Mixed anionic and cationic polyphosphazene complexes for effective gene delivery to glioblastoma in vitro and in vivo

Gene delivery vectors that are safe, efficient and affordable could significantly enhance the prospects for genetic-based therapies. Here we describe an approach to such vectors, using new variants of polyphosphazene materials and describe the synthesis of a series of degradable polyphosphazenes with both cationic and anionic side-chains, and report their use as mixed polyelectrolyte complexes for DNA and RNA delivery to glioblastoma cells in vitro and in vivo. Precursor poly(allylamino-phosphazene)s were converted to cationic and anionic derivatives via a,w-thiolated alkylamines and alkylcarboxylates, respectively. Simultaneous co-incubation of alkylamine- and alkylcarboxylate-poly(phosphazenes) with nucleic acids generated polyelectrolyte complexes which were more compact than poly(alkylamino-phosphazene):DNA analogues but with similar positive surface charges. Screening of a series of these complexes for transfection of U87MG glioblastoma cells, showed that 6-mercaptohexanoic acid substituted poly(phosphazene)s mixed in the polycation/DNA complexes resulted in the highest luciferase expression in the cells. These data were consistent with an increased buffering capability of the 6-mercaptohexanoic acid substituted polymer across the early endosomal pH range in comparison with other anionic side-chain substituted polymers. Transfection assays in 3D spheroid models and in subcutaneous xenograft U87MG tumours confirmed higher transgene expression for these mixed cationic and anionic poly(phosphazene)s compared to the related poly(alkylamino-phosphazene)-DNA complexes, and also to PEI-DNA complexes. Extension of the approach to siRNA delivery showed that the mixed cationic and anionic poly(phosphazene)s were able to silence a gene encoding for a kinase implicated in tumour progression (DYRK1A), resulting in a reduced renewal ability of U87MG cells in vitro and in delay of tumour growth in a xenograft model