Exosome-derived Menin from cancer-associated fibroblasts promotes gastric cancer progression by activating the HSPA6/JNK/JunD pathway and inducing EMT

Abstract

Background

Our previous studies found that Menin was highly expressed in gastric cancer (GC) and could promote GC progression. Tumor microenvironment (TME), including cancer-associated fibroblasts (CAFs) and their exosomes plays pivotal roles in GC. It remains unclear whether exosomes derived from CAFs influence GC by delivering Menin.

Methods

Primary CAFs and normal fibroblasts (NFs) were isolated from fresh GC tissues, and co-cultured with GC cells. After Men1 expression in CAFs and NFs was modulated, exosomes were extracted via ultracentrifugation and mixed with GC cells. Next, GC cell biological behaviors were assessed in vitro. A nude mouse model of lung metastasis was established, and a small animal in vivo imaging system was used to monitor the effects of exosomes on metastasis. HSPA6/JNK/JunD pathway components and EMT-related molecules were detected by Western blot.

Results

Menin was highly expressed in CAFs and in their exosomes. Co-culturing of CAFs with GC cells promoted the proliferation, invasion and migration of GC cells. After Men1 was knocked down in CAFs, exosomes derived from these CAFs inhibited the progression of GC both in vitro and in vivo. Conversely, after overexpressing Men1, exosomes from NFs promoted the progression of GC both in vitro and in vivo. The HSPA6/JNK/JunD pathway and EMT in GC cells were activated when GC cells were co-cultured with CAFs or exosomes from Menin-overexpressing NFs.

Conclusion

CAFs can promote GC progression by delivering Menin-containing exosomes, which activates the HSPA6/JNK/JunD pathway and induces EMT. Targeting Menin within CAFs and GC cells and blocking the delivery of Menin by exosomes may provide novel strategies for GC treatment.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12967-025-07539-3.

Introduction

GC ranks among the most lethal malignancies worldwide, imposing a severe disease burden in China characterized by “high incidence, high mortality, and a low rate of early diagnosis“ [1]. Despite the continuous advancement of therapeutic approaches such as surgical resection, neoadjuvant chemotherapy, and targeted therapy [2, 3], patients with advanced GC still confront the predicament of high recurrence rate, significant risk of distant metastasis, and poor prognosis, with the 5-year survival rate remaining persistently below 30% [4, 5]. The crux of this clinical bottleneck lies in the fact that current understanding of the molecular mechanisms underlying GC progression remains focused on the malignant transformation of tumor cells themselves, while the complex intercellular communication network within the TME has not been systematically characterized. Therefore, in-depth exploration of the key molecular pathways through which the TME regulates GC invasion and metastasis has become an urgent need for the development of novel therapeutic strategies.

The TME is a dynamic complex composed of various components including tumor cells, immune cells, endothelial cells, and fibroblasts [6]. Among these, CAFs represent the most prominent functional cells in the tumor stroma [7]. Compared with NFs, CAFs not only directly drive tumor progression by secreting cytokines (e.g., TGF-β, IL-6) and by remodeling the extracellular matrix (e.g., collagen, fibronectin) but also mediate long-distance intercellular communication through the release of extracellular vesicles (EVs) [8]. As a functionally well-defined subset of EVs, exosomes (with a diameter of 40–150 nm), which rely on their cargo of bioactive molecules such as proteins, RNAs, and lipids, have been confirmed as core mediators through which CAFs regulate the maintenance of GC cell stemness, chemoresistance, and immune evasion [9–12]. For instance, existing studies have revealed that CAFs-derived exosomes can activate the β-catenin/c-Myc pathway by delivering CCT6A or induce M2 polarization of macrophages by transferring miR-4253, thereby promoting the malignant phenotype of GC [13, 14]. However, whether other key effector molecules exist in CAFs-derived exosomes and how they precisely regulate the invasive and metastatic phenotypes of GC cells remain to be further explored.

Menin, a multifunctional transcriptional regulatory protein encoded by the Men1 gene, exhibits a remarkable tissue-specific role in tumors [15]. In diseases such as pancreatic neuroendocrine tumors, Menin functions as a tumor suppressor due to its own deletion or mutation; in contrast, in hepatocellular carcinoma and colorectal cancer, Menin exerts oncogenic effects by regulating pathways such as the Hedgehog and Wnt pathways [16–20]. Our previous study demonstrated for the first time that Menin is highly expressed in GC tissues and that it can enhance the invasive capacity of GC cells by activating the HSPA6/JNK/JunD pathway. Nevertheless, critical scientific questions remain unanswered: Does this regulatory process depend on the involvement of the TME? Can CAFs deliver Menin to GC cells via exosomes, thereby amplifying its oncogenic effects?

In order to answer these questions, we performed a series of experiments, aiming to elucidate the critical role of CAFs-derived exosomal Menin in GC progression and to provide experimental evidence for the development of novel GC therapeutic strategies targeting the TME.

Materials and methods

Isolation, culture, and immortalization of primary fibroblasts

Primary fibroblasts were isolated from tumor tissues and paired normal gastric mucosa tissues of GC patients. Minced tissues were digested in Dulbecco’s Modified Eagle Medium (DMEM, Gibco, USA)/Ham’s F-12 medium (Gibco, USA) containing 1 mg/mL Collagenase IV (Solarbio, Beijing, China). The digested cells were washed and then resuspended in DMEM/Ham’s F-12 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. Subsequently, they were incubated at 37 °C in a humidified atmosphere with 5% CO₂. All experiments were conducted using cells between passages 3 and 6 to ensure phenotypic stability. For immortalization, primary CAFs and NFs were transfected with SV40 plasmids. Western blot analysis was used to assess Menin expression before and after transfection, confirming successful transfection without affecting Menin expression levels (Supplementary Fig. 1A-B).

