Exosomal ANXA2 facilitates ovarian cancer peritoneal metastasis by activating peritoneal mesothelial cells through binding with TLR2

Abstract

Background

Peritoneal dissemination of ovarian cancer (OvCa) can be largely attributed to the formation of a metastatic microenvironment driven by tumoral exosomes. Here, we aimed to elucidate the mechanisms through which exosomal annexin A2 (ANXA2) derived from OvCa cells induces an HPMC phenotypic shift in favour of peritoneal metastasis.

Methods

Immunohistochemistry and orthotopic and intraperitoneal OvCa xenograft mouse models were used to clarify the relationship between tumour ANXA2 expression and peritoneal metastasis. Exosomes were isolated from OvCa cell lines via ultracentrifugation. Functional experiments on cell proliferation and motility, and western blot were performed to investigate the activation of HPMCs and its impact on tumour cell in vitro. High-throughput transcriptional sequencing and rescue experiments in which ANXA2 inhibitor (LCKLSL) or the toll-like receptor 2 (TLR2) inhibitor (C29) was used to co-culture the HPMCs with exosome were employed to identify the crucial functional molecules through which exosomal ANXA2 activates HPMCs. The impact of exosomal ANXA2-activated HPMCs on tumour progression was assessed via functional experiments.

Results

Primary OvCa samples with high ANXA2 expression exhibited a stronger tendency to metastasize to the abdominal cavity. Tumoral ANXA2 promoted OvCa peritoneal metastasis through the secretion of exosomes carrying ANXA2. ANXA2-loaded exosomes activated HPMCs through exosomal ANXA2 binding to TLR2, shifting the phenotype of HPMCs towards mesenchymal cells, increasing their migration and invasion capacities, and elevating the expression of lipocalin 2 (LCN2). High LCN2 expression in HPMCs promoted OvCa cell adhesion, proliferation, motility, and lipid metabolism reprogramming.

Conclusion

Exosomal ANXA2 secreted by tumour cells activates HPMCs and induces the expression of LCN2, which in turn promotes the peritoneal metastasis of OvCa.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12964-024-01987-y.

Background

Owing to rapid and asymptomatic progression, ovarian cancer (OvCa) patients are usually diagnosed at an advanced stage with peritoneal metastasis [1], which poses a considerable challenge for treatment. Therefore, there is an urgent need to explore the biological nature of peritoneal metastasis in OvCa and to identify novel therapeutic targets.

The peritoneal metastasis microenvironment represents a distinct tumour microenvironment characterized by the presence of tumour cells, human peritoneal mesothelial cells (HPMCs), and a limited population of stromal cells, including adipocytes, fibroblasts, endothelial cells, and macrophages, within the peritoneal cavity. It has received widespread attention, particularly in OvCa and gastrointestinal malignancies [2–4]. HPMC refers to a monolayer of cells that line the peritoneal cavity and was previously considered a mechanical barrier against bacterial and tumour adhesion [5]. However, cancer cells are able to induce mesothelial–mesenchymal transition (MMT) [6, 7] and educate HPMCs into cancer-associated mesothelial cells (CAMs). CAMs, in turn, promote tumour adhesion, invasion, vascularization and growth [8, 9], playing important roles in tumour peritoneal metastasis. Numerous studies have shown that tumour-derived exosomes induce HPMCs to undergo MMT through the delivery of microRNAs [10, 11]. Exosomes, vesicles that originate from endosomes and range in size from 30 to 200 nm in diameter and harbour nucleic acids, proteins, lipids and so on, play crucial roles in intercellular communication through mechanisms such as endocytosis, membrane fusion, or receptor-mediated interactions with recipient cells [12]. However, the specific role and underlying mechanisms by which OvCa-derived exosomes activate HPMCs to promote metastasis remain to be elucidated.

Exosomal Annexin A2 (ANXA2) has been reported to regulate the phenotype of the mesothelial cell line HMrSV5 in terms of cell migration, invasion, apoptosis and fibrosis [13], which triggered our interest in investigating its role in MMT. ANXA2, a member of the calcium-dependent phospholipid-binding protein family, is a scaffolding protein with versatile binding capacities on the surface of endothelial cells, immune cells and other cells, orchestrating haemostasis, inflammatory responses, and tissue fibrogenesis [14]. Notably, ANXA2 has been found in several types of cancer cells and tumour-derived exosomes, such as those from breast cancer, lung cancer, and glioma [15–17], where exosomal ANXA2 regulates tumour progression and the immunotherapy response through mediating communication between cancer cells and non-malignant cells in the tumour microenvironment, such as macrophages and astrocytes. In OvCa, ANXA2 promotes epithelial mesenchymal transition and cell adhesion and can serve as a target for immunotherapy [18–20], but its role and mechanism in interactions with HPMCs and MMT as well as the prometastatic peritoneal microenvironment remain unexplored.

In this work, we demonstrate the promoting effects of ANXA2-rich exosomes from OvCa on MMT and tumour intraperitoneal metastasis and identify Toll-like receptor 2 (TLR2) on the surface of HPMCs as a mediator linking exosomal ANXA2 to the phenotypic shift of HPMCs, providing a new basis for developing strategies that interfere with the peritoneal microenvironment to alleviate OvCa metastasis.

Materials and methods

Immunohistochemistry (IHC)

IHC staining for ANXA2 was conducted in fifty-six archival paraffin-embedded human OvCa samples (Supplementary Table 1) and xenografts derived from ES-2 cells. All human tumours were collected between October 2015 and July 2019 and pathologically diagnosed as primary OvCa, while those that received preoperative antitumour treatment were excluded. The use of the archival human samples was approved by the Institute Ethical Committee (IORG No: IORG0003571). The tumours isolated from the mice were formalin fixed and paraffin embedded immediately after removal. In addition to ANXA2, Ki-67, a marker of cell proliferation activity, was also detected by IHC in xenografts. For the IHC assays, the paraffin sections were deparaffinized and heated to expose the antigens using sodium citrate antigen repair solution. The staining was conducted with the aid of an IHC kit (PV-9000, ZSGB-Bio) according to the manufacturer’s instructions. Briefly, endogenous peroxidase activity was quenched via the use of an endogenous peroxidase blocker. The samples were incubated with anti-ANXA2 or anti-Ki67 antibodies overnight at 4 °C, followed by incubation with biotinylated goat anti-rabbit/mouse IgG for 30 min at 37 °C. The complexes were visualized using the chromogen diaminobenzidine (DAB). Finally, the tissue sections were counterstained with haematoxylin. In addition, mouse paraffin sections were stained with haematoxylin and eosin (H&E) for histological evaluation. Five bright field images (magnification ×400) were randomly taken per sample for quantification. The ANXA2 IHC results were calculated by multiplying the intensity score (none = 0, faint = 1, moderate = 2, strong = 3) by the percentage of the staining area (< 5% = 0, 5–25% = 1, 26–50% = 2, 51–75% = 3, > 75% = 4) for human tissues and were calculated via the IHC Toolbox in ImageJ for mouse tissues. Ki67-positive nuclei (DAB-brown) and total cell nuclei were counted manually to quantify the percentage of Ki67-positive cells. The antibodies used are listed in Supplementary Table 2.

OvCa cell lines and culture

OVCAR3 (human ovarian high-grade serous carcinoma cells) and ES-2 (human ovarian clear cell carcinoma cells) cells were obtained from the China Center for Type Culture Collection (Wuhan University, Wuhan, China). ES-2-HM (ES-2 highly metastatic) is an omental highly metastatic ES-2 subline obtained from a spontaneous metastatic model [21]. Luciferase-labelled ES-2 cells (ES-2-Luc) were established and maintained as previously described [22] and used for monitoring of xenografts during animal experiments. All of the cell lines were authenticated by short tandem repeat profiling in the past three years, which confirmed the absence of cross-contamination. All of the cells were cultured in DMEM/F12 supplemented with 10% foetal bovine serum (FBS, Gibco) at 37 °C with 5% CO₂.

Cell transfection

For long-term silencing of ANXA2 in OvCa, pGLVU6-shRNA-CMV-luciferase-Puro lentiviral vectors containing short hairpin RNA targeting ANXA2 (shANXA2) and nontargeting shRNA (shNC) synthesized by GenePharma (Shanghai, China) were used to transfect OVCAR3 and ES-2 cells. The sequences of the shRNAs used are listed in Supplementary Table 3. Lentiviral particles encoding LCN2 (Ubi-MCS-3FLAG-SV40-EGFP-IRES-puro-LCN2) were synthesized by GENECHEM (Shanghai, China) to augment the expression of LCN2 in HPMCs, and an empty vector not expressing LCN2 served as a negative control. Transfection was conducted according to the manufacturer’s instructions, and a moderate volume of lentivirus (according to the multiplicity of infection) and polybrene (200 µg/ml) was added to the culture medium. ES-2 and OVCAR3 cells were further selected by puromycin for two weeks, then the stable transfected ES-2 and OVCAR3 cell were verified by Western blot and qRT-PCR for ANXA2 level. And the transfected HPMCs did not undergo clonal selection because of the vulnerability of the primary cells. The efficacies of transfections were verified by Western blot and qRT-PCR. The ANXA2-knockdown group served as the KD group, the LCN2-overexpressing group served as the OE group, and the negative control groups were designated the NC groups.

Animal experiments

To investigate the role of ANXA2 expressed by tumour cells in OvCa metastasis, we established orthotopic ovarian transplantation tumour models using ES-2-Luc [23]. The cells transfected with shANXA2 (ES-2-ANXA2-KD) or shNC (ES-2-ANXA2-NC) were subcutaneously injected into nude mice, and the tumour blocks were harvested for subsequent orthotopic OvCa modelling. Eight female BALB/c nude mice aged 4 weeks (Beijing Vital River Laboratory Animal Co.) were randomly divided into two groups to receive tissue implantation into the left ovarian bursa (ANXA2-NC and ANXA2-KD). After implantation, the mice were monitored by a LUMIN II in vivo imaging system (Caliper, USA) twice a week. After 3 weeks, the mice were sacrificed, and their abdominal organs were isolated to detect bioluminescence ex vivo.

To investigate the role of exosome-loaded ANXA2 in tumour metastasis, 12 female BALB/c nude mice aged 4 weeks (Beijing Vital River Laboratory Animal Co.) were randomly divided into three groups and subjected to intraperitoneal injections of phosphate-buffered saline (PBS), exosomes (50 µg/200 µl) from ES-2-NC cells (exo^NC) or exosomes from ES-2-KD cells (exo^KD). They were injected every three days, four times in total, and then 2 × 10⁶ ES-2-Luc cells were intraperitoneally injected. Bioluminescence was quantified by a LUMIN II in vivo imaging system (Caliper, USA) at 4, 7, 10, 16, and 22 days after inoculation, and the mice were sacrificed after the last in vivo imaging was completed on day 22. The abdominal organs of these mice were isolated to detect bioluminescence ex vivo. Ovarian tissues and omental tissues were excised and fixed with 4% paraformaldehyde for paraffin embedding, followed by IHC. The animal welfare and experiments conformed to the guidelines for the care and use of laboratory animals and were performed according to the guidelines and approval of the Ethics Committee of Wuhan Youdu Biotechnology Co., Ltd.

