Caveolin-1 drives immunosuppression in esophageal squamous cell carcinoma by enhancing exosome secretion and macrophage M2 polarization via inhibition of MVB autophagic degradation
Ke-Rong Zhai, Pei-Lin Zhao, Zi-Han Wang, Hai-Ming Feng, Zhen-Qing Li, Hui-Rong Huang, Ning Yang, Zhi-Peng Su, Bai-Qiang Cui, Tie-Niu Song, Bin Li
TL;DR
Caveolin-1 in esophageal cancer boosts exosome release and M2 macrophage polarization, creating an immunosuppressive environment that helps the tumor evade the immune system.
Contribution
Identifies caveolin-1 as a novel regulator of exosome secretion and macrophage polarization in esophageal squamous cell carcinoma.
Findings
Caveolin-1 knockdown reduces exosome secretion and impairs M2 macrophage polarization.
CAV1 inhibits autophagy via the PI3K/AKT/mTOR pathway to maintain exosome secretion.
CAV1 expression correlates with M2 macrophage infiltration and poor prognosis in ESCC patients.
Abstract
Exosomes are pivotal mediators of molecular transfer and intercellular communication, orchestrating interactions between tumor cells and the tumor microenvironment (TME). However, the impact of exosomes derived from esophageal squamous cell carcinoma (ESCC) on macrophages and the mechanisms regulating their secretion remain poorly understood. experiments were performed to evaluate the effects of ESCC-derived exosomes on macrophage polarization. Proteomics, bioinformatics analyses, and functional validation were employed to identify key molecules regulating exosome secretion. Stable knockdown cell lines for the candidate molecules, together with co-culture assays, were used to evaluate their effects on exosome secretion and macrophage polarization. Western blotting, transmission electron microscopy, and immunofluorescence were conducted to investigate the underlying mechanisms.…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8- —http://dx.doi.org/10.13039/501100001809National Natural Science Foundation of China
- —Science and Technology Key Research and Development Program of Gansu Province
- —Key Talent Project of Gansu Province
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsExtracellular vesicles in disease · Caveolin-1 and cellular processes · Immune cells in cancer
Introduction
Esophageal cancer remains a major global health challenge, ranking as the seventh most common malignancy worldwide with more than 470,000 new cases annually [1]. In 2022, approximately 445,000 esophageal cancer–related deaths were reported worldwide [2, 3], and more than 90% are esophageal squamous cell carcinoma (ESCC) [1]. The poor prognosis of ESCC is mainly due to its aggressive biology, therapeutic resistance, and an immunosuppressive tumor microenvironment (TME), resulting in a 5-year survival rate below 30% [4]. Although immune checkpoint inhibitors have achieved significant progress in several cancers, a large proportion of ESCC patients show little or no response and derive limited benefit. Immune evasion within the TME is thought to underlie this poor efficacy [5], highlighting the need to elucidate its mechanisms to develop novel therapeutic strategies and improve patient outcomes in ESCC.
Primary tumors create a permissive TME within affected organs and tissues [6]. Infiltration of inflammatory cells, particularly tumor-associated macrophages (TAMs), plays a critical role in shaping the TME [7]. As the most abundant immune cells in the TME, TAMs are highly plastic and often polarize into classically activated (M1) or alternatively activated (M2) phenotypes in response to TME–derived signals [8]. Accumulating evidence shows that M2-like macrophages suppress anti-tumor immunity, foster an immunosuppressive TME, and promote tumor progression [7, 9]. In ESCC, M2 macrophage infiltration is markedly increased compared with adjacent non-tumorous tissues [10] and strongly correlates with advanced stage, lymph node metastasis, and poor patient survival [11]. Thus, targeting infiltrating M2-like TAMs represents a promising therapeutic approach in ESCC.
Exosomes are nanoscale extracellular vesicles (30–160 nm in diameter) secreted through exocytosis. They carry bioactive molecules such as nucleic acids, proteins, and lipids derived from parent cells, and can be internalized by recipient cells to alter their phenotype and biological functions. Acting as key mediators of molecular transfer and intercellular communication, exosomes play a pivotal role in interactions between tumor cells and the TME [12]. Our previous study demonstrated that ESCC-derived exosomes promote the generation of M2-like macrophages through high mobility group box 1 (HMGB1), thereby suppressing CD8^+^ T cell–mediated immunity and facilitating ESCC immune evasion [13]. O’Brien et al. reported that targeting the exosome release pathway may serve as a promising antitumor therapeutic strategy [14]. However, the mechanisms regulating tumor-derived exosome release remain poorly understood, and no studies have specifically addressed the regulatory pathways involved in ESCC.
Caveolin-1 (CAV1) is an essential scaffold protein for caveolae formation and a key structural component of membrane lipid rafts. It regulates intracellular vesicle trafficking, exocytosis, and signal transduction [15], and has been implicated in tumorigenesis across diverse cancer types [16]. CAV1 has also been reported to promote the biogenesis and secretion of extracellular vesicles [17]. Kou et al. found that in mesenchymal stem cells, Fas interacts with Fas-associated phosphatase-1 (Fap-1) and CAV1 to activate an NSF–SNARE-mediated membrane fusion mechanism, thereby facilitating extracellular vesicle release [18]. However, the precise role and underlying mechanisms of CAV1 in exosome secretion remain unclear. This study investigates the role of CAV1 in exosome release from ESCC cells, elucidates its contribution to immune escape, and provides a rationale for identifying CAV1 as a potential therapeutic target and predictive biomarker.
Materials and methods
Clinical samples and tissue microarray (TMA) analysis
This study was approved by the Ethics Committee of The Second Hospital & Clinical Medical School, Lanzhou University (No.: 2020 A-032), and written informed consent was obtained from all participants. Liquid nitrogen-frozen ESCC tissues and matched adjacent normal tissues were collected from 10 randomly selected patients who underwent surgical resection at the Department of Thoracic Surgery, The Second Hospital & Clinical Medical School, Lanzhou University (82 Linxia Road, Lanzhou, Gansu, China; Postal Code 730030), and used for real-time quantitative polymerase chain reaction (RT-qPCR) and Western blotting (WB) analyses. All cancer tissue specimens were confirmed as ESCC by postoperative pathological diagnosis.
In addition, a human ESCC TMA (HEsoS180Su10) purchased from Shanghai Outdo Biotech Co., Ltd. (Shanghai, China) was used to evaluate CAV1 expression in tumor cells and CD163 expression in the tumor stroma. The TMA comprised 112 ESCC tumor tissues and 68 adjacent normal tissues. Clinicopathological data, including age, sex, clinical stage, differentiation grade, tumor size, and follow-up information, were available with patient consent. Multiplex immunofluorescence (IF) staining of the TMA was performed using antibodies against CAV1 (CST, #3267; 1:500), CD163 (Proteintech, #16646-1-AP; 1:1,000), and pan Cytokeratin (HUABIO, #HA601094; 1:500). TMA slides were scanned and analyzed for positive cell counts, positive staining areas, and colocalization using HALO^®^ image analysis software (Indica Labs, USA).
Cell culture and reagents
The ESCC cell lines KYSE150 and TE-1, together with the human monocytic leukemia cell line THP-1, were purchased from the Chinese Academy of Sciences (Shanghai, China). Cells were cultured in RPMI 1640 medium (Gibco, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, USA) and 1% penicillin–streptomycin (Gibco, USA) under conditions of 5% CO_2_ at 37 °C. For THP-1, the culture medium was further supplemented with 0.05 mM 2-mercaptoethanol (Gibco, USA).