Immunofluorescence staining

Fibroblasts or GC tissue samples were fixed in 4% paraformaldehyde for 15 min, permeabilized with 0.2% Triton X-100 (Solarbio, Beijing, China) for 10 min, and then rinsed with phosphate-buffered saline (PBS). Samples were blocked with 5% bovine serum albumin (BSA, Sigma-Aldrich, USA) for 30 min, followed by incubation with primary antibodies against α-SMA (Proteintech, Cat# 23081-1-AP, RRID: AB_2815024, 1:200), FAP (Proteintech, Cat# 84018-4-RR, 1:200), and Menin (Bethyl, Cat# A300-105 A, RRID: AB_2143306, 1:500) at 4 °C overnight in the dark. After washing, samples were incubated with Alexa Fluor 488-conjugated secondary antibodies (Invitrogen, USA) for 1 h at room temperature in the dark. Subsequently, nuclei were counterstained with 4’,6-diamidino-2-phenylindole (DAPI) for 25 min. Images were captured using an inverted fluorescence microscope (Nikon, Japan).

Flow cytometry analysis

Cells were fixed in 4% paraformaldehyde for 15 min at room temperature, rinsed three times with PBS, and incubated with primary antibodies against α-SMA (Proteintech, Cat# 23081-1-AP, RRID: AB_2815024) and FAP (Proteintech, Cat# 84018-4-RR) for 1 h at room temperature in the dark. After washing, cells were incubated with Alexa Fluor 488-conjugated secondary antibodies (Invitrogen, USA) for 30 min under the same conditions. Fluorescence-labeled cells were analyzed using a flow cytometer (Beckman Coulter, USA).

RNA-seq transcriptome sequencing and bioinformatics analysis

RNA-sequencing (RNA-seq) was performed by Guangzhou Gedian Biotechnology Co., Ltd. (Guangzhou, China). PolyA-tailed eukaryotic mRNA was enriched using Oligo(dT) magnetic beads. cDNA was synthesized using random primers, and the second-strand cDNA was generated using DNA polymerase I. The cDNA library was purified using AMPure XP magnetic beads (Beckman Coulter, USA). RNA purity and concentration were assessed with a NanoPhotometer (IMPLEN, Germany) and Qubit 2.0 fluorometer (Invitrogen, USA), respectively. RNA integrity was evaluated using an Agilent 2100 bioanalyzer (Agilent Technologies, USA). High-throughput sequencing was conducted on an Illumina HiSeq™ 2500 system to generate raw reads. All downstream analyses were performed in R (v4.3.3, 2024). Differential expression analysis was carried out using DESeq2 (v1.42.0), with genes exhibiting |log₂FC| >1 and FDR < 0.05 considered significantly differentially expressed. Gene Ontology (GO) enrichment analysis and Gene Set Enrichment Analysis (GSEA) were performed using clusterProfiler (v4.10.0) and fgsea (v1.24.0), respectively.

Isolation and identification of exosomes

Fibroblasts were cultured in serum-free DMEM for 24–48 h. Exosomes were isolated from cell supernatants using differential ultracentrifugation combined with 0.22-µm filtration. Briefly, cell supernatants were subjected to sequential centrifugation at 4 °C (300×g for 5 min, 2000×g for 10 min, and 10,000×g for 30 min) to remove cell debris and apoptotic bodies. The supernatant was filtered through a 0.22-µm filter (Millipore, USA) and ultracentrifuged at 120,000×g for 90 min. The exosome pellet was resuspended in sterile PBS and stored at -80 °C. Transmission electron microscopy (TEM; Tecnai G20 TWIN, FEI, USA) was used to observe exosome morphology. Western blot analysis was performed to detect exosome-specific markers (CD63 and TSG101) in three biological replicates.

Cell culture

Human GC cell lines AGS (RRID: CVCL_0139) and MKN-28 (RRID: CVCL_1416) were obtained from Wuhan Pricella Life Technology Co., Ltd. (Wuhan, China) and authenticated using short tandem repeat (STR) profiling. AGS cells were cultured in Ham’s F-12 medium (Gibco, USA) supplemented with 10% FBS (Biological Industries, Israel), 50 µg/mL streptomycin, and 50 U/mL penicillin (Solarbio, Beijing, China). MKN-28 cells were maintained in RPMI-1640 medium (Gibco-BRI, USA) containing 10% FBS, 50 µg/mL streptomycin, and 50 U/mL penicillin. All cells were cultured at 37 °C in a humidified atmosphere with 5% CO₂.

Quantitative real-time PCR (qRT-PCR)

Total RNA was extracted from three biological replicates of each sample using the NucleoZOL RNA Isolation Kit (MACHEREY-NAGEL, Germany). Reverse transcription of 1 µg total RNA into cDNA was performed using HiScript III RT SuperMix for qPCR (+ gDNA wiper) (Vazyme, China). qRT-PCR was conducted on a 7500 Real-Time PCR System (Applied Biosystems, USA) using ChamQ SYBR qPCR Master Mix (Vazyme, China). The thermal cycling conditions were as follows: initial denaturation at 95 °C for 30 s, followed by 40 cycles of 95 °C for 10 s and 60 °C for 30 s. GAPDH was used as an internal reference gene. Primer sequences (synthesized by Sangon Biotech, Shanghai, China) are listed in Table 1.

Table 1.

Primer sequences used for qRT-PCR

Name	Sequence
Men1-F	5’-CCCATGTACTTACAAGCCGATAT-3’
Men1-R	5’-CCCCATTTTCGTAAACACACTC-3’
ACTA2-F(α-SMA)	5’-GTGTTGCCCCTGAAGAGCAT-3’
ACTA2-R(α-SMA)	5’-GCTGGGACATTGAAAGTCTCA-3’
FAP-F	5’-ATGAGCTTCCTCGTCCAATTCA-3’
FAP-R	5’-AGACCACCAGAGAGCATATTTTG-3’
GAPDH-F	5’-GTGGCCGAGGACTTTGATTG-3’
GAPDH-R	5’-CCTGTAACAACGCATCTCATATT-3’

Lentivirus-mediated RNA interference (RNAi) and overexpression

Small interfering RNA (siRNA) fragments targeting Men1 and Men1 overexpression sequences were inserted into lentiviral vectors (Jikai Gene Technology Co., Ltd., Shanghai, China). The sequences for Men1 silencing and overexpression are provided in Supplementary File 1. CAFs and NFs were infected with lentiviruses at a multiplicity of infection (MOI) of 100 for 72 h. Transfection efficiency was evaluated using fluorescence microscopy. Stable cell lines were selected using 2 µg/mL puromycin (Biosharp, China).