Isolation and culture of primary HPMCs

Eighteen omental surgery samples obtained from patients diagnosed with OvCa or borderline ovarian tumours (Supplementary Table 4) were used to isolate primary HPMCs. Informed consent was obtained from each patient, and the protocol was approved by the Ethical Committee (IORG No: IORG0003571). HPMCs were isolated following established protocols [24]. Briefly, omental tissues were washed with sterile PBS, cut into 0.5–1 cm² pieces, incubated with 0.25% trypsin (containing 0.02% EDTA) for 25 min at 37 °C, filtered through 100-mesh sieves, and centrifuged at 300 × g for 10 min. Finally, the cell sediment was suspended in DMEM/F12 medium supplemented with 10% foetal bovine serum (FBS; Gibco) and incubated at 37 °C with 5% CO₂. The HPMCs used in this study were between the second and fifth passages.

Co-immunoprecipitation (Co-IP)

A Co-IP assay was used to investigate the binding relationship between ANXA2 and TLR2 proteins. Initially, HPMCs were cultured with OvCa cell exosomes in the presence or absence of LCKLSL (HY-P2333A, MedChemExpress), an inhibitor of ANXA2 [25, 26], for 72 h. Then 1 × 10⁷ HPMCs were then washed with cold PBS and lysed with 1 ml IP lysis buffer (P0013, Beyotime) on ice for 1 h in the presence or absence of LCKLSL. The cell suspension was subsequently centrifuged at 12,000 × g for 15 min, and 100 µl of the supernatant was collected as input. The remaining supernatant was divided into two equal parts and incubated with equal amounts of IgG or the specific antibody respectively on a shaking table for an overnight period at 4 °C. Protein A/G agarose beads were then employed to capture the antigen-antibody complexes for 4 h at 4 °C. Thereafter, the beads were washed with PBS, and the complexes were boiled and subjected to Western blot analysis.

Flow cytometry assays for lipid accumulation in OvCa cells

OvCa cells were seeded in 6-well plates (1 × 10⁵ per well) and treated with mixed culture medium (50% conditional medium from HPMCs mixed with 50% complete medium) for 24 h. Thereafter, the OvCa cells were harvested via trypsinization and washed with PBS. The cells were stained with 10 µM BODIPY™ 493/503 (HY-W090090, MedChemExpress) in PBS for 30 min and washed with PBS, and the fluorescence intensity was measured via a flow cytometer (BD, LSRForstessaX-20 Special Order Product) with excitation at 488 nm and emission at 500–550 nm. The FITC-A fluorescence intensities were analysed via FlowJo software (10.4.1), and the results are expressed as the mean fluorescence intensities. Each assay was performed with three biological replicates.

Statistical analysis

All of the statistical data were analysed using Statistical Product and Service Solutions (SPSS) software version 22.0 (IBM SPSS, USA) and GraphPad Prism software version 8.0 (GraphPad Software, USA). The overall survival probability was determined via the Kaplan–Meier method, and high and low ANXA2 expression levels were grouped on the basis of the optimal cut-off value of the ANXA2 IHC score. Differences were evaluated via the log-rank test. All of the data were subjected to hypothesis testing prior to analysis. Data meeting the assumptions were analysed via two-tailed independent t tests or one-way ANOVA. Data not meeting the assumptions were analysed via the Mann‒Whitney U test or the Kruskal‒Wallis test. Parametric tests are reported as the means ± SDs, whereas nonparametric tests are reported as medians ± interquartile ranges. P < 0.05 was considered statistically significant.

Results

ANXA2 promotes the peritoneal metastasis of ovarian cancer

To explore whether ANXA2 expression is associated with OvCa metastasis, ANXA2 in human OvCa samples was analysed via IHC (Fig. 1A). Notably, higher ANXA2 IHC scores were significantly correlated with advanced FIGO stages, the presence of peritoneal metastasis (all P < 0.05; Fig. 1B), and poorer overall survival (P < 0.05; Fig. 1C). In vitro, Western blot analysis revealed higher ANXA2 levels in ES-2-HM than in its parent counterpart (Fig. S1A). To analyse the function of ANXA2 in vivo, ANXA2 was knocked down in ES-2 and OVCAR3 cells via shRNAs (Fig. S1B and Fig. S1C), and ES-2-NC and ES-2-KD cells labelled with luciferase were used to construct orthotopic ovarian transplantation tumour models in BALB/c-nude mice, which allowed for in vivo tracing of tumour cell signals. The bioluminescence signal of the ANXA2-NC group was significantly greater than that of the ANXA2-KD group after day 18. All four mice in the ANXA2-NC group presented signals at the site contralateral to the ovary, which was primarily inoculated with tumour tissue, whereas only one in the ANXA2-KD group presented with metastasis on the 18th day (Fig. 1D and E). Bioluminescence imaging of the abdominal organs isolated from the mice after sacrifice on the 21st day confirmed the in vivo observations. The total intensity of signals from all of the ex vivo organs was significantly greater in the ANXA2-NC group than in the ANXA2-KD group, which was attributed mainly to the increased signal at metastatic sites, such as the omentum and right ovary, rather than at the primary tumour site on the left ovary (Fig. 1F and G). Finally, IHC confirmed the low expression of ANXA2 in ANXA2-KD tumours (Fig. S1D). These results demonstrate that OvCa cells with high ANXA2 expression exhibit a stronger tendency to metastasize.

Fig. 1.

