Sensitive Hg2+ Sensing via Quenching the Fluorescence of the Complex between Polythymine and 5,10,15,20-tetrakis(N-methyl-4-pyridyl) Porphyrin (TMPyP)

The interaction between polythymine (dTn) and 5,10,15,20-tetrakis(N-methyl-4-pyridyl) porphyrin (TMPyP) was systematically studied using various techniques. dTn remarkably enhanced the fluorescence intensity of TMPyP as compared to other oligonucleotides. The enhanced fluorescence intensity and the shift of the emission peaks were ascribed to the formation of a π-π complex between TMPyP and dTn. And the quenching of the dTn-enhanced fluorescence by Hg2+ through a synergistic effect occurs due to the heavy atom effect. The binding of Hg2+ to TMPyP plays an important role in the Hg-TMPyP-dT30 ternary complex formation. A TMPyP-dT30-based Hg2+ sensor was developed with a dynamic range of Hg2+ from 5 nM to 100 nM. The detection limit of 1.3 nM was low enough for Hg2+ determination. The sensor also exhibited good selectivity against other metal ions. Experiments for tap water and river water demonstrated that the detection method was applicable for Hg2+ determination in real samples. The Hg2+ sensor based on oligonucleotide dT30-enhanced TMPyP fluorescence was fast and low-cost, presenting a promising platform for practical Hg2+ determination.


Introduction
As one of the most important conjugated organic molecules, porphyrins play a crucial role in the metabolism of living organisms [1]. For example, porphyrins are responsible for the production of singlet oxygen, which could damage DNA in tumor cells [2]. The interaction between porphyrins or their derivatives and nucleic acids has been extensively investigated [3][4][5], in which three binding types including intercalation, external or groove binding, and outside stacking were involved [6]. 5,10,15,20-tetrakis(N-methyl-4-pyridyl)porphyrin (TMPyP) is a water-soluble cationic porphyrin, which contains a porphyrin core and N-methylpyridinium side chains [7]. Due to their planarity and hydrophobicity, the porphyrin rings can intercalate into the base pairs of DNA and in the same time, the positively charged side chains can electrostatically interact with the negatively charged nucleic acids [8]. For example, evidenced by nuclear magnetic resonance spectroscopy, TMPyP can intercalate into GC-rich regions of duplex DNA [9,10]. However, the binding of TMPyP to the major groove of the AT-rich regions has also been proposed [11]. TMPyP has been reported to interact with triplex DNA, in which the third strand could inhibit the assembly of TMPyP with duplex DNA [12]. TMPyP can also bind to the G-quadruplex structure through external stacking rather than intercalation between

Procedure
The TMPyP-ODN complex was formed by mixing 1 µM TMPyP and 2 µM ODN, followed by dilution to 200 µL with Tris-HCl buffer and stirring for 50 min. For the Hg 2+ assay, Hg 2+ with various concentrations was added into the solution of TMPyP-dT 30 complex under stirring for 1 min. For the fluorescence, measurement with excitation was 422 nm and emission was at 660 nm. The fluorescence lifetimes were determined from the data obtained from a compact FluoTime 100 fluorescence lifetime spectrometer. The data could be analyzed by exponential fits. The solution of Hg 2+ , TMPyP and dT 30 were mixed by 1 µM Hg 2+ and 1 µM TMPyP for Hg-TMPyP; 1 µM TMPyP and 2 µM dT 30 for TMPyP-dT 30 ; 1 µM Hg 2+ , 1 µM TMPyP and 2 µM dT 30 for Hg-TMPyP-dT 30. The solutions were followed by dilution to 200 µL with Tris-HCl buffer and stirring for 120 min. All the lifetime measurements were with an excitation at 422 nm.

