Controlling electron rebound within four-base π-stacks in Z-DNA by changing the sugar moiety from deoxy- to ribonucleotide.

Charge transfer through DNA is of great interest because of the potential of DNA to be a building block for nanoelectronic sensors and devices. The photochemical reaction of 5-halouracil has been used for probing charge-transfer processes along DNA. We previously reported on unique charge transfer following photochemical reaction of 5-bromouracil within four-base π-stacks in Z-DNA. In this study, we incorporated a guanosine instead of a deoxyguanosine into Z-DNA, and found that electron transfer occurs in a different mechanism through four-base π-stacks.

Long-loop G-quadruplexes are misfolded population minorities with fast transition kinetics in human telomeric sequences.

Single-stranded guanine (G)-rich sequences at the 3' end of human telomeres provide ample opportunities for physiologically relevant structures, such as G-quadruplexes, to form and interconvert. Population equilibrium in this long sequence is expected to be intricate and beyond the resolution of ensemble-average techniques, such as circular dichroism, NMR, or X-ray crystallography. By combining a force-jump method at the single-molecular level and a statistical population deconvolution at the sub-nanometer resolution, we reveal a complex population network with unprecedented transition dynamics in human telomeric sequences that contain four to eight TTAGGG repeats. Our kinetic data firmly establish that G-triplexes are intermediates to G-quadruplexes while long-loop G-quadruplexes are misfolded population minorities whose formation and disassembly are faster than G-triplexes or regular G-quadruplexes. The existence of misfolded DNA supports the emerging view that structural and kinetic complexities of DNA can rival those of RNA or proteins. While G-quadruplexes are the most prevalent species in all the sequences studied, the abundance of a misfolded G-quadruplex in a particular telomeric sequence decreases with an increase in the loop length or the number of long-loops in the structure. These population patterns support the prediction that in the full-length 3' overhang of human telomeres, G-quadruplexes with shortest TTA loops would be the most dominant species, which justifies the modeling role of regular G-quadruplexes in the investigation of telomeric structures.

Photoreactivities of 5-bromouracil-containing RNAs.

5-Bromouracil ((Br)U) was incorporated into three types of synthetic RNA and the products of the photoirradiated (Br)U-containing RNAs were investigated using HPLC and MS analysis. The photoirradiation of r(GCA(Br)UGC)(2) and r(CGAA(Br)UUGC)/r(GCAAUUCG) in A-form RNA produced the corresponding 2'-keto adenosine ((keto)A) product at the 5'-neighboring nucleotide, such as r(GC(keto)AUGC) and r(CGA(keto)AUUGC), respectively. The photoirradiation of r(CGCG(Br)UGCG)/r(C(m)GCAC(m)GCG) in Z-form RNA produced the 2'-keto guanosine ((keto)G) product r(CGC(keto)GUGCG), whereas almost no products were observed from the photoirradiation of r(CGCG(Br)UGCG)/r(C(m)GCAC(m)GCG) in A-form RNA. The present results indicate clearly that hydrogen (H) abstraction by the photochemically generated uracil-5-yl radical selectively occurs at the C2' position to provide a 2'-keto RNA product.

Single strand DNA catenane synthesis using the formation of G-quadruplex structure.

DNA is a good material for constructing nanostructures such as DNA origami. One of the challenges in this field is constructing a topologically complex structure. Here, we synthesized a DNA catenane through the formation of a G-quadruplex structure. The formation of the DNA catenane was investigated by gel electrophoresis. Interestingly, the synthesized DNA catenane was destroyed by heat treatment. Because conventional methods to construct DNA catenane include enzymatic ligation or chemical reactions, DNA is cyclized by covalent bond connection and never destroyed by heat treatment. To our knowledge, this is the first report of the synthesis of DNA catenane without using covalent bonds. Our novel way of synthesizing DNA catenane may be of use in easily recoverable DNA topological labeling.

Intramolecular folding in three tandem guanine repeats of human telomeric DNA.

Intramolecular folding in three tandem guanine repeats of human telomeric DNA has been investigated using optical-tweezers, MD simulation and circular dichroism. A mechanically and thermodynamically stable species in this sequence shows a structure consistent with a triplex conformation. A similar species has also been observed to coexist with a G-quadruplex in a DNA sequence with four tandem guanine repeats.