Western blot analysis

Three independent biological replicates were prepared for each experimental group. Whole-cell lysates were extracted from fibroblasts, exosomes, AGS, and MKN-28 cells using RIPA lysis buffer (Solarbio, Beijing, China) containing 1 mM phenylmethanesulfonyl fluoride (PMSF) and phosphatase inhibitors. Protein concentration was determined using a bicinchoninic acid (BCA) protein assay kit (Beyotime, Shanghai, China). Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene fluoride (PVDF) membranes (Merck Millipore, Germany). Membranes were blocked with 5% skim milk for 30 min and incubated with primary antibodies overnight at 4 °C. Primary antibodies used included: anti-Menin (Bethyl, Cat# A300-105 A, RRID: AB_2143306, 1:8000), anti-α-SMA (Proteintech, Cat# 23081-1-AP, RRID: AB_2815024, 1:4000), anti-FAP (Proteintech, Cat# 84018-4-RR, 1:1000), anti-TSG101, anti-CD63, anti-HSPA6 (Proteintech, Cat# 13616-1-AP, RRID: AB_2120122, 1:4000), anti-JNK (Proteintech, Cat# 66210-1-Ig, RRID: AB_2881601, 1:3000), anti-phospho-JNK (Tyr185) (Proteintech, Cat# 80024-1-RR, RRID: AB_2882943, 1:1000), anti-JunD (Abcam, Cat# ab181615, RRID: AB_2864350, 1:2000), anti-phospho-JunD (S255) (Abcam, Cat# ab139180, RRID: AB_3661974, 1:4000), anti-E-cadherin (Proteintech, Cat# 20874-1-AP, RRID: AB_10697811, 1:10,000), anti-vimentin (Proteintech, Cat# 60330-1-Ig, RRID: AB_2881439, 1:20,000), anti-Snail, anti-MMP2 (Proteintech, Cat# 10373-2-AP, RRID: AB_2250823, 1:1000), anti-MMP9 (Proteintech, Cat# 27306-1-AP, RRID: AB_2880837, 1:1000), and anti-GAPDH (Proteintech, Cat# 60004-1-Ig, RRID: AB_2107436, 1:20,000). After washing with Tris-buffered saline with Tween-20 (TBST), membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (Proteintech Group) for 1 h at room temperature. Protein bands were visualized using an enhanced chemiluminescence (ECL) kit (Thermo Fisher Scientific, USA). Image analysis was performed using ImageJ (version 1.8.0, NIH, USA), and protein expression levels were normalized to GAPDH.

Exosome endocytosis assay

Exosomes were labeled with 10 µg/mL 3,3’-dioctadecyloxacarbocyanine perchlorate (DIO; Beyotime, Shanghai, China) for 30 min at room temperature. Excess dye was removed by washing three times with PBS, and labeled exosomes were resuspended in serum-free medium. GC cells were incubated with labeled exosomes for 6 h at 37 °C, fixed with 4% paraformaldehyde, and observed under a fluorescence microscope (Olympus, Japan).

Cell co-culture

Transwell plates (Corning, USA) were used for the co-culture of GC cells with NFs/CAFs, as well as with exosomes derived from NFs/CAFs. NFs/CAFs were seeded in the upper chamber, and GC cells were plated in the lower chamber. After 48 h of co-culture, GC cells in the lower chamber were harvested for subsequent experiments.

Cell counting kit-8 (CCK-8) assay

AGS and MKN-28 cells were seeded in 96-well plates at a density of 5 × 10³ cells/well in 100 µL complete medium, with three biological replicates per group. At 0, 24, 48, and 72 h after seeding, 10 µL CCK-8 reagent (Solarbio, Beijing, China) was added to each well, and cells were incubated at 37 °C in the dark for 1 h. Absorbance at 450 nm was measured using a microplate reader (Bio-Rad, USA).

EdU assay

The BeyoClick™ EdU-555 Cell Proliferation Assay Kit (Beyotime, Shanghai, China) was used to evaluate cell proliferation. EdU was added to the culture medium of AGS and MKN-28 cells and incubated for 2 h. Cells were then fixed, permeabilized, and stained with DAPI. Proliferating cells (EdU-positive) were stained red, and all nuclei (DAPI-positive) were stained blue. Images were captured using an inverted fluorescence microscope (Nikon, Japan). The proliferation rate was calculated as the ratio of EdU-positive cells to DAPI-positive cells, and results are presented as the mean ± standard deviation (SD) of three biological replicates.

Invasion and migration assays

Invasion and migration assays were performed using 24-well Transwell plates (Corning, USA). For the invasion assay, the upper chamber was coated with 200 µg/mL Matrigel (BD Biosciences, USA). GC cells were resuspended in serum-free medium at a density of 2 × 10⁵ cells/mL, and 100 µL cell suspension was added to the upper chamber. The lower chamber contained medium supplemented with 10% FBS as a chemoattractant. After 24 h of incubation, cells that migrated to the lower chamber were fixed with 95% ethanol and stained with 0.1% crystal violet. Non-invading cells on the upper membrane surface were removed using PBS and a cotton swab. Invading/migrating cells were counted under an inverted microscope (Olympus, Japan). The migration assay followed the same protocol, except the upper chamber was not coated with Matrigel. All experiments were performed with three biological replicates.

