Inhibition of Gene Expression Inside Cells by Peptide Nucleic Acids: Effect of mRNA Target Sequence, Mismatched Bases, and PNA Length Donald F. Doyle, ‡,§ Dwaine A. Braasch, ‡ Carla G. Simmons, Bethany A. Janowski, and David R. Corey* Departments of Pharmacology and Biochemistry, UniVersity of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines BouleVard, Dallas, Texas, 75390-9041 ReceiVed August 31, 2000; ReVised Manuscript ReceiVed October 20, 2000 ABSTRACT: Genome sequencing has revealed thousands of novel genes, placing renewed emphasis on chemical approaches for controlling gene expression. Antisense oligomers designed directly from the information generated by sequencing are one option for achieving this control. Here we explore the rules governing the inhibition of gene expression by peptide nucleic acids (PNAs) inside cells. PNAs are a DNA/RNA mimic in which the phosphate deoxyribose backbone has been replaced by uncharged linkages. Binding to complementary sequences is not hindered by electrostatic repulsion and is characterized by high rates of association and elevated affinities. Here we test the hypothesis that the favorable properties of PNAs offer advantages for recognition of mRNA and antisense inhibition of gene expression in vivo. We have targeted 27 PNAs to 18 different sites throughout the 5′-untranslated region (5′-UTR), start site, and coding regions of luciferase mRNA. PNAs were introduced into living cells in culture as PNA- DNA-lipid complexes, providing a convenient high throughput method for cellular delivery. We find that PNAs targeted to the terminus of the 5′-UTR are potent and sequence-specific antisense agents. PNAs fifteen to eighteen bases in length were optimal inhibitors. The introduction of one or two mismatches abolished inhibition, and complementary PNAs targeted to the sense strand were also inactive. In striking contrast to effective inhibition by PNAs directed to the terminal region, PNAs complementary to other sites within the 5′-UTR do not inhibit gene expression. We also observe no inhibition by PNAs complementary to the start site or rest of the coding region, nor do we detect inhibition by PNAs that are highly C/G rich and possess extremely high affinities for their target sequences. Our results suggest that PNAs can block binding of the translation machinery but are less able to block the progress of the ribosome along mRNA. The high specificity of antisense inhibition by PNAs emphasizes both the promise and the challenges for PNAs as antisense agents and provides general guidelines for using PNAs to probe the molecular recognition of biological targets inside cells. Genome sequencing has revealed the identities of many genes that encode proteins whose functions are unknown. One strategy for understanding the cellular roles of these proteins is to inhibit gene expression using synthetic compounds. Oligonucleotides and oligonucleotide mimics offer important advantages as synthetic tools for selective gene inhibition because knowledge of the target gene sequence provides sufficient information for inhibitor design (1). Synthesis of oligomers can be performed by straight- forward protocols, and many chemical options are available to tailor the oligomer properties for selected applications. Once synthesized, the effects of fully complementary and mismatch-containing oligomers can be compared, providing an important control to confirm that observed effects are due to inhibition of target gene expression rather than to unintended interactions with other cellular components. Oligonucleotides also represent an exquisitely specific tool for investigating nucleic acid structure and function inside cells under physiologically relevant conditions. Using them in this capacity, however, will require a basic understanding of the rules governing their recognition of targets inside cells and how these rules vary depending on the chemical characteristics of the type of oligomer being used. Despite their inherent advantages, the control gene expres- sion by oligonucleotides is often confounded by nonspecific interactions and by an inability to readily identify susceptible target sites (2). These problems have led to the development of sophisticated chemical modifications and screening meth- odologies. Modified DNA and RNA oligonucleotides have been shown to effectively inhibit gene expression in cell culture, and high throughput screening methods now allow the rapid identification of inhibitory oligonucleotides (3). These advances have led to successful treatment of disease, with one phosphorothioate-modified DNA oligomer ap- proved to treat cytomegalovirus (CMV) retinitis (4), and other DNA and DNA-2′-O-alkyl RNA chimeric oligomers being tested in clinical trials (5). Thus it is clear that oligonucleotides can be used to sequence-selectively block gene expression within cells. However, oligomers that † This work was supported by grants from the National Institutes of Health R01GM 60624, the Robert A. Welch Foundation (I-1244), and the Texas Advanced Technology Program (18603). * To whom correspondence should be addressed. E-mail: corey@ chop.swmed.edu. Phone: (214) 648-5096. Fax: (214) 648-5095. ‡ Both authors contributed equally. § Current address: School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, GA 30332-0400. 53 Biochemistry 2001, 40, 53-64 10.1021/bi0020630 CCC: $20.00 © 2001 American Chemical Society Published on Web 12/13/2000