ANXA2 promotes the peritoneal metastasis of ovarian cancer (OvCa). A Representative image of ANXA2 immunohistochemistry (IHC) staining in primary tumour tissues. Scale bar = 50 μm. B Left: Comparison of ANXA2 IHC scores of primary tumour tissues between OvCa patients with metastasis (N = 32) and those without metastasis (N = 24). Right: Comparison of ANXA2 IHC scores of primary tumour tissues between OvCa patients with FIGO stages I-II (N = 16) and those with FIGO stages III-IV (N = 40). Data are presented as the median ± interquartile range, p values were calculated via the Mann–Whitney U test, *p < 0.05. C Kaplan–Meier overall survival analysis of a cohort of OvCa patients (N =56, log rank P = 0.0468) stratified by low or high ANXA2 expression (based on the optimal ANXA2 IHC score cut-off value). D Dynamic detection of bioluminescence signals from ES-2-ANXA2-NC and ES-2-ANXA2-KD cells via in vivo animal imaging (n = 4). E Total bioluminescence signal at different observation points detected by in vivo animal imaging. The total bioluminescence intensity measured on the last day was compared via the Mann–Whitney U test, *p < 0.05. F Bioluminescence signal detection in nude mice by isolating the abdominal organs after sacrifice (n = 4). G Histogram showing the bioluminescence signals ex vivo of all abdominal organs, the primary tumour (left ovary), the omentum and the right ovary after sacrifice. Data are presented as the median ± interquartile range, p values were calculated via the Mann–Whitney U test, *p < 0.05; ns, not significant

Tumoral ANXA2-loaded exosomes promote peritoneal metastasis

Next, we investigated whether ANXA2 can be loaded in exosomes derived from OvCa cells and the role of exosomal ANXA2 in metastasis. Exosomes derived from OvCa cells were isolated and characterized. The nanoparticle tracking analysis revealed that the diameter of the vesicles predominantly ranged from 30 to 200 nm (Fig. 2A, Fig. S2A). Additionally, these vesicles displayed a cup-shaped morphology with a bilayer membrane structure (Fig. 2B, Fig. S2B) and expressed CD9, CD63, and TSG101 but not calnexin (Fig. 2C, Fig. S2C), confirming successful exosome isolation. The level of exosomal ANXA2 protein cargo was significantly greater in the exosomes derived from ES-2-HM cells (exo^ES−2−HM) than in those derived from ES-2 cells (exo^ES−2) (Fig. S2D), suggesting a potential role for exosomal ANXA2 in promoting metastasis. For verification, the ability of exosomes derived from ES-2 cells with ANXA2-KD (exo^KD) or ANXA2-NC (exo^NC) to promote peritoneal metastasis was assessed. The exosomal ANXA2 protein vanished after ANXA2-KD (Fig. S2E). BALB/c nude mice were intraperitoneally injected with PBS, ES-2-derived exo^NC, or exo^KD, followed by ES-2-Luc cell injection to assess the effects on peritoneal metastasis (Fig. 2D). Luminescence intensity analysis revealed higher levels in the exo^NC group than in the PBS and exo^KD groups from the 16th day (Fig. 2E, F). Ex vivo imaging of abdominal organs isolated from nude mice revealed that the exo^NC group presented increased luminescence intensity in total intra-abdominal organs and omentum (Fig. 2G, H). Moreover, the exo^NC group presented a greater proportion of Ki67-positive tumour cells in omental metastases than the PBS and exo^KD groups did (Fig. 2I). These results strongly suggest that tumoral ANXA2 promotes OvCa peritoneal metastasis through the secretion of exosomes carrying ANXA2.

Fig. 2.

Tumoral ANXA2-loaded exosomes promote peritoneal metastasis. A Size distribution profile of exosomes isolated from ES-2 cells (exo ES−2 ) evaluated via nanoparticle tracking analysis. B Representative transmission electron micrograph of exo ES−2 . Scale bar = 100 nm. C Western blot analysis of the positive exosomal markers CD63, TSG101, CD9, the negative marker calnexin as well as ANXA2 and GAPGH in total cell lysates (TCLs) and exosomes (exos) derived from ES-2 cells. D Schematic diagram of the experimental procedures used to explore the effect of exosome activation of the premetastatic microenvironment in nude mice injected intraperitoneally with PBS, exo ES−2−NC or exo ES−2−KD. E Dynamic detection of bioluminescence signals from ES-2-Luc cells after intraperitoneal injection into nude mice via in vivo animal imaging (n = 4). F Total bioluminescence signal at different observation points detected by in vivo animal imaging. The total bioluminescence intensity measured on the last day was compared via one-way ANOVA, *p < 0.05. G Bioluminescence signal detection in nude mice by isolating the abdominal organs after sacrifice (n = 4). H Histograms showing the bioluminescence signals of all abdominal organ and omentum tissues after ex vivo sacrifice. The data are presented as the mean ± SD, p values were calculated via one-way ANOVA, *p < 0.05; ns, not significant. I Representative images and statistical analysis of the IHC staining of Ki67 in omental metastatic tumours. Scale bar = 50 μm. The data are presented as the mean ± SD, p values were calculated via one-way ANOVA, *** p < 0.001

Exosomal ANXA2 activates HPMCs

To investigate the effects of ANXA2-loaded exosomes on HPMCs, which may contribute to peritoneal metastasis, we isolated primary HPMCs from the omentum for in vitro experiments. Primary HPMCs initially exhibited a reticular growth pattern, transitioning to a cobblestone appearance upon reaching confluence under light microscopy, and immunofluorescence staining confirmed the presence of cytokeratin 8 and vimentin, while CD31 was absent (Fig. S3A), verifying their identity as HPMCs [27, 28]. PKH-67-labelled OVCAR3-derived exosomes were internalized and accumulated around the nuclei of HPMCs after 24 h of coculture, indicating the uptake of OvCa-derived exosomes by HPMCs (Fig. 3A). The HPMCs were then treated with exosomes derived from cancer cells with ANXA2 knockdown (exo^ES−2−KD and exo^OVCAR3−KD), exosomes derived from negative control cells (exo^ES−2−NC and exo^OVCAR3−NC) or PBS. Compared with the PBS-treated HPMCs, the HPMCs cocultured with exo^ES−2−NC or exo^OVCAR3−NC presented decreased E-cadherin and calretinin expression but increased vimentin expression (Fig. 3B, S3B), accompanied by increased in vitro migration and invasion abilities (Fig. 3C and D), whereas these effects were notably weaker in the HPMCs cocultured with exo^ES−2−KD or exo^OVCAR3−KD.

