Cyclodextrin-Based Metal-Organic Nanotube as Fluorescent Probe for Selective Turn-On Detection of Hydrogen Sulfide in Living Cells Based on H2S-Involved Coordination Mechanism

Hydrogen sulfide (H2S) has been considered as the third biologically gaseous messenger (gasotransmitter) after nitric oxide (NO) and carbon monoxide (CO). Fluorescent detection of H2S in living cells is very important to human health because it has been found that the abnormal levels of H2S in human body can cause Alzheimer’s disease, cancers and diabetes. Herein, we develop a cyclodextrin-based metal-organic nanotube, CD-MONT-2, possessing a {Pb14} metallamacrocycle for efficient detection of H2S. CD-MONT-2′ (the guest-free form of CD-MONT-2) exhibits turn-on detection of H2S with high selectivity and moderate sensitivity when the material was dissolved in DMSO solution. Significantly, CD-MONT-2′ can act as a fluorescent turn-on probe for highly selective detection of H2S in living cells. The sensing mechanism in the present work is based on the coordination of H2S as the auxochromic group to the central Pb(II) ion to enhance the fluorescence intensity, which is studied for the first time.

Hydrogen sulfide (H 2 S), a colourless toxic gas with rotten egg smell, possesses double-sided nature 1 . On the one hand, H 2 S is known as a dangerous industrial pollutant for many years 2 . Because of the properties of forming explosive mixtures in the air, and causing an explosion under fire or heat, the H 2 S gas has received a growing universal attention in the aspect of safety 3,4 . On the other hand, along with nitric oxide (NO) and carbon monoxide (CO), the H 2 S gas has been recognized as a third gaseous transmitter gas in the human body recently 5 . In vivo, H 2 S is generated by endogenous enzymes [6][7][8] (such as, cystathionine β-synthase (CBS), cystathionine γ-lyase (CSE), or 3-mercaptopyruvate sulfurtransferase (MPST)) in many organs (e.g., heart, brain, kidneys, nervous system, etc.) and tissues (e.g., adipose tissues, etc.) [9][10][11][12] . The abnormal levels of generated H 2 S were related to Alzheimer's disease, cancers, diabetic complications and Down's syndrome [13][14][15] . Hence, more and more attention has been drawn to the sensitive and selective detection of H 2 S which is chosen as a target in biological systems 16 .
In the past decade, a variety of fluorescent probes were developed for rapid detection of H 2 S. Generally, the design strategies are highly dependent on the chemical properties of the physiologically active species 17,18 . On the basis of current research, the sensing mechanism for the fluorescent probes of H 2 S detection can be classified into three types [19][20][21][22] : (i) H 2 S reductive reactions; (ii) H 2 S nucleophilic reactions and (iii) metal sulfide precipitation reactions. Most of reported results are focused on design and synthesis of organic molecules with desired functional groups to detect H 2 S based on [23][24][25] (i) and (ii) reactions, seldom are metal-organic frameworks (MOFs) Scientific RepoRts | 6:21951 | DOI: 10.1038/srep21951 or metal-organic nanotubes (MONTs). In general, MOFs or MONTs with both high selectivity and fluorescence turn-on in response to H 2 S are very rare 17,26,27 .
In the previous work, we described a cyclodextrin-based Pb(II) metal-organic nanotube (CD-MONT-2) exhibiting temperature-dependent fluorescence and adsorption of I 2 molecules 28 . The excellent fluorescent property of CD-MONT-2 and the high affinity of Pb(II) to S atom prompted us to study its potential in fluorescent detection of H 2 S. Herein, we report CD-MONT-2′ (the guest-free sample of CD-MONT-2) as a fluorescence turn-on probe for H 2 S detection. Significantly, CD-MONT-2′ can detect H 2 S in living cells with high selectivity and moderate sensitivity. Furthermore, in the present work, a new sensing mechanism that H 2 S molecules act as auxocchromic groups to interact with the central chromophore to enhance the fluorescence emission is discovered for the first time, which is quite different from previous results.