To inhibit exosome release or autophagy, ESCC cells were treated with 5 µM GW4869 (MCE, #HY-19363, USA) for 24 h or 5 mM 3-methyladenine (3-MA; AbMole, #M2296, USA) for 12 h, respectively. Autophagy was induced with 100 nM rapamycin (Rapa; AbMole, #M1768, USA) for 24 h.
Macrophage induction
THP-1 cells were differentiated into M0-type macrophages by treatment with 100 ng/mL phorbol myristate acetate (PMA; Sigma–Aldrich, #16561-29-8, USA) for 24 h. For exosome co-culture, purified exosomes were added to the culture medium of M0 macrophages. After 48 h, cells were harvested for RNA extraction, fixed for IF staining, or co-cultured with CD8^+^ T cells.
Lentiviral-mediated gene knockdown and rescue experiments
Lentiviral vectors for CAV1, RAB34 and TSPAN1 knockdown were generated by cloning short hairpin RNA (shRNA) sequences specific to each gene into the pSIH1-CMV-GFP-Puro-H1-shRNA vector. Corresponding non-targeting control vectors (pSIH1-CMV-GFP-Puro-H1-shRNA-NC) were constructed in parallel. For CAV1 knockdown, shRNA1 targets the 3′ untranslated region (3′UTR) region of CAV1. The core targeting sequences of the shRNAs are provided in Supplementary Table S1. KYSE150 cells were transduced with the viruses at a multiplicity of infection (MOI) of 10. Transduced cells were selected and maintained in medium containing 2.5 µg/mL puromycin (AbMole, #M3637, USA). Knockdown efficiency was verified by RT-qPCR and WB.
For rescue experiments, the full-length human CAV1 coding sequence (CDS), lacking the 3′UTR, was cloned into a lentiviral expression vector under the control of the EF1α promoter with a blasticidin resistance cassette (Plenti-EF1α-hCAV1-P2A-BSD). An empty lentiviral vector with the same backbone and blasticidin resistance cassette but lacking the CAV1 insert was used as the rescue control. Stable CAV1-KD1 cells were subsequently infected with either the CAV1 rescue lentivirus or the empty control lentivirus, followed by blasticidin selection. Cells stably expressing both shRNA and rescue constructs were maintained under dual antibiotic selection. Because the shRNA targets the 3′UTR of endogenous CAV1, the re-expressed CAV1 CDS lacking the 3′UTR is resistant to shRNA-mediated knockdown. Re-expression efficiency was confirmed by WB. The full sequence of the CAV1 rescue construct is listed in the Supplementary Table S1.
Isolation and expansion of human CD8⁺ T cells
Peripheral blood mononuclear cells (PBMCs) were isolated from healthy donors by density gradient centrifugation using Lymphoprep™ (Cytiva, USA). Heparinized human peripheral blood was diluted 1:1 with phosphate buffered saline (PBS) and layered onto Lymphoprep™ solution, followed by centrifugation at 400 × g for 30 min. Mononuclear cells from the interphase were collected, washed with PBS to remove residual platelets and separation solution, and immediately used for subsequent experiments.
CD8^+^ T cells were purified from PBMCs using the EasySep™ Human CD8^+^ T Cell Isolation Kit (STEMCELL Technologies, #19661, Canada). Briefly, the PBMC suspension was adjusted to 1 × 10^7^ cells/mL, incubated with the CD8^+^ T cell isolation cocktail for 5 min, and then treated with RapidSpheres™ magnetic beads. The tube was placed on the EasySep™ Magnet (STEMCELL Technologies, #18000, Canada) for 3 min, and the supernatant enriched in CD8^+^ T cells was collected. The purity of the isolated CD8^+^ T cells was assessed by flow cytometry using anti-human CD3 and CD8 antibodies and routinely exceeded 90%. The sorted CD8^+^ T cells were cultured in ImmunoCult™-XF T cell expansion medium supplemented with CD3/CD28 T cell activators (STEMCELL Technologies, #10971, Canada) and interleukin-2 (IL-2, STEMCELL Technologies, #78036, Canada) to initiate T cell activation and support their expansion.
Isolation and differentiation of PBMC-derived M0 macrophages
PBMCs were isolated as described above, and CD14^+^ monocytes were purified using anti-CD14 magnetic beads (EasySep™ Human CD14 Positive Selection Kit, STEMCELL Technologies, #17858, Canada). Monocytes were cultured in RPMI 1640 containing 100 ng/mL macrophage colony-stimulating factor (M-CSF; STEMCELL Technologies, #78150, Canada) for 5 days to induce differentiation into M0 macrophages. Differentiation and purity were confirmed by flow cytometry using CD14, CD68, and CD11b antibodies (BioLegend, USA), and cells expressing macrophage markers were used for subsequent experiments.
Exosome isolation and identification
Exosomes were isolated and purified from human ESCC cell culture supernatants by differential ultracentrifugation, following the protocol described by Hua et al. [19]. Cells were cultured in RPMI 1640 medium supplemented with 10% exosome-depleted FBS (prepared by ultracentrifugation at 100,000 × g for 8 h) for 48 h. The supernatant was collected and centrifuged at 300 × g for 10 min to remove cells, followed by sequential centrifugation at 2,000 × g for 10 min and 10,000 × g for 30 min to eliminate residual cells and debris. The supernatant was filtered through a 0.22 μm filter (Biosharp, #BS-PES-22, China) and ultracentrifuged at 100,000 × g for 70 min (Beckman Coulter, USA). The pellet was resuspended in PBS and ultracentrifuged again at 100,000 × g for 70 min. The final exosome precipitate was collected and resuspended in PBS.
Exosomal protein concentration was measured using a bicinchoninic acid (BCA) protein assay kit (Solarbio, #PC0020, China). Exosome size distribution and particle concentration were analyzed by nanoparticle tracking analysis (NTA, Zetaview, Germany). Morphological features were examined by transmission electron microscope (JEOL, Japan), and exosomal marker proteins were detected by WB.
Co-culture assays
To examine the impact of ESCC cells on macrophage polarization, GW4869-pretreated ESCC cells and PMA-induced M0 macrophages from THP-1 cells were co-cultured using a Transwell^®^ co-culture system with a 0.4 μm polyester membrane (Corning, #3450, USA). To evaluate the influence of macrophages on CD8^+^ T cell function, PBS- or exosome-treated macrophages were co-cultured with CD8^+^ T cells in the Transwell system. After 48 h of co-culture, cells from the lower chamber were harvested for further analysis.
In vivo xenograft tumor model
All animal procedures were approved by the Animal Experiment Ethics Committee of The Second Hospital & Clinical Medical School, Lanzhou University (No.: D2020-25) and were conducted in accordance with the NIH guidelines. Female BALB/c nude mice (3–4 weeks old) were obtained from Huachuang Sino Pharmaceutical Technology Co., Ltd. (China). KYSE150 cell lines stably transfected with control vectors or CAV1 shRNA constructs (shRNA1 and shRNA2) were used for animal experiments. Mice were randomly divided into three groups (n = 7 per group): CAV1-Ctrl, CAV1-KD1, and CAV1-KD2. KYSE150 cells (6 × 10^6^) were mixed with THP-1–derived macrophages at a 5:1 ratio and subcutaneously injected into the right axilla of nude mice [20]. Tumor volume was calculated as V = (L × W^2^)/2, where V represents tumor volume (mm^3^), L is the longest diameter (mm), and W is the perpendicular shorter diameter (mm). After 4 weeks, mice were euthanized, and tumor tissues were harvested for IF staining to evaluate macrophage polarization.