Fig. 3.

Exosomal ANXA2 derived from OvCa activates HPMCs. A Representative image of the uptake of PKH67-labelled exosomes by HPMCs. Red, cytomembrane. Green, exosomes. Blue, cell nuclei. Scale bar = 100 μm. B Western blot analysis of E-cadherin, vimentin and calretinin levels in HPMCs (left) treated with PBS or 50 µg/ml exosomes derived from ES-2 (exo ES−2−NC or exo ES−2−KD ) for 72 h and HPMCs (right) treated with PBS or 50 µg/ml exosomes derived from OVCAR3 (exo OVCAR3−NC or exo OVCAR3−KD ) for 72 h. Wound healing assays (C) and Transwell assays (D) were performed to detect changes in the migratory and invasive potential of HPMCs treated with PBS or 50 µg/ml exo OVCAR3−NC or exo OVCAR3−KD and that of HPMCs treated with PBS or 50 µg/ml exo ES−2−NC or exo ES−2−KD . Left, representative images; scale bar = 100 μm. Right, histograms present the means ± SDs, n = 3, p values were calculated via one-way ANOVA, **p < 0.01; ***p < 0.001. E After treatment with DMSO or 50 µg/ml exosomes derived from OVCAR3 (exo OVCAR3 ) or ES-2 (exo ES−2 ) cells with or without 40 µM LCKLSL, an inhibitor of ANXA2, for 72 h, the E-cadherin, vimentin and calretinin levels in HPMCs were analysed by Western blot. F Transwell assays were performed to detect changes in the migratory and invasive potential of HPMCs after treatment with DMSO, 50 µg/ml exo OVCAR3 or exo ES−2 with or without 40 µM LCKLSL for 72 h. Data are presented as the mean ± SD, n = 3, p values were calculated via one-way ANOVA, ** p < 0.01; *** p < 0.001; **** p < 0.0001

Considering that ANXA2 KD in cancer cells may lead to changes in exosomal cargos other than a decrease in exosomal ANXA2, we introduced the ANXA2 inhibitor LCKLSL into cocultures of exosomes and HPMCs to block the potential interaction between ANXA2 and HPMCs. We found that the effect of LCKLSL on vimentin expression in HPMCs showed concentration dependence (Fig. S3C) and the 40µM LCKLSL could effectively rescue the exosome-induced alterations in the HPMCs in terms of calretinin and vimentin expression but not E-cadherin (Fig. 3E, S3D), as well as the migration and invasion abilities (Fig. 3F). These results indicate that exosomal ANXA2 derived from OvCa cells induces a phenotypic shift towards mesenchymal cells in HPMCs.

Exosomal ANXA2-educated HPMCs facilitate tumour cell adhesion, proliferation and motility

Given that tumour cell adherence to HPMCs, colonization, and invasion into the tissue beneath HPMCs are critical steps for peritoneal metastasis, we assessed how HPMCs educated with ANXA2-loaded exosomes affect the abilities of tumour cells. HPMCs were pretreated with PBS, exo^NC, or exo^KD. For adhesion assays, pretreated HPMCs were cultured to form monolayers, after which cancer cells dyed with calcein AM were added to the culture system (Fig. 4A). There were significantly more ES-2 and OVCAR3 cells adhered to the exo^NC-treated HPMCs than to the HPMCs treated with PBS or exo^KD (Fig. 4B). For proliferation, migration and invasion assays, serum-free culture medium from pretreated HPMCs (HPMC-CM) was collected to treat ES-2 and OVCAR3 cells (Fig. 4A). Compared with the OvCa cells cultured in complete medium or those cultured in the CM from PBS or exo^KD pre-treated HPMCs, those cultured in the CM from exo^NC treated HPMCs showed stronger proliferation, invasion and migration ability (Fig. 4C and D). These findings indicate that exosome loaded ANXA2-stimulated HPMCs enhance OvCa cell adhesion, proliferation, migration, and invasion.

Fig. 4.

Exosomal ANXA2-educated HPMCs promote OvCa cell adhesion, proliferation and motility. A Schematic representation of the methodologies utilized to investigate the effects of HPMCs on OvCa cells. B Adhesion assays were conducted to assess changes in OvCa cell adhesion to HPMCs treated with PBS or 50 µg/ml exosomes derived from OVCAR3 cells (exo OVCAR3−NC and exo OVCAR3−KD ) or exosomes derived from ES-2 cells (exo ES−2−NC and exo ES−2−KD ) for 72 h. Upper, representative images, scale bar = 100 μm; lower, histograms present the means ± SDs, n = 3, p values were calculated via one-way ANOVA, **p < 0.01; ***p < 0.005. C An EdU assay was performed to detect changes in the proliferative capacity of OvCa cells treated with PBS, CM from PBS or 50 µg/ml exosome-pretreated HPMCs. Left, representative images of OVCAR3 cells; scale bar = 100 μm. Right, histograms present the means ± SDs, n = 3, p values were calculated via one-way ANOVA, *p < 0.05; **p < 0.01. D Transwell assays were performed to detect changes in the migratory and invasive potential of OvCa cells treated with complete medium (control) or CM from PBS- or 50 µg/ml exosome-pretreated HPMCs. Left, representative images of OVCAR3 cells; scale bar = 100 μm. Right, histograms present the means ± SDs, n = 3, p values were calculated via one-way ANOVA, **p < 0.01; ***p < 0.001; ****p < 0.0001

