2025 年 73 巻 2 号 p. 103-107
Here, a DNA cleavage reagent (1-(anthracen-9-ylmethyl)-1,5,9-triazacyclododecane = Ant-[12]aneN3) was designed and synthesized, and its DNA photocleavage activity under UV irradiation at λ = 365 nm was evaluated. Ant-[12]aneN3 is a molecule containing anthracene as the photosensitizer and [12]aneN3 ( = 1,5,9-triazacyclododecane) as the DNA-interacting component. The cyclic polyamine [12]aneN3 could coordinate with zinc ions (ZnII) and affect DNA cleavage activity. Therefore, when Ant-[12]aneN3 reacted with Zn(NO3)‧6H2O, the product was not a ZnII complex but an N-protonated form of Ant-[12]aneN3. In DNA cleavage experiments with the pUC19 plasmid, Ant-[12]aneN3 also showed DNA photocleavage activity in a ZnII-independent manner. That is, [12]aneN3 enhances the DNA photocleavage activity of anthracene in a ZnII-independent manner, unlike bpa (bis(2-picolyl)amine), which was previously reported to enhance DNA cleavage activity by chelating ZnII. Under physiological conditions, the nitrogen atoms of [12]aneN3 appear protonated without the addition of ZnII salts and showed an affinity for the negatively charged DNA. The results of this study may facilitate the design of effective DNA cleavage reagents.
Polyamines, small basic molecules of aliphatic hydrocarbons, have attracted considerable attention because of their association with cell growth,1) differentiation,2) regeneration,3) and malignant transformation.4) The biological activity of polyamines is often explained using their interactions with acidic molecules in cells.5,6) Among others, polyamine interactions with DNA and RNA have influenced the design of various bioengineering reagents.7–9)
In this study, we designed DNA photocleavants with cyclic polyamine structures. In our previous study, we demonstrated the possibility of sequence-specific DNA cleavage when anthracene is attached to the metal-ligand structure.10) Anthracene moieties function as photosensitizers and can control biological activity.11) In addition, controlling ligand chirality may prove advantageous for DNA cleavage.12) The bpa [bis-(2-picolyl)amine], used in metal coordination, showed enhanced DNA photocleavage activity when coordinated with zinc ions (ZnII).13) Although bpa is a small molecule containing 3 nitrogen atoms, it is not exactly a polyamine. However, we focused on the structural similarities between bpa and polyamines, which are both nitrogen-containing structures. In other words, it was hypothesized that converting bpa into polyamines could improve the DNA affinity of previously developed anthracene–bpa derivatives, potentially leading to highly potent DNA photocleavage reagents. Because polyamines have multiple nitrogen atoms, they can coordinate transition metal ions, such as ZnII, depending on the reaction conditions.14–16) However, the effect of ZnII on the DNA affinity of polyamines remains unknown. Therefore, in the molecular design of this study, we used cyclic polyamines, which are advantageous for coordinating ZnII, to investigate the effect of ZnII. We selected 1,5,9-triazacyclododecane ([12]aneN3) as the cyclic polyamine. The [12]aneN3 is the same tridentate ligand as bpa and has similar binding constants to ZnII: the ZnII binding constant of [12]aneN3 (log K = 8.417)) slightly exceeds that of bpa (log K = 6.818)). For these reasons, we designed and synthesized 1-(anthracen-9-ylmethyl)-1,5,9-triazacyclododecane (Ant-[12]aneN3) and inves-tigated its DNA photocleavage activity.
We synthesized a photosensitive cyclic polyamine: Ant-[12]aneN3 (1). Notably, [12]aneN3 has three secondary amines that can introduce an anthracene moiety. To limit the reaction of the amines, two synthesis strategies were investigated with reference to the synthesis of [12]aneN3 derivatives reported by Choi et al.19) and Long et al.20) The reaction conditions for each step are shown in Chart 1.
