Adipose tissue mesenchymal stromal cells as therapeutic vehicles against glioblastoma

Abstract

Lately adipose tissue mesenchymal stem cells (hAMSCs) have emerged as cellular vehicles for therapy of solid tumors, due to their ease of isolation and manipulation, and wound/tumor homing capacity. HAMSCs have been successfully used in suicide gene therapy, employing the prodrug activating system based on Herpes simplex virus type I thymidine kinase (HSV-TK)/ganciclovir (GCV). In the current study we demonstrate an effective model of glioblastoma therapy based on the use of genetically modified hAMSCs and in vivo monitoring of tumor and therapeutic cells. Due to the capacity of photons to pass through living tissue, non-invasive monitoring by bioluminescence imaging has become a cutting edge technology for the study of ongoing biological processes in small lab animals. This technique uses photoproteins as cell reporters that generate light photons as byproducts of chemical reaction. Luciferases catalyze the oxidation of a substrate (luciferin) in the presence of ATP and oxygen to generate oxiluciferin, ADP and light photons. In spite of the apparent opacity of tissues, light can be detected at several millimeters of depth in live animals. Combination of several types of luciferases allows simultaneous monitoring of different cell populations proliferation or differentiation. We stably transduced hAMSCs for expression of Renilla luciferase, HSV-TK and red fluorescent protein, generating RLuc-R-TK-AMSC and U87MG human malignant glioma cells for expression of Firefly luciferase and green fluorescent protein, generating Pluc-G-U87 cells. SCID mice were stereotactically implanted in the brain first with Pluc-G-U87 and RLuc-R-TK-AMSC afterwards. Mice were subjected to GCV treatment and the therapeutic process was evaluated in real time. Tumor response was monitored in vivo by BLI. Therapeutic cell differentiation was assessed by labeling the above Renilla luciferase expressing hAMSCs with a Firefly luciferase reporter regulated by the CD31, endothelial specific promoter and in vivo monitorization. Endothelial lineage differentiation of hAMSC was impaired, by Notch1 shRNA, and therapeutic effect was assessed by BLI monitorization of tumor response. In our model of therapy, we show that tumor size can be continuously monitored by BLI and is significantly reduced (99,9% relative to control untreated tumours) by repeated inoculations in the tumours with thymidine kinase expressing hAMSCs followed by the prodrug ganciclovir. Moreover treatment resulted in a significant prolongation of survival time. In addition, the combination of BLI and confocal microscopy analysis of therapeutic cells suggests that efficient tumor eradication results from hAMSCs homing to tumor vessels, where they differentiate to endothelial cell lineage, intensifying their cytotoxic effect by destroying tumor vasculature and negating nutrient supply. Besides, hAMSCs endothelial differentiation inhibition resulted in an inefficient therapeutic effect compared to normal hAMSC (64% vs 6% respectively). Close association between hAMSCs and gliomas stem cells integrated in the tumor vascular system seems to be essential for an effective tumor reduction. We suggest that efficient tumor eradication is due to hAMSCs endothelial differentiation and tube location that intensifies its cytotoxic function, destroying tumor vasculature and inhibiting nutrient supply to tumor cells. Thus we propose adipose tissue hAMSCs as useful vehicles for clinical applications to deliver localized therapy to glioma surgical borders after tumor resection.