Toxicogenomics of non-viral vectors for gene therapy: a microarray study of lipofectin- and oligofectamine-induced gene expression changes in human epithelial cells

Abstract

Of the non-viral vectors, cationic lipid (CL) formulations are the most widely studied for the delivery of genes, antisense oligonucleotides and gene silencing nucleic acids such as small interfering RNAs. However, little is known about the impact of these delivery systems on global gene expression in target cells. In an attempt to study the geno-compatibility of CL formulations in target cells, we have used microarrays to examine the effect of Lipofectin and Oligofectamine on the gene expression profiles of human A431 epithelial cells. Using the manufacturer's recommended CL concentrations routinely used for gene delivery, cDNA microarray expression profiling revealed marked changes in the expression of several genes for both Lipofectin- and Oligofectamine-treated cells. Data from the 200 spot arrays housing 160 different genes indicated that Lipofectin or Oligofectamine treatment of A431 cells resulted in more than 2-fold altered expression of 10 and 27 genes, respectively. The downstream functional consequences of CL-induced gene expression alterations led to an increased tendency of cells to enter early apoptosis as assessed by annexin V-FITC flow cytometry analyses. This effect was greater for Oligofectamine than Lipofectin. Observed gene expression changes were not sufficient to induce any significant DNA damage as assessed by single cell gel electrophoresis (COMET) assay. These data highlight the fact that inadvertent gene expression changes can be induced by the delivery formulation alone and that these may, ultimately, have important safety implications for the use of these non-viral vectors in gene-based therapies. Also, the induced non-target gene changes should be taken into consideration in gene therapy or gene silencing experiments using CL formulations where they may potentially mask or interfere with the desired genotype and/or phenotype end-points.