Metal-Mediated Base Pairs in Nucleic Acids with Purine- and Pyrimidine-Derived Nucleosides

Part of the Metal Ions in Life Sciences book series (MILS, volume 10)


Metal-mediated base pairs are transition metal complexes formed from complementary nucleosides within nucleic acid double helices. Instead of relying on hydrogen bonds, they are stabilized by coordinative bonds. The nucleosides acting as ligands do not necessarily have to be artificial. In fact, several examples are known of naturally occurring nucleobases (e.g., thymine, cytosine) capable of forming stable metal-mediated base pairs that are highly selective towards certain metal ions. This chapter provides a comprehensive overview of metal-mediated base pairs formed from natural nucleosides or from closely related artificial nucleosides that are pyrimidine or purine derivatives. It addresses the different strategies that lead to the development of these base pairs. The article focuses on structural models for metal-mediated base pairs, their experimental characterization within a nucleic acid, and on their possible applications.


cytosine deazaadenine M-DNA metal-mediated base pairs polymerase sensor thymine 

1 Introduction to Metal-Mediated Base Pairs

The interaction of metal salts with nucleic acids has long been investigated (see refs in [1]). Early experiments focused on the reaction of DNA from natural sources or of polynucleotides with transition metal ions such as Co2+, Ni2+, Mn2+, Zn2+, Cd2+, Cu2+, Hg2+, or Ag+ [2]. It became clear that the metal ions are extremely diverse in their reaction with nucleic acids [3,4], and that in particular Hg2+ and Ag+ were able to bind strongly to the nucleobases without unwinding the double helices and without binding to the phosphate groups [4]. Replacement of protons from the nucleobases was suggested as one possible binding mode, termed “type II binding” in the case of Ag+ [5]. The fact that one Ag+ was included per two nucleotide residues, accompanied by proton displacement, was attributed to the conversion of a N–H···N hydrogen bond within a complementary base pair to two coordinative bonds in N-Ag-N [5]. In the case of Hg2+, experimental data pointed towards the preferential incorporation of the metal ion between the deprotonated N3 atoms of two (mismatched) thymidine residues [6].

The advent of automated solid-phase synthesis of oligonucleotides enabled a much more detailed picture of the metal-binding properties to be obtained by allowing sequence-specific studies. Numerous metal-mediated base pairs, i.e., base pairs in which hydrogen bonds are formally replaced by coordinative bonds to metal ions, have been generated site-specifically within nucleic acid double helices [7, 8, 9]. Hence, it became possible to investigate in more detail the above mentioned metal-mediated base pairs comprising natural nucleotides. In addition, artificial nucleotides have been developed with a larger affinity towards metal ions, increasing significantly the scope of this type of functionalized nucleic acid. Using artificial nucleosides the structures of metal-modified DNA double helices could be determined [10,11]. In this chapter, metal-mediated base pairs with natural nucleosides as well as artificial purine- or pyrimidine-derived nucleosides will be discussed. Selected examples for metal-mediated base pairs formed from artificial nucleosides without close structural resemblance to their natural counterparts are dealt with in Chapter 10 of this book [12].

2 M-DNA: A Metal Ion Complex of DNA

2.1 Formation of M-DNA

So-called M-DNA has been reported to form from regular B-DNA, both bacterial and synthetic, upon treatment with divalent cations such as Zn2+, Ni2+, or Co2+ at elevated pH values (pH > 8) [13,14]. Reaction with EDTA converts M-DNA back to B-DNA. The properties of this metal complex of DNA are quite distinct from those of B-DNA. For example, the well-established intercalator ethidium does not intercalate into M-DNA [13]. In fact, this observation has led to the development of a convenient assay for M-DNA formation based on the decrease in fluorescence that accompanies M-DNA formation. NMR studies suggest that the imino protons of thymine and guanine are absent in M-DNA [13], even though according to their pKa values of 10.5 and 10.0 (values given for the respective 5’-phosphates) [15] they should still be protonated under slightly alkaline conditions. A detailed study revealed that each metal ion displaces one proton from the original B-DNA [14]. Interestingly, the UV and CD spectra of M-DNA almost resemble those of B-DNA, suggesting similar structures [13]. Based on these observations, base pairing schemes were suggested for M-DNA, in which one imino (or amino) proton of a base pair is replaced by a metal ion, leading to metal-mediated base pairs.

2.2 Structural Models

In addition to the originally proposed models for the metal-mediated base pairs (Scheme 1a, b) [14], combined computational and experimental modelling stu­dies have identified other potential geometries [16]. Two of these are shown in Schemes (1c) and (1d).
Scheme 1

Base pairing schemes suggested for M-DNA (R, R’ = DNA backbone). (a) Replacement of the thymine and guanine imino protons by Zn2+ [14]. (b) Replacement of an adenine and cytosine amino proton by Zn2+ [14]. (c) Replacement of the guanine imino proton by Zn2+ [17]. (d) Loss of two hydrogen bonds without proton displacement by Zn2+ [18].

According to the base pairing scheme shown in Scheme (1a), the metal ion replaces an imino proton of either thymine (A:T base pair) or guanine (G:C base pair). It is additionally coordinated to one aqua ligand residing in the minor groove [14]. Alternatively, the metal ion was suggested to displace an amino proton of either adenine (A:T base pair) or cytosine (G:C base pair) as shown in Scheme (1b). No aqua ligand is necessary to achieve a four-fold coordination of the metal ion in this model. However, the complementary nucleobase is required to be present in its rare enol form [14]. Taking into account the slightly alkaline conditions, a different model has been put forward for a metal-mediated G:C base pair in which the Zn2+ ion is coordinated by hydroxido ligands rather than an aqua ligand (Scheme 1c) [17]. Based on the single-crystal X-ray diffraction analysis of a Zn2+ complex of 1-methylcytosine, a model for metal cross-links occurring during DNA melting in the presence of Zn2+ ions had been suggested [18] that was later proposed as a feasible base pairing scheme for M-DNA [16]. However, in this model, no proton displacement by Zn2+takes place (Scheme 1d). All of these models contain a divalent cation coordinated by four donor atoms. In the case of Zn2+, this typically translates into a tetrahedral coordination environment. According to a computational study, only the base pair shown in Scheme (1d) can adopt a planar geometry after full geometrical optimization [16], which is usually considered important for the formation of a stable DNA double helix with stacking neighboring base pairs. However, this base pair still contains the imino proton, and the resulting duplex would need to be parallel-stranded [16]. Both of these requirements are not in agreement with the experimental results.

Interestingly, the energy required to “flatten” the Zn2+-mediated G:C base pair shown in Scheme 1(a) is relatively low (<8 kcal/mol), suggesting that it indeed represents a plausible model for building an antiparallel double helix [16]. Unfortunately, no structural information has yet been obtained for M-DNA. Single-crystal X-ray diffraction analyses of oligonucleotides crystallized in the presence of Zn2+, Co2+, or Ni2+ under slightly alkaline conditions show a coordination of the metal ions to the N7 positions of the terminal guanine residues rather than an incorporation along the helical axis [19].

