Synthesis and structural characterization of stable branched DNA g-quadruplexes using the trebler phosphoramidite

Abstract

Guanine (G)-rich sequences can form a noncanonical four-stranded structure known as the G-quadruplex. G-quadruplex structures are interesting because of their potential biological properties and use in nanosciences. Here, we describe a method to prepare highly stable G-quadruplexes by linking four G-rich DNA strands to form a monomolecular G-quadruplex. In this method, one strand is synthesized first, and then a trebler molecule is added to simultaneously assemble the remaining three strands. This approach allows the introduction of specific modifications in only one of the strands. As a proof of concept, we prepared a quadruplex where one of the chains includes a change in polarity. A hybrid quadruplex is observed in ammonium acetate solutions, whereas in the presence of sodium or potassium, a parallel G-quadruplex structure is formed. In addition to the expected monomolecular quadruplexes, we observed the presence of dimeric G-quadruplex structures. We also applied the method to prepare G-quadruplexes containing a single 8-aminoguanine substitution and found that this single base stabilizes the G-quadruplex structure when located at an internal position.

Keywords: 8-aminoguanines; DNA structures; G-quadruplexs; branched oligonucleotides; oligonucleotides.

Scheme 1

Outline of the method for the preparation of G-quadruplexes proposed in this study (see Table 2 in the Experimental Section for oligonucleotide sequences). *G denotes the position of 8-aminoguanine residues. For standard DNA synthesis, a cycle for each nucleotide addition consists of the following steps: 1) 3 % trichloroacetic acid/dichloromethane; 2) 5′-DMT-nucleoside-3′-phosphoramidite, tetrazole; 3) capping with acetic acid and N-methylimidazole; 4) oxidation with 0.01 m iodine solution. The same steps are applied in the reversed DNA synthesis but with the use of 3′-DMT-nucleoside-5′-phosphoramidite as monomers.

Figure 1

CD spectra of oligonucleotide 1 (left), 2 (middle) and 3 (right) dissolved in water (—), 5 mM KCl (—), 100 mM NaCl (—) and 100 mM NH₄OAc (—).

Figure 2

Left: ESI-MS of a quadruplex formed by oligonucleotides 1 (a), 2 (b) and 3 (c) and the distribution of the number of NH₄⁺ ions preserved in the G-quadruplex at −6 charge state; the mass spectra were smoothed using a mean function, 2*10 channels, using MassLynx 4.0. Center: 2D ESI-MS and drift time distribution of oligonucleotide 1. Right: ESI-MS of a dimer formed by oligonucleotides 1 (d), 2 (e) and 3 (f); the mass spectra were smoothed using a mean function, 2*30 channels, using MassLynx 4.0.

Figure 3

Top: General scheme of a tetra-end-linked quadruplex showing the numeration of the residues as mentioned in the text. Bottom: exchangeable proton region of ¹H NMR spectra at different temperatures of oligonucleotides 1, 2 and 3 in 10 mM sodium phosphate buffer (upper rows) or 10 mM potassium phosphate buffer (lower rows).

Figure 4

Fragments of NOESY spectra (250 ms mixing time) of oligonucleotide 1 at 5 °C (bottom) and 45 °C (top) in 10 mM potassium phosphate (pH 7).

Figure 5

Hypothetical dimerization of the quadruplex-forming structures.