Interaction between ODNs and TMPyP
The interaction between TMPyP and ODNs was characterized by fluorescence and UV-vis absorption spectroscopy. As shown in Figure 2A, upon excitation at 422 nm, TMPyP exhibited two broad peaks at around 665 nm and 715 nm (curve a). With the addition of dG 6 TG 6 , the fluorescence of TMPyP decreased due to the electron transfer from guanine to TMPyP (curve b) [20]. However, the incorporation of dA 13 (curve c), dC 13 (curve d) and dT 13 (curve e) increased the fluorescence of TMPyP, and the emission wavelengths were shifted to 660 nm and 720 nm. The enhanced fluorescence intensity and the shift of the emission peaks were ascribed to the formation of a π-π complex between TMPyP and ODNs [20]. The above results were consistent with those in the cases of mononucleotides. Mononucleotides interacted differently with TMPyP, in which adenine, thymine, and cytosine increased the fluorescence intensity of TMPyP, while guanine substantially quenched the fluorescence intensity of TMPyP [20]. The higher fluorescence of TMPyP by dT 13 than that by dA 13 or dC 13 indicated that TMPyP binded to dT 13 with stronger affinity [21].
The TMPyP-ODN complex was formed by mixing 1 μM TMPyP and 2 μM ODN, followed by dilution to 200 μL with Tris-HCl buffer and stirring for 50 min. For the Hg 2+ assay, Hg 2+ with various concentrations was added into the solution of TMPyP-dT30 complex under stirring for 1 min. For the fluorescence, measurement with excitation was 422 nm and emission was at 660 nm. The fluorescence lifetimes were determined from the data obtained from a compact FluoTime 100 fluorescence lifetime spectrometer. The data could be analyzed by exponential fits. The solution of Hg 2+ , TMPyP and dT30 were mixed by 1 μM Hg 2+ and 1 μM TMPyP for Hg-TMPyP; 1 μM TMPyP and 2 μM dT30 for TMPyP-dT30; 1 μM Hg 2+ , 1 μM TMPyP and 2 μM dT30 for Hg-TMPyP-dT30. The solutions were followed by dilution to 200 μL with Tris-HCl buffer and stirring for 120 min. All the lifetime measurements were with an excitation at 422 nm.

Interaction between ODNs and TMPyP
The interaction between TMPyP and ODNs was characterized by fluorescence and UV-vis absorption spectroscopy. As shown in Figure 2A, upon excitation at 422 nm, TMPyP exhibited two broad peaks at around 665 nm and 715 nm (curve a). With the addition of dG6TG6, the fluorescence of TMPyP decreased due to the electron transfer from guanine to TMPyP (curve b) [20]. However, the incorporation of dA13 (curve c), dC13 (curve d) and dT13 (curve e) increased the fluorescence of TMPyP, and the emission wavelengths were shifted to 660 nm and 720 nm. The enhanced fluorescence intensity and the shift of the emission peaks were ascribed to the formation of a π-π complex between TMPyP and ODNs [20]. The above results were consistent with those in the cases of mononucleotides. Mononucleotides interacted differently with TMPyP, in which adenine, thymine, and cytosine increased the fluorescence intensity of TMPyP, while guanine substantially quenched the fluorescence intensity of TMPyP [20]. The higher fluorescence of TMPyP by dT13 than that by dA13 or dC13 indicated that TMPyP binded to dT13 with stronger affinity [21].
TMPyP displayed a characteristic absorption peak at 422 nm known as the Soret band (curve a in Figure 2B), which originated from the S0-S2 transition [22]. The incorporation of dG6TG6 (curve b), dA13 (curve c), dC13 (curve d) and dT13 (curve e) decreased the absorption of TMPyP, and the absorbance peak of TMPyP was red-shifted to 426 nm and 431 nm in the case of dC13 (curve d) and dT13 (curve e), respectively. The larger bathochromic shift in curve e suggested the higher binding affinity between TMPyP and dT13.  TMPyP displayed a characteristic absorption peak at 422 nm known as the Soret band (curve a in Figure 2B), which originated from the S 0 -S 2 transition [22]. The incorporation of dG 6 TG 6 (curve b), dA 13 (curve c), dC 13 (curve d) and dT 13 (curve e) decreased the absorption of TMPyP, and the absorbance peak of TMPyP was red-shifted to 426 nm and 431 nm in the case of dC 13 (curve d) and dT 13 (curve e), respectively. The larger bathochromic shift in curve e suggested the higher binding affinity between TMPyP and dT 13 .
The length of polythymine also influenced the fluorescence enhancement of TMPyP ( Figure 3). The incorporation of different lengths of polythymine resulted in increased fluorescence and began to level off with the length of dT 30 . Thus, the length of dT 30 might have been more favorable for π-π stacking between TMPyP and polythymine. to level off with the length of dT30. Thus, the length of dT30 might have been more favorable for π-π stacking between TMPyP and polythymine.