2.3 Physical Properties

M-DNA has gained a lot of attention because of reports that it supports metallic-like conduction [20]. Based on such electronic properties, the creation of DNA-based devices for nanoelectronic applications was suggested to be feasible. Evidence to support molecular wire behavior of M-DNA was collected using a variety of methods. For example, current-voltage curves of bundles of M-DNA duplexes placed between two electrodes separated by a gap of at least 1 μm have been measured directly [20]. Moreover, electrochemical investigations of self-assembled DNA monolayers on gold electrodes showed an increased electron transfer rate between different solute redox probes and the gold surface for M-DNA in comparison with B-DNA [21,22]. In particular, a differential behavior towards base mismatches was observed for B-DNA and M-DNA that was exploited for mismatch detection [23]. Finally, the attachment of donor and acceptor fluorophores at opposite ends of an M-DNA duplex enabled the investigation of charge transfer through the double helix. The observed fluorescence quenching was interpreted in terms of an electron hopping mechanism for electron transfer [24].

However, it needs to be stated that the enhanced electrical conductivity of M-DNA is heavily disputed. A magnetic study of the electronic states showed no substantial differences between B-DNA and M-DNA except for Mn2+-doped DNA. Nonetheless, even in the latter case a charge transport through the metalated DNA was not observed [25]. Instead, it was suggested that the divalent cations replace the Na+ ions shielding the negative charge of the phosphate backbone. A topographic characterization by atomic force microscopy showed that M-DNA is about five times shorter and ten times higher than regular B-DNA [26]. Electrostatic experiments performed with M-DNA adsorbed on mica surface showed an insulating behavior. The results of the electrochemical characterization have also been interpreted differently: The apparent enhancement of electrical conductivity of M-DNA could be attributed to the mere presence of divalent cations. Increased charge neutralization by these cations facilitates a penetration of the DNA monolayer by the negatively charged redox mediator, hence leading to the observed increase in conductivity [27]. The release of ethidium from B-DNA upon formation of M-DNA might be a result of the relatively low solubility of M-DNA: Upon precipitation under the conditions of M-DNA formation, ethidium could actually just dissociate from the precipitating B-DNA [28]. The displacement of the protons usually involved in hydrogen bonding could at least in part be explained by a (transient) coordination of the divalent cations to the nucleobases. From model chemistry it is well-known that the attachment of a metal ion to a nucleobase results in an acidification of both endocyclic imino and exocyclic amino groups [15]. Finally, the fluorescence quenching observed upon M-DNA formation could be attributed to its more compact size, resulting in either intermolecular FRET between aggregated double helices or intramolecular FRET as a result of metal-mediated fraying of the helix ends [28].

The prospect of self-assembling nanoscale electronic architectures based on DNA is certainly a fascinating one [29]. At present, it is not clear whether this goal can be achieved with M-DNA. Nonetheless, the studies of M-DNA have aroused interest in an investigation of metal-mediated base pairs with both natural and artificial nucleobases.

3 Metal-Mediated Base Pairs with Pyrimidine-Derived Nucleobases

3.1 Hg2+-Mediated T:T Base Pairs

3.1.1 History of the T–Hg–T Base Pair

In the early 1950s, a decrease in DNA viscosity was observed upon the addition of HgCl2 [30]. First attempts of an explanation involved a decrease in the overall size of the nucleic acid caused by a diminishing electrostatic repulsion between the phosphate groups due to the positively charged Hg2+, potentially accompanied by intramolecular cross-links of the type P–O–Hg–O–P [30]. After it became clear that the nucleobases act as the binding sites for Hg2+ rather than the phosphate groups [31], the formation of a 2:1 complex was proposed, containing two deprotonated thymine residues and a central mercuric ion [6]. It was suggested that these base pairs (Scheme 2) were formed between two proximate thyminate residues of opposite strands by means of a chain shift mechanism [32].
Scheme 2

Chemical structure of a T–Hg–T base pair (R, R’ = DNA backbone).

The molecular structure of the complex between deprotonated 1-methylthymine and Hg2+ strongly supports the proposed metal-mediated base pair (Figure 1). However, the structure obtained by single crystal X-ray diffraction analysis contains the nucleobases in a transoid orientation whereas a cisoid arrangement of the glycosidic bonds is required for an incorporation into a B-type duplex [33].
Figure 1

Molecular structure of [Hg(1-methylthyminate)2] [33].

3.1.2 Characterization of the T–Hg–T Base Pair within a DNA Duplex

As mentioned above, the advent of solid phase oligonucleotide synthesis enabled sequence-specific studies of the metal-binding behavior of the various nucleobases. Accordingly, the selective formation of T–Hg–T base pairs inside a DNA duplex could be investigated. Using hairpin structures comprising a loop of two or three neigh­boring thymine residues, a rearrangement towards a regular duplex was observed upon the addition of Hg2+. This rearrangement was accompanied by the formation of T–Hg–T base pairs, as could be shown by 1H NMR spectroscopy [34]. Interestingly, a hairpin containing four thymine residues in the loop did not rearrange towards a regular duplex upon the addition of Hg2+. Instead, one intramolecular T–Hg–T cross-link between the first and the last thymine residues was observed [34], indicating a complex relationship between oligonucleotide sequence and conformation.

A quantitative study regarding the thermal stabilization of a T–Hg–T base pair indicated that the addition of Hg2+ to a DNA duplex containing one T:T mismatch results in an increase of the melting temperature Tm of 10°C (from 37 to 47°C) [35]. Control experiments using a regular A:T base pair instead of the mismatch showed a Tm of 44°C, indicating that the metal-mediated base pair is more stable than the naturally occurring A:T pair [35]. An isothermal titration calorimetry study revealed that the binding constant for specific binding of Hg2+ to a T:T mismatch amounts to almost 106 M–1[36].

1H NMR spectroscopy was applied to investigate in more detail the formation of this metal-mediated base pair. By using a duplex containing two T:T mismatches, the formation of T–Hg–T base pairs could be followed by observing the resonances of the thymine imino protons. In the absence of Hg2+, four imino proton resonances were detected in the corresponding chemical shift region (Figure 2a, 10.4–11.1 ppm) [35]. Addition of Hg2+ led to a decrease in signal intensity, indicating a substitution of the imino protons by Hg2+ and hence the formation of a metal-mediated base pair (Figure 2c). In experiments using approximately half the amount of metal ions required for base pair formation, three independent species were observed in the 1H NMR spectrum, i.e., the Hg2+-free state and two partially Hg2+-bound states (Figure 2b) [35]. This clearly indicates that the metal ions are tightly bound and that the dissociation and association processes occur slowly on the NMR timescale. Hence, T–Hg–T base pairs are formed in an equilibrium process (Figure 2d) [35]. Moreover, the two metal-mediated base pairs within the duplex are not formed cooperatively but independently from each other.
Figure 2

Imino proton region of the 1H NMR spectra of an oligonucleotide double helix. (a) In the absence of Hg2+; (b) in the presence of Hg2+ (0.8 equivalents with respect to the duplex); (c) in the presence of Hg2+ (2.0 equivalents with respect to the duplex); (d) equilibrium system between the duplex and Hg2+. Reprinted with permission from [35]; copyright 2006 American Chemical Society.