Reagents and conditions (a) Diethyl malonate, NaOEt, EtOH, reflux, 7 d. (b) 9-Anthracenylmethyl chloride, triethylamine, CHCl3, reflux, overnight. (c) 1) BH3–THF, THF, reflux, 18 h. 2) HCl, reflux, 3 h, disassembled. (d) 1,3-Bis(tosyloxy)propane, NaBH4, Toluene, r.t., 3 d. (e) 0.75 mol/L NaOH aq. (containing 33 % (v/v) EtOH), reflux, 24 h.
Here, 9-(anthracen-9-ylmethyl)-1,5,9-triazacyclododecane-2,4-dione (Ant-2,4-dioxo-[12]aneN3) (3) was successfully synthesized via route 1 by introducing the Ant moiety into 1,5,9-triazacyclododecane-2,4-dione19) (2,4-dioxo-[12]aneN3) (2). However, the preparation of Ant-[12]aneN3 (1) by the reduction reaction of Ant-2,4-dioxo-[12]aneN3 (3) was unsuccessful. We also tried to reduce 3 with LiAlH4 under different reaction conditions. However, the synthesis of 1 by reduction of 3 could not be achieved. Thus, we abandoned route 1 because the amide reduction did not proceed via this route. The starting material used in route 2 was 1,5,7-triazabicyclo-[4.4.0]dec-5-ene (TBD).21) In the synthesis of 1, the conversion of TBD into its tricyclic intermediate, 1,5,9-triazatricyclo-[7.3.1.05.13]tridecane (4), limited the nitrogen atoms with which 9-anthracenylmethyl chloride could react. Compound 5 was converted into 1 by the addition of 0.75 mol/L NaOH in a mixture of H2O and EtOH (2 : 1) with prolonged reflux. The crude product of 1 was purified via amino-functionalized silica gel chromatography.
Complex Formation StoichiometryThe ZnII chelating properties of Ant-[12]aneN3 (1) were investigated by 1H-NMR titrations using methanol-d4. With the addition of increasing amounts of zinc nitrate (Zn(NO3)2), most of the protons of 1 exhibited a gradual downfield shift (Fig. 1B). The chemical shift changes (Δδ, Hz) of the Ha protons (Fig. 1A) corresponding to the Zn(NO3)2 (0.05–1 equivalent (equiv.)) were plotted on the vertical axis, and the chemical equivalent of the Zn(NO3)2 was plotted on the horizontal axis (Fig. 1C). The selected proton (Ha) peaks are referenced because 1) they were separated from other peaks, including solvent-derived peaks, and 2) the Δδ was highest when Zn(NO3)2 was added. The change in chemical shift did not require specific reaction conditions, such as heating, and was stable for more than 24 h after preparation. The Δδ values did not change with the addition of Zn(NO3)2 above 0.45 equiv. This suggested that the reaction between the ligand and ZnII converged at a 2 : 1 stoichiometry. An increase in the amount of Zn(NO3)2 to 3.0 equiv. relative to the ligand did not change the chemical shift any more than when 0.5 equiv. was added (data not shown). No new significant peaks were observed when excess Zn(NO3)2 was added. Previously, Kimura et al. reported when ZnII was chelated with [12]aneN3, the ZnII was located at the apical position of the triangular pyramid formed by the 3 nitrogen atoms of the polyamine and ZnII.22) The crystal structure of the metal ion complexed with [18]aneN6 ( = 1,4,7,10,13,16-hexaazacyclooctadecane), a macrocyclic polyamine with a hexadentate ligand, has been reported at a 1:1 stoichiometry.23) Thus, the complex formation by 2 ligand molecules with ZnII as the central metal is inferred (putative structure II, Chart 2).
1H-NMR titration of Ant-[12]aneN3 (final 2.0 mmol/L) in methanol-d4 with different amounts of zinc nitrate hexahydrate.
