RNAi-mediated knockdown of E2F2 inhibits tumorigenicity of human glioblastoma cells.

In a previous genome-wide expression profiling study, we identified E2F2 as a hyperexpressed gene in stem-like cells of distinct glioblastoma multiforme (GBM) specimens. Since the encoded E2F2 transcription factor has been implicated in both tumor suppression and tumor development, we conducted a functional study to investigate the pertinence of E2F2 to human gliomagenesis. E2F2 expression was knocked down by transfecting U87MG cells with plasmids carrying a specific silencing shRNA. Upon E2F2 silencing, in vitro cell proliferation was significantly reduced, as indicated by a time-course analysis of viable tumor cells. Anchorage-independent cell growth was also significantly inhibited after E2F2 silencing, based on cell colony formation in soft agar. Subcutaneous and orthotopic xenograft models of GBM in nude mice also indicated inhibition of tumor development in vivo, following E2F2 silencing. As expression of the E2F2 gene is associated with glioblastoma stem cells and is involved in the transformation of human astrocytes, the present findings suggest that E2F2 is involved in gliomagenesis and could be explored as a potential therapeutic target in malignant gliomas.

Upregulation of E2F1 in cerebellar neuroprogenitor cells and cell cycle arrest during postnatal brain development.

In the developing cerebellum, proliferation of granular neuroprogenitor (GNP) cells lasts until the early postnatal stages when terminal maturation of the cerebellar cortex occurs. GNPs are considered cell targets for neoplastic transformation, and disturbances in cerebellar GNP cell proliferation may contribute to the development of pediatric medulloblastoma. At the molecular level, proliferation of GNPs is regulated through an orchestrated action of the SHH, NOTCH, and WNT pathways, but the underlying mechanisms still need to be dissected. Here, we report that expression of the E2F1 transcription factor in rat GNPs is inversely correlated with cell proliferation rate during postnatal development, as opposed to its traditional SHH-dependent induction of cell cycle. Proliferation of GNPs peaked at postnatal day 3 (P3), with a subsequent continuing decrease in proliferation rates occurring until P12. Such gradual decline in proliferating neuroprogenitors paralleled the extent of cerebellum maturation confirmed by histological analysis with cresyl violet staining and temporal expression profiling of SHH, NOTCH2, and WNT4 genes. A time course analysis of E2F1 expression in GNPs revealed significantly increased levels at P12, correlating with decreased cell proliferation. Expression of the cell cycle inhibitor p18 ( Ink4c ), a target of E2F1, was also significantly higher at P12. Conversely, increased E2F1 expression did not correlate with either SMAC/DIABLO and BCL2 expression profiles or apoptosis of cerebellar cells. Altogether, these results suggest that E2F1 may also be involved in the inhibition of GNP proliferation during rat postnatal development despite its conventional mitogenic effects.

Contamination of mesenchymal stem-cells with fibroblasts accelerates neurodegeneration in an experimental model of Parkinson's disease.

Pre-clinical studies have supported the use of mesenchymal stem cells (MSC) to treat highly prevalent neurodegenerative diseases such as Parkinson's disease (PD) but preliminary trials have reported controversial results. In a rat model of PD induced by MPTP neurotoxin, we first observed a significant bilateral preservation of dopaminergic neurons in the substantia nigra and prevention of motor deficits typically observed in PD such as hypokinesia, catalepsy, and bradykinesia, following intracerebral administration of human umbilical cord-derived MSC (UC-MSC) early after MPTP injury. However, surprisingly, administration of fibroblasts, mesenchymal cells without stem cell properties, as a xenotransplantation control was highly detrimental, causing significant neurodegeneration and motor dysfunction independently of MPTP. This observation prompted us to further investigate the consequences of transplanting a MSC preparation contaminated with fibroblasts, a plausible circumstance in cell therapy since both cell types display similar immunophenotype and can be manipulated in vitro under the same conditions. Here we show for the first time, using the same experimental model and protocol, that transplantation of UC-MSC induced potent neuroprotection in the brain resulting in clinical benefit. However, co-transplantation of UC-MSC with fibroblasts reverted therapeutic efficacy and caused opposite damaging effects, significantly exacerbating neurodegeneration and motor deficits in MPTP-exposed rats. Besides providing a rationale for testing UC-MSC transplantation in early phases of PD aiming at delaying disease progression, our pre-clinical study suggests that fibroblasts may be common cell contaminants affecting purity of MSC preparations and clinical outcome in stem cell therapy protocols, which might also explain discrepant clinical results.

Human glioblastoma cells display mesenchymal stem cell features and form intracranial tumors in immunocompetent rats.

Isolation of highly tumorigenic stem-like cells from human glioblastoma specimens and cell lines has been focusing on their neural stem cells properties or capacity to efflux fluorescent dyes. Here, we report that, under standard culture conditions, human glioblastoma cells of the U87MG cell line display a predominant mesenchymal phenotype and share some of the in vitro properties of mesenchymal stem cells. Moreover, these cells were capable of forming tumors in immunocompetent rats. Infiltrative intracranial tumors could be detected 15 to 30 days post-stereotaxic cell injection within the motor cortex. Tumors were comprised by pleomorphic and mitotically active cells and displayed necrotic and hemorrhagic foci, which are common features of human glioblastomas. This rather unexpected in vivo tumorigenesis in the absence of immune suppression more closely mimics the physiological milieu encountered by tumor cells and could be explored as a xenograft orthotopic model of human glioblastomas to address new therapeutic approaches, particularly those involving immune effector mechanisms.