LCN2-high HPMCs promote OvCa progression and reprogram lipid metabolism

To clarify how exosomal ANXA2-educated HPMCs promote tumour progression, we conducted RNA-seq analysis on HPMCs treated with exo^OVCAR3−NC or exo^OVCAR3−KD. We identified 124 differentially expressed genes (DEGs), including 60 upregulated genes and 64 downregulated genes (Fig. 5A), which were enriched in cell adhesion and lipid metabolic processes (Fig. 5B). We focused on the DEGs identified via RNA-seq as mentioned above and validated their differential expression through qRT‒PCR. We found that lipocalin-2 (LCN2) was the gene with the most significant difference between exo^NC-treated HPMCs and exo^KD-treated HPMCs via both RNA-seq and qRT‒PCR (Fig. S4A). Compared with exo^KD or PBS, both exo^NC from ES-2 cells and OVCAR3 cells induced significant upregulation of LCN2 expression in HPMCs (Fig. 5C).

Fig. 5.

LCN2-high HPMCs promote OvCa progression and reprogram lipid metabolism. A Volcano plot showing the differential gene transcriptional profiles among three pairs of HPMCs treated with 50 µg/ml exo OVCAR3−NC or exo OVCAR3−KD . Red points, overexpressed genes in exo OVCAR3−NC -treated HPMCs; green points, downregulated genes. B Gene Ontology enrichment analysis of differentially expressed genes. BP, biological process; CC, cellular; MF, molecular function. C qRT‒PCR analysis of LCN2 expression in HPMCs treated with PBS or 50 µg/ml exosomes with different ANXA2 expression levels derived from OVCAR3 (exo OVCAR3−NC or exo OVCAR3−KD ) and ES-2 (exo ES−2−NC or exo ES−2−KD ) cells for 48 h. Data are presented as the mean ± SD, n = 3, p values were calculated by one-way ANOVA, ****p < 0.0001. D Transwell assays were performed to detect changes in the migratory and invasive potential of OvCa cells treated with complete medium (control) or CM from HPMC LCN2−NC or HPMC LCN2−OE . The data are presented as the mean ± SD, n = 3, p values were calculated via one-way ANOVA, **p < 0.01; ***p < 0.001; ****p < 0.0001. E An EdU assay was performed to detect changes in the proliferative capacity of OvCa cells treated with complete medium (control) or CM from HPMC LCN2−NC or HPMC LCN2−OE . The data are presented as the mean ± SD, n = 3, p values were calculated via one-way ANOVA, ***p < 0.001; ****p < 0.0001. F Flow cytometry was performed to detect lipid accumulation in OVCAR3 and ES-2 cells treated with complete medium (control) or CM derived from HPMC LCN2−NC or HPMC LCN2−OE . Left, representative images; right, histograms present the means ± SDs, n =3, p values were calculated via one-way ANOVA, **p < 0.01; ***p < 0.001; ****p < 0.0001. G, H Flow cytometry was performed to detect lipid accumulation in OVCAR3 and ES-2 cells treated with complete medium (control) or CM derived from HPMCs treated with PBS, exo NC or exo KD . Left, representative images; right, histogram presents the means ± SDs, n = 3, p values were calculated via one-way ANOVA, *p < 0.05; **p < 0.01; ***p < 0.001. I qRT‒PCR was used to analyse the expression of lipid metabolism-related genes in OVCAR3 and ES-2 cells treated with complete medium (control) or CM from HPMC LCN2−NC or HPMC LCN2−OE . The data are presented as the mean ± SD, n = 3, p values were calculated via one-way ANOVA, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; ns, not significant

To further investigate the impact of LCN2 expression in HPMCs on OvCa cells, we established HPMCs with stable upregulation of LCN2 (HPMC^LCN2−OE) and a negative control (HPMC^LCN2−NC) and confirmed high LCN2 expression via WB (Fig. S4B). Compared with complete medium or the CM from HPMC^LCN2−NC, the CM from HPMC^LCN2−OE significantly enhanced tumour invasion, migration and proliferation (Fig. 5D and E). Given the significant role of LCN2 in lipid transport, lipid accumulation in OVCAR3 and ES-2 cells treated with CM from HPMC^LCN2−NC or HPMC^LCN2−OE, as well as CM from HPMCs pretreated with PBS, exo^NC, or exo^KD, was examined via flow cytometry. Compared with the OvCa cells cultured in complete medium or CM from HPMC^LCN2−NC, those cultured in CM from HPMC^LCN2−OE presented greater lipid accumulation (Fig. 5F). Similarly, compared with the tumour cells cultured in complete medium or those cultured in the CM from PBS or exo^KD pre-treated HPMCs, those cultured in the CM from exo^NC treated HPMCs resulted in greater lipid accumulation (Fig. 5G, H). Furthermore, we analysed the mRNA expression levels of lipid metabolism-related genes via qRT‒PCR and revealed the differential expression of several genes involved in lipid synthesis, catabolism, and lipid droplet formation (Fig. 5I). These results suggest that HPMCs with high LCN2 expression may promote tumour progression and reprogram lipid metabolism in tumours.

