An analysis reveals that cancer cells remotely prepare distant sites for tumour spread in an organ-specific manner, by deploying organ-seeking extracellular vesicles. See Article p.329
The metastatic dissemination of cancer cells from their site of origin through the bloodstream to distant organs is a major cause of cancer-related deaths. This process is not random1; instead, certain populations of cancer cells preferentially seek out and colonize specific organs2, under the control of a range of molecular programs3. Such homing implicitly involves interactions between cancer cells that escape the primary tumour, sometimes known as seeds, and the microenvironment, or 'soil', of target sites1. But less intuitive is the discovery by Hoshino et al.4, described on page 329 of this issue, that seeds can influence the soil before their arrival, sending out extracellular vesicles called exosomes that precondition specific organs for metastatic invasion.
There is growing support for the provocative notion that a build-up of systemic responses to a primary tumour might precede, and even enable, the eruption of metastatic cancer. These responses might involve complex alterations in the body's vascular, coagulation and inflammatory systems — for example, cancer-related changes in the composition of soluble proteins, in cell populations3 or in the characteristics of exosomes5 in the blood.
Hoshino et al. define exosomes as small extracellular vesicles6 — membrane-bounded compartments that transport proteins, lipids and nucleic acids7 from one cell to another, and which can travel considerable distances in bodily fluids or the bloodstream. This information-transfer process has attracted considerable interest in cancer research, because some extracellular vesicles carry cancer-causing genes called oncogenes, or oncogenic proteins that promote cancer formation and disease progression8.
The involvement of extracellular vesicles, including exosomes, in metastasis has been studied for some time9,10, and contributes to several key events that prepare a distant site for colonization — a process called premetastatic niche formation11. For example, in a mouse model of melanoma, contact between exosomes and the capillary wall triggers vascular permeability, which enables cancer cells to escape from the blood vessel into a new site5. In addition, these exosomes can transfer the oncogenic MET receptor protein to circulating blood cells called myeloid cells, altering the cells' behaviour such that they condition premetastatic sites for subsequent colonization by cancer cells5. In pancreatic cancer, circulating exosomes transfer migration inhibitory factor protein to immune cells called Kupffer cells in the liver, triggering a cascade of events that results in premetastatic niche formation12.
Although these results indicate that exosomes can promote metastasis in general, whether and how exosomes are involved in organ-specific metastasis has not been extensively investigated. To explore this question, Hoshino et al. asked whether cancer-cell types known to preferentially home to the lung, liver, brain or bone might produce exosomes that selectively interact with the same organ. Remarkably, this is precisely what they observed. When exosomes from cancer cells were injected into mice, they became lodged in the organ to which those cancer cells are prone to metastasize. Furthermore, the organ-seeking exosomes interacted with different cell types. For instance, exosomes that targeted the lung became lodged in the epithelial cells that line the organ's interior, whereas liver-targeting exosomes entered Kupffer cells.
Hoshino et al. injected mice with exosomes followed by cancer cells from the same cell line, and demonstrated that the exosomes promoted organ-specific metastatic growth. They then made a tantalizing observation — exosomes taken from breast-cancer cells that metastasize to the lung could redirect another cancer-cell population to disseminate in the lung, when it would normally home to the bone. This discovery strengthens the notion that the metastatic characteristics of cancer cells are not autonomous, but can instead be influenced by external factors.
The authors provide several clues to how exosomes orchestrate organ-specific metastasis. They found that exosomes targeting different sites displayed different cell-adhesion receptor proteins called integrins on their surface. The integrin profile of each exosome subtype facilitated its uptake into organs in which an abundance of ligand for that integrin was produced. For instance, αvβ5 integrin directed exosomes to the liver, whereas α6β4 promoted homing to the lung (Fig. 1). Furthermore, inhibiting the exosomal expression or binding of integrins limited organ-specific metastasis. Finally, the authors found evidence that invasion of target organs by exosomes triggered the production of S100 proteins, which promote inflammation and cell migration, and activation of the protein Src — responses that precondition host cells for metastasis.
These fascinating observations expand our understanding of organ-specific metastasis. However, further investigation is required to establish whether and how this knowledge can be put to practical use. The authors demonstrate that integrin expression might predict metastatic spread, pointing to the possibility that exosomal integrin profiles could be used in cancer diagnostics. Their data also indicate that integrin inhibitors might curtail metastatic spread to specific organs. But in many cases, advanced cancers disseminate to several sites3, limiting the potential of therapeutics that work in an organ-specific manner.
It is also worth considering that the molecular pathways that induce metastasis, both exosome-dependent and -independent, are probably extremely diverse. As such, they might be triggered by many context-specific factors: the activation of differentiation pathways in cancer cells; the emergence of a particular molecular subtype within a tumour; therapeutic interventions and more. For example, the incidence of brain metastasis differs between molecular subtypes of breast cancer, and tends to be higher in those driven by the oncogenic protein ERBB2, even after effective treatment with ERBB2 inhibitors13. Whether, and how, ERBB2, its antagonists and the therapies used to treat ERBB2-driven cancers might influence the emission of organ-seeking exosomes is unknown, and is of great interest. Similarly, inflammation, abnormal clotting and other cancer-associated changes in physiology might interfere with the organ-seeking mechanism of exosomes and cells — and must be taken into account when analysing routes of metastasis. Thus, much remains to be understood about the fascinating part that organ-specific exosomes might play in fertilizing the metastatic soil in different human cancers.Footnote 1
Paget, S. Lancet 1, 571–573 (1889).
Fidler, I. J. Nature Rev. Cancer 3, 453–458 (2003).
Nguyen, D. X., Bos, P. D. & Massagué, J. Nature Rev. Cancer 9, 274–284 (2009).
Hoshino, A. et al. Nature 527, 329–335 (2015).
Peinado, H. et al. Nature Med. 18, 883–891 (2012).
Lötvall, J. et al. J. Extracell. Vesicles 3, 26913 (2014).
Colombo, M., Raposo, G. & Théry, C. Annu. Rev. Cell Dev. Biol. 30, 255–289 (2014).
Rak, J. Front. Pharmacol. 4, 21 (2013).
Hood, J. L., San, R. S. & Wickline, S. A. Cancer Res. 71, 3792–3801 (2011).
Poste, G. & Nicolson, G. L. Proc. Natl Acad. Sci. USA 77, 399–403 (1980).
Kaplan, R. N. et al. Nature 438, 820–827 (2005).
Costa-Silva, B. et al. Nature Cell Biol. 17, 816–826 (2015).
Steeg, P. S., Camphausen, K. A. & Smith, Q. R. Nature Rev. Cancer 11, 352–363 (2011).
About this article
Primary Thromboprophylaxis in Pancreatic Cancer Patients: Why Clinical Practice Guidelines Should Be Implemented
Cellular Oncology (2020)
Biochimica et Biophysica Acta (BBA) - Reviews on Cancer (2019)