Transmission electron microscopy (TEM)
The cell samples were fixed in 2.5% glutaraldehyde (Ted Pella, #18427, USA) for 2 h, washed three times with 0.1 M phosphate buffer, and subsequently fixed in 1% osmium acid (Ted Pella, #18456, USA) at 4 °C for 2 h. After dehydration through a graded ethanol series, the samples were embedded in Epon-Araldite resin (Ted Pella, #18010, USA) for infiltration and placed in molds for polymerization. Semithin sections were first prepared for positioning, followed by ultrathin sectioning for microstructural analysis. The sections were counterstained with 3% uranyl acetate (Electron Microscopy Sciences, #22400, USA) and 2.7% lead citrate (HEAD biotechnology, #HD17810, China), and finally examined using the JEM-1400 transmission electron microscope (JEOL, Japan).
Exosome labeling and internalization
Purified exosomes were labeled with PKH26 (red) kit (Sigma–Aldrich, #MIDI26, USA). Exosomes were resuspended in 1 mL of Diluent C, mixed with 4 µL PKH26 dye prepared in 1 mL Diluent C, and incubated for 5 min at room temperature. To bind excess dye, 2 mL of FBS was added. Labeled exosomes were collected by centrifuging at 100,000 × g for 70 min to remove unbound dye, resuspended in serum-free medium, and co-cultured with THP-1–derived M0 macrophages for 12 h. After incubation, cells were fixed with 4% paraformaldehyde and stained with 4,6-Diamidino-2-phenylindole dihydrochloride (DAPI; Solarbio, #S2110, China) and 3,3’-Dioctadecyloxacarbocyanine perchlorate (DiO; UElandy, #N4021, China) to visualize nuclei and membranes. Exosome uptake was assessed using fluorescence microscopy (Zeiss, Germany).
Protein sequencing
Tandem Mass Tag (TMT)-based quantitative proteomic analysis was conducted by Beijing Novogene Technology Co. Total proteins were extracted from harvested cells or isolated exosomes, followed by tryptic digestion, TMT peptide labeling, and quality control. Labeled peptides were analyzed by LC–MS/MS on a Q Exactive™ HF-X platform.
Exosomal MiRNA sequencing
Exosomal small RNAs were extracted from purified exosomes and assessed for quality. Libraries were constructed and sequenced on an Illumina platform by Genewiz (Suzhou, China). Differentially expressed miRNAs were identified based on normalized sequencing data.
T-cell proliferation assays
CD8^+^ T cells were resuspended in PBS at a final concentration of 1 × 10^6^ cells/mL, and labeled with 5 µM carboxyfluorescein diacetate succinimidyl ester (CFSA) working solution (Invitrogen, #C34570, USA), followed by incubation at 37 °C in the dark for 20 min. The staining reaction was quenched by adding five volumes of pre-chilled complete medium containing 10% FBS. After washing with PBS to remove unbound dye, stained cells were seeded into the lower chamber of a Transwell system (Corning, USA). M0 macrophages derived from THP-1 cells were placed in the upper chamber, and the two cell types were co-cultured for 48 h. CD8^+^ T cells were subsequently collected and analyzed by flow cytometry. Proliferation was assessed by CFSE dye dilution in daughter cells, and proliferation peaks were analyzed using FlowJo™ software (v10.4; BD Life Sciences, USA).
Flow cytometry
To analyze CD8^+^ T cell surface markers, cells were resuspended in FACS buffer and counted. Cells were incubated with Fc receptor blocking solution (BioLegend, #422301, USA) at 4 °C for 15 min, followed by staining with anti-human CD3 and CD8 antibodies (BioLegend, #344855 and #344705, USA) at 4 °C in the dark for 30 min. After washing, cells were immediately analyzed on the CytoFLEX flow cytometer (Beckman Coulter, USA). The proportion of CD8^+^ T cells within the CD3^+^ population was quantified using FlowJo™ software (v10.4; BD Life Sciences, USA).
RNA extraction and qRT-PCR
Total RNA was extracted by TRIzol reagent (Invitrogen, #15596026CN, USA), and RNA concentration was measured with NanoDrop (Thermo Fisher Scientific, USA). Equal amounts of RNA were subjected to one-step qRT-PCR using the BeyoFast™ SYBR Green One-Step qRT-PCR Kit (Beyotime, #D7268M, China) and gene-specific primers on an Applied Biosystems 7500 Fast Real-Time PCR System (Applied Biosystems, USA). GAPDH was used as the housekeeping gene, and relative gene expression levels were calculated using the 2^−ΔΔCT^ method. The primers used in this study are described in Supplementary Table S2.
WB
Protein extracts were resolved by 8–12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Following blocking with 5% non-fat milk (BD Difco, #232100, USA) or a commercial blocking solution (Yoche, #YWB0501, China), membranes were incubated overnight at 4 °C with primary antibodies against Calnexin (Proteintech, #66903-1-Ig; 1:10,000), CD9 (Proteintech, #60232-1-Ig; 1:30,000), CD63 (Proteintech, #67605-1-Ig; 1:12,000), TSG101 (Proteintech, #67381-1-Ig; 1:40,000), CAV1 (HUABIO, #ET1603-1; 1:2000), RAB34 (ABclonal, #A10332; 1:2000), TSPAN1 (Proteintech, #16058-1-AP; 1:500), EEA1 (ABclonal, #A0592; 1:1000), LC3B (ABclonal, #A19665; 1:2000), p62 (Selleck, #F0106; 1:10000), p-PI3K (Abmart, #TA3241; 1:500), PI3K (Abmart, #T40115; 1:1000), p-Akt (Abmart, #T40067; 1:1000), pan-Akt (Selleck, #F0004; 1:1000), p-mTOR (Selleck, #F2584; 1:5000), mTOR (Selleck, #F0169; 1:1000), p-AMPK (Selleck, #F0151; 1:1000), AMPK (Selleck, #F0269; 1:1000), and β-actin (Proteintech, #66009-1-Ig; 1:50,000). After incubation with horseradish peroxidase (HRP)-conjugated secondary antibodies (Proteintech, #SA00001-1 or #SA00001-2; 1:5,000) for 1 h at room temperature, protein bands were visualized with an enhanced chemiluminescence (ECL) kit (Biosharp, #BL523, China) using a chemiluminescence imaging system (Tanon, China).
IF staining
Cells were fixed with 4% paraformaldehyde (Biosharp, #BL539A, China) for 15 min, and permeabilized with 0.1% Triton X-100 (Beyotime, #P0096, China) for 10 min. After washing, cells were blocked with QuickBlock™ blocking solution (Beyotime, #P0260, China) and incubated with primary antibodies overnight at 4 °C. Fluorescent secondary antibodies (Proteintech, #RGAM005 or #RGAR004; 1:600) were applied and incubated for 1 h. After staining with DAPI, samples were imaged using the Zeiss LSM880 confocal microscope (ZEISS, Germany). Fluorescent intensity and colocalization were quantified using ImageJ software (v1.53a; NIH, USA).
Enzyme-linked immunosorbent assay (ELISA)
The concentrations of IL-10, transforming growth factor-β1 (TGF-β1), IL-6, granzyme B (GZMB), and interferon-γ (IFN-γ) in the supernatant were measured using human ELISA kits (BOSTER, China). All procedures were performed according to the manufacturer’s instructions.
Cell proliferation assay
Proliferation of CAV1-Ctrl and CAV1-KD cells was evaluated using a Cell Counting Kit-8 (CCK-8; AbMole, #M4839, USA). Cells (5 × 10^3^/well) were seeded into 96-well plates (n = 5 per group). At the indicated time points, CCK-8 reagent was added and incubated for 1 h at 37 °C, and absorbance was measured at 450 nm.