Fluorescence Quenching of TMPyP-dT30 Complex by Hg 2+
In comparison with the fluorescence of TMPyP at 665 nm and 715 nm (curve a in Figure 4A), the fluorescence of the TMPyP-dT30 complex at 660 nm and 720 nm (curve b in Figure 4A) was significantly enhanced. With the addition of Hg 2+ , the fluorescence of the TMPyP-dT30 complex at 660 nm was greatly quenched and in the same time, a tiny and broad emission peak at 635 nm was attained (curve c). The emission peak at 635 nm was ascribed to that of Hg-TMPyP, as evidenced by the emission of Hg-TMPyP in curve d. Porphyrins have been reported to bind to Hg 2+ with the formation of metalloporphyrin [23], and the fluorescence of porphyrins could be quenched by Hg 2+ due to the heavy atom effect [24]. Based on the results above, a new ternary complex of Hg-TMPyP-dT30 might have formed [25], providing the possibility for a Hg 2+ assay.
To further demonstrate the formation of the ternary complex, UV-vis absorption spectra were investigated ( Figure 4B). With the addition of dT30, the Soret band of TMPyP at 422 nm (curve a) was red-shifted to 431 nm (curve b). The electrostatic interaction and π-π stacking facilitated the formation of the TMPyP-dT30 complex [20]. The formation of the ternary complex of Hg-TMPyP-dT30 led to a larger bathochromic shift of the Soret band to 462 nm, and in the same time, a new and small peak appeared at 437 nm (curve c). The new peak originated from the absorption of Hg-TMPyP (curve d) [26].
The conformational change of dT30 induced by TMPyP was studied by CD measurements ( Figure 4C). The dT30 exhibited a negative peak at 250 nm and a positive peak at 275 nm (curve a). Upon adding TMPyP to the solution of dT30, both peaks decreased slightly and in the same time, a new positive peak at 430 nm appeared (curve b), which indicated the binding of TMPyP to dT30 through π-π stacking [27,28]. Note that TMPyP did not show any detectable CD signal. With the incorporation of Hg 2+ to the TMPyP-dT30 complex, the 275 nm-peak disappeared and the negative peak at 250 nm shifted to 260 nm (curve c). Furthermore, the 430 nm-peak disappeared. The results above suggested that the binding of TMPyP to dT30 was interrupted by Hg 2+ . Such an interruption might have been ascribed to the alteration of the planarity of the TMPyP core induced by Hg 2+ [5]. When adding Hg 2+ to dT30 solution, the formation of the T-Hg-T complex was characterized by the negative peak at 255 nm and the positive peak at 280 nm (curve d). No positive peak at 280 nm in curve c was shown compared to the T-Hg-T complex, which indicated the different structure of dT30 in T-Hg-T and the Hg-TMPyP-dT30 complex.
The fluorescence lifetime measurement provided deep insight into the ternary complex formation. The lifetime of TMPyP determined from the decay curve was 4.6 ns, consistent with that reported previously [26]. The lifetime was increased to 10.2 ns upon formation of TMPyP-dT30. However, the ternary complex of Hg-TMPyP-dT30 possessed a much shorter lifetime of 2.0 ns. Upon