A single-molecule platform for investigation of interactions between G-quadruplexes and small-molecule ligands.

Ligands that stabilize the formation of telomeric DNA G-quadruplexes have potential as cancer treatments, because the G-quadruplex structure cannot be extended by telomerase, an enzyme over-expressed in many cancer cells. Understanding the kinetic, thermodynamic and mechanical properties of small-molecule binding to these structures is therefore important, but classical ensemble assays are unable to measure these simultaneously. Here, we have used a laser tweezers method to investigate such interactions. With a force jump approach, we observe that pyridostatin promotes the folding of telomeric G-quadruplexes. The increased mechanical stability of pyridostatin-bound G-quadruplex permits the determination of a dissociation constant K(d) of 490 ± 80 nM. The free-energy change of binding obtained from a Hess-like process provides an identical K(d) for pyridostatin and a K(d) of 42 ± 3 µM for a weaker ligand RR110. We anticipate that this single-molecule platform can provide detailed insights into the mechanical, kinetic and thermodynamic properties of liganded bio-macromolecules, which have biological relevance.

Folding pathways of human telomeric type-1 and type-2 G-quadruplex structures.

We have investigated new folding pathways of human telomeric type-1 and type-2 G-quadruplex conformations via intermediate hairpin and triplex structures. The stabilization energies calculated by ab initio methods evidenced the formation of a hairpin structure with Hoogsteen GG base pairs. Further calculations revealed that the G-triplet is more stable than the hairpin conformation and equally stable when compared to the G-tetrad. This indicated the possibility of a triplex intermediate. The overall folding is facilitated by K(+) association in each step, as it decreases the electrostatic repulsion. The K(+) binding site was identified by molecular dynamics simulations. We then focused on the syn/anti arrangement and found that the anti conformation of deoxyguanosine is more stable than the syn conformation, which indicated that folding would increase the number of anti conformations. The K(+) binding to a hairpin near the second lateral TTA loop was found to be preferable, considering entropic effects. Stacking of G-tetrads with the same conformation (anti/anti or syn/syn) is more stable than mixed stacking (anti/syn and vice versa). These results suggest the formation of type-1 and type-2 G-quadruplex structures with the possibility of hairpin and triplex intermediates.

Visualization of dynamic conformational switching of the G-quadruplex in a DNA nanostructure.

We herein report the real-time observation of G-quadruplex formation by monitoring the G-quadruplex-induced global change of two duplexes incorporated in a DNA nanoscaffold. The introduced G-rich strands formed an interstrand (3 + 1) G-quadruplex structure in the presence of K(+), and the formed four-stranded structure was disrupted by removal of K(+). These conformational changes were visualized in a nanoscaffold in real-time with fast-scanning atomic force microscopy.

Overview of formation of G-quadruplex structures.

There are many structures that can be adopted by nucleic acids other than the Watson-Crick duplex. In particular, a noncanonical four-stranded topology, called a G-quadruplex, is of great interest because of its roles in key biological processes such as the maintenance of telomeres and regulation of gene transcription. This review describes the condition for forming the G-quadruplex structure, G-quadruplex-forming sequences, and methods for studying the structures.

A chiral wedge molecule inhibits telomerase activity.

In addition to the Watson-Crick double helix, secondary DNA structures are thought to play important roles in a variety of biological processes. One important example is the G-quadruplex structure that is formed at the chromosome ends, which inhibits telomerase activity by blocking its access to telomeres. G-quadruplex structures represent a new class of molecular targets for DNA-interactive compounds that may be useful to target telomeres. Here, we reported the first example of enantioselective recognition of quadruplex DNA by a chiral cyclic helicene. We propose a new ligand-binding cleft between two telomeric human G-quadruplexes linked by a TTA linker. We found that the cyclic helicene M1 exhibited potent inhibitory activity against telomerase.

The orientation of the ends of G-quadruplex structures investigated using end-extended oligonucleotides.