Wound healing assay

Cells were cultured in 6-well plates until reaching 90% confluence. A scratch was created in the cell monolayer using a 200 µL pipette tip. Cells were rinsed three times with PBS, and low-serum or serum-free medium was added. Images were captured at 0 and 24 h, and the scratch area was quantified using ImageJ. Three biological replicates were performed for each group.

Lung metastasis model and in vivo imaging

Four- to six-week-old BALB/c nude mice (15–20 g) were used to establish a lung metastasis model. AGS cells were co-cultured with fibroblasts from different groups (control, NFs, Men1-overexpressing NFs, CAFs, Men1-knockdown CAFs). A total of 1 × 10⁶ AGS cells were injected into mice via the tail vein. Fourteen days after injection, 15 mg/mL D-luciferin potassium salt (Beyotime Biotech, Shanghai) was administered intraperitoneally at a dose of 10 µL/g body weight. After 10–20 min (to allow maximum fluorescence stabilization), pulmonary fluorescence signals were detected using an in vivo imaging system (PerkinElmer/IVIS Spectrum, USA).

Histological analysis and immunohistochemistry

Lung tissues were fixed, embedded in paraffin, and sectioned. Hematoxylin and eosin (HE) staining was performed for histological examination. For immunohistochemistry, sections were deparaffinized in xylene, rehydrated through a graded ethanol series, and subjected to antigen retrieval in pH 9.0 EDTA buffer. After blocking, sections were incubated with primary antibodies against Menin (Bethyl, Cat# A300-105 A, RRID: AB_2143306, 1:500), HSPA6 (Proteintech, Cat# 13616-1-AP, RRID: AB_2120122, 1:200), phospho-JNK (Proteintech, Cat# 80024-1-RR, RRID: AB_2882943, 1:200), and phospho-JunD (Abcam, Cat# ab139180, RRID: AB_3661974, 1:200) at 4 °C overnight. HRP-conjugated secondary antibodies and a 3,3’-diaminobenzidine (DAB) staining kit (Zhongshan Golden Bridge, Beijing, China) were used for color development. Sections were counterstained with hematoxylin, mounted, and observed under a light microscope.

Statistical analysis

All data were analyzed using SPSS 23.0 software (IBM, Armonk, NY, USA; RRID: SCR_002865). Comparisons between groups were performed using independent-samples t-tests or one-way analysis of variance (ANOVA), followed by LSD or Tukey’s HSD post hoc tests if necessary. Graphs were generated using GraphPad Prism 9.0 software, and results are presented as the mean ± SD. A P-value < 0.05 was considered statistically significant.

Results

Menin is highly expressed in CAFs and CAFs-derived exosomes in GC

To investigate whether CAFs and their derived exosomes mediate GC progression by delivering Menin, we first isolated and identified primary CAFs and NFs from fresh GC tissues. Under light microscopy, both CAFs and NFs exhibited a spindle-shaped morphology (Fig. 1A). qRT-PCR analysis showed that the mRNA expression levels of α-SMA and FAP (specific markers of CAFs [21]) were 7.5-fold and 15.1-fold higher in CAFs than in NFs, respectively (Fig. 1B). Western blot analysis confirmed that the protein levels of α-SMA and FAP were 1.49-fold and 1.77-fold higher in CAFs than in NFs (Fig. 1C-D). Immunofluorescence staining further verified the increased expression of α-SMA and FAP in CAFs (Fig. 1E). Flow cytometry analysis showed that the purity of CAFs and NFs was 95.2% and 98.3%, respectively (Fig. 1F). To facilitate in vitro experiments, primary CAFs and NFs were immortalized by transfection with SV40 plasmids. Western blot analysis confirmed that Menin expression remained stable before and after immortalization, indicating that immortalized cells could be used for subsequent experiments (Supplementary Fig. 1A-B).

Fig. 1.

Identification and immortalization of primary CAFs and NFs from gastric cancer tissues. A. Light microscopy images of primary NFs and CAFs, showing spindle-shaped morphology. B. Identification of primary NFs and CAFs using qRT-PCR with the specific markers of α-SMA (P = 0.000245, t = 12.38) and FAP (P = 0.000088, t = 16.04). C-D. Identification of primary NFs and CAFs by using Western blot with the specific markers of α-SMA (P = 0.022252, t = 3.625) and FAP (P = 0.019006, t = 3.806) (C), and the quantification (D). E. Identification of primary NFs and CAFs by using immunofluorescence with the specific markers of α-SMA and FAP. F. Purity detection of primary NFs and CAFs by using flow cytometry with the specific marker of α-SMA. (NC: negative control, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001)

To identify differentially expressed genes between CAFs and NFs, RNA-seq was performed on primary CAFs and paired NFs from seven GC patients. The gene expression distribution of paired samples showed a median expression value of 3.67 (Fig. 2A), with stable and comparable expression levels across most samples, indicating good reproducibility. A heatmap of normalized gene expression (log₂(TPM + 1)) revealed distinct gene expression profiles between CAFs and NFs (Fig. 2B), which was further supported by hierarchical clustering. A volcano plot showed 24 upregulated and 35 downregulated genes in CAFs (|log₂FC| >1, FDR < 0.05) (Fig. 2C). GO enrichment analysis indicated that differentially expressed genes were enriched in biological processes related to tumor progression (e.g., “mesenchyme development,” “epithelial tube morphogenesis”), cellular components involved in exosome formation and transport (e.g., “endocytic vesicle membrane”), and molecular functions associated with extracellular matrix interactions (e.g., “scavenger receptor activity”) (Fig. 2D). GSEA revealed significant enrichment of the “EPITHELIAL_MESENCHYMAL_TRANSITION” pathway (Fig. 2E) and “MYOGENESIS” pathway (Fig. 2F) in CAFs, suggesting potential roles of CAFs in EMT induction and TME remodeling. Independent t-test and paired t-test analyses confirmed that Men1 expression was significantly higher in CAFs than in NFs (Fig. 2G-H). qRT-PCR and Western blot analyses further verified that Men1 mRNA and protein levels were 2.37-fold and 1.74-fold higher in CAFs than in NFs, respectively (Fig. 2I-K). Immunofluorescence staining showed increased expression of Menin and α-SMA in GC tissues compared with adjacent non-tumorous tissues (Fig. 2L).