Results and Discussion
Structure of CD-MONT-2. Colorless crystals of CD-MONT-2 were obtained under the guidance of reference 28 . The cyclodextrin-based Pb(II) metal-organic nanotube (CD-MONT-2) consists of coplanar {Pb 14 } metallamacrocycle surrounded by two β-cyclodextrin molecules, as shown in Fig. 1. In the CD-MONT-2, the dimensions of the chiral cavity are ca. 13.0 × 10.3 × 10.2 Å filled with cyclohexanol molecules. The uncoordinated solvates in the cavity can be fully removed by heating CD-MONT-2 at 120 °C for half an hour to generate guestfree form, CD-MONT-2′. The phase purity of bulk sample was further confirmed by comparison of the powder X-ray diffraction (PXRD) patterns of as-synthesized and activated sample (Supplementary Figure S1), which matched well with the simulated PXRD pattern from the single-crystal data. The following fluorescent measurements were based on CD-MONT-2′.
Fluorescent measurements of CD-MONT-2′. CD-MONT-2′ is slightly soluble in dimethylsulphoxide (DMSO), in which CD-MONT-2′ emits fluorescence at 409 nm (Φ = 0.02) upon the excitation at 330 nm (Supplementary Figure S2). To probe the fluorescent response of CD-MONT-2′ towards H 2 S, CD-MONT-2′ was dissolved in DMSO to make a 10 μ M stock solution, then the emission spectrum was recorded from 350 to 650 nm upon the excitation at 330 nm. CD-MONT-2′ in DMSO solution exhibits relatively weak fluorescence and keeps in the turn-off state due to the very dilute concentration. However, with the addition of H 2 S (1 mL) into the above solution, the fluorescence intensity shows a significant increase with time. Compared to the original one, almost 15 fold fluorescence enhancement is observed after 15 minutes, and no further increase occurs (Fig. 2a). The time-dependent fluorescence measurements for the addition of H 2 S into the stock solution reveal that CD-MONT-2′ in DMSO exhibits rapid response toward H 2 S, which is different from the reported Pb-based complexes 29 . To further probe the fluorescence turn-on response to sulfide, various concentrations of Na 2 S (0-10 μ M) were added to the stock, and the fluorescence spectra were recorded in Fig. 2b. Similarly, the fluorescence intensity clearly increases with the increasement of the concentration of Na 2 S, and almost becomes 4 times of original fluorescence intensity when the concentration of Na 2 S reaches 10 μ M.
In order to confirm the fluorescent selectivity to Na 2 S over other substances, various additional experiments were carried out by gradual addition of other sodium salts (such as Na 2 SO 4 , Na 2 S 2 O 3 , Na 2 SO 3 , NaNO 3 , NaHCO 3 , NaCl, NaClO, NaOAc, FeCl 2 , FeCl 3 and KHPO 4 ), reducing agents (glucose), thiol amino acids (GSH and L-cys), non-thiol amino acids (Gln, L-Thr, L-Trp, L-Tyr, Leu, L-Leu, L-asp and Gly), reactive nitrogen species (NO 2 − ), reactive oxygen species (H 2 O 2 and t BuOOH), and reactive sulfur species (TGA, THU and thiophene) 30,31 . And the spectra are shown in supplementary Figure S3-S5. The blank of only β-cyclodextrin in DMSO at 10 μ M with Na 2 S was also recorded, and the spectrum is shown in supplementary Figure S3o. The fluorescence intensities of (I-I 0 )/I 0 (where I 0 is the initial fluorescence intensity, and I is the fluorescence intensity after the addition of the analyte) spectra (λ = 409 nm) are displayed in Fig. 2c,d. The results reveal that the additions of those substances have little effect on the fluorescence intensity of CD-MONT-2′, indicating the high selectivity to Na 2 S over other substances through fluorescence enhancement. All these results demonstrate that CD-MONT-2′ exhibits fluorescence turn-on response to H 2 S molecule with high selectivity and moderate sensitivity 11 . Sensing mechanism. In the past decade, several MOF-based fluorescence turn-on probes on the detection of H 2 S were reported 19 . The reported sensing mechanism is mostly based on the H 2 S-involved organic reactions, through which the turn-off state of those material can be converted to turn-on state. Hence, there always exists a desired functional group in the MOF materials that can react with H 2 S to complete the conversion. However, in the metal-organic nanotube of CD-MONT-2′, there is no additional organic functional groups that can react with H 2 S to enhance the fluorescence emission. Therefore, the sensing mechanism in the present work should be different with the previous results. It is known that cyclodextrins are nonaromatic and fluorescence silent, as a result, the fluorescence emission of CD-MONT-2′ should be assigned to a metal-centered transition involving the s and p orbitals of Pb(II) ions 28,32 . Thus, the fluorescence enhancement should derive from the interactions between H 2 S molecules and Pb(II) ions due to the high affinity of S atom to Pb(II) ion (K sp of PbS: 1 × 10 −28 ) 33 . As an auxochrome, the coordination of H 2 S to Pb(II) ion significantly increases the fluorescence emission of CD-MONT-2′ (Fig. 3).
