Compared with LUAD, caveolin-1 was highly expressed in adjacent tissues. (D) Immunohistochemical detection of EGFR expression in LUAD and paracancerous tissue microarray. (E) Representative sample of EGFR expression in LUAD tissue and paracancerous tissue. (F) Compared with LUAD tissues, the expression of EGFR in adjacent tissues was significantly lower. (G) Correlation analysis showed that, caveolin-1 and EGFR expression levels in LUAD and adjacent tissues correlated negatively.(H) Kaplan-Meier survival analysis showed that, the overall survival time of caveolin-1 high expression group was prolonged. (I) Kaplan-Meier survival analysis found no significant difference between overall survival time of EGFR high and low expression group. ****Represents P < 0.0001.

Full size image

For caveolin-1 staining, the average total score of cancer cells and tumor stroma in cancer tissues served as the grouping criterion. Samples with a score ≤ 2 were classified as the caveolin-1 low-expression group, while those with a score > 2 were categorized as the high-expression group. For EGFR staining, samples with a score ≤ 2.5 were assigned to the low-expression group, and those with a score > 2.5 were designated as the high-expression group.

In the microarray cohort, LUAD patients were divided into 14 individuals in the low caveolin-1 expression group and 76 in the high-expression group. Clinical characteristics, including age, gender, pathological grade, and TNM stage, were statistically analyzed between the two groups, revealing no significant differences (Table 4). Kaplan-Meier survival analysis indicated that higher caveolin-1 expression levels were associated with prolonged overall survival (P = 0.036), as depicted in Fig. 6H.

Table 4 Comparison of clinical characteristics between caveolin-1 low and high expression groups.
Full size table

Based on EGFR expression levels, after excluding one missing data, the LUAD patients in the chip cohort were divided into EGFR low expression group (n = 27) and EGFR high expression group (n = 62). Statistical analysis of clinical characteristics showed that there were no statistically significant differences between the two groups (Table 5). Kaplan-meier survival analysis showed that there was no significant difference in overall survival between the two groups (P = 0.879), as shown in Fig. 6I.

Table 5 Comparison of clinical characteristics between EGFR low and high expression groups.
Full size table

CAV1 regulates EGFR expression in LUAD cell lines

We detected the expression of Cav-1 and EGFR in LUAD cell lines (A549, H1299, H1975 and PC-9) and Beas-2b cell lines using western blotting (Fig. 7A, B). To investigate whether CAV1 could regulate cell growth in LUAD cells, we used CAV1 siRNA to down-regulate CAV1 expression in PC-9 cells. Our western blotting results showed that CAV1 siRNA transfection inhibited the protein levels of CAV1 in PC-9 cells (P < 0.0001) and increased the protein levels of EGFR (P = 0.0299) (Fig. 7C). In the subsequent study, we selected CAV1 siRNA1 to deplete CAV1 expression in PC-9 cells.

Fig. 7
figure 7

Expression of EGFR in LUAD cells after Cav-1 knockdown or overexpression. (A) Expression levels of Cav-1 and EGFR were determined by western blotting in Beas-2b, A549, H1299, H1975 and PC-9 cell lines. (B) Correlation analysis of Cav-1 and EGFR expression levels in each cell lines. (C) The expression levels of Cav-1 and EGFR were determined by western blotting in PC-9 cells with CAV1 knockdown. (D) The expression levels of Cav-1 and EGFR were determined by western blotting in H1299 cells with CAV1 knockdown. (E) The expression levels of Cav-1 and EGFR were determined by western blotting in H1975 cells with CAV1 knockdown. (F) The expression levels of Cav-1 and EGFR were determined by western blotting in A549 cells with CAV1 overexpressing. *Represents P < 0.05. **Represents P < 0.01. ***Represents P < 0.001. ****Represents P < 0.0001.

Full size image

In the H1299 cell line, siRNA was transfected to knock down CAV1 expression. Western blotting showed that compared with the NC siRNA group, the expression level of Cav-1 protein in the CAV1 siRNA group decreased (P = 0.0004) and the level of EGFR increased (P = 0.0068), as exhibited in Fig. 7D.

