The Effect of pDNA Quality on Gene Transfer Outcome In Vivo.

Bazzani RP, Pringle IA, Schleef M, Schmeer M, Cheng SH, Scheule RK, Hyde SC, Gill DR

Molecular Therapy, 17 S1 S20


The American Society of Gene Therapy Annual Conference, San Diego, 2009

A potential problem of gene therapy is the induction of a host immune response after delivery of gene transfer agents. Our work investigates non-viral vectors, including plasmid DNA (pDNA) complexed with Genzyme lipid GL67 (GL67A) for lung delivery.

Previously it has been shown that unmethylated CpG motifs present in bacterially derived pDNA have an immunostimulatory effect in mammalian cells. Hyde et al. (2008 Nature Biotechnology 26:549) showed that the use of a CpG-free pDNA expression vector resulted in reduced inflammation; and CpG-free plasmids are currently being tested clinically in the lungs of individuals with cystic fibrosis. In mouse lung models we have occasionally been able to detect low-level, residual inflammation even when CpG-free pDNA was used. We hypothesised that residual bacterial chromosomal DNA (chrDNA) present in pDNA samples could be acting as a source of unmethylated CpG motifs, resulting in low-level inflammation in vivo.

We developed a TaqMan PCR assay to quantify possible chrDNA contamination present in pDNA samples used for in vivo studies. Plasmid produced by different manufacturers and different production methods, were analysed for: (i) chrDNA content by TaqMan PCR, (ii) plasmid forms by capillary gel electrophoresis, (iii) in vivo transgene expression by TaqMan, and (iv) levels of induced cytokines by Bio Plex (Multi plex cytokine assay, Bio-Rad, Hemel Hempstead, UK). Analysis of various pDNA samples showed that although levels of plasmid forms (supercoiled monomer, supercoiled dimer, and open circular) were similar, the amount of residual chrDNA varied considerably between different suppliers and production processes. To evaluate the pDNA samples in vivo we used intranasal instillation of GL67A/pDNA complexes (80µg pDNA/100µl) into the lungs of female BALB/c mice. After 24h bronchoalveolar lavage fluid was analysed for inflammation markers and transgene expression was measured in whole lung homogenates. Samples of pDNA with the highest chrDNA contamination (>0.9% chrDNA of total DNA) showed a significant increase in CpG-related cytokine levels compared with samples with low (<0.04%) chrDNA contamination (p<0.01, ANOVA/PLSD).

Furthermore, pDNA samples with low chrDNA contamination led to higher levels of transgene expression (p<0.05, ANOVA/PLSD). These results show that minimising residual bacterial chrDNA contamination in pDNA preparations can be a key factor in reducing inflammation and maximising transgene expression in vivo.