Fig. 2.

Differential gene expression screening and functional analysis of CAFs and NFs in gastric cancer. A. Boxplot of gene expression across 7 paired NFs and CAFs isolated from gastric cancer. B. Heatmap of differential gene expression between NFs and CAFs (log2(TPM + 1) normalization). C. Volcano plot of upregulated (red) and downregulated (blue) genes between NFs and CAFs (|log₂FC| >1 and FDR < 0.05). D. GO enrichment analysis of differentially expressed genes. E. GSEA enrichment in the “EPITHELIAL_MESENCHYMAL_TRANSITION” pathway in CAFs. F. GSEA enrichment in “MYOGENESIS” pathway in CAFs. G. t-test analysis of Men1 expression tested by RNA-seq in CAFs vs. in NFs (P = 0.0028, t = 4.852). H. Paired t-test for Men1 expression tested by RNA-seq in CAFs vs. in NFs (P = 0.0081, t = 3.501). I. The Men1 expression verified by qRT-PCR in CAFs and NFs (P = 0.0008, t = 8.999). J-K. The α-SMA (P = 0.021795, t = 3.649) and Menin (P = 0.043882, t = 2.905) expression verified by Western blot in CAFs and NFs (J), and the quantification (K). L. Immunofluorescence Assay for Menin and α-SMA in Gastric Cancer and Adjacent Non-tumorous (NT) Tissues. M. The identification of exosomes isolated from CAFs under TEM. N-O. The identification of exosomes with the specific markers of TSG101 and CD 63, and the Menin (P = 0.0055, t = 5.464) expression of exosomes detected by Western blot (N), and the quantification (O). (TEM: Transmission electron microscopy, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001)

Exosomes isolated from CAFs and NFs exhibited typical bilayer vesicular structures under TEM (Fig. 2M). Western blot analysis confirmed the expression of exosome-specific markers (CD63 and TSG101) and showed that Menin was present in exosomes derived from both CAFs and NFs, with significantly higher levels in CAFs-derived exosomes (Fig. 2N-O). These results demonstrate that Menin is highly expressed in CAFs and their derived exosomes.

CAFs promote the proliferation, invasion, and migration of GC cells

CCK-8 assays revealed that CAFs significantly enhanced the proliferation of AGS and MKN-28 cells (Fig. 3A-B), which was further confirmed by EdU assays (Fig. 3C-D). The proliferation rate of AGS and MKN-28 cells co-cultured with CAFs was 0.53 ± 0.03 and 0.50 ± 0.06, respectively, which was significantly higher than that of cells co-cultured with NFs (P < 0.05). Transwell assays showed that CAFs significantly increased the migration and invasion of GC cells (Fig. 3E-F). Specifically, the migration rate of AGS and MKN-28 cells was increased by 1.55-fold and 1.63-fold, and the invasion rate was increased by 1.43-fold and 1.51-fold, respectively, compared with the NFs co-culture group. Wound healing assays further confirmed that CAFs accelerated the wound closure of GC cells (Fig. 3G-H). Western blot analysis showed that Menin protein levels were significantly higher in AGS and MKN-28 cells co-cultured with CAFs than in those co-cultured with NFs or the control group (Fig. 3I-K). These results indicate that CAFs promote the proliferation, invasion, and migration of GC cells, which is likely associated with the high expression of Menin in CAFs.

Fig. 3.

Co-culturing with CAFs enhances the proliferation, invasion and migration of gastric cancer cells. A-B. The proliferation of AGS (P = 0.0115, q = 3.782) and MKN-28 (P = 0.0065, q = 3.919) cells detected by CCK-8 assay after co-culturing with NFs and CAFs. C-D. The proliferation of AGS (P < 0.0001, F = 131.2) and MKN-28 (P < 0.0001, F = 83.85) cells detected by EdU incorporation assay after co-culturing with NFs and CAFs (C), and the quantification (D). E-F. The migration (AGS P = 0.0003, F = 44.37; MKN-28 P < 0.0001, F = 143.2) and invasion (AGS P < 0.0001, F = 105.7; MKN-28 P = 0.0004, F = 37.56) of gastric cancer cells detected by Transwell assay after co-culturing with NFs and CAFs (E), and the quantification (F). G-H. The migration of AGS (P < 0.0001, F = 298.2) and MKN-28 (P < 0.0001, F = 75.77) cells detected by Wound healing assay after co-culturing with NFs and CAFs (G), and the quantification (H). I-K. Western blot analysis of Menin expression in AGS (P = 0.0264, t = 3.434) and MKN-28 (P = 0.02, t = 3.748) cells after co-culturing with NFs and CAFs (I-J), and the quantification (K). (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001)

CAFs promote GC progression via Menin-containing exosomes

To investigate the mechanism by which CAFs promote GC progression, we constructed Men1-knockdown CAFs (CAFs-sh-Men1) and Men1-overexpressing NFs (NFs-OE-Men1) using lentiviral vectors (Supplementary Fig. 1C-E). qRT-PCR and Western blot analyses confirmed the efficiency of Men1 knockdown and overexpression (Supplementary Fig. 1F-H). Menin expression in exosomes derived from CAFs-sh-Men1 cells was 0.58-fold lower than that in control CAFs-derived exosomes, while Menin expression in exosomes derived from NFs-OE-Men1 cells was 2.38-fold higher than that in control NFs-derived exosomes (Fig. 4A-B). Fluorescence microscopy showed that GC cells efficiently internalized DIO-labeled exosomes derived from CAFs (Fig. 4C).