5-ethynyl-2′-deoxyuridine (EdU) incorporation assay
Cell proliferation was assessed by EdU incorporation assay using an EdU Cell Proliferation Kit (Abbkine, #KTA2031, China). CAV1-Ctrl and CAV1-KD cells were incubated with 10 µM EdU for 2 h at 37 °C. After fixation and permeabilization, nuclei were counterstained with Hoechst 33,342, and EdU-positive cells were visualized by fluorescence microscopy. The proportion of EdU-positive cells was quantified using ImageJ software (v1.53a; NIH, USA).
Ki-67 immunohistochemical staining
Subcutaneous xenograft tumor tissues derived from the CAV1-Ctrl and CAV1-KD groups were excised, fixed, paraffin-embedded, and sectioned. Sections were incubated with a primary antibody against Ki-67 (Cell Signaling Technology, #9449; 1:1000) at 4 °C overnight. Immunoreactivity was visualized using a DAB detection kit (ZSGB-BIO, #ZLI-9018, China), followed by hematoxylin counterstaining. Ki-67-positive nuclei were examined under a light microscope and quantified using Fiji software (v1.54p; NIH, USA).
Statistical analysis
All statistical analyses and figures were performed using R (v4.3.2; R Foundation for Statistical Computing, Austria) and GraphPad Prism (v10.1.1; GraphPad Software, USA). Data are presented as mean ± standard deviation or median (interquartile range, 25th–75th percentile). For normally distributed data, differences between two groups were assessed using Student’s t-test, while differences among three or more groups were evaluated using one-way ANOVA followed by Tukey’s post hoc test. For non-normally distributed data, differences between two groups were assessed using the Mann–Whitney U test, and differences among three or more groups were evaluated using the Kruskal–Wallis test followed by Dunn’s multiple comparison test. Correlations between continuous variables were analyzed using Pearson or Spearman correlation. Differences in categorical variables were assessed using Pearson’s chi-square test or Fisher’s exact test. Survival analyses were performed using Kaplan–Meier curves, with differences between groups assessed by the log-rank test. Two-sided P < 0.05 were considered statistically significant.
Results
ESCC-derived exosomes promote macrophage M2 polarization in a concentration-dependent manner
To investigate the role of exosomes in modulating the tumor immune microenvironment, the human THP-1 cell line was used as a representative monocyte-macrophage model to evaluate the effects of ECSS-derived exosomes on macrophage polarization. THP-1 monocytes were differentiated into M0 macrophages following PMA treatment, as confirmed by morphological changes from a round suspension state to adherent cells (Supplementary Figure S1A) and by increased mRNA expression of macrophage markers CD11b and CD68 (P < 0.001; Supplementary Figure S1B). When co-cultured with exosomes derived from KYSE150 and TE-1 cells respectively, PKH26-labeled exosomes (red) were clearly detected within macrophage cytoplasm, confirming their uptake (Fig. 1A). Compared with PBS-treated controls, exosome-treated macrophages exhibited significantly higher mRNA expression of M2 markers (CD163, CD206, and Arginase-1 (Arg1)), while M1 markers (IL-1β and inducible nitric oxide synthase (iNOS)) remained unchanged (Fig. 1B). ELISA analysis showed significantly higher levels of M2 markers (TGF-β1 and IL-10) in the supernatant of the exosome-treated group, while no significant change was observed in the M1 marker IL-6 (Fig. 1C–E). These findings were further supported by IF analyses, which demonstrated enhanced M2 marker expression (Fig. 1F).
Fig. 1ESCC-derived exosomes induce macrophage M2 polarization in a dose-dependent manner. (A) IF imaging showing internalization of ESCC-derived exosomes in macrophages; Scale bar = 5 μm. (B) qRT-PCR of M1 (IL-1β, iNOS) and M2 (CD163, CD206, Arg-1) markers after PBS or exosomes treatment (n = 6 per group). (C–E) ELISA of M1 (IL-6) and M2 (TGF-β1, IL-10) marker levels (n = 6 per group). (F) Representative IF images of M2 markers (CD163, CD206) in macrophages (n = 3 per group); Scale bar = 20 μm. (G–I) qRT-PCR of M2 markers in macrophages treated with high, medium, or low exosome concentrations (n = 3 per group). (J,** L)** qRT-PCR of M1 and M2 markers in macrophages co-cultured with ESCC cells pretreated with PBS, DMSO, or GW4869 (n = 6 per group). (K,** M)** ELISA of M1 and M2 marker levels in macrophages co-cultured with ESCC cells (n = 3 per group). Data are presented as mean ± SD. Arg1 mRNA data in (D) were analyzed using the Kruskal–Wallis test with Dunn’s post hoc test. Other comparisons were performed using Student’s t-test or one-way ANOVA with Tukey’s post hoc test. ns, not significant; *P < 0.05; **P < 0.01; ***P < 0.001
Importantly, M2 polarization was concentration-dependent, with higher exosome doses inducing stronger effects (Fig. 1G–I), whereas the expression of M1 markers remained largely unchanged (Supplementary Figure S1C, D). Moreover, pretreatment of ESCC cells with GW4869, an inhibitor of exosome synthesis and release, significantly impaired the ability of their exosomes to drive M2 polarization (Fig. 1J–M). Together, these results indicate that ESCC-derived exosomes are efficiently internalized by macrophages and promote M2 polarization in a dose-dependent manner.
Increased exosome release underlies enhanced macrophage M2 polarization by KYSE150 cells
Exosomes were isolated from the supernatants of KYSE150 and TE-1 cells by differential centrifugation (Fig. 2A). TEM revealed typical disc- or cup-shaped vesicles (30–160 nm) in both groups (Fig. 2B). WB confirmed the presence of exosomal markers (TSG101, CD63, CD9) and the absence of the negative marker Calnexin, with higher levels of positive markers in KYSE150-derived exosomes (Fig. 2C). NTA showed that KYSE150-derived exosomes had a diameter of 124.2 ± 4.9 nm and an average concentration of 3.8 × 10^10^ particles/mL, whereas TE-1-derived exosomes measured 141.9 ± 5.1 nm with an average concentration of 1.6 × 10^10^ particles/mL. Both groups exhibited normal size distributions (Fig. 2D), but exosome concentrations were significantly higher in KYSE150 than in TE-1 (Fig. 2D, E). These findings confirmed the exosomal nature of the isolated particles. Consistently, BCA analysis further demonstrated that KYSE150 cells produced more exosomes under identical conditions (Fig. 2F).
Fig. 2KYSE150 cells release more exosomes. (A) Schematic of the exosome isolation process. (B) Representative TEM images showing the morphology and size of ESCC-derived exosomes; Scale bar = 100 nm. (C) WB of exosome markers (TSG101, CD63, CD9) and negative marker Calnexin (n = 7 per group). (D,** E)** NTA analysis of particle size distribution and concentration in KYSE150 and TE-1 cell-derived exosomes (n = 3 per group). (F) BCA assay of protein content in KYSE150 and TE-1 cell-derived exosomes (n = 28 per group). Data are presented as mean ± SD and analyzed using Student’s t-test. ns, not significant; *P < 0.05; **P < 0.01; ***P < 0.001
Exosome-induced M2 polarization of macrophages was concentration dependent. Notably, exosomes isolated from equal cell numbers and the same volume of conditioned medium were more abundant in the KYSE150 group than in the TE-1 group, and these KYSE150-derived exosomes exhibited a stronger ability to induce M2 polarization in macrophages. To determine whether this difference was mainly due to exosome quantity rather than compositional variations, we standardized the concentrations of exosomes from both cell lines using NTA before co-culturing them with macrophages. At equal concentrations, KYSE150-derived exosomes still induced slightly stronger M2 polarization, but the difference between the two groups was reduced (Supplementary Figure S2). These results indicate that differences in exosome secretion capacity are the primary factor underlying their effects on macrophage polarization.
