Characterising the Effect of Hypoxic Conditions on Extracellular Vesicle Release and Function in Triple-Negative Breast Cancer

Abstract

Hypoxia is a hallmark of many solid tumours and is strongly associated with cancer progression and metastasis. Triple-negative breast cancer (TNBC), an aggressive subtype lacking targeted therapies, exhibits more pronounced levels of tumour hypoxia than other breast cancer forms. Hypoxia-inducible factors (HIFs), particularly HIF-1α and HIF-2α, mediate cellular adaptation to low oxygen and drive gene expression programmes linked to poor clinical outcomes. Extracellular vesicles (EVs) have emerged as key mediators of tumour adaptation, facilitating intercellular communication by transporting bioactive cargo that promotes proliferation, migration, invasion, and epithelial–mesenchymal transition (EMT). Despite growing interest in EVs, studies have been hindered by inconsistencies in hypoxia modelling, including variations in exposure duration and reliance on single cell lines. Most commonly, studies apply acute hypoxia models of 24–48 hours; however, such short durations may not fully capture the biological responses occurring in the tumour microenvironment. This study incorporates extended hypoxic exposures up to 72 hours, allowing for a more comprehensive assessment of dynamic and time-dependent EV responses that better reflect in vivo tumour conditions. To address existing gaps, EV secretion was investigated in four TNBC cell lines - BT-549, Hs578T, MDA-MB-231, and MDA-MB-468 - under well-defined hypoxic conditions (1% O₂) and extended timepoints. Culture conditions were carefully standardised, and serum-free media were used to eliminate contamination from foetal bovine serum-derived EVs. EVs were separated into 10K and 120K subpopulations using differential ultracentrifugation, followed by characterisation through Bradford assay, immunoblotting, imaging flow cytometry, nanoparticle tracking analysis, and transmission electron microscopy. Findings demonstrated distinct, cell line-specific EV responses to hypoxia over time. Functional assays conducted using EVs collected at 48 hours revealed that hypoxic EVs from Hs578T, MDA-MB-231, and MDA-MB-468 cell lines significantly enhanced migration and invasion in recipient cells, suggesting that the phenotypic influence of EVs cargo may be more relevant than the overall quantity of EVs released. In contrast, molecular profiling of EVs collected at 72 hours showed that MDA-MB-468 cells exhibited a pronounced increase in small EV release, along with selective enrichment of EMT-associated proteins. ELISA analysis confirmed elevated levels of E-cadherin and interleukin-8 in the 72-hour EVs, supporting a role for hypoxia-induced EV cargo in modulating tumour cell behaviour. Conversely, MDA-MB-231 cells showed no significant change in EV release under hypoxia, despite previous reports indicating otherwise. The inclusion of extended hypoxic timepoints was crucial in uncovering these temporal dynamics, allowing for a more comprehensive understanding of EVs behaviour and content that would not have been captured using only short-term exposure models. To investigate EVs regulation mechanisms, three HIF inhibitors were assessed: acriflavine (a dual HIF-1α/HIF-2α inhibitor), PX-478 (a selective HIF-1α inhibitor), and MK-6482 (belzutifan), a selective HIF-2α inhibitor currently undergoing clinical trials. Drug tolerability was determined using acid phosphatase assays, with MK-6482 selected for further experiments due to its favourable viability profile. Immunoblotting confirmed effective suppression of HIF-2α in MK-6482-treated Hs578T cells. Subsequent EVs separation revealed a significant reduction in total and small EV secretion without notable cytotoxicity, suggesting a regulatory role of HIF-2α in EV release. Interestingly, MK-6482-treated cells also showed enhanced proliferation, indicating a possible shift towards HIF-1α-driven activity, which merits further investigation. To explore the translational relevance of EVs in breast cancer, six EVs separation methods were evaluated using serum from breast cancer patients and age-matched controls. While differential ultracentrifugation remains the standard method for EVs separation in research, its technical complexity, time-consuming protocols, and need for large sample volumes hinder its applicability in clinical settings. To address this, a CD63/CD81/CD9 immunobead-based enrichment method was identified as the most suitable for clinical use, offering high reproducibility, sensitive detection of EVs markers, and minimal albumin and lipoprotein contamination. Application of this method to an independent serum cohort revealed significantly higher levels of gremlin-1 in EVs from breast cancer patients compared to controls, highlighting their diagnostic potential and supporting the feasibility of blood-based EVs profiling for clinical applications. In conclusion, this study provides a robust, multi-cell line, time-resolved analysis of hypoxia-induced EV secretion in TNBC. It demonstrates the value of extended hypoxia modelling, identifies HIF-2α as a modulator of EVs dynamics, and establishes a clinically applicable EVs enrichment method. These findings contribute to the understanding of hypoxia–EV interplay and offer valuable insights into EVs biology with translational relevance for biomarker discovery and tumour profiling