IIBBA   05544
INSTITUTO DE INVESTIGACIONES BIOQUIMICAS DE BUENOS AIRES
Unidad Ejecutora - UE
congresos y reuniones científicas
Título:
Adenoviral oncolytic Capacity is defined by tumor cell-stroma Interaction.
Autor/es:
LOPEZ MV, VIALE D, CAFFERATA EGA, GOULD D, CHERNAJOVSKY Y AND PODHAJCER OL
Lugar:
Boston, Massachusetts, USA
Reunión:
Congreso; American Society of Gene Therapy's 11th Annual Meeting; 2008
Institución organizadora:
American Society of Gene Therapy
Resumen:
Adenoviral oncolytic Capacity is defined by tumor cell-stroma  Interaction. Veronica M Lopez1, Diego L Viale1, Eduardo A Cafferata1, David Gould2, Yuty Chernajovsky2, Osvaldo L Podhajcer1. 1Laboratory of Molecular and Cellular Therapy, Leloir Institute-CONICET, Ciudad Autonoma de Buenos Aires, Buenos Aires, Argentina. 2Bone and Joint research Unit, School of Medicine and Dentistry, University of London, United Kingdom. Tumor growth is essentially the result of a continuous cross-talk between malignant and tumor-associated stromal cells. Tumor-associated endothelial cells and fibroblasts are not merely a scaffold but are actively engaged in tumor growth. Therefore, we designed two novel CRAds to target both the malignant and stromal cell compartiments. We selected SPARC promoter since SPARC is over-expressed in malignant and tumor associated-stromal cells in human melanoma and other cancers. By using luciferase expression as a reporter gene, we performed a detailed analysis of the activity and specificity of different fragments of the SPARC promoter. A promoter sequence extending from -513 bp to +35 bp named F512 showed the best ratio of activity Vs. specificity in human melanoma cells compared to non-melanoma malignant and normal cells. We constructed a CRAd in which the E1A expression was driven by F512 promoter (AdF-512) and a second CRAd, Ad(I)-F512-TK that was armed with the herpes simplex thymidine kinase gene. Both CRAds were cytotoxic on a panel of melanoma cells but was unable to replicate in normal cells from different sources (melanocytes, keratinocytes, mammary and colon). CRAds exhibited also a certain cytotoxic effect on transformed-human microendothelial cells HMEC-1 and fetal fibroblasts, but had no lytic effct on adult fibroblasts. Ad(I)-F512-TK combined with GCV significantly enhanced the in vitro cytotoxic capacity of the CRAd both in melanoma and in HMEc-1 cells that express SPARC . Ad-F512 treatment of nude mice carrying established human melanomas inhibited tumor growth in 7 out 9 mice compared to none in the control group treated with Ad-b-gal. However, Ad-F512 was much less effective on established tumors composed of a mix of melanoma and stromal cells expressing SPARC (endothelial or fibroblasts) demonstrating that the presence of stromal cells imposed a restriction on viral activity. Interestingly, Ad(I)-F512-TK combined with GCV exhibited better efficacy than Ad-F512 on mixed tumors suggesting that adding TK/GCV can improve viral lytic activity. Remarkably, treatment of pancreatic tumors made of mix of MIA Paca 2 cells and HMEC-1 cells inhibited tumor growth in 5 out of 6 mice, demonstrating that in contrast to what occurred with melanoma tumors, the presence of stromal cells improved viral therapeutic efficacy on pancreatic tumors. Preliminary evidence indicate that conditioned media obtained from stromal cells enhanced F512 activity in pancreatic cells, while reduced F512  activity in melanoma cells. This data suggest that stromal cells can play a key in defining the therapeutic efficacy of a CRAd. Moreover, based on the paradigm of targeting malignant and stromal cells we propose d that this novel CRAd might be considered as a potential therapeutic tool for different cancedrtypes.