To further confirm the above sensing mechanism, the UV-Vis absorbance spectra, FTIR spectra and 1 H NMR spectra were recorded for CD-MONT-2′ in DMSO before and after addition of H 2 S. The absorption band of CD-MONT-2′ in DMSO appears at around 266 nm (ε = 5.68 × 10 4 M −1 cm −1 ), which could be assigned to the transition from 6s 2 to 6sp involving the lone pairs on the Pb(II) 34 . When H 2 S was added into the DMSO solution containing CD-MONT-2′, the absorption band enhances and shows a red-shift (Fig. 4a), which may be derived from the attachment of "S" to Pb(II) 34 . As a common auxochrome and donor, the connection of "S" could always shift the absorption to a longer wavelength and increase the absorption intensity [35][36][37] . In contrast, the addition of other substances only enhances the absorption intensity slightly and shows almost no shift. These results indicate that the sensing mechanism for CD-MONT-2′ is based on the coordination of S atom to Pb(II) to increase the electron transfer to enhance the fluorescence intensity [38][39][40][41][42] .
Moreover, the FTIR spectra of CD-MONT-2′ in DMSO solution before and after treated with H 2 S or Na 2 S were carried out (Fig. 4b). The absorption peaks remain unchanged, except that there is a slight difference around 1250 cm −1 . The broad peak around 3450 cm −1 can be assigned to the stretching vibration of adsorbed water and hydroxyl groups in β-CD molecule. The peak around 1655 cm −1 is attributed to the O-H bending vibration of  (Fig. 5a). Similarly, the fluorescence intensity obviously increases with the increasement of Na 2 S, and becomes almost 4 times of original fluorescence intensity when the concentration of Na 2 S reaches 10 μ M (Fig. 5b, detection limit 0.058 μ M, the figure is shown in Supplementary Figure S9). Moreover, in order to confirm the fluorescent selectivity, various additional experiments were carried out by gradual addition of other inorganic salts (such as NaNO 3 , NaHCO 3 , NaClO, FeCl 2 , FeCl 3 and KHPO 4 ), reducing agents (glucose), thiol amino acids (GSH and L-cys, which are known to reduce to generate off-target H 2 S detection under the action of enzymes 51    were also recorded, and the spectra are shown in Supplementary Figure S10 o and Figure S12 respectively. The results demonstrate that CD-MONT-2′ exhibits fluorescence turn-on response to H 2 S molecule in the cell growth environment with high selectivity and moderate sensitivity, possessing the potential in real-time intracellular H 2 S imaging. Hence, the CD-MONT-2′ may be utilized to living cell imaging to sulphide. To test the viability and proliferation of the living cell, the MTT assay on HeLa cells was performed 50 (Fig. 5d). The cell viability is not lower than 80% until the concentration of CD-MONT-2′ reaches 20 μ M, indicating the low toxicity at the concentration of 0.1 μ M. The HeLa cells were incubated with 10 μ M probe for 15 minutes at 37 °C in a 5% CO 2 atmosphere, and washed with PBS for three times to remove the residual probe. Then fresh PBS containing various concentrations of Na 2 S were respectively added into the treated HeLa cells and incubated for 15 minutes. The fluorescent image of control one shows that CD-MONT-2′ probe could enter inside the cell and result in the weak blue fluorescent signal. However, with the increase of sulphide (Na 2 S) concentration from 1 to 100 μ M, the signal intensity increases obviously (Fig. 6). The strong blue fluorescent signal is observed when the sulphide concentration reaches 100 μ M. These results confirm that CD-MONT-2′ is active as a probe for sulphide and can be applied in living cell imaging.