CAV1 identified as a key regulator of exosome secretion in ESCC cells
Exosome secretion is a multistep process regulated by various molecular factors. It starts with plasma membrane invagination forming early endosomes, which mature into late endosomes and then multivesicular bodies (MVBs). Within MVBs, endosomal membranes invaginate to generate intraluminal vesicles (ILVs). Mature MVBs fuse with the plasma membrane to release ILVs as exosomes into the extracellular space. Alternatively, MVBs can fuse with autophagosomes or lysosomes, leading to degradation [12]. Based on previous studies, we summarized 43 regulatory factors implicated in exosome secretion (Supplementary Table S3).
To explore the molecular mechanisms regulating exosome secretion in ESCC, we performed TMT-based quantitative proteomics to compare protein expression profiles between KYSE150 (high secretion) and TE-1 cells (low secretion). Differentially expressed proteins were identified with a fold-change > 2.0 (|log2FC| ≥ 1.5, and P < 0.01) and visualized in heatmap (Fig. 3A) and volcano plot (Fig. 3B). A total of 102 proteins were upregulated in KYSE150 cells. Notably, no significant differences were observed in canonical exosome-regulatory proteins between the two groups. However, gene set enrichment analysis (GSEA) revealed significant enrichment of the “regulation of intracellular transport” pathway (GO: 0032386) in KYSE150 cells (normalized enrichment score (NES) = 1.718, P = 1.089 × 10^− 4^, FDR q = 4.504 × 10^− 4^) (Fig. 3C; Additional file 3). Among all significantly enriched GO Biological Process terms, this pathway exhibited a clear positive enrichment signal. Although it was not the pathway with the highest NES, the vesicle trafficking and transport processes included in this pathway are directly relevant to exosome biogenesis and secretion. In contrast, no GO terms directly associated with exosome biogenesis or secretion were significantly enriched in TE-1 cells. ESCC may possess a unique regulatory network governing exosome release. GO enrichment analysis identified differential proteins between KYSE150 and TE-1 cells as significantly enriched in the “ER-to-Golgi transport vesicle membrane”, “transport vesicle membrane”, and “secretory granule membrane” (Supplementary Figure S3A). Venn analysis of our proteomic dataset with the exosome gene set identified 64 overlapping proteins, including CAV1, Ras-related protein Rab-34 (RAB34), and Tetraspanin-1 (TSPAN1) (Fig. 3D). To date, no studies have directly reported roles for these proteins in exosome secretion. However, CAV1, a known regulator of extracellular vesicle biogenesis and secretion [17], may also contribute to exosome release. RAB34 belongs to the Rab GTPase family, several members of which (e.g., Rab11, Rab27a, Rab27b) regulate exosome release. Tetraspanins are transmembrane proteins enriched in exosomes, and TSPAN1, as one member, may regulate exosome release through an endosomal sorting complex required for transport (ESCRT)-independent pathway [21]. Based on proteomic data and prior literature, CAV1, RAB34, and TSPAN1 were prioritized as candidates. Their expression in KYSE150 and TE-1 cells was validated by WB and qRT-PCR (Supplementary Figure S3B, C).
Fig. 3CAV1 regulates exosome secretion in ESCC. (A,** B)** Heatmap and volcano plot of differentially expressed proteins between KYSE150 and TE-1 cells. (C) GSEA showing significant positive enrichment of “regulation of intracellular transport” in KYSE150 cells. (D) Venn diagram showing overlapping proteins between the “exosome gene product set” and the differentially expressed protein set. (E) WB of CAV1 expression in CAV1-KD cells and controls (n = 3 per group). β-actin used for normalization. (F) qRT-PCR of CAV1 expression in CAV1-KD cells and controls (n = 6 per group). (G) NTA analysis of exosome particle concentrations in KYSE150 cells after knockdown of CAV1, RAB34, or TSPAN1 (n = 3 per group). (H) Representative particle size distribution of exosomes from CAV1-KD cells. (I) BCA assay of exosome protein content after knockdown of the three genes (n = 3 per group). (J) Representative TEM image of exosomes from CAV1-KD cells; Scale bar = 100 nm. (K,** L)** NTA analysis of exosome particle concentrations (n = 3 per group). (M) BCA assay of exosome protein content after CAV1 rescue in CAV1-KD1 cells (n = 3 per group). Data are presented as mean ± SD and analyzed using one-way ANOVA followed by Tukey’s post hoc test. ns, not significant; **P < 0.01; ***P < 0.001
To assess their regulatory roles, stable knockdown KYSE150 cell lines were generated for CAV1, RAB34, and TSPAN1. Knockdown efficiency was confirmed by WB and qRT-PCR (Fig. 3E, F; Supplementary Figure S3D–G). Exosomes were then isolated from equal cell numbers cultured under identical conditions and from equal volumes of supernatant. NTA revealed that exosome particle numbers were markedly reduced in the CAV1-KD group (1.6 × 10^10^ particles/mL) compared with controls (4.4 × 10^10^ particles/mL, P = 0.003) (Fig. 3G). Consistently, BCA assays confirmed that exosomal protein content was reduced in the CAV1-KD group (Fig. 3I). In contrast, exosome yields in the RAB34- and TSPAN1-KD groups did not differ significantly from controls. Furthermore, NTA and TEM analyses (Fig. 3H, J) revealed that exosome size and morphology remained unchanged after CAV1 knockdown. We further performed CAV1 rescue in CAV1-KD1 cells. WB showed that re-expression of wild-type CAV1 (CAV1-KD1 + CAV1-WT) effectively restored CAV1 protein levels, comparable to the CAV1-Ctrl group (Supplementary Figure S3H). Correspondingly, NTA and BCA indicated that both the particle number and total protein content of exosomes were significantly increased following CAV1 rescue compared with the knockdown group (Fig. 3K–M). Collectively, these results identify CAV1 as a key regulator of exosome secretion in ESCC cells.
CAV1-dependent exosome secretion drives macrophage M2 polarization and suppresses CD8+ T cell function
To test whether CAV1 regulates macrophage polarization via exosomes and thereby influences the tumor immune microenvironment, exosomes were isolated from equal volumes of conditioned medium collected from equal numbers of CAV1-Ctrl and CAV1-KD cells, and co-incubated with THP-1–derived M0 macrophages respectively. M2 macrophage markers were assessed by qRT-PCR, ELISA, and IF (Fig. 4A–D). Exosomes from CAV1-KD cells showed a reduced ability to induce M2 polarization compared with controls. In addition, Exosomes from the CAV1 rescue group significantly restored M2 polarization of macrophages compared with the CAV1-KD group, with M2 marker levels comparable to those in the CAV1-Ctrl group (Fig. 4E–G). The roles of RAB34 and TSPAN1 were evaluated under the same conditions, but knockdown of RAB34 or TSPAN1 individually did not significantly affect exosome-mediated M2 polarization (Supplementary Figure S4A, B).