Exosomal ANXA2 binding to TLR2 induces LCN2 expression and activation of HPMCs

It has been reported that LCN2 can be upregulated by Toll-like receptor (TLR) ligands in immune cells [29]. These findings suggest that the activation of TLRs in HPMCs may promote the expression of LCN2. To further confirm these findings, we assessed the expression of the TLR family in untreated HPMCs and found that, among the ten TLRs tested, TLR2 presented the highest mRNA level (Fig. 6A). The high expression of TLR2 in HPMCs was further confirmed by immunofluorescence (Fig. S5A). Given the subcellular location of ANXA2 on the plasma membrane, we speculated that exosomal ANXA2 is located on the exosome membrane. The results of the dot blot confirmed this speculation, as after Tween 20 was used to permeabilize the exosome membrane, there was no significant increase in the expression of ANXA2 or the exosome membrane protein CD63 (Fig. 6B). Therefore, we focused on the role of TLR2 in the interaction of HPMCs with exosomal ANXA2. The Immunofluorescence co-localization revealed that in exosome treated HPMCs, the fluorescence of ANXA2 and TLR2 was highly overlapping in the plasma membrane (Fig. 6C). We subsequently confirmed the binding interaction between ANXA2 and TLR2 through co-immunoprecipitation (Co-IP) (Fig. 6D). Furthermore, exo^NC, but not exo^KD, upregulated the mRNA expression of IL-1, IL-6, and IL-8, the downstream cytokines of TLR2, in HPMCs (Fig. S5B). Moreover, inhibition of TLR2 signalling with C29 partially rescued the exosome-induced upregulation of vimentin expression and downregulation of calretinin expression in HPMCs (Fig. 6E, S5C), as did the upregulation of LCN2, IL1, IL6, and IL8 expression in HPMCs (Fig. 6F, S5D). Further Co-IP assays demonstrated that blocking ANXA2 with LCKLSL attenuated the ANXA2-TLR2 interaction (Fig. 6G), and LCKLSL partially rescued the exosome-induced upregulation of LCN2 (Fig. 6H), as well as the upregulation of IL1, IL6, and IL8 in HPMCs (Fig. S5E), indicating that exosomal ANXA2 binds to TLR2 to induce HPMC activation and LCN2 expression.

Fig. 6.

Exosomal ANXA2 binds to TLR2, activating HPMCs and upregulating LCN2 expression in HPMCs. A qRT‒PCR assay showing the mRNA expression levels of the TLR family in HPMCs. B Dot blot analysis showing that ANXA2 is localized at the membrane of exosomes. C Immunofluorescence colocalization of ANXA2 and TLR2 in HPMCs. Scale bar = 100 μm. D Coimmunoprecipitation (co-IP) assays were used to identify interactions between ANXA2 and TLR2 in HPMCs after coculture with exosomes for 72 h. E After being treated with PBS or exosomes with or without 50 µM C29, an inhibitor of TLR2, for 72 h, the E-cadherin, vimentin and calretinin levels in HPMCs were analysed by Western blotting. F qRT‒PCR analysis of LCN2 expression in HPMCs treated with DMSO or exosomes with or without 50 µM C29. The data are presented as the mean ± SD, n = 3, p values were calculated via one-way ANOVA, *** p < 0.001; ****p < 0.0001. G Co-IP assays were used to identify the interaction between ANXA2 and TLR2 in HPMCs in the presence or absence of 40 µM LCKLSL after coculture with exosomes for 72 h. H qRT‒PCR analysis of LCN2 expression in HPMCs treated with DMSO or exosomes with or without 40 µM LCKLSL. The data are presented as the mean ± SD, n = 3, p values were calculated via one-way ANOVA, ***p < 0.001; ****p < 0.0001

Discussion

Patients with peritoneal dissemination of cancer face limited curative treatment options. Therefore, there is an urgent need to explore the biological mechanism of cancer metastasis and develop novel targeted therapies. The intricate and dynamic interaction between tumour cells and HPMCs plays a crucial role in peritoneal metastasis [2, 30]. This study revealed that the binding of tumour-derived exosomal ANXA2 to TLR2 activates HPMCs, which in turn facilitates the progression and lipid metabolism reprogramming of OvCa cells. These findings suggest that disrupting the communication between tumour cells and HPMCs via exosomal ANXA2 could represent a promising therapeutic strategy.

Exosomes mediate cell-to-cell communication and play important roles in shaping the premetastatic microenvironment of distant metastases [23]. ANXA2 has been reported to be highly enriched in various tumour exosomes, such as those from breast cancer, lung cancer, endometrial cancer, glioma and OvCa [13, 15, 17, 31–33]. Exosomal ANXA2 not only serves as a cancer biomarker but also exhibits diverse functions depending on the recipient cell type. For instance, Song YX et al. [17] demonstrated that glioma-derived extracellular vesicles containing ANXA2 enhance the uptake of these vesicles, thereby facilitating angiogenesis. In breast cancer, exosomal ANXA2 has been shown to promote angiogenesis in a tissue plasminogen activator (tPA)-dependent manner, and induce macrophage-mediated activation of the p38MAPK, NF-κB, and STAT3 signalling pathways, leading to an increased secretion of IL6 and TNFα [16]. And In lung cancer, tumoral exosomal ANXA2 facilitates the activation of neuroastrocyte cells CP-H122 cells [15], highlighting its role in modulating the tumour immune microenvironment. In OvCa, Gao L et al. [13] reported that exosomal ANXA2 promoted MMT and degradation of the extracellular matrix of HMrSV5 cells through the PI3K/AKT/mTOR pathway. In this study, we demonstrated that exosomal ANXA2 derived from OvCa cells could dock to TLR2 in HPMCs to promote MMT and upregulate LCN2, which in turn facilitated OvCa peritoneal metastasis. These findings suggest that HPMCs play important roles in OvCa peritoneal metastasis; furthermore, exosomal ANXA2 and TLR2 could be potential therapeutic targets for OvCa peritoneal metastasis.

