Selective recognition of c-MYC Pu22 G-quadruplex by a fluorescent probe

Abstract

Nucleic acid mimics of fluorescent proteins can be valuable tools to locate and image functional biomolecules in cells. Stacking between the internal G-quartet, formed in the mimics, and the exogenous fluorophore probes constitutes the basis for fluorescence emission. The precision of recognition depends upon probes selectively targeting the specific G-quadruplex in the mimics. However, the design of probes recognizing a G-quadruplex with high selectivity in vitro and in vivo remains a challenge. Through structure-based screening and optimization, we identified a light-up fluorescent probe, 9CI that selectively recognizes c-MYC Pu22 G-quadruplex both in vitro and ex vivo. Upon binding, the biocompatible probe emits both blue and green fluorescence with the excitation at 405 nm. With 9CI and c-MYC Pu22 G-quadruplex complex as the fluorescent response core, a DNA mimic of fluorescent proteins was constructed, which succeeded in locating a functional aptamer on the cellular periphery. The recognition mechanism analysis suggested the high selectivity and strong fluorescence response was attributed to the entire recognition process consisting of the kinetic match, dynamic interaction, and the final stacking. This study implies both the single stacking state and the dynamic recognition process are crucial for designing fluorescent probes or ligands with high selectivity for a specific G-quadruplex structure.

Scheme 1.

Structure of 6-(2-(anthracen-9-ylmethylene) hydrazinyl)-N²,N⁴-diphenyl-1,3,5-triazine-2,4-diamine (9CI) and its division into sub-structures. The triazine ring is labeled as A, and B represents the anthracene ring. C and D represent the phenyl rings. Nitrogen atoms connecting the triazine ring and the phenyl rings are labeled as 1 and 2, respectively.

Figure 1.

Structures of compounds 1–9 and their fluorescence response to eighteen DNA oligonucleotides. The excitation and emission wavelengths of these compounds are listed in Supplementary Table S2.

Figure 2.

(A) Cartoon representation of superimposed c-MYC Pu22 G-quadruplex structures in apo state (brown, model 1 in the NMR structure, PDB 1xav) and with quindoline (purple, model 1 in the NMR structure, PDB 2l7v). The two quindoline molecules are shown in cyan. (B) Superimposed structures without ligands. (C) Connolly surface of c-MYC Pu22 G-quadruplex structure from PDB 2l7v. The red spheres highlight the central platform on top of 3′ end G-quartet and at the bottom of 5′ end G-quartet, while green, yellow and blue spheres represent the locations of grooves connecting to the central platform, respectively.

Figure 3.

(A) Fluorescence emission spectra of 0.5 μM 9CI in the presence of various oligonucleotides (0.5 μM), λ_ex = 405 nm. Fluorescence emission of compound 9CI alone is shown in black. 0.5 μM oligonucleotide samples were in 10 mM K₂HPO₄/KH₂PO₄ buffer solution with 100 mM KCl at pH 7.0. (B) Fluorescence intensity enhancement of 0.5 μM 9CI against concentration of different DNA oligonucleotides. (C) Fluorescence intensity of 9Cl in the presence of c-MYC Pu18, c-MYC Pu22, c-MYC Pu24 and c-MYC Pu27. Both 9Cl and DNA are 0.5 μM. λ_ex = 405 nm, λ_em = 472 nm.

Figure 4.

Electrophoresis of DNA oligonucleotides on 20% non-denaturing acrylamide gel. (A) Gel stained by SYBR Gold after electrophoresis; (B) DNA samples were premixed with 20 μM 9CI before the electrophoresis. SYBR Gold used for staining was diluted ten thousand times from original stocking solution with TBE (Tris-borate-EDTA) buffer. 80 pmol DNAs were loaded into the different wells except for c-MYC Pu22 (160 pmol) in lane 3. Lane 1: DNA marker with 22 nucleotides; lane 2: c-MYC Pu22 (80 pmol); lane 3: c-MYC Pu22 (160 pmol); lane 4: TTA; lane 5: c-KIT1; lane 6: c-KIT2; lane 7: VEGF; lane 8: TBA; lane 9: CT4; lane 10: 19AT.

Figure 5.

Visualization of 9CI mixed with DNA oligonucleotides under UV light. (A) 5 μM 9CI and 5 μM DNA samples in 10 mM KH₂PO₄/K₂HPO₄ buffer with 100 mM KCl at pH 7.0. (B) Lane 1: 200 μM c-MYC Pu22 DNA; lane 2: 50 μM 9CI; lane 3: 25 μM c-MYC Pu22 DNA + 25 μM 9CI; lane 4: 50 μM c-MYC Pu22 DNA + 50 μM 9CI; lane 5: 100 μM c-MYC Pu22 DNA + 100 μM 9CI; lane 6: 200 μM c-MYC Pu22 DNA + 200 μM 9CI.

Figure 6.

Fluorescence images of A549 cells transfected with 40 pmol c-MYC Pu22 DNA G-quadruplex forming oligonucleotide labelled with 5′-Cy5 using lipofectamine 2000 and then incubated with 5 μM 9Cl for 2 hours successively. (A) Fluorescence signal collected between 655–755 nm at λ_ex = 635 nm, (B) fluorescence signal collected between 425–470 nm at λ_ex = 405 nm, (C) fluorescence signal collected between 500 and 545 nm at λ_ex = 405 nm, (D) merged image of (A) and (C), (E) merged images of (A), (B), (C) and the bright field. Scale bar, 20 μm.

Figure 7.

Schematic illustration of locating aptamers in cells by DNA mimic of fluorescent proteins with the 9CI–c-MYC Pu22 G-quadruplex complex as the fluorescent response core.

Figure 8.

Confocal microscope images of the treated A549 cells and HepG2 cells at 405 nm excitation. A549 cells was incubated with SL1-M for 30 min at room temperature, and washed by 1 ml DPBS twice. Then, the cells were incubated with 500 nM MFP (c-MYC Pu22-MFP/ TTA-MFP) and 5 μM 9CI for 10 min. Fluorescence was collected after the cells were washed with DPBS three times again. (A) Fluorescence collected at 425–470 nm emission channel, (B) fluorescence collected at 500–545 nm emission channel, (C) merged image of (A) and (B) and bright field. Scale bar, 10 μm.

Figure 9.

Fluorescence emission spectra of 0.5 μM 9CI with the addition of 0.5 μM c-MYC Pu22 DNA G-quadruplex oligonucleotide and the mutant oligonucleotides, λ_ex = 405 nm. Fluorescence emission of 9CI alone is shown in black, and 0.5 μM oligonucleotides were prepared in 10 mM K₂HPO₄/KH₂PO₄, 100 mM KCl, pH 7.0.

Figure 10.

Dynamic recognition between 9CI and c-MYC Pu22 G-quadruplex explored by molecular dynamic simulations.

Figure 11.

Schematic illustration for the induced fitting between 9CI and c-MYC Pu22 G-quadruplex.