Horizontal Transfer of Malignant Traits and the Involvement of Extracellular Vesicles in Metastasis

Abstract

Metastases are responsible for the vast majority of cancer deaths, yet most therapeutic efforts have focused on targeting and interrupting tumor growth rather than impairing the metastatic process. Traditionally, cancer metastasis is attributed to the dissemination of neoplastic cells from the primary tumor to distant organs through blood and lymphatic circulation. A thorough understanding of the metastatic process is essential to develop new therapeutic strategies that improve cancer survival. Since Paget's original description of the "Seed and Soil" hypothesis over a hundred years ago, alternative theories and new players have been proposed. In particular, the role of extracellular vesicles (EVs) released by cancer cells and their uptake by neighboring cells or at distinct anatomical sites has been explored. Here, we will outline and discuss these alternative theories and emphasize the horizontal transfer of EV-associated biomolecules as a possibly major event leading to cell transformation and the induction of metastases. We will also highlight the recently discovered intracellular pathway used by EVs to deliver their cargoes into the nucleus of recipient cells, which is a potential target for novel anti-metastatic strategies.

Keywords: cancer; exosomes; extracellular vesicles; metastasis; microenvironment; nucleoplasmic reticulum; oncogene; tumor suppressor gene.

Figure 1

EVs derived from metastatic colon cancer cells promote the morphological transformation of non-aggressive cancer cells. (A,B) The established isogenic colorectal cancer cell lines consisting of highly metastatic SW620 cells and non-metastatic SW480 cells provide a cellular model in which EVs (exosomes and/or microvesicles (MVs)) derived from the former can transform the latter and impact their malignant properties, notably by changing their migration patterns from mesenchymal to amoeboid motility [127,128]. The latter phenotypic alteration can be prevented either by intercepting CD9⁺ EVs derived from SW620 cells using a Fab fragment antibody directed against CD9 that blocks their internalization by SW480 cells (A), or by blocking the nuclear transfer of cargo carried by endocytosed EVs using drugs such as itraconazole (ICZ) or PRR851 (B), which act on the tripartite protein complex required for the translocation of late endosomes into the NR (for more details, see Section 10). EE, early endosome; LE, late endosome; Fab, fragment antigen-binding; NEI, nuclear envelope invagination. Modified from Ref. [124].

Figure 2

Target cells efficiently internalize EVs derived from colorectal cancer patient sera, and upon inoculation into mice give rise to tumors displaying an intestinal adenocarcinoma phenotype. (A–E) Sera prepared from patients with colorectal cancer and liver metastasis or EVs isolated from them (A) as described in Ref. [70] were incubated with oncosuppressor-deficient cells such as BRCA1-KO fibroblasts for a period of two weeks (B). Afterwards, cells were harvested and incorporated in Matrigel matrix (C) prior to their subcutaneous injection into NOD-SCID mice (D). Four weeks after cell transplantation, the animals were sacrificed and the xenotransplant excised (E). (F) Tissues containing growing tumor lesions were processed for Hematoxylin and Eosin (H&E staining) and/or immunohistochemistry for various markers, as indicated. Note that fibroblasts changed their fate (i.e., loss of Vimentin), and expressed markers of highly proliferative intestinal adenocarcinoma (Ki67, carcinoembryonic antigen (CEA), cytokeratin (CK)20, homeobox transcription factor CDX-2, and anion exchanger (AE)1/AE3), while being negative for the epithelial marker CK7. Micrographs were similar to those presented in Ref. [180]. Scale bar: 100 μm.

Figure 3

Cancer-cell-derived EVs transfer malignant traits in vivo. (A–D) EVs (about 4 × 10⁶) enriched by differential centrifugation from the conditioned medium of murine MC38 colon adenocarcinoma cells (A, for technical details see Ref. [179]) were injected in the lateral tail vein of NOD-SCID mice every other day for 5 weeks (B). Four weeks later, the animals were sacrificed and lungs were collected (C). Tissues were processed for H&E staining and/or immunohistochemistry for various markers, as indicated (D). The tumors that developed had a proliferative phenotype, as indicated by Ki67 labeling, and had characteristics of poorly differentiated colon cancer, as they were positive for CK20, CDX2 and AE1/AE3 markers, and of lung cancer, as marked by CK7, TTF1 and Napsin. Micrographs were similar to those presented in Ref. [179]. Scale bars: 50 µm.

Figure 4

Potential roles of EVs and their cargo in the horizontal transfer of metastatic traits in vivo. (A) The primary tumor cells release factors, either naked or encapsulated in EVs, that travel through the lymphatic system into the regional lymph nodes. (B) In the lymph nodes, the development of immunity may scavenge these factors with the subsequent inhibition of the metastatic process. Alternatively, tolerance might ensue with subsequent uptake of cancer factors and integration within lymph node cells. (C) If oncotolerance is established, oncofactors can freely travel through the bloodstream, undetected by the immune system, and reach a PMN. (D) At the PMN, three distinct scenarios can occur: 1. resident cells may be refractory to the uptake of the oncogenic factors (no penetration); 2. receptive putative target cells may express specific receptors or others proteins that facilitate oncofactors’ entry/penetration and integration into the genome without activation of gene transcription (penetration (a)); or 3. oncofactors may be taken up by target cells, with subsequent integration and immediate expression of cancer genes that would determine synchronous metastasis (penetration (b)). In the case of penetration (a), a subsequent decline in immune function would lead to the reactivation of the integrated genes with malignant transformation of the dormant cell and initiation of late metastasis. Adapted from Ref. [153]. The illustration is not drawn to scale.

Figure 5

Representation of nuclear envelope invagination-associated late endosomes containing endocytosed EVs and their implications for EV-mediated signaling. (A) After endocytosis, EVs move into the endosomal system, and in Rab7⁺ late endosomes they are translocated by a microtubule-dependent mechanism into the NR, especially into type II NEIs, which are often found in close proximity to the nucleolus. Nuclear translocation of late endosomes containing endocytosed EVs requires the interaction of VOR complex proteins: outer nuclear membrane (ONM)-associated VAP-A (yellow), lipid-binding ORP3 (green) and endosome-associated Rab7 (red). A lamin-rich proteinaceous meshwork underlies the inner nuclear membrane (INM), and nuclear pore complexes are found in type II NEIs [72,73]. (B) Late endosomes containing endocytosed EVs tether to the ONM via the VOR complex in which cytoplasmic ORP3 binds through its FFAT motif (FFAT being an acronym for two phenylalanines (FF) in an acidic tract) to the membrane protein VAP-A. The pleckstrin homology domain (PH) of ORP3 may mediate its binding to the late endosomal membrane. The domain interaction of Rab7 with the VAP-A-ORP3 complex needs to be defined. After the EV late endosomes fusion, which may be stimulated by a low pH environment, their cargoes are released into the cytoplasmic core of the NEI. Such a restricted space could promote interactions between EV cargoes (e.g., proteins or miRNAs) and their host targets, especially those exported from the nucleus (e.g., mRNA) (a). Alternatively, potential docking (?) of late endosomes to nuclear pores via the VOR complex could further promote the nuclear import of EV cargoes in which importin β1, including that carried by EVs, could play a role, as suggested by importazole treatment [72,73] (b). Modified from Ref. [20].