Similar to the role of TLR2 in promoting MMT described in this study, Wu et al. [34] demonstrated that the activation of TLR2 on HPMCs promoted the transformation of HPMCs into cancer-associated fibroblasts-like cells in gastric cancer, by activating the downstream signaling molecule NF-κB. In addition, TLR2 triggers the release of inflammatory and fibrotic mediators, which play important roles in peritonitis and peritoneal fibrosis. Specifically, in tuberculous pleural effusion, TLR2 mediates zonula occludens (ZO)−1 downregulation in HPMCs, increasing the permeability of mesothelial cells [35], and induces vascular endothelial growth factor (VEGF) overproduction in HPMCs, which increases vascular endothelial permeability by phosphorylating adhesion junction proteins and interfering with tight junction proteins in tissues [36, 37]. However, it is unclear whether exosomal ANXA2 activating TLR2 increases the mesothelial cells permeability and peritoneal permeability. Nonetheless, Li W et al. [38] reported ANXA2 in endothelial cell membrane played a role in reducing trans-endothelial permeability and maintaining cerebrovascular integrity, and Cheng C et al. [39] reported recombinant ANXA2 significantly reduced early blood-brain barrier disruption after traumatic brain injury. The role of exosomal ANXA2 in peritoneal permeability deserves further investigation, which will contribute to a more comprehensive understanding of its role in OvCa peritoneal metastasis.

Besides MMT occurrence, the upregulation of LCN2 expression in activated HMPCs is also noteworthy. Previous studies have shown that LCN2 expression could be regulated by TLR ligands [29], and our research confirms this finding again. Additionally, similar to previous reports that LCN2 expression is regulated by the NF-κB and JAK-STAT pathways [40], the RNA-Seq in this study identified differential expression of pathway-related genes, such as IRF5, IRAK2, GBP5, TNFSF15 and OXTR, suggesting that these pathways may be involved in exosomal ANXA2-mediated LCN2 upregulation. LCN2 is a 25 kDa secreted protein served as transporter of small lipophilic molecules. In this study, LCN2-high HPMCs promoted OvCa progression, suggesting that LCN2 may be an important molecule connecting the tumour microenvironment and cancer cells. A few studies have investigated the effect of LCN2 on the tumour immune microenvironment. Rui et al. [41] demonstrated that high LCN2 expression in T cells induced T-cell apoptosis and affected tumour progression through immune cytokines and cholesterol metabolism. Li et al. [42] demonstrated that tumour-associated astrocytes promoted Sonic Hedgehog medulloblastoma progression by secreting LCN2. In addition to its immunoregulatory role, numerous studies have shown that LCN2 is closely related to lipid metabolism. Liu et al. [43] reported that LCN2 mediated lipid accumulation in cardiomyocytes in sepsis-induced myocardial injury. Kim et al. [44] reported that the knockdown of LCN2 improved polychlorinated biphenyl-induced lipid accumulation in vitro. In addition, LCN2 deficiency decreased peroxisome proliferator-activated receptor γ (PPARγ) gene expression in adipocytes [45] and was associated with lipid metabolism disorders [46]. Therefore, we focused on exploring the effect of LCN2 expression in HPMCs on lipid metabolism in cancer cells. Interestingly, LCN2-high HPMCs promoted lipid accumulation and affected the expression of several genes related to lipid synthesis, catabolism, and droplet formation in OvCa cells. Cancer cells have a strong affinity for lipids and cholesterol. Cholesterol accumulation, lipid uptake and endogenous de novo fatty acid synthesis facilitates tumour growth and progression. The accumulated lipids and cholesterol form lipid droplets in cancer cells, which participate in cancer anoikis resistance, spreading and progression [47, 48]. Although the effect of LCN2-high HPMCs on lipid metabolism reprogramming in OvCa was initially revealed, the precise mechanism has yet to be elucidated and warrants future investigation.

Conclusion

This study revealed a relationship between ANXA2 and tumour peritoneal metastasis through immunohistochemistry and in vivo experiments and demonstrated that exosomal ANXA2 promotes tumour peritoneal metastasis. Mechanistically, exosomal ANXA2 activated HPMCs by binding to TLR2, which triggered the upregulation of LCN2 in HPMCs. Furthermore, LCN2-high HPMCs promoted proliferation, migration, and invasion of OvCa cells and reprogrammed their lipid metabolism. In conclusion, the interaction between tumour cells and HPMCs mediated by ANXA2–TLR2–LCN2 promotes peritoneal metastasis in OvCa, suggesting that this pathway may be a potential therapeutic target for OvCa peritoneal metastasis.

Abbreviations

OvCa

Ovarian cancer

HPMC

Human peritoneal mesothelial cell

MMT

Mesothelial-mesenchymal transition

CAM

Cancer-associated mesothelial cells

CM

Conditioned medium

DEGs

D

ifferentially expressed genes

Authors’ contributions

ZW and LG performed the study concept and design; JZ and HL performed cellular and molecular experiments and wrote the paper; LG supervised the project and performed the development of methodology and revision of the paper; TL and XL performed the animal experiments; QW and TL collected the clinical samples and analyzed the clinical data; JZ, QW and JC provided acquisition, analysis and interpretation of data; JZ performed statistical analysis; XY, JC and LG provided technical and material support. All authors read and approved the final manuscript.

Funding

This research was funded by the National Natural Science Foundation of China (No. 82203803) and the Natural Science Foundation of Hubei Province of China (No. 2022CFB147).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This study was performed in accordance with the Declaration of Helsinki. All clinical specimens and information involved in this study provided informed consent before collection and were approved by The Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology. All animal study procedures were approved by The Animal Care and Use Committee of Wuhan Youdu Biotechnology Co., Ltd.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.