To determine the metal chelating structure, several analyses were performed on the solid (1a) precipitate from the reaction of 1 with Zn(NO3)2 at a 1 : 1 stoichiometry. Flame atomic absorption spectrometry (FAAS) revealed that the Zn content in the 1 g/L of 1a dissolved in nitric acid was less than 0.1 mg/L (low limit of detection). For putative structure II, the calculated zinc content in the 1 g/L of 1a solution was 72 mg/L; thus, the FAAS results suggested the absence of zinc atoms in 1a. The results of the CHN elemental analysis (accuracy = ±0.3%) of 1a were as follows: C 66.37, H 7.66, N 12.78. The molecular formula for 1a predicted from the results was [1 + H]+(NO3–)‧(H2O)0.5: Calcd C 66.46, H 7.67, N 12.92. Considering that Zn(NO3)2 can provide two molecules of nitrate ions, the reaction stoichiometry of the ligand to that of Zn(NO3)2 suggested in Fig. 1C should be reasonable. The 1H-NMR (dimethyl sulfoxide (DMSO)-d6) analysis of 1a revealed a broad peak at 7.26 ppm, which disappeared upon the addition of D2O (Supplementary Fig. S1). The broad peak exhibited an integrated intensity of about 3H and was not observed in Ant-[12]aneN3 only. The NMR results support the structure prediction in 1a, in which a secondary nitrogen atom of 1 is protonated (putative structure III, Chart 2). The X-ray crystallographic analysis of the crystals, albeit at high R values (R1 = 0.3181, wR2 = 0.6618), obtained by the recrystallization of 1a supported the present structure prediction (Supplementary Fig. S2). Based on the above-mentioned analysis, Ant-[12]aneN3 (1) reacted with Zn(NO3)2 (0.5 equiv.) in MeOH and was protonated to yield a nitrate salt of probable structure III (Chart 2).
DNA Cleavage ActivityThe DNA cleavage activity was evaluated using pUC19 plasmid DNA. The DNA cleavage activities of the following compounds were compared: i.e., compounds 1, 3, 6–8 (Fig. 2A). Compound 6 is a bpa-substituted 9-methylanthracene derivative that has previously shown DNA photocleavage activity in the presence of ZnII.13) Compounds 7 and 8 are substructures of Ant-[12]aneN3. Compounds 6 and 8 were prepared according to a previous study.13) Instead of bpa, diethylamine (DEA) was used as the starting material for the synthesis of compound 8. Compound 7 was synthesized according to the method by Choi et al.19)
Reactants: pUC19 plasmid DNA strand (0.48 μg), zinc nitrate, final 200 μM, and ligands (L1, L2, and L3) (final 200 μM). Reaction conditions: The reactants were dissolved in 50 mM Tris-borate buffer (pH 8.4) at 20°C under atmospheric conditions. For photoreactive DNA cleavage, the light (wavelength = 365 nm) was continuously irradiated in microtubes using the SLUV handy UV lamp within 10 min.
First, Compound 1 (Ant-[12]aneN3) showed DNA cleavage activity within 10 min of UV irradiation (λ = 365 nm) in the presence of Zn(NO3)2 (lane 2). Ant-[12]aneN3 did not show DNA cleavage activity after 10 min of incubation in the dark in the presence of Zn(NO3)2 (lane 1). However, Ant-[12]aneN3 exhibited DNA cleavage activity when exposed to UV in the absence of Zn(NO3)2 (lane 3). When neither compound was added, Zn(NO3)2 showed no DNA photocleavage activity (lane 4). Similar to Ant-[12]aneN3, compound 3 showed no DNA cleavage activity in the dark (lane 5). Conversely, whether Zn(NO3)2 was added (lane 6) or not (lane 7), the DNA cleavage activities of compound 3 were weaker than those of Ant-[12]aneN3 when exposed to UV. Compound 6 (Ant-bpa), known for its ZnII-enhanced DNA photocleavage activity, showed DNA cleavage activity comparable to that of Ant-[12]aneN3 under identical conditions when irradiated with UV in the presence of Zn(NO3)2 (Fig. 2B lane 8, Supplementary Fig. S3). Compound 7 ([12]aneN3), the cyclic polyamine moiety of Ant-[12]aneN3, showed no DNA photocleavage activity in the presence of Zn(NO3)2 (lane 9). Compound 8 (Ant-DEA), an N-(9-anthracenyl)methyl derivative of DEA, contains a photosensitive moiety substituted on the backbone but only one nitrogen atom. Compound 8 displayed weaker DNA photocleavage activity than Ant-[12]aneN3 and Ant-bpa under the same conditions, yet its DNA photocleavage activity was comparable to that of compound 3 (lanes 6 and 10). Next, the UV irradiation time for DNA cleavage was investigated using Ant-[12]aneN3 as a ligand (Fig. 2C). Although UV was irradiated for 1, 2, 5, and 10 min, a faint form I band was present after 5 min irradiation and no such band was observed after 10 min irradiation (lanes 11–14). The UV irradiation time dependence of DNA cleavage was not affected by the presence or absence of Zn(NO3)2 (lanes 15–18).
