Dinuclear Zinc (II) Complexes of Macrocyclic Polyamine Ligands Containing an Imidazolium Bridge: Synthesis, Characterization, and Their Interaction with Plasmid DNA

Two novel macrocyclic polyamine ligands and their dinuclear zinc (II) complexes were synthesized and characterized. Their interaction with plasmid DNA was studied by gel electrophoresis and fluorescence quenching experiment. The result showed that these complexes could bind DNA efficiently under physiological conditions.


Introduction
Genetic engineering has brought new challenges and opportunities for medicine and biomedical research, whereas DNA strands would be damaged in cellular environment [1][2][3]. The damage of DNA would cause mutations and genomic instabilities that may contribute to a variety of human genetic diseases. Thus DNA protection may play a significant role in bioanalysis and delivery. Though a few methods have been taken to protect DNA [4][5][6][7][8], there were very few reports on small molecules that could protect DNA. He et al. found out that protonation of the amino-modified silica nanoparticles could protect DNA due to the ability to enrich the negatively charged DNA strands by the positive charges on these materials [4]. Imidazolium ionic liquids with cations were thus expected to bind DNA sequences. Additionally, ionic liquids applied in biological chemistry were studied recently [9,10]. Ohno and colleagues prepared ionic liquid-robed DNAs by using the cations from imidazole ring to fix the phosphate group of DNA [11,12].
Herein, two imidazolium ionic liquids containing dinuclear cyclen moiety and their zinc (II) complexes were first synthesized and characterized. Their interaction with pUC19 plasmid DNA was studied. The results revealed that these complexes binding with DNA showed different characteristics from previous results in our group [21][22][23] due to the cation from imidazolium moiety.

Preparation of dinuclear zinc (II) complexes
The synthetic route of two zinc (II) complexes with imidazolium salt bridge was shown in Scheme 1. (Boc) 2 O could selectively protect cyclen to obtain tri-boc-protected cyclen. Compound 3 was obtained by the reaction between excessive p-or m-bis(bromomethyl) benzene and tri-bocprotected cyclen. Desired product 5 was obtained through two steps from imidazole and 3 by alkylation and formation of imidazolum cation. The Boc-protective groups were removed by adding dropwise the solution of trifluoroacetic acid (TFA) in dichloromethane. The target zinc (II) complexes with imidazolium salt groups were prepared from 6 with Zn(ClO 4 ) 2 in ethanol solution. 1a and 1b were characterized by elemental analysis.

Interaction between ligands or their zinc (II) complexes and plasmid pUC19 DNA
The interaction of macrocyclic polyamine metallic complexes with DNA was reported previously [21][22][23]. These complexes cleaved supercoiled plasmid DNA to produce open-circular form. Nevertheless the interaction of imidazolium ligands 6a-b and their zinc (II) complexes 1 with DNA was different from the reported models. The experimental results were shown in Figure 1 Agarose gel electrophoresis demonstrated that ligands 6 and complexes 1 could bind with DNA. As shown in Figure 1, plasmid DNA control moved in the electric field (Lane 1), and DNA-1a complexes were retained around the sample well (Lanes 13 and 14). The reason why the DNA-1a complexes did not move toward the positive electrode lied in their charge and the large size [24]. On the other hand, lanes 15-18 showed that with the concentration of 1a decreased, plasmid DNA moved toward the positive electrode. The interaction of 1b with DNA showed similar phenomena as 1a due to their analogous molecule structure (Lanes 19-24).    Obviously zinc(II) complexes were more reactive than the free ligands. This might be due to that the cation density increased with the existence of zinc(II), it was more available for complexes to bind DNA with anion.
To investigate whether the macrocyclic polyamine imidazolium salt ligands and their zinc (II) complexes were efficient reagents in protecting DNA from being cleaved enzymatically, 0.1 U of DNaseI was added respectively to the plasmid DNA and the DNA complexes containing 7 µg/ml of DNA. The results were shown in Figure 3. DNaseI could cleave DNA completely (Lane 2), which was used as control. It is obvious that both ligands 6 and complexes 1 could retard the cleavage of DNA efficiently (Lanes 3-6). It was illustrated that DNA could be protected by those compounds to avoid being cleaved. There might be a few possible reasons for the DNA protection based on imidazolium salt. The first one is that the positive charge on the imidazolium group kept Mg 2+ away from the positively charged imidazolium salt. This would retard the enzymatic cleavage process, in which Mg 2+ is needed. The second one is that DNA surface binding with imidazolium salt resulted in a variation of the DNA structure due to the size effect [4]. A typical imidazoluim 1-ethyl-3-methylimidazolium bromide (EMI) can not bind DNA, as shown in Figure 4. Clearly, macropolyamine moiety takes a significant role when ligands or complexes interact with DNA. Polyamine could be protonated then possesses more cation density to bind DNA with anion, maybe so ligands 6 and complexes 1 are more active than imidazolium without macropolyamine moiety.

