All chemicals used for these investigations were of analytical grade purity. L-Cysteine (Cys, minimum 98.5%) was purchased from Sigma Aldrich Chemical Company and used as received. MSD (5′-TTCTTTCTTCCCCTTGTTTGTT-3′) was synthesized and purified by HPLC (Takara Biotech. Co., Dalian). SG (10,000×) was purchased from Invitrogen Inc. A stock solution of 400× was prepared with DMSO/water (volume 1:1) before use. Milli-Q water (18.2 MΩ cm) was used in all procedures. The fluorescence measurements were recorded at room temperature (RT) on a Perkin Elmer LS-55 spectrophotometer equipped with a xenon lamp excitation source. Fluorescence spectra were measured at an excitation wavelength of 490 nm and the emission range from 500 to 650 nm with the excitation and emission slit widths set at 5 nm.

10 nM MSD was first incubated with 70 nM Hg²⁺ solution in 2.0 mL of 10 mM MOPS (3-(N-morpholino)propanesulfonic acid) buffer containing 0.1 M NaNO₃ (pH 7.50). 5 μL of 25 × SG was then added to the solution. After incubation for two minutes, different amounts of Cys were added to the solution. The mixture was then immediately used for the fluorescence measurements. For selectivity analysis, various kinds of amino acids with final concentration of 140 nM were used instead of Cys.

MSD, including seven T-T mismatches, could form a stable hairpin structure upon combination with Hg²⁺, which could be specifically stained by SG and produce a high fluorescence signal. However, in the presence of Cys, Hg²⁺ was extracted from the T-Hg²⁺-T structure because Cys can bind to Hg²⁺ with higher affinity (the formation constant for Cys to Hg²⁺ is ca. 10⁴²) [31]. As a result, the hairpin structure dehybridizes and the fluorescence intensity of SG decreases. Scheme 1 depicts the designed fluorescence method for Cys detection. The detection can be completed in less than 5 min.

As shown in Figure 1, the fluorescence intensity of MSD/Hg²⁺/SG is 116.7 a.u., while MSD/Hg²⁺/SG/Cys is about 15 a.u. when the concentration of Cys is 140 nM.

In previous study, Liu et al. [30] found that MSD kinetically accomplishes the hairpin structure at [Hg²⁺]/7[MSD] = 1. One equivalent of MSD thus requires seven equivalents of Hg²⁺ to completely form the hairpin structure due to the T-Hg²⁺-T chemistry. So we fixed the concentration of MSD to be 10 nM and Hg²⁺ 70 nM, and then the concentration of SG was optimized. Figure 2 shows the fluorescence spectra of MSD and MSD-Hg²⁺ upon addition of different concentration of SG.

We can see from the results that the fluorescence intensity of SG increases both in the absence of Hg²⁺ [Figure 2(a)] and in the presence of Hg²⁺ [Figure 2(b)]. The highest ratio of the fluorescence intensity of SG-Hg²⁺-MSD to that of SG-MSD is 21.8 [Figure 2(c)], at which the concentration of SG is 1.225 × 10⁻⁷ M, so we selected the concentration of SG to be 1.225 × 10⁻⁷ M when the concentration of MSD is 10 nM and Hg²⁺ 70 nM.

To evaluate the sensitivity of the assay, different concentrations of Cys were mixed with MSD/Hg²⁺/SG, and then the fluorescence intensity was immediately measured. As shown in Figure 3(a), the fluorescence intensity of SG decreases gradually upon adding increased concentrations of Cys.

When the concentration of Cys is two-fold higher than that of Hg²⁺, the fluorescence intensity decreases very slowly due to formation of a 2:1 Cys/Hg²⁺ adduct [31]. This implies that almost all the Hg²⁺ has been extracted from T-Hg²⁺-T complex when the concentration of Cys is two fold higher than that of Hg²⁺. When we increased the concentration of Cys further, we found that the fluorescence intensity decreased very slowly. Perhaps some Cys forms Hg(Cys)₃ complexes. By measuring the fluorescence intensity at the emission maximum of SG-Hg²⁺-MSD-Cys, a linear response of fluorescence intensity vs. [Cys] was observed in the range 7–84 nM [Figure 3(b) inset]. The detection limit may be estimated from Equation (1): (1) LOD = 3 × S 0 Swhere S₀ is the standard deviation of the blank and S is the sensitivity. 3.39 nM was the experimentally estimated detection limit for Cys, which was lower than most reported Cys sensors. Table 1 lists a comparison of methods for the determination of Cys and confirms the results.

What is more important, our method is really fast. In the first step of this assay, SG staining was finished in 2 min, which was proved to be fully adequate for the SG binding by a previous study [30]. In the second step, after the addition of Cys, the fluorescence spectra was measured immediately, and as anticipated, the fluorescence intensity remained stable for a long time (data not shown), because Cys can bind to Hg²⁺ very strongly [31].

We further find that we can obtain different dynamic ranges by changing the concentration of MSD/Hg²⁺. For example, when we fixed the concentration of MSD to be 50 nM and that of Hg²⁺ to be 350 nM (MSD kinetically forms the hairpin structure at [Hg²⁺]/7[MSD] = 1 [30]), a linear dynamic range for Cys from 50 to 250 nM was obtained [Figure 4(a) inset]. When we changed the concentration of MSD and Hg²⁺ to be 100 nM and 700 nM, a very wide linear dynamic range for Cys from 50 to 1,000 nM was obtained [Figure 4(b) inset]. The concentration of SG was 6.125 × 10⁻⁷ M [Figure 4(a)] and 1.225 × 10⁻⁶ M [Figure 4(b)], respectively, which was determined by the same method mentioned above (Section 3.2). When the concentration of MSD/Hg²⁺ changed, different concentration of Cys was needed to extract Hg²⁺ from T-Hg²⁺-T structure due to the 2:1 Cys/Hg²⁺ adduct, which leads to the different dynamic ranges for Cys.

To challenge the assay’s selectivity, other amino acids and the Cys-containing tripeptide glutathione (GSH) at a concentration of 140 nM were analyzed. The results are shown in Figure 5. It is clear that only Cys and GSH show significant fluorescence intensity changes. In contrast to the enormous fluorescence decrease observed for Cys and GSH, there is very little fluorescence change observed in the presence of other amino acids. A Cys-containing tripeptide GSH (γ-Glu-Cys-Gly) showed almost the same fluorescence decrease as Cys, indicating that Hg²⁺ can be extracted from the T-Hg²⁺-T structure upon adding GSH. This is because GSH can also form a very stable 2:1 GSH/Hg²⁺ complex with Hg²⁺ [34–36]. Hg²⁺ is well known to have an affinity to certain N-type ligands such as amino acids [37], but the amino acids, except for Cys, cannot extract Hg²⁺ from the thymine-Hg²⁺-thymine complex. The binding affinity of Hg²⁺ to T-T mismatch sites thus appears to be stronger enough than that of Hg²⁺ to all of the amino acids studied, except Cys. We also found from Figure 5 that a sulfur-containing amino acid like methionine did not result in a significant fluorescence decrease. The formation constant for Cys to Hg²⁺ (about 10⁴²) is much higher than that of methionine to Hg²⁺ (about 10^17.6) [31]. The method presented here therefore shows very high specificity to Cys.