The first direct proof for the formation of a T–Hg–T base pair within a double helix was obtained by 15N NMR spectroscopy. By incorporating 15N-labelled thymine residues at different positions within oligonucleotide strands, metal ion binding to these residues could be unequivocally confirmed [37]: In the presence of Hg2+, a significant downfield shift of the 15N resonances of the endocyclic N3 atoms of thymine by ∼30 ppm was observed. Such a significant shift can only be explained by the substitution of the imino proton by Hg2+. Most interestingly, when both thymine residues within the metal-mediated base pair were 15N-labelled a splitting of the 15N resonance (2.4 Hz) was observed (Figure 3). This splitting was attributed to a \( {}^{\text{2}}{J}_{{}^{\text{15}}\text{N}{,}^{\text{15}}\text{N}}\) coupling across the central metal ion, providing direct proof for the formation of the T–Hg–T base pair [37].
Figure 3

Detail of the 15N NMR spectrum of the partially 15N-labelled oligonucleotide duplex (R, R’ = DNA backbone). The \( {J}_{\text{N}{,}^{\text{15}}\text{N}}\)coupling of 2.4 Hz is highlighted by arrows. Adapted with permission from [37]; copyright 2007 American Chemical Society.

3.1.3 Application in Hg2+ Sensors

With Hg2+ being a toxic environmental pollutant, its reliable detection from aqueous solution is highly desirable. Accordingly, the formation of T–Hg–T base pairs has been exploited for the development of DNA-based Hg2+-sensor systems. In the first example of this type, a thymine-rich oligonucleotide functionalized with a fluorophore at its 3’-end and a quencher at its 5’-end was applied [38]. In the absence of Hg2+ the oligonucleotide adopts a randomly coiled single-stranded conformation. However, in the presence of Hg2+ a hairpin structure comprising several T–Hg–T base pairs is adopted. In this conformation, the fluorescence is quenched with increasing concentration of Hg2+ because fluorophore and quencher are located close to each other [38]. To ensure a high specificity of the sensor, control experiments with other metal ions were performed, such as Ca2+, Mg2+, Cu2+, Fe2+, Cd2+, Pb2+, Zn2+, Ni2+, Mn2+, Co2+. None of these influence significantly the fluorescence spectra. Hence, this Hg2+ sensor is sensitive enough to detect small amounts of Hg2+ in solutions containing an excess of other metal ions [38]. By combining the T–Hg–T base pairs with Au nanoparticle-functionalized reporter DNA [39] or electrochemiluminescent tags [40], the detection limit can be lowered down to the picomolar concentration range.

As mercury cations can exist in two oxidation states, the T–Hg–T base pairs have also been applied as sensors for the redox environment [41]. Here, an oligonucleotide containing only thymine residues was labelled with fluorophore and quencher as described above. In the presence of Hg2+ a hairpin structure with metal-mediated base pairs is adopted, leading to a quenched fluorescence. Upon introduction of a reductive environment (via the addition of Na2SO3), Hg2+ is reduced to Hg\( {}_{2}^{2+}\). As the formation of Hg2+-mediated base pairs is no longer possible, the folded structure is converted back to the randomly coiled conformation, leading again to a recovering fluorescence [41].

3.1.4 U–Hg–U, an RNA Analogue

Apart from a gold-mediated G:C base pair accidentally formed during the crystallization of a nucleic acid [42], only one metal-mediated base pair has been reported as yet for RNA [43]: In analogy to the well-established T–Hg–T base pair, RNA duplexes containing uracilate–Hg2+–uracilate (U–Hg–U) base pairs have been synthesized. As the presence or absence of a methyl group at the C5 position represents the only difference between thymine and uracil, a similar complexation behavior towards Hg2+ was expected.

So far, five RNA duplexes have been described, comprising stretches of up to 20 consecutive uracil residues, thereby being able to potentially form up to 20 Hg2+-mediated base pairs [43]. A combination of different spectroscopic techniques and dynamic light scattering (DLS) were applied to demonstrate that the N3 imino proton of uracil is substituted by Hg2+[43], in analogy to the situation described above for thymine. The RNA oligonucleotides under investigation were either regular duplexes with U:U mismatches (most likely in the form of a wobble pair) or hairpins with the uracil residues located in the loop. The latter case resembles the situation described in Section 3.1.2, where the formation of T–Hg–T base pairs leads to a hairpin-to-duplex transition [34]. For the RNA hairpins, this conformational change was followed by DLS and DOSY NMR spectroscopy, as double helix and hairpin show significantly different hydrodynamic radii rH: Upon the addition of Hg2+, rHincreases by about 30%, indicating that a conformational change takes place from hairpin to a regular duplex with U–Hg–U base pairs [43].

This conformational change was further monitored by NMR spectroscopic techniques other than DOSY. In particular, the 1H,1H-NOESY spectrum of the hairpin r(GGAGCGCGUUUUUUCGCGCUCC) comprising six uracil residues in the loop displays well-resolved and dispersed peaks in the absence of Hg2+ [43]. The fact that a sequential walk could be followed through the entire sequence indicates the formation of a well-structured loop. The A-type conformation of the stem region was confirmed by characteristic cross-peaks [43]. Upon the addition of Hg2+, the 1H NMR spectra changed in the same manner as previously observed for DNA duplexes comprising T–Hg–T base pairs: The signal intensity of the N3 imino proton resonances of the uracil residues involved in metal-mediated base pairing decreased [43]. Moreover, a study applying a 13C-labelled oligonucleotide confirmed that only those uracil residues not involved in Watson-Crick base pairing (i.e., all uracil residues except of U20) bind to Hg2+. As can be seen from the 1H,13C-HSQC spectrum (Figure 4), the resonances of the central uracil residues shift significantly after the addition of Hg2+, indicating the formation of U–Hg–U base pairs, whereas the resonance of U20 is least affected [43].
Figure 4

Section of the 1H,13C-HSQC spectra of r(GGAGCGCGUUUUUUCGCGCUCC) in the absence (left) and presence (right) of Hg2+, showing the 1J correlation between H6 and C6 of all uracil residues. Reprinted with permission from [43], copyright 2008 Elsevier.

In analogy to the T–Hg–T base pairs, the formation of U–Hg–U base pairs is accompanied by an increase in thermal stability [43]. CD spectroscopy was used to confirm the A-type helical conformation (Figure 5). This study showed that U–Hg–U base pairs are formed in an equilibrium process, as indicated by the appearance of isosbestic points upon the stepwise addition of Hg2+. Once all U–Hg–U base pairs are formed, excess Hg2+ binds unspecifically to other coordination sites of the duplex [43].
Figure 5

CD spectra of r(GGAGCGCGUUUUUUCGCGCUCC), confirming an A-type duplex conformation in the absence and presence of up to one equivalent of Hg2+. Here, one equivalent is defined as the amount of metal ion required to saturate every possible metal-mediated base pair. The isosbestic points observed up to the addition of 1 equivalent of Hg2+ are highlighted by circles. Reprinted with permission from [43], copyright 2008 Elsevier.

Interestingly, Hg2+ can be removed from the U–Hg–U base pairs by treating the RNA duplexes with a chelating resin, restoring the previous nucleic acid conformation [43]. This behavior is unlike that found for T–Hg–T base pairs [37] (and also for the artificial imidazole–Ag+–imidazole base pairs) [11], where the metal ions cannot be removed by using chelating agents. Hence, this different reactivity might be attributed to the A-type conformation of the RNA duplex, in which the metal ions are not aligned along the helical axis but twisted around it.

The incorporation of U–Hg–U base pairs into RNA duplexes demonstrates that this nucleic acid should not be disregarded in the context of metal-mediated base pairs. As a result of the different helical conformation compared with DNA, the scope of this type of nucleic acid functionalization can be significantly broadened by including RNA.

