Barriers to horizontal cell transformation by extracellular vesicles containing oncogenic H-ras

Extracellular vesicles (EVs) enable the exit of regulatory, mutant and oncogenic macromolecules (proteins, RNA and DNA) from their parental tumor cells and uptake of this material by unrelated cellular populations. Among the resulting biological effects of interest is the notion that cancer-derived EVs may mediate horizontal transformation of normal cells through transfer of mutant genes, including mutant ras. Here, we report that H-ras-mediated transformation of intestinal epithelial cells (IEC-18) results in the emission of exosome-like EVs containing genomic DNA, HRAS oncoprotein and transcript. However, EV-mediated horizontal transformation of non-transformed cells (epithelial, astrocytic, fibroblastic and endothelial) is transient, limited or absent due to barrier mechanisms that curtail the uptake, retention and function of oncogenic H-ras in recipient cells. Thus, epithelial cells and astrocytes are resistant to EV uptake, unless they undergo malignant transformation. In contrast, primary and immortalized fibroblasts are susceptible to the EV uptake, retention of H-ras DNA and phenotypic transformation, but these effects are transient and fail to produce a permanent tumorigenic conversion of these cells in vitro and in vivo, even after several months of observation. Increased exposure to EVs isolated from H-ras-transformed cancer cells, but not to those from their indolent counterparts, triggers demise of recipient fibroblasts. Uptake of H-ras-containing EVs stimulates but fails to transform primary endothelial cells. Thus, we suggest that intercellular transfer of oncogenes exerts regulatory rather than transforming influence on recipient cells, while cancer cells may often act as preferential EV recipients.


Supplementary Figure S3: Secreted factors from RAS-3 cancer cells are unable to recapitulate the effects of endogenous mutant H-ras expression on the EV uptake by non-tumorigenic IEC-18 cells. H-ras transformation changes
cellular secretome, thereby contributing to the expression of malignant phenotype. We asked whether these paracrine secreted factors (e.g. growth factors, enzymes) may regulate EV uptake in IEC-18 cells similarly to the effects exerted by the endogenous mutant H-ras. To accomplish this, RAS-3 EVs labelled with PKH26 were added to IEC-18 cultures in the presence or absence of the soluble fraction of RAS-3 conditioned medium (CM). FACS analysis revealed that this treatment did not change EV uptake, and thereby did not recapitulate the effects of endogenous H-ras expression (See Figure 2). Figure S4: MEK/MAPK pathway activity is not required for EV uptake by H-ras-transformed cancer cells. Enforced HRAS transformation in IEC-18 cells (RAS-3 clone) leads to the onset of EV uptake. Since HRAS activates the MAPK cascade, which has been implicated in EV endocytosis, we asked whether blockade of the MEK/MAPK pathway using the pharmacological inhibitor PD98059 leads to the reversal of HRAS-mediated EV uptake (See Figure 2). RAS-3 cells were pre-treated for 24 hours with PD98059 at 50 µM or with vehicle and subsequently exposed to PKH26-labelled EVs purified from RAS-3 cell cultures. No change in the avid transfer of EV fluorescence was observed under these conditions, suggesting that MEK/MAPK activity is not essential for oncogeneinduced uptake of EVs. Figure S6: Studies interrogating the ability of extracellular oncogenic activity associated with RAS-3 cells to act as a potential trigger of horizontal transformation in vivo. (A) Experimental design: Immune-deficient SCID mice were used as recipients of subcutaneous grafts consisting of 2 × 10 6 viable RAT-1 cells pre-treated in culture with RAS-3-derived EVs, as in Figure 2. Alternatively, RAT-1 cells were injected in mixture with mitotically inactive but viable RAS-3 cells (5 × 10 4 ), which had been pre-treated with Mitomycin C (MitoC). In a similar manner, injections of MEFp53-/-primary fibroblasts were carried out. Finally, Mitomycin C-treated RAS-3 cells were injected alone, in which case the recipients of their related extracellular transforming activity would be only the host (mouse) cells at the site of injection or systemically. Untreated RAS-3, RAT-1 and MEFp53-/-cells were injected as controls. (B) Cumulative survival of mice upon injection of the aforementioned cellular preparations (Kaplan-Meier curves). All mice injected with RAS-3 cells rapidly developed aggressive tumors, while spontaneous growth of RAT-1 cells occurred in a fraction of mice and after prolonged latency. No increase in tumor formation due to horizontal transformation or intercellular transfer of oncogenic activity from RAS-3 cells to non-transformed recipients was observed in these experiments (see Table 1).

Immunofluorescence
Cultured cells were fixed for 10 minutes in 4% paraformaldehyde, permeabilized with 0.2% Triton X-100 and stained overnight at 4°C with antibodies. This was followed by their respective secondary Alexa Fluor antibodies (Invitrogen). Imaris software (Bitplane) was used for the analysis of confocal images.