Fig. 4CAV1-KD exosomes reduce macrophage M2 polarization and enhance CD8^+^ T-cell activity. (A–C) qRT-PCR and ELISA of M2 markers in macrophages after treatment with CAV1-Ctrl or CAV1-KD exosomes (n = 3 per group). (D) Representative IF images of M2 markers in macrophages after CAV1-Ctrl or CAV1-KD exosome treatment (n = 3 per group); Scale bar = 50 μm. (E–G) Macrophages treated with CAV1-Ctrl, CAV1-KD, or CAV1 rescue exosomes were analyzed for M2 marker expression by ELISA and qRT-PCR (n = 3 per group). (H) Representative flow cytometric assessment of the purity of primary macrophages. (I–K) qRT-PCR and ELISA of M2 markers in primary macrophages after treatment with CAV1-Ctrl or CAV1-KD exosomes (n = 3 per group). (L) Schematic workflow of co-culturing macrophages treated with control or CAV1-KD exosomes with CD8^+^ T cells. (M) Representative flow cytometric assessment of the purity of isolated CD8^+^ T cells (n = 3 per group). (N) Representative CFSE assay assessing CD8^+^ T-cell proliferation in PBS, CAV1-Ctrl, and CAV1-KD groups (n = 3 per group). (O,** P)** ELISA of IFN-γ and GZMB levels in CD8^+^ T-cell supernatants from PBS, CAV1-Ctrl, and CAV1-KD groups (n = 3 per group). Data are presented as mean ± SD and analyzed using one-way ANOVA followed by Tukey’s post hoc test. ns, not significant; *P < 0.05; **P < 0.01; ***P < 0.001
To extend our findings beyond the THP-1 cell line model and evaluate their translational relevance, we investigated the effect in human primary macrophages. CD14^+^ monocytes were isolated from PBMCs of healthy donors and differentiated into M0 macrophages, as confirmed by flow cytometry (Fig. 4H). Under identical conditions, qRT-PCR and ELISA showed that primary macrophages treated with exosomes from the CAV1-KD group exhibited significantly reduced M2 marker expression compared with the CAV1-Ctrl group (Fig. 4I–K). These results confirm that the impaired exosome-mediated M2 polarization due to CAV1 knockdown is reproduced in physiologically relevant primary human macrophages.
To assess the impact of CAV1 knockdown on the composition of exosomes, we performed proteomic and miRNA sequencing of exosomes from CAV1-Ctrl and CAV1-KD1 cells. The results demonstrated that overall cargo composition was highly similar (detailed data are provided in Additional Files 4 and 5). To further separate the effects of exosome secretion from cargo differences on macrophage polarization, exosomes from CAV1-Ctrl and CAV1-KD1 cells were normalized by NTA and co-cultured with M0 macrophages. Under these equal-quantity conditions, exosomes from the two groups showed no significant difference in their ability to induce M2 polarization (Supplementary Figure S4C).
To further assess whether CAV1 influences CD8^+^ T cell function through exosome-mediated macrophage polarization, macrophages were treated with PBS, CAV1-Ctrl exosomes, or CAV1-KD exosomes, and then co-cultured with CD8^+^ T cells. The experimental workflow is illustrated in Fig. 4L. Flow cytometry confirmed that the purity of isolated CD8^+^ T cells exceeded 90% (Fig. 4M). CD8^+^ T cells in the CAV1-Ctrl group showed significantly reduced proliferation compared with PBS controls (Fig. 4N). ELISA results further revealed markedly reduced secretion of effector molecules IFN-γ and GZMB in this group (Fig. 4O, P). Notably, CD8^+^ T cells co-cultured with macrophages induced by CAV1-KD exosomes exhibited restored proliferation and cytotoxicity (Fig. 4N–P). These findings indicate that CAV1 promotes exosome secretion in ESCC cells, driving macrophage polarization toward the M2 phenotype and consequently suppressing CD8^+^ T cell proliferation and cytotoxic function.
CAV1 maintains exosome secretion by regulating autophagy via the PI3K/Akt/mTOR pathway
Reduced exosome secretion can result from impaired formation of MVBs or ILVs, or from enhanced degradation of MVBs through fusion with autophagosomes or lysosomes. To investigate how CAV1 regulates exosome secretion, we first assessed the early endosome marker EEA1 in CAV1-KD and control cells. WB analysis revealed no significant difference (Fig. 5A). In CAV1-KD cells, re-expression of CAV1 markedly reversed changes in autophagy-related proteins (Supplementary Figure S5A). TEM demonstrated that MVB and ILV morphology and abundance were unaffected by CAV1 knockdown (Fig. 5B–D), indicating that CAV1 does not regulate exosome secretion via MVB biogenesis. Exosome secretion is closely linked to autophagy. Under stress conditions, autophagy is activated and intracellular components are sequestered into double-membrane autophagosomes. Autophagosomes fuse with MVBs to form amphisomes, which are subsequently degraded by lysosomes [21, 22]. To test whether autophagy contributes to CAV1-mediated regulation of exosome secretion, we assessed autophagy markers. CAV1-KD cells showed a higher LC3B-II/I ratio and reduced p62 expression compared with controls (Fig. 5A), indicating autophagy activation. TEM analysis further showed that CAV1 knockdown increased the number of autophagic vacuoles (AVs) and enhanced their fusion with MVBs (Fig. 5E–G).
Fig. 5CAV1-KD induces autophagy-mediated MVB degradation and reduces M2 polarization (A) WB of EEA1, LC3B-II/I, and p62 in control and CAV1-KD cells (n = 3 per group). β-actin used for normalization. (B–D) Representative TEM images showing MVB and ILV morphology and abundance in control and CAV1-KD cells (n = 3 per group). Green arrows indicate MVBs; Scale bar = 200 nm. (E–G) Representative TEM images showing AV abundance and MVB–AV fusion events in control and CAV1-KD cells (n = 3 per group). Enlarged image of CAV1-Ctrl shows normal MVB structure, while those of CAV1-KD1 and CAV1-KD2 cells show MVBs fused with AVs. Red arrows indicate AVs; Scale bar = 1 μm. (H) Representative IF images of MVB and AV localization in CAV1-Ctrl, Rapa-treated CAV1-Ctrl, CAV1-KD, and 3-MA-treated CAV1-KD cells (n = 3 per group); Scale bar = 10 μm. Colocalization analysis was performed using ImageJ. (I) NTA analysis of exosome particle concentration (n = 3 per group). (J) BCA assay for exosomal protein content (n = 3 per group). (K–M) qRT-PCR of M2 markers in macrophages treated with exosomes (n = 3 per group). (N) WB of CAV1, EEA1, LC3B-II/I, and p62 (n = 3 per group). β-actin used for normalization. (O) WB of PI3K, AKT, mTOR, and phosphorylated forms in control and CAV1-KD cells (n = 3 per group). β-actin used for normalization. Data are presented as mean ± SD or median and analyzed by one-way ANOVA followed by Tukey’s post hoc test. ns, not significant; *P < 0.05; **P < 0.01; ***P < 0.001
To confirm that CAV1 regulates exosome secretion through autophagy and thereby influences macrophage M2 polarization, CAV1-Ctrl cells were treated with Rapa, an mTOR phosphorylation inhibitor that promotes autophagosome formation, while CAV1-KD cells were treated with 3-MA, an autophagy inhibitor. IF showed that in CAV1-Ctrl cells, Rapa increased AVs and enhanced colocalization of CD63^+^ MVBs with LC3B^+^ AVs (Fig. 5H). Exosomes isolated from equal cell numbers and equal supernatant volumes were quantified using NTA and BCA. Exosome particle size remained unchanged across groups. However, Rapa pretreatment reduced exosome secretion compared with untreated CAV1-Ctrl cells (Fig. 5I, J), and exosomes from Rapa-treated CAV1-Ctrl cells induced weaker M2 polarization in THP-1–derived macrophages (Fig. 5K–M). No significant differences in EEA1 protein levels were observed among groups. In CAV1-Ctrl cells, Rapa increased the LC3B-II/I ratio and decreased p62 expression (Fig. 5N). Conversely, in CAV1-KD cells, 3-MA decreased autophagosome formation and MVB–AV colocalization, markedly inhibiting autophagic flux, as indicated by reduced LC3B-II/I ratios and p62 accumulation. Exosome secretion was significantly increased, and their ability to induce M2 polarization in macrophages was partially restored (Fig. 5H–N). Collectively, these findings demonstrate that CAV1 regulates exosome secretion by limiting autophagic degradation of MVBs, thereby promoting macrophage M2 polarization.
