Classical gene targeting employs natural homologous recombination for a gene correction using a specially designed and artificially delivered DNA construct but the method is very inefficient. On the other hand, small DNA fragments in the form of tiny chromatin-like particles naturally present in blood plasma can spontaneously penetrate into human cells and cell nuclei. We hypothesized that these natural DNA nanoparticles with recombinagenic free ends might be effective agents for gene replacement therapy. We demonstrate that a mixture of small fragments of total human chromatin from nonmutant cells added to a culture medium without transfection agents efficiently repaired a 47 base pair deletion in the CASP3 gene in 30% of treated human MCF7 breast cancer cells, as shown by restoration of caspase-3 apoptotic function and CASP3 DNA and mRNA structure. Such an innate gene replacement mechanism might function naturally in an organism using its own apoptotic DNA fragments. This mechanism might enable human cancer cell phenotype normalization in the presence of excess normal cells.Gene targeting (GT) modifies a gene in its chromosomal location, preserving existing mechanisms to regulate its function in cells, and thereby holds great promise as a medical treatment strategy. 1-3 GT employs endogenous homologous recombination (HR) and homologous sequences contained in targeting DNA constructs for modification of a target sequence in genomic DNA. The GT used in reverse genetics studies requires artificial transfection, sophisticated targeting vectors, and drug-based selection to overcome the extremely low efficiency of the method, making it impractical for in vivo or therapeutic use. 2 Earlier it was found that the GT yield increases up to 100-fold by increasing the homology length in a donor consruct above 10 kbp, but still remains insufficient for in vivo applications. 4 However, it was also shown that the homology length required for HR in mammals is only around 100-400 bp, 5-7 suggesting that small DNA constructs might be effective. Indeed, the surprisingly high 1-10% repair efficiency demonstrated by a GT method called small fragment homologous replacement (SFHR) using DNA constructs a few hundred bp in length 8 might indicate that the small size of those constructs is favorable for GT. Free DNA ends, resulting from cutting of either targeting plasmid or target DNA in the region of common homology, are highly recombinagenic, enhancing HR locally in both yeast 9 and mammalian 10,11 cells. Recently this approach was used for repairing a point mutation using simultaneous transfection of a plasmid containing a short homologous donor DNA (1.5 kbp) and a construct expressing artificial zinc-finger nucleases specifically inducing DSBs in proximity to the mutation, which led to 20% efficiency of the target repair. 12 Another GT strategy called "ends-out" or gene replacement allows repair of extended genetic mutations without generating potentially mutagenic DSBs by using recombinant linear DNA in which the two ends ...