In the H1975 cell line, siRNA was transfected to knock down CAV1 expression. Western blotting showed that compared to the NC siRNA group, the expression of Cav-1 protein in the CAV1 siRNA group decreased (P < 0.0001), and the level of EGFR increased (P = 0.0213), as demonstrated in Fig. 7E.

The plasmid for overexpressing CAV1 was constructed and then transfected into A549 cells. The results of Western blotting demonstrated that the recombinant Flag protein was expressed, and the level of EGFR decreased in the group with CAV1 overexpression compared to the group with the vector control (P = 0.0042), as illustrated in Fig. 7F.

CAV1 regulates cell proliferation, invasion and apoptosis in LUAD cells

The CCK-8 assay showed that the proliferation ability of the CAV1 siRNA group in PC-9 and H1975 cells was enhanced compared with that of the NC siRNA group. In the H1299 cell line, there was no significant difference in cell proliferation between the two groups. In the A549 cell line, the cell proliferation ability of the CAV1 overexpression group was weakened compared with the vector control group (Fig. 8A).

Fig. 8
figure 8

Cav-1 inhibits LUAD cell proliferation, migration and invasion, while increases cell apoptosis. (A) The effect of CAV1 knockdown and overexpression on LUAD cell proliferation was assessed by CCK8 assay. (B) The effect of CAV1 knockdown and overexpression on LUAD cell proliferation was assessed by clone formation assay. (C) The effect of CAV1 knockdown and overexpression on LUAD cell migration was assessed by scratch assay. (D) Transwell assay was used to detect the effect of CAV1 knockdown and overexpression on LUAD cell invasion. (E) The cell apoptosis of CAV1 knockdown and overexpressing PC-9, H1299 and A549 cells was analyzed by flow cytometry.

Full size image

Colony formation assay showed that compared with the vector control group, the colonies of CAV1 overexpression group were smaller, less dense and lighter, so that the cell proliferation ability was weakened. In H1299, H1975 and PC-9 cell lines, compared with the NC siRNA groups, the colonies of the CAV1 siRNA groups were larger and darker, which revealed that the cell proliferation abilities were enhanced (Fig. 8B). These findings suggested that down-regulation of CAV1 could promote cell growth in LUAD cells, while up-regulation of CAV1 could suppress cell growth in LUAD cells.

To validate the role of CAV1 in cell migration, wound healing assay was used to detect the migratory activity in LUAD cells after down-regulation or up-regulation of CAV1. The results showed that the CAV1 siRNA group had stronger cell migration ability than the NC siRNA group in PC-9, H1299 and H1975 cell lines. In the A549 cell line, the scratch healing process of the CAV1 overexpression group was slower compared with the vector control group (Fig. 8C). In conclusion, down-regulation of CAV1 promoted the migration of LUAD cells, while up-regulation of CAV1 inhibited cell migration.

To determine whether down-regulation or up-regulation of CAV1 could impact cell invasion in LUAD cells, we used transwell chamber assays to measure cell invasion in LUAD cells after siRNA or plasmid transfection. As shown in Fig. 8D, transwell assays showed that the invasion ability of the CAV1 siRNA group was stronger than that of the NC siRNA group in PC-9, H1299 and H1975 cell lines. In the A549 cell line, compared with the vector control group, the invasion ability of the CAV1 overexpression group was weaker. Taken together, these findings suggest that down-regulation of CAV1 promotes invasive activity in LUAD cells, while up-regulation of CAV1 inhibits cell invasion.

To detect whether CAV1 governs cell apoptosis in LUAD cells, cell apoptosis was measured after transfection. The Annexin V-FITC/PI apoptosis assay was used to measure the percentage of apoptotic cells. The results showed that in PC-9 and H1299 cell lines, the proportion of apoptotic cells in the CAV1 siRNA group was lower than that in the NC siRNA group. In the A549 cell line, after cisplatin treatment, the apoptosis rate of the CAV1 overexpression group was higher than that of the vector control group (Fig. 8E). These results indicated that up-regulation of CAV1 enhanced cell apoptosis, which could contribute to the inhibition of cell growth in LUAD cells.

