Affiliations 

  • 1 Department of Neurology, The Massachusetts General Hospital, and NeuroDiscovery Center, Harvard Medical School, Boston, MA, USA
  • 2 University of Massachusetts Medical School, Worcester, MA, USA
  • 3 Molecular Pathology Division, The Massachusetts General Hospital, Boston, MA, USA
  • 4 Department of Neurology, The Massachusetts General Hospital, and NeuroDiscovery Center, Harvard Medical School, Boston, MA, USA. badr.christian@mgh.harvard.edu
  • 5 Department of Neurology, The Massachusetts General Hospital, and NeuroDiscovery Center, Harvard Medical School, Boston, MA, USA. cmaguire@mgh.harvard.edu
J Neurooncol, 2018 Sep;139(2):293-305.
PMID: 29767307 DOI: 10.1007/s11060-018-2889-2

Abstract

The malignant primary brain tumor, glioblastoma (GBM) is generally incurable. New approaches are desperately needed. Adeno-associated virus (AAV) vector-mediated delivery of anti-tumor transgenes is a promising strategy, however direct injection leads to focal transgene spread in tumor and rapid tumor division dilutes out the extra-chromosomal AAV genome, limiting duration of transgene expression. Intravenous (IV) injection gives widespread distribution of AAV in normal brain, however poor transgene expression in tumor, and high expression in non-target cells which may lead to ineffective therapy and high toxicity, respectively. Delivery of transgenes encoding secreted, anti-tumor proteins to tumor stromal cells may provide a more stable and localized reservoir of therapy as they are more differentiated than fast-dividing tumor cells. Reactive astrocytes and tumor-associated macrophage/microglia (TAMs) are stromal cells that comprise a large portion of the tumor mass and are associated with tumorigenesis. In mouse models of GBM, we used IV delivery of exosome-associated AAV vectors driving green fluorescent protein expression by specific promoters (NF-κB-responsive promoter and a truncated glial fibrillary acidic protein promoter), to obtain targeted transduction of TAMs and reactive astrocytes, respectively, while avoiding transgene expression in the periphery. We used our approach to express the potent, yet toxic anti-tumor cytokine, interferon beta, in tumor stroma of a mouse model of GBM, and achieved a modest, yet significant enhancement in survival compared to controls. Noninvasive genetic modification of tumor microenvironment represents a promising approach for therapy against cancers. Additionally, the vectors described here may facilitate basic research in the study of tumor stromal cells in situ.

* Title and MeSH Headings from MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine.