Currently in vivo gene delivery by synthetic vectors is m). Intraventricular injection of complexes formulated in hindered by the limited diffusibility of complexes in extraglu-cose showed the complexes to be highly diffusible in cellular fluids and matrices. Here we show that certain the cerebrospinal fluid of newborn and adult mice, diffusing formulations of plasmid DNA with linear polyethylenimine from a single site of injection throughout the entire brain (22 kDa PEI, ExGene 500) can produce complexes that ventricular spaces. Transfection efficiency was followed by are sufficiently small and stable in physiological fluids so histochemistry of ␤-galactosidase activity and double as to provide high diffusibility. When plasmid DNA was forimmunocytochemistry was used to identify the cells transmulated with 22 kDa PEI in 5% glucose, it produced a fected. Transgene expression was found in both neurons homogeneous population of complexes with mean diamand glia adjacent to ventricular spaces. Thus, this method eters ranging from 30 to 100 nm according to the amount of formulation is promising for in vivo work and may well of PEI used. In contrast, formulation in physiological saline be adaptable to other vectors and physiological models. produced complexes an order of magnitude greater (у1 Keywords: cationic polymers; polyethylenimine; brain; non-viral; gene transfer Gene transfer into the central nervous system (CNS) could provide a valuable technique for exploring the function of neural systems and a potential basis for therapy of neural disease. Until recently, the CNS appeared to present insurmountable barriers to the transfer of genes. First, the blood-brain barrier tends to preclude vascular delivery.1 Second, the principally postmitotic neuronal population prevents the use of retroviruses. Thus, to circumvent these problems and to obtain substantial levels of foreign gene expression in the CNS, progress must be made with vectors suitable for intrathecal delivery.Ideally vectors for gene transfer in the brain should have the following characteristics: they should not incur any safety risks (neither infectious, nor toxic), they should be easy to produce and verify, they should not limit the size of genes to be vectorised and they should be able to transduce post-mitotic neurons. Non-viral vectors can potentially fulfil all of these criteria. [2][3][4][5][6][7] In particular the branched cationic polymer polyethylenimine (PEI) has already been shown to be a versatile and efficient vector in the mammalian brain. In this study, we chose to examine the effect of formulation procedures (glucose or saline solutions) on the size and in vivo transfection activity of a mixture of linear polymers (repeated units of CH 2 -CH 2 -NH) with a mean MW of 22 kDa (Exgene 500; Euromedex, Souffleweyersheim, France). As the complexes formed in glucose were shown to be an order of magnitude smaller than those formed in saline, we examined the diffusibility and gene transfer efficiency of these complexes following injection into the...

Uncontrolled interactions of gene vectors and drug carriers in and with an in vivo environment pose serious limitations to their applicability. In order to reduce such interactions we have designed, synthesized and applied novel copolymers of poly(ethylene glycol) and reactive linkers which are derivatized with anionic peptides after copolymerization. The anionic copolymer derivatives are used to coat positively charged nonviral gene vectors by electrostatic interactions. The copolymer coat confers to polyelectrolyte colloids of DNA and polycations steric stabilization in their minimal size and prevents salt-and serum albumin-induced aggregation.

No abstract