CAV1 degrades EGFR in LUAD cell lines

To indicate whether EGFR expression was regulated by CAV1 at the transcriptional level, we used RT-qPCR to detect EGFR mRNA levels in LAUD cells with the CAV1 knockdown. After overexpression of CAV1 in A549 cells, both the levels of CAV1 mRNA and EGFR mRNA increased in the CAV1 overexpression group compared to the vector control group (Fig. 9A). After transfection of siRNA in H1299, H1975 and PC-9 cells, the levels of CAV1 mRNA and EGFR mRNA in the CAV1 siRNA group were reduced compared to the NC siRNA group (Fig. 9B–D). Therefore, CAV1 may not directly inhibit EGFR protein synthesis, and further studies are required to investigate the mechanism of CAV1 in regulating EGFR, particularly in terms of protein degradation.

Fig. 9
figure 9

Cav-1 regulates EGFR degradation at the protein level and mediates the downstream AKT/STAT3 signaling pathway, Bax/Caspase-3/Bcl-2 pathway and epithelial-mesenchymal transition (EMT). (A) CAV1 and EGFR mRNA levels after transfection of CAV1 overexpression plasmid in A549 cell line were detected by RT-qPCR. (B) CAV1 and EGFR mRNA levels after transfection of siRNA in H1299 cell line were detected by RT-qPCR. (C) CAV1 and EGFR mRNA changes after transfection of CAV1 siRNA in H1975 cell line were detected by RT-qPCR. (D) CAV1 and EGFR mRNA levels after transfection of CAV1 siRNA in PC-9 cell line were detected by RT-qPCR. (E) Results of CHX chase experiment of CAV1 overexpressing group in A549 cell line. (F) Results of CHX chase experiment of CAV1 knockdown group in PC-9 cell line. (G) Western blot analyses of the levels of caveolin-1, EGFR, p-AKT, AKT, p-STAT3 and STAT3 in CAV1-knockdown and CAV1-overexpressing LUAD cells. (H) Western blot analyses of the levels of caveolin-1, Bax, caspase-3, cleaved caspase-3 and Bcl-2 in CAV1-knockdown and CAV1-overexpressing LUAD cells. (I) Western blot analyses of the levels of caveolin-1, E-cadherin, N-cadherin and vimentin in CAV1-knockdown and CAV1-overexpressing LUAD cells.

Full size image

The CHX chase assay was used to demonstrate whether Cav-1 could regulate EGFR degradation in LUAD cells. CHX was added to inhibit protein synthesis of EGFR, so the degradation of EGFR protein could be detected and observed subsequently. In A549 cells, EGFR degradation was increased and accelerated in the CAV1 overexpression group compared with the vector control group (Fig. 9E). The result indicated that overexpression of CAV1 could contribute to the degradation of EGFR on the cell membrane. In the PC-9 cell line, the addition of CHX reduced and delayed EGFR degradation in the CAV1 siRNA group compared with the NC siRNA group (Fig. 9F). In conclusion, up-regulation of CAV1 promoted EGFR degradation in LUAD cells, while down-regulation of CAV1 obstructed the process.

Caveolin-1 regulates EGFR/AKT/STAT3 pathway in LUAD cells

To determine whether CAV1 could regulate the phosphorylation of downstream signalling pathways of EGFR in LUAD cells, western blot analysis was conducted to detect the levels of p-AKT and p-STAT3 in LUAD cells after CAV1 siRNA or overexpression plasmid transfection. In PC-9, H1975 and H1299 cell lines, the levels of p-AKT and p-STAT3 were increased in the CAV1 siRNA group compared to the NC siRNA group. In the A549 cell line, p-AKT and p-STAT3 levels were decreased in the CAV1 overexpression group compared to the vector control group (Fig. 9G). These results suggest that CAV1 can decrease EGFR expression, subsequently suppressing the phosphorylation of AKT and STAT3 in LUAD cells.