3.2 Ag+-Mediated C:C Base Pairs

Apart from thymine and uracil, the third naturally occurring pyrimidine nucleobase cytosine has also been utilized in the construction of metal-mediated base pairs. In contrast to the previously discussed examples, metal-mediated self-pairs of cytosine are specifically formed in the presence of Ag+.

The reversible binding of Ag+to oligonucleotides containing cytosine-rich regions has been known for about 50 years [44]. In the course of some of the first studies, the interaction of Ag+ with poly-d(C) was investigated by potentiometric and photometric methods [3,5]. Even though the metal-binding sites were not clear at that point, the formation of Ag+-mediated base pairs in the so-called “type II binding” was suggested as a result of the substitution of a proton involved in hydrogen bonding by a metal ion. In contrast to the molecular structure of [Hg(1-methylthyminate)2] (Figure 1), the complex [Ag(1-MeC)](NO3) synthesized from 1-methylcytosine (1-MeC) and AgNO3 shows a 1:1 stoichiometry between ligand and metal ion [45,46]. In this dinuclear complex, Ag+ is bound by two cytosine residues in an N3–Ag–O2 fashion and, due to the coordination of the nitrate counter ions, displays a distorted trigonal coordination sphere. Based on this structure, models were suggested for Ag+-mediated base pairs, e.g., in nucleic acid regions of high G:C content [45].

The Ag+-mediated cytosine self-pair C–Ag–C was recently discovered as another example for a metal-meditated self-pair of naturally occurring nucleobases [47]. By using an oligonucleotide double helix comprising a single C:C mismatch within otherwise regular Watson-Crick base pairs, Ag+ was shown to be the only metal ion significantly stabilizing the duplex, with an increase of the melting temperature Tm of 8°C [47]. Other metal ions such as Hg2+, Cu2+, Ni2+, Pd2+, Co2+, Mn2+, Zn2+, Pb2+, Cd2+, Mg2+, Ca2+, Fe2+, Fe3+, or Ru3+ had no notable effects on Tm, making the C:C mismatch highly selective for the binding of Ag+. An investigation by means of 1H NMR spectroscopy confirmed the formation of C–Ag–C base pairs, as is obvious by significant changes in the imino proton region of the spectra. In analogy to what had been observed for T–Hg–T base pairs [35], a spectrum recorded at a lower Ag+ concentration revealed the existence of two independent species, namely the metalated and non-metalated duplex [47]. The binding constant for specific binding of Ag+ to a C:C mismatch within a regular DNA duplex lies in the range of 105M–1 [48].

In these examples, the formation of a Watson-Crick duplex containing a single C:C mismatch enforces a cisoid orientation of the nucleobases inside the C–Ag–C base pair (Scheme 3a). However, the corresponding transoid base pair is expected to be of higher stability, due to diminished electrostatic repulsion and the possibility to form an additional hydrogen bond (Scheme 3b). In fact, CD and UV spectroscopic investigations of d(C8) containing only cytosine moieties suggest the preferred formation of transoid C–Ag–C base pairs if the strand orientation is not predetermined by surrounding nucleobases [49]. Importantly, a comparison of the CD spectra in the presence of Ag+ and under slightly acidic conditions, respectively, rules out that the addition of Ag+ leads to an i-motif-like structure.
Scheme 3

Postulated structures of the Ag+-mediated base pairs C–Ag–C (cisoid and transoid) and C–Ag–5MiC (R, R’ = DNA backbone) [47,49,50]. The latter base pair probably also contains one hydrogen bond. Coordinative bonds are drawn as arrows and hydrogen bonds as broken lines.

The preferred orientation of a C–Ag–C base pair was probed by a series of experimental and theoretical investigations. For instance, various Watson-Crick duplexes differing only in a single base mismatch comprising cytosine and/or 5-methylisocytosine (5MiC) (Scheme 3c) were synthesized and their melting temperature determined [50]. The formation of the C–Ag–5MiC base pair (ΔTm = +18.9°C) results in a significantly larger increase in Tm than that of the cisoid C–Ag–C base pair (ΔTm = +4.8°C). As the C–Ag–5MiC base pair geometry closely resembles that of the transoid C–Ag–C one, these data strongly suggest that the latter is the thermodynamically more stable one. Moreover, experiments with DNA duplexes covalently locked in either a parallel or an antiparallel orientation proved the stabilization of parallel-stranded duplexes by transoid C–Ag–C base pairs [51]. Very recently, dispersion-corrected DFT calculations have been carried out revealing that the transoid C–Ag–C is indeed stabilized by one hydrogen bond and therefore about 7 kcal mol–1 more stable than the corresponding cisoid one [52].

Applications of the Ag+- mediated cytosine self-pair include the oligonucleotide-based detection of this metal ion. In analogy to the Hg2+ sensor mentioned in Section 3.1.3, a turn-off fluorescent sensor utilizing the quenching of fluorescence as a result of a conformational change of an oligonucleotide upon the formation of C–Ag–C base pairs has been reported [47]. A similar turn-on sensor based on a fluorescein-labelled single-stranded oligonucleotide has also been reported, taking advantage of the fact that graphene oxide selectively adsorbs and quenches the labelled single-stranded DNA probe but not the duplex comprising C–Ag–C base pairs [53]. Both sensors are highly selective towards Ag+ with detection limits in the nanomolar range.

3.3 Chemically Modified Pyrimidines

3.3.1 5-Substituted Uracil Derivatives

As the naturally occurring pyrimidine nucleobases have been found to be excellent ligands for the formation of metal-mediated base pairs, several chemically modified pyrimidine derivatives have been investigated in an attempt to control the stability and selectivity towards different metal ions. A promising approach for such a fine-tuning is based on the use of 5-substituted uracil derivatives [54]. The pKa value of the amide proton of these pyrimidine derivatives varies from 6.5 to 9.8, depending on the electron-withdrawing effect of the substituent in the 5-position (Scheme 4). As the formation of a metal-mediated base pair requires the deprotonation of this site, a change in pKa should directly influence the stability of such a base pair. Moreover, the affinity towards metal ions depends on the nucleophilic character of the ligand which often goes along with its basicity.
Scheme 4

5-Substituted uracil derivatives used for metal-mediated base pairs (5MU: 5-methyl­uracil = thymine; 5BrU: 5-bromouracil; 5FU: 5-fluorouracil; 5CNU: 5-cyanouracil; R = DNA backbone) [54]. The pKavalues of the amide protons are shown as well (for R = CH3).

The investigation of oligonucleotide duplexes containing one base mismatch each of the 5-substituted uracil derivatives revealed significant differences in the reactivity towards Hg2+ and Ag+ in the pH range of 5.5–9. For example, in the case of a T:T mismatch the addition of Ag+ does not lead to a change of Tm (within error limits), whereas the addition of Hg2+ results in a duplex stabilization of ΔTm≈ 7°C almost irrespective of the pH [54]. In contrast, 5-fluorouracil (5FU) shows a completely different reactivity. Under slightly acidic conditions, the presence of Hg2+ results in a significant stabilization of the duplex (ΔTm = 7°C) whereas the addition of Ag+ does not affect Tm at all. By changing the pH to slightly basic conditions, the reactivity is reversed: Now Ag+ stabilizes the duplex more significantly than Hg2+Tm = 14°C vs. 6°C) [54].