Fluorescence Quenching of TMPyP-dT 30 Complex by Hg 2+
In comparison with the fluorescence of TMPyP at 665 nm and 715 nm (curve a in Figure 4A), the fluorescence of the TMPyP-dT 30 complex at 660 nm and 720 nm (curve b in Figure 4A) was significantly enhanced. With the addition of Hg 2+ , the fluorescence of the TMPyP-dT 30 complex at 660 nm was greatly quenched and in the same time, a tiny and broad emission peak at 635 nm was attained (curve c). The emission peak at 635 nm was ascribed to that of Hg-TMPyP, as evidenced by the emission of Hg-TMPyP in curve d. Porphyrins have been reported to bind to Hg 2+ with the formation of metalloporphyrin [23], and the fluorescence of porphyrins could be quenched by Hg 2+ due to the heavy atom effect [24]. Based on the results above, a new ternary complex of Hg-TMPyP-dT 30 might have formed [25], providing the possibility for a Hg 2+ assay. adding Hg 2+ to the solution of TMPyP, a lifetime of 1.2 ns was observed, indicating the binding of Hg 2+ to TMPyP [26]. As reported previously, the binding of TMPyP to dT30 possessed high binding affinity (binding constant of 10 7 M −1 ), while the coordination of Hg 2+ to dT30 was a slow dynamic process at low concentrations of Hg 2+ [29]. The binding of Hg 2+ to TMPyP plays an important role in the Hg-TMPyP-dT30 ternary complex formation. Upon addition of Hg 2+ to the solution of TMPyP-dC30, the ternary complex of Hg-TMPyP-dC30 was also formed, as evidenced by the quenching of the fluorescence of TMPyP-dC30 by Hg 2+ (Figure 4D). Considering the relatively lower affinity between Hg 2+ and dC30 than that between Hg 2+ and TMPyP [26,30], the coordination of Hg 2+ to TMPyP was essential for the formation of the Hg-TMPyP-dC30 complex. The quenching of the fluorescence of TMPyP-dT30 by Hg 2+ provided the possibility for a Hg 2+ assay.