Caveolin-1 regulates Bax/caspase-3/Bcl-2 pathway in LUAD cells

To investigate whether CAV1 regulates the apoptosis signaling pathway in LUAD cells, Western blot analysis was performed to evaluate the protein levels of Bax, cleaved caspase-3, and Bcl-2 in LUAD cells following transfection with CAV1 siRNA or an overexpression plasmid. In the A549 cell line, cisplatin was administered to induce apoptosis. Compared with the vector control group, the expression levels of Bax, caspase-3, cleaved caspase-3, and Bcl-2 were significantly increased in the CAV1 overexpression group. In the H1975 cell line, the expression levels of Bax, caspase-3, cleaved caspase-3, and Bcl-2 were lower in the CAV1 siRNA group than in the NC siRNA group after cisplatin treatment. Similarly, in the H1299 cell line, the expression levels of these proteins were reduced in the CAV1 siRNA group compared to the NC siRNA group. In the PC-9 cell line, the expression levels of Bax, cleaved caspase-3, and Bcl-2 were also decreased in the CAV1 siRNA group relative to the NC siRNA group. These findings indicate that caveolin-1 positively regulates the Bax/Caspase-3/Bcl-2 pathway (Fig. 9H), thereby playing a critical role in apoptosis in LUAD cells.

Caveolin-1 influences epithelial-mesenchymal transition (EMT)

Western blotting was employed to assess the expression of EMT-related molecular markers in LUAD cells. In the H1975 cell line, the expression of E-cadherin was downregulated in the CAV1 siRNA group compared to the NC siRNA group, while the expression levels of N-cadherin and Vimentin were upregulated. In other cell lines, the results exhibited some variability. In the A549 cell line, the expression level of N-cadherin was elevated in the CAV1 overexpression group compared to the vector control group. In the H1299 cell line, the expression levels of N-cadherin and Vimentin were increased in the CAV1 siRNA group compared with the NC siRNA group, while the expression of E-cadherin was not significantly changed. In the PC-9 cell line, compared with NC siRNA group, the expression of E-cadherin and N-cadherin increased, and the expression level of vimentin decreased in CAV1 siRNA group (Fig. 9I).

CAV1 was knocked down using ShRNA to investigate its anti-proliferative, anti-invasive, anti-migratory, and pro-apoptotic effects in the PC-9 cell line

In this study, stable cell lines (shNC, shCAV1-1, and shCAV1-2) were successfully screened and constructed. Western blot analysis confirmed the knockdown efficiency of CAV1 in the shNC, shCAV1-1, and shCAV1-2 stably transfected cell lines, while also detecting the expression levels of EGFR protein. The results demonstrated that compared with the shNC group, the protein expression levels of CAV1 in the shCAV1-1 and shCAV1-2 groups were significantly reduced, whereas the protein levels of EGFR were markedly increased (Fig. 10A). RT-qPCR was employed to validate the changes in CAV1 and EGFR mRNA expression in the shNC, shCAV1-1, and shCAV1-2 stable cell lines. The findings revealed that compared with the shNC group, the mRNA levels of CAV1 were significantly decreased in the shCAV1-1 and shCAV1-2 groups, while the mRNA levels of EGFR were significantly elevated (Fig. 10B). Western blotting was further utilized to examine the phosphorylation status of proteins in the EGFR/AKT/STAT3 signaling pathway in the shNC, shCAV1-1, and shCAV1-2 groups. The results indicated that the levels of p-AKT and p-STAT3 in the shCAV1-1 group were higher than those in the shNC group, and p-AKT levels were also increased in the shCAV1-2 group (Fig. 10C). Additionally, western blotting was used to assess the expression levels of markers in the Bax/Caspase-3/Bcl-2 pathway in the shNC, shCAV1-1, and shCAV1-2 groups. The data showed that compared with the shNC group, the expression levels of Bax were increased in the shCAV1-1 and shCAV1-2 groups, while the expression levels of cleaved caspase-3 were decreased, and Bcl-2 levels were reduced (Fig. 10C).

