Poly(amidoamine)s as potential endosomolytic polymers: evaluation in vitro and body distribution in normal and...

Fusogenic peptides derived from viral coat proteins cause perturbation of the endosomal membrane and are often used to improve the transfection efficiency of non-viral vectors in vitro. However, fusogenic peptides have limited potential for use in vivo due to their inherent immunogenicity. Totally synthetic polymers that are endosomolytic should circumvent this problem and could be useful as components of non-viral delivery systems as long as they do not immediately localise in the liver after intravenous (i.v.) injection. Linear poly(amidoamine) polymers (PAAs) having amido- and tertiary amino-groups along the main polymer undergo pH-dependent conformational change and thus provide an ideal opportunity for design of polymers that display membrane activity at low pH. Here we describe four PAAs, ISA 1 (Mn = 6900 Da) and ISA 23 (Mn = 10,500 Da) and their analogues ISA 4 and ISA 22 (Mn approximately 8000 Da) containing approximately 1 mol% 2-p-hydroxyphenyl ethylamine to allow radioiodination and thus monitoring of their biodistribution. In vitro cytotoxicity was assessed by MTT assay after incubation of PAAs with B16F10 and Mewo cell lines. The IC50 values observed for all PAAs were > 2 mg/mL in comparison with poly(L-lysine) which displayed an IC50 in the range 0.01-0.1 mg/mL. At pH 7.4 none of the PAAs studied was haemolytic at 1 h at concentrations below 3 mg/mL. PAAs were subsequently incubated with rat red blood cells for 24 h (1 mg/mL) at different pHs. In contrast to poly(L-lysine) which was haemolytic at pH 7.4, 6.5 and 5.5, none of the PAAs was lytic at pH 7.4, but they became membrane active at lower pH (approximately 45% for ISA 4, 50% for ISA 22 and 90% for ISA 23). These observations were substantiated by SEM and confirm the pH-dependence of membrane activity. After i.v. injection to rats 125I-labelled ISA 4 was immediately taken up by the liver (> 80% recovered dose at 1 h) whereas 125I-labelled ISA 22 was not (liver uptake was < 10% recovered dose at 5 h). Furthermore, biodistribution studies in mice bearing subcutaneous B16F10 melanoma showed that 125I-labelled ISA 22 was still accumulating in tumour tissue after 5 h (2.5% dose/g). PAAs have potential as endosomolytic agents and quantitation of the endosome to cytoplasm transfer is warranted after i.v. administration.