Cytotoxic chemotherapy can effectively treat invasive breast cancer, but studies in mice have suggested that it can also have pro-metastatic effects. Keklikoglou et al. now show that, in mouse models of chemoresistant breast cancer, paclitaxel and doxorubicin trigger the production of tumour-derived extracellular vesicles (EVs) with pro-metastatic properties.

Credit: Oxana Grivina/Alamy Stock Vector

The authors first treated mammary tumour-bearing mice with paclitaxel, observing that it only modestly inhibited primary tumour growth. However, compared with vehicle, paclitaxel increased the incidence and size of lung metastases or the pulmonary seeding of mammary cancer cells, depending on the mouse model.

EVs can aid the metastasis of primary tumours by influencing cells associated with the pre-metastatic niche. Here, compared with controls, EVs isolated from 4T1 mammary tumour cells (4T1 cells) or murine mammary tumours following paclitaxel treatment increased the number of metastatic nodules in tumour-free mice challenged with metastasis-forming cancer cells. Thus, paclitaxel-induced EVs appear to promote breast cancer metastasis. As EVs derived from murine mammary tumours following doxorubicin treatment also enhanced lung metastasis in mice, this phenomenon is not paclitaxel-specific.

In assessing how chemotherapy-induced EVs promote breast cancer metastasis the authors first observed that paclitaxel increased the release of EVs from mouse and human cancer cell lines as well as the number of EVs in the blood of mice with 4T1 cell-derived tumours. Furthermore, proteomic analysis revealed that EVs derived from paclitaxel-treated 4T1 cells contained higher levels of the Ca2+-binding membrane-associated protein annexin A6 than EVs from vehicle-treated 4T1 cells. Paclitaxel and doxorubicin promoted the loading of annexin A6 into EVs from tumour-derived (but not non-transformed) cells, and annexin A6-deficient 4T1 cells did not promote lung colonization in mice in response to paclitaxel or doxorubicin. Thus, chemotherapy-induced EVs require annexin A6 to promote lung metastasis.

CC-chemokine ligand 2 (CCL2) can reportedly enhance breast cancer metastasis that is assisted by lymphocyte antigen 6C (LY6C)+CC-chemokine receptor 2 (CCR2)+ monocytes. In line with this theory, treatment of tumour-bearing (but not tumour-free) mice with paclitaxel increased the expression of CCL2 and the number of LY6C+ monocytes in the lungs. EVs isolated from paclitaxel-treated mice with tumours also increased the abundance of LY6C+CCR2+ monocytes in the lungs of tumour-free mice. As paclitaxel-induced EVs did not promote pulmonary metastasis in Ccr2-knockout mice or increase the expression of Ccl2 or Ccr2, or the number of LY6C+ monocytes, in the lungs of annexin A6-null mice, annexin A6 may initiate metastasis by expanding LY6C+CCR2+ monocytes.

Finally, looking at the biology of cells in the lung pre-metastatic niche, the authors observed that paclitaxel promoted the internalization of EVs by lung endothelial cells. By using murine endothelial bEnd.3 cells, the authors found that annexin A6 was transferred from paclitaxel-induced EVs to bEnd.3 cells, where it co-localized with the nuclear factor-κB (NF-κB) subunit p65. NF-κB can activate the transcription of Ccl2 and, here, paclitaxel-induced EVs increased NF-κB activity and Ccl2 expression in bEnd.3 cells in an annexin A6-dependent manner.

paclitaxel-induced EVs ... promote breast cancer metastasis

Taking all of their data into account, the authors propose that chemotherapy-induced EVs transfer annexin A6 to pulmonary endothelial cells, where it promotes the upregulation of CCL2 and the accumulation of LY6C+ monocytes to allow tumours to colonize at metastatic sites. As annexin A6 levels were increased in the plasma of patients with breast cancer undergoing neoadjuvant chemotherapy, chemotherapy-induced EVs are potential biomarkers of pulmonary metastasis.