Fig. 10
figure 10

Construction and phenotype experiments of PC-9 cell line stably transfected with shCAV1. (A) The knockdown of CAV1 in shNC, shCAV1-1 and shCAV1-2 stably transfected cell lines was verified by western blotting, and the expression of EGFR protein was detected. (B) RT-qPCR was used to detect the changes of CAV1 and EGFR mRNA in the shNC, shCAV1-1 and shCAV1-2 groups. (C) Western blotting was used to detect the protein phosphorylation of EGFR/AKT/STAT3 pathway and expressions of Bax/Caspase-3/Bcl-2 pathway markers in shNC, shCAV1-1 and shCAV1-2 groups. (D) CCK-8 experiment showed that compared with the shNC group, the proliferation ability of shCav1-1 and shCav1-2 groups was enhanced, and the difference was statistically significant. (E) The clone formation assay showed that compared with the shNC group, the shCAV1-1 and shCAV1-2 groups had an increased number of colonies, larger shape, and darker color. (F) Transwell assay showed that compared with the shNC group, the invasion and migration ability of shCAV1-1 and shCAV1-2 group was enhanced, and the difference was statistically significant. (G) Scratch experiments in shNC, shCAV1-1 and shCAV1-2 groups. (H) Scratch experiments showed that the migration ability of shCAV1-1 and shCAV1-2 groups was significantly enhanced compared with that of shNC group. (I) Apoptosis experiments in shNC, shCAV1-1 and shCAV1-2 groups. Ns represents no significant difference. *Represents P < 0.05, **represents P < 0.01, ***represents P < 0.001, and ****represents P < 0.0001.

Full size image

Regarding cell proliferation experiments, both CCK-8 assays and colony formation assays revealed that compared with the shNC group, the proliferative capacity of the shCAV1-1 and shCAV1-2 groups was enhanced, and the difference was statistically significant (Fig. 10D, E). From the perspective of cell invasion assays, transwell experiments demonstrated that compared with the shNC group, the invasive capabilities of cells in the shCAV1-1 and shCAV1-2 groups were significantly enhanced, with statistically significant differences (Fig. 10F).

Regarding cell migration experiments, both transwell assays and scratch tests revealed that compared with the shNC group, the migratory capabilities of cells in the shCAV1-1 and shCAV1-2 groups were also significantly enhanced, with statistically significant differences (Fig. 10G).

Apoptosis assays indicated that 24 h after cisplatin-induced apoptosis, a lower proportion of early apoptotic cells was detected in the shCAV1-1 and shCAV1-2 groups compared with the shNC group (Fig. 10I).

Caveolin-1 has anti-tumor effect in vivo

To verify the anti-tumor effects of caveolin-1 in vivo, subcutaneous tumor formation experiments were conducted in nude mice. Tumor volumes were measured and calculated, revealing that compared with the shNC group, the tumor volumes in the shCAV1-1 and shCAV1-2 groups were significantly larger, with statistically significant differences (Fig. 11A). On the 28th day post-transplantation, subcutaneous xenograft tumors were harvested, and tumor weights were measured and compared. The results showed that compared with the shNC group, tumors in the shCAV1-1 and shCAV1-2 groups were significantly heavier, with statistically significant differences (Fig. 11B, C).

Fig. 11
figure 11

Results of subcutaneous tumor formation experiment in nude mice. (A) Tumor volume measurement and calculation of nude mice. Compared with the shNC group, the tumor volume of the shCAV1-1 and shCAV1-2 groups was larger, and the difference was statistically significant. (B) The tumor weight of the shCAV1-1 and shCAV1-2 groups was heavier than that of the shNC group, and the difference was statistically significant. (C) The appearance and size of tumors in shNC, shCAV1-1 and shCAV1-2 groups. (D) HE staining and immunohistochemical comparison of Caveolin-1, EGFR and Ki-67 in shNC, shCAV1-1 and shCAV1-2 groups. (E) Immunohistochemical score comparison of Caveolin-1, EGFR and Ki-67 in shNC, shCAV1-1 and shCAV1-2 groups. Compared with the shNC group, the expression of Caveolin-1 in the tumors of the shCAV1-1 and shCAV1-2 groups decreased, the expression of EGFR increased, and the Ki-67 index increased, with statistically significant differences. *Represents P < 0.05, **represents P < 0.01, and ****represents P < 0.0001.

Full size image

Tumor tissues were confirmed by HE staining under a microscope. Immunohistochemical analyses further revealed that compared with the shNC group, the expression levels of caveolin-1 were decreased, while those of EGFR and Ki-67 were increased in the shCAV1-1 and shCAV1-2 groups, with statistically significant differences (Fig. 11D, E).

