Short LNA-modified oligonucleotide probes as efficient disruptors of DNA G-quadruplexes

Abstract

G-quadruplexes (G4s) are well known non-canonical DNA secondary structures that can form in human cells. Most of the tools available to investigate G4-biology rely on small molecule ligands that stabilise these structures. However, the development of probes that disrupt G4s is equally important to study their biology. In this study, we investigated the disruption of G4s using Locked Nucleic Acids (LNA) as invader probes. We demonstrated that strategic positioning of LNA-modifications within short oligonucleotides (10 nts.) can significantly accelerate the rate of G4-disruption. Single-molecule experiments revealed that short LNA-probes can promote disruption of G4s with mechanical stability sufficient to stall polymerases. We corroborated this using a single-step extension assay, revealing that short LNA-probes can relieve replication dependent polymerase-stalling at G4 sites. We further demonstrated the potential of such LNA-based probes to study G4-biology in cells. By using a dual-luciferase assay, we found that short LNA probes can enhance the expression of c-KIT to levels similar to those observed when the c-KIT promoter is mutated to prevent the formation of the c-KIT1 G4. Collectively, our data suggest a potential use of rationally designed LNA-modified oligonucleotides as an accessible chemical-biology tool for disrupting individual G4s and interrogating their biological functions in cells.

Figure 1.

Illustration of a G4-structure and its potential disruption by duplex formation with its reverse complementary strand. (A) Schematic illustration of a G-tetrad, which is the core of a G4-structure. Four guanines interact in a tetrameric planar structure held together by Hoogsteen H-bonding, which can be further stabilised by alkali cations (M⁺), such as K⁺. (B) Stacking of three G-tetrads leads to the formation of a G4-structure, which can adopt either parallel or antiparallel topologies depending on the relative orientation of the four DNA strands in the G4-structure, as indicated by the arrows in the scheme. (C) G4-are formed from single-stranded DNA sequences. In the presence of their C-rich reverse complement, G4-folded sequences can be invaded to form a duplex DNA, which can be leveraged to disrupt the G4. (D) Chemical structures of DNA and LNA nucleosides. In LNA, an additional methylene (i.e. -CH₂-) bridge shown in red ‘locks’ the nucleoside sugar conformation, significantly improving Watson–Crick H-bonding efficiency and DNA duplex stability.

Figure 2.

(A) Schematic depicting the principles of the FRET G4-disruption assay. c-KIT1 G4 dually labelled with FAM (donor) and TAMRA (acceptor) fluorophores is disrupted in presence of DNA or LNA-modified reverse complementary strand. When the donor fluorophore (FAM) is excited, there is a Föster energy transfer to the acceptor fluorophore (TAMRA), which is sensitive to the distance between the donor and the acceptor. As the G4 gets disrupted to form a duplex with the LNA/DNA probes, the distance between the donor and the acceptor increases, resulting in a concomitant decrease in the energy transfer, which increases donor-emission. (B) Half-life of c-KIT1 obtained using the FRET G4-disruption assay when treated with unmodified DNA or KIT_LNA1—4 probes. (C) Disruption half-lives of hTelo G4 when treated with full-length DNA (hTelo_DNA1) or LNA (hTelo_LNA1 or hTelo_LNA2). Experiments were performed in three independent replicates with SEM range plotted. *, ** and *** represent statistically significant results at P< 0.05, P< 0.01 and P< 0.001, respectively. ns represents differences that are not statistically significant (P> 0.05).

Figure 3.

(A) Bar plot depicting c-KIT1 G4 half-life measured with short LNA/DNA probes. Treatment of c-KIT1 G4 with KIT_LNA_short resulted in a significant reduction in the G4-disruption half-life compared to the full-length analogue KIT_LNA2, whereas treatment with KIT_DNA_short showed no evidence of G4-disruption (no change in FRET emission over time). (B) Treatment of hTelo G4 with the short probe hTelo_LNA_short resulted in half-lives comparable to those obtained with the full-length LNA probe (hTelo_LNA1). The short DNA probe hTelo_DNA_short significantly increased half-lives. Experiments were performed in three independent replicates and plotted with corresponding SEM intervals. **** and *** represent statistical significance at P< 0.0001 and P< 0.001. ns represents statistically non-significant difference (P> 0.05).

Figure 4.

(A) Schematic of disrupting an ssDNA that contains multiple units of telomeric G4s without (top) and with (bottom) G4 disrupting DNA or LNA probes. (B) Force vs extension curves (left, solid and dotted traces depict stretching and relaxing processes, respectively) and ΔL versus force curves (right) for the same molecule in a 10 mM Tris buffer (pH 7.4) supplemented with 100 mM KCl. Blue, buffer only. Pink and purple represent 100 nM hTelo_DNA_short and hTelo_LNA_short, respectively. Arrows indicate F_1/2, where 50% GQ structures were mechanically disrupted. Bar diagrams of (C) total number of disrupted GQ and (D) F_1/2 values in different solutions. Blue: buffer only; Pink or Purple depict the same buffer added with 100 nM hTelo_DNA_short or hTelo_LNA_short, respectively. Bar charts represent mean values from at least N= 5 independent measurements. Error bars represent standard error of mean (SEM). *, ** or *** represent statistical significance at P< 0.05, P< 0.01 and P< 0.001, respectively.

Figure 5.

(A) Schematic representation of the template and primer used for the single-extension assay. (B) Urea-PAGE showing distribution of the full-length and the stalled products when extending the template in the presence or absence of either KIT_LNA_short or KIT_DNA_short under different conditions. (C) Quantified intensity ratio of the full-length vs the stalled product band. Higher ratio suggests greater extension efficiency and lower polymerase stalling. Error bars represent standard error of mean (SEM). Means were calculated from N= 3 independent experiments. Asterisk * represents statistical significance (P< 0.05) whereas ns represents results that are not statistically significant (P> 0.05).

Figure 6.

(A) The pc-KIT plasmid used in the dual luciferase reporter assay with c-KIT promoter region controlling Renilla expression. (B) Transfection of the plasmid alone or co-transfection with DNA or LNA probes in HEK293T revealed significant differences in the ability of the probes to perturb relative gene-expressions, as assessed by the ratio of the Renilla/Firefly expression levels under the different conditions tested. Error bars represent Standard error of Mean (SEM). Means were computed from N= 3 independent biological experiments. Asterisks * and ** represent statistical significance at P< 0.05 and P< 0.01, respectively, whereas ns represents results that are not statistically significant (P> 0.05).