CAV1 has been associated with the Akt/mTOR signaling pathway in multiple diseases [23]. Given the key roles of PI3K, Akt, AMPK, and mTOR in autophagy, we examined their activity to further elucidate CAV1 regulation. In CAV1-KD cells, phosphorylation of PI3K, Akt, and mTOR was reduced (Fig. 5O), whereas p-AMPK, unexpectedly, was decreased rather than increased (Supplementary Figure S5B). Thus, CAV1 appears to suppress autophagy by sustaining PI3K/Akt/mTOR signaling, thereby limiting MVB degradation and maintaining exosome secretion.
CAV1 knockdown suppresses tumor growth and macrophage M2 polarization in vivo
A subcutaneous tumor model was established in BALB/c nude mice to evaluate the role of CAV1 in tumor growth and macrophage infiltration and M2 polarization in vivo. CAV1-Ctrl or CAV1-KD ESCC cells were co-mixed with THP-1–derived macrophages and subcutaneously implanted into mice (Fig. 6A). After four weeks, tumor volume was smaller in the CAV1-KD group than in the control group (Fig. 6B, C). IF analysis further revealed reduced infiltration of CD68^+^CD163^+^ macrophages in the CAV1-KD group compared with controls (Fig. 6D).
Fig. 6CAV1 knockdown suppresses xenograft growth and M2 macrophage infiltration in vivo. (A) Schematic of the subcutaneous xenograft model established in nude mice. (B) Representative images of subcutaneous xenografts generated from co-cultures of control or CAV1-KD cells with THP-1-derived macrophages (n = 7 per group). (C) Tumor volumes of subcutaneous xenografts (n = 7 per group). (D) Representative IF images showing M0 (CD68) and M2 (CD163) marker expression in tumor tissues; Scale bar: 20 μm (n = 3 per group). Colocalization analysis was performed using ImageJ. (E) Representative images and quantification of EdU incorporation in CAV1-Ctrl and CAV1-KD cells; Scale bar: 200 μm (n = 6 per group). (F) Cell proliferation assessed by CCK-8 in CAV1-Ctrl and CAV1-KD cells (n = 5 per group). (G) Representative Ki67 immunohistochemical staining images and quantification of Ki67-positive cells in subcutaneous xenograft tumors; Scale bar: 200 μm (n = 6 per group). Data are presented as mean ± SD and analyzed by one-way ANOVA followed by Tukey’s post hoc test. ns, not significant; ***P < 0.001
Given the significant differences in in vivo tumor growth, we next examined whether tumor cell proliferation was affected. In vitro EdU incorporation and CCK-8 assays showed no significant differences between CAV1-Ctrl and CAV1-KD cells (Fig. 6E, F). Consistently, Ki67 immunohistochemistry of subcutaneous xenograft tumors revealed comparable proliferative activity among groups (Fig. 6G).
Collectively, these data indicate that the tumor growth inhibition observed upon CAV1 knockdown is not primarily attributable to altered tumor cell proliferation, but is more likely associated with changes in M2 macrophage polarization.
High CAV1 expression correlates with stromal M2 macrophage infiltration and poor prognosis in ESCC
CAV1 expression was significantly elevated in ESCC tumor tissues relative to matched adjacent normal tissues (Fig. 7A, B). To assess the associations of CAV1 expression with the tumor immune microenvironment and patient prognosis, ESCC tumor samples and matched adjacent normal tissues were analyzed using multicolor IF on TMAs. CAV1 expression density was higher in the tumor parenchyma (1003 ± 1130 number/mm^2^) than in adjacent normal tissues (620 ± 326 number/mm^2^) (P = 0.027; Fig. 7C, D), and CAV1 expression in the tumor parenchyma was positively correlated with stromal CD163 expression (r = 0.505, P < 0.001; Fig. 7E, F). Kaplan-Meier survival analysis showed that patients with high CAV1 expression exhibited reduced overall survival compared with those with low expression (log-rank P = 0.016; Fig. 7G). Univariate Cox regression analysis further confirmed that high CAV1 expression was associated with poor prognosis (HR = 1.77, 95% CI: 1.10–2.85, P = 0.020; Supplementary Table S4). Conversely, CAV1 expression was not significantly associated with age, sex, tumor size, histological grade, or TNM stage (**Supplementary Table **S5). In summary, elevated CAV1 expression is associated with increased stromal M2 macrophage infiltration and may serve as an adverse prognostic marker in ESCC.
Fig. 7. High CAV1 correlates with increased stromal M2 macrophage infiltration and poor prognosis. (A) WB of CAV1 in patient tumor (T) and paired adjacent non-tumor (N) tissues (n = 4). (B) qRT-PCR of CAV1 in tumor and paired adjacent non-tumor tissues (n = 6). (C) Multicolor IF of CAV1 density in tumor and paired adjacent non-tumor tissues (n = 43 per group). (D) Representative IF images of CAV1 in tumor and paired adjacent non-tumor tissues; Scale bar: 50 μm. (E) Representative multicolor IF analysis of CD163 levels in CAV1-high and CAV1-low tumors; Scale bar: 50 μm. (F) Positive correlation between CAV1 expression in tumor parenchyma and stromal CD163 in our cohort (n = 71). (G) Kaplan–Meier survival analysis of patients with high versus low CAV1 expression (n = 79). (H) Schematic illustration of the mechanism by which CAV1 promotes immune evasion in ESCC. Data are shown as mean ± SD. Paired t-tests were used for paired samples, and Spearman correlation coefficients were used to assess the correlation. ns = not significant; *P < 0.05; **P < 0.01
Discussion
Although immunotherapy has advanced the treatment of several malignancies, including ESCC, overall response rates remain modest, largely due to immune evasion mechanisms orchestrated by the TME. Among these, M2-polarized TAMs are of particular interest for their central role in tumor progression and metastasis. M2 macrophages promote immune evasion by modulating PD-1/PD-L1 expression through immunosuppressive cytokines such as TGF-β and IL-10 [24, 25], and by releasing anti-inflammatory mediators including CC Motif Chemokine Ligand 18 (CCL18) and Arg1, ultimately driving disease progression [10]. Clinical evidence consistently shows that high M2 macrophage infiltration in tumor tissues correlates with poor patient outcomes [26]. Therefore, elucidating the mechanisms underlying tumor cell-induced M2 polarization and clarifying tumor–macrophage interactions are crucial for guiding new therapeutic strategies that target the immunosuppressive microenvironment.