In addition to supplementing cells with extraneous sources of sulphide, our experiments further focus on biothiols, such as the amino acid glutathione (GSH) and L-cysteine (L-cys), which can act as potential sulphide sources [51][52][53][54][55] . After 15 minutes of incubation, addition of both thiol species (200 μ M GSH or L-cys in PBS) elicits a brighter fluorescent response (see Fig. 7). The significant responses indicate that the CD-MONT-2′ probe could detect not only external sulphides supplemented to the cell cultures, but also sulphides produced by the cells in vivo.

Discussion
The design and synthesis of fluorescent turn-on probes for rapid detection of H 2 S in living cells is an active field in material chemistry and cell biology 9,17 . The development of coordination chemistry in the past decades opened a new avenue in searching fluorescent materials for selective detection of H 2 S. Actually, most of fluorescent coordination complexes including metal-organic frameworks show turn-off response towards H 2 S 29 , functional coordination complex-based probes with fluorescent turn-on response towards H 2 S are quite rare. Up to date, several MOF-based fluorescent turn-on probes have been synthesized and applied in the detection of H 2 S based on reduction/precipitation mechanism 19 . In the present work, the sensing mechanism is based on the coordination of H 2 S (as auxochromic group) to Pb(II) ion to enhance the fluorescent emission. To the best of our knowledge, this is the first fluorescence turn-on probe that can selectively detect H 2 S in living cells based on H 2 S-involved coordination mechanism.
On the other hand, one of the significant bottlenecks in detection of H 2 S in living cells is the toxicity of the fluorescent probe. Most of organic ligands used in the assembly of coordination complexes or metal-organic frameworks are limited to non-renewable petrochemical feedstocks and somewhat toxic. Recently, Stoddart and co-workers reported a series of MOFs composed of an edible natural product, γ-cyclodextrin [56][57][58][59] . In our work, CD-MONT-2′ was assembled by use of β-cyclodextrin, which is non-toxic and increases its practical application in the fluorescent detection of H 2 S in living cells.

Conclusions
In conclusion, a fluorescent metal-organic nanotube based on β-cyclodextrin for the detection of H 2 S has been developed and described. The newly developed fluorescent probe can detect H 2 S through fluorescence turn-on  fashion with high selectivity and moderate sensitivity. Furthermore, the sensing mechanism is based on the coordination of H 2 S to the central metal ions of the probe to tune the fluorescence intensity, which is quite different from the results reported previously. Significantly, the use of nontoxic β-cyclodextrin ligand in the probe makes it more advantage in the practical application. Our study may provide a new way in design and synthesis of new functional material on fluorescence turn-on detection of H 2 S in living cells.

Methods
Materials and Physical Measurements. Materials. All chemicals and solvents were purchased and used as received without further purification. Water used in living cell experiments were processed with a Millipore Milli-Q system (18.2 M Ω ·cm). Thioglycollic cid (TGA) and thiourea (THU) were purchased. t BuOOH could also be used to induce ROS in biological systems 31 . The ONOO − source was generated by the reaction of H 2 O 2 , H 2 SO 4 , NaNO 2 and MnO 2 . The concentration is obtained by UV-Vis at 302 nm 60 .
Physical Measurements. Fluorescence spectra were recorded with a Hitachi F-7000 fluorescence spectrophotometer. The powder X-ray diffraction data were obtained on a Philips X' Pert with Cu-Kα radiation (λ = 0.15418 nm). FTIR spectra were collected on a Bruker VERTEX-70 spectrometer in the 4000 − 600 cm −1 region. The optical absorption spectra were measured on a UV-vis spectrometer (Specord 205, Analytik Jena) in the range of 200 to 600 nm. 1 H NMR spectra were recorded on a Bruker AVANCE-400 NMR Spectrometer in d 6 -DMSO.