Fig. 4.

CAFs promote the proliferation, invasion and migration of gastric cancer cells through secreting Menin-containing exosomes. A-B. Western blot results of Menin expression in exosomes extracted from NFs-OE-Men1 (P < 0.0123, t = 4.34) and CAFs-sh-Men1 (P < 0.0042, t = 5.869) cells (A), and the quantification (B). C. Fluorescence uptake assay showing the uptake of DIO-labeled exosomes in gastric cancer cells. D-E. The proliferation of AGS (NFs group P = 0.0152, t = 4.072; CAFs group P = 0.0002, t = 12.33) and MKN-28 (NFs group P = 0.002, t = 7.145; CAFs group P = 0.0023, t = 6.926) cells after co-culturing with exosomes from NFs-OE-Men1 and CAFs-sh-Men1 detected by CCK-8 assay. F-G. The proliferation of AGS (NFs group P = 0.0273, t = 3.401; CAFs group P < 0.0001, t = 16.04) and MKN-28 (NFs group P = 0.0126, t = 4.309; CAFs group P < 0.0001, t = 21.75) cells after co-culturing with exosomes from NFs-OE-Men1 and CAFs-sh-Men1 detected by EdU incorporation assay (F), and the quantification (G). H-I. The migration and invasion of AGS (NFs group migration p = 0.001, t = 8.613; CAFs group migration P = 0.0343, t = 3.157; NFs group invasion P = 0.0137, t = 4.198; CAFs group invasion P = 0.0151, t = 4.081) and MKN-28 (NFs group migration P = 0.0032, t = 6.317; CAFs group migration P < 0.0011, t = 8.415; NFs group invasion P = 0.0002, t = 13.26; CAFs group invasion P < 0.0018, t = 7.403) cells after co-culturing with exosomes from NFs-OE-Men1 and CAFs-sh-Men1 detected by Transwell assay (H), and the quantification (I). J-K. The migration of AGS (NFs group P = 0.0273, t = 3.401; CAFs group P < 0.0001, t = 16.04) and MKN-28 (NFs group P = 0.0126, t = 4.309; CAFs group P < 0.0001, t = 21.75) cells after co-culturing with exosomes from NFs-OE-Men1 and CAFs-sh-Men1 detected by Wound healing assay (J), and the quantification (K). (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001)

To validate whether exosomal Menin directly modulates the malignant biological behaviors of GC cells, we treated AGS and MKN-28 cells with exosomes isolated from different fibroblast groups (NFs, NFs-OE-Men1, CAFs, and CAFs-sh-Men1) and subsequently evaluated cell proliferation, migration, and invasion.

CCK-8 and EdU assays showed that exosomes derived from NFs-OE-Men1 cells significantly promoted the proliferation of GC cells, while exosomes derived from CAFs-sh-Men1 cells inhibited GC cell proliferation (Fig. 4D-G). Transwell assays demonstrated that NFs-OE-Men1 exosomes enhanced the migration and invasion of GC cells, whereas CAFs-sh-Men1 exosomes had the opposite effect (Fig. 4H-I). Wound healing assays further confirmed that NFs-OE-Men1 exosomes accelerated wound closure, while CAFs-sh-Men1 exosomes delayed it (Fig. 4J-K). These results indicate that CAFs promote GC progression by secreting Menin-containing exosomes.

Menin-containing exosomes activate the HSPA6/JNK/JunD pathway and induce EMT in GC cells

GSEA results suggested that CAFs may promote GC progression by inducing EMT (Fig. 2E). Western blot analysis showed that co-culture of GC cells with NFs-OE-Men1 exosomes increased the expression of mesenchymal markers (Snail, Vimentin, MMP2, MMP9) and decreased the expression of the epithelial marker E-cadherin (Fig. 5A-B). In contrast, co-culture with CAFs-sh-Men1 exosomes downregulated the expression of mesenchymal markers and upregulated E-cadherin expression (Fig. 5A-C).

Fig. 5.

Menin-containing exosomes from CAFs promote the progression of gastric cancer through activating the HSPA6/JNK/JunD pathway and inducing EMT. C. Western blot analysis of EMT-related markers in gastric cancer cells after co-culturing with exosomes from NFs-OE-Men1 (Menin P = 0.019003, t = 3.806; E-cadherin P = 0.000012, t = 26.61; MMP2 P = 0.037423, t = 3.066; MMP9 P = 0.033413, t = 3.184; Vimentin P = 0.023377, t = 3.570; Snail P = 0.024516, t = 3.517) and CAFs-sh-Men1 (Menin P = 0.014440, t = 4.135; E-cadherin P = 0.000535, t = 10.13; MMP2 P = 0.006700, t = 5.160; MMP9 P = 0.002362, t = 6.861; Vimentin P = 0.049498, t = 2.786; Snail P = 0.019514, t = 3.775) (A), and the quantification (B-C). D-F. Western blot analysis of HSPA6/JNK/JunD pathway in gastric cancer cells after co-culturing with exosomes from NFs-OE-Men1 (Menin P = 0.019, t = 3.806; HSPA6 P = 0.005177, t = 5.544; JNK P = 0.505273, t = 0.7310; p-JNK P = 0.017783, t = 3.884; JUND P = 0.150165, t = 1.777; p-JUND P = 0.024534, t = 3.516) and CAFs-sh-Men1 (Menin P = 0.01444, t = 4.135; HSPA6 P = 0.03192, t = 3.232; JNK P = 0.433354, t = 0.8701; p-JNK P = 0.000065, t = 17.32; JUND P = 0.149762, t = 1.780; p-JUND P = 0.031861, t = 3.234) (D), and the quantification (E-F). (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001)

Western blot analysis also showed that NFs-OE-Men1 exosomes increased HSPA6 expression and enhanced the phosphorylation of JNK and JunD in GC cells (Fig. 5D-E). Conversely, CAFs-sh-Men1 exosomes decreased HSPA6 expression and reduced the phosphorylation of JNK and JunD (Fig. 5D-F). These results demonstrate that Menin-containing exosomes from CAFs promote GC progression by activating the HSPA6/JNK/JunD pathway and inducing EMT.