Human telomere DNA is of intense interest because of its role in the biology of both cancer and aging. The single-stranded telomere terminus can adopt the structure of a G-quadruplex, which is of particular important for anticancer drug discovery many researchers have reported various G-quadruplex structures in the human telomere. Although the human telomere consists of a number of tandem repeats, higher-order G-quadruplex structures are less discussed due to the complexity of the structures. Here we examined the orientation of the ends of the G-quadruplex structures with consideration given to higher-order structures. We prepared end-extended and (Br)G-substituted oligonucleotides. Native PAGE analysis, CD measurements and NMR spectroscopy showed that the ends of stable G-quadruplex structures point in opposite directions. Our results indicate that the human telomere DNA is likely to form rod-like higher-order structures. This may provide important information for understanding telomere structure and the development of telomere G-quadruplex-binding molecules as telomerase inhibitors.

Structural stability analysis of the intermediates in the folding pathway of human telomeric hybrid-1 G-quadruplex based on fragment molecular orbital method.

The human telomeric DNA sequence d[AGGG(TTAGGG)(3)] has been found to form different type of G-quadruplex structure based on NMR(1), X-ray crystallography(2) and circular dichroism (CD). Recently human telomeric hybrid-1 G-quadruplex structure in K(+) solution has been revealed by CD and NMR(3,4,5). However, folding pathway of G-quadruplex structures is not clear to date. It is important to elucidate the intermediate structure of human telomeric hybrid-1 G-quadruplex for drug discovery in addition to having essential knowledge of telomere. In this study, we designed two types of triplex intermediate model from hybrid-1 NMR structure and evaluated their stabilities with ab initio Fragment Molecular Orbital (FMO) method(6,7,8). The folding pathways of human telomeric hybrid-1 G-quadruplex structure are discussed.

Orientation of ends of G-quadruplex structure investigated with end-extended oligonucleotides.

The human telomere terminus can adopt the structure of a G-quadruplex. This structure has become an attractive target for anticancer drugs, because it effectively inhibits telomerase activity. In this study, we investigated the orientation of both 5' and 3' ends of the stable G-quadruplex structure. To verify the orientation, we designed end-extended G-quadruplex forming oligonucleotides. We carried out gel electrophoresis and the NMR analysis and found that the ends of the stable G-quadruplex structure are located on opposite faces of each of the quadruplexes.

Folding pathways of hybrid-1 and hybrid-2 G-quadruplex structures.

The G-rich sequence in the human telomeric DNA can form the G-quadruplex structure. The G-quadruplex structure has become an attractive target for the anticancer drugs, because it effectively inhibits telomerase activity. Recently, the human telomere G-quadruplex in K(+) solution has been determined as a hybrid structure. This structure is called hybrid-1. More recently, the hybrid-2 G-quadruplex has been determined by NMR. Hybrid-2 G-quadruplex differs from hybrid-1 in its loop arrangement and strand orientation. Here, we propose the folding pathways of hybrid-1 and hybrid-2 G-quadruplex structures. In hybrid-1, we proposed two pathways. In one pathway, the random coil forms triplex-1; in another pathway, the random coil forms chair-1. Similarly, we proposed two pathways in hybrid-2. In one pathway, the random coil forms triplex-2; in another pathway, the random coil forms chair-2. The folding pathways of human telomeric hybrid-1 G-quadruplex and hybrid-2 G-quadruplex structures were proposed using MO calculation and molecular modelling.

G-quadruplex structures of human telomere DNA examined by single molecule FRET and BrG-substitution.

The human telomere is known to form the G-quadruplex structure to inhibit the activity of telomerase. Its detailed structure has been of great interest. Recently, two kinds of the parallel-antiparallel hybrid structures have been specified in K(+) solution. However, the G-quadruplex structure is generally thought to be in equilibrium among different structures. Here, we describe the single-pair fluorescence resonance energy transfer (sp-FRET) experiments on telomere samples with bromoguanine (BrG)-substitutions, which control the G-quadruplex structures, at different positions and one without any substitution. The observed FRET distributions were decomposed into five components and the relative population of these components depended on the BrG-substitution positions. In order to consistently explain the variety of conformations, we proposed a novel structural model, the so-called triple-strand-core model. On the basis of this model, the components of the FRET distributions were attributed to the mixed-chair hybrid structures, which were reported recently, and chair-type antiparallel structures, which can be predicted from this model. The FRET efficiencies of these structures were explained in terms of partially broken structures due to steric hindrance and inappropriate capping. This basic model also consistently explains experimental results reported previously. Furthermore, using this model, the folding pathway of the hybrid structures and T-loop formation can be predicted.