Systemic miRNA-7 delivery inhibits tumor angiogenesis and growth in murine xenograft glioblastoma.

Tumor-angiogenesis is the multi-factorial process of sprouting of endothelial cells (EC) into micro-vessels to provide tumor cells with nutrients and oxygen. To explore miRNAs as therapeutic angiogenesis-inhibitors, we performed a functional screen to identify miRNAs that are able to decrease EC viability. We identified miRNA-7 (miR-7) as a potent negative regulator of angiogenesis. Introduction of miR-7 in EC resulted in strongly reduced cell viability, tube formation, sprouting and migration. Application of miR-7 in the chick chorioallantoic membrane assay led to a profound reduction of vascularization, similar to anti-angiogenic drug sunitinib. Local administration of miR-7 in an in vivo murine neuroblastoma tumor model significantly inhibited angiogenesis and tumor growth. Finally, systemic administration of miR-7 using a novel integrin-targeted biodegradable polymeric nanoparticles that targets both EC and tumor cells, strongly reduced angiogenesis and tumor proliferation in mice with human glioblastoma xenografts. Transcriptome analysis of miR-7 transfected EC in combination with in silico target prediction resulted in the identification of OGT as novel target gene of miR-7. Our study provides a comprehensive validation of miR-7 as novel anti-angiogenic therapeutic miRNA that can be systemically delivered to both EC and tumor cells and offers promise for miR-7 as novel anti-tumor therapeutic.

Colorectal cancer candidate biomarkers identified by tissue secretome proteome profiling.

Colorectal cancer (CRC) is a major health problem. Biomarkers associated with molecular changes in cancer cells can aid early detection, diagnosis, prognosis, therapy selection, and disease monitoring. Tumor tissue secretomes are a rich source of candidate biomarkers. To identify CRC protein biomarkers, secretomes of four pairs of human CRC tissue and patient-matched normal colon tissue samples, and secretomes of five CRC cell lines were analyzed by GeLC-MS/MS. Subsequent data analysis was based on label-free spectral counting, Ingenuity Pathway Analysis, Secretome/SignalP, STRING and Cytoscape, resulting in 2703 protein identifications in the tissue secretomes, of which 409 proteins were significantly more present in CRC samples than in controls. Biomarker selection of 76 candidates was based on consistent and abundant over-representation in cancer- compared to control-secretomes, and presumed neoplastic origin. Overlap analysis with previously obtained datasets revealed 21 biomarkers suited for early detection of CRC. Immunohistochemistry confirmed overexpression in CRC of one candidate marker (MCM5). In conclusion, a human reference dataset of 76 candidate biomarkers was identified for which we illustrate that combination with existing pre-clinical datasets allows pre-selection of biomarkers for blood- or stool-based assays to support clinical management of CRC. Further dedicated validation studies are required to demonstrate their clinical applicability. BIOLOGICAL SIGNIFICANCE: Tissue secretome proteomes are a rich source of candidate biomarkers. Several secretome proteome datasets have been obtained from pre-clinical in vitro and in vivo colorectal cancer (CRC) model systems, yielding promising CRC biomarkers obtained under well-defined experimentally controlled conditions. However, which of these biomarker proteins are actually secreted by human CRC samples was not known. To our knowledge, this is the first study that directly compares secretome proteomes from clinically relevant human CRC tissues to patient-matched normal colon tissues. We identified 76 human CRC protein biomarkers that may facilitate blood-based or stool-based assay development to support clinical management of CRC. Overlap analysis with datasets from well-defined pre-clinical studies helps to determine what clinical application suits these human CRC biomarkers best, i.e. early detection, diagnosis, prognosis, therapy selection, and/or disease monitoring of CRC. This is demonstrated for a CRC mouse model dataset, revealing 21 human CRC biomarkers suited for early detection of CRC.