Discussion

The genes CAV1, and CAV2, which encode human Cav-1 and Cav-2 localise in D7S522 of chromosome 7q31.118. Caveolins are a family of integral membrane proteins that weight between 21 and 25 kDa. They are vital constituents of detergent-resistant cholesterol-lipid-rich membranes, including caveolae and lipid rafts. Cav-1 binds to various signalling molecules, including epidermal growth factor receptor (EGFR), G-proteins, Src family tyrosine kinases, protein kinase C, and endothelial nitric oxide synthase. Additionally, it has the ability to functionally inactivate the enzymatic activity of the above-mentioned molecules. Therefore, the role of Cav-1 has been demonstrated in various biological behaviours, including cancer progression, tumour suppressor function19, apoptosis20, signal transduction, vesicular trafficking, potocytosis and transcytosis. The current study suggests that Cav-1 could act as potent targets for effective drug therapy. Notably, when Lewis lung cancer cells were transplanted into CAV1 knockout mice, there was a significant increase in tumour growth rates and vascular permeability, which may be associated with VE-cadherin mediated VEGFR-2 phosphorylation21. CAV1 may regulate cell proliferation, migration and invasion of A549 cells through controlling of HOTAIR expression22. According to Cox regression analysis, increased expression of Cav-1 in NSCLC tumour cells was identified as a prognostic factor for overall survival23. However, several studies have demonstrated that Cav-1 expression is related to poor prognosis and advanced pathologic stage in squamous cell lung cancer24 and LUAD25, which has made the role of Cav-1 in the genesis and development of lung cancer more complicated and controversial. Cav-1 immunoreactivity in lung cancer depends on the tumour histotype26. The different results may be on account of different roles and expression levels27 of Cav-1 in various stages of disease28, tumour types and cell lines29.

EGFR is a transmembrane glycoprotein receptor for members of the EGF family, and it belongs to the protein kinase superfamily. It serves as a central hub for many other pathways.7 The primary downstream pathways are Ras-Raf-MEK-ERK1/2 and PI3K-Akt-mTOR-STAT3 which are associated with cell differentiation and proliferation. It has been reported that EGFR interacts with the caveolin scaffolding domain through a motif located in the kinase domain, especially in residues 898–90530. Khan EM et al. found that under oxidative stress, EGFR was able to undergo clathrin-independent endocytosis, and sorted to the perinuclear region where EGFR could remain active31. In recent years, studies on the relationship between EGFR and Cav-1 in the fields of autophagy32, cancer, and trauma33 have emerged. The regulatory association between EGFR and Cav-1 has been studied in tumours of other systems. For instance, in head and neck squamous cell carcinoma, cancer cells with overexpressed CAV1 exhibited improved viability when exposed to the EGFR targeting antibody cetuximab in combination with radiotherapy. This resistance was mediated by the Cav-1/EREG/YAP axis34. There was a significant negative correlation between EGFR and Cav-1 expression in thyroid cancer35. In breast cancer MCF-7 cell line, the level of EGFR phosphorylation was decreased in the cells overexpressing Cav-1 compared with parental cells, while the levels of total EGFR expression were similar36. This finding is in line with the results of the current study, which demonstrated that Cav-1 regulated the EGFR/AKT signalling pathway in several LUAD cell lines, including A549, H1299, H1975 and PC-9. However, Cui Y et al. showed that knockdown of CAV1 down-regulated EGFR phosphorylation, which may enhance sensitivity to EGFR-TKIs. Based on the findings, it is speculated that Cav-1 could serve as an indicator of the reduced effectiveness of EGFR-TKIs in patients with EGFR-mutated LUAD37. The results of this study were not consistent with those of our study, and further experiments related to TKIs should be carried out to verify the results.