Synthesis of CD-MONT-2. β-CD (0.10 mmol, 115 mg) and PbCl 2 (0.80 mmol, 225 mg) were suspended in distilled water (30 mL) and stirred at 80 °C for an hour. After cooled to room temperature, the precipitate was separated from the mixture. The obtained solution was transformed to five 6 mL of glass tubes, then 3 mL cyclohexanol and trimethylamine were layered onto the solution in each tube. The glass tubes were sealed and heated at 110 °C for 3 days. A lot of colourless rod-like crystals were collected by filtration, washed with distilled water and dried in air (yield: 78%).

Fluorescent experiments.
All fluorescent measurements were carried out at room temperature on a Hitachi F7000 fluorescence spectrophotometer. Samples were excited at 330 nm with the excitation and emission slit widths set at 20 and 10 nm, respectively. The emission spectrum was scanned from 350 to 650 nm with 1200 nm min −1 . The photomultiplier voltage was set at 400 V. Accordingly, the probe was dissolved in dimethylsulphoxide (DMSO) to make a 10 μ M stock solution and the added substances were dissolved in DMSO as well. The stock was diluented by PBS for 100 times to obtain the concentration of 0.1 μ M and the added substances were dissolved in PBS. The H 2 S was made by the reaction of FeS and H 2 SO 4 and collected in the 250 mL flask for more than 30 min. To test the time-dependent properties, 1 mL H 2 S gas was taken out from the flask and bubbled into 1 mL corresponding solution.
MTT Cytotoxicity assay. HeLa cells were grown up in DMEM media with 10% FBS and penicillin/streptomycin. Cells were allowed to grow to 80% confluency before being collected using trypsin. Cells were transferred into a 96-well plate (Corning), and then incubated overnight at 37 °C in a 5% CO 2 atmosphere. A serial dilution on CD-MONT-2′ was performed in DMEM media, with 10 μ L added to each well to give final concentrations of 5, 7.5, 10, 15 and 20 μ M probe. Cells were allowed to incubate for 24 h. Wells containing only cells and only DMSO were also set up to serve as positive and negative controls. To test cell viability and proliferation, the MTT assay was performed. Briefly, after incubation for the indicated times, 10 μ L of MTT solution (5 mg/mL) was added to each well, and the cells were incubated for further 4 h at 37 °C. The precipitated formazan was dissolved in 150 μ L of dimethyl sulfoxide. The absorbance at 490 nm (A490) was measured using a microplate autoreader (Molecular Devices, M2e). Note that the wells without cells acted as the blank during the A490 measurement.
Cellular imaging experiments. HeLa cells were grown as previously described. The cells were seeded onto 12 mm sterile coverslips in a 24-well plate (Coring) and allowed to grow to 80% confluency at 37 °C in a 5% CO 2 atmosphere. At this time, a final concentration of 0.1 μ M CD-MONT-2′ was added to the cells and incubated for 15 minutes at the previous conditions. Media was then removed, and PBS was added to remove the probe left in solution and optimize the background signal. The sulphur source was then added (Na 2 S, GSH, or L-cys) to the desired concentration and cells were incubated for 10-15 minutes at room temperature before imaging. And then, the cells were fixed for 20 min in 200 μ L 4% paraformaldehyde (the fluorescent images of HeLa cells without being fixed are showed in Figure S13). After fix-ation, the cells were washed thrice with PBS. The coverslip with fixed cells was topped by a glass slide with a drop of 10 μ L of glycerol/PBS (v/v = 1:1) and placed above the objective on a fluorescence microscope.
All imaging experiments were performed on a Leica DMI3000B Inverted fluorescence microscopic. Excitation and emission were monitored using blue fiter provided with the scope. Imaging was performed with the × 20 dry objectives which are provided with the scope. Images were captured using Leica Application Suite software.