The differential binding of these metal ions was corroborated by a mass spectrometric study from a solution containing both metal ions, directly proving the specific binding of Hg2+ at pH 7 and of Ag+ at pH 9, respectively [54]. Interestingly, a UV titration suggested that the base pair stabilized by Ag+ contains two metal ions (Scheme 5). This finding did not come as a total surprise, as it had previously been predicted based on the single crystal X-ray diffraction analysis of the complex [Ag(1-methyluracilate)] [55]. In summary, a carefully designed manipulation of a nucleobase at a position that is not involved in metal-ion binding can affect the stability of the resulting metal-mediated base pairs and concomitantly the affinity towards different metal ions.
Scheme 5

Proposed formation of 5FU–Ag25FU (R = DNA backbone) [54].

3.3.2 4-Substituted Pyrimidinone Derivatives

All base pairs discussed so far in Section 3 contain a metal ion with linear coordination geometry (Hg2+ or Ag+). Even though metal complexes are known that reveal the ability of thymine and cytosine to act as chelating ligands [56,57], no proof exists of comparable metal-mediated self-pairs of pyrimidines within a nucleic acid duplex. Hence, to extend the concept of pyrimidine self-pairs to metal ions preferring a more sophisticated coordination geometry, a functionalization of the metal-binding site of the nucleobase is necessary. In this context, the synthesis and metal-binding properties of oligonucleotides containing the artificial nucleobase 4-(2’-pyridyl)-pyrimidinone (PyrPy) have been reported [58]. This nucleobase is a cytosine derivative containing a pyridyl substituent attached to C4 instead of the amino group. Oligonucleotide duplexes containing the PyrPy:PyrPymismatch as well as combinations of PyrPy with naturally occurring nucleobases have been investigated. In the absence of metal ions, PyrPy has a destabilizing effect on the duplex (ΔTm = –15.1°C compared to a regular G:C base pair). However, in the presence of Ni2+ the formation of a PyrPy–Ni–PyrPy base pair (Figure 6a) was delineated from a significant increase in Tm. The artificial PyrPy–Ni–PyrPy base pair is formed in a highly selective manner, as neither the presence of other divalent metal ions (Co2+, Cu2+, Zn2+, Fe2+, Mn2+) nor a replacement of one of the PyrPy residues by a natural nucleobase results in a significant increase of Tm. As a result, the artificial PyrPy–Ni–PyrPy base pair has a stability and a mismatch discrimination comparable to those of natural Watson-Crick pairs [58].
Figure 6

Schematic representations of (a) the PyrPy–Ni–PyrPy base pair [58] and (b) the PyrBipy–Ag–4Py base pair [59]; (c) superposition of the geometry-optimized structure of PyrBipy–Ag–Py on a natural A:T base pair (R, R’ = DNA backbone). The distances between the glycosidic bonds are indicated by arrows. This figure has been adapted from [59] with permission of The Royal Society of Chemistry; copyright 2005.

Following the strategy of increasing the number of metal-binding sites on a ­cytosine derivative to modify its metal-binding properties, a 2,2’-bipyridine residue was attached at C4 instead of a pyridine ring, leading to a nucleobase with a terpyridine-derived binding motif (PyrBipy) (Figure 6b) [59]. As PyrBipy is a tridentate ligand, the formation of a planar base pair requires a monodentate ligand in the complementary position, e.g., a pyridyl moiety glycosylated at C3 (3Py) or C4 (4Py), respectively. Interestingly, only the combination of PyrBipy with 4Py leads to the formation of a stable Ag+- mediated base pair (ΔTm = +12.9°C) [59]. The addition of other metal ions or the choice of a different nucleobase complementary to PyrBipy (e.g., PyrBipy, 3Py, G, A, C, T) does not result in a remarkable duplex stabilization [59]. The different stability of 3Py and 4Py is a prime example for the importance of base pair dimension and geometry. As shown by the geometry-optimized structure of a PyrBipy–Ag–Py base pair, the distance between the glycosidic bonds is larger for PyrBipy–Ag–4Py and therefore resembles more closely the distance found in natural DNA double helices (Figure 6c) [59].

4 Metal-Mediated Base Pairs with Purine-Derived Nucleobases

4.1 1-Deazaadenine and 1,3-Dideazaadenine

Metal-mediated base pairs containing 1-deazaadenine (c1A) or 1,3-dideazaadenine (c1,3A) are special inasmuch as they adopt a Hoogsteen geometry. As the purine bases are defunctionalized at their Watson-Crick edge, the base pairs cannot be incorporated into a Watson-Crick duplex. What makes them special is that they have a thermal stability comparable to that of natural base pairs [60], contrary to most other metal-mediated base pairs. This provides the opportunity to create DNA duplexes containing a large number of metal-mediated base pairs without losing the reversibility of the hybridization process. In principle, metal-mediated Hoogsteen base pairs can also be formed from unmodified purine bases, but at current examples for such base pairs are limited to triple-helical systems [61] or unusual left-handed DNA duplexes [62]. Hence, 1-deaza- and 1,3-dideazapurine derivatives are required for the exclusive formation of a metal-modified Hoogsteen double helix.

Hoogsteen duplexes from 1-deazaadenine and thymine adopt an established structural conformation in the absence of transition metal ions (Scheme 6a) [63]. In agreement with other metal-mediated base pairs formally derived by the substitution of an imino proton by a metal ion, Ag+- mediated T–Ag–c1A base pairs could be devised (Scheme 6b) [60]. For this, different oligonucleotides containing 1-deazaadenine were synthesized and investigated in the presence of complementary thymine-rich sequences and Ag+. For an equimolar mixture of d(T20) and d(c1A19A) the specific binding of one metal ion per artificial base pair was observed, accompanied by an increase in melting temperature of ΔTm > 36°C (approx. 2°C per base pair) [60]. As a result, Tm is still low enough to ensure a completely reversible duplex formation in aqueous solution. CD spectroscopic experiments revealed only marginal differences in the absence and presence of Ag+, indicating that the insertion of a metal ion into the Hoogsteen base pair does not affect the overall duplex structure [60].
Scheme 6

Artificial Hoogsteen base pairs formed from (a) thymine and 1-deazaadenine (c1A) [63]; (b) thyminate, Ag+, and 1-deazaadenine [60]; (c) thyminate, Ag+, and 1,3-dideazaadenine (c1,3A) (R, R’ = DNA backbone) [64].

The use of 1,3-dideazaadenine as an artificial nucleobase was investigated as well [64]. At first glance a reactivity similar to that of 1-deazaadenine could be expected, as both nucleobases comprise an identical Hoogsteen edge. However, the nucleobases show different base-pairing properties in the absence of transition metal ions, as 1,3-dideazaadenine does not form stable Hoogsteen base pairs with thymine [65]. A differential reactivity was also established in the presence of Ag+. The self-complementary oligonucleotide sequences d(c1,3AT)9 and d(c1,3A9T9) ­comprising 1,3-dideazaadenine (c1,3A) form duplexes consisting of doubly Ag+-mediated T–Ag2–c1,3A base pairs (Scheme 6c). This can be explained by the more basic exocyclic amino group of 1,3-dideazaadenine as compared to 1-deazaadenine [66], enabling it to bind an additional metal ion. The sequence d(c1,3AT)9 with alternating purine and pyrimidine nucleosides (Tm = 82°C) was found to adopt a more stable duplex structure than d(c1,3A9T9) (Tm = 62°C) in the presence of Ag+ [64]. The formation of stable metalated duplexes was further proven by dynamic light scattering revealing an increase in the hydrodynamic radius after the addition of Ag+ and by mass spectrometric experiments. The doubly metalated base pair was investigated by dispersion-corrected DFT calculations, confirming that it is energetically favored over the corresponding non- and singly-metalated base pairs [64]. In the geometry-optimized structure of T–Ag2–c1,3A, the two Ag+ ions are in close vicinity (2.88 Å), suggesting an argentophilic interaction [64]. This interaction contributes some 16 kcal mol–1 to the overall stability of the base pair. Its presence is supported by the fact that, upon excitation at 270 nm, the oligonucleotide duplexes exhibit fluorescence at ∼465 nm in the presence of Ag+[64].