AKT is a serine/threonine protein kinase with a molecular weight of approximately 60 kDa, which plays a crucial role in the phosphatidylinositol-3-kinase (PI3K) signalling pathway. Overexpression and phosphorylation-induced activation of AKT are a significant event can regulate the function of many downstream proteins and participate in cell survival, proliferation, migration, metabolism and angiogenesis38. David O et al. verified that the high expression of p-AKT was an independent prognostic factor of NSCLC patients, and was associated with a limited survival time39. This study showed that Cav-1 can act as a negative regulator of the EGFR/AKT signalling pathway. That is to say, Cav-1 suppresses AKT phosphorylation in cells by antagonising EGFR, thereby inhibiting the growth of LUAD cells and promotes tumour cell apoptosis. The signal transducer and activator of transcription 3 (STAT3) is one of the key downstream signalling mediators of EGFR activation and is considered to be one of the key factors in tumour formation40. Previous studies have shown that the overexpression of p-STAT3 is associated with prognosis of several human tumours41. However, the association of p-STAT3 with lung cancer is still not fully defined. A systematic review has shown that high expression of p-STAT3 is significantly associated with poor overall survival, later TNM stage and lymph node metastasis in patients with lung cancer42. The results of this study showed that the levels of EGFR and p-STAT3 were increased after CAV1 knockdown, while the proliferation and invasion abilities of tumour cells were enhanced, and apoptosis was reduced. Consistent with the conclusion of this study, the above clinical study results were supported and clarified by the basic experimental aspects.

Apoptosis proceeds through a cascade of caspase activation, encompassing both exogenous and endogenous pathways43. The exogenous pathway-mediated apoptosis involves the receptor Fas-associated protein with death domain (FADD), which activates the initiator caspases such as caspase-8 or -10, subsequently leading to the activation of effector caspases including caspase-3, -6, -7, and - 9. In contrast, the endogenous pathway is triggered by stress, radiotherapy, and chemotherapy drugs, involving mitochondrial release of cytochrome C that forms a complex with apoptotic protease-activating factor-1 (APAF-1) and caspase-9, thereby activating effector caspases, particularly caspase-344. The regulation of apoptosis is a highly intricate process involving multiple genes, including Bcl-2 and its related family members such as Bcl-xL, Bcl-xs, Bad, and Bax. Bcl-2 and Bcl-xL exert anti-apoptotic effects partly by binding to pro-apoptotic proteins like Bax. Members of the Bcl-2 family regulate the initiation of the mitochondrial apoptotic pathway45. Bcl-2, a mitochondrial membrane protein, functions at various levels of the apoptotic cascade primarily to inhibit apoptosis and enhance cell survival. Overexpression of Bcl-2 protein has been associated with increased carcinogenic potential and adverse responses to lung cancer treatment. While Bcl-2 inhibits the release of cytochrome C, Bax promotes its release, thereby promoting caspase-3 expression46. Bax promotes apoptosis by forming channels with voltage-dependent anion channels (VDACs) and adenine nucleotide translocators (ANTs) on the mitochondrial membrane, thereby facilitating cytochrome C release. Caspase-3 serves as one of the primary effectors of apoptosis, and its activation systematically degrades cellular components through the cleavage of key proteins, particularly poly ADP-ribose polymerase (PARP). This study demonstrates that CAV1 upregulates Bax expression, leading to caspase-3 activation and potentially playing a pivotal role in inducing apoptosis in LUAD cells. Elevated Bax levels trigger apoptosis, whereas Bcl-2 functions as a downstream gene transcriptionally regulated by the STAT3 signaling pathway47. In this investigation, the expression level of Bcl-2 was influenced by STAT3 signaling. When Bcl-2 expression was upregulated or overexpressed, cells exhibited resistance to Bax-induced apoptosis. Jame-Martinez LA et al. reported that Bax is more critical for cell apoptosis than Bcl-248. Consistent with this finding, our study observed elevated Bcl-2 levels, highlighting the interplay between pro-apoptotic Bax and anti-apoptotic Bcl-2 in regulating apoptosis. Modulating the activities of these signaling molecules may offer novel therapeutic strategies for LUAD.