Menin-containing exosomes promote GC metastasis in vivo

A nude mouse model of lung metastasis was established to evaluate the in vivo effect of Menin-containing exosomes. In vivo imaging showed that CAFs-derived exosomes promoted lung metastasis, while CAFs-sh-Men1 exosomes inhibited it (Fig. 6A-D). HE staining and immunohistochemical analysis confirmed that mice treated with CAFs had more and larger metastatic lesions in the lungs, with increased expression of Menin, HSPA6, phospho-JNK, and phospho-JunD in metastatic foci. In contrast, the CAFs-sh-Men1 group had fewer and smaller metastatic lesions, with decreased Menin expression (Fig. 6E). These results confirm that Menin-containing exosomes from CAFs promote GC metastasis in vivo.

Fig. 6.

Menin-containing exosomes from CAFs promote the lung metastasis of gastric cancer in nude mice. A-B. Gross appearance of the mice 14 days after the caudal vein injection, and the lung tissues acquired from the mice. C-D. The tumor fluorescence intensity of pulmonary metastases tested by a small animal in vivo imaging system in the mice (C), and the results (NFs group P = 0.0463, t = 0.0463; CAFs group P < 0.0001, t = 7.175) were quantified (D). E. HE staining and immunohistochemical analysis of lung tissues in the mice. (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001)

Collectively, our findings suggest that CAFs secrete Menin-containing exosomes, which are internalized by GC cells, activate the HSPA6/JNK/JunD pathway, induce EMT, and ultimately promote GC progression (Fig. 7).

Fig. 7.

Working model elaborating how Menin-containing exosomes from CAFs promote gastric cancer progression. CAFs secrete exosomes enriched in Menin, which are taken up by gastric cancer cells. Within these cells, Menin upregulates EMT-related genes (Snail, Vimentin, MMP2, and MMP9) and downregulates E-cadherin to drive the EMT process. At the same time, Menin induces HSPA6 expression and enhances JNK phosphorylation, thereby activating JunD. The activated HSPA6/JNK/JunD axis further increases MMP9 expression. Collectively, these molecular changes drive the EMT, leading to more invasive and metastatic phenotypes and ultimately facilitating gastric cancer progression

Discussion

TME constitutes a dynamic ecosystem in which stromal cells—most notably CAFs—interact with tumor cells to drive the progression of various cancers, including GC [22]. These intercellular interactions facilitate the exchange of bioactive molecules through exosomes that act as critical mediators of tumor-stroma crosstalk [23]. In the present study, we demonstrate that Menin is highly expressed in CAFs isolated from GC tissues and in the exosomes secreted by these CAFs. Exosomes derived from these CAFs deliver Menin to GC cells, thereby promoting the proliferation, migration, and invasion of GC cells as well as inducing EMT in vitro, while also enhancing lung metastasis of GC in vivo. Mechanistically, this pro-tumor effect is mediated by the activation of the HSPA6/JNK/JunD signaling pathway, consistent with our previous findings regarding the oncogenic function of Menin in GC. These results underscore the protumorigenic role of exosomal Menin in mediating CAFs-GC crosstalk, thereby highlighting potential therapeutic strategies that target this CAFs-exosomal Menin-GC signaling axis.

As the most abundant stromal cell population in solid tumors, CAFs orchestrate a tumor-supportive microenvironmental niche by secreting bioactive factors that enhance the aggressiveness of tumor cells [24]. Our data further illustrate this concept. We show that CAFs promote malignant behaviors of GC cells via exosomes enriched in Menin and these exosomes are internalized by recipient GC cells, thereby amplifying their malignant phenotypic traits [25, 26]. Although our experiments focused on Menin as the primary functional cargo, it is reasonable to postulate that other molecular components within CAFs-derived exosomes synergize with Menin to potentiate the progression of GC. For example, exosomal CCT6A derived from CAFs has been shown to interact with β-catenin, thereby enhancing the tumorigenic potential of GC cells [13]; meanwhile, exosomal miR-4253 promotes the polarization of macrophages toward an M2-like phenotype, which in turn fosters a protumorigenic microenvironment [14]. Future studies should therefore profile additional molecular cargos within Menin-containing exosomes to elucidate the potential cooperative effects between these cargos and Menin.

The role of Menin in cancer is context-dependent: it functions as a tumor suppressor in endocrine tumors (e.g., pancreatic neuroendocrine neoplasms) [27], yet acts as an oncogene in other cancer types, including hepatocellular carcinoma, colorectal cancer, and now GC [18, 28]. Our findings further reinforce the protumorigenic activity of Menin in GC, wherein Menin is upregulated in CAFs and subsequently transmitted to GC cells via exosomes to activate downstream signaling pathways. However, this conclusion contrasts with a previous report, which indicated that Menin expression is low in certain GC samples and that Menin exerts a suppressive effect on GC cell proliferation through its interaction with IQGAP1 [29]. This discrepancy may stem from sample heterogeneity—including differences in patient cohorts (e.g., tumor subtypes, disease stages, or ethnic backgrounds)—methodological variations (e.g., protocols for tissue processing or types of detection assays), or the failure to capture microenvironmental influences in isolated cell line models. Furthermore, the conclusion that Menin exerts a suppressive effect on GC cell proliferation in the previous report is supported by incomplete experimental evidence. For instance, the study may have only evaluated the impact of Menin on cell proliferation using basic assays (e.g.,MTT assay) without investigating its effects on other key malignant phenotypes of GC cells, such as invasion, migration, or EMT—processes that are tightly linked to GC progression and metastasis. Additionally, it may not have validated the functional role of Menin-IQGAP1 interaction in vivo (e.g., using animal models of GC growth or metastasis) or explored whether this interaction is modulated by TME factors. Without comprehensive assessment of these critical aspects, the conclusion regarding Menin’s suppressive effect on GC proliferation remains insufficiently robust and may not reflect the full functional spectrum of Menin in the context of GC.