Detailed EV isolation methods
EVs were obtained at standard (2 100 mm petri dishes, each containing 9 ml of conditioned media and 0.65 µg/µl of EV protein) or quadruple concentrations by ultracentrifugation as previously described [1][2][3][4][5]. Briefly, conditioned media was centrifuged at 400-g for 10 minutes to remove cell debris, followed by filtration through a 0.2 µm PES filter (#565-0020, Thermo Fisher Scientific). The filtrate was ultracentrifuged at 110,000-g for 1 hour to isolate EVs. For sucrose gradient centrifugation [5], the filtrate was centrifuged at 110,000-g for 1 hour, and the pellets were resuspended in 2 ml of 20 mM HEPES, 2 M sucrose. The resulting samples were transferred to the bottom of the Beckman SW41 centrifuge tube and slowly layered with a continuous gradient of sucrose, from 2 to 0.25 M. The samples were ultracentrifuged in a swinging bucket rotor for 17 hours at 210,000-g with the brake set on low. After centrifugation, the samples in each tube were separated into 10 fractions of 1 ml each, starting from the top. Each fraction was transferred into a 3-ml tube, and mixed with 2 ml of HEPES (20 mM), centrifuged at 110,000-g for 1 hour, resuspended in 50 μl of PBS. EV protein extracts were obtained by directly adding radioimmunoprecipitation assay buffer (RIPA) to the pellet after the spin. Similarly, EV nucleic acid extracts were obtained by either adding Trizol (RNA) or proteinase K-treated lysis buffer (DNA). For the total number of particles, conditioned media were loaded onto the nanoparticle tracking analysis system (NTA; #NS500, Nanosight) and five recordings of 30 seconds were obtained and processed using NTA software.

Detailed histology (H&E staining)
Tumor samples from xenograft experiments were fixed in 4% paraformaldehyde immediately upon their resection from the mice. A series of automated processing steps were then executed on the samples by the Leica TP 1050 tissue processor. The resulting paraffin-embedded blocks were sectioned into 4 µm thick tissue sections using the American Optical microtome and were placed on pre-coated microscope slides. These slides were de-waxed in xylene and then hydrated in a series of washes from 95% ethanol to 50% ethanol. The slides were visualized by haematoxylin and eosin (H&E) staining, which was done by incubating them with 1.5% haematoxylin, pH 2.5, Blueing solution and eosin solution.
In our study, EVs released from RAS-3 cells exerted pro-survival effects in the context of primary human endothelial cells. This is of interest as processes of tumor angiogenesis are often dissimilar from those occurring during normal vascular development, and may include changes in endothelial cells that resemble genetic transformation [47]. For example, tumorrelated endothelial cells may exhibit aneuploidy [48], or oncogenic features of their adjacent cancer cells [49,50]. The impact of oncogenic EVs on these events is uncertain but cultured endothelial cells exposed to EV-associated oncogenic EGFR acquire the ability to activate vascular endothelial growth factor receptor 2 (VEGFR2) in an apparently autocrine manner [4]. In this sense, cancerrelated EVs may be viewed as unique mediators of abnormal tumor angiogenesis [6].
In our experiments, the effects of H-ras-containing EVs on HUVEC cells were consistent with their postulated role in tumor angiogenesis [4,26]. This observation is in line with the mounting evidence for EV-mediated cell-cell communication, including intercellular transfer of potent oncogenic, signal-transducing and regulatory molecules to the vascular and other cellular compartments, as reviewed in [6].
As mentioned earlier, EV cargo may contain bioactive tumor suppressors [41,42], bioactive regulatory proteins [43], microRNA [44], mRNA [24,26,45], singlestranded DNA [28], as well as histones, chromatin and gDNA [3,29,30]. It is of note that our study documents that mutant H-ras induces emission of genomic DNA as cargo of cancer-related EVs. The modulatory role of these molecular entities in cancer progression and metastasis is increasingly well documented in vitro and in vivo [4,15,16,26]. Interestingly, we observed no obvious effect of EVs purified from IEC-18 conditioned medium.
Studies are underway to elucidate pathways governing EV uptake by various cell types, including the role of endocytosis mediated by proteoglycan and other mechanisms [51,52]. The regulation of the intracellular fate and processing of different EV-related macromolecules by various types of recipient cells is of considerable interest but remains largely unclear. Published reports suggest cell-specific intracellular degradation [53], re-expression [2] or secretion [22] of cargo molecules, but the fate of EV-related DNA remains to be determined. It is possible that elimination of this material may involve removal of cells harboring EVrelated DNA, as suggested by detrimental effects of the excessive EV uptake [54].
Finally, while we used transfer of H-ras gDNA to document molecular transfer between donor RAS-3 cells and different EV recipients, it is likely that multiple molecular perturbations downstream of RAS and secondary to H-ras transformation (e.g. due to genetic instability and drift) may influence the content and biological activity of RAS-3-derived EVs. For example, several RAS-regulated target and effector proteins, transcripts and microRNA and other non-coding RNA may contribute to changes we observed following the RAS-3 EV uptake [40]. Implicitly, some of these macromolecules may be essential for H-ras-dependent transformation and undergo intercellular transfer as cargo of EVs. These considerations notwithstanding, permanent transformation was not observed in any of the RAS-3-EV recipient cells analysed in our study. Thus, in cases where horizontal transformation occurs experimentally or in vivo upon contact of cancer and stromal cells is probably relatively rare and may entail viral transmission, epigenetic reprogramming or other secondary changes rather than direct transfer and reactivation of genomic sequences carrying mutant oncogenes. It is also possible that cellular responses to exogenous DNA may activate pathways of sequestration, elimination and stress response reminiscent of those triggered by viral infection. Indeed, we observed toxicity of EVs containing genomic DNA and derived from cancer cells, when added at high concentrations, but this was not the case for comparable levels of EVs derived from non-transformed cells and devoid of DNA. Mutant RAS represents one of the most potent transforming influences and a paradigm of molecular genesis of human cancers. While intrinsic barriers block these effects by provoking demise of normal cells exposed to mutant RAS signalling, so too, as shown in our study, do various normal cells populations mount biological barriers against transformation through horizontal transfer of RAS as cargo of extracellular vesicles.