The base pairs formed from 1-deazaadenine and 1,3-dideazaadenine, respectively, represent another example for the potential that lies in the modification of the base pairing properties by subtly changing the nucleobases.

4.2 6-Substituted Purine Derivatives

In analogy to the pyrimidinone-derived nucleobases (Section 3.3.2), the affinity of purine bases towards metal ions can be enhanced by the functionalization with a heterocyclic aryl moiety, too. For example, a modified adenine derivative PurPy (6-(2’-pyridyl)-purine) containing a 2-pyridyl residue instead of an amino group at the C6 position has been reported (Scheme 7a) [67]. In oligonucleotides containing one self-pair of this artificial nucleoside, only Ni2+ was shown to form a stable metal-mediated base pair (PurPy–Ni–PurPy), whereas the addition of other metal ions (Co2+, Cu2+, Zn2+, Ag+, Fe2+, Mn2+, Eu3+, Pd2+) does not result in a stabilization of the duplex. Moreover, the choice of a different nucleobase complementary to PurPy (e.g., G, A, C, T) does not result in a remarkable duplex stabilization [67]. Quantum chemical calculations revealed an almost perfect overlap of the glycosidic bond positions of the artificial PurPy–Ni–PurPy and the natural A:T base pair within an antiparallel Watson-Crick duplex [67].
Scheme 7

Schematic representation of (a) the PurPy–Ni–PurPy base pair [67], (b) the PurBipy–Ag–4Py base pair [68], and (c) the Pur2,6Py–Ag–3Py base pair [68] (R, R’ = DNA backbone).

The attachment of one bipyridyl residue or two pyridyl moieties to a purine ring leads to terpyridine-like binding motifs. Accordingly, oligonucleotides containing the purine-derived nucleobases Pur2,6Py and PurBipy (Scheme 7b, c) have been investigated [68]. In analogy to the bipyridine-functionalized pyrimidinone PyrBipy (Figure 6b), these nucleobases form stable Ag+-mediated base pairs with pyridine. Because of the different geometrical orientation of the terpyridine-like motif, Pur2,6Py preferentially forms metal-mediated base pairs with 3Py, whereas PurBipy prefers 4Py [68].

5 Processing of Metal-Mediated Base Pairs by Polymerases

As it is known that polymerases can recognize modified substrates and catalyze a replication reaction involving the formation of artificial base pairs, attempts have been made to directly incorporate metal-mediated base pairs into DNA duplexes by in vitro transcription in the presence of the appropriate metal ion. In the case of the T–Hg–T base pair, a primer-extension reaction using various DNA polymerases has been reported to take place, incorporating thymidine 5’-triphosphate (TTP) opposite to a thymine residue in the template strand in the presence of Hg2+[69]. It has been proposed that this finding suggests a potential mechanism for the mutagenicity of Hg2+ [69]. As a result of the highly specific incorporation of Hg2+ into the T–Hg–T base pair, the addition of other metal ions like Mn2+, Fe2+, Fe3+, Co2+, Cu2+, Zn2+, Pb2+, Ni2+, or Au+ does not lead to the incorporation of TTP [69]. In another example, Taq polymerase has been shown to be capable of bypassing T–Hg–T and C–Ag–C base pairs in the primer region [70]. Hence, the amplification products are formed only in the presence of the appropriate metal ion. By applying this strategy, molecular logic gates have been devised that are triggered by Ag+ and Hg2+, respectively [70]. Both studies indicate that metal-mediated base pairs with a geometry similar to that of the natural base pairs can be processed by polymerases, suggesting that an expansion of the genetic alphabet might be possible via a metal-mediated enzymatic incorporation of artificial nucleobases into nucleic acids [69].

6 Concluding Remarks and Outlook

Metal-mediated base pairs from natural nucleosides (or close derivatives thereof) have come a long way. Originally proposed to explain changes in the physical properties of natural DNA that was observed in the presence of certain transition metal ions, they are nowadays discussed and utilized in the context of sophisticated DNA-based applications such as metal ion sensors [71,72], sensors for the redox environment [41], molecular logic gates [70], the detection of single nucleotide polymorphisms [73], long-distance charge transfer [74], and DNA-based machines like bipedal walkers and steppers [75]. This wealth of possible applications will undoubtedly increase even further with the development of new metal-mediated base pairs including entirely artificial nucleosides [12].

Several studies have underlined the potential that lies in fine-tuning the metal-binding properties of nucleosides by replacing single atoms or entire groups of atoms not directly involved in the binding of the metal ions [54,64,76]. This approach will certainly play a major role in diversifying the range of nucleosides available for metal binding. Factors contributing to the stabilization of nucleic acid duplexes with metal-mediated base pairs, many of which display an increased positive charge along the helical axis, are only about to emerge. In this context, a non-covalent metallophilic attraction might play a significant role in stabilizing consecutive metal-mediated base pairs comprising metal ions with a closed-shell electronic structure [77]. The argentophilic interaction has already been shown to be a main contributor to the stability of doubly Ag+-mediated base pairs [64].

Finally, the formation of metal-mediated base pairs is not limited to DNA but has also been found for RNA [43] and the nucleic acid analogues GNA (glycol nucleic acid) [78] and PNA (peptide nucleic acid) [79]. Hence, the possibility to combine particular properties of a nucleic acid with the functionality introduced by a metal-mediated base pair will enable the access to supramolecules that are structurally similar yet display chemically and physically distinct properties.



cytosine, (deoxy)cytidine










circular dichroism


dynamic light scattering


diffusion-ordered spectroscopy


density functional theory




fluorescence resonance energy transfer


guanine, (deoxy)guanosine


glycol nucleic acid


heteronuclear single quantum coherence






nuclear Overhauser enhancement spectroscopy


peptide nucleic acid
















hydrodynamic radius


thymine, thymidine




thymidine 5’-triphosphate


uracil, uridine













Financial support of our research by the Deutsche Forschungsgemeinschaft is gratefully acknowledged.