Calibration Curve of The Hg 2+ Assay
The fluorescence intensity of TMPyP-dT30 at 660 nm gradually decreased with an increased Hg 2+ concentration ( Figure 5A). The dependence of (F0 − F)/F0 on the concentration of Hg 2+ was shown in Figure 5B, where F0 and F represented the fluorescence intensities in the absence and presence of Hg 2+ , respectively. The inset of Figure 5B depicted the linear portion of the curve with  To further demonstrate the formation of the ternary complex , UV-vis absorption spectra were investigated ( Figure 4B). With the addition of dT 30 , the Soret band of TMPyP at 422 nm (curve a) was red-shifted to 431 nm (curve b). The electrostatic interaction and π-π stacking facilitated the formation of the TMPyP-dT 30 complex [20]. The formation of the ternary complex of Hg-TMPyP-dT 30 led to a larger bathochromic shift of the Soret band to 462 nm, and in the same time, a new and small peak appeared at 437 nm (curve c). The new peak originated from the absorption of Hg-TMPyP (curve d) [26].
The conformational change of dT 30 induced by TMPyP was studied by CD measurements ( Figure 4C). The dT 30 exhibited a negative peak at 250 nm and a positive peak at 275 nm (curve a). Upon adding TMPyP to the solution of dT 30 , both peaks decreased slightly and in the same time, a new positive peak at 430 nm appeared (curve b), which indicated the binding of TMPyP to dT 30 through π-π stacking [27,28]. Note that TMPyP did not show any detectable CD signal. With the incorporation of Hg 2+ to the TMPyP-dT 30 complex, the 275 nm-peak disappeared and the negative peak at 250 nm shifted to 260 nm (curve c). Furthermore, the 430 nm-peak disappeared. The results above suggested that the binding of TMPyP to dT 30 was interrupted by Hg 2+ . Such an interruption might have been ascribed to the alteration of the planarity of the TMPyP core induced by Hg 2+ [5]. When adding Hg 2+ to dT 30 solution, the formation of the T-Hg-T complex was characterized by the negative peak at 255 nm and the positive peak at 280 nm (curve d). No positive peak at 280 nm in curve c was shown compared to the T-Hg-T complex, which indicated the different structure of dT 30 in T-Hg-T and the Hg-TMPyP-dT 30 complex.
The fluorescence lifetime measurement provided deep insight into the ternary complex formation. The lifetime of TMPyP determined from the decay curve was 4.6 ns, consistent with that reported previously [26]. The lifetime was increased to 10.2 ns upon formation of TMPyP-dT 30 . However, the ternary complex of Hg-TMPyP-dT 30 possessed a much shorter lifetime of 2.0 ns. Upon adding Hg 2+ to the solution of TMPyP, a lifetime of 1.2 ns was observed, indicating the binding of Hg 2+ to TMPyP [26].
As reported previously, the binding of TMPyP to dT 30 possessed high binding affinity (binding constant of 10 7 M −1 ), while the coordination of Hg 2+ to dT 30 was a slow dynamic process at low concentrations of Hg 2+ [29]. The binding of Hg 2+ to TMPyP plays an important role in the Hg-TMPyP-dT 30 ternary complex formation. Upon addition of Hg 2+ to the solution of TMPyP-dC 30 , the ternary complex of Hg-TMPyP-dC 30 was also formed, as evidenced by the quenching of the fluorescence of TMPyP-dC 30 by Hg 2+ (Figure 4D). Considering the relatively lower affinity between Hg 2+ and dC 30 than that between Hg 2+ and TMPyP [26,30], the coordination of Hg 2+ to TMPyP was essential for the formation of the Hg-TMPyP-dC 30 complex. The quenching of the fluorescence of TMPyP-dT 30 by Hg 2+ provided the possibility for a Hg 2+ assay.

Calibration Curve of The Hg 2+ Assay
The fluorescence intensity of TMPyP-dT 30 at 660 nm gradually decreased with an increased Hg 2+ concentration ( Figure 5A). The dependence of (F 0 − F)/F 0 on the concentration of Hg 2+ was shown in Figure 5B, where F 0 and F represented the fluorescence intensities in the absence and presence of Hg 2+ , respectively. The inset of Figure 5B depicted the linear portion of the curve with Hg 2+ concentrations ranging from 5 nM to 100 nM and the linear regression equation was presented as (F 0 − F)/F 0 = 3.22 [Hg 2+ ] (µM) + 0.05 (R 2 = 0.9605). The detection limit was estimated to be 1.3 nM, being much lower than those by the 8-amino BODIPY-based fluorescence assay (49 nM) [31], Raman spectroscopic method (10 nM) [32], colorimetric assay employing plasmonic gold nanoparticles (50 nM [33] and 17.3 nM [34]), colorimetric assay based on Cu 2-x Se nanoparticles (2.7nM) [35] and the fluorescence polarization assay based on CdTe/CdS quantum dots (8.6 nM) [36]. Such a concentration level was also comparable with those by fluorescent assay based on phosphorothioate RNA modifications (1.7 nM) [37]. The comparison of the proposed sensor with other methods is shown in Table 1. Because the maximum permissible level of Hg 2+ in drinking water and food, set by the U.S. Environmental Protection Agency and World Health Organization, is at 2 ppb (10 nM) [38], the sensing protocol serves as a viable alternative for a practical Hg 2+ assay.