Fluorescence quenching experiment
Fluorescence quenching experiment was performed in Tris-HCl 100 mM, pH 7.4 at room temperature to measure the binding affinities of compounds (6 and 1) with DNA. The extent of binding between the complexes and DNA could be determined by the fluorescence quenching of ethidium bromide intercalated to DNA.
According to the equation % 100 1    Figure 5 and Figure 6 showed that the fluorescence intensity decreased obviously associated with the increase of the concentration of complexes 1, which illustrated that complexes 1 could displace ethidium bromide effectively. Meanwhile, ligands 6 quenched the system fluorescence intensity with less efficiency comparing to 1, which showed that ethidium bromide is displaced partly by ligands 6. C 50 value describes the concentration of certain compound when this compound could cause a 50 % decrease (the value of r in Figure 5 and Figure 6 decrease to 50 %) in fluorescence intensity, and this value could show DNA-binding activity of certain compound. According to Figure 5 and Figure 6, the C 50 values were the concentration 60, 60, 4, 3 M respectively. These results indicated that the free ligands were less active in the DNA-binding process than those of the metal complexes. The results were in accord with those of gel electrophoresis experiments.

Conclusions
In this paper, imidazolium ionic liquids containing dinuclear cyclen moiety and their zinc (II) complexes using an m-or p-xylyl linkage were synthesized and characterized. Their interaction with DNA was detected by the method of agarose gel electrophoresis and fluorescence quenching. The results showed that DNA could be protected efficiently by the ligands and their zinc (II) complexes. Moreover, zinc (II) complexes were much more active than the free ligands. Imidazolium containing dinuclear cyclen moiety show much more interaction active with DNA than imidazolium without cyclen moiety. It will be useful in DNA separation, purification, and detection, and possibly in genetic engineering and gene therapy. Further studies are in progress.

General procedure for the synthesis of TFA salts of compounds 6a-b.
Trifluoroacetic acid (1.2 mmol) was added dropwise to a solution of 5 (0.1 mmol) in CH 2 Cl 2 (10 ml) at 0 ºC under N 2 atmosphere. The whole mixture was stirred for 4 h. Then the reaction mixture was concentrated under reduced pressure. The remaining yellow oil liquid was washed three times with CH 2 Cl 2 (5 ml) to obtain 6·TFA.

General procedure for the synthesis of complexes 1a-b
The trifluoroacetic acid salts of ligands 6a-b (0.1 mmol) were dissolved, respectively in the 5 ml of ethanol and adjusted the aqueous solution to alkaline (pH > 2) with 50 % aqueous NaOH. The solutions were extracted with CH 2 Cl 2 (4×15 ml). The combined organic layer was dried overnight by anhydrous Na 2 SO 4 and the solutions were concentrated to obtain a white oils 6a-b. To the ethanol solutions (5 ml) of 6a-b, equimolar amount of salts Zn(ClO 4 ) 2 in 5 ml of ethanol were added and the mixture were stirred at room temperature overnight. After filtration, the solids were washed with ethanol (2×5 ml), recrystallized from ethanol/H 2 O (3 : 1), and dried in vacuum to give pure zinc complexes.

Interaction between ligands or their zinc (II) complexes and plasmid pUC19 DNA
Interaction between ligands or their zinc (II) complexes with plasmid pUC19 DNA was monitored by agarose gel electrophoresis. In a typical experiment, supercoiled pUC19 DNA (10 l, 0.025 g/l) in Tris-HCl (100 mM, pH 7.4) was treated with different concentration of ligands 6 or their zinc (II) complexes 1, followed by dilution with the Tris-HCl buffer to a total volume of 35 µl. The samples were then incubated at 37 ºC for 1 h, and loaded on a 1 % agarose gel containing 1.0 g/ml ethidium bromide. Electrophoresis was carried out at 40 V for 30 min in TAE buffer. Bands were visualized by UV light and photographed followed by the estimation of the intensity of the DNA bands using a Gel Documentation System.

Fluorescence quenching experiments
All experiments were performed at room temperature in buffered aqueous solution (Tris-HCl 100 mM, pH 7.4). CT (calf thymus) DNA solution with optical density more than 1.8 at 260 nm, the concentration is 1 g/ml. 10 µl DNA solution and 80 µl ethidium bromide with the concentration of 1 g/ml was added, followed by different volume of 0.5 mM complexes (1a, 1b) or 5 mM ligands (6a, 6b) solution, then buffer was added and adjusted the whole volume to 2.5 ml. After reacting for 0.5 h, the fluorescence intensities were measured by fluorescence spectrophotometer.