  1. 1.
    J. Müller, Metallomics 2010, 2, 318–327.CrossRefPubMedGoogle Scholar
  2. 2.
    G. L. Eichhorn, Y. A. Shin, J. Am. Chem. Soc. 1968, 90, 7323–7328.CrossRefPubMedGoogle Scholar
  3. 3.
    G. L. Eichhorn, J. J. Butzow, P. Clark, E. Tarien, Biopolymers 1967, 5, 283–296.CrossRefPubMedGoogle Scholar
  4. 4.
    R. M. Izatt, J. J. Christensen, J. H. Rytting, Chem. Rev. 1971, 71, 439–481.CrossRefPubMedGoogle Scholar
  5. 5.
    R. H. Jensen, N. Davidson, Biopolymers 1966, 4, 17–32.CrossRefGoogle Scholar
  6. 6.
    S. Katz, Nature 1962, 194, 569.CrossRefGoogle Scholar
  7. 7.
    J. Müller, Eur. J. Inorg. Chem. 2008, 3749–3763.Google Scholar
  8. 8.
    G. H. Clever, C. Kaul, T. Carell, Angew. Chem. Int. Ed. 2007, 46, 6226–6236.CrossRefGoogle Scholar
  9. 9.
    K. Tanaka, M. Shionoya, Coord. Chem. Rev. 2007, 251, 2732–2742.CrossRefGoogle Scholar
  10. 10.
    S. Atwell, E. Meggers, G. Spraggon, P. G. Schultz, J. Am. Chem. Soc. 2001, 123, 12364–12367.CrossRefPubMedGoogle Scholar
  11. 11.
    S. Johannsen, N. Megger, D. Böhme, R. K. O. Sigel, J. Müller, Nat. Chem. 2010, 2, 229–234.CrossRefPubMedGoogle Scholar
  12. 12.
    G. H. Clever, M. Shionoya, Chapter 10 of this book.Google Scholar
  13. 13.
    J. S. Lee, L. J. P. Latimer, R. S. Reid, Biochem. Cell Biol. 1993, 71, 162–168.CrossRefPubMedGoogle Scholar
  14. 14.
    P. Aich, S. L. Labiuk, L. W. Tari, L. J. T. Delbaere, W. J. Roesler, K. J. Falk, R. P. Steer, J. S. Lee, J. Mol. Biol. 1999, 294, 477–485.CrossRefPubMedGoogle Scholar
  15. 15.
    B. Lippert, Prog. Inorg. Chem. 2005, 54, 385–447.CrossRefGoogle Scholar
  16. 16.
    M. Fuentes-Cabrera, B. G. Sumpter, J. E. Šponer, J. Šponer, L. Petit, J. C. Wells, J. Phys. Chem. B 2007, 111, 870–879.CrossRefPubMedGoogle Scholar
  17. 17.
    S. S. Alexandre, J. M. Soler, L. Seijo, F. Zamora, Phys. Rev. B 2006, 73, 205112.CrossRefGoogle Scholar
  18. 18.
    E. C. Fusch, B. Lippert, J. Am. Chem. Soc. 1994, 116, 7204–7209.CrossRefGoogle Scholar
  19. 19.
    S. L. Labiuk, L. T. J. Delbaere, J. S. Lee, J. Biol. Inorg. Chem. 2003, 8, 715–720.CrossRefPubMedGoogle Scholar
  20. 20.
    A. Rakitin, P. Aich, C. Papadopoulos, Y. Kobzar, A. S. Vedeneev, J. S. Lee, J. M. Xu, Phys. Rev. Lett. 2001, 86, 3670–3673.CrossRefPubMedGoogle Scholar
  21. 21.
    C.-Z. Li, Y.-T. Long, H.-B. Kraatz, J. S. Lee, J. Phys. Chem. B 2003, 107, 2291–2296.CrossRefGoogle Scholar
  22. 22.
    M. J. Dinsmore, J. S. Lee, J. Electroanal. Chem. 2008, 617, 71–77.CrossRefGoogle Scholar
  23. 23.
    X. Li, Y. Zhou, T. C. Sutherland, B. Baker, J. S. Lee, H.-B. Kraatz, Anal. Chem. 2005, 77, 5766–5769.CrossRefPubMedGoogle Scholar
  24. 24.
    S. D. Wettig, D. O. Wood, P. Aich, J. S. Lee, J. Inorg. Biochem. 2005, 99, 2093–2101.CrossRefPubMedGoogle Scholar
  25. 25.
    K. Mizoguchi, S. Tanaka, T. Ogawa, N. Shiobara, H. Sakamoto, Phys. Rev. B 2005, 72, 0033106.CrossRefGoogle Scholar
  26. 26.
    F. Moreno-Herrero, P. Herrero, F. Moreno, J. Colchero, C. Gómez-Navarro, J. Gómez-Herrero, A. M. Baró, Nanotechnology 2003, 14, 128–133.CrossRefGoogle Scholar
  27. 27.
    B. Liu, A. J. Bard, C.-Z. Li, H.-B. Kraatz, J. Phys. Chem. B 2005, 109, 5193–5198.CrossRefPubMedGoogle Scholar
  28. 28.
    B. Q. Spring, R. M. Clegg, J. Phys. Chem. B 2007, 111, 10040–10052.CrossRefPubMedGoogle Scholar
  29. 29.
    G. H. Clever, M. Shionoya, Coord. Chem. Rev. 2010, 254, 2391–2402.CrossRefGoogle Scholar
  30. 30.
    S. Katz, J. Am. Chem. Soc. 1952, 74, 2238–2245.CrossRefGoogle Scholar
  31. 31.
    C. A. Thomas, J. Am. Chem. Soc. 1954, 76, 6032–6034.CrossRefGoogle Scholar
  32. 32.
    S. Katz, Biochim. Biophys. Acta 1963, 68, 240–253.CrossRefPubMedGoogle Scholar
  33. 33.
    L. D. Kosturko, C. Folzer, R. F. Stewart, Biochemistry 1974, 13, 3949–3952.CrossRefPubMedGoogle Scholar
  34. 34.
    Z. Kuklenyik, L. G. Marzilli, Inorg. Chem. 1996, 35, 5654–5662.CrossRefPubMedGoogle Scholar
  35. 35.
    Y. Miyake, H. Togashi, M. Tashiro, H. Yamaguchi, S. Oda, M. Kudo, Y. Tanaka, Y. Kondo, R. Sawa, T. Fujimoto, T. Machinami, A. Ono, J. Am. Chem. Soc. 2006, 128, 2172–2173.CrossRefPubMedGoogle Scholar
  36. 36.
    H. Torigoe, A. Ono, T. Kozasa, Chem. Eur. J. 2010, 16, 13218–13225.CrossRefPubMedGoogle Scholar
  37. 37.
    Y. Tanaka, S. Oda, H. Yamaguchi, Y. Kondo, C. Kojima, A. Ono, J. Am. Chem. Soc. 2007, 129, 244–245.CrossRefPubMedGoogle Scholar
  38. 38.
    A. Ono, H. Togashi, Angew. Chem. Int. Ed. 2004, 43, 4300–4302.CrossRefGoogle Scholar
  39. 39.
    R.-M. Kong, X.-B. Zhang, L.-L. Zhang, X.-Y. Jin, S.-Y. Huan, G.-L. Shen, R.-Q. Yu, Chem. Commun. 2009, 5633–5635.Google Scholar
  40. 40.
    C.-X. Tang, Y. Zhao, X.-W. He, X.-B. Yin, Chem. Commun. 2010, 46, 9022–9024.CrossRefGoogle Scholar
  41. 41.
    Y. Miyake, A. Ono, Tetrahedron Lett. 2005, 46, 2441–2443.CrossRefGoogle Scholar
  42. 42.