Selectivity of the Assay
The selectivity of the method was evaluated by the addition of various metal ions of 4.0 μM ( Figure 6). Hg 2+ quenched about 90% of the fluorescence intensity of the TMPyP-dT30 complex at 660 nm. However, other metal ions, such as Cu 2+ , Fe 3+ , Fe 2+ , Cd 2+ , Cr 3+ , Mg 2+ , Mn 2+ and Ni 2+ exerted negligible influence on the fluorescence intensity of TMPyP-dT30. Pb 2+ caused 30% quenching of the fluorescence intensity due to the formation of the unstable Pb 2+ -TMPyP-dT30 [39]. The proposed fluorescence method possessed good selectivity toward Hg 2+ determination.

Selectivity of the Assay
The selectivity of the method was evaluated by the addition of various metal ions of 4.0 µM ( Figure 6). Hg 2+ quenched about 90% of the fluorescence intensity of the TMPyP-dT 30 complex at 660 nm. However, other metal ions, such as Cu 2+ , Fe 3+ , Fe 2+ , Cd 2+ , Cr 3+ , Mg 2+ , Mn 2+ and Ni 2+ exerted negligible influence on the fluorescence intensity of TMPyP-dT 30 . Pb 2+ caused 30% quenching of the fluorescence intensity due to the formation of the unstable Pb 2+ -TMPyP-dT 30 [39]. The proposed fluorescence method possessed good selectivity toward Hg 2+ determination.

Practical Samples for the Hg 2+ Assay
The practical application of this method for Hg 2+ determination was evaluated by the standard addition method through adding different concentration of Hg 2+ in tap water and river water. As shown in Table 2, the recoveries were between 96-105%. The recovery suggested that the method was largely free from the matrix effect of the complex real water samples and could be served for Hg 2+ determination in tap water and river water.

Practical Samples for the Hg 2+ Assay
The practical application of this method for Hg 2+ determination was evaluated by the standard addition method through adding different concentration of Hg 2+ in tap water and river water. As shown in Table 2, the recoveries were between 96-105%. The recovery suggested that the method was largely free from the matrix effect of the complex real water samples and could be served for Hg 2+ determination in tap water and river water.

Conclusions
The fluorescence quenching capability of Hg 2+ on the TMPyP-dT30 has been proposed, enabling a sensitive and selective Hg 2+ assay. The formation of the Hg-TMPyP-dT30 ternary complex was characterized by various techniques. The mechanism inherent in the fluorescence quenching of TMPyP-dT30 by Hg 2+ involved a synergistic effect. The detection limit of the proposed method was obtained as 1.3 nM. Experiments for tap water and river water demonstrated that the detection method was applicable for Hg 2+ determination in real samples. The Hg 2+ sensor based on oligonucleotide dT30-enhanced TMPyP fluorescence was fast and low-cost, presenting a promising platform for practical Hg 2+ determination. In addition, this is the first time dT30 enhanced TMPyP fluorescence intensity was used for Hg 2+ determination. The sensing protocol may provide useful information on the interaction among porphyrins, DNA and heavy metal ions.

Conclusions
The fluorescence quenching capability of Hg 2+ on the TMPyP-dT 30 has been proposed, enabling a sensitive and selective Hg 2+ assay. The formation of the Hg-TMPyP-dT 30 ternary complex was characterized by various techniques. The mechanism inherent in the fluorescence quenching of TMPyP-dT 30 by Hg 2+ involved a synergistic effect. The detection limit of the proposed method was obtained as 1.3 nM. Experiments for tap water and river water demonstrated that the detection method was applicable for Hg 2+ determination in real samples. The Hg 2+ sensor based on oligonucleotide dT 30 -enhanced TMPyP fluorescence was fast and low-cost, presenting a promising platform for practical Hg 2+ determination. In addition, this is the first time dT 30 enhanced TMPyP fluorescence intensity was used for Hg 2+ determination. The sensing protocol may provide useful information on the interaction among porphyrins, DNA and heavy metal ions.