We have examined the abilities of three complementary γ-peptide nucleic acid (γPNA) oligomers to invade an RNA G-quadruplex and potently inhibit translation of a luciferase reporter transcript containing the quadruplex-forming sequence (QFS) within its 5'-untranslated region. All three γPNA oligomers bind with low nanomolar affinities to an RNA oligonucleotide containing the QFS. However, while all probes inhibit translation with low to midnanomolar IC50 values, the γPNA designed to hybridize to the first two G-tracts of the QFS and adjacent 5'-overhanging nucleotides was 5-6 times more potent than probes directed to either the 3'-end or internal regions of the target at 37 °C. This position-dependent effect was eliminated after the probes and target were preincubated at an elevated temperature prior to translation, demonstrating that kinetic effects exert significant control over quadruplex invasion and translation inhibition. We also found that antisense γPNAs exhibited similarly potent effects against luciferase reporter transcripts bearing QFS motifs having G2, G3, or G4 tracts. Finally, our results indicate that γPNA oligomers exhibit selectivity and/or potency higher than those of other antisense molecules such as standard PNA and 2'-OMe RNA previously reported to target G-quadruplexes in RNA.

No abstract

High affinity nucleic acid analogues such as gammaPNA (γPNA) are capable of invading stable secondary and tertiary structures in DNA and RNA targets but are susceptible to off-target binding to mismatch-containing sequences. We introduced a hairpin secondary structure into a γPNA oligomer to enhance hybridization selectivity compared with a hairpin-free analogue. The hairpin structure features a five base PNA mask that covers the proximal five bases of the γPNA probe, leaving an additional five γPNA bases available as a toehold for target hybridization. Surface plasmon resonance experiments demonstrated that the hairpin probe exhibited slower on-rates and faster off-rates (i.e., lower affinity) compared with the linear probe but improved single mismatch discrimination by up to a factor of five, due primarily to slower on-rates for mismatch vs. perfect match targets. The ability to discriminate against single mismatches was also determined in a cell-free mRNA translation assay using a luciferase reporter gene, where the hairpin probe was two-fold more selective than the linear probe. These results validate the hairpin design and present a generalizable approach to improving hybridization selectivity.