The results indicate that Ant-[12]aneN3 is as potent as compound 6 and more potent than compound 3 in terms of DNA photocleavage activity (Fig. 2B). Interestingly, the DNA photocleavage activity of Ant-[12]aneN3 and compound 3 was not enhanced by the addition of Zn(NO3)2. At present, we believe that the [12]aneN3 moiety of Ant-[12]aneN3 does not coordinate ZnII. Therefore, the DNA photocleavage activity is not affected by the addition of Zn(NO3)2, and it is interesting that Ant-[12]aneN3 exhibits DNA photocleavage activity without ZnII. It can be concluded that the presence of ZnII does not affect the DNA photocleavage activity of Ant-[12]aneN3. Linear polyamines, such as spermidine and spermine, are known to bind DNA.24) For example, the 3 pKa values of spermidine are 8.25, 9.71, and 10.90, indicating that its amines are protonated under physiological conditions.25) That is, the amines have an affinity for the phosphate moiety of DNA. It is possible that [12]aneN3, a saturated amine like spermidine, also protonates in Tris-borate buffer (pH 8.4), contributing to the increased affinity for DNA in Ant-[12]aneN3. The prediction that protonated amines increase affinity for DNA strands also explains why the DNA cleavage activity of compound 3 is weaker than that of Ant-[12]aneN3. Compound 3 has a structure where both secondary amines of Ant-[12]aneN3 are converted to amides, leaving the amide nitrogen unprotonated. However, the weak ZnII-independent DNA photocleavage activity of compound 3, similar to that of Ant-[12]aneN3, may be influenced by N-protonation. In other words, protonation of the tertiary amine substituted by the 9-(anthracenyl)methyl group is expected to increase the DNA affinity. This assumption is supported by the fact that the DNA photocleavage activity of compound 8 was comparable to that of compound 3 (lanes 6 and 10). In conclusion, the cyclic polyamine [12]aneN3 is expected to complement the DNA selectivity of anthracene and enhance its efficacy as a DNA photocleavage reagent in a ZnII-independent manner, unlike bpa.
Here, Ant-[12]aneN3 was designed and synthesized for the development of a DNA photocleavage reagent. In Ant-[12]aneN3, the cyclic triamine [12]aneN3 moiety thought that it could influence biological activity by coordinating ZnII. However, the reactant of Ant-[12]aneN3 (1) with Zn(NO3)‧6H2O was not a ZnII complex, but the N-protonated form ([1 + H]+) of Ant-[12]aneN3. The [12]aneN3 without anthracene substitution forms complexes with ZnII.22) Thus, the introduction of anthracene reduced the ZnII coordination ability of [12]aneN3. We have recently prepared ZnII complexes of Ant-[12]aneN4, an analog of Ant-[12]aneN3, and determined their X-ray crystal structures.26) The [12]aneN4 is a tetradentate cyclic polyamine, 1,4,7,10-tetraazacyclododecane (cyclen).27) This result with Ant-[12]aneN4 refutes the notion that anthracene completely impairs the ZnII coordination ability of cyclic polyamines. Rather, the loss of ZnII coordination potential of [12]aneN3 in Ant-[12]aneN3 may be rare.
Ant-[12]aneN3 exhibited DNA photocleavage activity in a ZnII-independent manner. Under physiological conditions, the nitrogen atoms of [12]aneN3 are protonated without the addition of ZnII salts, which have an affinity for DNA and may enhance the DNA photocleavage activity of anthracene.
JSPS KAKENHI Grant JP23K14339 to Y.I. is acknowledged.
The authors declare no conflict of interest.
This article contains supplementary materials.
