Gene Therapy (2002) 9, 686–690  2002 Nature Publishing Group All rights reserved 0969-7128/02 $25.00 www.nature.com/gt Modification of hepatic genomic DNA using RNA/DNA oligonucleotides BT Kren, Z Chen 1 , R Felsheim 1,4 , N Roy Chowdhury 3 , J Roy Chowdhury 3 and CJ Steer 1,2 1 Department of Medicine, University of Minnesota, Minneapolis, MN, USA; 2 Department of Genetics, Cell Biology and Development, University of Minnesota, Minneapolis, MN, USA; and 3 Marion Bessin Liver Research Center, Albert Einstein College of Medicine, Bronx, NY, USA The ideal gene therapy is one that repairs the precise gen- etic defect without additional modification of the genome. Such a strategy has been developed for correcting single nucleotide mutations by using RNA/DNA oligonucleotides, or chimeraplasts. This approach for in situ repair is based on the delivery of exogenous DNA designed to mediate gen- omic base conversion, insertion, or deletion, thereby, cor- recting the genetic mutation. Using in vivo delivery systems to hepatocytes via the asialoglycoprotein receptor, we tar- geted rat liver DNA and successfully modified the genomic sequence by chimeraplasty. The changes in both the hepatic genes, and their associated phenotypes remained stable for 2 years. In addition, we also examined the potential to alter sequence defects in mitochondrial DNA. Therefore, we Keywords: RNA/DNA oligonucleotide; mitochondrial; gene therapy; mismatch repair; site-directed conversion; single-point mutation Inherited metabolic diseases impose a significant world- wide burden both financially and socially. Although there have been advances in treating these disorders, alternative therapeutic approaches are being developed to correct the responsible genetic defects. Gene aug- mentation using viral vectors and transgene expression to replace an absent or defective protein has shown lim- ited success in treating a variety of genetic disorders. In fact, the mechanisms by which cells and organisms pro- tect against viral infections have presented significant barriers for long-term stable expression of the introduced transgene. In part, due to the disadvantages of using bio- logical vectors, a variety of different approaches for non- viral gene delivery have been developed. 1 In particular, targeting of nonviral delivery systems to specific tissues by receptor/ligand interactions has been exploited for in vivo gene transfer. 2 By targeting the asialoglycoprotein receptor (ASGPR), gene delivery to hepatocytes has been significantly improved both in vitro 3 and in vivo. 4,5 Thus, we have focused on using nonviral approaches to correct the genetic defect in hepatocytes using galactosylated Correspondence: BT Kren, Department of Medicine, Mayo Mail Code 36, 420 Delaware Street SE, Minneapolis, Minnesota 55455, USA 4 Present address: Biological Process Technology Institute, University of Minnesota, Saint Paul, MN 55108, USA determined whether mitochondria possess the enzymatic machinery for chimeraplast-mediated DNA changes. Using an in vitro DNA repair assay of mutagenized plasmids and an Escherichia coli readout system, we showed that extracts from highly purified rat liver mitochondria have the essential enzymatic activity to mediate precise single-nucleotide changes at a frequency similar to liver nuclear extracts. Moreover, single-stranded oligonucleotides carrying a single nucleotide mismatch with the target sequence were capable of promoting gene conversion using either mitochondrial or nuclear extracts. Several approaches now exist for the pre- cise repair of genetic mutations using either single-stranded or RNA/DNA chimeric oligonucleotides. Gene Therapy (2002) 9, 686–690. DOI: 10.1038/sj/gt/3301762 compounds attached directly to polycations, or incorpor- ated in liposomes. 6 The strategy of in situ genomic repair using chimeric RNA/DNA molecules (chimeraplasts) developed from studies investigating molecular aspects of DNA repair. It became apparent that there was a significant increase in pairing efficiency between an oligonucleotide 50 bases and a genomic DNA target if RNA replaced DNA in a portion of the targeting oligonucleotide. 7,8 The original chimeric 68-mer design incorporated 10 2’-O-methylated RNA residues flanking each side of a 5 bp stretch of DNA, poly-T hairpin loops, a 3’ GC clamp and a comp- lementary all-DNA strand resulting in a stable, nuclease- resistant duplex molecule. The 25 bp homology segment between the chimeraplast and its genomic target is con- structed such that it includes one mismatch with the DNA sequence of the gene. This engineered mismatch permits the directed alteration of the genomic target using the chimeraplast as a template for the intended cor- rection of target sequence. It is postulated that the ‘mis- matched’ chimeraplast complexes with the genomic DNA and creates the illusion of a genetic mutation leading to the recruitment of endogenous repair functions. 9,10 This novel approach of gene repair, based on in vitro experimental evidence, was shown by Yoon et al 11 and Cole-Strauss et al 12 to be capable of introducing targeted single base conversions in episomal and genomic DNA in cultured cells. We then reported that the chimeraplast