Nonviral Vectors for the Delivery of Small Interfering RNAs to the CNS

Inmaculada Posadas; Francisco Javier Guerra; Valentín Ceña


Nanomedicine. 2010;5(8):1219-1236. 

In This Article

PEI Derivatives

Polyethylenimine is an organic polymer with high cationic-charge density potential owing to the fact that every third atom, in the case of branched PEI, is a nitrogen that takes part in a secondary amino group that can be protonated and therefore can bind nucleic acids electrostatically. PEI will spontaneously adhere and condense DNA to form toroidal complexes that are readily endocytosed by cells. In addition, PEI is able to retain an important buffering capacity at virtually any pH, protecting DNA and escaping from lysosomal nuclease degradation. The polymer exists in two forms: the linear form, obtained from cationic polymerization from a 2-substituted 2-oxazoline monomer; and the branched form, produced by cationic polymerization of aziridine monomers. It has been established that PEIs with branched structure condense nucleic acids to a greater extent than linear PEIs.[91]

PEI transfection efficiency was shown to be higher when low-MW PEIs were used.[92] In addition, preparing PEI/DNA complexes in 5% glucose instead of in NaCl resulted in higher transfection.[93] In fact, increasing NaCl concentration markedly decreases transfection efficiency.[94] This effect seems to be related to the diameter of the particles formed that reaches approximately 100 nm in the presence of glucose and above 1 µm when formed in the presence of NaCl.[93]

Transfection of Neurons in vitro

As soon as PEI was described as an efficient vector to deliver DNA and oligonucleotides inside cells, PEI transfection efficiency was tested in postnatal primary neurons, including primary cultures of granule cells of the cerebellum, hippocampal pyramidal neurons, primary sensory neurons of the dorsal root ganglia and sympathetic neurons from the superior cervical ganglia.[95]

The initial experiments demonstrated that the transfection efficiency was low: approximately 1% for cerebellar granular cells and approximately 20% for sympathetic neurons from the superior cervical ganglia.[95] In addition, neuronal toxicity was apparent for PEI concentrations above 150 µM. In spite of the low transfection efficiency, PEI-mediated transfection was compatible with electrophysiological techniques, making this approach suitable for single cell studies, especially on ionic channels.[95]

These initial experiments highlighted the two main problems that have been found for the use of PEIs to transfect neuronal cells: low transfection efficiency and toxicity.[96,97]

One of the main problems described for PEIs is that they can be toxic to certain types of cells (including neurons) depending on MW, structure and concentration of the molecule used. The size of the particles appears to be related to transfection efficiency and toxicity. In fact, it has been described that high-MW PEIs induced rapid necrotic-like changes resulting from perturbation of the plasma membrane, followed by activation of a mitochondrial-mediated apoptosis in Jurkat cells, human umbilical vein endothelial cells and THLE3 hepatocyte-like cells.[98] By contrast, low-MW PEIs (<2000 Da) displayed much less toxicity but almost no transfection efficiency.[99] PEI toxicity appears to involve the disruption of the endosome/lysosome complex.[100,101] Excess proton accumulation in endosomes, due to the buffer capacity of protonable polymers, leads to counterion and water accumulation resulting in osmolysis. Accordingly, many efforts have been made to improve biocompatibility of high-MW PEIs by modifying their solubility, biodegradability and chemical homogeneity.[102,103]

As mentioned previously, particle size also plays a relevant role in transfection efficiency. Linear PEI polymers show lower levels of transfection than branched PEIs. Using linear 22 kDa PEI, a transfection efficiency of approximately 9% in primary sympathetic neurons was obtained,[104] while it rose to 15% in rat hypothalamic neurons transfected with a branched PEI (600–800 kDa).[16] This better transfection of branched PEIs appears to be related to a larger size of the resulting particle complex of linear PEIs, compared with the branched form of the polymer with similar MW. For example, transfection efficiencies of 25 kDa linear PEI (L-PEI 25) in rat cerebral glial cells and rat cortical neurons was smaller than that obtained for 25 kDa branched PEIs (PEI 25). The particle size formed with L-PEI 25 was 782 nm, while the particle size formed with PEI 25 was 70 nm.[105] This would support the principle that, in general, particles too large may not be effectively endocytosed within the cell.

