Gene therapy has the potential to treat many diseases, and adeno-associated virus (AAV) vectors in particular have shown great promise for gene delivery. AAV is a parvovirus with a single-stranded, linear DNA genome of 4.7 kb, and vectors can package up to ˜4.5 kb of foreign DNA between the viral terminal inverted repeats. AAV can function as a gene targeting vector by introducing modifications in endogenous genetic loci at their normal chromosomal locations by homologous recombination. However, these vectors can integrate randomly, with the potential of insertional mutagenesis or proto-oncogene activation, complicating the recovery of gene-targeted cells. In theory, the frequency of random integrations can be reduced by adding a negative selection marker in the form of a short-interfering RNA (siRNA) cassette to the gene targeting vector, just outside the vector homology arms utilized in homologous recombination. The siRNA cassette, then, should only be included in randomly integrated proviruses. To this end, six siRNA vectors were constructed, three of which contained unique RNA-interference (RNAi) sequences to target the nuclear lamin gene LMNB1, and three of which contained unique RNAi sequences to target the cell cycle gene CDK2. The efficacy of the RNAi constructs was tested by transfecting 293D and HeLa cells, and preliminary data suggests that these novel DNA constructs have the potential to reduce the frequency of AAV vector random integration in gene targeting. Given that AAV vectors are already being used in several clinical gene-therapy trials, improving the specificity of gene targeting is an important area of investigation.