Mechanistically, exosomal Menin upregulates the expression of HSPA6 and promotes the phosphorylation of JNK and JunD; concurrently, it induces the expression of EMT-related markers (e.g., Snail, Vimentin, MMP2, and MMP9) while downregulating the expression of the epithelial marker E-cadherin. Existing literature suggests that HSPA6—a member of the heat shock protein (HSP) family—promotes malignant progression in gliomas and breast cancer [30, 31]. Our team’s preliminary experimental data, particularly the results from dual-luciferase reporter gene assays, provide clear evidence that Menin enhances the transcriptional activity of the HSPA6 gene, supporting its direct regulation of HSPA6 transcription. Subsequently, we conducted an in-depth analysis of the downstream pathways of HSPA6 by querying the KEGG database, and verified the JNK/JunD pathway through Western blot experiments. This observation supports the notion that the JNK/JunD signaling axis serves as a key conduit for Menin-driven malignant aggressiveness, given that JNK activation has been previously linked to EMT induction and tumor cell invasion in GC [32–34]. However, the specificity of this regulatory cascade remains unclear, as our study prioritized the investigation of the HSPA6/JNK/JunD axis based on the results of RNA-seq, without systematically validating other potential signaling pathways.

To gain deeper mechanistic insights, future experiments should investigate whether inhibiting JNK or JunD—either through pharmacological means (e.g., using the JNK inhibitor SP600125) or genetic approaches (e.g., siRNA-mediated knockdown)—can block exosomal Menin-induced EMT and tumor cell invasion. Supporting evidence includes studies showing that JNK inhibition reverses EMT in GC models driven by NKCC1 or HER2 overexpression [35, 36], as well as reports demonstrating that JNK blockade attenuates exosome-mediated EMT in other cancer types [35, 36]. Additionally, pharmacological disruption of exosome uptake—for example, using dynamin inhibitors, which block endocytic processes—could also reverse these malignant phenotypes, given that exosomes depend on endocytosis to deliver their cargos into recipient cells [9].

Our experiments observed that NFs-OE-Men1 and their secreted Menin-enriched exosomes also promote GC progression, which confirms our conclusion regarding the oncogenic properties of Menin. If Menin’s oncogenic role is so critical in driving GC progression via exosomal delivery, then could the regulatory axis of “Menin-containing exosomes and GC cells” be further validated in more physiologically relevant models that better reflect the in vivo TME? Our in vitro co-culture and exosome-based models effectively recapitulate interactions between CAFs/NFs and GC cells, but they only partially mimic the complexity of the in vivo TME—specifically, they lack immune cells, vascular components, and hypoxic microenvironments [37, 38]. Advanced experimental models—such as 3D organoids or microfluidic systems—better approximate these in vivo dynamic processes [39, 40], thereby highlighting the limitations of our 2D cell culture and Transwell-based experimental setups. In our nude mouse model of GC lung metastasis, we observed enhanced metastatic capacity of GC cells; however, due to the immunodeficient nature of nude mice, we were unable to evaluate immune cell infiltration into metastatic foci or immune responses to the tumor [41–43]. Given the established immunomodulatory roles of CAFs, future studies using immunocompetent mouse models should assess changes in specific immune cell populations—such as T cells or macrophages—and investigate how these changes interact with emerging immunotherapeutic strategies for GC [26].

From a therapeutic perspective, targeting Menin itself or the exosomal transport of Menin holds considerable promise for GC treatment, as this approach could potentially synergize with existing Menin-targeted inhibitors—such as Menin-MLL interaction blockers [44]. However, several challenges need to be addressed: first, the risk of off-target effects; second, the need to achieve delivery specificity to CAFs or exosomes; and third, the potential development of therapeutic resistance driven by TME heterogeneity. Notably, even subtle dysregulation in the translational regulation of Menin or exosomal transport could significantly amplify therapeutic responses [45, 46], underscoring the importance of precision therapeutic approaches. Thus, the identification of effective therapeutic targets—such as Menin in the CAF-exosomal Menin-GC axis—is crucial for advancing GC treatment [47].

In summary, Menin is highly expressed in CAFs and their exosomes in GC tissues, and CAFs promote the progression of GC by secreting Menin-containing exosomes, which are uptaked by GC cells to activate the HSPA6/JNK/JunD pathway and EMT-related signals.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Shan-hu Wang: Drafted the manuscript and completed some of the experiments. Jin-xun Jiang: Data collection and manuscript revision. Kai-tian Zheng: Completed part of the experiments. Shi-jie Zhang: Completed part of the experiments. Tian-de Chen: Statistical analysis. Zhen Wang: Designed, supervised and guided the study, and revised the manuscript.

Funding

This study was supported by the National Natural Science Foundation of China (82002492), the Natural Science Foundation of Guangxi (2018GXNSFBA281159), the Training Program for Excellent Medical Talent of the First Affiliated Hospital of Guangxi Medical University (202207), the Innovation Team of the First Affiliated Hospital of Guangxi Medical University (YYZS2022004), and the First-class discipline innovation-driven talent program of Guangxi Medical University.

Data availability

The raw data of the experiments can be provided in case.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of the First Affiliated Hospital of Guangxi Medical University. Institutional review board approval was obtained, and all the animal experiments were performed in accordance with the institutional guidelines of the Institutional Animal Care and Use Committee of the First Affiliated Hospital of Guangxi Medical University, Nanning, China.

Consent for publication

All authors consent to publication.

Competing interests

The authors declare that they do not have any competing financial interests or personal relationships that could have appeared to influence the work reported in this study.