DNA–RNA Hybrid Quadruplexes Reveal Interactions that Favor RNA Parallel Topologies

Abstract

A DNA G-quadruplex adopting a (3+1)-hybrid structure was substituted at its 5'-tetrad by riboguanosine (rG) analogs. Incorporation of anti-favoring rG at appropriate syn-positions of the 5'-outer tetrad induced conformational rearrangements to yield a quadruplex featuring a 5'-tetrad with reversed polarity. A high-resolution structure of a disubstituted quadruplex variant as well as direct NMR experimental evidence reveals a non-conventional C-H⋅⋅⋅O hydrogen bond in a medium groove between the 2'-OH of an rG residue adopting a C2'-endo sugar pucker and H8 of a 3'-neighboring anti-G residue. In contrast, a C3'-endo sugar conformation for another guanine ribonucleotide prevents formation of a corresponding hydrogen bond but relocates its 2'-OH substituent from the quadruplex narrow groove into a medium groove. Both the formation of favorable CHO hydrogen bridges and unfavorable interactions of the 2'-hydroxyl group in a narrow groove will promote RNA folding into a parallel topology featuring all-anti core residues and four grooves of medium size.

Manipulating DNA G‐Quadruplex Structures by Using Guanosine Analogues

“…Recently, the ODN quadruplex has been subjected to further modifications also employing F araG and rG as anti ‐favoring surrogates for incorporation into unmatching syn positions of its 5′‐tetrad. In analogy to F rG substitutions, a tetrad polarity flip was promoted in both cases . Noticeably, stabilities of F araG‐modified quadruplexes were hardly compromised, in contrast to F rG modifications with their destabilizing effect depending on number and position of substitution sites .…”

“…Although unconventional hydrogen bonds appear to play a crucial role in determining conformational equilibria, it is difficult to predict their formation and impact on the stability of specific G4 topologies reliably. Thus, the expected participation of fluorine in a sequential hydrogen bond to H8 was not observed when F rG was incorporated into the 5′‐tetrad of the ODN quadruplex, yet an O2′ ⋅⋅⋅ H8−C8 interaction was found for corresponding rG modifications (Figure B) . Moreover, whereas the anti F araG residue showed an intraresidue hydrogen bond when incorporated into central or outer positions of the ODN G4, a sequential F2′ ⋅⋅⋅ H8−C8 pseudo‐hydrogen bond was proposed on the basis of a high‐resolution NMR structure of an F araG‐modified two‐repeat telomeric quadruplex with a parallel fold (Figure C) .…”

“…Remarkably, the coexistence of two stable quadruplex structures with either an anti or a syn conformation for 2′‐F‐modified G incorporated into the central tetrad of the ODN G4 illustrates that both F rG and F araG might also assume the disfavored syn glycosidic torsion angle in a conserved topology. Several examples suggest that the stability of different conformations both for 2′‐F‐ and for 2′‐OH‐substituted G analogues, which might possibly determine distinct global conformational rearrangements, seems to be partially linked to the formation of stabilizing unconventional hydrogen bonds between the 2′‐substituent in an S ‐puckered ribose unit as hydrogen bond acceptor and guanine C8−H8 (Figure ) . Strikingly, in the two coexisting ODN topologies with a single modification in the central tetrad (Scheme ), both syn F rG and syn F araG in the native G4, but also anti F rG in the rearranged G4, participate in a sequential F2′ ⋅⋅⋅ H8−C8 hydrogen bond to the 3′‐adjacent guanine base .…”

“…Not only backbone but also groove interactions seem to affect the conformations of introduced G analogues with their influence on quadruplex transitions. Structural data suggest that a quadruplex narrow groove does not readily host sugar or base substituents . For instance, F rG and rG, when incorporated into the ODN or HT 5′‐position of a G‐tract bordering a narrow and medium groove, were each found to adopt a C3′‐ endo sugar pucker.…”

“…It is this conformation that orients the 2′‐fluoro or 2′‐hydroxy substituent away from the narrow groove and into the medium groove, likely avoiding unfavorable interactions within the narrow groove (Figure A). In contrast, G analogues of other G‐tracts are S‐puckered with the 2′‐substituent turned upwards, allowing for sequential pseudo‐hydrogen bonds within a medium groove (Figure B) …”