    E. Ennifar, P. Walter, P. Dumas, Nucleic Acids Res. 2003, 31, 2671–2682.CrossRefPubMedGoogle Scholar
  43. 43.
    S. Johannsen, S. Paulus, N. Düpre, J. Müller, R. K. O. Sigel, J. Inorg. Biochem. 2008, 102, 1141–1151.CrossRefPubMedGoogle Scholar
  44. 44.
    T. Yamane, N. Davidson, Biochim. Biophys. Acta 1962, 55, 609–621.CrossRefPubMedGoogle Scholar
  45. 45.
    L. G. Marzilli, T. J. Kistenmacher, M. Rossi, J. Am. Chem. Soc. 1977, 99, 2797–2798.CrossRefPubMedGoogle Scholar
  46. 46.
    T. J. Kistenmacher, M. Rossi, L. G. Marzilli, Inorg. Chem. 1979, 18, 240–244.CrossRefGoogle Scholar
  47. 47.
    A. Ono, S. Cao, H. Togashi, M. Tashiro, T. Fujimoto, T. Machinami, S. Oda, Y. Miyake, I. Okamoto, Y. Tanaka, Chem. Commun. 2008, 4825–4827.Google Scholar
  48. 48.
    H. Torigoe, Y. Miyakawa, A. Ono, T. Kozasa, Nucleosides Nucleotides Nucleic Acids 2011, 30, 149–167.CrossRefPubMedGoogle Scholar
  49. 49.
    D. A. Megger, J. Müller, Nucleosides Nucleotides Nucleic Acids 2010, 29, 27–38.CrossRefPubMedGoogle Scholar
  50. 50.
    H. Urata, E. Yamaguchi, Y. Nakamura, S.-i. Wada, Chem. Commun. 2011, 47, 941–943.CrossRefGoogle Scholar
  51. 51.
    T. Ono, K. Yoshida, Y. Saotome, R. Sakabe, I. Okamoto, A. Ono, Chem. Commun. 2011, 47, 1542–1544.CrossRefGoogle Scholar
  52. 52.
    D. A. Megger, C. Fonseca Guerra, F. M. Bickelhaupt, J. Müller, J. Inorg. Biochem. 2011, 105, 1398–1404.Google Scholar
  53. 53.
    Y. Wen, F. Xing, S. He, S. Song, L. Wang, Y. Long, D. Li, C. Fan, Chem. Commun. 2010, 46, 2596–2598.CrossRefGoogle Scholar
  54. 54.
    I. Okamoto, K. Iwamoto, Y. Watanabe, Y. Miyake, A. Ono, Angew. Chem. Int. Ed. 2009, 48, 1648–1651.CrossRefGoogle Scholar
  55. 55.
    K. Aoki, W. Saenger, Acta Cryst. 1984, C40, 775–778.Google Scholar
  56. 56.
    B. Lippert, Coord. Chem. Rev. 2000, 200–202, 487–516.CrossRefGoogle Scholar
  57. 57.
    J. Ruiz, M. D. Villa, V. Rodríguez, N. Cutillas, C. Vicente, G. López, D. Bautista, Inorg. Chem. 2007, 46, 5448–5449.CrossRefPubMedGoogle Scholar
  58. 58.
    C. Switzer, D. Shin, Chem. Commun. 2005, 1342–1344.Google Scholar
  59. 59.
    D. Shin, C. Switzer, Chem. Commun. 2007, 4401–4403.Google Scholar
  60. 60.
    F.-A. Polonius, J. Müller, Angew. Chem. Int. Ed. 2007, 46, 5602–5604.CrossRefGoogle Scholar
  61. 61.
    T. Ihara, T. Ishii, N. Araki, A. W. Wilson, A. Jyo, J. Am. Chem. Soc. 2009, 131, 3826–3827.CrossRefPubMedGoogle Scholar
  62. 62.
    M. G. Santangelo, P. M. Antoni, B. Spingler, G. Jeschke, ChemPhysChem 2010, 11, 599–606.CrossRefPubMedGoogle Scholar
  63. 63.
    F. Seela, T. Wenzel, Helv. Chim. Acta 1994, 77, 1485–1499.CrossRefGoogle Scholar
  64. 64.
    D. A. Megger, C. Fonseca Guerra, J. Hoffmann, B. Brutschy, F. M. Bickelhaupt, J. Müller, Chem. Eur. J. 2011, 17, 6533–6544.Google Scholar
  65. 65.
    F. Seela, T. Wenzel, Helv. Chim. Acta 1995, 78, 833–846.CrossRefGoogle Scholar
  66. 66.
    L. E. Kapinos, A. Holý, J. Günter, H. Sigel, Inorg. Chem. 2001, 40, 2500–2508.CrossRefPubMedGoogle Scholar
  67. 67.
    C. Switzer, S. Sinha, P. H. Kim, B. D. Heuberger, Angew. Chem. Int. Ed. 2005, 44, 1529–1532.CrossRefGoogle Scholar
  68. 68.
    B. D. Heuberger, D. Shin, C. Switzer, Org. Lett. 2008, 10, 1091–1094.CrossRefPubMedGoogle Scholar
  69. 69.
    H. Urata, E. Yamaguchi, T. Funai, Y. Matsumura, S.-i. Wada, Angew. Chem. Int. Ed. 2010, 49, 6516–6519.CrossRefGoogle Scholar
  70. 70.
    K. S. Park, C. Jung, H. G. Park, Angew. Chem. Int. Ed. 2010, 49, 9757–9760.CrossRefGoogle Scholar
  71. 71.
    R. Freeman, T. Finder, I. Willner, Angew. Chem. Int. Ed. 2009, 48, 7818–7821.CrossRefGoogle Scholar
  72. 72.
    N. Kanayama, T. Takarada, M. Maeda, Chem. Commun. 2011, 47, 2077–2079.CrossRefGoogle Scholar
  73. 73.
    H. Torigoe, T. Kozasa, A. Ono, Nucleic Acids Symp. Ser. 2006, 50, 89–90.CrossRefGoogle Scholar
  74. 74.
    J. Joseph, G. B. Schuster, Org. Lett. 2007, 9, 1843–1846.CrossRefPubMedGoogle Scholar
  75. 75.
    Z.-G. Wang, J. Elbaz, I. Willner, Nano Lett. 2011, 11, 304–309.CrossRefPubMedGoogle Scholar
  76. 76.
    J. Müller, D. Böhme, P. Lax, M. Morell Cerdà, M. Roitzsch, Chem. Eur. J. 2005, 11, 6246–6253.CrossRefPubMedGoogle Scholar
  77. 77.
    L. Benda, M. Straka, Y. Tanaka, V. Sychrovský, Phys. Chem. Chem. Phys. 2011, 13, 100–103.CrossRefPubMedGoogle Scholar
  78. 78.
    M. K. Schlegel, L. Zhang, N. Pagano, E. Meggers, Org. Biomol. Chem. 2009, 7, 476–482.CrossRefPubMedGoogle Scholar
  79. 79.
    R. M. Franzini, R. M. Watson, G. K. Patra, R. M. Breece, D. L. Tierney, M. P. Hendrich, C. Achim, Inorg. Chem. 2006, 45, 9798–9811.CrossRefPubMedGoogle Scholar

Copyright information

© Springer Science+Business Media B.V. 2012

Authors and Affiliations

  • Dominik A. Megger
    • 1
  • Nicole Megger
    • 1
  • Jens Müller
    • 1
  1. 1.Institute for Inorganic and Analytical ChemistryUniversity of MünsterMünsterGermany

Personalised recommendations