Glioblastoma microvesicles transport RNA and proteins that promote tumour growth and provide diagnostic biomarkers

Abstract

Glioblastoma tumour cells release microvesicles (exosomes) containing mRNA, miRNA and angiogenic proteins. These microvesicles are taken up by normal host cells, such as brain microvascular endothelial cells. By incorporating an mRNA for a reporter protein into these microvesicles, we demonstrate that messages delivered by microvesicles are translated by recipient cells. These microvesicles are also enriched in angiogenic proteins and stimulate tubule formation by endothelial cells. Tumour-derived microvesicles therefore serve as a means of delivering genetic information and proteins to recipient cells in the tumour environment. Glioblastoma microvesicles also stimulated proliferation of a human glioma cell line, indicating a self-promoting aspect. Messenger RNA mutant/variants and miRNAs characteristic of gliomas could be detected in serum microvesicles of glioblastoma patients. The tumour-specific EGFRvIII was detected in serum microvesicles from 7 out of 25 glioblastoma patients. Thus, tumour-derived microvesicles may provide diagnostic information and aid in therapeutic decisions for cancer patients through a blood test.

Figure 1. Glioblastoma cells produce microvesicles containing RNA

Scanning EM image of a primary glioblastoma cell (bar = 10 μm). (b) Higher magnification showing the microvesicles on the cell surface. Vesicles can be binned into diameters of around 50 nm and 500 nm (bar = 1 μm). (c) Microvesicles were exposed to RNase A or mock-treated prior to RNA isolation and levels of RNA determined (n = 5). (d) Bioanalyzer data shows the size distribution of total RNA extracted from primary glioblastoma cells and (e) microvesicles isolated from them. The smallest peak represents an internal standard. The two prominent peaks in (d) (arrows) represent 18S (left) and 28S (right) ribosomal RNA, absent in microvesicles.

Figure 2. Characterization of the microvesicle RNA

(a, b) Scatterplots of mRNA levels in the microvesicles compared to donor cells from two different experiments. Linear regressions showed that levels in cells versus microvesicles were not well correlated. (c, d) In contrast, mRNA intensities in two different cell or two different microvesicle preparations were closely correlated. (e) 3426 genes were found to be more than 5-fold differentially distributed in the microvesicles as compared to the cells from which they were derived (p-value <0.01). (f) The biological process ontology of the 500 most abundant mRNA species in the microvesicles is displayed. (g) The intensity of microvesicle RNAs belonging to ontologies related to tumor growth is shown with the x-axis representing the number of mRNA transcripts present in the ontology. The median intensity levels on the arrays were 182. (h) Levels of mature miRNAs in microvesicles and glioblastoma cells from two different patients (GBM1 and GBM2) were analysed using quantitative miRNA RT-PCR. The cycle threshold (Ct) value is presented as the mean ± SEM (n = 4).

Figure 3. Glioblastoma microvesicles can deliver functional RNA to HBMVECs

(a) Purified microvesicles were labelled with membrane dye PKH67 (green) and added to HBMVECs. The microvesicles were internalised into endosome-like structures within an hr. (b) Microvesicles were isolated from glioblastoma cells stably expressing Gluc. RNA extraction and RTPCR of Gluc and GAPDH mRNAs showed that both were incorporated into microvesicles. (c) Microvesicles were then added to HBMVECs and incubated for 24 hrs. The Gluc activity was measured in the medium at 0, 15 and 24 hrs after microvesicle addition and normalized to Gluc activity in microvesicles. The results are presented as the mean ± SEM (n = 4).

Figure 4. Glioblastoma microvesicles stimulate angiogenesis in vitro and contain angiogenic proteins

a) HBMVECs were cultured on Matrigel^™ in basal medium (EBM) alone, or supplemented with GBM microvesicles (EBM+MV) or angiogenic factors (EGM). Tubule formation was measured after 16 hrs as average tubule length ± SEM compared to cells grown in EBM (n = 6). (b) Total protein from primary glioblastoma cells and microvesicles (MV) from them (1 mg each) was analysed on a human angiogenesis antibody array. (c) The arrays were scanned and the intensities analysed with the ImageJ software (n = 4).