We previously reported that ESCC-derived exosomes transfer HMGB1 to induce TAM M2 polarization, thereby fostering an immunosuppressive microenvironment [13]. While exosomes are recognized as key mediators of tumor immune evasion, the mechanisms governing their release remain poorly defined. This study reveals for the first time that CAV1 is a key regulator of exosome secretion in ESCC. Using proteomic profiling, bioinformatics, and functional assays, we showed that ESCC-derived exosomes promote macrophage M2 polarization in a dose-dependent manner. CAV1 knockdown markedly reduced exosome secretion without altering MVB biogenesis, vesicle morphology, or particle size. It also impaired the capacity of exosomes to induce macrophage M2 polarization, thereby enhancing CD8^+^ T cell proliferation and cytotoxicity. Re-expression of CAV1 restored both exosome secretion and the impaired ability of macrophages to undergo M2 polarization. Notably, CAV1 knockdown had little effect on exosome cargo composition. When exosome numbers were normalized, exosomes from both groups induced M2 polarization similarly. These results indicate that CAV1 regulates macrophage polarization primarily through changes in exosome secretion rather than major alterations in cargo. Mechanistically, CAV1 sustains PI3K/AKT/mTOR signaling to suppress autophagy, limiting MVB–AV fusion and degradation. This process maintains exosome secretion from ESCC cells and promotes the establishment of an immunosuppressive microenvironment (Fig. 7H). We have created a graphical abstract summarizing the role of CAV1 in exosome secretion and exosome-mediated macrophage polarization. It integrates cellular-level molecular mechanisms, animal model data, and clinically relevant evidence.
Peinado et al. reported that RAB27a knockdown decreased exosome secretion without affecting their cargo composition, thereby impairing the ability of tumor cells to recruit and “educate” bone marrow–derived cells and disrupting pre-metastatic niche formation [27]. In 4T1 cells, exosome secretion levels closely correlate with tumor growth and metastasis [28]. Beyond pharmacological inhibition, Marleau et al. proposed physically removing circulating tumor exosomes via hemofiltration to suppress tumor progression [29]. These findings underscore the importance of elucidating the molecular mechanisms governing exosome release in ESCC. In this study, we found that KYSE150 cells induce stronger macrophage M2 polarization than TE-1 cells, primarily due to higher exosome secretion rather than differences in exosomal contents. Although both are ESCC-derived, KYSE150 cells exhibit a highly malignant phenotype with low differentiation. In contrast, TE-1 cells were derived from well-differentiated ESCC tissue from a 58-year-old Japanese male patient and exhibit low proliferative activity [30, 31]. Cellular metabolic activity, mechanical stress, and oxidative stress have been reported to influence exosome secretion. Highly invasive tumor cells typically display stronger metabolic activity and stress responses, which can relieve intracellular pressure to maintain survival and also supply the energy needed for exosome biosynthesis and release [32]. These differences may underlie the markedly higher exosome secretion capacity of KYSE150 compared with TE-1. This distinction makes them valuable comparative models for investigating the regulation of exosome secretion in ESCC.
As the main structural component of caveolae, CAV1 regulates membrane invagination and vesicular trafficking [17]. Saquel et al. indicates that CAV1 enhances extracellular vesicle release under mechanical stress, facilitating intercellular communication and tumor progression [33]. However, whether CAV1 contributes specifically to exosome biogenesis or secretion remains unclear. Previous studies have implicated CAV1 regulates autophagy, mainly through its effects on energy metabolism and tumor signaling pathways [34]. In colorectal cancer, loss of CAV1 induces autophagy through activation of the AMPK–TP53/p53 axis [35], whereas Guan et al. reported that CAV1 overexpression alleviates paclitaxel resistance by inhibiting PI3K–AKT–JNK-dependent autophagy in osteosarcoma [36]. Given that CAV1 may interact with signaling pathways such as PI3K/AKT and mTOR to regulate autophagy, we examined changes in PI3K, AKT, AMPK, and mTOR activity after CAV1 knockdown.
CAV1 exhibits dual roles, acting as either a tumor suppressor or an oncogene in different tumor types [37]. Previous studies have suggested that CAV1 also plays a significant role in immune regulation. Zhang et al. used bioinformatics analyses and showed that ESCC patients with low CAV1 expression exhibited a higher tumor mutation burden and greater responsiveness to immune checkpoint inhibitors [38]. CAV1 also regulates F-actin and PAK1 to promote immune synapse formation, a process essential for T-cell activation and effective anti-tumor responses [39]. In a clinical study of triple-negative breast cancer, Godina et al. reported that elevated CAV1 expression was strongly associated with epithelial–mesenchymal transition (EMT) and an immunosuppressive TME [40]. Our animal experiments further confirmed the immunomodulatory role of CAV1. Knockdown of CAV1 significantly suppressed tumor growth in nude mice, and this inhibitory effect was not attributable to changes in tumor cell proliferation. In our cohort of 80 clinical samples, CAV1 exhibited a tumor-promoting role in ESCC. Its expression was higher in tumor tissues compared with adjacent normal tissues. More importantly, elevated CAV1 expression was strongly associated with increased M2 macrophage infiltration and poor patient prognosis, reinforcing CAV1 as an independent prognostic biomarker in ESCC.
However, it should be noted that Liu et al. (60 samples) reported a positive correlation between CAV1 expression and TNM stage in ESCC [41]. Ando et al. (47 samples) observed a significant association between CAV1 immunohistochemical staining intensity and tumor differentiation [42]. In contrast, our study found no significant correlation between CAV1 expression and either histological grade or TNM stage. This discrepancy may arise from differences in detection methods, evaluation criteria, or sensitivity. Our study employed multicolor IF combined with PANCK staining to distinguish tumor parenchyma from stromal regions and to quantitatively assess CAV1 expression density and intensity within the tumor parenchyma, whereas the aforementioned studies used conventional immunohistochemistry, which has limitations in distinguishing tumor cells from stromal areas. Differences in patient cohorts, such as sample size, genetic background, geographic distribution, and treatment modalities, may also contribute to inconsistent findings.
These findings suggest a potential strategy to target tumor–immune cell communication and remodel the immunosuppressive TME. We validated the effect of ESCC cell–derived exosomes on macrophage polarization in a THP-1–derived model and observed similar results in human primary macrophages, strengthening the biological relevance and clinical translatability of this regulatory axis. Although CAV1 knockdown significantly alleviated the immunosuppressive TME and inhibited tumor growth in this study, directly targeting CAV1 is challenging due to its essential roles in normal tissues and the risk of off-target effects. A more feasible approach may involve targeting downstream processes mediated by CAV1, such as the biogenesis and release of tumor-derived exosomes, enabling precise modulation of the TME while preserving CAV1 physiological functions. Future studies should further validate the relationship between CAV1 expression and clinicopathological features in immune-competent models and larger multicenter cohorts, and assess the safety and efficacy of targeting this axis as a therapeutic strategy.
Conclusion
Our study provides the first evidence that CAV1 regulates exosome secretion in ESCC by suppressing autophagy, thereby promoting macrophage M2 polarization and fostering the establishment of an immunosuppressive TME. These findings enhance our understanding of tumor-macrophage crosstalk and support the potential of CAV1 as a prognostic biomarker in ESCC. Future studies are warranted to confirm the role of CAV1 in exosome secretion and its impact on the TME across diverse cell lines and animal models.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary Material 1: This file contains all supplementary tables and figures cited in the main manuscript.
Supplementary Material 2: This file contains the uncropped original Western blot images.
Supplementary Material 3: The full GSEA results for all differentially enriched proteins in KYSE150 and TE-1 cells.
Supplementary Material 4: This file contains the complete proteomic profiles of exosomes from CAV1-Ctrl and CAV1-KD1 cells.
Supplementary Material 5: This file contains the miRNA sequencing profiles of exosomes from CAV1-Ctrl and CAV1-KD1 cells.