Different approaches have been used to increase PEI transfection efficiency, including linking low-MW PEI (600 Da) to β-cyclodextrin, a commonly used cyclic oligosaccharide drug carrier,[106] which markedly increases the transfection efficiency of naive low-MW PEI in neurons without signs of toxicity.[107]

Another strategy consists of combining bioactive peptides to PEI in order to enhance gene delivery. Tet1-PEI, a 12-mer peptide that binds to neurons at their presynaptic terminals and is retrogradely transported in neurons,[108] and PEI-HGP, a viral-derived peptide from HIV-1 with lytic activity,[109] have both been used, combined with branched 25 KDa PEI, to deliver plasmid and siRNA to neuron-like cells. Both peptides increased plasmid and siRNA transfection efficiency in PC12 cells.[110] In addition, selective targeting to neurons has been attempted using the 50 kDa nontoxic fragment from tetanus toxin, grafted to PEI via a bifunctional PEG reactive for the thiol moieties present in the complex surface.[111] Conversely, a water-soluble lipopolymer, which combines low-MW PEI (PEI1800) and cholesterol, increased transfection efficiency approximately eightfold in mouse neuronal progenitor cells[112] by favoring its integration into low-density lipoprotein, thus allowing the entry of vectors inside cells via low-density lipoprotein receptor-mediated endocytosis.[113]

Transfection of Neurons in vivo

Direct intracranial brain injection of DNA/PEI complexes produces transgene expression levels comparable to those obtained with the lenti- or adeno-viral vectors in the newborn and adult mouse brain, and in the newborn and adult rat brain.[92–94,114] However, intravenous administration of transfection vectors highly reduces transfection efficiency due to particle aggregation, unspecific interactions of these particles with blood components, such as plasma proteins, uptake by the reticuloendothelial system or opsonization by the innate immune system. In addition, an excess of positive charge can strongly activate the complement system.[115]

To increase transfection, the coupling of hydrophilic polymers, such as PEG, to DNA/PEI complexes has been investigated. Covalent coupling of PEG (PEGylation) resulted in spherical particles with a diameter of 40 nm, indicating that PEGylation protects from particle aggregation[116] in agreement with data previously reported for liposomes.[117] It must be taken into account that, for in vivo use, highly concentrated solutions of DNA complexes are usually required, resulting in formation of aggregates with concentrations of DNA above 50 µg/ml.[118] PEGylation not only prevented complexes aggregating, but also prevented undesirable effects reported for non-PEGylated complexes, including erythrocyte aggregation and binding of complexes to plasma proteins, such as IgM, fibronectin and complement C3.[117] Injecting the PEGylated and non-PEGylated complexes into the tail vein of mice resulted in a rapid clearance of non-PEGylated complexes from blood, whereas PEGylated complexes showed prolonged circulation times in the bloodstream. Besides, when non-PEGylated complexes were injected, a high acute toxicity was observed (50% of the animals died within 30 min after injection, with signs of acute lung embolism), whereas PEGylated complexes were well tolerated and did not accumulate in lungs.[117]

Another strategy developed to increase complex solubility in vivo consisted of coupling PEIs with a cationic carrier, such as polyethylene oxide or Pluronic 123 (P123). However, intravenous administration of this compound into the tail vein of C57Bl/6 mice resulted in luciferase expression in liver, heart, spleen and lung, whilst no activity was detected in the brain.[119] In fact, despite the efforts aimed to increase complex solubility, none of the data reported described an efficient PEI-based transfection vector able to cross the BBB when administered intravenously.

However, lumbar intratechal injection of PEGgylated PEI can deliver the gene to regions distant from the injection site.[120] Cationic copolymers were synthesized by grafting PEG to the amino groups of 2 and 25 kDa branched PEIs and coupled to DNA in 5% glucose. PEG–PEI/DNA complexes obtained were relatively uniform and smaller than 200 nm.[116] Lumbar intratechal injection of PEG–PEI/DNA complexes improved transfection efficiency by threefold compared with that mediated by PEI. Transgene expression mediated by the PEGylated complexes was detected in lumbar, thoracic and spinal cord, but also in brainstem, cerebellum, cerebral cortex and basal nuclei/diencephalons, while non-PEGylated PEI did not show transgene expression in these areas.[120] This method provides a viable strategy for CNS transfection, but it is still too traumatic to be used in human therapy.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.