Caveolin-1 mediates tumor invasion and metastasis, often associated with epithelial-mesenchymal transition (EMT). Studies in other systemic tumors have elucidated the role of caveolin-1 in regulating EMT from diverse perspectives, with results varying based on tumor and cell type. In bladder cancer, caveolin-1 induces EMT by activating the PI3K/AKT signaling pathway, upregulating Slug expression, and promoting metastasis49. In hepatocellular carcinoma, caveolin-1 overexpression significantly increases vimentin expression while decreasing E-cadherin expression, resulting in EMT changes and impacting tumor cell migration and invasion abilities50. In gastric cancer, hypoxia-induced reduction in caveolin-1 function enhances EGFR activation, subsequently activating STAT3. This cascade results in the downregulation of E-cadherin and upregulation of mesenchymal markers (Slug, α-SMA, N-cadherin, and vimentin)51. Herein, we demonstrate that CAV1 knockdown in the H1975 cell line induces EMT, thereby promoting tumor invasion. In A549, H1299, and PC-9 cell lines, N-cadherin expression was elevated; however, the results for other molecular markers were inconsistent across these lines. EMT is not merely a binary process but encompasses a spectrum of intermediate states, with different cell lines potentially residing at distinct stages of EMT progression. In complete EMT (cEMT), cells fully lose their epithelial markers and acquire mesenchymal markers. By contrast, in partial EMT (pEMT), cells co-express both epithelial and mesenchymal markers, resulting in a hybrid phenotype that may exhibit enhanced aggressiveness. The variability in EMT outcomes arises from the synergistic interplay between intrinsic cellular properties (e.g., transcription factor expression profiles) and extrinsic factors (e.g., microenvironmental cues and inducing signals). Differential activation of transcription factors across various cell lines contributes to the heterogeneity of EMT phenotypes. Additionally, distinct signaling pathways, such as the TGF-β pathway and FAK signaling, demonstrate differential responsiveness in different cell lines. These observations suggest potential inherent differences among cell lines and tumor subtypes. In addition to EMT, it is plausible that caveolin-1 mediates tumor cell invasion and migration through alternative pathways. Furthermore, mesenchymal-epithelial transition (MET) may also occur during tumorigenesis and progression.

Ki-67 is a DNA-binding nuclear protein expressed during the active phases of the cell cycle. It serves as a well-established prognostic marker in various malignancies, including breast cancer and mantle cell lymphoma. Studies indicate that high Ki-67 expression negatively correlates with disease-free survival, recurrence-free survival, and overall survival in non-small cell lung cancer52,53,54. A meta-analyses on Ki-67 staining in lung cancer confirm its role as a poor prognostic factor for recurrence and survival outcomes55. Ki-67 is widely used as a marker of cellular proliferation to assess tumor growth dynamics56. Several studies have shown that the Ki-67 index independently predicts tumor progression. In this study, shRNA-mediated CAV1 knockdown in vivo resulted in upregulated Ki-67 expression, consistent with the observed trend in tumor growth among groups, thus supporting enhanced cellular proliferation.

Caveolin-1 plays a critical role in various biological mechanisms, including cell senescence, microenvironment regulation, inflammatory signaling pathways, and iron metabolism. The upregulation of cav-1, along with alterations in plasma membrane components, serves as a key marker for the morphological changes associated with cell senescence. Cav-1 has been shown to induce cell senescence and inhibit tumor angiogenesis in colorectal cancer57. Additionally, cav-1 regulates tissue responses and influences tumor invasion and metastasis by mechanically modulating the microenvironmental structure58. As a mechanosensor, cav-1 responds to diverse mechanical stimuli from the cellular microenvironment, thereby participating in integrin-mediated inflammatory signaling pathways59. Furthermore, Cav-1 inhibits excessive accumulation of Fe2+ in hepatocytes by activating the FTL/FTH pathway, which alleviates liver injury60. Low pH conditions promote the secretion of exosomes containing caveolin-161. Hydrostatic pressure ranging from 0 to 20 mmHg enhances the migration speed and invasiveness of lung cancer cells (CL1-5 and A549), increases the expression level of cav-1, and elevates the phosphorylation of ERK1/262. In conclusion, further investigations into the underlying mechanisms are essential for elucidating the precise role of cav-1 in LUAD.

Conclusion

In summary, we have elucidated that the CAV1/EGFR/AKT/STAT3 signaling pathway plays a critical role in mediating tumor growth and provided robust evidence supporting the function of CAV1 in suppressing LUAD proliferation and metastasis. These findings may enhance our understanding of the mechanisms underlying CAV1’s role in LUAD progression and its potential implications for therapeutic strategies.